From 9105cae261bb29d4235b47e86c99c11909a4a3fb Mon Sep 17 00:00:00 2001 From: Martin Abelson Sahlen Date: Thu, 16 Jan 2025 10:27:26 +0200 Subject: [PATCH 001/286] Allow empty options for BigQuery (#1657) Co-authored-by: Martin Abelson Sahlen --- src/parser/mod.rs | 2 +- tests/sqlparser_bigquery.rs | 6 ++++++ 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ac764a535..861a392d4 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7341,7 +7341,7 @@ impl<'a> Parser<'a> { pub fn parse_options(&mut self, keyword: Keyword) -> Result, ParserError> { if self.parse_keyword(keyword) { self.expect_token(&Token::LParen)?; - let options = self.parse_comma_separated(Parser::parse_sql_option)?; + let options = self.parse_comma_separated0(Parser::parse_sql_option, Token::RParen)?; self.expect_token(&Token::RParen)?; Ok(options) } else { diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 9dfabc014..a173a6cc9 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -473,6 +473,12 @@ fn parse_create_table_with_options() { r#"description = "table option description")"# ); bigquery().verified_stmt(sql); + + let sql = "CREATE TABLE foo (x INT64) OPTIONS()"; + bigquery().verified_stmt(sql); + + let sql = "CREATE TABLE db.schema.test (x INT64 OPTIONS(description = 'An optional INTEGER field')) OPTIONS()"; + bigquery().verified_stmt(sql); } #[test] From 474150006fc8dd299bfa212321206263de0f4290 Mon Sep 17 00:00:00 2001 From: AvivDavid-Satori <107786696+AvivDavid-Satori@users.noreply.github.com> Date: Thu, 16 Jan 2025 11:09:53 +0200 Subject: [PATCH 002/286] Add support for parsing RAISERROR (#1656) --- src/ast/mod.rs | 50 ++++++++++++++++++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 3 +++ src/parser/mod.rs | 41 ++++++++++++++++++++++++++++++++ tests/sqlparser_mssql.rs | 33 ++++++++++++++++++++++++++ 5 files changed, 128 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 2d79f7d6b..e6499e14a 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3441,6 +3441,38 @@ pub enum Statement { /// /// See SetSessionParam(SetSessionParamKind), + /// RaiseError (MSSQL) + /// RAISERROR ( { msg_id | msg_str | @local_variable } + /// { , severity , state } + /// [ , argument [ , ...n ] ] ) + /// [ WITH option [ , ...n ] ] + /// See + RaisError { + message: Box, + severity: Box, + state: Box, + arguments: Vec, + options: Vec, + }, +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum RaisErrorOption { + Log, + NoWait, + SetError, +} + +impl fmt::Display for RaisErrorOption { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + RaisErrorOption::Log => write!(f, "LOG"), + RaisErrorOption::NoWait => write!(f, "NOWAIT"), + RaisErrorOption::SetError => write!(f, "SETERROR"), + } + } } impl fmt::Display for Statement { @@ -5026,6 +5058,24 @@ impl fmt::Display for Statement { Statement::RenameTable(rename_tables) => { write!(f, "RENAME TABLE {}", display_comma_separated(rename_tables)) } + Statement::RaisError { + message, + severity, + state, + arguments, + options, + } => { + write!(f, "RAISERROR({message}, {severity}, {state}")?; + if !arguments.is_empty() { + write!(f, ", {}", display_comma_separated(arguments))?; + } + write!(f, ")")?; + if !options.is_empty() { + write!(f, " WITH {}", display_comma_separated(options))?; + } + Ok(()) + } + Statement::List(command) => write!(f, "LIST {command}"), Statement::Remove(command) => write!(f, "REMOVE {command}"), Statement::SetSessionParam(kind) => write!(f, "SET {kind}"), diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 183bebf8c..6f89cd0db 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -495,6 +495,7 @@ impl Spanned for Statement { Statement::LoadData { .. } => Span::empty(), Statement::UNLISTEN { .. } => Span::empty(), Statement::RenameTable { .. } => Span::empty(), + Statement::RaisError { .. } => Span::empty(), Statement::List(..) | Statement::Remove(..) => Span::empty(), Statement::SetSessionParam { .. } => Span::empty(), } diff --git a/src/keywords.rs b/src/keywords.rs index eb9e3ea6f..6b09a877a 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -468,6 +468,7 @@ define_keywords!( LOCATION, LOCK, LOCKED, + LOG, LOGIN, LOGS, LONGBLOB, @@ -636,6 +637,7 @@ define_keywords!( QUARTER, QUERY, QUOTE, + RAISERROR, RANGE, RANK, RAW, @@ -728,6 +730,7 @@ define_keywords!( SESSION, SESSION_USER, SET, + SETERROR, SETS, SETTINGS, SHARE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 861a392d4..f443b6e42 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -579,6 +579,7 @@ impl<'a> Parser<'a> { Keyword::SAVEPOINT => self.parse_savepoint(), Keyword::RELEASE => self.parse_release(), Keyword::COMMIT => self.parse_commit(), + Keyword::RAISERROR => Ok(self.parse_raiserror()?), Keyword::ROLLBACK => self.parse_rollback(), Keyword::ASSERT => self.parse_assert(), // `PREPARE`, `EXECUTE` and `DEALLOCATE` are Postgres-specific @@ -13150,6 +13151,46 @@ impl<'a> Parser<'a> { } } + /// Parse a 'RAISERROR' statement + pub fn parse_raiserror(&mut self) -> Result { + self.expect_token(&Token::LParen)?; + let message = Box::new(self.parse_expr()?); + self.expect_token(&Token::Comma)?; + let severity = Box::new(self.parse_expr()?); + self.expect_token(&Token::Comma)?; + let state = Box::new(self.parse_expr()?); + let arguments = if self.consume_token(&Token::Comma) { + self.parse_comma_separated(Parser::parse_expr)? + } else { + vec![] + }; + self.expect_token(&Token::RParen)?; + let options = if self.parse_keyword(Keyword::WITH) { + self.parse_comma_separated(Parser::parse_raiserror_option)? + } else { + vec![] + }; + Ok(Statement::RaisError { + message, + severity, + state, + arguments, + options, + }) + } + + pub fn parse_raiserror_option(&mut self) -> Result { + match self.expect_one_of_keywords(&[Keyword::LOG, Keyword::NOWAIT, Keyword::SETERROR])? { + Keyword::LOG => Ok(RaisErrorOption::Log), + Keyword::NOWAIT => Ok(RaisErrorOption::NoWait), + Keyword::SETERROR => Ok(RaisErrorOption::SetError), + _ => self.expected( + "LOG, NOWAIT OR SETERROR raiserror option", + self.peek_token(), + ), + } + } + pub fn parse_deallocate(&mut self) -> Result { let prepare = self.parse_keyword(Keyword::PREPARE); let name = self.parse_identifier()?; diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 567cd5382..a0ac8a4d8 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1250,6 +1250,39 @@ fn parse_mssql_declare() { ); } +#[test] +fn test_parse_raiserror() { + let sql = r#"RAISERROR('This is a test', 16, 1)"#; + let s = ms().verified_stmt(sql); + assert_eq!( + s, + Statement::RaisError { + message: Box::new(Expr::Value(Value::SingleQuotedString( + "This is a test".to_string() + ))), + severity: Box::new(Expr::Value(Value::Number("16".parse().unwrap(), false))), + state: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + arguments: vec![], + options: vec![], + } + ); + + let sql = r#"RAISERROR('This is a test', 16, 1) WITH NOWAIT"#; + let _ = ms().verified_stmt(sql); + + let sql = r#"RAISERROR('This is a test', 16, 1, 'ARG') WITH SETERROR, LOG"#; + let _ = ms().verified_stmt(sql); + + let sql = r#"RAISERROR(N'This is message %s %d.', 10, 1, N'number', 5)"#; + let _ = ms().verified_stmt(sql); + + let sql = r#"RAISERROR(N'<<%*.*s>>', 10, 1, 7, 3, N'abcde')"#; + let _ = ms().verified_stmt(sql); + + let sql = r#"RAISERROR(@ErrorMessage, @ErrorSeverity, @ErrorState)"#; + let _ = ms().verified_stmt(sql); +} + #[test] fn parse_use() { let valid_object_names = [ From b4b5576dd4bcf5b386665f9010b6c3e94465eeaf Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 16 Jan 2025 18:50:30 +0200 Subject: [PATCH 003/286] Add support for Snowflake column aliases that use SQL keywords (#1632) --- src/dialect/mod.rs | 14 ++++ src/dialect/snowflake.rs | 45 ++++++++++++ src/parser/mod.rs | 131 ++++++++++++++++++++--------------- tests/sqlparser_snowflake.rs | 29 ++++++++ 4 files changed, 163 insertions(+), 56 deletions(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 64dbc4b1b..c69253b76 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -820,6 +820,20 @@ pub trait Dialect: Debug + Any { fn supports_set_stmt_without_operator(&self) -> bool { false } + + /// Returns true if the specified keyword should be parsed as a select item alias. + /// When explicit is true, the keyword is preceded by an `AS` word. Parser is provided + /// to enable looking ahead if needed. + fn is_select_item_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { + explicit || !keywords::RESERVED_FOR_COLUMN_ALIAS.contains(kw) + } + + /// Returns true if the specified keyword should be parsed as a table factor alias. + /// When explicit is true, the keyword is preceded by an `AS` word. Parser is provided + /// to enable looking ahead if needed. + fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { + explicit || !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 6b8380d63..f6e9c9eba 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -251,6 +251,51 @@ impl Dialect for SnowflakeDialect { fn supports_partiql(&self) -> bool { true } + + fn is_select_item_alias(&self, explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { + explicit + || match kw { + // The following keywords can be considered an alias as long as + // they are not followed by other tokens that may change their meaning + // e.g. `SELECT * EXCEPT (col1) FROM tbl` + Keyword::EXCEPT + // e.g. `SELECT 1 LIMIT 5` + | Keyword::LIMIT + // e.g. `SELECT 1 OFFSET 5 ROWS` + | Keyword::OFFSET + // e.g. `INSERT INTO t SELECT 1 RETURNING *` + | Keyword::RETURNING if !matches!(parser.peek_token_ref().token, Token::Comma | Token::EOF) => + { + false + } + + // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` + // which would give it a different meanins, for example: `SELECT 1 FETCH FIRST 10 ROWS` - not an alias + Keyword::FETCH + if parser.peek_keyword(Keyword::FIRST) || parser.peek_keyword(Keyword::NEXT) => + { + false + } + + // Reserved keywords by the Snowflake dialect, which seem to be less strictive + // than what is listed in `keywords::RESERVED_FOR_COLUMN_ALIAS`. The following + // keywords were tested with the this statement: `SELECT 1 `. + Keyword::FROM + | Keyword::GROUP + | Keyword::HAVING + | Keyword::INTERSECT + | Keyword::INTO + | Keyword::MINUS + | Keyword::ORDER + | Keyword::SELECT + | Keyword::UNION + | Keyword::WHERE + | Keyword::WITH => false, + + // Any other word is considered an alias + _ => true, + } + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f443b6e42..f34a5d742 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8838,38 +8838,76 @@ impl<'a> Parser<'a> { Ok(IdentWithAlias { ident, alias }) } - /// Parse `AS identifier` (or simply `identifier` if it's not a reserved keyword) - /// Some examples with aliases: `SELECT 1 foo`, `SELECT COUNT(*) AS cnt`, - /// `SELECT ... FROM t1 foo, t2 bar`, `SELECT ... FROM (...) AS bar` + /// Optionally parses an alias for a select list item + fn maybe_parse_select_item_alias(&mut self) -> Result, ParserError> { + fn validator(explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { + parser.dialect.is_select_item_alias(explicit, kw, parser) + } + self.parse_optional_alias_inner(None, validator) + } + + /// Optionally parses an alias for a table like in `... FROM generate_series(1, 10) AS t (col)`. + /// In this case, the alias is allowed to optionally name the columns in the table, in + /// addition to the table itself. + pub fn maybe_parse_table_alias(&mut self) -> Result, ParserError> { + fn validator(explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { + parser.dialect.is_table_factor_alias(explicit, kw, parser) + } + match self.parse_optional_alias_inner(None, validator)? { + Some(name) => { + let columns = self.parse_table_alias_column_defs()?; + Ok(Some(TableAlias { name, columns })) + } + None => Ok(None), + } + } + + /// Wrapper for parse_optional_alias_inner, left for backwards-compatibility + /// but new flows should use the context-specific methods such as `maybe_parse_select_item_alias` + /// and `maybe_parse_table_alias`. pub fn parse_optional_alias( &mut self, reserved_kwds: &[Keyword], ) -> Result, ParserError> { + fn validator(_explicit: bool, _kw: &Keyword, _parser: &mut Parser) -> bool { + false + } + self.parse_optional_alias_inner(Some(reserved_kwds), validator) + } + + /// Parses an optional alias after a SQL element such as a select list item + /// or a table name. + /// + /// This method accepts an optional list of reserved keywords or a function + /// to call to validate if a keyword should be parsed as an alias, to allow + /// callers to customize the parsing logic based on their context. + fn parse_optional_alias_inner( + &mut self, + reserved_kwds: Option<&[Keyword]>, + validator: F, + ) -> Result, ParserError> + where + F: Fn(bool, &Keyword, &mut Parser) -> bool, + { let after_as = self.parse_keyword(Keyword::AS); + let next_token = self.next_token(); match next_token.token { - // Accept any identifier after `AS` (though many dialects have restrictions on - // keywords that may appear here). If there's no `AS`: don't parse keywords, - // which may start a construct allowed in this position, to be parsed as aliases. - // (For example, in `FROM t1 JOIN` the `JOIN` will always be parsed as a keyword, - // not an alias.) - Token::Word(w) if after_as || !reserved_kwds.contains(&w.keyword) => { + // By default, if a word is located after the `AS` keyword we consider it an alias + // as long as it's not reserved. + Token::Word(w) + if after_as || reserved_kwds.is_some_and(|x| !x.contains(&w.keyword)) => + { Ok(Some(w.into_ident(next_token.span))) } - // MSSQL supports single-quoted strings as aliases for columns - // We accept them as table aliases too, although MSSQL does not. - // - // Note, that this conflicts with an obscure rule from the SQL - // standard, which we don't implement: - // https://crate.io/docs/sql-99/en/latest/chapters/07.html#character-string-literal-s - // "[Obscure Rule] SQL allows you to break a long up into two or more smaller s, split by a that includes a newline - // character. When it sees such a , your DBMS will - // ignore the and treat the multiple strings as - // a single ." + // This pattern allows for customizing the acceptance of words as aliases based on the caller's + // context, such as to what SQL element this word is a potential alias of (select item alias, table name + // alias, etc.) or dialect-specific logic that goes beyond a simple list of reserved keywords. + Token::Word(w) if validator(after_as, &w.keyword, self) => { + Ok(Some(w.into_ident(next_token.span))) + } + // For backwards-compatibility, we accept quoted strings as aliases regardless of the context. Token::SingleQuotedString(s) => Ok(Some(Ident::with_quote('\'', s))), - // Support for MySql dialect double-quoted string, `AS "HOUR"` for example Token::DoubleQuotedString(s) => Ok(Some(Ident::with_quote('\"', s))), _ => { if after_as { @@ -8881,23 +8919,6 @@ impl<'a> Parser<'a> { } } - /// Parse `AS identifier` when the AS is describing a table-valued object, - /// like in `... FROM generate_series(1, 10) AS t (col)`. In this case - /// the alias is allowed to optionally name the columns in the table, in - /// addition to the table itself. - pub fn parse_optional_table_alias( - &mut self, - reserved_kwds: &[Keyword], - ) -> Result, ParserError> { - match self.parse_optional_alias(reserved_kwds)? { - Some(name) => { - let columns = self.parse_table_alias_column_defs()?; - Ok(Some(TableAlias { name, columns })) - } - None => Ok(None), - } - } - pub fn parse_optional_group_by(&mut self) -> Result, ParserError> { if self.parse_keywords(&[Keyword::GROUP, Keyword::BY]) { let expressions = if self.parse_keyword(Keyword::ALL) { @@ -10899,7 +10920,7 @@ impl<'a> Parser<'a> { let name = self.parse_object_name(false)?; self.expect_token(&Token::LParen)?; let args = self.parse_optional_args()?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Function { lateral: true, name, @@ -10912,7 +10933,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; let expr = self.parse_expr()?; self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::TableFunction { expr, alias }) } else if self.consume_token(&Token::LParen) { // A left paren introduces either a derived table (i.e., a subquery) @@ -10961,7 +10982,7 @@ impl<'a> Parser<'a> { #[allow(clippy::if_same_then_else)] if !table_and_joins.joins.is_empty() { self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::NestedJoin { table_with_joins: Box::new(table_and_joins), alias, @@ -10974,7 +10995,7 @@ impl<'a> Parser<'a> { // (B): `table_and_joins` (what we found inside the parentheses) // is a nested join `(foo JOIN bar)`, not followed by other joins. self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::NestedJoin { table_with_joins: Box::new(table_and_joins), alias, @@ -10988,9 +11009,7 @@ impl<'a> Parser<'a> { // [AS alias])`) as well. self.expect_token(&Token::RParen)?; - if let Some(outer_alias) = - self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)? - { + if let Some(outer_alias) = self.maybe_parse_table_alias()? { // Snowflake also allows specifying an alias *after* parens // e.g. `FROM (mytable) AS alias` match &mut table_and_joins.relation { @@ -11043,7 +11062,7 @@ impl<'a> Parser<'a> { // SELECT * FROM VALUES (1, 'a'), (2, 'b') AS t (col1, col2) // where there are no parentheses around the VALUES clause. let values = SetExpr::Values(self.parse_values(false)?); - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Derived { lateral: false, subquery: Box::new(Query { @@ -11069,7 +11088,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::RParen)?; let with_ordinality = self.parse_keywords(&[Keyword::WITH, Keyword::ORDINALITY]); - let alias = match self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS) { + let alias = match self.maybe_parse_table_alias() { Ok(Some(alias)) => Some(alias), Ok(None) => None, Err(e) => return Err(e), @@ -11106,7 +11125,7 @@ impl<'a> Parser<'a> { let columns = self.parse_comma_separated(Parser::parse_json_table_column_def)?; self.expect_token(&Token::RParen)?; self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::JsonTable { json_expr, json_path, @@ -11151,7 +11170,7 @@ impl<'a> Parser<'a> { } } - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; // MSSQL-specific table hints: let mut with_hints = vec![]; @@ -11329,7 +11348,7 @@ impl<'a> Parser<'a> { } else { Vec::new() }; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::OpenJsonTable { json_expr, json_path, @@ -11428,7 +11447,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::MatchRecognize { table: Box::new(table), @@ -11672,7 +11691,7 @@ impl<'a> Parser<'a> { ) -> Result { let subquery = self.parse_query()?; self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Derived { lateral: match lateral { Lateral => true, @@ -11766,7 +11785,7 @@ impl<'a> Parser<'a> { }; self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Pivot { table: Box::new(table), aggregate_functions, @@ -11788,7 +11807,7 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::IN)?; let columns = self.parse_parenthesized_column_list(Mandatory, false)?; self.expect_token(&Token::RParen)?; - let alias = self.parse_optional_table_alias(keywords::RESERVED_FOR_TABLE_ALIAS)?; + let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Unpivot { table: Box::new(table), value, @@ -12614,7 +12633,7 @@ impl<'a> Parser<'a> { }) } expr => self - .parse_optional_alias(keywords::RESERVED_FOR_COLUMN_ALIAS) + .maybe_parse_select_item_alias() .map(|alias| match alias { Some(alias) => SelectItem::ExprWithAlias { expr, alias }, None => SelectItem::UnnamedExpr(expr), diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 112aa5264..fe6439b36 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3022,3 +3022,32 @@ fn parse_ls_and_rm() { snowflake().verified_stmt(r#"LIST @"STAGE_WITH_QUOTES""#); } + +#[test] +fn test_sql_keywords_as_select_item_aliases() { + // Some keywords that should be parsed as an alias + let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT"]; + for kw in unreserved_kws { + snowflake() + .one_statement_parses_to(&format!("SELECT 1 {kw}"), &format!("SELECT 1 AS {kw}")); + } + + // Some keywords that should not be parsed as an alias + let reserved_kws = vec![ + "FROM", + "GROUP", + "HAVING", + "INTERSECT", + "INTO", + "ORDER", + "SELECT", + "UNION", + "WHERE", + "WITH", + ]; + for kw in reserved_kws { + assert!(snowflake() + .parse_sql_statements(&format!("SELECT 1 {kw}")) + .is_err()); + } +} From 3eeb9160eadbe01e10c011cf144ef0fbd3f578d6 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Fri, 17 Jan 2025 08:13:12 +0100 Subject: [PATCH 004/286] fix parsing of `INSERT INTO ... SELECT ... RETURNING ` (#1661) --- src/keywords.rs | 1 + tests/sqlparser_common.rs | 21 +++++++++++++++++++++ 2 files changed, 22 insertions(+) diff --git a/src/keywords.rs b/src/keywords.rs index 6b09a877a..cc56329f4 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -946,6 +946,7 @@ pub const RESERVED_FOR_TABLE_ALIAS: &[Keyword] = &[ Keyword::GLOBAL, Keyword::ANTI, Keyword::SEMI, + Keyword::RETURNING, // for MSSQL-specific OUTER APPLY (seems reserved in most dialects) Keyword::OUTER, Keyword::SET, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 07a30bc08..0c44fb849 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -265,6 +265,27 @@ fn parse_insert_select_returning() { } } +#[test] +fn parse_insert_select_from_returning() { + let sql = "INSERT INTO table1 SELECT * FROM table2 RETURNING id"; + match verified_stmt(sql) { + Statement::Insert(Insert { + table: TableObject::TableName(table_name), + source: Some(source), + returning: Some(returning), + .. + }) => { + assert_eq!("table1", table_name.to_string()); + assert!(matches!(*source.body, SetExpr::Select(_))); + assert_eq!( + returning, + vec![SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("id"))),] + ); + } + bad_stmt => unreachable!("Expected valid insert, got {:?}", bad_stmt), + } +} + #[test] fn parse_returning_as_column_alias() { verified_stmt("SELECT 1 AS RETURNING"); From e9498d538aea161283767acc441d70d612dddc00 Mon Sep 17 00:00:00 2001 From: Alexander Beedie Date: Fri, 17 Jan 2025 13:59:47 +0400 Subject: [PATCH 005/286] Add support for `IS [NOT] [form] NORMALIZED` (#1655) Co-authored-by: Alexander Beedie --- src/ast/mod.rs | 30 ++++++++++++-- src/ast/query.rs | 4 +- src/ast/spans.rs | 7 +++- src/ast/value.rs | 29 ++++++++++++++ src/keywords.rs | 5 +++ src/parser/mod.rs | 41 ++++++++++++++++--- tests/sqlparser_common.rs | 84 +++++++++++++++++++++++++++++++++++++-- tests/sqlparser_mysql.rs | 2 +- 8 files changed, 185 insertions(+), 17 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index e6499e14a..7992596e3 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -83,7 +83,7 @@ pub use self::trigger::{ pub use self::value::{ escape_double_quote_string, escape_quoted_string, DateTimeField, DollarQuotedString, - TrimWhereField, Value, + NormalizationForm, TrimWhereField, Value, }; use crate::ast::helpers::stmt_data_loading::{ @@ -653,6 +653,12 @@ pub enum Expr { IsDistinctFrom(Box, Box), /// `IS NOT DISTINCT FROM` operator IsNotDistinctFrom(Box, Box), + /// ` IS [ NOT ] [ form ] NORMALIZED` + IsNormalized { + expr: Box, + form: Option, + negated: bool, + }, /// `[ NOT ] IN (val1, val2, ...)` InList { expr: Box, @@ -1118,7 +1124,7 @@ impl fmt::Display for LambdaFunction { /// `OneOrManyWithParens` implements `Deref` and `IntoIterator`, /// so you can call slice methods on it and iterate over items /// # Examples -/// Acessing as a slice: +/// Accessing as a slice: /// ``` /// # use sqlparser::ast::OneOrManyWithParens; /// let one = OneOrManyWithParens::One("a"); @@ -1419,6 +1425,24 @@ impl fmt::Display for Expr { if *regexp { "REGEXP" } else { "RLIKE" }, pattern ), + Expr::IsNormalized { + expr, + form, + negated, + } => { + let not_ = if *negated { "NOT " } else { "" }; + if form.is_none() { + write!(f, "{} IS {}NORMALIZED", expr, not_) + } else { + write!( + f, + "{} IS {}{} NORMALIZED", + expr, + not_, + form.as_ref().unwrap() + ) + } + } Expr::SimilarTo { negated, expr, @@ -7799,7 +7823,7 @@ where /// ```sql /// EXPLAIN (ANALYZE, VERBOSE TRUE, FORMAT TEXT) SELECT * FROM my_table; /// -/// VACCUM (VERBOSE, ANALYZE ON, PARALLEL 10) my_table; +/// VACUUM (VERBOSE, ANALYZE ON, PARALLEL 10) my_table; /// ``` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/query.rs b/src/ast/query.rs index e7020ae23..9bcdc2e74 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2821,10 +2821,10 @@ impl fmt::Display for ValueTableMode { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum UpdateTableFromKind { - /// Update Statment where the 'FROM' clause is before the 'SET' keyword (Supported by Snowflake) + /// Update Statement where the 'FROM' clause is before the 'SET' keyword (Supported by Snowflake) /// For Example: `UPDATE FROM t1 SET t1.name='aaa'` BeforeSet(TableWithJoins), - /// Update Statment where the 'FROM' clause is after the 'SET' keyword (Which is the standard way) + /// Update Statement where the 'FROM' clause is after the 'SET' keyword (Which is the standard way) /// For Example: `UPDATE SET t1.name='aaa' FROM t1` AfterSet(TableWithJoins), } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 6f89cd0db..2a5a75b49 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1325,6 +1325,12 @@ impl Spanned for Expr { escape_char: _, any: _, } => expr.span().union(&pattern.span()), + Expr::RLike { .. } => Span::empty(), + Expr::IsNormalized { + expr, + form: _, + negated: _, + } => expr.span(), Expr::SimilarTo { negated: _, expr, @@ -1360,7 +1366,6 @@ impl Spanned for Expr { Expr::Array(array) => array.span(), Expr::MatchAgainst { .. } => Span::empty(), Expr::JsonAccess { value, path } => value.span().union(&path.span()), - Expr::RLike { .. } => Span::empty(), Expr::AnyOp { left, compare_op: _, diff --git a/src/ast/value.rs b/src/ast/value.rs index 45cc06a07..1b16646be 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -270,6 +270,35 @@ impl fmt::Display for DateTimeField { } } +#[derive(Debug, Clone, PartialEq, Eq, Ord, PartialOrd, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// The Unicode Standard defines four normalization forms, which are intended to eliminate +/// certain distinctions between visually or functionally identical characters. +/// +/// See [Unicode Normalization Forms](https://unicode.org/reports/tr15/) for details. +pub enum NormalizationForm { + /// Canonical Decomposition, followed by Canonical Composition. + NFC, + /// Canonical Decomposition. + NFD, + /// Compatibility Decomposition, followed by Canonical Composition. + NFKC, + /// Compatibility Decomposition. + NFKD, +} + +impl fmt::Display for NormalizationForm { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + NormalizationForm::NFC => write!(f, "NFC"), + NormalizationForm::NFD => write!(f, "NFD"), + NormalizationForm::NFKC => write!(f, "NFKC"), + NormalizationForm::NFKD => write!(f, "NFKD"), + } + } +} + pub struct EscapeQuotedString<'a> { string: &'a str, quote: char, diff --git a/src/keywords.rs b/src/keywords.rs index cc56329f4..0d1bab9e5 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -530,6 +530,10 @@ define_keywords!( NESTED, NEW, NEXT, + NFC, + NFD, + NFKC, + NFKD, NO, NOBYPASSRLS, NOCREATEDB, @@ -540,6 +544,7 @@ define_keywords!( NOORDER, NOREPLICATION, NORMALIZE, + NORMALIZED, NOSCAN, NOSUPERUSER, NOT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f34a5d742..5cacfda98 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3184,9 +3184,11 @@ impl<'a> Parser<'a> { { let expr2 = self.parse_expr()?; Ok(Expr::IsNotDistinctFrom(Box::new(expr), Box::new(expr2))) + } else if let Ok(is_normalized) = self.parse_unicode_is_normalized(expr) { + Ok(is_normalized) } else { self.expected( - "[NOT] NULL or TRUE|FALSE or [NOT] DISTINCT FROM after IS", + "[NOT] NULL | TRUE | FALSE | DISTINCT | [form] NORMALIZED FROM after IS", self.peek_token(), ) } @@ -3851,7 +3853,7 @@ impl<'a> Parser<'a> { /// If the current token is the `expected` keyword, consume the token. /// Otherwise, return an error. /// - // todo deprecate infavor of expected_keyword_is + // todo deprecate in favor of expected_keyword_is pub fn expect_keyword(&mut self, expected: Keyword) -> Result { if self.parse_keyword(expected) { Ok(self.get_current_token().clone()) @@ -8453,6 +8455,33 @@ impl<'a> Parser<'a> { } } + /// Parse a literal unicode normalization clause + pub fn parse_unicode_is_normalized(&mut self, expr: Expr) -> Result { + let neg = self.parse_keyword(Keyword::NOT); + let normalized_form = self.maybe_parse(|parser| { + match parser.parse_one_of_keywords(&[ + Keyword::NFC, + Keyword::NFD, + Keyword::NFKC, + Keyword::NFKD, + ]) { + Some(Keyword::NFC) => Ok(NormalizationForm::NFC), + Some(Keyword::NFD) => Ok(NormalizationForm::NFD), + Some(Keyword::NFKC) => Ok(NormalizationForm::NFKC), + Some(Keyword::NFKD) => Ok(NormalizationForm::NFKD), + _ => parser.expected("unicode normalization form", parser.peek_token()), + } + })?; + if self.parse_keyword(Keyword::NORMALIZED) { + return Ok(Expr::IsNormalized { + expr: Box::new(expr), + form: normalized_form, + negated: neg, + }); + } + self.expected("unicode normalization form", self.peek_token()) + } + pub fn parse_enum_values(&mut self) -> Result, ParserError> { self.expect_token(&Token::LParen)?; let values = self.parse_comma_separated(|parser| { @@ -8979,7 +9008,7 @@ impl<'a> Parser<'a> { } } - /// Parse a table object for insetion + /// Parse a table object for insertion /// e.g. `some_database.some_table` or `FUNCTION some_table_func(...)` pub fn parse_table_object(&mut self) -> Result { if self.dialect.supports_insert_table_function() && self.parse_keyword(Keyword::FUNCTION) { @@ -11887,7 +11916,7 @@ impl<'a> Parser<'a> { } else { let mut name = self.parse_grantee_name()?; if self.consume_token(&Token::Colon) { - // Redshift supports namespace prefix for extenrnal users and groups: + // Redshift supports namespace prefix for external users and groups: // : or : // https://docs.aws.amazon.com/redshift/latest/mgmt/redshift-iam-access-control-native-idp.html let ident = self.parse_identifier()?; @@ -12883,7 +12912,7 @@ impl<'a> Parser<'a> { Ok(WithFill { from, to, step }) } - // Parse a set of comma seperated INTERPOLATE expressions (ClickHouse dialect) + // Parse a set of comma separated INTERPOLATE expressions (ClickHouse dialect) // that follow the INTERPOLATE keyword in an ORDER BY clause with the WITH FILL modifier pub fn parse_interpolations(&mut self) -> Result, ParserError> { if !self.parse_keyword(Keyword::INTERPOLATE) { @@ -14432,7 +14461,7 @@ mod tests { assert_eq!( ast, Err(ParserError::ParserError( - "Expected: [NOT] NULL or TRUE|FALSE or [NOT] DISTINCT FROM after IS, found: a at Line: 1, Column: 16" + "Expected: [NOT] NULL | TRUE | FALSE | DISTINCT | [form] NORMALIZED FROM after IS, found: a at Line: 1, Column: 16" .to_string() )) ); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 0c44fb849..49588a581 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4600,7 +4600,7 @@ fn run_explain_analyze( expected_verbose: bool, expected_analyze: bool, expected_format: Option, - exepcted_options: Option>, + expected_options: Option>, ) { match dialect.verified_stmt(query) { Statement::Explain { @@ -4616,7 +4616,7 @@ fn run_explain_analyze( assert_eq!(verbose, expected_verbose); assert_eq!(analyze, expected_analyze); assert_eq!(format, expected_format); - assert_eq!(options, exepcted_options); + assert_eq!(options, expected_options); assert!(!query_plan); assert!(!estimate); assert_eq!("SELECT sqrt(id) FROM foo", statement.to_string()); @@ -9317,6 +9317,46 @@ fn parse_is_boolean() { verified_expr(sql) ); + let sql = "a IS NORMALIZED"; + assert_eq!( + IsNormalized { + expr: Box::new(Identifier(Ident::new("a"))), + form: None, + negated: false, + }, + verified_expr(sql) + ); + + let sql = "a IS NOT NORMALIZED"; + assert_eq!( + IsNormalized { + expr: Box::new(Identifier(Ident::new("a"))), + form: None, + negated: true, + }, + verified_expr(sql) + ); + + let sql = "a IS NFKC NORMALIZED"; + assert_eq!( + IsNormalized { + expr: Box::new(Identifier(Ident::new("a"))), + form: Some(NormalizationForm::NFKC), + negated: false, + }, + verified_expr(sql) + ); + + let sql = "a IS NOT NFKD NORMALIZED"; + assert_eq!( + IsNormalized { + expr: Box::new(Identifier(Ident::new("a"))), + form: Some(NormalizationForm::NFKD), + negated: true, + }, + verified_expr(sql) + ); + let sql = "a IS UNKNOWN"; assert_eq!( IsUnknown(Box::new(Identifier(Ident::new("a")))), @@ -9335,6 +9375,12 @@ fn parse_is_boolean() { verified_stmt("SELECT f FROM foo WHERE field IS FALSE"); verified_stmt("SELECT f FROM foo WHERE field IS NOT FALSE"); + verified_stmt("SELECT f FROM foo WHERE field IS NORMALIZED"); + verified_stmt("SELECT f FROM foo WHERE field IS NFC NORMALIZED"); + verified_stmt("SELECT f FROM foo WHERE field IS NFD NORMALIZED"); + verified_stmt("SELECT f FROM foo WHERE field IS NOT NORMALIZED"); + verified_stmt("SELECT f FROM foo WHERE field IS NOT NFKC NORMALIZED"); + verified_stmt("SELECT f FROM foo WHERE field IS UNKNOWN"); verified_stmt("SELECT f FROM foo WHERE field IS NOT UNKNOWN"); @@ -9342,7 +9388,37 @@ fn parse_is_boolean() { let res = parse_sql_statements(sql); assert_eq!( ParserError::ParserError( - "Expected: [NOT] NULL or TRUE|FALSE or [NOT] DISTINCT FROM after IS, found: 0" + "Expected: [NOT] NULL | TRUE | FALSE | DISTINCT | [form] NORMALIZED FROM after IS, found: 0" + .to_string() + ), + res.unwrap_err() + ); + + let sql = "SELECT s, s IS XYZ NORMALIZED FROM foo"; + let res = parse_sql_statements(sql); + assert_eq!( + ParserError::ParserError( + "Expected: [NOT] NULL | TRUE | FALSE | DISTINCT | [form] NORMALIZED FROM after IS, found: XYZ" + .to_string() + ), + res.unwrap_err() + ); + + let sql = "SELECT s, s IS NFKC FROM foo"; + let res = parse_sql_statements(sql); + assert_eq!( + ParserError::ParserError( + "Expected: [NOT] NULL | TRUE | FALSE | DISTINCT | [form] NORMALIZED FROM after IS, found: FROM" + .to_string() + ), + res.unwrap_err() + ); + + let sql = "SELECT s, s IS TRIM(' NFKC ') FROM foo"; + let res = parse_sql_statements(sql); + assert_eq!( + ParserError::ParserError( + "Expected: [NOT] NULL | TRUE | FALSE | DISTINCT | [form] NORMALIZED FROM after IS, found: TRIM" .to_string() ), res.unwrap_err() @@ -13003,7 +13079,7 @@ fn test_trailing_commas_in_from() { let sql = "SELECT a FROM b, WHERE c = 1"; let _ = dialects.parse_sql_statements(sql).unwrap(); - // nasted + // nested let sql = "SELECT 1, 2 FROM (SELECT * FROM t,),"; let _ = dialects.parse_sql_statements(sql).unwrap(); diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index dcf3f57fe..e93ac5695 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2572,7 +2572,7 @@ fn parse_kill() { } #[test] -fn parse_table_colum_option_on_update() { +fn parse_table_column_option_on_update() { let sql1 = "CREATE TABLE foo (`modification_time` DATETIME ON UPDATE CURRENT_TIMESTAMP())"; match mysql().verified_stmt(sql1) { Statement::CreateTable(CreateTable { name, columns, .. }) => { From 44df6d6f92e8bba1cbc6d4b57c99f8f126b786fd Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sun, 19 Jan 2025 11:43:45 +0100 Subject: [PATCH 006/286] Add support for qualified column names in JOIN ... USING (#1663) --- src/ast/query.rs | 2 +- src/ast/spans.rs | 2 +- src/parser/mod.rs | 35 +++++++++++++++++++++++++++++++---- tests/sqlparser_common.rs | 3 ++- 4 files changed, 35 insertions(+), 7 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 9bcdc2e74..66c70b466 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2050,7 +2050,7 @@ pub enum JoinOperator { #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum JoinConstraint { On(Expr), - Using(Vec), + Using(Vec), Natural, None, } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 2a5a75b49..1ddd47d7f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2008,7 +2008,7 @@ impl Spanned for JoinConstraint { fn span(&self) -> Span { match self { JoinConstraint::On(expr) => expr.span(), - JoinConstraint::Using(vec) => union_spans(vec.iter().map(|i| i.span)), + JoinConstraint::Using(vec) => union_spans(vec.iter().map(|i| i.span())), JoinConstraint::Natural => Span::empty(), JoinConstraint::None => Span::empty(), } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 5cacfda98..a3adb0235 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9336,18 +9336,45 @@ impl<'a> Parser<'a> { }) } - /// Parse a parenthesized comma-separated list of unqualified, possibly quoted identifiers + /// Parses a parenthesized comma-separated list of unqualified, possibly quoted identifiers. + /// For example: `(col1, "col 2", ...)` pub fn parse_parenthesized_column_list( &mut self, optional: IsOptional, allow_empty: bool, ) -> Result, ParserError> { + self.parse_parenthesized_column_list_inner(optional, allow_empty, |p| p.parse_identifier()) + } + + /// Parses a parenthesized comma-separated list of qualified, possibly quoted identifiers. + /// For example: `(db1.sc1.tbl1.col1, db1.sc1.tbl1."col 2", ...)` + pub fn parse_parenthesized_qualified_column_list( + &mut self, + optional: IsOptional, + allow_empty: bool, + ) -> Result, ParserError> { + self.parse_parenthesized_column_list_inner(optional, allow_empty, |p| { + p.parse_object_name(true) + }) + } + + /// Parses a parenthesized comma-separated list of columns using + /// the provided function to parse each element. + fn parse_parenthesized_column_list_inner( + &mut self, + optional: IsOptional, + allow_empty: bool, + mut f: F, + ) -> Result, ParserError> + where + F: FnMut(&mut Parser) -> Result, + { if self.consume_token(&Token::LParen) { if allow_empty && self.peek_token().token == Token::RParen { self.next_token(); Ok(vec![]) } else { - let cols = self.parse_comma_separated(|p| p.parse_identifier())?; + let cols = self.parse_comma_separated(|p| f(p))?; self.expect_token(&Token::RParen)?; Ok(cols) } @@ -9358,7 +9385,7 @@ impl<'a> Parser<'a> { } } - /// Parse a parenthesized comma-separated list of table alias column definitions. + /// Parses a parenthesized comma-separated list of table alias column definitions. fn parse_table_alias_column_defs(&mut self) -> Result, ParserError> { if self.consume_token(&Token::LParen) { let cols = self.parse_comma_separated(|p| { @@ -11853,7 +11880,7 @@ impl<'a> Parser<'a> { let constraint = self.parse_expr()?; Ok(JoinConstraint::On(constraint)) } else if self.parse_keyword(Keyword::USING) { - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_qualified_column_list(Mandatory, false)?; Ok(JoinConstraint::Using(columns)) } else { Ok(JoinConstraint::None) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 49588a581..7271d6375 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -6541,7 +6541,7 @@ fn parse_joins_using() { sample: None, }, global: false, - join_operator: f(JoinConstraint::Using(vec!["c1".into()])), + join_operator: f(JoinConstraint::Using(vec![ObjectName(vec!["c1".into()])])), } } // Test parsing of aliases @@ -6598,6 +6598,7 @@ fn parse_joins_using() { only(&verified_only_select("SELECT * FROM t1 FULL JOIN t2 USING(c1)").from).joins, vec![join_with_constraint("t2", None, JoinOperator::FullOuter)] ); + verified_stmt("SELECT * FROM tbl1 AS t1 JOIN tbl2 AS t2 USING(t2.col1)"); } #[test] From 5da702fc19f9dc73559d9a6f0408729f1121444a Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sun, 19 Jan 2025 12:08:45 +0100 Subject: [PATCH 007/286] Add support for Snowflake AT/BEFORE (#1667) --- src/ast/query.rs | 6 ++++++ src/dialect/bigquery.rs | 5 +++++ src/dialect/mod.rs | 6 ++++++ src/dialect/mssql.rs | 5 +++++ src/dialect/snowflake.rs | 5 +++++ src/parser/mod.rs | 26 ++++++++++++++------------ tests/sqlparser_snowflake.rs | 7 +++++++ 7 files changed, 48 insertions(+), 12 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 66c70b466..4053dd239 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1873,13 +1873,19 @@ impl fmt::Display for TableAliasColumnDef { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum TableVersion { + /// When the table version is defined using `FOR SYSTEM_TIME AS OF`. + /// For example: `SELECT * FROM tbl FOR SYSTEM_TIME AS OF TIMESTAMP_SUB(CURRENT_TIMESTAMP(), INTERVAL 1 HOUR)` ForSystemTimeAsOf(Expr), + /// When the table version is defined using a function. + /// For example: `SELECT * FROM tbl AT(TIMESTAMP => '2020-08-14 09:30:00')` + Function(Expr), } impl Display for TableVersion { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { TableVersion::ForSystemTimeAsOf(e) => write!(f, " FOR SYSTEM_TIME AS OF {e}")?, + TableVersion::Function(func) => write!(f, " {func}")?, } Ok(()) } diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index 66d7d2061..e92169a35 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -77,4 +77,9 @@ impl Dialect for BigQueryDialect { fn supports_struct_literal(&self) -> bool { true } + + // See + fn supports_timestamp_versioning(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index c69253b76..119bb3cf7 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -834,6 +834,12 @@ pub trait Dialect: Debug + Any { fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { explicit || !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) } + + /// Returns true if this dialect supports querying historical table data + /// by specifying which version of the data to query. + fn supports_timestamp_versioning(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 67a648944..7d8611cb0 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -90,4 +90,9 @@ impl Dialect for MsSqlDialect { fn supports_set_stmt_without_operator(&self) -> bool { true } + + /// See: + fn supports_timestamp_versioning(&self) -> bool { + true + } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index f6e9c9eba..78237acd0 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -296,6 +296,11 @@ impl Dialect for SnowflakeDialect { _ => true, } } + + /// See: + fn supports_timestamp_versioning(&self) -> bool { + true + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a3adb0235..51bbcfabd 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11208,7 +11208,7 @@ impl<'a> Parser<'a> { }; // Parse potential version qualifier - let version = self.parse_table_version()?; + let version = self.maybe_parse_table_version()?; // Postgres, MSSQL, ClickHouse: table-valued functions: let args = if self.consume_token(&Token::LParen) { @@ -11639,18 +11639,20 @@ impl<'a> Parser<'a> { } } - /// Parse a given table version specifier. - /// - /// For now it only supports timestamp versioning for BigQuery and MSSQL dialects. - pub fn parse_table_version(&mut self) -> Result, ParserError> { - if dialect_of!(self is BigQueryDialect | MsSqlDialect) - && self.parse_keywords(&[Keyword::FOR, Keyword::SYSTEM_TIME, Keyword::AS, Keyword::OF]) - { - let expr = self.parse_expr()?; - Ok(Some(TableVersion::ForSystemTimeAsOf(expr))) - } else { - Ok(None) + /// Parses a the timestamp version specifier (i.e. query historical data) + pub fn maybe_parse_table_version(&mut self) -> Result, ParserError> { + if self.dialect.supports_timestamp_versioning() { + if self.parse_keywords(&[Keyword::FOR, Keyword::SYSTEM_TIME, Keyword::AS, Keyword::OF]) + { + let expr = self.parse_expr()?; + return Ok(Some(TableVersion::ForSystemTimeAsOf(expr))); + } else if self.peek_keyword(Keyword::AT) || self.peek_keyword(Keyword::BEFORE) { + let func_name = self.parse_object_name(true)?; + let func = self.parse_function(func_name)?; + return Ok(Some(TableVersion::Function(func))); + } } + Ok(None) } /// Parses MySQL's JSON_TABLE column definition. diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index fe6439b36..0c4bdf149 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3051,3 +3051,10 @@ fn test_sql_keywords_as_select_item_aliases() { .is_err()); } } + +#[test] +fn test_timetravel_at_before() { + snowflake().verified_only_select("SELECT * FROM tbl AT(TIMESTAMP => '2024-12-15 00:00:00')"); + snowflake() + .verified_only_select("SELECT * FROM tbl BEFORE(TIMESTAMP => '2024-12-15 00:00:00')"); +} From 17d5610f098c085e5e89d8edc71df1949defe76d Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Mon, 20 Jan 2025 11:32:12 -0500 Subject: [PATCH 008/286] Update verson to 0.54.0 and update changelog (#1668) --- CHANGELOG.md | 6 ++- Cargo.toml | 2 +- changelog/0.54.0.md | 118 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 124 insertions(+), 2 deletions(-) create mode 100644 changelog/0.54.0.md diff --git a/CHANGELOG.md b/CHANGELOG.md index ec74bf633..d1c55b285 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -18,13 +18,17 @@ --> # Changelog -All notable changes to this project will be documented in this file. +All notable changes to this project will be documented in one of the linked +files. This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). + Given that the parser produces a typed AST, any changes to the AST will technically be breaking and thus will result in a `0.(N+1)` version. - Unreleased: Check https://github.com/sqlparser-rs/sqlparser-rs/commits/main for undocumented changes. +- `0.54.0`: [changelog/0.54.0.md](changelog/0.54.0.md) +- `0.53.0`: [changelog/0.53.0.md](changelog/0.53.0.md) - `0.52.0`: [changelog/0.52.0.md](changelog/0.52.0.md) - `0.51.0` and earlier: [changelog/0.51.0-pre.md](changelog/0.51.0-pre.md) diff --git a/Cargo.toml b/Cargo.toml index 8ff0ceb55..06fed2c68 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.53.0" +version = "0.54.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/changelog/0.54.0.md b/changelog/0.54.0.md new file mode 100644 index 000000000..c0a63ae45 --- /dev/null +++ b/changelog/0.54.0.md @@ -0,0 +1,118 @@ + + +# sqlparser-rs 0.54.0 Changelog + +This release consists of 57 commits from 24 contributors. See credits at the end of this changelog for more information. + +**Implemented enhancements:** + +- feat: support `INSERT INTO [TABLE] FUNCTION` of Clickhouse [#1633](https://github.com/apache/datafusion-sqlparser-rs/pull/1633) (byte-sourcerer) + +**Other:** + +- Run cargo fmt on `derive` crate [#1595](https://github.com/apache/datafusion-sqlparser-rs/pull/1595) (alamb) +- Add Apache license header to spans.rs [#1594](https://github.com/apache/datafusion-sqlparser-rs/pull/1594) (alamb) +- Add support for BigQuery `ANY TYPE` data type [#1602](https://github.com/apache/datafusion-sqlparser-rs/pull/1602) (MartinSahlen) +- Add support for TABLESAMPLE [#1580](https://github.com/apache/datafusion-sqlparser-rs/pull/1580) (yoavcloud) +- Redshift: Fix parsing for quoted numbered columns [#1576](https://github.com/apache/datafusion-sqlparser-rs/pull/1576) (7phs) +- Add the alter table ON COMMIT option to Snowflake [#1606](https://github.com/apache/datafusion-sqlparser-rs/pull/1606) (yoavcloud) +- Support parsing `EXPLAIN ESTIMATE` of Clickhouse [#1605](https://github.com/apache/datafusion-sqlparser-rs/pull/1605) (byte-sourcerer) +- Fix BigQuery hyphenated ObjectName with numbers [#1598](https://github.com/apache/datafusion-sqlparser-rs/pull/1598) (ayman-sigma) +- Fix test compilation issue [#1609](https://github.com/apache/datafusion-sqlparser-rs/pull/1609) (iffyio) +- Allow foreign table constraint without columns [#1608](https://github.com/apache/datafusion-sqlparser-rs/pull/1608) (ramnivas) +- Support optional table for `ANALYZE` statement [#1599](https://github.com/apache/datafusion-sqlparser-rs/pull/1599) (yuyang-ok) +- Support DOUBLE data types with precision for Mysql [#1611](https://github.com/apache/datafusion-sqlparser-rs/pull/1611) (artorias1024) +- Add `#[recursive]` [#1522](https://github.com/apache/datafusion-sqlparser-rs/pull/1522) (blaginin) +- Support arbitrary composite access expressions [#1600](https://github.com/apache/datafusion-sqlparser-rs/pull/1600) (ayman-sigma) +- Consolidate `MapAccess`, and `Subscript` into `CompoundExpr` to handle the complex field access chain [#1551](https://github.com/apache/datafusion-sqlparser-rs/pull/1551) (goldmedal) +- Handle empty projection in Postgres SELECT statements [#1613](https://github.com/apache/datafusion-sqlparser-rs/pull/1613) (tobyhede) +- Merge composite and compound expr test cases [#1615](https://github.com/apache/datafusion-sqlparser-rs/pull/1615) (iffyio) +- Support Snowflake Update-From-Select [#1604](https://github.com/apache/datafusion-sqlparser-rs/pull/1604) (yuval-illumex) +- Improve parsing performance by reducing token cloning [#1587](https://github.com/apache/datafusion-sqlparser-rs/pull/1587) (davisp) +- Improve Parser documentation [#1617](https://github.com/apache/datafusion-sqlparser-rs/pull/1617) (alamb) +- Add support for DROP EXTENSION [#1610](https://github.com/apache/datafusion-sqlparser-rs/pull/1610) (ramnivas) +- Refactor advancing token to avoid duplication, avoid borrow checker issues [#1618](https://github.com/apache/datafusion-sqlparser-rs/pull/1618) (alamb) +- Fix the parsing result for the special double number [#1621](https://github.com/apache/datafusion-sqlparser-rs/pull/1621) (goldmedal) +- SQLite: Allow dollar signs in placeholder names [#1620](https://github.com/apache/datafusion-sqlparser-rs/pull/1620) (hansott) +- Improve error for an unexpected token after DROP [#1623](https://github.com/apache/datafusion-sqlparser-rs/pull/1623) (ramnivas) +- Fix `sqlparser_bench` benchmark compilation [#1625](https://github.com/apache/datafusion-sqlparser-rs/pull/1625) (alamb) +- Improve parsing speed by avoiding some clones in parse_identifier [#1624](https://github.com/apache/datafusion-sqlparser-rs/pull/1624) (alamb) +- Simplify `parse_keyword_apis` more [#1626](https://github.com/apache/datafusion-sqlparser-rs/pull/1626) (alamb) +- Test benchmarks and Improve benchmark README.md [#1627](https://github.com/apache/datafusion-sqlparser-rs/pull/1627) (alamb) +- Add support for MYSQL's `RENAME TABLE` [#1616](https://github.com/apache/datafusion-sqlparser-rs/pull/1616) (wugeer) +- Correctly tokenize nested comments [#1629](https://github.com/apache/datafusion-sqlparser-rs/pull/1629) (hansott) +- Add support for USE SECONDARY ROLE (vs. ROLES) [#1637](https://github.com/apache/datafusion-sqlparser-rs/pull/1637) (yoavcloud) +- Add support for various Snowflake grantees [#1640](https://github.com/apache/datafusion-sqlparser-rs/pull/1640) (yoavcloud) +- Add support for the SQL OVERLAPS predicate [#1638](https://github.com/apache/datafusion-sqlparser-rs/pull/1638) (yoavcloud) +- Add support for Snowflake LIST and REMOVE [#1639](https://github.com/apache/datafusion-sqlparser-rs/pull/1639) (yoavcloud) +- Add support for MySQL's INSERT INTO ... SET syntax [#1641](https://github.com/apache/datafusion-sqlparser-rs/pull/1641) (yoavcloud) +- Start new line if \r in Postgres dialect [#1647](https://github.com/apache/datafusion-sqlparser-rs/pull/1647) (hansott) +- Support pluralized time units [#1630](https://github.com/apache/datafusion-sqlparser-rs/pull/1630) (wugeer) +- Replace `ReferentialAction` enum in `DROP` statements [#1648](https://github.com/apache/datafusion-sqlparser-rs/pull/1648) (stepancheg) +- Add support for MS-SQL BEGIN/END TRY/CATCH [#1649](https://github.com/apache/datafusion-sqlparser-rs/pull/1649) (yoavcloud) +- Fix MySQL parsing of GRANT, REVOKE, and CREATE VIEW [#1538](https://github.com/apache/datafusion-sqlparser-rs/pull/1538) (mvzink) +- Add support for the Snowflake MINUS set operator [#1652](https://github.com/apache/datafusion-sqlparser-rs/pull/1652) (yoavcloud) +- ALTER TABLE DROP {COLUMN|CONSTRAINT} RESTRICT [#1651](https://github.com/apache/datafusion-sqlparser-rs/pull/1651) (stepancheg) +- Add support for ClickHouse `FORMAT` on `INSERT` [#1628](https://github.com/apache/datafusion-sqlparser-rs/pull/1628) (bombsimon) +- MsSQL SET for session params [#1646](https://github.com/apache/datafusion-sqlparser-rs/pull/1646) (yoavcloud) +- Correctly look for end delimiter dollar quoted string [#1650](https://github.com/apache/datafusion-sqlparser-rs/pull/1650) (hansott) +- Support single line comments starting with '#' for Hive [#1654](https://github.com/apache/datafusion-sqlparser-rs/pull/1654) (wugeer) +- Support trailing commas in `FROM` clause [#1645](https://github.com/apache/datafusion-sqlparser-rs/pull/1645) (barsela1) +- Allow empty options for BigQuery [#1657](https://github.com/apache/datafusion-sqlparser-rs/pull/1657) (MartinSahlen) +- Add support for parsing RAISERROR [#1656](https://github.com/apache/datafusion-sqlparser-rs/pull/1656) (AvivDavid-Satori) +- Add support for Snowflake column aliases that use SQL keywords [#1632](https://github.com/apache/datafusion-sqlparser-rs/pull/1632) (yoavcloud) +- fix parsing of `INSERT INTO ... SELECT ... RETURNING ` [#1661](https://github.com/apache/datafusion-sqlparser-rs/pull/1661) (lovasoa) +- Add support for `IS [NOT] [form] NORMALIZED` [#1655](https://github.com/apache/datafusion-sqlparser-rs/pull/1655) (alexander-beedie) +- Add support for qualified column names in JOIN ... USING [#1663](https://github.com/apache/datafusion-sqlparser-rs/pull/1663) (yoavcloud) +- Add support for Snowflake AT/BEFORE [#1667](https://github.com/apache/datafusion-sqlparser-rs/pull/1667) (yoavcloud) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 13 Yoav Cohen + 9 Andrew Lamb + 4 Hans Ott + 3 Ramnivas Laddad + 3 wugeer + 2 Ayman Elkfrawy + 2 Ifeanyi Ubah + 2 Jax Liu + 2 Martin Abelson Sahlen + 2 Stepan Koltsov + 2 cjw + 1 Aleksei Piianin + 1 Alexander Beedie + 1 AvivDavid-Satori + 1 Dmitrii Blaginin + 1 Michael Victor Zink + 1 Ophir LOJKINE + 1 Paul J. Davis + 1 Simon Sawert + 1 Toby Hede + 1 Yuval Shkolar + 1 artorias1024 + 1 bar sela + 1 yuyang +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + From e5bc3dfad879ea15dcbe28b85d92697193298f08 Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Mon, 20 Jan 2025 14:19:14 -0500 Subject: [PATCH 009/286] Update rat_exclude_file.txt (#1670) --- dev/release/rat_exclude_files.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/dev/release/rat_exclude_files.txt b/dev/release/rat_exclude_files.txt index a567eda9c..562eec2f1 100644 --- a/dev/release/rat_exclude_files.txt +++ b/dev/release/rat_exclude_files.txt @@ -3,4 +3,5 @@ .tool-versions dev/release/rat_exclude_files.txt fuzz/.gitignore +sqlparser_bench/img/flamegraph.svg From 183274e274dc0f98f4946e0eb55d5d4dcf63cadb Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 20 Jan 2025 20:20:41 +0100 Subject: [PATCH 010/286] Add support for Snowflake account privileges (#1666) --- src/ast/mod.rs | 274 +++++++++++++++++++++++++++++++- src/keywords.rs | 41 +++++ src/parser/mod.rs | 296 +++++++++++++++++++++++++++-------- tests/sqlparser_common.rs | 5 +- tests/sqlparser_snowflake.rs | 106 +++++++++++++ 5 files changed, 651 insertions(+), 71 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 7992596e3..5c06d7196 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -5453,29 +5453,107 @@ impl fmt::Display for FetchDirection { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum Action { + AddSearchOptimization, + Apply { + apply_type: ActionApplyType, + }, + ApplyBudget, + AttachListing, + AttachPolicy, + Audit, + BindServiceEndpoint, Connect, - Create, + Create { + obj_type: Option, + }, Delete, - Execute, - Insert { columns: Option> }, - References { columns: Option> }, - Select { columns: Option> }, + EvolveSchema, + Execute { + obj_type: Option, + }, + Failover, + ImportedPrivileges, + ImportShare, + Insert { + columns: Option>, + }, + Manage { + manage_type: ActionManageType, + }, + ManageReleases, + ManageVersions, + Modify { + modify_type: ActionModifyType, + }, + Monitor { + monitor_type: ActionMonitorType, + }, + Operate, + OverrideShareRestrictions, + Ownership, + PurchaseDataExchangeListing, + Read, + ReadSession, + References { + columns: Option>, + }, + Replicate, + ResolveAll, + Select { + columns: Option>, + }, Temporary, Trigger, Truncate, - Update { columns: Option> }, + Update { + columns: Option>, + }, Usage, } impl fmt::Display for Action { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { + Action::AddSearchOptimization => f.write_str("ADD SEARCH OPTIMIZATION")?, + Action::Apply { apply_type } => write!(f, "APPLY {apply_type}")?, + Action::ApplyBudget => f.write_str("APPLY BUDGET")?, + Action::AttachListing => f.write_str("ATTACH LISTING")?, + Action::AttachPolicy => f.write_str("ATTACH POLICY")?, + Action::Audit => f.write_str("AUDIT")?, + Action::BindServiceEndpoint => f.write_str("BIND SERVICE ENDPOINT")?, Action::Connect => f.write_str("CONNECT")?, - Action::Create => f.write_str("CREATE")?, + Action::Create { obj_type } => { + f.write_str("CREATE")?; + if let Some(obj_type) = obj_type { + write!(f, " {obj_type}")? + } + } Action::Delete => f.write_str("DELETE")?, - Action::Execute => f.write_str("EXECUTE")?, + Action::EvolveSchema => f.write_str("EVOLVE SCHEMA")?, + Action::Execute { obj_type } => { + f.write_str("EXECUTE")?; + if let Some(obj_type) = obj_type { + write!(f, " {obj_type}")? + } + } + Action::Failover => f.write_str("FAILOVER")?, + Action::ImportedPrivileges => f.write_str("IMPORTED PRIVILEGES")?, + Action::ImportShare => f.write_str("IMPORT SHARE")?, Action::Insert { .. } => f.write_str("INSERT")?, + Action::Manage { manage_type } => write!(f, "MANAGE {manage_type}")?, + Action::ManageReleases => f.write_str("MANAGE RELEASES")?, + Action::ManageVersions => f.write_str("MANAGE VERSIONS")?, + Action::Modify { modify_type } => write!(f, "MODIFY {modify_type}")?, + Action::Monitor { monitor_type } => write!(f, "MONITOR {monitor_type}")?, + Action::Operate => f.write_str("OPERATE")?, + Action::OverrideShareRestrictions => f.write_str("OVERRIDE SHARE RESTRICTIONS")?, + Action::Ownership => f.write_str("OWNERSHIP")?, + Action::PurchaseDataExchangeListing => f.write_str("PURCHASE DATA EXCHANGE LISTING")?, + Action::Read => f.write_str("READ")?, + Action::ReadSession => f.write_str("READ SESSION")?, Action::References { .. } => f.write_str("REFERENCES")?, + Action::Replicate => f.write_str("REPLICATE")?, + Action::ResolveAll => f.write_str("RESOLVE ALL")?, Action::Select { .. } => f.write_str("SELECT")?, Action::Temporary => f.write_str("TEMPORARY")?, Action::Trigger => f.write_str("TRIGGER")?, @@ -5498,6 +5576,186 @@ impl fmt::Display for Action { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// See +/// under `globalPrivileges` in the `CREATE` privilege. +pub enum ActionCreateObjectType { + Account, + Application, + ApplicationPackage, + ComputePool, + DataExchangeListing, + Database, + ExternalVolume, + FailoverGroup, + Integration, + NetworkPolicy, + OrganiationListing, + ReplicationGroup, + Role, + Share, + User, + Warehouse, +} + +impl fmt::Display for ActionCreateObjectType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ActionCreateObjectType::Account => write!(f, "ACCOUNT"), + ActionCreateObjectType::Application => write!(f, "APPLICATION"), + ActionCreateObjectType::ApplicationPackage => write!(f, "APPLICATION PACKAGE"), + ActionCreateObjectType::ComputePool => write!(f, "COMPUTE POOL"), + ActionCreateObjectType::DataExchangeListing => write!(f, "DATA EXCHANGE LISTING"), + ActionCreateObjectType::Database => write!(f, "DATABASE"), + ActionCreateObjectType::ExternalVolume => write!(f, "EXTERNAL VOLUME"), + ActionCreateObjectType::FailoverGroup => write!(f, "FAILOVER GROUP"), + ActionCreateObjectType::Integration => write!(f, "INTEGRATION"), + ActionCreateObjectType::NetworkPolicy => write!(f, "NETWORK POLICY"), + ActionCreateObjectType::OrganiationListing => write!(f, "ORGANIZATION LISTING"), + ActionCreateObjectType::ReplicationGroup => write!(f, "REPLICATION GROUP"), + ActionCreateObjectType::Role => write!(f, "ROLE"), + ActionCreateObjectType::Share => write!(f, "SHARE"), + ActionCreateObjectType::User => write!(f, "USER"), + ActionCreateObjectType::Warehouse => write!(f, "WAREHOUSE"), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// See +/// under `globalPrivileges` in the `APPLY` privilege. +pub enum ActionApplyType { + AggregationPolicy, + AuthenticationPolicy, + JoinPolicy, + MaskingPolicy, + PackagesPolicy, + PasswordPolicy, + ProjectionPolicy, + RowAccessPolicy, + SessionPolicy, + Tag, +} + +impl fmt::Display for ActionApplyType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ActionApplyType::AggregationPolicy => write!(f, "AGGREGATION POLICY"), + ActionApplyType::AuthenticationPolicy => write!(f, "AUTHENTICATION POLICY"), + ActionApplyType::JoinPolicy => write!(f, "JOIN POLICY"), + ActionApplyType::MaskingPolicy => write!(f, "MASKING POLICY"), + ActionApplyType::PackagesPolicy => write!(f, "PACKAGES POLICY"), + ActionApplyType::PasswordPolicy => write!(f, "PASSWORD POLICY"), + ActionApplyType::ProjectionPolicy => write!(f, "PROJECTION POLICY"), + ActionApplyType::RowAccessPolicy => write!(f, "ROW ACCESS POLICY"), + ActionApplyType::SessionPolicy => write!(f, "SESSION POLICY"), + ActionApplyType::Tag => write!(f, "TAG"), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// See +/// under `globalPrivileges` in the `EXECUTE` privilege. +pub enum ActionExecuteObjectType { + Alert, + DataMetricFunction, + ManagedAlert, + ManagedTask, + Task, +} + +impl fmt::Display for ActionExecuteObjectType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ActionExecuteObjectType::Alert => write!(f, "ALERT"), + ActionExecuteObjectType::DataMetricFunction => write!(f, "DATA METRIC FUNCTION"), + ActionExecuteObjectType::ManagedAlert => write!(f, "MANAGED ALERT"), + ActionExecuteObjectType::ManagedTask => write!(f, "MANAGED TASK"), + ActionExecuteObjectType::Task => write!(f, "TASK"), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// See +/// under `globalPrivileges` in the `MANAGE` privilege. +pub enum ActionManageType { + AccountSupportCases, + EventSharing, + Grants, + ListingAutoFulfillment, + OrganizationSupportCases, + UserSupportCases, + Warehouses, +} + +impl fmt::Display for ActionManageType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ActionManageType::AccountSupportCases => write!(f, "ACCOUNT SUPPORT CASES"), + ActionManageType::EventSharing => write!(f, "EVENT SHARING"), + ActionManageType::Grants => write!(f, "GRANTS"), + ActionManageType::ListingAutoFulfillment => write!(f, "LISTING AUTO FULFILLMENT"), + ActionManageType::OrganizationSupportCases => write!(f, "ORGANIZATION SUPPORT CASES"), + ActionManageType::UserSupportCases => write!(f, "USER SUPPORT CASES"), + ActionManageType::Warehouses => write!(f, "WAREHOUSES"), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// See +/// under `globalPrivileges` in the `MODIFY` privilege. +pub enum ActionModifyType { + LogLevel, + TraceLevel, + SessionLogLevel, + SessionTraceLevel, +} + +impl fmt::Display for ActionModifyType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ActionModifyType::LogLevel => write!(f, "LOG LEVEL"), + ActionModifyType::TraceLevel => write!(f, "TRACE LEVEL"), + ActionModifyType::SessionLogLevel => write!(f, "SESSION LOG LEVEL"), + ActionModifyType::SessionTraceLevel => write!(f, "SESSION TRACE LEVEL"), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// See +/// under `globalPrivileges` in the `MONITOR` privilege. +pub enum ActionMonitorType { + Execution, + Security, + Usage, +} + +impl fmt::Display for ActionMonitorType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ActionMonitorType::Execution => write!(f, "EXECUTION"), + ActionMonitorType::Security => write!(f, "SECURITY"), + ActionMonitorType::Usage => write!(f, "USAGE"), + } + } +} + /// The principal that receives the privileges #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/keywords.rs b/src/keywords.rs index 0d1bab9e5..68b040c08 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -84,6 +84,7 @@ define_keywords!( AFTER, AGAINST, AGGREGATION, + ALERT, ALGORITHM, ALIAS, ALL, @@ -96,6 +97,7 @@ define_keywords!( ANY, APPLICATION, APPLY, + APPLYBUDGET, ARCHIVE, ARE, ARRAY, @@ -109,6 +111,8 @@ define_keywords!( AT, ATOMIC, ATTACH, + AUDIT, + AUTHENTICATION, AUTHORIZATION, AUTO, AUTOINCREMENT, @@ -127,6 +131,7 @@ define_keywords!( BIGINT, BIGNUMERIC, BINARY, + BIND, BINDING, BIT, BLOB, @@ -150,6 +155,7 @@ define_keywords!( CASCADE, CASCADED, CASE, + CASES, CAST, CATALOG, CATCH, @@ -305,12 +311,15 @@ define_keywords!( ESTIMATE, EVENT, EVERY, + EVOLVE, EXCEPT, EXCEPTION, + EXCHANGE, EXCLUDE, EXCLUSIVE, EXEC, EXECUTE, + EXECUTION, EXISTS, EXP, EXPANSION, @@ -322,6 +331,7 @@ define_keywords!( EXTERNAL, EXTRACT, FAIL, + FAILOVER, FALSE, FETCH, FIELDS, @@ -357,6 +367,7 @@ define_keywords!( FREEZE, FROM, FSCK, + FULFILLMENT, FULL, FULLTEXT, FUNCTION, @@ -394,6 +405,8 @@ define_keywords!( ILIKE, IMMEDIATE, IMMUTABLE, + IMPORT, + IMPORTED, IN, INCLUDE, INCLUDE_NULL_VALUES, @@ -421,6 +434,7 @@ define_keywords!( INT64, INT8, INTEGER, + INTEGRATION, INTERPOLATE, INTERSECT, INTERSECTION, @@ -460,6 +474,7 @@ define_keywords!( LINES, LIST, LISTEN, + LISTING, LN, LOAD, LOCAL, @@ -478,6 +493,8 @@ define_keywords!( LOW_PRIORITY, LS, MACRO, + MANAGE, + MANAGED, MANAGEDLOCATION, MAP, MASKING, @@ -499,6 +516,7 @@ define_keywords!( MERGE, METADATA, METHOD, + METRIC, MICROSECOND, MICROSECONDS, MILLENIUM, @@ -515,6 +533,7 @@ define_keywords!( MODIFIES, MODIFY, MODULE, + MONITOR, MONTH, MONTHS, MSCK, @@ -528,6 +547,7 @@ define_keywords!( NCHAR, NCLOB, NESTED, + NETWORK, NEW, NEXT, NFC, @@ -575,7 +595,9 @@ define_keywords!( ONLY, OPEN, OPENJSON, + OPERATE, OPERATOR, + OPTIMIZATION, OPTIMIZE, OPTIMIZER_COSTS, OPTION, @@ -584,6 +606,7 @@ define_keywords!( ORC, ORDER, ORDINALITY, + ORGANIZATION, OUT, OUTER, OUTPUTFORMAT, @@ -591,9 +614,13 @@ define_keywords!( OVERFLOW, OVERLAPS, OVERLAY, + OVERRIDE, OVERWRITE, OWNED, OWNER, + OWNERSHIP, + PACKAGE, + PACKAGES, PARALLEL, PARAMETER, PARQUET, @@ -618,6 +645,7 @@ define_keywords!( PLAN, PLANS, POLICY, + POOL, PORTION, POSITION, POSITION_REGEX, @@ -637,6 +665,7 @@ define_keywords!( PROGRAM, PROJECTION, PUBLIC, + PURCHASE, PURGE, QUALIFY, QUARTER, @@ -670,6 +699,7 @@ define_keywords!( RELATIVE, RELAY, RELEASE, + RELEASES, REMOTE, REMOVE, RENAME, @@ -678,12 +708,15 @@ define_keywords!( REPEATABLE, REPLACE, REPLICA, + REPLICATE, REPLICATION, RESET, + RESOLVE, RESPECT, RESTART, RESTRICT, RESTRICTED, + RESTRICTIONS, RESTRICTIVE, RESULT, RESULTSET, @@ -732,6 +765,7 @@ define_keywords!( SERDE, SERDEPROPERTIES, SERIALIZABLE, + SERVICE, SESSION, SESSION_USER, SET, @@ -739,6 +773,7 @@ define_keywords!( SETS, SETTINGS, SHARE, + SHARING, SHOW, SIMILAR, SKIP, @@ -782,6 +817,7 @@ define_keywords!( SUM, SUPER, SUPERUSER, + SUPPORT, SUSPEND, SWAP, SYMMETRIC, @@ -794,6 +830,7 @@ define_keywords!( TABLESAMPLE, TAG, TARGET, + TASK, TBLPROPERTIES, TEMP, TEMPORARY, @@ -819,6 +856,7 @@ define_keywords!( TO, TOP, TOTALS, + TRACE, TRAILING, TRANSACTION, TRANSIENT, @@ -885,11 +923,14 @@ define_keywords!( VERBOSE, VERSION, VERSIONING, + VERSIONS, VIEW, VIEWS, VIRTUAL, VOLATILE, + VOLUME, WAREHOUSE, + WAREHOUSES, WEEK, WEEKS, WHEN, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 51bbcfabd..355e520ab 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -48,9 +48,6 @@ pub enum ParserError { RecursionLimitExceeded, } -// avoid clippy type_complexity warnings -type ParsedAction = (Keyword, Option>); - // Use `Parser::expected` instead, if possible macro_rules! parser_err { ($MSG:expr, $loc:expr) => { @@ -3950,7 +3947,7 @@ impl<'a> Parser<'a> { ) } - pub fn parse_actions_list(&mut self) -> Result, ParserError> { + pub fn parse_actions_list(&mut self) -> Result, ParserError> { let mut values = vec![]; loop { values.push(self.parse_grant_permission()?); @@ -11980,37 +11977,8 @@ impl<'a> Parser<'a> { with_privileges_keyword: self.parse_keyword(Keyword::PRIVILEGES), } } else { - let (actions, err): (Vec<_>, Vec<_>) = self - .parse_actions_list()? - .into_iter() - .map(|(kw, columns)| match kw { - Keyword::DELETE => Ok(Action::Delete), - Keyword::INSERT => Ok(Action::Insert { columns }), - Keyword::REFERENCES => Ok(Action::References { columns }), - Keyword::SELECT => Ok(Action::Select { columns }), - Keyword::TRIGGER => Ok(Action::Trigger), - Keyword::TRUNCATE => Ok(Action::Truncate), - Keyword::UPDATE => Ok(Action::Update { columns }), - Keyword::USAGE => Ok(Action::Usage), - Keyword::CONNECT => Ok(Action::Connect), - Keyword::CREATE => Ok(Action::Create), - Keyword::EXECUTE => Ok(Action::Execute), - Keyword::TEMPORARY => Ok(Action::Temporary), - // This will cover all future added keywords to - // parse_grant_permission and unhandled in this - // match - _ => Err(kw), - }) - .partition(Result::is_ok); - - if !err.is_empty() { - let errors: Vec = err.into_iter().filter_map(|x| x.err()).collect(); - return Err(ParserError::ParserError(format!( - "INTERNAL ERROR: GRANT/REVOKE unexpected keyword(s) - {errors:?}" - ))); - } - let act = actions.into_iter().filter_map(|x| x.ok()).collect(); - Privileges::Actions(act) + let actions = self.parse_actions_list()?; + Privileges::Actions(actions) }; self.expect_keyword_is(Keyword::ON)?; @@ -12049,38 +12017,244 @@ impl<'a> Parser<'a> { Ok((privileges, objects)) } - pub fn parse_grant_permission(&mut self) -> Result { - if let Some(kw) = self.parse_one_of_keywords(&[ - Keyword::CONNECT, - Keyword::CREATE, - Keyword::DELETE, - Keyword::EXECUTE, - Keyword::INSERT, - Keyword::REFERENCES, - Keyword::SELECT, - Keyword::TEMPORARY, - Keyword::TRIGGER, - Keyword::TRUNCATE, - Keyword::UPDATE, - Keyword::USAGE, + pub fn parse_grant_permission(&mut self) -> Result { + fn parse_columns(parser: &mut Parser) -> Result>, ParserError> { + let columns = parser.parse_parenthesized_column_list(Optional, false)?; + if columns.is_empty() { + Ok(None) + } else { + Ok(Some(columns)) + } + } + + // Multi-word privileges + if self.parse_keywords(&[Keyword::IMPORTED, Keyword::PRIVILEGES]) { + Ok(Action::ImportedPrivileges) + } else if self.parse_keywords(&[Keyword::ADD, Keyword::SEARCH, Keyword::OPTIMIZATION]) { + Ok(Action::AddSearchOptimization) + } else if self.parse_keywords(&[Keyword::ATTACH, Keyword::LISTING]) { + Ok(Action::AttachListing) + } else if self.parse_keywords(&[Keyword::ATTACH, Keyword::POLICY]) { + Ok(Action::AttachPolicy) + } else if self.parse_keywords(&[Keyword::BIND, Keyword::SERVICE, Keyword::ENDPOINT]) { + Ok(Action::BindServiceEndpoint) + } else if self.parse_keywords(&[Keyword::EVOLVE, Keyword::SCHEMA]) { + Ok(Action::EvolveSchema) + } else if self.parse_keywords(&[Keyword::IMPORT, Keyword::SHARE]) { + Ok(Action::ImportShare) + } else if self.parse_keywords(&[Keyword::MANAGE, Keyword::VERSIONS]) { + Ok(Action::ManageVersions) + } else if self.parse_keywords(&[Keyword::MANAGE, Keyword::RELEASES]) { + Ok(Action::ManageReleases) + } else if self.parse_keywords(&[Keyword::OVERRIDE, Keyword::SHARE, Keyword::RESTRICTIONS]) { + Ok(Action::OverrideShareRestrictions) + } else if self.parse_keywords(&[ + Keyword::PURCHASE, + Keyword::DATA, + Keyword::EXCHANGE, + Keyword::LISTING, ]) { - let columns = match kw { - Keyword::INSERT | Keyword::REFERENCES | Keyword::SELECT | Keyword::UPDATE => { - let columns = self.parse_parenthesized_column_list(Optional, false)?; - if columns.is_empty() { - None - } else { - Some(columns) - } - } - _ => None, - }; - Ok((kw, columns)) + Ok(Action::PurchaseDataExchangeListing) + } else if self.parse_keywords(&[Keyword::RESOLVE, Keyword::ALL]) { + Ok(Action::ResolveAll) + } else if self.parse_keywords(&[Keyword::READ, Keyword::SESSION]) { + Ok(Action::ReadSession) + + // Single-word privileges + } else if self.parse_keyword(Keyword::APPLY) { + let apply_type = self.parse_action_apply_type()?; + Ok(Action::Apply { apply_type }) + } else if self.parse_keyword(Keyword::APPLYBUDGET) { + Ok(Action::ApplyBudget) + } else if self.parse_keyword(Keyword::AUDIT) { + Ok(Action::Audit) + } else if self.parse_keyword(Keyword::CONNECT) { + Ok(Action::Connect) + } else if self.parse_keyword(Keyword::CREATE) { + let obj_type = self.maybe_parse_action_create_object_type(); + Ok(Action::Create { obj_type }) + } else if self.parse_keyword(Keyword::DELETE) { + Ok(Action::Delete) + } else if self.parse_keyword(Keyword::EXECUTE) { + let obj_type = self.maybe_parse_action_execute_obj_type(); + Ok(Action::Execute { obj_type }) + } else if self.parse_keyword(Keyword::FAILOVER) { + Ok(Action::Failover) + } else if self.parse_keyword(Keyword::INSERT) { + Ok(Action::Insert { + columns: parse_columns(self)?, + }) + } else if self.parse_keyword(Keyword::MANAGE) { + let manage_type = self.parse_action_manage_type()?; + Ok(Action::Manage { manage_type }) + } else if self.parse_keyword(Keyword::MODIFY) { + let modify_type = self.parse_action_modify_type()?; + Ok(Action::Modify { modify_type }) + } else if self.parse_keyword(Keyword::MONITOR) { + let monitor_type = self.parse_action_monitor_type()?; + Ok(Action::Monitor { monitor_type }) + } else if self.parse_keyword(Keyword::OPERATE) { + Ok(Action::Operate) + } else if self.parse_keyword(Keyword::REFERENCES) { + Ok(Action::References { + columns: parse_columns(self)?, + }) + } else if self.parse_keyword(Keyword::READ) { + Ok(Action::Read) + } else if self.parse_keyword(Keyword::REPLICATE) { + Ok(Action::Replicate) + } else if self.parse_keyword(Keyword::SELECT) { + Ok(Action::Select { + columns: parse_columns(self)?, + }) + } else if self.parse_keyword(Keyword::TEMPORARY) { + Ok(Action::Temporary) + } else if self.parse_keyword(Keyword::TRIGGER) { + Ok(Action::Trigger) + } else if self.parse_keyword(Keyword::TRUNCATE) { + Ok(Action::Truncate) + } else if self.parse_keyword(Keyword::UPDATE) { + Ok(Action::Update { + columns: parse_columns(self)?, + }) + } else if self.parse_keyword(Keyword::USAGE) { + Ok(Action::Usage) + } else if self.parse_keyword(Keyword::OWNERSHIP) { + Ok(Action::Ownership) } else { self.expected("a privilege keyword", self.peek_token())? } } + fn maybe_parse_action_create_object_type(&mut self) -> Option { + // Multi-word object types + if self.parse_keywords(&[Keyword::APPLICATION, Keyword::PACKAGE]) { + Some(ActionCreateObjectType::ApplicationPackage) + } else if self.parse_keywords(&[Keyword::COMPUTE, Keyword::POOL]) { + Some(ActionCreateObjectType::ComputePool) + } else if self.parse_keywords(&[Keyword::DATA, Keyword::EXCHANGE, Keyword::LISTING]) { + Some(ActionCreateObjectType::DataExchangeListing) + } else if self.parse_keywords(&[Keyword::EXTERNAL, Keyword::VOLUME]) { + Some(ActionCreateObjectType::ExternalVolume) + } else if self.parse_keywords(&[Keyword::FAILOVER, Keyword::GROUP]) { + Some(ActionCreateObjectType::FailoverGroup) + } else if self.parse_keywords(&[Keyword::NETWORK, Keyword::POLICY]) { + Some(ActionCreateObjectType::NetworkPolicy) + } else if self.parse_keywords(&[Keyword::ORGANIZATION, Keyword::LISTING]) { + Some(ActionCreateObjectType::OrganiationListing) + } else if self.parse_keywords(&[Keyword::REPLICATION, Keyword::GROUP]) { + Some(ActionCreateObjectType::ReplicationGroup) + } + // Single-word object types + else if self.parse_keyword(Keyword::ACCOUNT) { + Some(ActionCreateObjectType::Account) + } else if self.parse_keyword(Keyword::APPLICATION) { + Some(ActionCreateObjectType::Application) + } else if self.parse_keyword(Keyword::DATABASE) { + Some(ActionCreateObjectType::Database) + } else if self.parse_keyword(Keyword::INTEGRATION) { + Some(ActionCreateObjectType::Integration) + } else if self.parse_keyword(Keyword::ROLE) { + Some(ActionCreateObjectType::Role) + } else if self.parse_keyword(Keyword::SHARE) { + Some(ActionCreateObjectType::Share) + } else if self.parse_keyword(Keyword::USER) { + Some(ActionCreateObjectType::User) + } else if self.parse_keyword(Keyword::WAREHOUSE) { + Some(ActionCreateObjectType::Warehouse) + } else { + None + } + } + + fn parse_action_apply_type(&mut self) -> Result { + if self.parse_keywords(&[Keyword::AGGREGATION, Keyword::POLICY]) { + Ok(ActionApplyType::AggregationPolicy) + } else if self.parse_keywords(&[Keyword::AUTHENTICATION, Keyword::POLICY]) { + Ok(ActionApplyType::AuthenticationPolicy) + } else if self.parse_keywords(&[Keyword::JOIN, Keyword::POLICY]) { + Ok(ActionApplyType::JoinPolicy) + } else if self.parse_keywords(&[Keyword::MASKING, Keyword::POLICY]) { + Ok(ActionApplyType::MaskingPolicy) + } else if self.parse_keywords(&[Keyword::PACKAGES, Keyword::POLICY]) { + Ok(ActionApplyType::PackagesPolicy) + } else if self.parse_keywords(&[Keyword::PASSWORD, Keyword::POLICY]) { + Ok(ActionApplyType::PasswordPolicy) + } else if self.parse_keywords(&[Keyword::PROJECTION, Keyword::POLICY]) { + Ok(ActionApplyType::ProjectionPolicy) + } else if self.parse_keywords(&[Keyword::ROW, Keyword::ACCESS, Keyword::POLICY]) { + Ok(ActionApplyType::RowAccessPolicy) + } else if self.parse_keywords(&[Keyword::SESSION, Keyword::POLICY]) { + Ok(ActionApplyType::SessionPolicy) + } else if self.parse_keyword(Keyword::TAG) { + Ok(ActionApplyType::Tag) + } else { + self.expected("GRANT APPLY type", self.peek_token()) + } + } + + fn maybe_parse_action_execute_obj_type(&mut self) -> Option { + if self.parse_keywords(&[Keyword::DATA, Keyword::METRIC, Keyword::FUNCTION]) { + Some(ActionExecuteObjectType::DataMetricFunction) + } else if self.parse_keywords(&[Keyword::MANAGED, Keyword::ALERT]) { + Some(ActionExecuteObjectType::ManagedAlert) + } else if self.parse_keywords(&[Keyword::MANAGED, Keyword::TASK]) { + Some(ActionExecuteObjectType::ManagedTask) + } else if self.parse_keyword(Keyword::ALERT) { + Some(ActionExecuteObjectType::Alert) + } else if self.parse_keyword(Keyword::TASK) { + Some(ActionExecuteObjectType::Task) + } else { + None + } + } + + fn parse_action_manage_type(&mut self) -> Result { + if self.parse_keywords(&[Keyword::ACCOUNT, Keyword::SUPPORT, Keyword::CASES]) { + Ok(ActionManageType::AccountSupportCases) + } else if self.parse_keywords(&[Keyword::EVENT, Keyword::SHARING]) { + Ok(ActionManageType::EventSharing) + } else if self.parse_keywords(&[Keyword::LISTING, Keyword::AUTO, Keyword::FULFILLMENT]) { + Ok(ActionManageType::ListingAutoFulfillment) + } else if self.parse_keywords(&[Keyword::ORGANIZATION, Keyword::SUPPORT, Keyword::CASES]) { + Ok(ActionManageType::OrganizationSupportCases) + } else if self.parse_keywords(&[Keyword::USER, Keyword::SUPPORT, Keyword::CASES]) { + Ok(ActionManageType::UserSupportCases) + } else if self.parse_keyword(Keyword::GRANTS) { + Ok(ActionManageType::Grants) + } else if self.parse_keyword(Keyword::WAREHOUSES) { + Ok(ActionManageType::Warehouses) + } else { + self.expected("GRANT MANAGE type", self.peek_token()) + } + } + + fn parse_action_modify_type(&mut self) -> Result { + if self.parse_keywords(&[Keyword::LOG, Keyword::LEVEL]) { + Ok(ActionModifyType::LogLevel) + } else if self.parse_keywords(&[Keyword::TRACE, Keyword::LEVEL]) { + Ok(ActionModifyType::TraceLevel) + } else if self.parse_keywords(&[Keyword::SESSION, Keyword::LOG, Keyword::LEVEL]) { + Ok(ActionModifyType::SessionLogLevel) + } else if self.parse_keywords(&[Keyword::SESSION, Keyword::TRACE, Keyword::LEVEL]) { + Ok(ActionModifyType::SessionTraceLevel) + } else { + self.expected("GRANT MODIFY type", self.peek_token()) + } + } + + fn parse_action_monitor_type(&mut self) -> Result { + if self.parse_keyword(Keyword::EXECUTION) { + Ok(ActionMonitorType::Execution) + } else if self.parse_keyword(Keyword::SECURITY) { + Ok(ActionMonitorType::Security) + } else if self.parse_keyword(Keyword::USAGE) { + Ok(ActionMonitorType::Usage) + } else { + self.expected("GRANT MONITOR type", self.peek_token()) + } + } + pub fn parse_grantee_name(&mut self) -> Result { let mut name = self.parse_object_name(false)?; if self.dialect.supports_user_host_grantee() diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 7271d6375..ebd6bef03 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8500,8 +8500,8 @@ fn parse_grant() { Action::References { columns: None }, Action::Trigger, Action::Connect, - Action::Create, - Action::Execute, + Action::Create { obj_type: None }, + Action::Execute { obj_type: None }, Action::Temporary, ], actions @@ -8616,6 +8616,7 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO SHARE share1"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO a:b"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO GROUP group1"); + verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST"); } #[test] diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 0c4bdf149..324c45e86 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3058,3 +3058,109 @@ fn test_timetravel_at_before() { snowflake() .verified_only_select("SELECT * FROM tbl BEFORE(TIMESTAMP => '2024-12-15 00:00:00')"); } + +#[test] +fn test_grant_account_privileges() { + let privileges = vec![ + "ALL", + "ALL PRIVILEGES", + "ATTACH POLICY", + "AUDIT", + "BIND SERVICE ENDPOINT", + "IMPORT SHARE", + "OVERRIDE SHARE RESTRICTIONS", + "PURCHASE DATA EXCHANGE LISTING", + "RESOLVE ALL", + "READ SESSION", + ]; + let with_grant_options = vec!["", " WITH GRANT OPTION"]; + + for p in &privileges { + for wgo in &with_grant_options { + let sql = format!("GRANT {p} ON ACCOUNT TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } + + let create_object_types = vec![ + "ACCOUNT", + "APPLICATION", + "APPLICATION PACKAGE", + "COMPUTE POOL", + "DATA EXCHANGE LISTING", + "DATABASE", + "EXTERNAL VOLUME", + "FAILOVER GROUP", + "INTEGRATION", + "NETWORK POLICY", + "ORGANIZATION LISTING", + "REPLICATION GROUP", + "ROLE", + "SHARE", + "USER", + "WAREHOUSE", + ]; + for t in &create_object_types { + for wgo in &with_grant_options { + let sql = format!("GRANT CREATE {t} ON ACCOUNT TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } + + let apply_types = vec![ + "AGGREGATION POLICY", + "AUTHENTICATION POLICY", + "JOIN POLICY", + "MASKING POLICY", + "PACKAGES POLICY", + "PASSWORD POLICY", + "PROJECTION POLICY", + "ROW ACCESS POLICY", + "SESSION POLICY", + "TAG", + ]; + for t in &apply_types { + for wgo in &with_grant_options { + let sql = format!("GRANT APPLY {t} ON ACCOUNT TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } + + let execute_types = vec![ + "ALERT", + "DATA METRIC FUNCTION", + "MANAGED ALERT", + "MANAGED TASK", + "TASK", + ]; + for t in &execute_types { + for wgo in &with_grant_options { + let sql = format!("GRANT EXECUTE {t} ON ACCOUNT TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } + + let manage_types = vec![ + "ACCOUNT SUPPORT CASES", + "EVENT SHARING", + "GRANTS", + "LISTING AUTO FULFILLMENT", + "ORGANIZATION SUPPORT CASES", + "USER SUPPORT CASES", + "WAREHOUSES", + ]; + for t in &manage_types { + for wgo in &with_grant_options { + let sql = format!("GRANT MANAGE {t} ON ACCOUNT TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } + + let monitor_types = vec!["EXECUTION", "SECURITY", "USAGE"]; + for t in &monitor_types { + for wgo in &with_grant_options { + let sql = format!("GRANT MONITOR {t} ON ACCOUNT TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } +} From c7c0de6551e09d897d42d86a7324b4b2051feb2f Mon Sep 17 00:00:00 2001 From: Denys Tsomenko Date: Mon, 20 Jan 2025 22:39:44 +0200 Subject: [PATCH 011/286] Add support for Create Iceberg Table statement for Snowflake parser (#1664) --- src/ast/dml.rs | 48 +++++++++++++++++- src/ast/helpers/stmt_create_table.rs | 65 +++++++++++++++++++++++- src/ast/mod.rs | 23 +++++++++ src/ast/spans.rs | 6 +++ src/dialect/snowflake.rs | 51 ++++++++++++++++++- src/keywords.rs | 7 +++ tests/sqlparser_duckdb.rs | 8 ++- tests/sqlparser_mssql.rs | 12 +++++ tests/sqlparser_postgres.rs | 6 +++ tests/sqlparser_snowflake.rs | 75 ++++++++++++++++++++++++++++ 10 files changed, 296 insertions(+), 5 deletions(-) diff --git a/src/ast/dml.rs b/src/ast/dml.rs index de555c109..8cfc67414 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -36,7 +36,8 @@ use super::{ CommentDef, Expr, FileFormat, FromTable, HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, InsertAliases, MysqlInsertPriority, ObjectName, OnCommit, OnInsert, OneOrManyWithParens, OrderByExpr, Query, RowAccessPolicy, SelectItem, Setting, SqlOption, - SqliteOnConflict, TableEngine, TableObject, TableWithJoins, Tag, WrappedCollection, + SqliteOnConflict, StorageSerializationPolicy, TableEngine, TableObject, TableWithJoins, Tag, + WrappedCollection, }; /// CREATE INDEX statement. @@ -117,6 +118,7 @@ pub struct CreateTable { pub if_not_exists: bool, pub transient: bool, pub volatile: bool, + pub iceberg: bool, /// Table name #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] pub name: ObjectName, @@ -192,6 +194,21 @@ pub struct CreateTable { /// Snowflake "WITH TAG" clause /// pub with_tags: Option>, + /// Snowflake "EXTERNAL_VOLUME" clause for Iceberg tables + /// + pub external_volume: Option, + /// Snowflake "BASE_LOCATION" clause for Iceberg tables + /// + pub base_location: Option, + /// Snowflake "CATALOG" clause for Iceberg tables + /// + pub catalog: Option, + /// Snowflake "CATALOG_SYNC" clause for Iceberg tables + /// + pub catalog_sync: Option, + /// Snowflake "STORAGE_SERIALIZATION_POLICY" clause for Iceberg tables + /// + pub storage_serialization_policy: Option, } impl Display for CreateTable { @@ -205,7 +222,7 @@ impl Display for CreateTable { // `CREATE TABLE t (a INT) AS SELECT a from t2` write!( f, - "CREATE {or_replace}{external}{global}{temporary}{transient}{volatile}TABLE {if_not_exists}{name}", + "CREATE {or_replace}{external}{global}{temporary}{transient}{volatile}{iceberg}TABLE {if_not_exists}{name}", or_replace = if self.or_replace { "OR REPLACE " } else { "" }, external = if self.external { "EXTERNAL " } else { "" }, global = self.global @@ -221,6 +238,8 @@ impl Display for CreateTable { temporary = if self.temporary { "TEMPORARY " } else { "" }, transient = if self.transient { "TRANSIENT " } else { "" }, volatile = if self.volatile { "VOLATILE " } else { "" }, + // Only for Snowflake + iceberg = if self.iceberg { "ICEBERG " } else { "" }, name = self.name, )?; if let Some(on_cluster) = &self.on_cluster { @@ -382,6 +401,31 @@ impl Display for CreateTable { )?; } + if let Some(external_volume) = self.external_volume.as_ref() { + write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; + } + + if let Some(catalog) = self.catalog.as_ref() { + write!(f, " CATALOG = '{catalog}'")?; + } + + if self.iceberg { + if let Some(base_location) = self.base_location.as_ref() { + write!(f, " BASE_LOCATION = '{base_location}'")?; + } + } + + if let Some(catalog_sync) = self.catalog_sync.as_ref() { + write!(f, " CATALOG_SYNC = '{catalog_sync}'")?; + } + + if let Some(storage_serialization_policy) = self.storage_serialization_policy.as_ref() { + write!( + f, + " STORAGE_SERIALIZATION_POLICY = {storage_serialization_policy}" + )?; + } + if self.copy_grants { write!(f, " COPY GRANTS")?; } diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index a3be57986..e7090cb86 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -28,7 +28,7 @@ use super::super::dml::CreateTable; use crate::ast::{ ClusteredBy, ColumnDef, CommentDef, Expr, FileFormat, HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, OneOrManyWithParens, Query, RowAccessPolicy, SqlOption, Statement, - TableConstraint, TableEngine, Tag, WrappedCollection, + StorageSerializationPolicy, TableConstraint, TableEngine, Tag, WrappedCollection, }; use crate::parser::ParserError; @@ -71,6 +71,7 @@ pub struct CreateTableBuilder { pub if_not_exists: bool, pub transient: bool, pub volatile: bool, + pub iceberg: bool, pub name: ObjectName, pub columns: Vec, pub constraints: Vec, @@ -107,6 +108,11 @@ pub struct CreateTableBuilder { pub with_aggregation_policy: Option, pub with_row_access_policy: Option, pub with_tags: Option>, + pub base_location: Option, + pub external_volume: Option, + pub catalog: Option, + pub catalog_sync: Option, + pub storage_serialization_policy: Option, } impl CreateTableBuilder { @@ -119,6 +125,7 @@ impl CreateTableBuilder { if_not_exists: false, transient: false, volatile: false, + iceberg: false, name, columns: vec![], constraints: vec![], @@ -155,6 +162,11 @@ impl CreateTableBuilder { with_aggregation_policy: None, with_row_access_policy: None, with_tags: None, + base_location: None, + external_volume: None, + catalog: None, + catalog_sync: None, + storage_serialization_policy: None, } } pub fn or_replace(mut self, or_replace: bool) -> Self { @@ -192,6 +204,11 @@ impl CreateTableBuilder { self } + pub fn iceberg(mut self, iceberg: bool) -> Self { + self.iceberg = iceberg; + self + } + pub fn columns(mut self, columns: Vec) -> Self { self.columns = columns; self @@ -371,6 +388,34 @@ impl CreateTableBuilder { self } + pub fn base_location(mut self, base_location: Option) -> Self { + self.base_location = base_location; + self + } + + pub fn external_volume(mut self, external_volume: Option) -> Self { + self.external_volume = external_volume; + self + } + + pub fn catalog(mut self, catalog: Option) -> Self { + self.catalog = catalog; + self + } + + pub fn catalog_sync(mut self, catalog_sync: Option) -> Self { + self.catalog_sync = catalog_sync; + self + } + + pub fn storage_serialization_policy( + mut self, + storage_serialization_policy: Option, + ) -> Self { + self.storage_serialization_policy = storage_serialization_policy; + self + } + pub fn build(self) -> Statement { Statement::CreateTable(CreateTable { or_replace: self.or_replace, @@ -380,6 +425,7 @@ impl CreateTableBuilder { if_not_exists: self.if_not_exists, transient: self.transient, volatile: self.volatile, + iceberg: self.iceberg, name: self.name, columns: self.columns, constraints: self.constraints, @@ -416,6 +462,11 @@ impl CreateTableBuilder { with_aggregation_policy: self.with_aggregation_policy, with_row_access_policy: self.with_row_access_policy, with_tags: self.with_tags, + base_location: self.base_location, + external_volume: self.external_volume, + catalog: self.catalog, + catalog_sync: self.catalog_sync, + storage_serialization_policy: self.storage_serialization_policy, }) } } @@ -435,6 +486,7 @@ impl TryFrom for CreateTableBuilder { if_not_exists, transient, volatile, + iceberg, name, columns, constraints, @@ -471,6 +523,11 @@ impl TryFrom for CreateTableBuilder { with_aggregation_policy, with_row_access_policy, with_tags, + base_location, + external_volume, + catalog, + catalog_sync, + storage_serialization_policy, }) => Ok(Self { or_replace, temporary, @@ -505,6 +562,7 @@ impl TryFrom for CreateTableBuilder { clustered_by, options, strict, + iceberg, copy_grants, enable_schema_evolution, change_tracking, @@ -515,6 +573,11 @@ impl TryFrom for CreateTableBuilder { with_row_access_policy, with_tags, volatile, + base_location, + external_volume, + catalog, + catalog_sync, + storage_serialization_policy, }), _ => Err(ParserError::ParserError(format!( "Expected create table statement, but received: {stmt}" diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 5c06d7196..2fc89e29b 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -8396,6 +8396,29 @@ impl fmt::Display for SessionParamValue { } } +/// Snowflake StorageSerializationPolicy for Iceberg Tables +/// ```sql +/// [ STORAGE_SERIALIZATION_POLICY = { COMPATIBLE | OPTIMIZED } ] +/// ``` +/// +/// +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum StorageSerializationPolicy { + Compatible, + Optimized, +} + +impl Display for StorageSerializationPolicy { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + StorageSerializationPolicy::Compatible => write!(f, "COMPATIBLE"), + StorageSerializationPolicy::Optimized => write!(f, "OPTIMIZED"), + } + } +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 1ddd47d7f..acd3987da 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -532,6 +532,7 @@ impl Spanned for CreateTable { if_not_exists: _, // bool transient: _, // bool volatile: _, // bool + iceberg: _, // bool, Snowflake specific name, columns, constraints, @@ -568,6 +569,11 @@ impl Spanned for CreateTable { with_aggregation_policy: _, // todo, Snowflake specific with_row_access_policy: _, // todo, Snowflake specific with_tags: _, // todo, Snowflake specific + external_volume: _, // todo, Snowflake specific + base_location: _, // todo, Snowflake specific + catalog: _, // todo, Snowflake specific + catalog_sync: _, // todo, Snowflake specific + storage_serialization_policy: _, // todo, Snowflake specific } = self; union_spans( diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 78237acd0..88e54016d 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -37,6 +37,7 @@ use alloc::string::String; use alloc::vec::Vec; #[cfg(not(feature = "std"))] use alloc::{format, vec}; +use sqlparser::ast::StorageSerializationPolicy; use super::keywords::RESERVED_FOR_IDENTIFIER; @@ -130,16 +131,19 @@ impl Dialect for SnowflakeDialect { let mut temporary = false; let mut volatile = false; let mut transient = false; + let mut iceberg = false; match parser.parse_one_of_keywords(&[ Keyword::TEMP, Keyword::TEMPORARY, Keyword::VOLATILE, Keyword::TRANSIENT, + Keyword::ICEBERG, ]) { Some(Keyword::TEMP | Keyword::TEMPORARY) => temporary = true, Some(Keyword::VOLATILE) => volatile = true, Some(Keyword::TRANSIENT) => transient = true, + Some(Keyword::ICEBERG) => iceberg = true, _ => {} } @@ -148,7 +152,7 @@ impl Dialect for SnowflakeDialect { return Some(parse_create_stage(or_replace, temporary, parser)); } else if parser.parse_keyword(Keyword::TABLE) { return Some(parse_create_table( - or_replace, global, temporary, volatile, transient, parser, + or_replace, global, temporary, volatile, transient, iceberg, parser, )); } else { // need to go back with the cursor @@ -325,12 +329,14 @@ fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result +/// pub fn parse_create_table( or_replace: bool, global: Option, temporary: bool, volatile: bool, transient: bool, + iceberg: bool, parser: &mut Parser, ) -> Result { let if_not_exists = parser.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); @@ -342,6 +348,7 @@ pub fn parse_create_table( .temporary(temporary) .transient(transient) .volatile(volatile) + .iceberg(iceberg) .global(global) .hive_formats(Some(Default::default())); @@ -468,6 +475,28 @@ pub fn parse_create_table( let on_commit = Some(parser.parse_create_table_on_commit()?); builder = builder.on_commit(on_commit); } + Keyword::EXTERNAL_VOLUME => { + parser.expect_token(&Token::Eq)?; + builder.external_volume = Some(parser.parse_literal_string()?); + } + Keyword::CATALOG => { + parser.expect_token(&Token::Eq)?; + builder.catalog = Some(parser.parse_literal_string()?); + } + Keyword::BASE_LOCATION => { + parser.expect_token(&Token::Eq)?; + builder.base_location = Some(parser.parse_literal_string()?); + } + Keyword::CATALOG_SYNC => { + parser.expect_token(&Token::Eq)?; + builder.catalog_sync = Some(parser.parse_literal_string()?); + } + Keyword::STORAGE_SERIALIZATION_POLICY => { + parser.expect_token(&Token::Eq)?; + + builder.storage_serialization_policy = + Some(parse_storage_serialization_policy(parser)?); + } _ => { return parser.expected("end of statement", next_token); } @@ -502,9 +531,29 @@ pub fn parse_create_table( } } + if iceberg && builder.base_location.is_none() { + return Err(ParserError::ParserError( + "BASE_LOCATION is required for ICEBERG tables".to_string(), + )); + } + Ok(builder.build()) } +pub fn parse_storage_serialization_policy( + parser: &mut Parser, +) -> Result { + let next_token = parser.next_token(); + match &next_token.token { + Token::Word(w) => match w.keyword { + Keyword::COMPATIBLE => Ok(StorageSerializationPolicy::Compatible), + Keyword::OPTIMIZED => Ok(StorageSerializationPolicy::Optimized), + _ => parser.expected("storage_serialization_policy", next_token), + }, + _ => parser.expected("storage_serialization_policy", next_token), + } +} + pub fn parse_create_stage( or_replace: bool, temporary: bool, diff --git a/src/keywords.rs b/src/keywords.rs index 68b040c08..02ce04988 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -121,6 +121,7 @@ define_keywords!( AVRO, BACKWARD, BASE64, + BASE_LOCATION, BEFORE, BEGIN, BEGIN_FRAME, @@ -158,6 +159,7 @@ define_keywords!( CASES, CAST, CATALOG, + CATALOG_SYNC, CATCH, CEIL, CEILING, @@ -191,6 +193,7 @@ define_keywords!( COMMENT, COMMIT, COMMITTED, + COMPATIBLE, COMPRESSION, COMPUTE, CONCURRENTLY, @@ -329,6 +332,7 @@ define_keywords!( EXTENDED, EXTENSION, EXTERNAL, + EXTERNAL_VOLUME, EXTRACT, FAIL, FAILOVER, @@ -397,6 +401,7 @@ define_keywords!( HOSTS, HOUR, HOURS, + ICEBERG, ID, IDENTITY, IDENTITY_INSERT, @@ -599,6 +604,7 @@ define_keywords!( OPERATOR, OPTIMIZATION, OPTIMIZE, + OPTIMIZED, OPTIMIZER_COSTS, OPTION, OPTIONS, @@ -806,6 +812,7 @@ define_keywords!( STDOUT, STEP, STORAGE_INTEGRATION, + STORAGE_SERIALIZATION_POLICY, STORED, STRICT, STRING, diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index db4ffb6f6..ca7f926a9 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -660,6 +660,7 @@ fn test_duckdb_union_datatype() { if_not_exists: Default::default(), transient: Default::default(), volatile: Default::default(), + iceberg: Default::default(), name: ObjectName(vec!["tbl1".into()]), columns: vec![ ColumnDef { @@ -737,7 +738,12 @@ fn test_duckdb_union_datatype() { default_ddl_collation: Default::default(), with_aggregation_policy: Default::default(), with_row_access_policy: Default::default(), - with_tags: Default::default() + with_tags: Default::default(), + base_location: Default::default(), + external_volume: Default::default(), + catalog: Default::default(), + catalog_sync: Default::default(), + storage_serialization_policy: Default::default(), }), stmt ); diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index a0ac8a4d8..da2b6160e 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1539,6 +1539,7 @@ fn parse_create_table_with_valid_options() { clustered_by: None, options: None, strict: false, + iceberg: false, copy_grants: false, enable_schema_evolution: None, change_tracking: None, @@ -1548,6 +1549,11 @@ fn parse_create_table_with_valid_options() { with_aggregation_policy: None, with_row_access_policy: None, with_tags: None, + base_location: None, + external_volume: None, + catalog: None, + catalog_sync: None, + storage_serialization_policy: None, }) ); } @@ -1641,6 +1647,7 @@ fn parse_create_table_with_identity_column() { if_not_exists: false, transient: false, volatile: false, + iceberg: false, name: ObjectName(vec![Ident { value: "mytable".to_string(), quote_style: None, @@ -1695,6 +1702,11 @@ fn parse_create_table_with_identity_column() { with_aggregation_policy: None, with_row_access_policy: None, with_tags: None, + base_location: None, + external_volume: None, + catalog: None, + catalog_sync: None, + storage_serialization_policy: None, }), ); } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 864fb5eb3..0fca4cec1 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5043,6 +5043,7 @@ fn parse_trigger_related_functions() { if_not_exists: false, transient: false, volatile: false, + iceberg: false, name: ObjectName(vec![Ident::new("emp")]), columns: vec![ ColumnDef { @@ -5109,6 +5110,11 @@ fn parse_trigger_related_functions() { with_aggregation_policy: None, with_row_access_policy: None, with_tags: None, + base_location: None, + external_volume: None, + catalog: None, + catalog_sync: None, + storage_serialization_policy: None, } ); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 324c45e86..3320400e9 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -849,6 +849,81 @@ fn test_snowflake_create_table_with_several_column_options() { } } +#[test] +fn test_snowflake_create_iceberg_table_all_options() { + match snowflake().verified_stmt("CREATE ICEBERG TABLE my_table (a INT, b INT) \ + CLUSTER BY (a, b) EXTERNAL_VOLUME = 'volume' CATALOG = 'SNOWFLAKE' BASE_LOCATION = 'relative/path' CATALOG_SYNC = 'OPEN_CATALOG' \ + STORAGE_SERIALIZATION_POLICY = COMPATIBLE COPY GRANTS CHANGE_TRACKING=TRUE DATA_RETENTION_TIME_IN_DAYS=5 MAX_DATA_EXTENSION_TIME_IN_DAYS=10 \ + WITH AGGREGATION POLICY policy_name WITH ROW ACCESS POLICY policy_name ON (a) WITH TAG (A='TAG A', B='TAG B')") { + Statement::CreateTable(CreateTable { + name, cluster_by, base_location, + external_volume, catalog, catalog_sync, + storage_serialization_policy, change_tracking, + copy_grants, data_retention_time_in_days, + max_data_extension_time_in_days, with_aggregation_policy, + with_row_access_policy, with_tags, .. + }) => { + assert_eq!("my_table", name.to_string()); + assert_eq!( + Some(WrappedCollection::Parentheses(vec![ + Ident::new("a"), + Ident::new("b"), + ])), + cluster_by + ); + assert_eq!("relative/path", base_location.unwrap()); + assert_eq!("volume", external_volume.unwrap()); + assert_eq!("SNOWFLAKE", catalog.unwrap()); + assert_eq!("OPEN_CATALOG", catalog_sync.unwrap()); + assert_eq!(StorageSerializationPolicy::Compatible, storage_serialization_policy.unwrap()); + assert!(change_tracking.unwrap()); + assert!(copy_grants); + assert_eq!(Some(5), data_retention_time_in_days); + assert_eq!(Some(10), max_data_extension_time_in_days); + assert_eq!( + Some("WITH ROW ACCESS POLICY policy_name ON (a)".to_string()), + with_row_access_policy.map(|policy| policy.to_string()) + ); + assert_eq!( + Some("policy_name".to_string()), + with_aggregation_policy.map(|name| name.to_string()) + ); + assert_eq!(Some(vec![ + Tag::new("A".into(), "TAG A".into()), + Tag::new("B".into(), "TAG B".into()), + ]), with_tags); + + } + _ => unreachable!(), + } +} + +#[test] +fn test_snowflake_create_iceberg_table() { + match snowflake() + .verified_stmt("CREATE ICEBERG TABLE my_table (a INT) BASE_LOCATION = 'relative_path'") + { + Statement::CreateTable(CreateTable { + name, + base_location, + .. + }) => { + assert_eq!("my_table", name.to_string()); + assert_eq!("relative_path", base_location.unwrap()); + } + _ => unreachable!(), + } +} + +#[test] +fn test_snowflake_create_iceberg_table_without_location() { + let res = snowflake().parse_sql_statements("CREATE ICEBERG TABLE my_table (a INT)"); + assert_eq!( + ParserError::ParserError("BASE_LOCATION is required for ICEBERG tables".to_string()), + res.unwrap_err() + ); +} + #[test] fn parse_sf_create_or_replace_view_with_comment_missing_equal() { assert!(snowflake_and_generic() From 4f7154288e0d6f3b324aee4a262babe5872191fc Mon Sep 17 00:00:00 2001 From: Hans Ott Date: Thu, 23 Jan 2025 17:16:53 +0100 Subject: [PATCH 012/286] National strings: check if dialect supports backslash escape (#1672) --- src/test_utils.rs | 14 +++++++++++++- src/tokenizer.rs | 32 +++++++++++++++++++++++++++++++- 2 files changed, 44 insertions(+), 2 deletions(-) diff --git a/src/test_utils.rs b/src/test_utils.rs index 914be7d9f..51e4fd748 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -33,7 +33,7 @@ use core::fmt::Debug; use crate::dialect::*; use crate::parser::{Parser, ParserError}; -use crate::tokenizer::Tokenizer; +use crate::tokenizer::{Token, Tokenizer}; use crate::{ast::*, parser::ParserOptions}; #[cfg(test)] @@ -237,6 +237,18 @@ impl TestedDialects { pub fn verified_expr(&self, sql: &str) -> Expr { self.expr_parses_to(sql, sql) } + + /// Check that the tokenizer returns the expected tokens for the given SQL. + pub fn tokenizes_to(&self, sql: &str, expected: Vec) { + self.dialects.iter().for_each(|dialect| { + let mut tokenizer = Tokenizer::new(&**dialect, sql); + if let Some(options) = &self.options { + tokenizer = tokenizer.with_unescape(options.unescape); + } + let tokens = tokenizer.tokenize().unwrap(); + assert_eq!(expected, tokens); + }); + } } /// Returns all available dialects. diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 39ca84c9f..08e233b66 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -971,7 +971,10 @@ impl<'a> Tokenizer<'a> { match chars.peek() { Some('\'') => { // N'...' - a - let s = self.tokenize_single_quoted_string(chars, '\'', true)?; + let backslash_escape = + self.dialect.supports_string_literal_backslash_escape(); + let s = + self.tokenize_single_quoted_string(chars, '\'', backslash_escape)?; Ok(Some(Token::NationalStringLiteral(s))) } _ => { @@ -2155,6 +2158,7 @@ mod tests { use crate::dialect::{ BigQueryDialect, ClickHouseDialect, HiveDialect, MsSqlDialect, MySqlDialect, SQLiteDialect, }; + use crate::test_utils::all_dialects_where; use core::fmt::Debug; #[test] @@ -3543,4 +3547,30 @@ mod tests { ]; compare(expected, tokens); } + + #[test] + fn test_national_strings_backslash_escape_not_supported() { + all_dialects_where(|dialect| !dialect.supports_string_literal_backslash_escape()) + .tokenizes_to( + "select n'''''\\'", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::NationalStringLiteral("''\\".to_string()), + ], + ); + } + + #[test] + fn test_national_strings_backslash_escape_supported() { + all_dialects_where(|dialect| dialect.supports_string_literal_backslash_escape()) + .tokenizes_to( + "select n'''''\\''", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::NationalStringLiteral("'''".to_string()), + ], + ); + } } From ef072be9e1b1ecbf8032bd2040131a9d5b00de5d Mon Sep 17 00:00:00 2001 From: Hans Ott Date: Fri, 24 Jan 2025 09:02:53 +0100 Subject: [PATCH 013/286] Only support escape literals for Postgres, Redshift and generic dialect (#1674) --- src/dialect/generic.rs | 4 ++++ src/dialect/mod.rs | 7 ++++++ src/dialect/postgresql.rs | 4 ++++ src/dialect/redshift.rs | 4 ++++ src/test_utils.rs | 6 ++++- src/tokenizer.rs | 46 ++++++++++++++++++++++++++++++++++++++- 6 files changed, 69 insertions(+), 2 deletions(-) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index d696861b5..4021b5753 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -139,4 +139,8 @@ impl Dialect for GenericDialect { fn supports_user_host_grantee(&self) -> bool { true } + + fn supports_string_escape_constant(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 119bb3cf7..79260326b 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -840,6 +840,13 @@ pub trait Dialect: Debug + Any { fn supports_timestamp_versioning(&self) -> bool { false } + + /// Returns true if this dialect supports the E'...' syntax for string literals + /// + /// Postgres: + fn supports_string_escape_constant(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 170b0a7c9..d4f2a032e 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -245,6 +245,10 @@ impl Dialect for PostgreSqlDialect { fn supports_nested_comments(&self) -> bool { true } + + fn supports_string_escape_constant(&self) -> bool { + true + } } pub fn parse_create(parser: &mut Parser) -> Option> { diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index 55405ba53..a4522bbf8 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -109,4 +109,8 @@ impl Dialect for RedshiftSqlDialect { fn supports_partiql(&self) -> bool { true } + + fn supports_string_escape_constant(&self) -> bool { + true + } } diff --git a/src/test_utils.rs b/src/test_utils.rs index 51e4fd748..1c322f654 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -240,13 +240,17 @@ impl TestedDialects { /// Check that the tokenizer returns the expected tokens for the given SQL. pub fn tokenizes_to(&self, sql: &str, expected: Vec) { + if self.dialects.is_empty() { + panic!("No dialects to test"); + } + self.dialects.iter().for_each(|dialect| { let mut tokenizer = Tokenizer::new(&**dialect, sql); if let Some(options) = &self.options { tokenizer = tokenizer.with_unescape(options.unescape); } let tokens = tokenizer.tokenize().unwrap(); - assert_eq!(expected, tokens); + assert_eq!(expected, tokens, "Tokenized differently for {:?}", dialect); }); } } diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 08e233b66..309f09d81 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -985,7 +985,7 @@ impl<'a> Tokenizer<'a> { } } // PostgreSQL accepts "escape" string constants, which are an extension to the SQL standard. - x @ 'e' | x @ 'E' => { + x @ 'e' | x @ 'E' if self.dialect.supports_string_escape_constant() => { let starting_loc = chars.location(); chars.next(); // consume, to check the next char match chars.peek() { @@ -3573,4 +3573,48 @@ mod tests { ], ); } + + #[test] + fn test_string_escape_constant_not_supported() { + all_dialects_where(|dialect| !dialect.supports_string_escape_constant()).tokenizes_to( + "select e'...'", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::make_word("e", None), + Token::SingleQuotedString("...".to_string()), + ], + ); + + all_dialects_where(|dialect| !dialect.supports_string_escape_constant()).tokenizes_to( + "select E'...'", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::make_word("E", None), + Token::SingleQuotedString("...".to_string()), + ], + ); + } + + #[test] + fn test_string_escape_constant_supported() { + all_dialects_where(|dialect| dialect.supports_string_escape_constant()).tokenizes_to( + "select e'\\''", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::EscapedStringLiteral("'".to_string()), + ], + ); + + all_dialects_where(|dialect| dialect.supports_string_escape_constant()).tokenizes_to( + "select E'\\''", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::EscapedStringLiteral("'".to_string()), + ], + ); + } } From fd6c98e9332cfe80f44acbe3a994309065c70946 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Sat, 25 Jan 2025 16:01:33 +0100 Subject: [PATCH 014/286] BigQuery: Support trailing commas in column definitions list (#1682) --- src/dialect/bigquery.rs | 5 ++++ src/dialect/mod.rs | 9 +++++- src/parser/mod.rs | 12 ++++++-- tests/sqlparser_common.rs | 59 ++++++++++++++++++++++++++++----------- 4 files changed, 65 insertions(+), 20 deletions(-) diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index e92169a35..716174391 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -31,6 +31,11 @@ impl Dialect for BigQueryDialect { true } + /// See + fn supports_column_definition_trailing_commas(&self) -> bool { + true + } + fn is_identifier_start(&self, ch: char) -> bool { ch.is_ascii_lowercase() || ch.is_ascii_uppercase() || ch == '_' } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 79260326b..9fc16cd56 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -405,11 +405,18 @@ pub trait Dialect: Debug + Any { } /// Returns true if the dialect supports trailing commas in the `FROM` clause of a `SELECT` statement. - /// /// Example: `SELECT 1 FROM T, U, LIMIT 1` + /// Example: `SELECT 1 FROM T, U, LIMIT 1` fn supports_from_trailing_commas(&self) -> bool { false } + /// Returns true if the dialect supports trailing commas in the + /// column definitions list of a `CREATE` statement. + /// Example: `CREATE TABLE T (x INT, y TEXT,)` + fn supports_column_definition_trailing_commas(&self) -> bool { + false + } + /// Returns true if the dialect supports double dot notation for object names /// /// Example diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 355e520ab..c5b222acf 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6718,7 +6718,11 @@ impl<'a> Parser<'a> { return self.expected("',' or ')' after column definition", self.peek_token()); }; - if rparen && (!comma || self.options.trailing_commas) { + if rparen + && (!comma + || self.dialect.supports_column_definition_trailing_commas() + || self.options.trailing_commas) + { let _ = self.consume_token(&Token::RParen); break; } @@ -9298,7 +9302,11 @@ impl<'a> Parser<'a> { self.next_token(); Ok(vec![]) } else { - let cols = self.parse_comma_separated(Parser::parse_view_column)?; + let cols = self.parse_comma_separated_with_trailing_commas( + Parser::parse_view_column, + self.dialect.supports_column_definition_trailing_commas(), + None, + )?; self.expect_token(&Token::RParen)?; Ok(cols) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index ebd6bef03..e1ef2f909 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -10189,15 +10189,19 @@ fn parse_trailing_comma() { "Expected: column name or constraint definition, found: )".to_string() ) ); + + let unsupported_dialects = all_dialects_where(|d| !d.supports_trailing_commas()); + assert_eq!( + unsupported_dialects + .parse_sql_statements("SELECT * FROM track ORDER BY milliseconds,") + .unwrap_err(), + ParserError::ParserError("Expected: an expression, found: EOF".to_string()) + ); } #[test] fn parse_projection_trailing_comma() { - // Some dialects allow trailing commas only in the projection - let trailing_commas = TestedDialects::new(vec![ - Box::new(SnowflakeDialect {}), - Box::new(BigQueryDialect {}), - ]); + let trailing_commas = all_dialects_where(|d| d.supports_projection_trailing_commas()); trailing_commas.one_statement_parses_to( "SELECT album_id, name, FROM track", @@ -10210,20 +10214,14 @@ fn parse_projection_trailing_comma() { trailing_commas.verified_stmt("SELECT DISTINCT ON (album_id) name FROM track"); + let unsupported_dialects = all_dialects_where(|d| { + !d.supports_projection_trailing_commas() && !d.supports_trailing_commas() + }); assert_eq!( - trailing_commas - .parse_sql_statements("SELECT * FROM track ORDER BY milliseconds,") + unsupported_dialects + .parse_sql_statements("SELECT album_id, name, FROM track") .unwrap_err(), - ParserError::ParserError("Expected: an expression, found: EOF".to_string()) - ); - - assert_eq!( - trailing_commas - .parse_sql_statements("CREATE TABLE employees (name text, age int,)") - .unwrap_err(), - ParserError::ParserError( - "Expected: column name or constraint definition, found: )".to_string() - ), + ParserError::ParserError("Expected an expression, found: FROM".to_string()) ); } @@ -13061,6 +13059,33 @@ fn parse_overlaps() { verified_stmt("SELECT (DATE '2016-01-10', DATE '2016-02-01') OVERLAPS (DATE '2016-01-20', DATE '2016-02-10')"); } +#[test] +fn parse_column_definition_trailing_commas() { + let dialects = all_dialects_where(|d| d.supports_column_definition_trailing_commas()); + + dialects.one_statement_parses_to("CREATE TABLE T (x INT64,)", "CREATE TABLE T (x INT64)"); + dialects.one_statement_parses_to( + "CREATE TABLE T (x INT64, y INT64, )", + "CREATE TABLE T (x INT64, y INT64)", + ); + dialects.one_statement_parses_to( + "CREATE VIEW T (x, y, ) AS SELECT 1", + "CREATE VIEW T (x, y) AS SELECT 1", + ); + + let unsupported_dialects = all_dialects_where(|d| { + !d.supports_projection_trailing_commas() && !d.supports_trailing_commas() + }); + assert_eq!( + unsupported_dialects + .parse_sql_statements("CREATE TABLE employees (name text, age int,)") + .unwrap_err(), + ParserError::ParserError( + "Expected: column name or constraint definition, found: )".to_string() + ), + ); +} + #[test] fn test_trailing_commas_in_from() { let dialects = all_dialects_where(|d| d.supports_from_trailing_commas()); From 211b15e790328272b0cf4ffd01f477f75fae7d42 Mon Sep 17 00:00:00 2001 From: Ayman Elkfrawy <120422207+ayman-sigma@users.noreply.github.com> Date: Sun, 26 Jan 2025 06:13:51 -0800 Subject: [PATCH 015/286] Enhance object name path segments (#1539) --- README.md | 2 +- src/ast/helpers/stmt_create_table.rs | 4 +- src/ast/mod.rs | 32 ++- src/ast/spans.rs | 40 ++-- src/ast/visitor.rs | 6 +- src/dialect/snowflake.rs | 2 +- src/parser/mod.rs | 93 +++++---- src/test_utils.rs | 6 +- tests/sqlparser_bigquery.rs | 31 +-- tests/sqlparser_clickhouse.rs | 39 ++-- tests/sqlparser_common.rs | 293 ++++++++++++++------------- tests/sqlparser_databricks.rs | 24 ++- tests/sqlparser_duckdb.rs | 24 +-- tests/sqlparser_hive.rs | 13 +- tests/sqlparser_mssql.rs | 44 ++-- tests/sqlparser_mysql.rs | 103 +++++----- tests/sqlparser_postgres.rs | 199 ++++++++++-------- tests/sqlparser_redshift.rs | 17 +- tests/sqlparser_snowflake.rs | 70 ++++--- tests/sqlparser_sqlite.rs | 8 +- 20 files changed, 584 insertions(+), 466 deletions(-) diff --git a/README.md b/README.md index 41a44d3d7..d18a76b50 100644 --- a/README.md +++ b/README.md @@ -53,7 +53,7 @@ println!("AST: {:?}", ast); This outputs ```rust -AST: [Query(Query { ctes: [], body: Select(Select { distinct: false, projection: [UnnamedExpr(Identifier("a")), UnnamedExpr(Identifier("b")), UnnamedExpr(Value(Long(123))), UnnamedExpr(Function(Function { name: ObjectName(["myfunc"]), args: [Identifier("b")], filter: None, over: None, distinct: false }))], from: [TableWithJoins { relation: Table { name: ObjectName(["table_1"]), alias: None, args: [], with_hints: [] }, joins: [] }], selection: Some(BinaryOp { left: BinaryOp { left: Identifier("a"), op: Gt, right: Identifier("b") }, op: And, right: BinaryOp { left: Identifier("b"), op: Lt, right: Value(Long(100)) } }), group_by: [], having: None }), order_by: [OrderByExpr { expr: Identifier("a"), asc: Some(false) }, OrderByExpr { expr: Identifier("b"), asc: None }], limit: None, offset: None, fetch: None })] +AST: [Query(Query { ctes: [], body: Select(Select { distinct: false, projection: [UnnamedExpr(Identifier("a")), UnnamedExpr(Identifier("b")), UnnamedExpr(Value(Long(123))), UnnamedExpr(Function(Function { name:ObjectName([Identifier(Ident { value: "myfunc", quote_style: None })]), args: [Identifier("b")], filter: None, over: None, distinct: false }))], from: [TableWithJoins { relation: Table { name: ObjectName([Identifier(Ident { value: "table_1", quote_style: None })]), alias: None, args: [], with_hints: [] }, joins: [] }], selection: Some(BinaryOp { left: BinaryOp { left: Identifier("a"), op: Gt, right: Identifier("b") }, op: And, right: BinaryOp { left: Identifier("b"), op: Lt, right: Value(Long(100)) } }), group_by: [], having: None }), order_by: [OrderByExpr { expr: Identifier("a"), asc: Some(false) }, OrderByExpr { expr: Identifier("b"), asc: None }], limit: None, offset: None, fetch: None })] ``` diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index e7090cb86..2a44cef3e 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -42,7 +42,7 @@ use crate::parser::ParserError; /// ```rust /// use sqlparser::ast::helpers::stmt_create_table::CreateTableBuilder; /// use sqlparser::ast::{ColumnDef, DataType, Ident, ObjectName}; -/// let builder = CreateTableBuilder::new(ObjectName(vec![Ident::new("table_name")])) +/// let builder = CreateTableBuilder::new(ObjectName::from(vec![Ident::new("table_name")])) /// .if_not_exists(true) /// .columns(vec![ColumnDef { /// name: Ident::new("c1"), @@ -602,7 +602,7 @@ mod tests { #[test] pub fn test_from_valid_statement() { - let builder = CreateTableBuilder::new(ObjectName(vec![Ident::new("table_name")])); + let builder = CreateTableBuilder::new(ObjectName::from(vec![Ident::new("table_name")])); let stmt = builder.clone().build(); diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 2fc89e29b..b473dc11f 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -267,7 +267,13 @@ impl fmt::Display for Ident { #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct ObjectName(pub Vec); +pub struct ObjectName(pub Vec); + +impl From> for ObjectName { + fn from(idents: Vec) -> Self { + ObjectName(idents.into_iter().map(ObjectNamePart::Identifier).collect()) + } +} impl fmt::Display for ObjectName { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { @@ -275,6 +281,30 @@ impl fmt::Display for ObjectName { } } +/// A single part of an ObjectName +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ObjectNamePart { + Identifier(Ident), +} + +impl ObjectNamePart { + pub fn as_ident(&self) -> Option<&Ident> { + match self { + ObjectNamePart::Identifier(ident) => Some(ident), + } + } +} + +impl fmt::Display for ObjectNamePart { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ObjectNamePart::Identifier(ident) => write!(f, "{}", ident), + } + } +} + /// Represents an Array Expression, either /// `ARRAY[..]`, or `[..]` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index acd3987da..5316bfbda 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -28,13 +28,13 @@ use super::{ FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, FunctionArguments, GroupByExpr, HavingBound, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, MatchRecognizePattern, - Measure, NamedWindowDefinition, ObjectName, Offset, OnConflict, OnConflictAction, OnInsert, - OrderBy, OrderByExpr, Partition, PivotValueSource, ProjectionSelect, Query, ReferentialAction, - RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, - SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, - TableConstraint, TableFactor, TableObject, TableOptionsClustered, TableWithJoins, - UpdateTableFromKind, Use, Value, Values, ViewColumnDef, WildcardAdditionalOptions, With, - WithFill, + Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, + OnConflictAction, OnInsert, OrderBy, OrderByExpr, Partition, PivotValueSource, + ProjectionSelect, Query, ReferentialAction, RenameSelectItem, ReplaceSelectElement, + ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, + SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, + TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, + WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -1358,7 +1358,7 @@ impl Spanned for Expr { .union_opt(&overlay_for.as_ref().map(|i| i.span())), Expr::Collate { expr, collation } => expr .span() - .union(&union_spans(collation.0.iter().map(|i| i.span))), + .union(&union_spans(collation.0.iter().map(|i| i.span()))), Expr::Nested(expr) => expr.span(), Expr::Value(value) => value.span(), Expr::TypedString { .. } => Span::empty(), @@ -1462,7 +1462,7 @@ impl Spanned for Expr { object_name .0 .iter() - .map(|i| i.span) + .map(|i| i.span()) .chain(iter::once(token.0.span)), ), Expr::OuterJoin(expr) => expr.span(), @@ -1507,7 +1507,15 @@ impl Spanned for ObjectName { fn span(&self) -> Span { let ObjectName(segments) = self; - union_spans(segments.iter().map(|i| i.span)) + union_spans(segments.iter().map(|i| i.span())) + } +} + +impl Spanned for ObjectNamePart { + fn span(&self) -> Span { + match self { + ObjectNamePart::Identifier(ident) => ident.span, + } } } @@ -1538,7 +1546,7 @@ impl Spanned for Function { union_spans( name.0 .iter() - .map(|i| i.span) + .map(|i| i.span()) .chain(iter::once(args.span())) .chain(iter::once(parameters.span())) .chain(filter.iter().map(|i| i.span())) @@ -1624,7 +1632,7 @@ impl Spanned for SelectItem { object_name .0 .iter() - .map(|i| i.span) + .map(|i| i.span()) .chain(iter::once(wildcard_additional_options.span())), ), SelectItem::Wildcard(wildcard_additional_options) => wildcard_additional_options.span(), @@ -1734,7 +1742,7 @@ impl Spanned for TableFactor { } => union_spans( name.0 .iter() - .map(|i| i.span) + .map(|i| i.span()) .chain(alias.as_ref().map(|alias| { union_spans( iter::once(alias.name.span) @@ -1779,7 +1787,7 @@ impl Spanned for TableFactor { } => union_spans( name.0 .iter() - .map(|i| i.span) + .map(|i| i.span()) .chain(args.iter().map(|i| i.span())) .chain(alias.as_ref().map(|alias| alias.span())), ), @@ -1930,7 +1938,7 @@ impl Spanned for FunctionArgExpr { match self { FunctionArgExpr::Expr(expr) => expr.span(), FunctionArgExpr::QualifiedWildcard(object_name) => { - union_spans(object_name.0.iter().map(|i| i.span)) + union_spans(object_name.0.iter().map(|i| i.span())) } FunctionArgExpr::Wildcard => Span::empty(), } @@ -2141,7 +2149,7 @@ impl Spanned for TableObject { fn span(&self) -> Span { match self { TableObject::TableName(ObjectName(segments)) => { - union_spans(segments.iter().map(|i| i.span)) + union_spans(segments.iter().map(|i| i.span())) } TableObject::TableFunction(func) => func.span(), } diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index c824ad2f3..457dbbaed 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -403,7 +403,7 @@ where /// ``` /// # use sqlparser::parser::Parser; /// # use sqlparser::dialect::GenericDialect; -/// # use sqlparser::ast::{ObjectName, visit_relations_mut}; +/// # use sqlparser::ast::{ObjectName, ObjectNamePart, Ident, visit_relations_mut}; /// # use core::ops::ControlFlow; /// let sql = "SELECT a FROM foo"; /// let mut statements = Parser::parse_sql(&GenericDialect{}, sql) @@ -411,7 +411,7 @@ where /// /// // visit statements, renaming table foo to bar /// visit_relations_mut(&mut statements, |table| { -/// table.0[0].value = table.0[0].value.replace("foo", "bar"); +/// table.0[0] = ObjectNamePart::Identifier(Ident::new("bar")); /// ControlFlow::<()>::Continue(()) /// }); /// @@ -529,7 +529,7 @@ where /// if matches!(expr, Expr::Identifier(col_name) if col_name.value == "x") { /// let old_expr = std::mem::replace(expr, Expr::Value(Value::Null)); /// *expr = Expr::Function(Function { -/// name: ObjectName(vec![Ident::new("f")]), +/// name: ObjectName::from(vec![Ident::new("f")]), /// uses_odbc_syntax: false, /// args: FunctionArguments::List(FunctionArgumentList { /// duplicate_treatment: None, diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 88e54016d..bd9afb191 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -651,7 +651,7 @@ pub fn parse_snowflake_stage_name(parser: &mut Parser) -> Result { parser.prev_token(); diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c5b222acf..9cc8f0620 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -988,7 +988,7 @@ impl<'a> Parser<'a> { } Token::Mul => { return Ok(Expr::QualifiedWildcard( - ObjectName(id_parts), + ObjectName::from(id_parts), AttachedToken(next_token), )); } @@ -1128,7 +1128,7 @@ impl<'a> Parser<'a> { if dialect_of!(self is PostgreSqlDialect | GenericDialect) => { Ok(Some(Expr::Function(Function { - name: ObjectName(vec![w.clone().into_ident(w_span)]), + name: ObjectName::from(vec![w.clone().into_ident(w_span)]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::None, @@ -1143,7 +1143,7 @@ impl<'a> Parser<'a> { | Keyword::CURRENT_DATE | Keyword::LOCALTIME | Keyword::LOCALTIMESTAMP => { - Ok(Some(self.parse_time_functions(ObjectName(vec![w.clone().into_ident(w_span)]))?)) + Ok(Some(self.parse_time_functions(ObjectName::from(vec![w.clone().into_ident(w_span)]))?)) } Keyword::CASE => Ok(Some(self.parse_case_expr()?)), Keyword::CONVERT => Ok(Some(self.parse_convert_expr(false)?)), @@ -1187,7 +1187,7 @@ impl<'a> Parser<'a> { let query = self.parse_query()?; self.expect_token(&Token::RParen)?; Ok(Some(Expr::Function(Function { - name: ObjectName(vec![w.clone().into_ident(w_span)]), + name: ObjectName::from(vec![w.clone().into_ident(w_span)]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::Subquery(query), @@ -1232,7 +1232,7 @@ impl<'a> Parser<'a> { if let Some(expr) = self.parse_outer_join_expr(&id_parts) { Ok(expr) } else { - let mut expr = self.parse_function(ObjectName(id_parts))?; + let mut expr = self.parse_function(ObjectName::from(id_parts))?; // consume all period if it's a method chain expr = self.try_parse_method(expr)?; let fields = vec![]; @@ -1553,7 +1553,7 @@ impl<'a> Parser<'a> { return self.expected("an identifier or a '*' after '.'", self.peek_token()); }; Ok(Expr::QualifiedWildcard( - ObjectName(Self::exprs_to_idents(root, chain)?), + ObjectName::from(Self::exprs_to_idents(root, chain)?), AttachedToken(wildcard_token), )) } else if self.peek_token().token == Token::LParen { @@ -1566,7 +1566,7 @@ impl<'a> Parser<'a> { if let Some(expr) = self.parse_outer_join_expr(&id_parts) { Ok(expr) } else { - self.parse_function(ObjectName(id_parts)) + self.parse_function(ObjectName::from(id_parts)) } } else { if Self::is_all_ident(&root, &chain) { @@ -1694,7 +1694,7 @@ impl<'a> Parser<'a> { Token::Word(word) => word.into_ident(tok.span), _ => return p.expected("identifier", tok), }; - let func = match p.parse_function(ObjectName(vec![name]))? { + let func = match p.parse_function(ObjectName::from(vec![name]))? { Expr::Function(func) => func, _ => return p.expected("function", p.peek_token()), }; @@ -2197,7 +2197,7 @@ impl<'a> Parser<'a> { Some(expr) => Ok(expr), // Snowflake supports `position` as an ordinary function call // without the special `IN` syntax. - None => self.parse_function(ObjectName(vec![ident])), + None => self.parse_function(ObjectName::from(vec![ident])), } } @@ -4044,6 +4044,21 @@ impl<'a> Parser<'a> { Ok(values) } + /// Parse a period-separated list of 1+ items accepted by `F` + fn parse_period_separated(&mut self, mut f: F) -> Result, ParserError> + where + F: FnMut(&mut Parser<'a>) -> Result, + { + let mut values = vec![]; + loop { + values.push(f(self)?); + if !self.consume_token(&Token::Period) { + break; + } + } + Ok(values) + } + /// Parse a keyword-separated list of 1+ items accepted by `F` pub fn parse_keyword_separated( &mut self, @@ -4757,7 +4772,9 @@ impl<'a> Parser<'a> { let mut data_type = self.parse_data_type()?; if let DataType::Custom(n, _) = &data_type { // the first token is actually a name - name = Some(n.0[0].clone()); + match n.0[0].clone() { + ObjectNamePart::Identifier(ident) => name = Some(ident), + } data_type = self.parse_data_type()?; } @@ -9063,7 +9080,7 @@ impl<'a> Parser<'a> { } } } - Ok(ObjectName(idents)) + Ok(ObjectName::from(idents)) } /// Parse a possibly qualified, possibly quoted identifier, e.g. @@ -9079,20 +9096,26 @@ impl<'a> Parser<'a> { // BigQuery accepts any number of quoted identifiers of a table name. // https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#quoted_identifiers if dialect_of!(self is BigQueryDialect) - && idents.iter().any(|ident| ident.value.contains('.')) + && idents.iter().any(|part| { + part.as_ident() + .is_some_and(|ident| ident.value.contains('.')) + }) { idents = idents .into_iter() - .flat_map(|ident| { - ident + .flat_map(|part| match part.as_ident() { + Some(ident) => ident .value .split('.') - .map(|value| Ident { - value: value.into(), - quote_style: ident.quote_style, - span: ident.span, + .map(|value| { + ObjectNamePart::Identifier(Ident { + value: value.into(), + quote_style: ident.quote_style, + span: ident.span, + }) }) - .collect::>() + .collect::>(), + None => vec![part], }) .collect() } @@ -10427,14 +10450,14 @@ impl<'a> Parser<'a> { } let variables = if self.parse_keywords(&[Keyword::TIME, Keyword::ZONE]) { - OneOrManyWithParens::One(ObjectName(vec!["TIMEZONE".into()])) + OneOrManyWithParens::One(ObjectName::from(vec!["TIMEZONE".into()])) } else if self.dialect.supports_parenthesized_set_variables() && self.consume_token(&Token::LParen) { let variables = OneOrManyWithParens::Many( self.parse_comma_separated(|parser: &mut Parser<'a>| parser.parse_identifier())? .into_iter() - .map(|ident| ObjectName(vec![ident])) + .map(|ident| ObjectName::from(vec![ident])) .collect(), ); self.expect_token(&Token::RParen)?; @@ -11770,7 +11793,7 @@ impl<'a> Parser<'a> { Token::Word(w) => Ok(w.value), _ => self.expected("a function identifier", self.peek_token()), }?; - let expr = self.parse_function(ObjectName(vec![Ident::new(function_name)]))?; + let expr = self.parse_function(ObjectName::from(vec![Ident::new(function_name)]))?; let alias = if self.parse_keyword(Keyword::AS) { Some(self.parse_identifier()?) } else { @@ -11819,7 +11842,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; let aggregate_functions = self.parse_comma_separated(Self::parse_aliased_function_call)?; self.expect_keyword_is(Keyword::FOR)?; - let value_column = self.parse_object_name(false)?.0; + let value_column = self.parse_period_separated(|p| p.parse_identifier())?; self.expect_keyword_is(Keyword::IN)?; self.expect_token(&Token::LParen)?; @@ -11955,10 +11978,9 @@ impl<'a> Parser<'a> { // https://docs.aws.amazon.com/redshift/latest/mgmt/redshift-iam-access-control-native-idp.html let ident = self.parse_identifier()?; if let GranteeName::ObjectName(namespace) = name { - name = GranteeName::ObjectName(ObjectName(vec![Ident::new(format!( - "{}:{}", - namespace, ident - ))])); + name = GranteeName::ObjectName(ObjectName::from(vec![Ident::new( + format!("{}:{}", namespace, ident), + )])); }; } Grantee { @@ -12267,9 +12289,10 @@ impl<'a> Parser<'a> { let mut name = self.parse_object_name(false)?; if self.dialect.supports_user_host_grantee() && name.0.len() == 1 + && name.0[0].as_ident().is_some() && self.consume_token(&Token::AtSign) { - let user = name.0.pop().unwrap(); + let user = name.0.pop().unwrap().as_ident().unwrap().clone(); let host = self.parse_identifier()?; Ok(GranteeName::UserHost { user, host }) } else { @@ -13781,7 +13804,7 @@ impl<'a> Parser<'a> { // [ OWNED BY { table_name.column_name | NONE } ] let owned_by = if self.parse_keywords(&[Keyword::OWNED, Keyword::BY]) { if self.parse_keywords(&[Keyword::NONE]) { - Some(ObjectName(vec![Ident::new("NONE")])) + Some(ObjectName::from(vec![Ident::new("NONE")])) } else { Some(self.parse_object_name(false)?) } @@ -14072,7 +14095,9 @@ impl<'a> Parser<'a> { .parse_one_of_keywords(&[Keyword::FROM, Keyword::IN]) .is_some() { - parent_name.0.insert(0, self.parse_identifier()?); + parent_name + .0 + .insert(0, ObjectNamePart::Identifier(self.parse_identifier()?)); } (None, Some(parent_name)) } @@ -14388,14 +14413,14 @@ mod tests { test_parse_data_type!( dialect, "GEOMETRY", - DataType::Custom(ObjectName(vec!["GEOMETRY".into()]), vec![]) + DataType::Custom(ObjectName::from(vec!["GEOMETRY".into()]), vec![]) ); test_parse_data_type!( dialect, "GEOMETRY(POINT)", DataType::Custom( - ObjectName(vec!["GEOMETRY".into()]), + ObjectName::from(vec!["GEOMETRY".into()]), vec!["POINT".to_string()] ) ); @@ -14404,7 +14429,7 @@ mod tests { dialect, "GEOMETRY(POINT, 4326)", DataType::Custom( - ObjectName(vec!["GEOMETRY".into()]), + ObjectName::from(vec!["GEOMETRY".into()]), vec!["POINT".to_string(), "4326".to_string()] ) ); @@ -14540,7 +14565,7 @@ mod tests { }}; } - let dummy_name = ObjectName(vec![Ident::new("dummy_name")]); + let dummy_name = ObjectName::from(vec![Ident::new("dummy_name")]); let dummy_authorization = Ident::new("dummy_authorization"); test_parse_schema_name!( diff --git a/src/test_utils.rs b/src/test_utils.rs index 1c322f654..f2e3adf09 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -353,7 +353,7 @@ pub fn table_alias(name: impl Into) -> Option { pub fn table(name: impl Into) -> TableFactor { TableFactor::Table { - name: ObjectName(vec![Ident::new(name.into())]), + name: ObjectName::from(vec![Ident::new(name.into())]), alias: None, args: None, with_hints: vec![], @@ -381,7 +381,7 @@ pub fn table_from_name(name: ObjectName) -> TableFactor { pub fn table_with_alias(name: impl Into, alias: impl Into) -> TableFactor { TableFactor::Table { - name: ObjectName(vec![Ident::new(name)]), + name: ObjectName::from(vec![Ident::new(name)]), alias: Some(TableAlias { name: Ident::new(alias), columns: vec![], @@ -406,7 +406,7 @@ pub fn join(relation: TableFactor) -> Join { pub fn call(function: &str, args: impl IntoIterator) -> Expr { Expr::Function(Function { - name: ObjectName(vec![Ident::new(function)]), + name: ObjectName::from(vec![Ident::new(function)]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index a173a6cc9..cbb963761 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -222,7 +222,7 @@ fn parse_delete_statement() { .. }) => { assert_eq!( - table_from_name(ObjectName(vec![Ident::with_quote('"', "table")])), + table_from_name(ObjectName::from(vec![Ident::with_quote('"', "table")])), from[0].relation ); } @@ -249,7 +249,7 @@ fn parse_create_view_with_options() { } => { assert_eq!( name, - ObjectName(vec![ + ObjectName::from(vec![ "myproject".into(), "mydataset".into(), "newview".into() @@ -356,7 +356,7 @@ fn parse_create_table_with_unquoted_hyphen() { Statement::CreateTable(CreateTable { name, columns, .. }) => { assert_eq!( name, - ObjectName(vec![ + ObjectName::from(vec![ "my-pro-ject".into(), "mydataset".into(), "mytable".into() @@ -397,7 +397,7 @@ fn parse_create_table_with_options() { }) => { assert_eq!( name, - ObjectName(vec!["mydataset".into(), "newtable".into()]) + ObjectName::from(vec!["mydataset".into(), "newtable".into()]) ); assert_eq!( vec![ @@ -486,7 +486,7 @@ fn parse_nested_data_types() { let sql = "CREATE TABLE table (x STRUCT, b BYTES(42)>, y ARRAY>)"; match bigquery_and_generic().one_statement_parses_to(sql, sql) { Statement::CreateTable(CreateTable { name, columns, .. }) => { - assert_eq!(name, ObjectName(vec!["table".into()])); + assert_eq!(name, ObjectName::from(vec!["table".into()])); assert_eq!( columns, vec![ @@ -1376,7 +1376,7 @@ fn parse_table_identifiers() { assert_eq!( select.from, vec![TableWithJoins { - relation: table_from_name(ObjectName(expected)), + relation: table_from_name(ObjectName::from(expected)), joins: vec![] },] ); @@ -1518,7 +1518,10 @@ fn parse_hyphenated_table_identifiers() { ) .from[0] .relation, - table_from_name(ObjectName(vec![Ident::new("foo-123"), Ident::new("bar")])), + table_from_name(ObjectName::from(vec![ + Ident::new("foo-123"), + Ident::new("bar") + ])), ); assert_eq!( @@ -1551,7 +1554,7 @@ fn parse_table_time_travel() { select.from, vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t1")]), + name: ObjectName::from(vec![Ident::new("t1")]), alias: None, args: None, with_hints: vec![], @@ -1630,11 +1633,11 @@ fn parse_merge() { let update_action = MergeAction::Update { assignments: vec![ Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new("a")])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new("a")])), value: Expr::Value(number("1")), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new("b")])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new("b")])), value: Expr::Value(number("2")), }, ], @@ -1650,7 +1653,7 @@ fn parse_merge() { assert!(!into); assert_eq!( TableFactor::Table { - name: ObjectName(vec![Ident::new("inventory")]), + name: ObjectName::from(vec![Ident::new("inventory")]), alias: Some(TableAlias { name: Ident::new("T"), columns: vec![], @@ -1667,7 +1670,7 @@ fn parse_merge() { ); assert_eq!( TableFactor::Table { - name: ObjectName(vec![Ident::new("newArrivals")]), + name: ObjectName::from(vec![Ident::new("newArrivals")]), alias: Some(TableAlias { name: Ident::new("S"), columns: vec![], @@ -1985,7 +1988,7 @@ fn parse_map_access_expr() { }), AccessExpr::Subscript(Subscript::Index { index: Expr::Function(Function { - name: ObjectName(vec![Ident::with_span( + name: ObjectName::from(vec![Ident::with_span( Span::new(Location::of(1, 11), Location::of(1, 22)), "safe_offset", )]), @@ -2037,7 +2040,7 @@ fn test_bigquery_create_function() { or_replace: true, temporary: true, if_not_exists: false, - name: ObjectName(vec![ + name: ObjectName::from(vec![ Ident::new("project1"), Ident::new("mydataset"), Ident::new("myfunction"), diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index fed4308fc..0f22db389 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -62,7 +62,7 @@ fn parse_map_access_expr() { })], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("foos")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("foos")])), joins: vec![], }], lateral_views: vec![], @@ -166,7 +166,10 @@ fn parse_delimited_identifiers() { version, .. } => { - assert_eq!(vec![Ident::with_quote('"', "a table")], name.0); + assert_eq!( + ObjectName::from(vec![Ident::with_quote('"', "a table")]), + name + ); assert_eq!(Ident::with_quote('"', "alias"), alias.unwrap().name); assert!(args.is_none()); assert!(with_hints.is_empty()); @@ -185,7 +188,7 @@ fn parse_delimited_identifiers() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::with_quote('"', "myfun")]), + name: ObjectName::from(vec![Ident::with_quote('"', "myfun")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -302,7 +305,7 @@ fn parse_alter_table_add_projection() { Statement::AlterTable { name, operations, .. } => { - assert_eq!(name, ObjectName(vec!["t0".into()])); + assert_eq!(name, ObjectName::from(vec!["t0".into()])); assert_eq!(1, operations.len()); assert_eq!( operations[0], @@ -372,7 +375,7 @@ fn parse_alter_table_drop_projection() { Statement::AlterTable { name, operations, .. } => { - assert_eq!(name, ObjectName(vec!["t0".into()])); + assert_eq!(name, ObjectName::from(vec!["t0".into()])); assert_eq!(1, operations.len()); assert_eq!( operations[0], @@ -405,7 +408,7 @@ fn parse_alter_table_clear_and_materialize_projection() { Statement::AlterTable { name, operations, .. } => { - assert_eq!(name, ObjectName(vec!["t0".into()])); + assert_eq!(name, ObjectName::from(vec!["t0".into()])); assert_eq!(1, operations.len()); assert_eq!( operations[0], @@ -549,7 +552,7 @@ fn parse_clickhouse_data_types() { match clickhouse_and_generic().one_statement_parses_to(sql, &canonical_sql) { Statement::CreateTable(CreateTable { name, columns, .. }) => { - assert_eq!(name, ObjectName(vec!["table".into()])); + assert_eq!(name, ObjectName::from(vec!["table".into()])); assert_eq!( columns, vec![ @@ -590,7 +593,7 @@ fn parse_create_table_with_nullable() { match clickhouse_and_generic().one_statement_parses_to(sql, &canonical_sql) { Statement::CreateTable(CreateTable { name, columns, .. }) => { - assert_eq!(name, ObjectName(vec!["table".into()])); + assert_eq!(name, ObjectName::from(vec!["table".into()])); assert_eq!( columns, vec![ @@ -639,7 +642,7 @@ fn parse_create_table_with_nested_data_types() { match clickhouse().one_statement_parses_to(sql, "") { Statement::CreateTable(CreateTable { name, columns, .. }) => { - assert_eq!(name, ObjectName(vec!["table".into()])); + assert_eq!(name, ObjectName::from(vec!["table".into()])); assert_eq!( columns, vec![ @@ -755,7 +758,7 @@ fn parse_create_table_with_primary_key() { }) ); fn assert_function(actual: &Function, name: &str, arg: &str) -> bool { - assert_eq!(actual.name, ObjectName(vec![Ident::new(name)])); + assert_eq!(actual.name, ObjectName::from(vec![Ident::new(name)])); assert_eq!( actual.args, FunctionArguments::List(FunctionArgumentList { @@ -814,7 +817,7 @@ fn parse_create_table_with_variant_default_expressions() { options: vec![ColumnOptionDef { name: None, option: ColumnOption::Materialized(Expr::Function(Function { - name: ObjectName(vec![Ident::new("now")]), + name: ObjectName::from(vec![Ident::new("now")]), uses_odbc_syntax: false, args: FunctionArguments::List(FunctionArgumentList { args: vec![], @@ -836,7 +839,7 @@ fn parse_create_table_with_variant_default_expressions() { options: vec![ColumnOptionDef { name: None, option: ColumnOption::Ephemeral(Some(Expr::Function(Function { - name: ObjectName(vec![Ident::new("now")]), + name: ObjectName::from(vec![Ident::new("now")]), uses_odbc_syntax: false, args: FunctionArguments::List(FunctionArgumentList { args: vec![], @@ -867,7 +870,7 @@ fn parse_create_table_with_variant_default_expressions() { options: vec![ColumnOptionDef { name: None, option: ColumnOption::Alias(Expr::Function(Function { - name: ObjectName(vec![Ident::new("toString")]), + name: ObjectName::from(vec![Ident::new("toString")]), uses_odbc_syntax: false, args: FunctionArguments::List(FunctionArgumentList { args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr( @@ -895,14 +898,14 @@ fn parse_create_table_with_variant_default_expressions() { fn parse_create_view_with_fields_data_types() { match clickhouse().verified_stmt(r#"CREATE VIEW v (i "int", f "String") AS SELECT * FROM t"#) { Statement::CreateView { name, columns, .. } => { - assert_eq!(name, ObjectName(vec!["v".into()])); + assert_eq!(name, ObjectName::from(vec!["v".into()])); assert_eq!( columns, vec![ ViewColumnDef { name: "i".into(), data_type: Some(DataType::Custom( - ObjectName(vec![Ident { + ObjectName::from(vec![Ident { value: "int".into(), quote_style: Some('"'), span: Span::empty(), @@ -914,7 +917,7 @@ fn parse_create_view_with_fields_data_types() { ViewColumnDef { name: "f".into(), data_type: Some(DataType::Custom( - ObjectName(vec![Ident { + ObjectName::from(vec![Ident { value: "String".into(), quote_style: Some('"'), span: Span::empty(), @@ -1355,7 +1358,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( clickhouse().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -1363,7 +1366,7 @@ fn parse_use() { // Test single identifier with different type of quotes assert_eq!( clickhouse().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index e1ef2f909..6897d44ae 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -160,7 +160,7 @@ fn parse_insert_default_values() { assert_eq!(source, None); assert_eq!( table_name, - TableObject::TableName(ObjectName(vec!["test_table".into()])) + TableObject::TableName(ObjectName::from(vec!["test_table".into()])) ); } _ => unreachable!(), @@ -188,7 +188,7 @@ fn parse_insert_default_values() { assert_eq!(source, None); assert_eq!( table_name, - TableObject::TableName(ObjectName(vec!["test_table".into()])) + TableObject::TableName(ObjectName::from(vec!["test_table".into()])) ); } _ => unreachable!(), @@ -216,7 +216,7 @@ fn parse_insert_default_values() { assert_eq!(source, None); assert_eq!( table_name, - TableObject::TableName(ObjectName(vec!["test_table".into()])) + TableObject::TableName(ObjectName::from(vec!["test_table".into()])) ); } _ => unreachable!(), @@ -343,15 +343,15 @@ fn parse_update() { assignments, vec![ Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["a".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec!["a".into()])), value: Expr::Value(number("1")), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["b".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec!["b".into()])), value: Expr::Value(number("2")), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["c".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec!["c".into()])), value: Expr::Value(number("3")), }, ] @@ -396,11 +396,11 @@ fn parse_update_set_from() { stmt, Statement::Update { table: TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("t1")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("t1")])), joins: vec![], }, assignments: vec![Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new("name")])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new("name")])), value: Expr::CompoundIdentifier(vec![Ident::new("t2"), Ident::new("name")]) }], from: Some(UpdateTableFromKind::AfterSet(TableWithJoins { @@ -419,7 +419,7 @@ fn parse_update_set_from() { ], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("t1")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("t1")])), joins: vec![], }], lateral_views: vec![], @@ -488,7 +488,7 @@ fn parse_update_with_table_alias() { assert_eq!( TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("users")]), + name: ObjectName::from(vec![Ident::new("users")]), alias: Some(TableAlias { name: Ident::new("u"), columns: vec![], @@ -507,7 +507,7 @@ fn parse_update_with_table_alias() { ); assert_eq!( vec![Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![ + target: AssignmentTarget::ColumnName(ObjectName::from(vec![ Ident::new("u"), Ident::new("username") ])), @@ -577,7 +577,7 @@ fn parse_select_with_table_alias() { select.from, vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("lineitem")]), + name: ObjectName::from(vec![Ident::new("lineitem")]), alias: Some(TableAlias { name: Ident::new("l"), columns: vec![ @@ -628,7 +628,7 @@ fn parse_delete_statement() { .. }) => { assert_eq!( - table_from_name(ObjectName(vec![Ident::with_quote('"', "table")])), + table_from_name(ObjectName::from(vec![Ident::with_quote('"', "table")])), from[0].relation ); } @@ -659,22 +659,22 @@ fn parse_delete_statement_for_multi_tables() { .. }) => { assert_eq!( - ObjectName(vec![Ident::new("schema1"), Ident::new("table1")]), + ObjectName::from(vec![Ident::new("schema1"), Ident::new("table1")]), tables[0] ); assert_eq!( - ObjectName(vec![Ident::new("schema2"), Ident::new("table2")]), + ObjectName::from(vec![Ident::new("schema2"), Ident::new("table2")]), tables[1] ); assert_eq!( - table_from_name(ObjectName(vec![ + table_from_name(ObjectName::from(vec![ Ident::new("schema1"), Ident::new("table1") ])), from[0].relation ); assert_eq!( - table_from_name(ObjectName(vec![ + table_from_name(ObjectName::from(vec![ Ident::new("schema2"), Ident::new("table2") ])), @@ -695,28 +695,28 @@ fn parse_delete_statement_for_multi_tables_with_using() { .. }) => { assert_eq!( - table_from_name(ObjectName(vec![ + table_from_name(ObjectName::from(vec![ Ident::new("schema1"), Ident::new("table1") ])), from[0].relation ); assert_eq!( - table_from_name(ObjectName(vec![ + table_from_name(ObjectName::from(vec![ Ident::new("schema2"), Ident::new("table2") ])), from[1].relation ); assert_eq!( - table_from_name(ObjectName(vec![ + table_from_name(ObjectName::from(vec![ Ident::new("schema1"), Ident::new("table1") ])), using[0].relation ); assert_eq!( - table_from_name(ObjectName(vec![ + table_from_name(ObjectName::from(vec![ Ident::new("schema2"), Ident::new("table2") ])), @@ -742,7 +742,7 @@ fn parse_where_delete_statement() { .. }) => { assert_eq!( - table_from_name(ObjectName(vec![Ident::new("foo")])), + table_from_name(ObjectName::from(vec![Ident::new("foo")])), from[0].relation, ); @@ -777,7 +777,7 @@ fn parse_where_delete_with_alias_statement() { }) => { assert_eq!( TableFactor::Table { - name: ObjectName(vec![Ident::new("basket")]), + name: ObjectName::from(vec![Ident::new("basket")]), alias: Some(TableAlias { name: Ident::new("a"), columns: vec![], @@ -795,7 +795,7 @@ fn parse_where_delete_with_alias_statement() { assert_eq!( Some(vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("basket")]), + name: ObjectName::from(vec![Ident::new("basket")]), alias: Some(TableAlias { name: Ident::new("b"), columns: vec![], @@ -962,7 +962,7 @@ fn parse_select_into() { temporary: false, unlogged: false, table: false, - name: ObjectName(vec![Ident::new("table0")]), + name: ObjectName::from(vec![Ident::new("table0")]), }, only(&select.into) ); @@ -995,7 +995,7 @@ fn parse_select_wildcard() { let select = verified_only_select(sql); assert_eq!( &SelectItem::QualifiedWildcard( - ObjectName(vec![Ident::new("foo")]), + ObjectName::from(vec![Ident::new("foo")]), WildcardAdditionalOptions::default() ), only(&select.projection) @@ -1005,7 +1005,7 @@ fn parse_select_wildcard() { let select = verified_only_select(sql); assert_eq!( &SelectItem::QualifiedWildcard( - ObjectName(vec![Ident::new("myschema"), Ident::new("mytable"),]), + ObjectName::from(vec![Ident::new("myschema"), Ident::new("mytable"),]), WildcardAdditionalOptions::default(), ), only(&select.projection) @@ -1082,7 +1082,7 @@ fn parse_select_count_wildcard() { let select = verified_only_select(sql); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("COUNT")]), + name: ObjectName::from(vec![Ident::new("COUNT")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -1105,7 +1105,7 @@ fn parse_select_count_distinct() { let select = verified_only_select(sql); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("COUNT")]), + name: ObjectName::from(vec![Ident::new("COUNT")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -2342,7 +2342,7 @@ fn parse_select_having() { assert_eq!( Some(Expr::BinaryOp { left: Box::new(Expr::Function(Function { - name: ObjectName(vec![Ident::new("COUNT")]), + name: ObjectName::from(vec![Ident::new("COUNT")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -2373,7 +2373,7 @@ fn parse_select_qualify() { assert_eq!( Some(Expr::BinaryOp { left: Box::new(Expr::Function(Function { - name: ObjectName(vec![Ident::new("ROW_NUMBER")]), + name: ObjectName::from(vec![Ident::new("ROW_NUMBER")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -2780,7 +2780,7 @@ fn parse_listagg() { assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("LISTAGG")]), + name: ObjectName::from(vec![Ident::new("LISTAGG")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -2935,7 +2935,10 @@ fn parse_window_function_null_treatment_arg() { let SelectItem::UnnamedExpr(Expr::Function(actual)) = &projection[i] else { unreachable!() }; - assert_eq!(ObjectName(vec![Ident::new("FIRST_VALUE")]), actual.name); + assert_eq!( + ObjectName::from(vec![Ident::new("FIRST_VALUE")]), + actual.name + ); let FunctionArguments::List(arg_list) = &actual.args else { panic!("expected argument list") }; @@ -3231,7 +3234,7 @@ fn parse_create_table() { options: vec![ColumnOptionDef { name: None, option: ColumnOption::ForeignKey { - foreign_table: ObjectName(vec!["othertable".into()]), + foreign_table: ObjectName::from(vec!["othertable".into()]), referred_columns: vec!["a".into(), "b".into()], on_delete: None, on_update: None, @@ -3246,7 +3249,7 @@ fn parse_create_table() { options: vec![ColumnOptionDef { name: None, option: ColumnOption::ForeignKey { - foreign_table: ObjectName(vec!["othertable2".into()]), + foreign_table: ObjectName::from(vec!["othertable2".into()]), referred_columns: vec![], on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::NoAction), @@ -3262,7 +3265,7 @@ fn parse_create_table() { TableConstraint::ForeignKey { name: Some("fkey".into()), columns: vec!["lat".into()], - foreign_table: ObjectName(vec!["othertable3".into()]), + foreign_table: ObjectName::from(vec!["othertable3".into()]), referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Restrict), on_update: None, @@ -3271,7 +3274,7 @@ fn parse_create_table() { TableConstraint::ForeignKey { name: Some("fkey2".into()), columns: vec!["lat".into()], - foreign_table: ObjectName(vec!["othertable4".into()]), + foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::NoAction), on_update: Some(ReferentialAction::Restrict), @@ -3280,7 +3283,7 @@ fn parse_create_table() { TableConstraint::ForeignKey { name: None, columns: vec!["lat".into()], - foreign_table: ObjectName(vec!["othertable4".into()]), + foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::SetDefault), @@ -3289,7 +3292,7 @@ fn parse_create_table() { TableConstraint::ForeignKey { name: None, columns: vec!["lng".into()], - foreign_table: ObjectName(vec!["othertable4".into()]), + foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["longitude".into()], on_delete: None, on_update: Some(ReferentialAction::SetNull), @@ -3388,7 +3391,7 @@ fn parse_create_table_with_constraint_characteristics() { TableConstraint::ForeignKey { name: Some("fkey".into()), columns: vec!["lat".into()], - foreign_table: ObjectName(vec!["othertable3".into()]), + foreign_table: ObjectName::from(vec!["othertable3".into()]), referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Restrict), on_update: None, @@ -3401,7 +3404,7 @@ fn parse_create_table_with_constraint_characteristics() { TableConstraint::ForeignKey { name: Some("fkey2".into()), columns: vec!["lat".into()], - foreign_table: ObjectName(vec!["othertable4".into()]), + foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::NoAction), on_update: Some(ReferentialAction::Restrict), @@ -3414,7 +3417,7 @@ fn parse_create_table_with_constraint_characteristics() { TableConstraint::ForeignKey { name: None, columns: vec!["lat".into()], - foreign_table: ObjectName(vec!["othertable4".into()]), + foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::SetDefault), @@ -3427,7 +3430,7 @@ fn parse_create_table_with_constraint_characteristics() { TableConstraint::ForeignKey { name: None, columns: vec!["lng".into()], - foreign_table: ObjectName(vec!["othertable4".into()]), + foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["longitude".into()], on_delete: None, on_update: Some(ReferentialAction::SetNull), @@ -3620,7 +3623,7 @@ fn parse_create_table_hive_array() { .. }) => { assert!(if_not_exists); - assert_eq!(name, ObjectName(vec!["something".into()])); + assert_eq!(name, ObjectName::from(vec!["something".into()])); assert_eq!( columns, vec![ @@ -3817,7 +3820,7 @@ fn parse_create_table_as_table() { match verified_stmt(sql1) { Statement::CreateTable(CreateTable { query, name, .. }) => { - assert_eq!(name, ObjectName(vec![Ident::new("new_table")])); + assert_eq!(name, ObjectName::from(vec![Ident::new("new_table")])); assert_eq!(query.unwrap(), expected_query1); } _ => unreachable!(), @@ -3844,7 +3847,7 @@ fn parse_create_table_as_table() { match verified_stmt(sql2) { Statement::CreateTable(CreateTable { query, name, .. }) => { - assert_eq!(name, ObjectName(vec![Ident::new("new_table")])); + assert_eq!(name, ObjectName::from(vec![Ident::new("new_table")])); assert_eq!(query.unwrap(), expected_query2); } _ => unreachable!(), @@ -3947,8 +3950,8 @@ fn parse_create_table_clone() { let sql = "CREATE OR REPLACE TABLE a CLONE a_tmp"; match verified_stmt(sql) { Statement::CreateTable(CreateTable { name, clone, .. }) => { - assert_eq!(ObjectName(vec![Ident::new("a")]), name); - assert_eq!(Some(ObjectName(vec![(Ident::new("a_tmp"))])), clone) + assert_eq!(ObjectName::from(vec![Ident::new("a")]), name); + assert_eq!(Some(ObjectName::from(vec![(Ident::new("a_tmp"))])), clone) } _ => unreachable!(), } @@ -4176,11 +4179,11 @@ fn parse_rename_table() { Statement::RenameTable(rename_tables) => { assert_eq!( vec![RenameTable { - old_name: ObjectName(vec![ + old_name: ObjectName::from(vec![ Ident::new("test".to_string()), Ident::new("test1".to_string()), ]), - new_name: ObjectName(vec![ + new_name: ObjectName::from(vec![ Ident::new("test_db".to_string()), Ident::new("test2".to_string()), ]), @@ -4198,16 +4201,16 @@ fn parse_rename_table() { assert_eq!( vec![ RenameTable { - old_name: ObjectName(vec![Ident::new("old_table1".to_string())]), - new_name: ObjectName(vec![Ident::new("new_table1".to_string())]), + old_name: ObjectName::from(vec![Ident::new("old_table1".to_string())]), + new_name: ObjectName::from(vec![Ident::new("new_table1".to_string())]), }, RenameTable { - old_name: ObjectName(vec![Ident::new("old_table2".to_string())]), - new_name: ObjectName(vec![Ident::new("new_table2".to_string())]), + old_name: ObjectName::from(vec![Ident::new("old_table2".to_string())]), + new_name: ObjectName::from(vec![Ident::new("new_table2".to_string())]), }, RenameTable { - old_name: ObjectName(vec![Ident::new("old_table3".to_string())]), - new_name: ObjectName(vec![Ident::new("new_table3".to_string())]), + old_name: ObjectName::from(vec![Ident::new("old_table3".to_string())]), + new_name: ObjectName::from(vec![Ident::new("new_table3".to_string())]), } ], rename_tables @@ -4802,7 +4805,7 @@ fn parse_named_argument_function() { assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("FUN")]), + name: ObjectName::from(vec![Ident::new("FUN")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -4842,7 +4845,7 @@ fn parse_named_argument_function_with_eq_operator() { .verified_only_select(sql); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("FUN")]), + name: ObjectName::from(vec![Ident::new("FUN")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -4917,7 +4920,7 @@ fn parse_window_functions() { assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("row_number")]), + name: ObjectName::from(vec![Ident::new("row_number")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -5044,7 +5047,7 @@ fn test_parse_named_window() { projection: vec![ SelectItem::ExprWithAlias { expr: Expr::Function(Function { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "MIN".to_string(), quote_style: None, span: Span::empty(), @@ -5079,7 +5082,7 @@ fn test_parse_named_window() { }, SelectItem::ExprWithAlias { expr: Expr::Function(Function { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "MAX".to_string(), quote_style: None, span: Span::empty(), @@ -5115,7 +5118,7 @@ fn test_parse_named_window() { ], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "aggregate_test_100".to_string(), quote_style: None, span: Span::empty(), @@ -5729,7 +5732,7 @@ fn parse_interval_and_or_xor() { }))], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "test".to_string(), quote_style: None, span: Span::empty(), @@ -6341,11 +6344,11 @@ fn parse_implicit_join() { assert_eq!( vec![ TableWithJoins { - relation: table_from_name(ObjectName(vec!["t1".into()])), + relation: table_from_name(ObjectName::from(vec!["t1".into()])), joins: vec![], }, TableWithJoins { - relation: table_from_name(ObjectName(vec!["t2".into()])), + relation: table_from_name(ObjectName::from(vec!["t2".into()])), joins: vec![], }, ], @@ -6357,17 +6360,17 @@ fn parse_implicit_join() { assert_eq!( vec![ TableWithJoins { - relation: table_from_name(ObjectName(vec!["t1a".into()])), + relation: table_from_name(ObjectName::from(vec!["t1a".into()])), joins: vec![Join { - relation: table_from_name(ObjectName(vec!["t1b".into()])), + relation: table_from_name(ObjectName::from(vec!["t1b".into()])), global: false, join_operator: JoinOperator::Inner(JoinConstraint::Natural), }], }, TableWithJoins { - relation: table_from_name(ObjectName(vec!["t2a".into()])), + relation: table_from_name(ObjectName::from(vec!["t2a".into()])), joins: vec![Join { - relation: table_from_name(ObjectName(vec!["t2b".into()])), + relation: table_from_name(ObjectName::from(vec!["t2b".into()])), global: false, join_operator: JoinOperator::Inner(JoinConstraint::Natural), }], @@ -6383,7 +6386,7 @@ fn parse_cross_join() { let select = verified_only_select(sql); assert_eq!( Join { - relation: table_from_name(ObjectName(vec![Ident::new("t2")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("t2")])), global: false, join_operator: JoinOperator::CrossJoin, }, @@ -6401,7 +6404,7 @@ fn parse_joins_on() { ) -> Join { Join { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new(relation.into())]), + name: ObjectName::from(vec![Ident::new(relation.into())]), alias, args: None, with_hints: vec![], @@ -6530,7 +6533,7 @@ fn parse_joins_using() { ) -> Join { Join { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new(relation.into())]), + name: ObjectName::from(vec![Ident::new(relation.into())]), alias, args: None, with_hints: vec![], @@ -6541,7 +6544,9 @@ fn parse_joins_using() { sample: None, }, global: false, - join_operator: f(JoinConstraint::Using(vec![ObjectName(vec!["c1".into()])])), + join_operator: f(JoinConstraint::Using(vec![ObjectName::from(vec![ + "c1".into() + ])])), } } // Test parsing of aliases @@ -6606,7 +6611,7 @@ fn parse_natural_join() { fn natural_join(f: impl Fn(JoinConstraint) -> JoinOperator, alias: Option) -> Join { Join { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t2")]), + name: ObjectName::from(vec![Ident::new("t2")]), alias, args: None, with_hints: vec![], @@ -6878,7 +6883,7 @@ fn parse_derived_tables() { }), }, joins: vec![Join { - relation: table_from_name(ObjectName(vec!["t2".into()])), + relation: table_from_name(ObjectName::from(vec!["t2".into()])), global: false, join_operator: JoinOperator::Inner(JoinConstraint::Natural), }], @@ -7826,7 +7831,7 @@ fn lateral_function() { top_before_distinct: false, into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "customer".to_string(), quote_style: None, span: Span::empty(), @@ -7834,7 +7839,7 @@ fn lateral_function() { joins: vec![Join { relation: TableFactor::Function { lateral: true, - name: ObjectName(vec!["generate_series".into()]), + name: ObjectName::from(vec!["generate_series".into()]), args: vec![ FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("1")))), FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::CompoundIdentifier( @@ -7995,7 +8000,7 @@ fn parse_set_variable() { assert!(!hivevar); assert_eq!( variables, - OneOrManyWithParens::One(ObjectName(vec!["SOMETHING".into()])) + OneOrManyWithParens::One(ObjectName::from(vec!["SOMETHING".into()])) ); assert_eq!( value, @@ -8019,9 +8024,9 @@ fn parse_set_variable() { assert_eq!( variables, OneOrManyWithParens::Many(vec![ - ObjectName(vec!["a".into()]), - ObjectName(vec!["b".into()]), - ObjectName(vec!["c".into()]), + ObjectName::from(vec!["a".into()]), + ObjectName::from(vec!["b".into()]), + ObjectName::from(vec!["c".into()]), ]) ); assert_eq!( @@ -8095,7 +8100,7 @@ fn parse_set_role_as_variable() { assert!(!hivevar); assert_eq!( variables, - OneOrManyWithParens::One(ObjectName(vec!["role".into()])) + OneOrManyWithParens::One(ObjectName::from(vec!["role".into()])) ); assert_eq!( value, @@ -8142,7 +8147,7 @@ fn parse_set_time_zone() { assert!(!hivevar); assert_eq!( variable, - OneOrManyWithParens::One(ObjectName(vec!["TIMEZONE".into()])) + OneOrManyWithParens::One(ObjectName::from(vec!["TIMEZONE".into()])) ); assert_eq!( value, @@ -8698,7 +8703,7 @@ fn parse_merge() { assert_eq!( table, TableFactor::Table { - name: ObjectName(vec![Ident::new("s"), Ident::new("bar")]), + name: ObjectName::from(vec![Ident::new("s"), Ident::new("bar")]), alias: Some(TableAlias { name: Ident::new("dest"), columns: vec![], @@ -8730,7 +8735,7 @@ fn parse_merge() { )], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![ + relation: table_from_name(ObjectName::from(vec![ Ident::new("s"), Ident::new("foo") ])), @@ -8844,7 +8849,7 @@ fn parse_merge() { action: MergeAction::Update { assignments: vec![ Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![ + target: AssignmentTarget::ColumnName(ObjectName::from(vec![ Ident::new("dest"), Ident::new("F") ])), @@ -8854,7 +8859,7 @@ fn parse_merge() { ]), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![ + target: AssignmentTarget::ColumnName(ObjectName::from(vec![ Ident::new("dest"), Ident::new("G") ])), @@ -8961,12 +8966,12 @@ fn test_lock_table() { let lock = ast.locks.pop().unwrap(); assert_eq!(lock.lock_type, LockType::Update); assert_eq!( - lock.of.unwrap().0, - vec![Ident { + lock.of.unwrap(), + ObjectName::from(vec![Ident { value: "school".to_string(), quote_style: None, span: Span::empty(), - }] + }]) ); assert!(lock.nonblock.is_none()); @@ -8976,12 +8981,12 @@ fn test_lock_table() { let lock = ast.locks.pop().unwrap(); assert_eq!(lock.lock_type, LockType::Share); assert_eq!( - lock.of.unwrap().0, - vec![Ident { + lock.of.unwrap(), + ObjectName::from(vec![Ident { value: "school".to_string(), quote_style: None, span: Span::empty(), - }] + }]) ); assert!(lock.nonblock.is_none()); @@ -8991,23 +8996,23 @@ fn test_lock_table() { let lock = ast.locks.remove(0); assert_eq!(lock.lock_type, LockType::Share); assert_eq!( - lock.of.unwrap().0, - vec![Ident { + lock.of.unwrap(), + ObjectName::from(vec![Ident { value: "school".to_string(), quote_style: None, span: Span::empty(), - }] + }]) ); assert!(lock.nonblock.is_none()); let lock = ast.locks.remove(0); assert_eq!(lock.lock_type, LockType::Update); assert_eq!( - lock.of.unwrap().0, - vec![Ident { + lock.of.unwrap(), + ObjectName::from(vec![Ident { value: "student".to_string(), quote_style: None, span: Span::empty(), - }] + }]) ); assert!(lock.nonblock.is_none()); } @@ -9020,12 +9025,12 @@ fn test_lock_nonblock() { let lock = ast.locks.pop().unwrap(); assert_eq!(lock.lock_type, LockType::Update); assert_eq!( - lock.of.unwrap().0, - vec![Ident { + lock.of.unwrap(), + ObjectName::from(vec![Ident { value: "school".to_string(), quote_style: None, span: Span::empty(), - }] + }]) ); assert_eq!(lock.nonblock.unwrap(), NonBlock::SkipLocked); @@ -9035,12 +9040,12 @@ fn test_lock_nonblock() { let lock = ast.locks.pop().unwrap(); assert_eq!(lock.lock_type, LockType::Share); assert_eq!( - lock.of.unwrap().0, - vec![Ident { + lock.of.unwrap(), + ObjectName::from(vec![Ident { value: "school".to_string(), quote_style: None, span: Span::empty(), - }] + }]) ); assert_eq!(lock.nonblock.unwrap(), NonBlock::Nowait); } @@ -9222,7 +9227,7 @@ fn parse_time_functions() { let sql = format!("SELECT {}()", func_name); let select = verified_only_select(&sql); let select_localtime_func_call_ast = Function { - name: ObjectName(vec![Ident::new(func_name)]), + name: ObjectName::from(vec![Ident::new(func_name)]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -9495,7 +9500,7 @@ fn parse_cache_table() { verified_stmt(format!("CACHE TABLE '{cache_table_name}'").as_str()), Statement::Cache { table_flag: None, - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: false, options: vec![], query: None, @@ -9505,8 +9510,8 @@ fn parse_cache_table() { assert_eq!( verified_stmt(format!("CACHE {table_flag} TABLE '{cache_table_name}'").as_str()), Statement::Cache { - table_flag: Some(ObjectName(vec![Ident::new(table_flag)])), - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_flag: Some(ObjectName::from(vec![Ident::new(table_flag)])), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: false, options: vec![], query: None, @@ -9521,8 +9526,8 @@ fn parse_cache_table() { .as_str() ), Statement::Cache { - table_flag: Some(ObjectName(vec![Ident::new(table_flag)])), - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_flag: Some(ObjectName::from(vec![Ident::new(table_flag)])), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: false, options: vec![ SqlOption::KeyValue { @@ -9546,8 +9551,8 @@ fn parse_cache_table() { .as_str() ), Statement::Cache { - table_flag: Some(ObjectName(vec![Ident::new(table_flag)])), - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_flag: Some(ObjectName::from(vec![Ident::new(table_flag)])), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: false, options: vec![ SqlOption::KeyValue { @@ -9571,8 +9576,8 @@ fn parse_cache_table() { .as_str() ), Statement::Cache { - table_flag: Some(ObjectName(vec![Ident::new(table_flag)])), - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_flag: Some(ObjectName::from(vec![Ident::new(table_flag)])), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: true, options: vec![ SqlOption::KeyValue { @@ -9591,8 +9596,8 @@ fn parse_cache_table() { assert_eq!( verified_stmt(format!("CACHE {table_flag} TABLE '{cache_table_name}' {sql}").as_str()), Statement::Cache { - table_flag: Some(ObjectName(vec![Ident::new(table_flag)])), - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_flag: Some(ObjectName::from(vec![Ident::new(table_flag)])), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: false, options: vec![], query: Some(query.clone().into()), @@ -9602,8 +9607,8 @@ fn parse_cache_table() { assert_eq!( verified_stmt(format!("CACHE {table_flag} TABLE '{cache_table_name}' AS {sql}").as_str()), Statement::Cache { - table_flag: Some(ObjectName(vec![Ident::new(table_flag)])), - table_name: ObjectName(vec![Ident::with_quote('\'', cache_table_name)]), + table_flag: Some(ObjectName::from(vec![Ident::new(table_flag)])), + table_name: ObjectName::from(vec![Ident::with_quote('\'', cache_table_name)]), has_as: true, options: vec![], query: Some(query.into()), @@ -9666,7 +9671,7 @@ fn parse_uncache_table() { assert_eq!( verified_stmt("UNCACHE TABLE 'table_name'"), Statement::UNCache { - table_name: ObjectName(vec![Ident::with_quote('\'', "table_name")]), + table_name: ObjectName::from(vec![Ident::with_quote('\'', "table_name")]), if_exists: false, } ); @@ -9674,7 +9679,7 @@ fn parse_uncache_table() { assert_eq!( verified_stmt("UNCACHE TABLE IF EXISTS 'table_name'"), Statement::UNCache { - table_name: ObjectName(vec![Ident::with_quote('\'', "table_name")]), + table_name: ObjectName::from(vec![Ident::with_quote('\'', "table_name")]), if_exists: true, } ); @@ -9881,7 +9886,7 @@ fn parse_pivot_table() { verified_only_select(sql).from[0].relation, Pivot { table: Box::new(TableFactor::Table { - name: ObjectName(vec![Ident::new("monthly_sales")]), + name: ObjectName::from(vec![Ident::new("monthly_sales")]), alias: Some(TableAlias { name: Ident::new("a"), columns: vec![] @@ -9957,7 +9962,7 @@ fn parse_unpivot_table() { verified_only_select(sql).from[0].relation, Unpivot { table: Box::new(TableFactor::Table { - name: ObjectName(vec![Ident::new("sales")]), + name: ObjectName::from(vec![Ident::new("sales")]), alias: Some(TableAlias { name: Ident::new("s"), columns: vec![] @@ -10028,7 +10033,7 @@ fn parse_pivot_unpivot_table() { Pivot { table: Box::new(Unpivot { table: Box::new(TableFactor::Table { - name: ObjectName(vec![Ident::new("census")]), + name: ObjectName::from(vec![Ident::new("census")]), alias: Some(TableAlias { name: Ident::new("c"), columns: vec![] @@ -10231,7 +10236,7 @@ fn parse_create_type() { verified_stmt("CREATE TYPE db.type_name AS (foo INT, bar TEXT COLLATE \"de_DE\")"); assert_eq!( Statement::CreateType { - name: ObjectName(vec![Ident::new("db"), Ident::new("type_name")]), + name: ObjectName::from(vec![Ident::new("db"), Ident::new("type_name")]), representation: UserDefinedTypeRepresentation::Composite { attributes: vec![ UserDefinedTypeCompositeAttributeDef { @@ -10242,7 +10247,7 @@ fn parse_create_type() { UserDefinedTypeCompositeAttributeDef { name: Ident::new("bar"), data_type: DataType::Text, - collation: Some(ObjectName(vec![Ident::with_quote('\"', "de_DE")])), + collation: Some(ObjectName::from(vec![Ident::with_quote('\"', "de_DE")])), } ] } @@ -10323,7 +10328,7 @@ fn parse_call() { )))], clauses: vec![], }), - name: ObjectName(vec![Ident::new("my_procedure")]), + name: ObjectName::from(vec![Ident::new("my_procedure")]), filter: None, null_treatment: None, over: None, @@ -10335,7 +10340,7 @@ fn parse_call() { #[test] fn parse_execute_stored_procedure() { let expected = Statement::Execute { - name: ObjectName(vec![ + name: ObjectName::from(vec![ Ident { value: "my_schema".to_string(), quote_style: None, @@ -10447,7 +10452,7 @@ fn parse_unload() { projection: vec![UnnamedExpr(Expr::Identifier(Ident::new("cola"))),], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("tab")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("tab")])), joins: vec![], }], lateral_views: vec![], @@ -10600,7 +10605,7 @@ fn parse_map_access_expr() { }), AccessExpr::Subscript(Subscript::Index { index: Expr::Function(Function { - name: ObjectName(vec![Ident::with_span( + name: ObjectName::from(vec![Ident::with_span( Span::new(Location::of(1, 11), Location::of(1, 22)), "safe_offset", )]), @@ -10641,7 +10646,7 @@ fn parse_connect_by() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("title"))), ], from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("employees")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("employees")])), joins: vec![], }], into: None, @@ -10721,7 +10726,7 @@ fn parse_connect_by() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("title"))), ], from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("employees")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("employees")])), joins: vec![], }], into: None, @@ -10797,7 +10802,7 @@ fn test_selective_aggregation() { .projection, vec![ SelectItem::UnnamedExpr(Expr::Function(Function { - name: ObjectName(vec![Ident::new("ARRAY_AGG")]), + name: ObjectName::from(vec![Ident::new("ARRAY_AGG")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -10816,7 +10821,7 @@ fn test_selective_aggregation() { })), SelectItem::ExprWithAlias { expr: Expr::Function(Function { - name: ObjectName(vec![Ident::new("ARRAY_AGG")]), + name: ObjectName::from(vec![Ident::new("ARRAY_AGG")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -10876,7 +10881,7 @@ fn test_match_recognize() { use MatchRecognizeSymbol::*; use RepetitionQuantifier::*; - let table = table_from_name(ObjectName(vec![Ident::new("my_table")])); + let table = table_from_name(ObjectName::from(vec![Ident::new("my_table")])); fn check(options: &str, expect: TableFactor) { let select = all_dialects_where(|d| d.supports_match_recognize()).verified_only_select( @@ -11260,7 +11265,7 @@ fn parse_odbc_scalar_function() { else { unreachable!("expected function") }; - assert_eq!(name, &ObjectName(vec![Ident::new("my_func")])); + assert_eq!(name, &ObjectName::from(vec![Ident::new("my_func")])); assert!(uses_odbc_syntax); matches!(args, FunctionArguments::List(l) if l.args.len() == 2); @@ -12327,7 +12332,7 @@ fn parse_load_data() { assert_eq!("/local/path/to/data.txt", inpath); assert_eq!(false, overwrite); assert_eq!( - ObjectName(vec![Ident::new("test"), Ident::new("my_table")]), + ObjectName::from(vec![Ident::new("test"), Ident::new("my_table")]), table_name ); assert_eq!(None, partitioned); @@ -12350,7 +12355,7 @@ fn parse_load_data() { assert_eq!(false, local); assert_eq!("/local/path/to/data.txt", inpath); assert_eq!(true, overwrite); - assert_eq!(ObjectName(vec![Ident::new("my_table")]), table_name); + assert_eq!(ObjectName::from(vec![Ident::new("my_table")]), table_name); assert_eq!(None, partitioned); assert_eq!(None, table_format); } @@ -12387,7 +12392,7 @@ fn parse_load_data() { assert_eq!("/local/path/to/data.txt", inpath); assert_eq!(false, overwrite); assert_eq!( - ObjectName(vec![Ident::new("test"), Ident::new("my_table")]), + ObjectName::from(vec![Ident::new("test"), Ident::new("my_table")]), table_name ); assert_eq!(None, partitioned); @@ -12425,7 +12430,7 @@ fn parse_load_data() { assert_eq!(true, local); assert_eq!("/local/path/to/data.txt", inpath); assert_eq!(false, overwrite); - assert_eq!(ObjectName(vec![Ident::new("my_table")]), table_name); + assert_eq!(ObjectName::from(vec![Ident::new("my_table")]), table_name); assert_eq!( Some(vec![ Expr::BinaryOp { @@ -12461,7 +12466,7 @@ fn parse_load_data() { assert_eq!("/local/path/to/data.txt", inpath); assert_eq!(true, overwrite); assert_eq!( - ObjectName(vec![Ident::new("good"), Ident::new("my_table")]), + ObjectName::from(vec![Ident::new("good"), Ident::new("my_table")]), table_name ); assert_eq!( @@ -12815,7 +12820,7 @@ fn parse_composite_access_expr() { verified_expr("f(a).b"), Expr::CompoundFieldAccess { root: Box::new(Expr::Function(Function { - name: ObjectName(vec![Ident::new("f")]), + name: ObjectName::from(vec![Ident::new("f")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -12839,7 +12844,7 @@ fn parse_composite_access_expr() { verified_expr("f(a).b.c"), Expr::CompoundFieldAccess { root: Box::new(Expr::Function(Function { - name: ObjectName(vec![Ident::new("f")]), + name: ObjectName::from(vec![Ident::new("f")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -12865,7 +12870,7 @@ fn parse_composite_access_expr() { let stmt = verified_only_select("SELECT f(a).b FROM t WHERE f(a).b IS NOT NULL"); let expr = Expr::CompoundFieldAccess { root: Box::new(Expr::Function(Function { - name: ObjectName(vec![Ident::new("f")]), + name: ObjectName::from(vec![Ident::new("f")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index b9ca55d13..be7588de2 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -185,7 +185,9 @@ fn test_values_clause() { "SELECT * FROM values", )); assert_eq!( - Some(&table_from_name(ObjectName(vec![Ident::new("values")]))), + Some(&table_from_name(ObjectName::from(vec![Ident::new( + "values" + )]))), query .body .as_select() @@ -205,7 +207,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( databricks().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -213,7 +215,7 @@ fn parse_use() { // Test single identifier with different type of quotes assert_eq!( databricks().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) @@ -225,21 +227,21 @@ fn parse_use() { // Test single identifier with keyword and different type of quotes assert_eq!( databricks().verified_stmt(&format!("USE CATALOG {0}my_catalog{0}", quote)), - Statement::Use(Use::Catalog(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Catalog(ObjectName::from(vec![Ident::with_quote( quote, "my_catalog".to_string(), )]))) ); assert_eq!( databricks().verified_stmt(&format!("USE DATABASE {0}my_database{0}", quote)), - Statement::Use(Use::Database(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Database(ObjectName::from(vec![Ident::with_quote( quote, "my_database".to_string(), )]))) ); assert_eq!( databricks().verified_stmt(&format!("USE SCHEMA {0}my_schema{0}", quote)), - Statement::Use(Use::Schema(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Schema(ObjectName::from(vec![Ident::with_quote( quote, "my_schema".to_string(), )]))) @@ -249,15 +251,19 @@ fn parse_use() { // Test single identifier with keyword and no quotes assert_eq!( databricks().verified_stmt("USE CATALOG my_catalog"), - Statement::Use(Use::Catalog(ObjectName(vec![Ident::new("my_catalog")]))) + Statement::Use(Use::Catalog(ObjectName::from(vec![Ident::new( + "my_catalog" + )]))) ); assert_eq!( databricks().verified_stmt("USE DATABASE my_schema"), - Statement::Use(Use::Database(ObjectName(vec![Ident::new("my_schema")]))) + Statement::Use(Use::Database(ObjectName::from(vec![Ident::new( + "my_schema" + )]))) ); assert_eq!( databricks().verified_stmt("USE SCHEMA my_schema"), - Statement::Use(Use::Schema(ObjectName(vec![Ident::new("my_schema")]))) + Statement::Use(Use::Schema(ObjectName::from(vec![Ident::new("my_schema")]))) ); // Test invalid syntax - missing identifier diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index ca7f926a9..aee6d654c 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -160,7 +160,7 @@ fn test_select_wildcard_with_exclude() { let select = duckdb().verified_only_select("SELECT name.* EXCLUDE department_id FROM employee_table"); let expected = SelectItem::QualifiedWildcard( - ObjectName(vec![Ident::new("name")]), + ObjectName::from(vec![Ident::new("name")]), WildcardAdditionalOptions { opt_exclude: Some(ExcludeSelectItem::Single(Ident::new("department_id"))), ..Default::default() @@ -191,7 +191,7 @@ fn test_create_macro() { let expected = Statement::CreateMacro { or_replace: false, temporary: false, - name: ObjectName(vec![Ident::new("schema"), Ident::new("add")]), + name: ObjectName::from(vec![Ident::new("schema"), Ident::new("add")]), args: Some(vec![MacroArg::new("a"), MacroArg::new("b")]), definition: MacroDefinition::Expr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("a"))), @@ -208,7 +208,7 @@ fn test_create_macro_default_args() { let expected = Statement::CreateMacro { or_replace: false, temporary: false, - name: ObjectName(vec![Ident::new("add_default")]), + name: ObjectName::from(vec![Ident::new("add_default")]), args: Some(vec![ MacroArg::new("a"), MacroArg { @@ -236,7 +236,7 @@ fn test_create_table_macro() { let expected = Statement::CreateMacro { or_replace: true, temporary: true, - name: ObjectName(vec![Ident::new("dynamic_table")]), + name: ObjectName::from(vec![Ident::new("dynamic_table")]), args: Some(vec![ MacroArg::new("col1_value"), MacroArg::new("col2_value"), @@ -268,7 +268,7 @@ fn test_select_union_by_name() { top_before_distinct: false, into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "capitals".to_string(), quote_style: None, span: Span::empty(), @@ -297,7 +297,7 @@ fn test_select_union_by_name() { top_before_distinct: false, into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "weather".to_string(), quote_style: None, span: Span::empty(), @@ -587,7 +587,7 @@ fn test_duckdb_named_argument_function_with_assignment_operator() { let select = duckdb_and_generic().verified_only_select(sql); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("FUN")]), + name: ObjectName::from(vec![Ident::new("FUN")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -661,7 +661,7 @@ fn test_duckdb_union_datatype() { transient: Default::default(), volatile: Default::default(), iceberg: Default::default(), - name: ObjectName(vec!["tbl1".into()]), + name: ObjectName::from(vec!["tbl1".into()]), columns: vec![ ColumnDef { name: "one".into(), @@ -765,7 +765,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( duckdb().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -773,7 +773,7 @@ fn parse_use() { // Test single identifier with different type of quotes assert_eq!( duckdb().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) @@ -785,7 +785,7 @@ fn parse_use() { // Test double identifier with different type of quotes assert_eq!( duckdb().verified_stmt(&format!("USE {0}CATALOG{0}.{0}my_schema{0}", quote)), - Statement::Use(Use::Object(ObjectName(vec![ + Statement::Use(Use::Object(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") ]))) @@ -794,7 +794,7 @@ fn parse_use() { // Test double identifier without quotes assert_eq!( duckdb().verified_stmt("USE mydb.my_schema"), - Statement::Use(Use::Object(ObjectName(vec![ + Statement::Use(Use::Object(ObjectName::from(vec![ Ident::new("mydb"), Ident::new("my_schema") ]))) diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 5349f1207..9c4e8f079 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -368,7 +368,7 @@ fn set_statement_with_minus() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![ + variables: OneOrManyWithParens::One(ObjectName::from(vec![ Ident::new("hive"), Ident::new("tez"), Ident::new("java"), @@ -461,7 +461,10 @@ fn parse_delimited_identifiers() { json_path: _, sample: _, } => { - assert_eq!(vec![Ident::with_quote('"', "a table")], name.0); + assert_eq!( + ObjectName::from(vec![Ident::with_quote('"', "a table")]), + name + ); assert_eq!(Ident::with_quote('"', "alias"), alias.unwrap().name); assert!(args.is_none()); assert!(with_hints.is_empty()); @@ -480,7 +483,7 @@ fn parse_delimited_identifiers() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::with_quote('"', "myfun")]), + name: ObjectName::from(vec![Ident::with_quote('"', "myfun")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -516,7 +519,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( hive().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -524,7 +527,7 @@ fn parse_use() { // Test single identifier with different type of quotes assert_eq!( hive().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index da2b6160e..3c4017590 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -63,7 +63,7 @@ fn parse_table_time_travel() { select.from, vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t1")]), + name: ObjectName::from(vec![Ident::new("t1")]), alias: None, args: None, with_hints: vec![], @@ -159,7 +159,7 @@ fn parse_create_procedure() { })) } ]), - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test".into(), quote_style: None, span: Span::empty(), @@ -211,7 +211,7 @@ fn parse_mssql_openjson() { assert_eq!( vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t_test_table")]), + name: ObjectName::from(vec![Ident::new("t_test_table")]), alias: Some(TableAlias { name: Ident::new("A"), columns: vec![] @@ -270,7 +270,7 @@ fn parse_mssql_openjson() { assert_eq!( vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t_test_table"),]), + name: ObjectName::from(vec![Ident::new("t_test_table"),]), alias: Some(TableAlias { name: Ident::new("A"), columns: vec![] @@ -329,8 +329,7 @@ fn parse_mssql_openjson() { assert_eq!( vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t_test_table")]), - + name: ObjectName::from(vec![Ident::new("t_test_table")]), alias: Some(TableAlias { name: Ident::new("A"), columns: vec![] @@ -389,7 +388,7 @@ fn parse_mssql_openjson() { assert_eq!( vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t_test_table")]), + name: ObjectName::from(vec![Ident::new("t_test_table")]), alias: Some(TableAlias { name: Ident::new("A"), columns: vec![] @@ -428,7 +427,7 @@ fn parse_mssql_openjson() { assert_eq!( vec![TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("t_test_table")]), + name: ObjectName::from(vec![Ident::new("t_test_table")]), alias: Some(TableAlias { name: Ident::new("A"), columns: vec![] @@ -532,7 +531,7 @@ fn parse_mssql_create_role() { assert_eq_vec(&["mssql"], &names); assert_eq!( authorization_owner, - Some(ObjectName(vec![Ident { + Some(ObjectName::from(vec![Ident { value: "helena".into(), quote_style: None, span: Span::empty(), @@ -619,7 +618,10 @@ fn parse_delimited_identifiers() { version, .. } => { - assert_eq!(vec![Ident::with_quote('"', "a table")], name.0); + assert_eq!( + ObjectName::from(vec![Ident::with_quote('"', "a table")]), + name + ); assert_eq!(Ident::with_quote('"', "alias"), alias.unwrap().name); assert!(args.is_none()); assert!(with_hints.is_empty()); @@ -638,7 +640,7 @@ fn parse_delimited_identifiers() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::with_quote('"', "myfun")]), + name: ObjectName::from(vec![Ident::with_quote('"', "myfun")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -671,11 +673,11 @@ fn parse_table_name_in_square_brackets() { let select = ms().verified_only_select(r#"SELECT [a column] FROM [a schema].[a table]"#); if let TableFactor::Table { name, .. } = only(select.from).relation { assert_eq!( - vec![ + ObjectName::from(vec![ Ident::with_quote('[', "a schema"), Ident::with_quote('[', "a table") - ], - name.0 + ]), + name ); } else { panic!("Expecting TableFactor::Table"); @@ -1086,7 +1088,7 @@ fn parse_substring_in_select() { })], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "test".to_string(), quote_style: None, span: Span::empty(), @@ -1204,7 +1206,7 @@ fn parse_mssql_declare() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![Ident::new("@bar")])), + variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("@bar")])), value: vec![Expr::Value(Value::Number("2".parse().unwrap(), false))], }, Statement::Query(Box::new(Query { @@ -1298,7 +1300,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( ms().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -1306,7 +1308,7 @@ fn parse_use() { // Test single identifier with different type of quotes assert_eq!( ms().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) @@ -1408,7 +1410,7 @@ fn parse_create_table_with_valid_options() { }, value: Expr::Function( Function { - name: ObjectName( + name: ObjectName::from( vec![ Ident { value: "HASH".to_string(), @@ -1472,7 +1474,7 @@ fn parse_create_table_with_valid_options() { if_not_exists: false, transient: false, volatile: false, - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "mytable".to_string(), quote_style: None, span: Span::empty(), @@ -1648,7 +1650,7 @@ fn parse_create_table_with_identity_column() { transient: false, volatile: false, iceberg: false, - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "mytable".to_string(), quote_style: None, span: Span::empty(), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index e93ac5695..fb72436ed 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -149,7 +149,7 @@ fn parse_flush() { read_lock: false, export: false, tables: vec![ - ObjectName(vec![ + ObjectName::from(vec![ Ident { value: "mek".to_string(), quote_style: Some('`'), @@ -161,7 +161,7 @@ fn parse_flush() { span: Span::empty(), } ]), - ObjectName(vec![Ident { + ObjectName::from(vec![Ident { value: "table2".to_string(), quote_style: None, span: Span::empty(), @@ -189,7 +189,7 @@ fn parse_flush() { read_lock: true, export: false, tables: vec![ - ObjectName(vec![ + ObjectName::from(vec![ Ident { value: "mek".to_string(), quote_style: Some('`'), @@ -201,7 +201,7 @@ fn parse_flush() { span: Span::empty(), } ]), - ObjectName(vec![Ident { + ObjectName::from(vec![Ident { value: "table2".to_string(), quote_style: None, span: Span::empty(), @@ -218,7 +218,7 @@ fn parse_flush() { read_lock: false, export: true, tables: vec![ - ObjectName(vec![ + ObjectName::from(vec![ Ident { value: "mek".to_string(), quote_style: Some('`'), @@ -230,7 +230,7 @@ fn parse_flush() { span: Span::empty(), } ]), - ObjectName(vec![Ident { + ObjectName::from(vec![Ident { value: "table2".to_string(), quote_style: None, span: Span::empty(), @@ -251,7 +251,7 @@ fn parse_show_columns() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mytable")])), + parent_name: Some(ObjectName::from(vec![Ident::new("mytable")])), }), filter_position: None, limit_from: None, @@ -269,7 +269,10 @@ fn parse_show_columns() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mydb"), Ident::new("mytable")])), + parent_name: Some(ObjectName::from(vec![ + Ident::new("mydb"), + Ident::new("mytable") + ])), }), filter_position: None, limit_from: None, @@ -287,7 +290,7 @@ fn parse_show_columns() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mytable")])), + parent_name: Some(ObjectName::from(vec![Ident::new("mytable")])), }), filter_position: None, limit_from: None, @@ -305,7 +308,7 @@ fn parse_show_columns() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mytable")])), + parent_name: Some(ObjectName::from(vec![Ident::new("mytable")])), }), filter_position: None, limit_from: None, @@ -323,7 +326,7 @@ fn parse_show_columns() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mytable")])), + parent_name: Some(ObjectName::from(vec![Ident::new("mytable")])), }), filter_position: Some(ShowStatementFilterPosition::Suffix( ShowStatementFilter::Like("pattern".into()) @@ -343,7 +346,7 @@ fn parse_show_columns() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mytable")])), + parent_name: Some(ObjectName::from(vec![Ident::new("mytable")])), }), filter_position: Some(ShowStatementFilterPosition::Suffix( ShowStatementFilter::Where(mysql_and_generic().verified_expr("1 = 2")) @@ -430,7 +433,7 @@ fn parse_show_tables() { show_in: Some(ShowStatementIn { clause: ShowStatementInClause::FROM, parent_type: None, - parent_name: Some(ObjectName(vec![Ident::new("mydb")])), + parent_name: Some(ObjectName::from(vec![Ident::new("mydb")])), }), filter_position: None } @@ -534,7 +537,7 @@ fn parse_show_extended_full() { #[test] fn parse_show_create() { - let obj_name = ObjectName(vec![Ident::new("myident")]); + let obj_name = ObjectName::from(vec![Ident::new("myident")]); for obj_type in &[ ShowCreateObject::Table, @@ -591,7 +594,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( mysql_and_generic().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -600,7 +603,7 @@ fn parse_use() { assert_eq!( mysql_and_generic() .verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) @@ -617,7 +620,7 @@ fn parse_set_variables() { Statement::SetVariable { local: true, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec!["autocommit".into()])), + variables: OneOrManyWithParens::One(ObjectName::from(vec!["autocommit".into()])), value: vec![Expr::Value(number("1"))], } ); @@ -1017,7 +1020,7 @@ fn parse_create_table_comment_character_set() { options: vec![ ColumnOptionDef { name: None, - option: ColumnOption::CharacterSet(ObjectName(vec![Ident::new( + option: ColumnOption::CharacterSet(ObjectName::from(vec![Ident::new( "utf8mb4" )])) }, @@ -1413,7 +1416,7 @@ fn parse_simple_insert() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("tasks")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("tasks")])), table_name ); assert_eq!(vec![Ident::new("title"), Ident::new("priority")], columns); @@ -1471,7 +1474,7 @@ fn parse_ignore_insert() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("tasks")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("tasks")])), table_name ); assert_eq!(vec![Ident::new("title"), Ident::new("priority")], columns); @@ -1518,7 +1521,7 @@ fn parse_priority_insert() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("tasks")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("tasks")])), table_name ); assert_eq!(vec![Ident::new("title"), Ident::new("priority")], columns); @@ -1562,7 +1565,7 @@ fn parse_priority_insert() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("tasks")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("tasks")])), table_name ); assert_eq!(vec![Ident::new("title"), Ident::new("priority")], columns); @@ -1607,14 +1610,14 @@ fn parse_insert_as() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::with_quote('`', "table")])), + TableObject::TableName(ObjectName::from(vec![Ident::with_quote('`', "table")])), table_name ); assert_eq!(vec![Ident::with_quote('`', "date")], columns); let insert_alias = insert_alias.unwrap(); assert_eq!( - ObjectName(vec![Ident::with_quote('`', "alias")]), + ObjectName::from(vec![Ident::with_quote('`', "alias")]), insert_alias.row_alias ); assert_eq!(Some(vec![]), insert_alias.col_aliases); @@ -1659,7 +1662,7 @@ fn parse_insert_as() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::with_quote('`', "table")])), + TableObject::TableName(ObjectName::from(vec![Ident::with_quote('`', "table")])), table_name ); assert_eq!( @@ -1668,7 +1671,7 @@ fn parse_insert_as() { ); let insert_alias = insert_alias.unwrap(); assert_eq!( - ObjectName(vec![Ident::with_quote('`', "alias")]), + ObjectName::from(vec![Ident::with_quote('`', "alias")]), insert_alias.row_alias ); assert_eq!( @@ -1719,7 +1722,7 @@ fn parse_replace_insert() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("tasks")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("tasks")])), table_name ); assert_eq!(vec![Ident::new("title"), Ident::new("priority")], columns); @@ -1766,7 +1769,7 @@ fn parse_empty_row_insert() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("tb")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("tb")])), table_name ); assert!(columns.is_empty()); @@ -1808,7 +1811,7 @@ fn parse_insert_with_on_duplicate_update() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("permission_groups")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("permission_groups")])), table_name ); assert_eq!( @@ -1855,31 +1858,31 @@ fn parse_insert_with_on_duplicate_update() { assert_eq!( Some(OnInsert::DuplicateKeyUpdate(vec![ Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new( + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new( "description".to_string() )])), value: call("VALUES", [Expr::Identifier(Ident::new("description"))]), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new( + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new( "perm_create".to_string() )])), value: call("VALUES", [Expr::Identifier(Ident::new("perm_create"))]), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new( + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new( "perm_read".to_string() )])), value: call("VALUES", [Expr::Identifier(Ident::new("perm_read"))]), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new( + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new( "perm_update".to_string() )])), value: call("VALUES", [Expr::Identifier(Ident::new("perm_update"))]), }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![Ident::new( + target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new( "perm_delete".to_string() )])), value: call("VALUES", [Expr::Identifier(Ident::new("perm_delete"))]), @@ -1910,7 +1913,7 @@ fn parse_select_with_numeric_prefix_column_name() { )))], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::with_quote( + relation: table_from_name(ObjectName::from(vec![Ident::with_quote( '"', "table" )])), joins: vec![] @@ -1962,7 +1965,7 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { ], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::with_quote( + relation: table_from_name(ObjectName::from(vec![Ident::with_quote( '"', "table" )])), joins: vec![] @@ -1997,7 +2000,7 @@ fn parse_insert_with_numeric_prefix_column_name() { .. }) => { assert_eq!( - TableObject::TableName(ObjectName(vec![Ident::new("s1"), Ident::new("t1")])), + TableObject::TableName(ObjectName::from(vec![Ident::new("s1"), Ident::new("t1")])), table_name ); assert_eq!(vec![Ident::new("123col_$@length123")], columns); @@ -2021,7 +2024,7 @@ fn parse_update_with_joins() { assert_eq!( TableWithJoins { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("orders")]), + name: ObjectName::from(vec![Ident::new("orders")]), alias: Some(TableAlias { name: Ident::new("o"), columns: vec![] @@ -2036,7 +2039,7 @@ fn parse_update_with_joins() { }, joins: vec![Join { relation: TableFactor::Table { - name: ObjectName(vec![Ident::new("customers")]), + name: ObjectName::from(vec![Ident::new("customers")]), alias: Some(TableAlias { name: Ident::new("c"), columns: vec![] @@ -2067,7 +2070,7 @@ fn parse_update_with_joins() { ); assert_eq!( vec![Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec![ + target: AssignmentTarget::ColumnName(ObjectName::from(vec![ Ident::new("o"), Ident::new("completed") ])), @@ -2255,7 +2258,7 @@ fn parse_alter_table_drop_primary_key() { #[test] fn parse_alter_table_change_column() { - let expected_name = ObjectName(vec![Ident::new("orders")]); + let expected_name = ObjectName::from(vec![Ident::new("orders")]); let expected_operation = AlterTableOperation::ChangeColumn { old_name: Ident::new("description"), new_name: Ident::new("desc"), @@ -2307,7 +2310,7 @@ fn parse_alter_table_change_column() { #[test] fn parse_alter_table_change_column_with_column_position() { - let expected_name = ObjectName(vec![Ident::new("orders")]); + let expected_name = ObjectName::from(vec![Ident::new("orders")]); let expected_operation_first = AlterTableOperation::ChangeColumn { old_name: Ident::new("description"), new_name: Ident::new("desc"), @@ -2355,7 +2358,7 @@ fn parse_alter_table_change_column_with_column_position() { #[test] fn parse_alter_table_modify_column() { - let expected_name = ObjectName(vec![Ident::new("orders")]); + let expected_name = ObjectName::from(vec![Ident::new("orders")]); let expected_operation = AlterTableOperation::ModifyColumn { col_name: Ident::new("description"), data_type: DataType::Text, @@ -2404,7 +2407,7 @@ fn parse_alter_table_modify_column() { #[test] fn parse_alter_table_modify_column_with_column_position() { - let expected_name = ObjectName(vec![Ident::new("orders")]); + let expected_name = ObjectName::from(vec![Ident::new("orders")]); let expected_operation_first = AlterTableOperation::ModifyColumn { col_name: Ident::new("description"), data_type: DataType::Text, @@ -2478,7 +2481,7 @@ fn parse_substring_in_select() { })], into: None, from: vec![TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident { + relation: table_from_name(ObjectName::from(vec![Ident { value: "test".to_string(), quote_style: None, span: Span::empty(), @@ -2873,10 +2876,12 @@ fn parse_create_table_with_column_collate() { vec![ColumnDef { name: Ident::new("id"), data_type: DataType::Text, - collation: Some(ObjectName(vec![Ident::new("utf8mb4_0900_ai_ci")])), + collation: Some(ObjectName::from(vec![Ident::new("utf8mb4_0900_ai_ci")])), options: vec![ColumnOptionDef { name: None, - option: ColumnOption::CharacterSet(ObjectName(vec![Ident::new("utf8mb4")])) + option: ColumnOption::CharacterSet(ObjectName::from(vec![Ident::new( + "utf8mb4" + )])) }], },], columns @@ -3039,7 +3044,7 @@ fn parse_grant() { ); assert_eq!( objects, - GrantObjects::Tables(vec![ObjectName(vec!["*".into(), "*".into()])]) + GrantObjects::Tables(vec![ObjectName::from(vec!["*".into(), "*".into()])]) ); assert!(!with_grant_option); assert!(granted_by.is_none()); @@ -3080,7 +3085,7 @@ fn parse_revoke() { ); assert_eq!( objects, - GrantObjects::Tables(vec![ObjectName(vec!["db1".into(), "*".into()])]) + GrantObjects::Tables(vec![ObjectName::from(vec!["db1".into(), "*".into()])]) ); if let [Grantee { grantee_type: GranteesType::None, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 0fca4cec1..b3eb4f10d 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -402,7 +402,7 @@ fn parse_create_table_with_defaults() { unit: None } )), - collation: Some(ObjectName(vec![Ident::with_quote('"', "es_ES")])), + collation: Some(ObjectName::from(vec![Ident::with_quote('"', "es_ES")])), options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -1040,7 +1040,7 @@ fn test_copy_from() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: false, @@ -1058,7 +1058,7 @@ fn test_copy_from() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: false, @@ -1076,7 +1076,7 @@ fn test_copy_from() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: false, @@ -1100,7 +1100,7 @@ fn test_copy_to() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: true, @@ -1118,7 +1118,7 @@ fn test_copy_to() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: true, @@ -1136,7 +1136,7 @@ fn test_copy_to() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: true, @@ -1177,7 +1177,7 @@ fn parse_copy_from() { pg_and_generic().one_statement_parses_to(sql, ""), Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["table".into()]), + table_name: ObjectName::from(vec!["table".into()]), columns: vec!["a".into(), "b".into()], }, to: false, @@ -1223,7 +1223,7 @@ fn parse_copy_to() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: true, @@ -1241,7 +1241,7 @@ fn parse_copy_to() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["country".into()]), + table_name: ObjectName::from(vec!["country".into()]), columns: vec![], }, to: true, @@ -1258,7 +1258,7 @@ fn parse_copy_to() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["country".into()]), + table_name: ObjectName::from(vec!["country".into()]), columns: vec![], }, to: true, @@ -1344,7 +1344,7 @@ fn parse_copy_from_before_v9_0() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: false, @@ -1373,7 +1373,7 @@ fn parse_copy_from_before_v9_0() { pg_and_generic().one_statement_parses_to(sql, ""), Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: false, @@ -1401,7 +1401,7 @@ fn parse_copy_to_before_v9_0() { stmt, Statement::Copy { source: CopySource::Table { - table_name: ObjectName(vec!["users".into()]), + table_name: ObjectName::from(vec!["users".into()]), columns: vec![], }, to: true, @@ -1433,7 +1433,7 @@ fn parse_set() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![Ident::new("a")])), + variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), value: vec![Expr::Identifier(Ident { value: "b".into(), quote_style: None, @@ -1448,7 +1448,7 @@ fn parse_set() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![Ident::new("a")])), + variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), value: vec![Expr::Value(Value::SingleQuotedString("b".into()))], } ); @@ -1459,7 +1459,7 @@ fn parse_set() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![Ident::new("a")])), + variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), value: vec![Expr::Value(number("0"))], } ); @@ -1470,7 +1470,7 @@ fn parse_set() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![Ident::new("a")])), + variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), value: vec![Expr::Identifier(Ident::new("DEFAULT"))], } ); @@ -1481,7 +1481,7 @@ fn parse_set() { Statement::SetVariable { local: true, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![Ident::new("a")])), + variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), value: vec![Expr::Identifier("b".into())], } ); @@ -1492,7 +1492,7 @@ fn parse_set() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![ + variables: OneOrManyWithParens::One(ObjectName::from(vec![ Ident::new("a"), Ident::new("b"), Ident::new("c") @@ -1514,7 +1514,7 @@ fn parse_set() { Statement::SetVariable { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName(vec![ + variables: OneOrManyWithParens::One(ObjectName::from(vec![ Ident::new("hive"), Ident::new("tez"), Ident::new("auto"), @@ -1658,7 +1658,7 @@ fn parse_execute() { assert_eq!( stmt, Statement::Execute { - name: ObjectName(vec!["a".into()]), + name: ObjectName::from(vec!["a".into()]), parameters: vec![], has_parentheses: false, using: vec![] @@ -1669,7 +1669,7 @@ fn parse_execute() { assert_eq!( stmt, Statement::Execute { - name: ObjectName(vec!["a".into()]), + name: ObjectName::from(vec!["a".into()]), parameters: vec![ Expr::Value(number("1")), Expr::Value(Value::SingleQuotedString("t".to_string())) @@ -1684,7 +1684,7 @@ fn parse_execute() { assert_eq!( stmt, Statement::Execute { - name: ObjectName(vec!["a".into()]), + name: ObjectName::from(vec!["a".into()]), parameters: vec![], has_parentheses: false, using: vec![ @@ -1793,7 +1793,9 @@ fn parse_pg_on_conflict() { assert_eq!( OnConflictAction::DoUpdate(DoUpdate { assignments: vec![Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["dname".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from( + vec!["dname".into()] + )), value: Expr::CompoundIdentifier(vec!["EXCLUDED".into(), "dname".into()]) },], selection: None @@ -1824,14 +1826,18 @@ fn parse_pg_on_conflict() { OnConflictAction::DoUpdate(DoUpdate { assignments: vec![ Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["dname".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec![ + "dname".into() + ])), value: Expr::CompoundIdentifier(vec![ "EXCLUDED".into(), "dname".into() ]) }, Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["area".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from(vec![ + "area".into() + ])), value: Expr::CompoundIdentifier(vec!["EXCLUDED".into(), "area".into()]) }, ], @@ -1881,7 +1887,9 @@ fn parse_pg_on_conflict() { assert_eq!( OnConflictAction::DoUpdate(DoUpdate { assignments: vec![Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["dname".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from( + vec!["dname".into()] + )), value: Expr::Value(Value::Placeholder("$1".to_string())) },], selection: Some(Expr::BinaryOp { @@ -1915,11 +1923,16 @@ fn parse_pg_on_conflict() { })), .. }) => { - assert_eq!(vec![Ident::from("distributors_did_pkey")], cname.0); + assert_eq!( + ObjectName::from(vec![Ident::from("distributors_did_pkey")]), + cname + ); assert_eq!( OnConflictAction::DoUpdate(DoUpdate { assignments: vec![Assignment { - target: AssignmentTarget::ColumnName(ObjectName(vec!["dname".into()])), + target: AssignmentTarget::ColumnName(ObjectName::from( + vec!["dname".into()] + )), value: Expr::Value(Value::Placeholder("$1".to_string())) },], selection: Some(Expr::BinaryOp { @@ -2624,7 +2637,7 @@ fn parse_array_subquery_expr() { let select = pg().verified_only_select(sql); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("ARRAY")]), + name: ObjectName::from(vec![Ident::new("ARRAY")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::Subquery(Box::new(Query { @@ -2963,7 +2976,10 @@ fn parse_json_table_is_not_reserved() { TableFactor::Table { name: ObjectName(name), .. - } => assert_eq!("JSON_TABLE", name[0].value), + } => assert_eq!( + ObjectNamePart::Identifier(Ident::new("JSON_TABLE")), + name[0] + ), other => panic!("Expected: JSON_TABLE to be parsed as a table name, but got {other:?}"), } } @@ -3004,7 +3020,7 @@ fn test_composite_value() { SelectItem::UnnamedExpr(Expr::CompositeAccess { key: Ident::new("n"), expr: Box::new(Expr::Nested(Box::new(Expr::Function(Function { - name: ObjectName(vec![ + name: ObjectName::from(vec![ Ident::new("information_schema"), Ident::new("_pg_expandarray") ]), @@ -3185,7 +3201,7 @@ fn parse_current_functions() { let select = pg_and_generic().verified_only_select(sql); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("CURRENT_CATALOG")]), + name: ObjectName::from(vec![Ident::new("CURRENT_CATALOG")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::None, @@ -3198,7 +3214,7 @@ fn parse_current_functions() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("CURRENT_USER")]), + name: ObjectName::from(vec![Ident::new("CURRENT_USER")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::None, @@ -3211,7 +3227,7 @@ fn parse_current_functions() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("SESSION_USER")]), + name: ObjectName::from(vec![Ident::new("SESSION_USER")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::None, @@ -3224,7 +3240,7 @@ fn parse_current_functions() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::new("USER")]), + name: ObjectName::from(vec![Ident::new("USER")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::None, @@ -3536,7 +3552,7 @@ fn parse_alter_role() { span: Span::empty(), }, operation: AlterRoleOperation::Set { - config_name: ObjectName(vec![Ident { + config_name: ObjectName::from(vec![Ident { value: "maintenance_work_mem".into(), quote_style: None, span: Span::empty(), @@ -3557,13 +3573,13 @@ fn parse_alter_role() { span: Span::empty(), }, operation: AlterRoleOperation::Set { - config_name: ObjectName(vec![Ident { + config_name: ObjectName::from(vec![Ident { value: "maintenance_work_mem".into(), quote_style: None, span: Span::empty(), }]), config_value: SetConfigValue::Value(Expr::Value(number("100000"))), - in_database: Some(ObjectName(vec![Ident { + in_database: Some(ObjectName::from(vec![Ident { value: "database_name".into(), quote_style: None, span: Span::empty(), @@ -3582,13 +3598,13 @@ fn parse_alter_role() { span: Span::empty(), }, operation: AlterRoleOperation::Set { - config_name: ObjectName(vec![Ident { + config_name: ObjectName::from(vec![Ident { value: "maintenance_work_mem".into(), quote_style: None, span: Span::empty(), }]), config_value: SetConfigValue::Value(Expr::Value(number("100000"))), - in_database: Some(ObjectName(vec![Ident { + in_database: Some(ObjectName::from(vec![Ident { value: "database_name".into(), quote_style: None, span: Span::empty(), @@ -3607,13 +3623,13 @@ fn parse_alter_role() { span: Span::empty(), }, operation: AlterRoleOperation::Set { - config_name: ObjectName(vec![Ident { + config_name: ObjectName::from(vec![Ident { value: "maintenance_work_mem".into(), quote_style: None, span: Span::empty(), }]), config_value: SetConfigValue::Default, - in_database: Some(ObjectName(vec![Ident { + in_database: Some(ObjectName::from(vec![Ident { value: "database_name".into(), quote_style: None, span: Span::empty(), @@ -3648,12 +3664,12 @@ fn parse_alter_role() { span: Span::empty(), }, operation: AlterRoleOperation::Reset { - config_name: ResetConfig::ConfigName(ObjectName(vec![Ident { + config_name: ResetConfig::ConfigName(ObjectName::from(vec![Ident { value: "maintenance_work_mem".into(), quote_style: None, span: Span::empty(), }])), - in_database: Some(ObjectName(vec![Ident { + in_database: Some(ObjectName::from(vec![Ident { value: "database_name".into(), quote_style: None, span: Span::empty(), @@ -3679,7 +3695,10 @@ fn parse_delimited_identifiers() { version, .. } => { - assert_eq!(vec![Ident::with_quote('"', "a table")], name.0); + assert_eq!( + ObjectName::from(vec![Ident::with_quote('"', "a table")]), + name + ); assert_eq!(Ident::with_quote('"', "alias"), alias.unwrap().name); assert!(args.is_none()); assert!(with_hints.is_empty()); @@ -3698,7 +3717,7 @@ fn parse_delimited_identifiers() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::with_quote('"', "myfun")]), + name: ObjectName::from(vec![Ident::with_quote('"', "myfun")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -3754,7 +3773,7 @@ fn parse_create_function() { Statement::CreateFunction(CreateFunction { or_replace: false, temporary: false, - name: ObjectName(vec![Ident::new("add")]), + name: ObjectName::from(vec![Ident::new("add")]), args: Some(vec![ OperateFunctionArg::unnamed(DataType::Integer(None)), OperateFunctionArg::unnamed(DataType::Integer(None)), @@ -3798,7 +3817,7 @@ fn parse_drop_function() { Statement::DropFunction { if_exists: true, func_desc: vec![FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_func".to_string(), quote_style: None, span: Span::empty(), @@ -3815,7 +3834,7 @@ fn parse_drop_function() { Statement::DropFunction { if_exists: true, func_desc: vec![FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_func".to_string(), quote_style: None, span: Span::empty(), @@ -3841,7 +3860,7 @@ fn parse_drop_function() { if_exists: true, func_desc: vec![ FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_func1".to_string(), quote_style: None, span: Span::empty(), @@ -3860,7 +3879,7 @@ fn parse_drop_function() { ]), }, FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_func2".to_string(), quote_style: None, span: Span::empty(), @@ -3892,7 +3911,7 @@ fn parse_drop_procedure() { Statement::DropProcedure { if_exists: true, proc_desc: vec![FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_proc".to_string(), quote_style: None, span: Span::empty(), @@ -3909,7 +3928,7 @@ fn parse_drop_procedure() { Statement::DropProcedure { if_exists: true, proc_desc: vec![FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_proc".to_string(), quote_style: None, span: Span::empty(), @@ -3935,7 +3954,7 @@ fn parse_drop_procedure() { if_exists: true, proc_desc: vec![ FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_proc1".to_string(), quote_style: None, span: Span::empty(), @@ -3954,7 +3973,7 @@ fn parse_drop_procedure() { ]), }, FunctionDesc { - name: ObjectName(vec![Ident { + name: ObjectName::from(vec![Ident { value: "test_proc2".to_string(), quote_style: None, span: Span::empty(), @@ -4136,7 +4155,7 @@ fn parse_select_group_by_cube() { #[test] fn parse_truncate() { let truncate = pg_and_generic().verified_stmt("TRUNCATE db.table_name"); - let table_name = ObjectName(vec![Ident::new("db"), Ident::new("table_name")]); + let table_name = ObjectName::from(vec![Ident::new("db"), Ident::new("table_name")]); let table_names = vec![TruncateTableTarget { name: table_name.clone(), }]; @@ -4159,7 +4178,7 @@ fn parse_truncate_with_options() { let truncate = pg_and_generic() .verified_stmt("TRUNCATE TABLE ONLY db.table_name RESTART IDENTITY CASCADE"); - let table_name = ObjectName(vec![Ident::new("db"), Ident::new("table_name")]); + let table_name = ObjectName::from(vec![Ident::new("db"), Ident::new("table_name")]); let table_names = vec![TruncateTableTarget { name: table_name.clone(), }]; @@ -4184,8 +4203,8 @@ fn parse_truncate_with_table_list() { "TRUNCATE TABLE db.table_name, db.other_table_name RESTART IDENTITY CASCADE", ); - let table_name_a = ObjectName(vec![Ident::new("db"), Ident::new("table_name")]); - let table_name_b = ObjectName(vec![Ident::new("db"), Ident::new("other_table_name")]); + let table_name_a = ObjectName::from(vec![Ident::new("db"), Ident::new("table_name")]); + let table_name_b = ObjectName::from(vec![Ident::new("db"), Ident::new("other_table_name")]); let table_names = vec![ TruncateTableTarget { @@ -4381,7 +4400,7 @@ fn test_simple_postgres_insert_with_alias() { or: None, ignore: false, into: true, - table: TableObject::TableName(ObjectName(vec![Ident { + table: TableObject::TableName(ObjectName::from(vec![Ident { value: "test_tables".to_string(), quote_style: None, span: Span::empty(), @@ -4451,7 +4470,7 @@ fn test_simple_postgres_insert_with_alias() { or: None, ignore: false, into: true, - table: TableObject::TableName(ObjectName(vec![Ident { + table: TableObject::TableName(ObjectName::from(vec![Ident { value: "test_tables".to_string(), quote_style: None, span: Span::empty(), @@ -4523,7 +4542,7 @@ fn test_simple_insert_with_quoted_alias() { or: None, ignore: false, into: true, - table: TableObject::TableName(ObjectName(vec![Ident { + table: TableObject::TableName(ObjectName::from(vec![Ident { value: "test_tables".to_string(), quote_style: None, span: Span::empty(), @@ -4720,10 +4739,10 @@ fn parse_create_simple_before_insert_trigger() { let expected = Statement::CreateTrigger { or_replace: false, is_constraint: false, - name: ObjectName(vec![Ident::new("check_insert")]), + name: ObjectName::from(vec![Ident::new("check_insert")]), period: TriggerPeriod::Before, events: vec![TriggerEvent::Insert], - table_name: ObjectName(vec![Ident::new("accounts")]), + table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], trigger_object: TriggerObject::Row, @@ -4732,7 +4751,7 @@ fn parse_create_simple_before_insert_trigger() { exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { - name: ObjectName(vec![Ident::new("check_account_insert")]), + name: ObjectName::from(vec![Ident::new("check_account_insert")]), args: None, }, }, @@ -4748,10 +4767,10 @@ fn parse_create_after_update_trigger_with_condition() { let expected = Statement::CreateTrigger { or_replace: false, is_constraint: false, - name: ObjectName(vec![Ident::new("check_update")]), + name: ObjectName::from(vec![Ident::new("check_update")]), period: TriggerPeriod::After, events: vec![TriggerEvent::Update(vec![])], - table_name: ObjectName(vec![Ident::new("accounts")]), + table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], trigger_object: TriggerObject::Row, @@ -4767,7 +4786,7 @@ fn parse_create_after_update_trigger_with_condition() { exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { - name: ObjectName(vec![Ident::new("check_account_update")]), + name: ObjectName::from(vec![Ident::new("check_account_update")]), args: None, }, }, @@ -4783,10 +4802,10 @@ fn parse_create_instead_of_delete_trigger() { let expected = Statement::CreateTrigger { or_replace: false, is_constraint: false, - name: ObjectName(vec![Ident::new("check_delete")]), + name: ObjectName::from(vec![Ident::new("check_delete")]), period: TriggerPeriod::InsteadOf, events: vec![TriggerEvent::Delete], - table_name: ObjectName(vec![Ident::new("accounts")]), + table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], trigger_object: TriggerObject::Row, @@ -4795,7 +4814,7 @@ fn parse_create_instead_of_delete_trigger() { exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { - name: ObjectName(vec![Ident::new("check_account_deletes")]), + name: ObjectName::from(vec![Ident::new("check_account_deletes")]), args: None, }, }, @@ -4811,14 +4830,14 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { let expected = Statement::CreateTrigger { or_replace: false, is_constraint: true, - name: ObjectName(vec![Ident::new("check_multiple_events")]), + name: ObjectName::from(vec![Ident::new("check_multiple_events")]), period: TriggerPeriod::Before, events: vec![ TriggerEvent::Insert, TriggerEvent::Update(vec![]), TriggerEvent::Delete, ], - table_name: ObjectName(vec![Ident::new("accounts")]), + table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], trigger_object: TriggerObject::Row, @@ -4827,7 +4846,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { - name: ObjectName(vec![Ident::new("check_account_changes")]), + name: ObjectName::from(vec![Ident::new("check_account_changes")]), args: None, }, }, @@ -4847,21 +4866,21 @@ fn parse_create_trigger_with_referencing() { let expected = Statement::CreateTrigger { or_replace: false, is_constraint: false, - name: ObjectName(vec![Ident::new("check_referencing")]), + name: ObjectName::from(vec![Ident::new("check_referencing")]), period: TriggerPeriod::Before, events: vec![TriggerEvent::Insert], - table_name: ObjectName(vec![Ident::new("accounts")]), + table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![ TriggerReferencing { refer_type: TriggerReferencingType::NewTable, is_as: true, - transition_relation_name: ObjectName(vec![Ident::new("new_accounts")]), + transition_relation_name: ObjectName::from(vec![Ident::new("new_accounts")]), }, TriggerReferencing { refer_type: TriggerReferencingType::OldTable, is_as: true, - transition_relation_name: ObjectName(vec![Ident::new("old_accounts")]), + transition_relation_name: ObjectName::from(vec![Ident::new("old_accounts")]), }, ], trigger_object: TriggerObject::Row, @@ -4870,7 +4889,7 @@ fn parse_create_trigger_with_referencing() { exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { - name: ObjectName(vec![Ident::new("check_account_referencing")]), + name: ObjectName::from(vec![Ident::new("check_account_referencing")]), args: None, }, }, @@ -4929,8 +4948,8 @@ fn parse_drop_trigger() { pg().verified_stmt(sql), Statement::DropTrigger { if_exists, - trigger_name: ObjectName(vec![Ident::new("check_update")]), - table_name: ObjectName(vec![Ident::new("table_name")]), + trigger_name: ObjectName::from(vec![Ident::new("check_update")]), + table_name: ObjectName::from(vec![Ident::new("table_name")]), option } ); @@ -5044,7 +5063,7 @@ fn parse_trigger_related_functions() { transient: false, volatile: false, iceberg: false, - name: ObjectName(vec![Ident::new("emp")]), + name: ObjectName::from(vec![Ident::new("emp")]), columns: vec![ ColumnDef { name: "empname".into(), @@ -5126,7 +5145,7 @@ fn parse_trigger_related_functions() { or_replace: false, temporary: false, if_not_exists: false, - name: ObjectName(vec![Ident::new("emp_stamp")]), + name: ObjectName::from(vec![Ident::new("emp_stamp")]), args: None, return_type: Some(DataType::Trigger), function_body: Some( @@ -5161,10 +5180,10 @@ fn parse_trigger_related_functions() { Statement::CreateTrigger { or_replace: false, is_constraint: false, - name: ObjectName(vec![Ident::new("emp_stamp")]), + name: ObjectName::from(vec![Ident::new("emp_stamp")]), period: TriggerPeriod::Before, events: vec![TriggerEvent::Insert, TriggerEvent::Update(vec![])], - table_name: ObjectName(vec![Ident::new("emp")]), + table_name: ObjectName::from(vec![Ident::new("emp")]), referenced_table_name: None, referencing: vec![], trigger_object: TriggerObject::Row, @@ -5173,7 +5192,7 @@ fn parse_trigger_related_functions() { exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { - name: ObjectName(vec![Ident::new("emp_stamp")]), + name: ObjectName::from(vec![Ident::new("emp_stamp")]), args: None, } }, @@ -5186,8 +5205,8 @@ fn parse_trigger_related_functions() { drop_trigger, Statement::DropTrigger { if_exists: false, - trigger_name: ObjectName(vec![Ident::new("emp_stamp")]), - table_name: ObjectName(vec![Ident::new("emp")]), + trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), + table_name: ObjectName::from(vec![Ident::new("emp")]), option: None } ); diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index 857d378bc..c4b897f01 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -39,7 +39,7 @@ fn test_square_brackets_over_db_schema_table_name() { assert_eq!( select.from[0], TableWithJoins { - relation: table_from_name(ObjectName(vec![ + relation: table_from_name(ObjectName::from(vec![ Ident { value: "test_schema".to_string(), quote_style: Some('['), @@ -81,7 +81,7 @@ fn test_double_quotes_over_db_schema_table_name() { assert_eq!( select.from[0], TableWithJoins { - relation: table_from_name(ObjectName(vec![ + relation: table_from_name(ObjectName::from(vec![ Ident { value: "test_schema".to_string(), quote_style: Some('"'), @@ -114,7 +114,10 @@ fn parse_delimited_identifiers() { version, .. } => { - assert_eq!(vec![Ident::with_quote('"', "a table")], name.0); + assert_eq!( + ObjectName::from(vec![Ident::with_quote('"', "a table")]), + name + ); assert_eq!(Ident::with_quote('"', "alias"), alias.unwrap().name); assert!(args.is_none()); assert!(with_hints.is_empty()); @@ -133,7 +136,7 @@ fn parse_delimited_identifiers() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::with_quote('"', "myfun")]), + name: ObjectName::from(vec![Ident::with_quote('"', "myfun")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -297,7 +300,7 @@ fn test_parse_json_path_from() { TableFactor::Table { name, json_path, .. } => { - assert_eq!(name, &ObjectName(vec![Ident::new("src")])); + assert_eq!(name, &ObjectName::from(vec![Ident::new("src")])); assert_eq!( json_path, &Some(JsonPath { @@ -321,7 +324,7 @@ fn test_parse_json_path_from() { TableFactor::Table { name, json_path, .. } => { - assert_eq!(name, &ObjectName(vec![Ident::new("src")])); + assert_eq!(name, &ObjectName::from(vec![Ident::new("src")])); assert_eq!( json_path, &Some(JsonPath { @@ -354,7 +357,7 @@ fn test_parse_json_path_from() { } => { assert_eq!( name, - &ObjectName(vec![Ident::new("src"), Ident::new("a"), Ident::new("b")]) + &ObjectName::from(vec![Ident::new("src"), Ident::new("a"), Ident::new("b")]) ); assert_eq!(json_path, &None); } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 3320400e9..2b2350936 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -634,7 +634,7 @@ fn test_snowflake_create_table_with_collated_column() { vec![ColumnDef { name: "a".into(), data_type: DataType::Text, - collation: Some(ObjectName(vec![Ident::with_quote('\'', "de_DE")])), + collation: Some(ObjectName::from(vec![Ident::with_quote('\'', "de_DE")])), options: vec![] },] ); @@ -818,7 +818,7 @@ fn test_snowflake_create_table_with_several_column_options() { ColumnDef { name: "b".into(), data_type: DataType::Text, - collation: Some(ObjectName(vec![Ident::with_quote('\'', "de_DE")])), + collation: Some(ObjectName::from(vec![Ident::with_quote('\'', "de_DE")])), options: vec![ ColumnOptionDef { name: None, @@ -1274,7 +1274,10 @@ fn parse_delimited_identifiers() { version, .. } => { - assert_eq!(vec![Ident::with_quote('"', "a table")], name.0); + assert_eq!( + ObjectName::from(vec![Ident::with_quote('"', "a table")]), + name + ); assert_eq!(Ident::with_quote('"', "alias"), alias.unwrap().name); assert!(args.is_none()); assert!(with_hints.is_empty()); @@ -1293,7 +1296,7 @@ fn parse_delimited_identifiers() { ); assert_eq!( &Expr::Function(Function { - name: ObjectName(vec![Ident::with_quote('"', "myfun")]), + name: ObjectName::from(vec![Ident::with_quote('"', "myfun")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -1365,7 +1368,7 @@ fn test_select_wildcard_with_exclude() { let select = snowflake_and_generic() .verified_only_select("SELECT name.* EXCLUDE department_id FROM employee_table"); let expected = SelectItem::QualifiedWildcard( - ObjectName(vec![Ident::new("name")]), + ObjectName::from(vec![Ident::new("name")]), WildcardAdditionalOptions { opt_exclude: Some(ExcludeSelectItem::Single(Ident::new("department_id"))), ..Default::default() @@ -1402,7 +1405,7 @@ fn test_select_wildcard_with_rename() { "SELECT name.* RENAME (department_id AS new_dep, employee_id AS new_emp) FROM employee_table", ); let expected = SelectItem::QualifiedWildcard( - ObjectName(vec![Ident::new("name")]), + ObjectName::from(vec![Ident::new("name")]), WildcardAdditionalOptions { opt_rename: Some(RenameSelectItem::Multiple(vec![ IdentWithAlias { @@ -1505,7 +1508,7 @@ fn test_alter_table_clustering() { Expr::Identifier(Ident::new("c1")), Expr::Identifier(Ident::with_quote('"', "c2")), Expr::Function(Function { - name: ObjectName(vec![Ident::new("TO_DATE")]), + name: ObjectName::from(vec![Ident::new("TO_DATE")]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -2034,11 +2037,11 @@ fn test_copy_into() { } => { assert_eq!( into, - ObjectName(vec![Ident::new("my_company"), Ident::new("emp_basic")]) + ObjectName::from(vec![Ident::new("my_company"), Ident::new("emp_basic")]) ); assert_eq!( from_stage, - ObjectName(vec![Ident::with_quote('\'', "gcs://mybucket/./../a.csv")]) + ObjectName::from(vec![Ident::with_quote('\'', "gcs://mybucket/./../a.csv")]) ); assert!(files.is_none()); assert!(pattern.is_none()); @@ -2069,7 +2072,7 @@ fn test_copy_into_with_stage_params() { //assert_eq!("s3://load/files/", stage_params.url.unwrap()); assert_eq!( from_stage, - ObjectName(vec![Ident::with_quote('\'', "s3://load/files/")]) + ObjectName::from(vec![Ident::with_quote('\'', "s3://load/files/")]) ); assert_eq!("myint", stage_params.storage_integration.unwrap()); assert_eq!( @@ -2128,7 +2131,7 @@ fn test_copy_into_with_stage_params() { } => { assert_eq!( from_stage, - ObjectName(vec![Ident::with_quote('\'', "s3://load/files/")]) + ObjectName::from(vec![Ident::with_quote('\'', "s3://load/files/")]) ); assert_eq!("myint", stage_params.storage_integration.unwrap()); } @@ -2182,7 +2185,7 @@ fn test_copy_into_with_transformations() { } => { assert_eq!( from_stage, - ObjectName(vec![Ident::new("@schema"), Ident::new("general_finished")]) + ObjectName::from(vec![Ident::new("@schema"), Ident::new("general_finished")]) ); assert_eq!( from_transformations.as_ref().unwrap()[0], @@ -2291,17 +2294,17 @@ fn test_snowflake_stage_object_names() { "@~/path", ]; let mut allowed_object_names = [ - ObjectName(vec![Ident::new("my_company"), Ident::new("emp_basic")]), - ObjectName(vec![Ident::new("@namespace"), Ident::new("%table_name")]), - ObjectName(vec![ + ObjectName::from(vec![Ident::new("my_company"), Ident::new("emp_basic")]), + ObjectName::from(vec![Ident::new("@namespace"), Ident::new("%table_name")]), + ObjectName::from(vec![ Ident::new("@namespace"), Ident::new("%table_name/path"), ]), - ObjectName(vec![ + ObjectName::from(vec![ Ident::new("@namespace"), Ident::new("stage_name/path"), ]), - ObjectName(vec![Ident::new("@~/path")]), + ObjectName::from(vec![Ident::new("@~/path")]), ]; for it in allowed_formatted_names @@ -2330,10 +2333,13 @@ fn test_snowflake_copy_into() { Statement::CopyIntoSnowflake { into, from_stage, .. } => { - assert_eq!(into, ObjectName(vec![Ident::new("a"), Ident::new("b")])); + assert_eq!( + into, + ObjectName::from(vec![Ident::new("a"), Ident::new("b")]) + ); assert_eq!( from_stage, - ObjectName(vec![Ident::new("@namespace"), Ident::new("stage_name")]) + ObjectName::from(vec![Ident::new("@namespace"), Ident::new("stage_name")]) ) } _ => unreachable!(), @@ -2350,14 +2356,14 @@ fn test_snowflake_copy_into_stage_name_ends_with_parens() { } => { assert_eq!( into, - ObjectName(vec![ + ObjectName::from(vec![ Ident::new("SCHEMA"), Ident::new("SOME_MONITORING_SYSTEM") ]) ); assert_eq!( from_stage, - ObjectName(vec![Ident::new("@schema"), Ident::new("general_finished")]) + ObjectName::from(vec![Ident::new("@schema"), Ident::new("general_finished")]) ) } _ => unreachable!(), @@ -2771,7 +2777,7 @@ fn parse_use() { // Test single identifier without quotes assert_eq!( snowflake().verified_stmt(&format!("USE {}", object_name)), - Statement::Use(Use::Object(ObjectName(vec![Ident::new( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) ); @@ -2779,7 +2785,7 @@ fn parse_use() { // Test single identifier with different type of quotes assert_eq!( snowflake().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), - Statement::Use(Use::Object(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), )]))) @@ -2791,7 +2797,7 @@ fn parse_use() { // Test double identifier with different type of quotes assert_eq!( snowflake().verified_stmt(&format!("USE {0}CATALOG{0}.{0}my_schema{0}", quote)), - Statement::Use(Use::Object(ObjectName(vec![ + Statement::Use(Use::Object(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") ]))) @@ -2800,7 +2806,7 @@ fn parse_use() { // Test double identifier without quotes assert_eq!( snowflake().verified_stmt("USE mydb.my_schema"), - Statement::Use(Use::Object(ObjectName(vec![ + Statement::Use(Use::Object(ObjectName::from(vec![ Ident::new("mydb"), Ident::new("my_schema") ]))) @@ -2810,35 +2816,35 @@ fn parse_use() { // Test single and double identifier with keyword and different type of quotes assert_eq!( snowflake().verified_stmt(&format!("USE DATABASE {0}my_database{0}", quote)), - Statement::Use(Use::Database(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Database(ObjectName::from(vec![Ident::with_quote( quote, "my_database".to_string(), )]))) ); assert_eq!( snowflake().verified_stmt(&format!("USE SCHEMA {0}my_schema{0}", quote)), - Statement::Use(Use::Schema(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Schema(ObjectName::from(vec![Ident::with_quote( quote, "my_schema".to_string(), )]))) ); assert_eq!( snowflake().verified_stmt(&format!("USE SCHEMA {0}CATALOG{0}.{0}my_schema{0}", quote)), - Statement::Use(Use::Schema(ObjectName(vec![ + Statement::Use(Use::Schema(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") ]))) ); assert_eq!( snowflake().verified_stmt(&format!("USE ROLE {0}my_role{0}", quote)), - Statement::Use(Use::Role(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Role(ObjectName::from(vec![Ident::with_quote( quote, "my_role".to_string(), )]))) ); assert_eq!( snowflake().verified_stmt(&format!("USE WAREHOUSE {0}my_wh{0}", quote)), - Statement::Use(Use::Warehouse(ObjectName(vec![Ident::with_quote( + Statement::Use(Use::Warehouse(ObjectName::from(vec![Ident::with_quote( quote, "my_wh".to_string(), )]))) @@ -3076,7 +3082,7 @@ fn parse_ls_and_rm() { .verified_stmt("LIST @SNOWFLAKE_KAFKA_CONNECTOR_externalDataLakeSnowflakeConnector_STAGE_call_tracker_stream/"); match statement { Statement::List(command) => { - assert_eq!(command.stage, ObjectName(vec!["@SNOWFLAKE_KAFKA_CONNECTOR_externalDataLakeSnowflakeConnector_STAGE_call_tracker_stream/".into()])); + assert_eq!(command.stage, ObjectName::from(vec!["@SNOWFLAKE_KAFKA_CONNECTOR_externalDataLakeSnowflakeConnector_STAGE_call_tracker_stream/".into()])); assert!(command.pattern.is_none()); } _ => unreachable!(), @@ -3088,7 +3094,7 @@ fn parse_ls_and_rm() { Statement::Remove(command) => { assert_eq!( command.stage, - ObjectName(vec!["@my_csv_stage/analysis/".into()]) + ObjectName::from(vec!["@my_csv_stage/analysis/".into()]) ); assert_eq!(command.pattern, Some(".*data_0.*".to_string())); } diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index edd1365f4..3a612f70a 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -418,7 +418,7 @@ fn parse_window_function_with_filter() { assert_eq!( select.projection, vec![SelectItem::UnnamedExpr(Expr::Function(Function { - name: ObjectName(vec![Ident::new(func_name)]), + name: ObjectName::from(vec![Ident::new(func_name)]), uses_odbc_syntax: false, parameters: FunctionArguments::None, args: FunctionArguments::List(FunctionArgumentList { @@ -469,8 +469,8 @@ fn parse_update_tuple_row_values() { or: None, assignments: vec![Assignment { target: AssignmentTarget::Tuple(vec![ - ObjectName(vec![Ident::new("a"),]), - ObjectName(vec![Ident::new("b"),]), + ObjectName::from(vec![Ident::new("a"),]), + ObjectName::from(vec![Ident::new("b"),]), ]), value: Expr::Tuple(vec![ Expr::Value(Value::Number("1".parse().unwrap(), false)), @@ -479,7 +479,7 @@ fn parse_update_tuple_row_values() { }], selection: None, table: TableWithJoins { - relation: table_from_name(ObjectName(vec![Ident::new("x")])), + relation: table_from_name(ObjectName::from(vec![Ident::new("x")])), joins: vec![], }, from: None, From cbe59a2d8b08cf01469f48e8f65bf74bf286f882 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sun, 26 Jan 2025 15:15:36 +0100 Subject: [PATCH 016/286] Enable GROUP BY exp for Snowflake dialect (#1683) --- src/dialect/snowflake.rs | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index bd9afb191..eb8ea4de9 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -305,6 +305,11 @@ impl Dialect for SnowflakeDialect { fn supports_timestamp_versioning(&self) -> bool { true } + + /// See: + fn supports_group_by_expr(&self) -> bool { + true + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { From 74163b148ed984cb73146699998f615f7b22a642 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sun, 26 Jan 2025 15:20:00 +0100 Subject: [PATCH 017/286] Add support for parsing empty dictionary expressions (#1684) --- src/parser/mod.rs | 3 ++- tests/sqlparser_common.rs | 2 ++ 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9cc8f0620..0d2973c7f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2854,7 +2854,8 @@ impl<'a> Parser<'a> { fn parse_duckdb_struct_literal(&mut self) -> Result { self.expect_token(&Token::LBrace)?; - let fields = self.parse_comma_separated(Self::parse_duckdb_dictionary_field)?; + let fields = + self.parse_comma_separated0(Self::parse_duckdb_dictionary_field, Token::RBrace)?; self.expect_token(&Token::RBrace)?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 6897d44ae..5a1e812d0 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11291,6 +11291,8 @@ fn test_dictionary_syntax() { ); } + check("{}", Expr::Dictionary(vec![])); + check( "{'Alberta': 'Edmonton', 'Manitoba': 'Winnipeg'}", Expr::Dictionary(vec![ From fdbe864d0d507068fcd666cebc4507e368503d9d Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Tue, 28 Jan 2025 08:45:14 +0100 Subject: [PATCH 018/286] Support multiple tables in `UPDATE FROM` clause (#1681) --- src/ast/mod.rs | 4 ++-- src/ast/query.rs | 4 ++-- src/ast/spans.rs | 9 +++++---- src/parser/mod.rs | 6 ++++-- tests/sqlparser_common.rs | 11 +++++++---- 5 files changed, 20 insertions(+), 14 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index b473dc11f..e64b7d3db 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3872,13 +3872,13 @@ impl fmt::Display for Statement { } write!(f, "{table}")?; if let Some(UpdateTableFromKind::BeforeSet(from)) = from { - write!(f, " FROM {from}")?; + write!(f, " FROM {}", display_comma_separated(from))?; } if !assignments.is_empty() { write!(f, " SET {}", display_comma_separated(assignments))?; } if let Some(UpdateTableFromKind::AfterSet(from)) = from { - write!(f, " FROM {from}")?; + write!(f, " FROM {}", display_comma_separated(from))?; } if let Some(selection) = selection { write!(f, " WHERE {selection}")?; diff --git a/src/ast/query.rs b/src/ast/query.rs index 4053dd239..e982c7f0d 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2829,8 +2829,8 @@ impl fmt::Display for ValueTableMode { pub enum UpdateTableFromKind { /// Update Statement where the 'FROM' clause is before the 'SET' keyword (Supported by Snowflake) /// For Example: `UPDATE FROM t1 SET t1.name='aaa'` - BeforeSet(TableWithJoins), + BeforeSet(Vec), /// Update Statement where the 'FROM' clause is after the 'SET' keyword (Which is the standard way) /// For Example: `UPDATE SET t1.name='aaa' FROM t1` - AfterSet(TableWithJoins), + AfterSet(Vec), } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 5316bfbda..aed1c6c2b 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2138,10 +2138,11 @@ impl Spanned for SelectInto { impl Spanned for UpdateTableFromKind { fn span(&self) -> Span { - match self { - UpdateTableFromKind::BeforeSet(from) => from.span(), - UpdateTableFromKind::AfterSet(from) => from.span(), - } + let from = match self { + UpdateTableFromKind::BeforeSet(from) => from, + UpdateTableFromKind::AfterSet(from) => from, + }; + union_spans(from.iter().map(|t| t.span())) } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0d2973c7f..c6e1eb196 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12617,7 +12617,7 @@ impl<'a> Parser<'a> { let table = self.parse_table_and_joins()?; let from_before_set = if self.parse_keyword(Keyword::FROM) { Some(UpdateTableFromKind::BeforeSet( - self.parse_table_and_joins()?, + self.parse_table_with_joins()?, )) } else { None @@ -12625,7 +12625,9 @@ impl<'a> Parser<'a> { self.expect_keyword(Keyword::SET)?; let assignments = self.parse_comma_separated(Parser::parse_assignment)?; let from = if from_before_set.is_none() && self.parse_keyword(Keyword::FROM) { - Some(UpdateTableFromKind::AfterSet(self.parse_table_and_joins()?)) + Some(UpdateTableFromKind::AfterSet( + self.parse_table_with_joins()?, + )) } else { from_before_set }; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5a1e812d0..5c11b2901 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -403,7 +403,7 @@ fn parse_update_set_from() { target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new("name")])), value: Expr::CompoundIdentifier(vec![Ident::new("t2"), Ident::new("name")]) }], - from: Some(UpdateTableFromKind::AfterSet(TableWithJoins { + from: Some(UpdateTableFromKind::AfterSet(vec![TableWithJoins { relation: TableFactor::Derived { lateral: false, subquery: Box::new(Query { @@ -455,7 +455,7 @@ fn parse_update_set_from() { }) }, joins: vec![] - })), + }])), selection: Some(Expr::BinaryOp { left: Box::new(Expr::CompoundIdentifier(vec![ Ident::new("t1"), @@ -471,6 +471,9 @@ fn parse_update_set_from() { or: None, } ); + + let sql = "UPDATE T SET a = b FROM U, (SELECT foo FROM V) AS W WHERE 1 = 1"; + dialects.verified_stmt(sql); } #[test] @@ -13051,8 +13054,8 @@ fn parse_select_without_projection() { #[test] fn parse_update_from_before_select() { - all_dialects() - .verified_stmt("UPDATE t1 FROM (SELECT name, id FROM t1 GROUP BY id) AS t2 SET name = t2.name WHERE t1.id = t2.id"); + verified_stmt("UPDATE t1 FROM (SELECT name, id FROM t1 GROUP BY id) AS t2 SET name = t2.name WHERE t1.id = t2.id"); + verified_stmt("UPDATE t1 FROM U, (SELECT id FROM V) AS W SET a = b WHERE 1 = 1"); let query = "UPDATE t1 FROM (SELECT name, id FROM t1 GROUP BY id) AS t2 SET name = t2.name FROM (SELECT name from t2) AS t2"; From f7b0812b01111c678c595f6f79b8d2f5cf5cb305 Mon Sep 17 00:00:00 2001 From: AvivDavid-Satori <107786696+AvivDavid-Satori@users.noreply.github.com> Date: Tue, 28 Jan 2025 10:41:03 +0200 Subject: [PATCH 019/286] Add support for mysql table hints (#1675) --- src/ast/mod.rs | 10 ++--- src/ast/query.rs | 82 +++++++++++++++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/dialect/mod.rs | 4 ++ src/dialect/mysql.rs | 14 +++++++ src/parser/mod.rs | 67 ++++++++++++++++++++++++++++++ src/test_utils.rs | 3 ++ tests/sqlparser_bigquery.rs | 3 ++ tests/sqlparser_common.rs | 78 +++++++++++++++++++++++++++++++++++ tests/sqlparser_hive.rs | 1 + tests/sqlparser_mssql.rs | 6 +++ tests/sqlparser_mysql.rs | 2 + 12 files changed, 266 insertions(+), 5 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index e64b7d3db..6917b7c96 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -69,11 +69,11 @@ pub use self::query::{ OrderBy, OrderByExpr, PivotValueSource, ProjectionSelect, Query, RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, Select, SelectInto, SelectItem, SetExpr, SetOperator, SetQuantifier, Setting, SymbolDefinition, Table, - TableAlias, TableAliasColumnDef, TableFactor, TableFunctionArgs, TableSample, - TableSampleBucket, TableSampleKind, TableSampleMethod, TableSampleModifier, - TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, TableSampleUnit, TableVersion, - TableWithJoins, Top, TopQuantity, UpdateTableFromKind, ValueTableMode, Values, - WildcardAdditionalOptions, With, WithFill, + TableAlias, TableAliasColumnDef, TableFactor, TableFunctionArgs, TableIndexHintForClause, + TableIndexHintType, TableIndexHints, TableIndexType, TableSample, TableSampleBucket, + TableSampleKind, TableSampleMethod, TableSampleModifier, TableSampleQuantity, TableSampleSeed, + TableSampleSeedModifier, TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, + UpdateTableFromKind, ValueTableMode, Values, WildcardAdditionalOptions, With, WithFill, }; pub use self::trigger::{ diff --git a/src/ast/query.rs b/src/ast/query.rs index e982c7f0d..09058f760 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -975,6 +975,81 @@ pub struct TableFunctionArgs { pub settings: Option>, } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum TableIndexHintType { + Use, + Ignore, + Force, +} + +impl fmt::Display for TableIndexHintType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.write_str(match self { + TableIndexHintType::Use => "USE", + TableIndexHintType::Ignore => "IGNORE", + TableIndexHintType::Force => "FORCE", + }) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum TableIndexType { + Index, + Key, +} + +impl fmt::Display for TableIndexType { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.write_str(match self { + TableIndexType::Index => "INDEX", + TableIndexType::Key => "KEY", + }) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum TableIndexHintForClause { + Join, + OrderBy, + GroupBy, +} + +impl fmt::Display for TableIndexHintForClause { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.write_str(match self { + TableIndexHintForClause::Join => "JOIN", + TableIndexHintForClause::OrderBy => "ORDER BY", + TableIndexHintForClause::GroupBy => "GROUP BY", + }) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct TableIndexHints { + pub hint_type: TableIndexHintType, + pub index_type: TableIndexType, + pub for_clause: Option, + pub index_names: Vec, +} + +impl fmt::Display for TableIndexHints { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{} {} ", self.hint_type, self.index_type)?; + if let Some(for_clause) = &self.for_clause { + write!(f, "FOR {} ", for_clause)?; + } + write!(f, "({})", display_comma_separated(&self.index_names)) + } +} + /// A table name or a parenthesized subquery with an optional alias #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -1009,6 +1084,9 @@ pub enum TableFactor { /// Optional table sample modifier /// See: sample: Option, + /// Optional index hints(mysql) + /// See: + index_hints: Vec, }, Derived { lateral: bool, @@ -1590,6 +1668,7 @@ impl fmt::Display for TableFactor { with_ordinality, json_path, sample, + index_hints, } => { write!(f, "{name}")?; if let Some(json_path) = json_path { @@ -1618,6 +1697,9 @@ impl fmt::Display for TableFactor { if let Some(alias) = alias { write!(f, " AS {alias}")?; } + if !index_hints.is_empty() { + write!(f, " {}", display_separated(index_hints, " "))?; + } if !with_hints.is_empty() { write!(f, " WITH ({})", display_comma_separated(with_hints))?; } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index aed1c6c2b..8f72c26f5 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1739,6 +1739,7 @@ impl Spanned for TableFactor { partitions: _, json_path: _, sample: _, + index_hints: _, } => union_spans( name.0 .iter() diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 9fc16cd56..6329c5cfc 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -854,6 +854,10 @@ pub trait Dialect: Debug + Any { fn supports_string_escape_constant(&self) -> bool { false } + /// Returns true if the dialect supports the table hints in the `FROM` clause. + fn supports_table_hints(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 535b4298a..a67fe67b0 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -25,6 +25,10 @@ use crate::{ parser::{Parser, ParserError}, }; +use super::keywords; + +const RESERVED_FOR_TABLE_ALIAS_MYSQL: &[Keyword] = &[Keyword::USE, Keyword::IGNORE, Keyword::FORCE]; + /// A [`Dialect`] for [MySQL](https://www.mysql.com/) #[derive(Debug)] pub struct MySqlDialect {} @@ -111,6 +115,16 @@ impl Dialect for MySqlDialect { fn supports_user_host_grantee(&self) -> bool { true } + + fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { + explicit + || (!keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) + && !RESERVED_FOR_TABLE_ALIAS_MYSQL.contains(kw)) + } + + fn supports_table_hints(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c6e1eb196..c8ff01f70 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8910,6 +8910,64 @@ impl<'a> Parser<'a> { } } + fn parse_table_index_hints(&mut self) -> Result, ParserError> { + let mut hints = vec![]; + while let Some(hint_type) = + self.parse_one_of_keywords(&[Keyword::USE, Keyword::IGNORE, Keyword::FORCE]) + { + let hint_type = match hint_type { + Keyword::USE => TableIndexHintType::Use, + Keyword::IGNORE => TableIndexHintType::Ignore, + Keyword::FORCE => TableIndexHintType::Force, + _ => { + return self.expected( + "expected to match USE/IGNORE/FORCE keyword", + self.peek_token(), + ) + } + }; + let index_type = match self.parse_one_of_keywords(&[Keyword::INDEX, Keyword::KEY]) { + Some(Keyword::INDEX) => TableIndexType::Index, + Some(Keyword::KEY) => TableIndexType::Key, + _ => { + return self.expected("expected to match INDEX/KEY keyword", self.peek_token()) + } + }; + let for_clause = if self.parse_keyword(Keyword::FOR) { + let clause = if self.parse_keyword(Keyword::JOIN) { + TableIndexHintForClause::Join + } else if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) { + TableIndexHintForClause::OrderBy + } else if self.parse_keywords(&[Keyword::GROUP, Keyword::BY]) { + TableIndexHintForClause::GroupBy + } else { + return self.expected( + "expected to match FOR/ORDER BY/GROUP BY table hint in for clause", + self.peek_token(), + ); + }; + Some(clause) + } else { + None + }; + + self.expect_token(&Token::LParen)?; + let index_names = if self.peek_token().token != Token::RParen { + self.parse_comma_separated(Parser::parse_identifier)? + } else { + vec![] + }; + self.expect_token(&Token::RParen)?; + hints.push(TableIndexHints { + hint_type, + index_type, + for_clause, + index_names, + }); + } + Ok(hints) + } + /// Wrapper for parse_optional_alias_inner, left for backwards-compatibility /// but new flows should use the context-specific methods such as `maybe_parse_select_item_alias` /// and `maybe_parse_table_alias`. @@ -11257,6 +11315,14 @@ impl<'a> Parser<'a> { let alias = self.maybe_parse_table_alias()?; + // MYSQL-specific table hints: + let index_hints = if self.dialect.supports_table_hints() { + self.maybe_parse(|p| p.parse_table_index_hints())? + .unwrap_or(vec![]) + } else { + vec![] + }; + // MSSQL-specific table hints: let mut with_hints = vec![]; if self.parse_keyword(Keyword::WITH) { @@ -11285,6 +11351,7 @@ impl<'a> Parser<'a> { with_ordinality, json_path, sample, + index_hints, }; while let Some(kw) = self.parse_one_of_keywords(&[Keyword::PIVOT, Keyword::UNPIVOT]) { diff --git a/src/test_utils.rs b/src/test_utils.rs index f2e3adf09..208984223 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -362,6 +362,7 @@ pub fn table(name: impl Into) -> TableFactor { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], } } @@ -376,6 +377,7 @@ pub fn table_from_name(name: ObjectName) -> TableFactor { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], } } @@ -393,6 +395,7 @@ pub fn table_with_alias(name: impl Into, alias: impl Into) -> Ta with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], } } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index cbb963761..45d87a8b8 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -1565,6 +1565,7 @@ fn parse_table_time_travel() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, joins: vec![] },] @@ -1665,6 +1666,7 @@ fn parse_merge() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, table ); @@ -1682,6 +1684,7 @@ fn parse_merge() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, source ); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5c11b2901..2489ce2d9 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -503,6 +503,7 @@ fn parse_update_with_table_alias() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, joins: vec![], }, @@ -596,6 +597,7 @@ fn parse_select_with_table_alias() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, joins: vec![], }] @@ -792,6 +794,7 @@ fn parse_where_delete_with_alias_statement() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, from[0].relation, ); @@ -810,6 +813,7 @@ fn parse_where_delete_with_alias_statement() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, joins: vec![], }]), @@ -6416,6 +6420,7 @@ fn parse_joins_on() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, global, join_operator: f(JoinConstraint::On(Expr::BinaryOp { @@ -6545,6 +6550,7 @@ fn parse_joins_using() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, global: false, join_operator: f(JoinConstraint::Using(vec![ObjectName::from(vec![ @@ -6623,6 +6629,7 @@ fn parse_natural_join() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, global: false, join_operator: f(JoinConstraint::Natural), @@ -8718,6 +8725,7 @@ fn parse_merge() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], } ); assert_eq!(table, table_no_into); @@ -9901,6 +9909,7 @@ fn parse_pivot_table() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }), aggregate_functions: vec![ expected_function("a", None), @@ -9977,6 +9986,7 @@ fn parse_unpivot_table() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }), value: Ident { value: "quantity".to_string(), @@ -10023,6 +10033,73 @@ fn parse_unpivot_table() { ); } +#[test] +fn parse_select_table_with_index_hints() { + let supported_dialects = all_dialects_where(|d| d.supports_table_hints()); + let s = supported_dialects.verified_only_select( + "SELECT * FROM t1 USE INDEX (i1) IGNORE INDEX FOR ORDER BY (i2) ORDER BY a", + ); + if let TableFactor::Table { index_hints, .. } = &s.from[0].relation { + assert_eq!( + vec![ + TableIndexHints { + hint_type: TableIndexHintType::Use, + index_names: vec!["i1".into()], + index_type: TableIndexType::Index, + for_clause: None, + }, + TableIndexHints { + hint_type: TableIndexHintType::Ignore, + index_names: vec!["i2".into()], + index_type: TableIndexType::Index, + for_clause: Some(TableIndexHintForClause::OrderBy), + }, + ], + *index_hints + ); + } else { + panic!("Expected TableFactor::Table"); + } + supported_dialects.verified_stmt("SELECT * FROM t1 USE INDEX (i1) USE INDEX (i1, i1)"); + supported_dialects.verified_stmt( + "SELECT * FROM t1 USE INDEX () IGNORE INDEX (i2) USE INDEX (i1) USE INDEX (i2)", + ); + supported_dialects.verified_stmt("SELECT * FROM t1 FORCE INDEX FOR JOIN (i2)"); + supported_dialects.verified_stmt("SELECT * FROM t1 IGNORE INDEX FOR JOIN (i2)"); + supported_dialects.verified_stmt( + "SELECT * FROM t USE INDEX (index1) IGNORE INDEX FOR ORDER BY (index1) IGNORE INDEX FOR GROUP BY (index1) WHERE A = B", + ); + + // Test that dialects that don't support table hints will keep parsing the USE as table alias + let sql = "SELECT * FROM T USE LIMIT 1"; + let unsupported_dialects = all_dialects_where(|d| !d.supports_table_hints()); + let select = unsupported_dialects + .verified_only_select_with_canonical(sql, "SELECT * FROM T AS USE LIMIT 1"); + assert_eq!( + select.from, + vec![TableWithJoins { + relation: TableFactor::Table { + name: ObjectName(vec![sqlparser::ast::ObjectNamePart::Identifier( + Ident::new("T") + )]), + alias: Some(TableAlias { + name: Ident::new("USE"), + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + joins: vec![], + }] + ); +} + #[test] fn parse_pivot_unpivot_table() { let sql = concat!( @@ -10048,6 +10125,7 @@ fn parse_pivot_unpivot_table() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }), value: Ident { value: "population".to_string(), diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 9c4e8f079..5d710b17d 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -460,6 +460,7 @@ fn parse_delimited_identifiers() { partitions: _, json_path: _, sample: _, + index_hints: _, } => { assert_eq!( ObjectName::from(vec![Ident::with_quote('"', "a table")]), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 3c4017590..9046e9e74 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -74,6 +74,7 @@ fn parse_table_time_travel() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![] }, joins: vec![] },] @@ -223,6 +224,7 @@ fn parse_mssql_openjson() { partitions: vec![], json_path: None, sample: None, + index_hints: vec![] }, joins: vec![Join { relation: TableFactor::OpenJsonTable { @@ -282,6 +284,7 @@ fn parse_mssql_openjson() { partitions: vec![], json_path: None, sample: None, + index_hints: vec![] }, joins: vec![Join { relation: TableFactor::OpenJsonTable { @@ -341,6 +344,7 @@ fn parse_mssql_openjson() { partitions: vec![], json_path: None, sample: None, + index_hints: vec![] }, joins: vec![Join { relation: TableFactor::OpenJsonTable { @@ -400,6 +404,7 @@ fn parse_mssql_openjson() { partitions: vec![], json_path: None, sample: None, + index_hints: vec![], }, joins: vec![Join { relation: TableFactor::OpenJsonTable { @@ -439,6 +444,7 @@ fn parse_mssql_openjson() { partitions: vec![], json_path: None, sample: None, + index_hints: vec![], }, joins: vec![Join { relation: TableFactor::OpenJsonTable { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index fb72436ed..501dce3e3 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2036,6 +2036,7 @@ fn parse_update_with_joins() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, joins: vec![Join { relation: TableFactor::Table { @@ -2051,6 +2052,7 @@ fn parse_update_with_joins() { with_ordinality: false, json_path: None, sample: None, + index_hints: vec![], }, global: false, join_operator: JoinOperator::Inner(JoinConstraint::On(Expr::BinaryOp { From 8de3a62948d3384b9c13a387b0039984d51752fb Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Tue, 28 Jan 2025 11:33:07 +0100 Subject: [PATCH 020/286] BigQuery: Add support for select expr star (#1680) --- src/ast/mod.rs | 13 +++--- src/ast/query.rs | 34 +++++++++++++-- src/ast/spans.rs | 18 +++++--- src/dialect/bigquery.rs | 5 +++ src/dialect/mod.rs | 11 +++++ src/parser/mod.rs | 30 ++++++++++--- tests/sqlparser_bigquery.rs | 11 +++-- tests/sqlparser_common.rs | 85 +++++++++++++++++++++++++++++++++++- tests/sqlparser_duckdb.rs | 2 +- tests/sqlparser_snowflake.rs | 4 +- 10 files changed, 183 insertions(+), 30 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 6917b7c96..d3a028b0b 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -68,12 +68,13 @@ pub use self::query::{ NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, OffsetRows, OpenJsonTableColumn, OrderBy, OrderByExpr, PivotValueSource, ProjectionSelect, Query, RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, Select, - SelectInto, SelectItem, SetExpr, SetOperator, SetQuantifier, Setting, SymbolDefinition, Table, - TableAlias, TableAliasColumnDef, TableFactor, TableFunctionArgs, TableIndexHintForClause, - TableIndexHintType, TableIndexHints, TableIndexType, TableSample, TableSampleBucket, - TableSampleKind, TableSampleMethod, TableSampleModifier, TableSampleQuantity, TableSampleSeed, - TableSampleSeedModifier, TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, - UpdateTableFromKind, ValueTableMode, Values, WildcardAdditionalOptions, With, WithFill, + SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, SetOperator, SetQuantifier, + Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, + TableFunctionArgs, TableIndexHintForClause, TableIndexHintType, TableIndexHints, + TableIndexType, TableSample, TableSampleBucket, TableSampleKind, TableSampleMethod, + TableSampleModifier, TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, + TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, UpdateTableFromKind, + ValueTableMode, Values, WildcardAdditionalOptions, With, WithFill, }; pub use self::trigger::{ diff --git a/src/ast/query.rs b/src/ast/query.rs index 09058f760..747d925c9 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -586,6 +586,20 @@ impl fmt::Display for Cte { } } +/// Represents an expression behind a wildcard expansion in a projection. +/// `SELECT T.* FROM T; +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum SelectItemQualifiedWildcardKind { + /// Expression is an object name. + /// e.g. `alias.*` or even `schema.table.*` + ObjectName(ObjectName), + /// Select star on an arbitrary expression. + /// e.g. `STRUCT('foo').*` + Expr(Expr), +} + /// One item of the comma-separated list following `SELECT` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -595,12 +609,24 @@ pub enum SelectItem { UnnamedExpr(Expr), /// An expression, followed by `[ AS ] alias` ExprWithAlias { expr: Expr, alias: Ident }, - /// `alias.*` or even `schema.table.*` - QualifiedWildcard(ObjectName, WildcardAdditionalOptions), + /// An expression, followed by a wildcard expansion. + /// e.g. `alias.*`, `STRUCT('foo').*` + QualifiedWildcard(SelectItemQualifiedWildcardKind, WildcardAdditionalOptions), /// An unqualified `*` Wildcard(WildcardAdditionalOptions), } +impl fmt::Display for SelectItemQualifiedWildcardKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match &self { + SelectItemQualifiedWildcardKind::ObjectName(object_name) => { + write!(f, "{object_name}.*") + } + SelectItemQualifiedWildcardKind::Expr(expr) => write!(f, "{expr}.*"), + } + } +} + /// Single aliased identifier /// /// # Syntax @@ -867,8 +893,8 @@ impl fmt::Display for SelectItem { match &self { SelectItem::UnnamedExpr(expr) => write!(f, "{expr}"), SelectItem::ExprWithAlias { expr, alias } => write!(f, "{expr} AS {alias}"), - SelectItem::QualifiedWildcard(prefix, additional_options) => { - write!(f, "{prefix}.*")?; + SelectItem::QualifiedWildcard(kind, additional_options) => { + write!(f, "{kind}")?; write!(f, "{additional_options}")?; Ok(()) } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 8f72c26f5..58fa27aa8 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -15,6 +15,7 @@ // specific language governing permissions and limitations // under the License. +use crate::ast::query::SelectItemQualifiedWildcardKind; use core::iter; use crate::tokenizer::Span; @@ -1623,16 +1624,23 @@ impl Spanned for JsonPathElem { } } +impl Spanned for SelectItemQualifiedWildcardKind { + fn span(&self) -> Span { + match self { + SelectItemQualifiedWildcardKind::ObjectName(object_name) => object_name.span(), + SelectItemQualifiedWildcardKind::Expr(expr) => expr.span(), + } + } +} + impl Spanned for SelectItem { fn span(&self) -> Span { match self { SelectItem::UnnamedExpr(expr) => expr.span(), SelectItem::ExprWithAlias { expr, alias } => expr.span().union(&alias.span), - SelectItem::QualifiedWildcard(object_name, wildcard_additional_options) => union_spans( - object_name - .0 - .iter() - .map(|i| i.span()) + SelectItem::QualifiedWildcard(kind, wildcard_additional_options) => union_spans( + [kind.span()] + .into_iter() .chain(iter::once(wildcard_additional_options.span())), ), SelectItem::Wildcard(wildcard_additional_options) => wildcard_additional_options.span(), diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index 716174391..b45754213 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -83,6 +83,11 @@ impl Dialect for BigQueryDialect { true } + /// See + fn supports_select_expr_star(&self) -> bool { + true + } + // See fn supports_timestamp_versioning(&self) -> bool { true diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 6329c5cfc..bc3c0c967 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -447,6 +447,17 @@ pub trait Dialect: Debug + Any { false } + /// Return true if the dialect supports wildcard expansion on + /// arbitrary expressions in projections. + /// + /// Example: + /// ```sql + /// SELECT STRUCT('foo').* FROM T + /// ``` + fn supports_select_expr_star(&self) -> bool { + false + } + /// Does the dialect support MySQL-style `'user'@'host'` grantee syntax? fn supports_user_host_grantee(&self) -> bool { false diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c8ff01f70..179c120bb 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1528,10 +1528,17 @@ impl<'a> Parser<'a> { // function array_agg traverses this control flow if dialect_of!(self is PostgreSqlDialect) { ending_wildcard = Some(next_token); - break; } else { - return self.expected("an identifier after '.'", next_token); + // Put back the consumed .* tokens before exiting. + // If this expression is being parsed in the + // context of a projection, then this could imply + // a wildcard expansion. For example: + // `SELECT STRUCT('foo').* FROM T` + self.prev_token(); // * + self.prev_token(); // . } + + break; } Token::SingleQuotedString(s) => { let expr = Expr::Identifier(Ident::with_quote('\'', s)); @@ -1568,18 +1575,18 @@ impl<'a> Parser<'a> { } else { self.parse_function(ObjectName::from(id_parts)) } + } else if chain.is_empty() { + Ok(root) } else { if Self::is_all_ident(&root, &chain) { return Ok(Expr::CompoundIdentifier(Self::exprs_to_idents( root, chain, )?)); } - if chain.is_empty() { - return Ok(root); - } + Ok(Expr::CompoundFieldAccess { root: Box::new(root), - access_chain: chain.clone(), + access_chain: chain, }) } } @@ -12935,7 +12942,7 @@ impl<'a> Parser<'a> { pub fn parse_select_item(&mut self) -> Result { match self.parse_wildcard_expr()? { Expr::QualifiedWildcard(prefix, token) => Ok(SelectItem::QualifiedWildcard( - prefix, + SelectItemQualifiedWildcardKind::ObjectName(prefix), self.parse_wildcard_additional_options(token.0)?, )), Expr::Wildcard(token) => Ok(SelectItem::Wildcard( @@ -12965,6 +12972,15 @@ impl<'a> Parser<'a> { alias, }) } + expr if self.dialect.supports_select_expr_star() + && self.consume_tokens(&[Token::Period, Token::Mul]) => + { + let wildcard_token = self.get_previous_token().clone(); + Ok(SelectItem::QualifiedWildcard( + SelectItemQualifiedWildcardKind::Expr(expr), + self.parse_wildcard_additional_options(wildcard_token)?, + )) + } expr => self .maybe_parse_select_item_alias() .map(|alias| match alias { diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 45d87a8b8..921a37a8a 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -1540,9 +1540,6 @@ fn parse_hyphenated_table_identifiers() { ])) }) ); - - let error_sql = "select foo-bar.* from foo-bar"; - assert!(bigquery().parse_sql_statements(error_sql).is_err()); } #[test] @@ -2204,6 +2201,14 @@ fn parse_extract_weekday() { ); } +#[test] +fn bigquery_select_expr_star() { + bigquery() + .verified_only_select("SELECT STRUCT((SELECT foo FROM T WHERE true)).* FROM T"); + bigquery().verified_only_select("SELECT [STRUCT('foo')][0].* EXCEPT (foo) FROM T"); + bigquery().verified_only_select("SELECT myfunc()[0].* FROM T"); +} + #[test] fn test_select_as_struct() { bigquery().verified_only_select("SELECT * FROM (SELECT AS VALUE STRUCT(123 AS a, false AS b))"); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 2489ce2d9..5dd74e1fa 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -1002,7 +1002,7 @@ fn parse_select_wildcard() { let select = verified_only_select(sql); assert_eq!( &SelectItem::QualifiedWildcard( - ObjectName::from(vec![Ident::new("foo")]), + SelectItemQualifiedWildcardKind::ObjectName(ObjectName::from(vec![Ident::new("foo")])), WildcardAdditionalOptions::default() ), only(&select.projection) @@ -1012,7 +1012,10 @@ fn parse_select_wildcard() { let select = verified_only_select(sql); assert_eq!( &SelectItem::QualifiedWildcard( - ObjectName::from(vec![Ident::new("myschema"), Ident::new("mytable"),]), + SelectItemQualifiedWildcardKind::ObjectName(ObjectName::from(vec![ + Ident::new("myschema"), + Ident::new("mytable"), + ])), WildcardAdditionalOptions::default(), ), only(&select.projection) @@ -1057,6 +1060,84 @@ fn parse_column_aliases() { one_statement_parses_to("SELECT a.col + 1 newname FROM foo AS a", sql); } +#[test] +fn parse_select_expr_star() { + let dialects = all_dialects_where(|d| d.supports_select_expr_star()); + + // Identifier wildcard expansion. + let select = dialects.verified_only_select("SELECT foo.bar.* FROM T"); + let SelectItem::QualifiedWildcard(SelectItemQualifiedWildcardKind::ObjectName(object_name), _) = + only(&select.projection) + else { + unreachable!( + "expected wildcard select item: got {:?}", + &select.projection[0] + ) + }; + assert_eq!( + &ObjectName::from( + ["foo", "bar"] + .into_iter() + .map(Ident::new) + .collect::>() + ), + object_name + ); + + // Arbitrary compound expression with wildcard expansion. + let select = dialects.verified_only_select("SELECT foo - bar.* FROM T"); + let SelectItem::QualifiedWildcard( + SelectItemQualifiedWildcardKind::Expr(Expr::BinaryOp { left, op, right }), + _, + ) = only(&select.projection) + else { + unreachable!( + "expected wildcard select item: got {:?}", + &select.projection[0] + ) + }; + let (Expr::Identifier(left), BinaryOperator::Minus, Expr::Identifier(right)) = + (left.as_ref(), op, right.as_ref()) + else { + unreachable!("expected binary op expr: got {:?}", &select.projection[0]) + }; + assert_eq!(&Ident::new("foo"), left); + assert_eq!(&Ident::new("bar"), right); + + // Arbitrary expression wildcard expansion. + let select = dialects.verified_only_select("SELECT myfunc().foo.* FROM T"); + let SelectItem::QualifiedWildcard( + SelectItemQualifiedWildcardKind::Expr(Expr::CompoundFieldAccess { root, access_chain }), + _, + ) = only(&select.projection) + else { + unreachable!("expected wildcard expr: got {:?}", &select.projection[0]) + }; + assert!(matches!(root.as_ref(), Expr::Function(_))); + assert_eq!(1, access_chain.len()); + assert!(matches!( + &access_chain[0], + AccessExpr::Dot(Expr::Identifier(_)) + )); + + dialects.one_statement_parses_to( + "SELECT 2. * 3 FROM T", + #[cfg(feature = "bigdecimal")] + "SELECT 2 * 3 FROM T", + #[cfg(not(feature = "bigdecimal"))] + "SELECT 2. * 3 FROM T", + ); + dialects.verified_only_select("SELECT myfunc().* FROM T"); + dialects.verified_only_select("SELECT myfunc().* EXCEPT (foo) FROM T"); + + // Invalid + let res = dialects.parse_sql_statements("SELECT foo.*.* FROM T"); + assert_eq!( + ParserError::ParserError("Expected: end of statement, found: .".to_string()), + res.unwrap_err() + ); +} + #[test] fn test_eof_after_as() { let res = parse_sql_statements("SELECT foo AS"); diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index aee6d654c..4289ebd1f 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -160,7 +160,7 @@ fn test_select_wildcard_with_exclude() { let select = duckdb().verified_only_select("SELECT name.* EXCLUDE department_id FROM employee_table"); let expected = SelectItem::QualifiedWildcard( - ObjectName::from(vec![Ident::new("name")]), + SelectItemQualifiedWildcardKind::ObjectName(ObjectName::from(vec![Ident::new("name")])), WildcardAdditionalOptions { opt_exclude: Some(ExcludeSelectItem::Single(Ident::new("department_id"))), ..Default::default() diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 2b2350936..f8b58dd15 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -1368,7 +1368,7 @@ fn test_select_wildcard_with_exclude() { let select = snowflake_and_generic() .verified_only_select("SELECT name.* EXCLUDE department_id FROM employee_table"); let expected = SelectItem::QualifiedWildcard( - ObjectName::from(vec![Ident::new("name")]), + SelectItemQualifiedWildcardKind::ObjectName(ObjectName::from(vec![Ident::new("name")])), WildcardAdditionalOptions { opt_exclude: Some(ExcludeSelectItem::Single(Ident::new("department_id"))), ..Default::default() @@ -1405,7 +1405,7 @@ fn test_select_wildcard_with_rename() { "SELECT name.* RENAME (department_id AS new_dep, employee_id AS new_emp) FROM employee_table", ); let expected = SelectItem::QualifiedWildcard( - ObjectName::from(vec![Ident::new("name")]), + SelectItemQualifiedWildcardKind::ObjectName(ObjectName::from(vec![Ident::new("name")])), WildcardAdditionalOptions { opt_rename: Some(RenameSelectItem::Multiple(vec![ IdentWithAlias { From 269967a6ac4f4d9799cccb6c97142823123ed2c5 Mon Sep 17 00:00:00 2001 From: Paul Grau Date: Tue, 28 Jan 2025 15:26:08 +0200 Subject: [PATCH 021/286] Support underscore separators in numbers for Clickhouse. Fixes #1659 (#1677) --- src/dialect/clickhouse.rs | 4 ++ src/dialect/mod.rs | 5 +++ src/dialect/postgresql.rs | 4 ++ src/tokenizer.rs | 74 +++++++++++++++++++++++++++++++++-- tests/sqlparser_clickhouse.rs | 15 +++++++ 5 files changed, 99 insertions(+), 3 deletions(-) diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs index 884dfcbcb..830b3da90 100644 --- a/src/dialect/clickhouse.rs +++ b/src/dialect/clickhouse.rs @@ -59,6 +59,10 @@ impl Dialect for ClickHouseDialect { true } + fn supports_numeric_literal_underscores(&self) -> bool { + true + } + // ClickHouse uses this for some FORMAT expressions in `INSERT` context, e.g. when inserting // with FORMAT JSONEachRow a raw JSON key-value expression is valid and expected. // diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index bc3c0c967..817f5f325 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -304,6 +304,11 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports numbers containing underscores, e.g. `10_000_000` + fn supports_numeric_literal_underscores(&self) -> bool { + false + } + /// Returns true if the dialects supports specifying null treatment /// as part of a window function's parameter list as opposed /// to after the parameter list. diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index d4f2a032e..5ce4250fb 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -249,6 +249,10 @@ impl Dialect for PostgreSqlDialect { fn supports_string_escape_constant(&self) -> bool { true } + + fn supports_numeric_literal_underscores(&self) -> bool { + true + } } pub fn parse_create(parser: &mut Parser) -> Option> { diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 309f09d81..7742e8fae 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -1136,12 +1136,24 @@ impl<'a> Tokenizer<'a> { } // numbers and period '0'..='9' | '.' => { - let mut s = peeking_take_while(chars, |ch| ch.is_ascii_digit()); + // Some dialects support underscore as number separator + // There can only be one at a time and it must be followed by another digit + let is_number_separator = |ch: char, next_char: Option| { + self.dialect.supports_numeric_literal_underscores() + && ch == '_' + && next_char.is_some_and(|next_ch| next_ch.is_ascii_hexdigit()) + }; + + let mut s = peeking_next_take_while(chars, |ch, next_ch| { + ch.is_ascii_digit() || is_number_separator(ch, next_ch) + }); // match binary literal that starts with 0x if s == "0" && chars.peek() == Some(&'x') { chars.next(); - let s2 = peeking_take_while(chars, |ch| ch.is_ascii_hexdigit()); + let s2 = peeking_next_take_while(chars, |ch, next_ch| { + ch.is_ascii_hexdigit() || is_number_separator(ch, next_ch) + }); return Ok(Some(Token::HexStringLiteral(s2))); } @@ -1150,7 +1162,10 @@ impl<'a> Tokenizer<'a> { s.push('.'); chars.next(); } - s += &peeking_take_while(chars, |ch| ch.is_ascii_digit()); + + s += &peeking_next_take_while(chars, |ch, next_ch| { + ch.is_ascii_digit() || is_number_separator(ch, next_ch) + }); // No number -> Token::Period if s == "." { @@ -1946,6 +1961,24 @@ fn peeking_take_while(chars: &mut State, mut predicate: impl FnMut(char) -> bool s } +/// Same as peeking_take_while, but also passes the next character to the predicate. +fn peeking_next_take_while( + chars: &mut State, + mut predicate: impl FnMut(char, Option) -> bool, +) -> String { + let mut s = String::new(); + while let Some(&ch) = chars.peek() { + let next_char = chars.peekable.clone().nth(1); + if predicate(ch, next_char) { + chars.next(); // consume + s.push(ch); + } else { + break; + } + } + s +} + fn unescape_single_quoted_string(chars: &mut State<'_>) -> Option { Unescape::new(chars).unescape() } @@ -2227,6 +2260,41 @@ mod tests { compare(expected, tokens); } + #[test] + fn tokenize_numeric_literal_underscore() { + let dialect = GenericDialect {}; + let sql = String::from("SELECT 10_000"); + let mut tokenizer = Tokenizer::new(&dialect, &sql); + let tokens = tokenizer.tokenize().unwrap(); + let expected = vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Number("10".to_string(), false), + Token::make_word("_000", None), + ]; + compare(expected, tokens); + + all_dialects_where(|dialect| dialect.supports_numeric_literal_underscores()).tokenizes_to( + "SELECT 10_000, _10_000, 10_00_, 10___0", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Number("10_000".to_string(), false), + Token::Comma, + Token::Whitespace(Whitespace::Space), + Token::make_word("_10_000", None), // leading underscore tokenizes as a word (parsed as column identifier) + Token::Comma, + Token::Whitespace(Whitespace::Space), + Token::Number("10_00".to_string(), false), + Token::make_word("_", None), // trailing underscores tokenizes as a word (syntax error in some dialects) + Token::Comma, + Token::Whitespace(Whitespace::Space), + Token::Number("10".to_string(), false), + Token::make_word("___0", None), // multiple underscores tokenizes as a word (syntax error in some dialects) + ], + ); + } + #[test] fn tokenize_select_exponent() { let sql = String::from("SELECT 1e10, 1e-10, 1e+10, 1ea, 1e-10a, 1e-10-10"); diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 0f22db389..5b0638a4b 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -1649,6 +1649,21 @@ fn parse_table_sample() { clickhouse().verified_stmt("SELECT * FROM tbl SAMPLE 1 / 10 OFFSET 1 / 2"); } +#[test] +fn parse_numbers_with_underscore() { + let canonical = if cfg!(feature = "bigdecimal") { + "SELECT 10000" + } else { + "SELECT 10_000" + }; + let select = clickhouse().verified_only_select_with_canonical("SELECT 10_000", canonical); + + assert_eq!( + select.projection, + vec![SelectItem::UnnamedExpr(Expr::Value(number("10_000")))] + ) +} + fn clickhouse() -> TestedDialects { TestedDialects::new(vec![Box::new(ClickHouseDialect {})]) } From 7980c866a3824af7e1937ffda274657b5dbae99d Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Wed, 29 Jan 2025 12:03:55 +0100 Subject: [PATCH 022/286] BigQuery: Fix column identifier reserved keywords list (#1678) --- src/dialect/bigquery.rs | 26 +++++++++++++++ src/dialect/mod.rs | 12 +++++-- src/parser/mod.rs | 65 +++++++++++++++++++++++++++---------- tests/sqlparser_bigquery.rs | 9 +++++ tests/sqlparser_common.rs | 42 ++++++++++++++++-------- 5 files changed, 120 insertions(+), 34 deletions(-) diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index b45754213..bb1a0d5ce 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -16,6 +16,28 @@ // under the License. use crate::dialect::Dialect; +use crate::keywords::Keyword; +use crate::parser::Parser; + +/// These keywords are disallowed as column identifiers. Such that +/// `SELECT 5 AS FROM T` is rejected by BigQuery. +const RESERVED_FOR_COLUMN_ALIAS: &[Keyword] = &[ + Keyword::WITH, + Keyword::SELECT, + Keyword::WHERE, + Keyword::GROUP, + Keyword::HAVING, + Keyword::ORDER, + Keyword::LATERAL, + Keyword::LIMIT, + Keyword::FETCH, + Keyword::UNION, + Keyword::EXCEPT, + Keyword::INTERSECT, + Keyword::FROM, + Keyword::INTO, + Keyword::END, +]; /// A [`Dialect`] for [Google Bigquery](https://cloud.google.com/bigquery/) #[derive(Debug, Default)] @@ -92,4 +114,8 @@ impl Dialect for BigQueryDialect { fn supports_timestamp_versioning(&self) -> bool { true } + + fn is_column_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { + !RESERVED_FOR_COLUMN_ALIAS.contains(kw) + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 817f5f325..b648869d2 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -804,7 +804,7 @@ pub trait Dialect: Debug + Any { keywords::RESERVED_FOR_IDENTIFIER.contains(&kw) } - // Returns reserved keywords when looking to parse a [TableFactor]. + /// Returns reserved keywords when looking to parse a `TableFactor`. /// See [Self::supports_from_trailing_commas] fn get_reserved_keywords_for_table_factor(&self) -> &[Keyword] { keywords::RESERVED_FOR_TABLE_FACTOR @@ -844,11 +844,17 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the specified keyword should be parsed as a column identifier. + /// See [keywords::RESERVED_FOR_COLUMN_ALIAS] + fn is_column_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { + !keywords::RESERVED_FOR_COLUMN_ALIAS.contains(kw) + } + /// Returns true if the specified keyword should be parsed as a select item alias. /// When explicit is true, the keyword is preceded by an `AS` word. Parser is provided /// to enable looking ahead if needed. - fn is_select_item_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { - explicit || !keywords::RESERVED_FOR_COLUMN_ALIAS.contains(kw) + fn is_select_item_alias(&self, explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { + explicit || self.is_column_alias(kw, parser) } /// Returns true if the specified keyword should be parsed as a table factor alias. diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 179c120bb..b09360078 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3951,7 +3951,7 @@ impl<'a> Parser<'a> { self.parse_comma_separated_with_trailing_commas( |p| p.parse_select_item(), trailing_commas, - None, + Self::is_reserved_for_column_alias, ) } @@ -3985,30 +3985,42 @@ impl<'a> Parser<'a> { self.parse_comma_separated_with_trailing_commas( Parser::parse_table_and_joins, trailing_commas, - Some(self.dialect.get_reserved_keywords_for_table_factor()), + |kw, _parser| { + self.dialect + .get_reserved_keywords_for_table_factor() + .contains(kw) + }, ) } /// Parse the comma of a comma-separated syntax element. + /// `R` is a predicate that should return true if the next + /// keyword is a reserved keyword. /// Allows for control over trailing commas + /// /// Returns true if there is a next element - fn is_parse_comma_separated_end_with_trailing_commas( + fn is_parse_comma_separated_end_with_trailing_commas( &mut self, trailing_commas: bool, - reserved_keywords: Option<&[Keyword]>, - ) -> bool { - let reserved_keywords = reserved_keywords.unwrap_or(keywords::RESERVED_FOR_COLUMN_ALIAS); + is_reserved_keyword: &R, + ) -> bool + where + R: Fn(&Keyword, &mut Parser) -> bool, + { if !self.consume_token(&Token::Comma) { true } else if trailing_commas { - let token = self.peek_token().token; - match token { - Token::Word(ref kw) if reserved_keywords.contains(&kw.keyword) => true, + let token = self.next_token().token; + let is_end = match token { + Token::Word(ref kw) if is_reserved_keyword(&kw.keyword, self) => true, Token::RParen | Token::SemiColon | Token::EOF | Token::RBracket | Token::RBrace => { true } _ => false, - } + }; + self.prev_token(); + + is_end } else { false } @@ -4017,7 +4029,10 @@ impl<'a> Parser<'a> { /// Parse the comma of a comma-separated syntax element. /// Returns true if there is a next element fn is_parse_comma_separated_end(&mut self) -> bool { - self.is_parse_comma_separated_end_with_trailing_commas(self.options.trailing_commas, None) + self.is_parse_comma_separated_end_with_trailing_commas( + self.options.trailing_commas, + &Self::is_reserved_for_column_alias, + ) } /// Parse a comma-separated list of 1+ items accepted by `F` @@ -4025,26 +4040,33 @@ impl<'a> Parser<'a> { where F: FnMut(&mut Parser<'a>) -> Result, { - self.parse_comma_separated_with_trailing_commas(f, self.options.trailing_commas, None) + self.parse_comma_separated_with_trailing_commas( + f, + self.options.trailing_commas, + Self::is_reserved_for_column_alias, + ) } - /// Parse a comma-separated list of 1+ items accepted by `F` - /// Allows for control over trailing commas - fn parse_comma_separated_with_trailing_commas( + /// Parse a comma-separated list of 1+ items accepted by `F`. + /// `R` is a predicate that should return true if the next + /// keyword is a reserved keyword. + /// Allows for control over trailing commas. + fn parse_comma_separated_with_trailing_commas( &mut self, mut f: F, trailing_commas: bool, - reserved_keywords: Option<&[Keyword]>, + is_reserved_keyword: R, ) -> Result, ParserError> where F: FnMut(&mut Parser<'a>) -> Result, + R: Fn(&Keyword, &mut Parser) -> bool, { let mut values = vec![]; loop { values.push(f(self)?); if self.is_parse_comma_separated_end_with_trailing_commas( trailing_commas, - reserved_keywords, + &is_reserved_keyword, ) { break; } @@ -4118,6 +4140,13 @@ impl<'a> Parser<'a> { self.parse_comma_separated(f) } + /// Default implementation of a predicate that returns true if + /// the specified keyword is reserved for column alias. + /// See [Dialect::is_column_alias] + fn is_reserved_for_column_alias(kw: &Keyword, parser: &mut Parser) -> bool { + !parser.dialect.is_column_alias(kw, parser) + } + /// Run a parser method `f`, reverting back to the current position if unsuccessful. /// Returns `None` if `f` returns an error pub fn maybe_parse(&mut self, f: F) -> Result, ParserError> @@ -9394,7 +9423,7 @@ impl<'a> Parser<'a> { let cols = self.parse_comma_separated_with_trailing_commas( Parser::parse_view_column, self.dialect.supports_column_definition_trailing_commas(), - None, + Self::is_reserved_for_column_alias, )?; self.expect_token(&Token::RParen)?; Ok(cols) diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 921a37a8a..853bffeef 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -213,6 +213,15 @@ fn parse_raw_literal() { ); } +#[test] +fn parse_big_query_non_reserved_column_alias() { + let sql = r#"SELECT OFFSET, EXPLAIN, ANALYZE, SORT, TOP, VIEW FROM T"#; + bigquery().verified_stmt(sql); + + let sql = r#"SELECT 1 AS OFFSET, 2 AS EXPLAIN, 3 AS ANALYZE FROM T"#; + bigquery().verified_stmt(sql); +} + #[test] fn parse_delete_statement() { let sql = "DELETE \"table\" WHERE 1"; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5dd74e1fa..119adc7e4 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -253,8 +253,13 @@ fn parse_insert_default_values() { #[test] fn parse_insert_select_returning() { - verified_stmt("INSERT INTO t SELECT 1 RETURNING 2"); - let stmt = verified_stmt("INSERT INTO t SELECT x RETURNING x AS y"); + // Dialects that support `RETURNING` as a column identifier do + // not support this syntax. + let dialects = + all_dialects_where(|d| !d.is_column_alias(&Keyword::RETURNING, &mut Parser::new(d))); + + dialects.verified_stmt("INSERT INTO t SELECT 1 RETURNING 2"); + let stmt = dialects.verified_stmt("INSERT INTO t SELECT x RETURNING x AS y"); match stmt { Statement::Insert(Insert { returning: Some(ret), @@ -6993,9 +6998,6 @@ fn parse_union_except_intersect_minus() { verified_stmt("SELECT 1 EXCEPT SELECT 2"); verified_stmt("SELECT 1 EXCEPT ALL SELECT 2"); verified_stmt("SELECT 1 EXCEPT DISTINCT SELECT 1"); - verified_stmt("SELECT 1 MINUS SELECT 2"); - verified_stmt("SELECT 1 MINUS ALL SELECT 2"); - verified_stmt("SELECT 1 MINUS DISTINCT SELECT 1"); verified_stmt("SELECT 1 INTERSECT SELECT 2"); verified_stmt("SELECT 1 INTERSECT ALL SELECT 2"); verified_stmt("SELECT 1 INTERSECT DISTINCT SELECT 1"); @@ -7014,6 +7016,13 @@ fn parse_union_except_intersect_minus() { verified_stmt("SELECT 1 AS x, 2 AS y INTERSECT BY NAME SELECT 9 AS y, 8 AS x"); verified_stmt("SELECT 1 AS x, 2 AS y INTERSECT ALL BY NAME SELECT 9 AS y, 8 AS x"); verified_stmt("SELECT 1 AS x, 2 AS y INTERSECT DISTINCT BY NAME SELECT 9 AS y, 8 AS x"); + + // Dialects that support `MINUS` as column identifier + // do not support `MINUS` as a set operator. + let dialects = all_dialects_where(|d| !d.is_column_alias(&Keyword::MINUS, &mut Parser::new(d))); + dialects.verified_stmt("SELECT 1 MINUS SELECT 2"); + dialects.verified_stmt("SELECT 1 MINUS ALL SELECT 2"); + dialects.verified_stmt("SELECT 1 MINUS DISTINCT SELECT 1"); } #[test] @@ -7690,19 +7699,26 @@ fn parse_invalid_subquery_without_parens() { #[test] fn parse_offset() { + // Dialects that support `OFFSET` as column identifiers + // don't support this syntax. + let dialects = + all_dialects_where(|d| !d.is_column_alias(&Keyword::OFFSET, &mut Parser::new(d))); + let expect = Some(Offset { value: Expr::Value(number("2")), rows: OffsetRows::Rows, }); - let ast = verified_query("SELECT foo FROM bar OFFSET 2 ROWS"); + let ast = dialects.verified_query("SELECT foo FROM bar OFFSET 2 ROWS"); assert_eq!(ast.offset, expect); - let ast = verified_query("SELECT foo FROM bar WHERE foo = 4 OFFSET 2 ROWS"); + let ast = dialects.verified_query("SELECT foo FROM bar WHERE foo = 4 OFFSET 2 ROWS"); assert_eq!(ast.offset, expect); - let ast = verified_query("SELECT foo FROM bar ORDER BY baz OFFSET 2 ROWS"); + let ast = dialects.verified_query("SELECT foo FROM bar ORDER BY baz OFFSET 2 ROWS"); assert_eq!(ast.offset, expect); - let ast = verified_query("SELECT foo FROM bar WHERE foo = 4 ORDER BY baz OFFSET 2 ROWS"); + let ast = + dialects.verified_query("SELECT foo FROM bar WHERE foo = 4 ORDER BY baz OFFSET 2 ROWS"); assert_eq!(ast.offset, expect); - let ast = verified_query("SELECT foo FROM (SELECT * FROM bar OFFSET 2 ROWS) OFFSET 2 ROWS"); + let ast = + dialects.verified_query("SELECT foo FROM (SELECT * FROM bar OFFSET 2 ROWS) OFFSET 2 ROWS"); assert_eq!(ast.offset, expect); match *ast.body { SetExpr::Select(s) => match only(s.from).relation { @@ -7713,7 +7729,7 @@ fn parse_offset() { }, _ => panic!("Test broke"), } - let ast = verified_query("SELECT 'foo' OFFSET 0 ROWS"); + let ast = dialects.verified_query("SELECT 'foo' OFFSET 0 ROWS"); assert_eq!( ast.offset, Some(Offset { @@ -7721,7 +7737,7 @@ fn parse_offset() { rows: OffsetRows::Rows, }) ); - let ast = verified_query("SELECT 'foo' OFFSET 1 ROW"); + let ast = dialects.verified_query("SELECT 'foo' OFFSET 1 ROW"); assert_eq!( ast.offset, Some(Offset { @@ -7729,7 +7745,7 @@ fn parse_offset() { rows: OffsetRows::Row, }) ); - let ast = verified_query("SELECT 'foo' OFFSET 1"); + let ast = dialects.verified_query("SELECT 'foo' OFFSET 1"); assert_eq!( ast.offset, Some(Offset { From 784605c9132d7e14f9e3b396c26d63af086ff885 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 29 Jan 2025 17:50:36 +0100 Subject: [PATCH 023/286] Fix bug when parsing a Snowflake stage with `;` suffix (#1688) --- src/dialect/snowflake.rs | 2 +- tests/sqlparser_snowflake.rs | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index eb8ea4de9..d775ffc36 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -625,7 +625,7 @@ pub fn parse_stage_name_identifier(parser: &mut Parser) -> Result break, + Token::Whitespace(_) | Token::SemiColon => break, Token::Period => { parser.prev_token(); break; diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index f8b58dd15..1310b984f 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3102,6 +3102,10 @@ fn parse_ls_and_rm() { }; snowflake().verified_stmt(r#"LIST @"STAGE_WITH_QUOTES""#); + // Semi-colon after stage name - should terminate the stage name + snowflake() + .parse_sql_statements("LIST @db1.schema1.stage1/dir1/;") + .unwrap(); } #[test] From 252fdbab823220e7ea29895d7c7230e49d6fb742 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Wed, 29 Jan 2025 22:15:57 -0800 Subject: [PATCH 024/286] Allow plain JOIN without turning it into INNER (#1692) --- src/ast/query.rs | 10 +++++++++- src/ast/spans.rs | 1 + src/parser/mod.rs | 8 ++++++-- src/test_utils.rs | 2 +- tests/sqlparser_bigquery.rs | 2 +- tests/sqlparser_common.rs | 31 ++++++++++++++++++++++--------- tests/sqlparser_mysql.rs | 2 +- tests/sqlparser_postgres.rs | 2 +- 8 files changed, 42 insertions(+), 16 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 747d925c9..239e14554 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2038,13 +2038,20 @@ impl fmt::Display for Join { } match &self.join_operator { - JoinOperator::Inner(constraint) => write!( + JoinOperator::Join(constraint) => write!( f, " {}JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) ), + JoinOperator::Inner(constraint) => write!( + f, + " {}INNER JOIN {}{}", + prefix(constraint), + self.relation, + suffix(constraint) + ), JoinOperator::LeftOuter(constraint) => write!( f, " {}LEFT JOIN {}{}", @@ -2128,6 +2135,7 @@ impl fmt::Display for Join { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum JoinOperator { + Join(JoinConstraint), Inner(JoinConstraint), LeftOuter(JoinConstraint), RightOuter(JoinConstraint), diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 58fa27aa8..8c46fc073 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2001,6 +2001,7 @@ impl Spanned for Join { impl Spanned for JoinOperator { fn span(&self) -> Span { match self { + JoinOperator::Join(join_constraint) => join_constraint.span(), JoinOperator::Inner(join_constraint) => join_constraint.span(), JoinOperator::LeftOuter(join_constraint) => join_constraint.span(), JoinOperator::RightOuter(join_constraint) => join_constraint.span(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b09360078..aff4c8d38 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11000,9 +11000,13 @@ impl<'a> Parser<'a> { let join_operator_type = match peek_keyword { Keyword::INNER | Keyword::JOIN => { - let _ = self.parse_keyword(Keyword::INNER); // [ INNER ] + let inner = self.parse_keyword(Keyword::INNER); // [ INNER ] self.expect_keyword_is(Keyword::JOIN)?; - JoinOperator::Inner + if inner { + JoinOperator::Inner + } else { + JoinOperator::Join + } } kw @ Keyword::LEFT | kw @ Keyword::RIGHT => { let _ = self.next_token(); // consume LEFT/RIGHT diff --git a/src/test_utils.rs b/src/test_utils.rs index 208984223..6270ac42b 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -403,7 +403,7 @@ pub fn join(relation: TableFactor) -> Join { Join { relation, global: false, - join_operator: JoinOperator::Inner(JoinConstraint::Natural), + join_operator: JoinOperator::Join(JoinConstraint::Natural), } } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 853bffeef..55de4801d 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -1602,7 +1602,7 @@ fn parse_join_constraint_unnest_alias() { with_ordinality: false, }, global: false, - join_operator: JoinOperator::Inner(JoinConstraint::On(Expr::BinaryOp { + join_operator: JoinOperator::Join(JoinConstraint::On(Expr::BinaryOp { left: Box::new(Expr::Identifier("c1".into())), op: BinaryOperator::Eq, right: Box::new(Expr::Identifier("c2".into())), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 119adc7e4..99660507d 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -6457,7 +6457,7 @@ fn parse_implicit_join() { joins: vec![Join { relation: table_from_name(ObjectName::from(vec!["t1b".into()])), global: false, - join_operator: JoinOperator::Inner(JoinConstraint::Natural), + join_operator: JoinOperator::Join(JoinConstraint::Natural), }], }, TableWithJoins { @@ -6465,7 +6465,7 @@ fn parse_implicit_join() { joins: vec![Join { relation: table_from_name(ObjectName::from(vec!["t2b".into()])), global: false, - join_operator: JoinOperator::Inner(JoinConstraint::Natural), + join_operator: JoinOperator::Join(JoinConstraint::Natural), }], }, ], @@ -6523,7 +6523,7 @@ fn parse_joins_on() { "t2", table_alias("foo"), false, - JoinOperator::Inner, + JoinOperator::Join, )] ); one_statement_parses_to( @@ -6533,7 +6533,7 @@ fn parse_joins_on() { // Test parsing of different join operators assert_eq!( only(&verified_only_select("SELECT * FROM t1 JOIN t2 ON c1 = c2").from).joins, - vec![join_with_constraint("t2", None, false, JoinOperator::Inner)] + vec![join_with_constraint("t2", None, false, JoinOperator::Join)] ); assert_eq!( only(&verified_only_select("SELECT * FROM t1 LEFT JOIN t2 ON c1 = c2").from).joins, @@ -6650,7 +6650,7 @@ fn parse_joins_using() { vec![join_with_constraint( "t2", table_alias("foo"), - JoinOperator::Inner, + JoinOperator::Join, )] ); one_statement_parses_to( @@ -6660,6 +6660,10 @@ fn parse_joins_using() { // Test parsing of different join operators assert_eq!( only(&verified_only_select("SELECT * FROM t1 JOIN t2 USING(c1)").from).joins, + vec![join_with_constraint("t2", None, JoinOperator::Join)] + ); + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 INNER JOIN t2 USING(c1)").from).joins, vec![join_with_constraint("t2", None, JoinOperator::Inner)] ); assert_eq!( @@ -6722,9 +6726,14 @@ fn parse_natural_join() { } } - // if not specified, inner join as default + // unspecified join assert_eq!( only(&verified_only_select("SELECT * FROM t1 NATURAL JOIN t2").from).joins, + vec![natural_join(JoinOperator::Join, None)] + ); + // inner join explicitly + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 NATURAL INNER JOIN t2").from).joins, vec![natural_join(JoinOperator::Inner, None)] ); // left join explicitly @@ -6748,7 +6757,7 @@ fn parse_natural_join() { // natural join another table with alias assert_eq!( only(&verified_only_select("SELECT * FROM t1 NATURAL JOIN t2 AS t3").from).joins, - vec![natural_join(JoinOperator::Inner, table_alias("t3"))] + vec![natural_join(JoinOperator::Join, table_alias("t3"))] ); let sql = "SELECT * FROM t1 natural"; @@ -6816,8 +6825,12 @@ fn parse_join_nesting() { #[test] fn parse_join_syntax_variants() { one_statement_parses_to( - "SELECT c1 FROM t1 INNER JOIN t2 USING(c1)", "SELECT c1 FROM t1 JOIN t2 USING(c1)", + "SELECT c1 FROM t1 JOIN t2 USING(c1)", + ); + one_statement_parses_to( + "SELECT c1 FROM t1 INNER JOIN t2 USING(c1)", + "SELECT c1 FROM t1 INNER JOIN t2 USING(c1)", ); one_statement_parses_to( "SELECT c1 FROM t1 LEFT OUTER JOIN t2 USING(c1)", @@ -6981,7 +6994,7 @@ fn parse_derived_tables() { joins: vec![Join { relation: table_from_name(ObjectName::from(vec!["t2".into()])), global: false, - join_operator: JoinOperator::Inner(JoinConstraint::Natural), + join_operator: JoinOperator::Join(JoinConstraint::Natural), }], }), alias: None, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 501dce3e3..2e6dfc72b 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2055,7 +2055,7 @@ fn parse_update_with_joins() { index_hints: vec![], }, global: false, - join_operator: JoinOperator::Inner(JoinConstraint::On(Expr::BinaryOp { + join_operator: JoinOperator::Join(JoinConstraint::On(Expr::BinaryOp { left: Box::new(Expr::CompoundIdentifier(vec![ Ident::new("o"), Ident::new("customer_id") diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index b3eb4f10d..9bccda11a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -4371,7 +4371,7 @@ fn parse_join_constraint_unnest_alias() { with_ordinality: false, }, global: false, - join_operator: JoinOperator::Inner(JoinConstraint::On(Expr::BinaryOp { + join_operator: JoinOperator::Join(JoinConstraint::On(Expr::BinaryOp { left: Box::new(Expr::Identifier("c1".into())), op: BinaryOperator::Eq, right: Box::new(Expr::Identifier("c2".into())), From a7e984099fdded3cbcfbf236a7b800b7ace4252f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?R=C3=A9my=20SAISSY?= Date: Thu, 30 Jan 2025 07:50:30 +0100 Subject: [PATCH 025/286] Fix DDL generation in case of an empty arguments function. (#1690) --- src/ast/ddl.rs | 2 ++ tests/sqlparser_postgres.rs | 1 + 2 files changed, 3 insertions(+) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 1b5ccda26..4fe6b0bb2 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2011,6 +2011,8 @@ impl fmt::Display for CreateFunction { )?; if let Some(args) = &self.args { write!(f, "({})", display_comma_separated(args))?; + } else { + write!(f, "()")?; } if let Some(return_type) = &self.return_type { write!(f, " RETURNS {return_type}")?; diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 9bccda11a..5a2f726af 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -3802,6 +3802,7 @@ fn parse_create_function_detailed() { pg_and_generic().verified_stmt("CREATE OR REPLACE FUNCTION add(a INTEGER, IN b INTEGER = 1) RETURNS INTEGER LANGUAGE SQL STABLE PARALLEL UNSAFE RETURN a + b"); pg_and_generic().verified_stmt("CREATE OR REPLACE FUNCTION add(a INTEGER, IN b INTEGER = 1) RETURNS INTEGER LANGUAGE SQL STABLE CALLED ON NULL INPUT PARALLEL UNSAFE RETURN a + b"); pg_and_generic().verified_stmt(r#"CREATE OR REPLACE FUNCTION increment(i INTEGER) RETURNS INTEGER LANGUAGE plpgsql AS $$ BEGIN RETURN i + 1; END; $$"#); + pg_and_generic().verified_stmt(r#"CREATE OR REPLACE FUNCTION no_arg() RETURNS VOID LANGUAGE plpgsql AS $$ BEGIN DELETE FROM my_table; END; $$"#); } #[test] fn parse_incorrect_create_function_parallel() { From 9c384a91940b4b621b02b0fee7d40b5ec2709bbd Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Thu, 30 Jan 2025 23:45:31 +0100 Subject: [PATCH 026/286] Fix `CREATE FUNCTION` round trip for Hive dialect (#1693) --- src/ast/ddl.rs | 2 -- src/parser/mod.rs | 11 +++++------ tests/sqlparser_postgres.rs | 2 +- 3 files changed, 6 insertions(+), 9 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 4fe6b0bb2..1b5ccda26 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2011,8 +2011,6 @@ impl fmt::Display for CreateFunction { )?; if let Some(args) = &self.args { write!(f, "({})", display_comma_separated(args))?; - } else { - write!(f, "()")?; } if let Some(return_type) = &self.return_type { write!(f, " RETURNS {return_type}")?; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index aff4c8d38..df5c19a38 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4553,14 +4553,13 @@ impl<'a> Parser<'a> { temporary: bool, ) -> Result { let name = self.parse_object_name(false)?; + self.expect_token(&Token::LParen)?; - let args = if self.consume_token(&Token::RParen) { - self.prev_token(); - None + let args = if Token::RParen != self.peek_token_ref().token { + self.parse_comma_separated(Parser::parse_function_arg)? } else { - Some(self.parse_comma_separated(Parser::parse_function_arg)?) + vec![] }; - self.expect_token(&Token::RParen)?; let return_type = if self.parse_keyword(Keyword::RETURNS) { @@ -4656,7 +4655,7 @@ impl<'a> Parser<'a> { or_replace, temporary, name, - args, + args: Some(args), return_type, behavior: body.behavior, called_on_null: body.called_on_null, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 5a2f726af..93af04980 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5147,7 +5147,7 @@ fn parse_trigger_related_functions() { temporary: false, if_not_exists: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), - args: None, + args: Some(vec![]), return_type: Some(DataType::Trigger), function_body: Some( CreateFunctionBody::AsBeforeOptions( From 94b2ff7191ee2d2a2134a0264461c65e9d683e72 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Fri, 31 Jan 2025 00:05:36 +0100 Subject: [PATCH 027/286] Make numeric literal underscore test dialect agnostic (#1685) --- tests/sqlparser_clickhouse.rs | 15 --------------- tests/sqlparser_common.rs | 18 ++++++++++++++++++ 2 files changed, 18 insertions(+), 15 deletions(-) diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 5b0638a4b..0f22db389 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -1649,21 +1649,6 @@ fn parse_table_sample() { clickhouse().verified_stmt("SELECT * FROM tbl SAMPLE 1 / 10 OFFSET 1 / 2"); } -#[test] -fn parse_numbers_with_underscore() { - let canonical = if cfg!(feature = "bigdecimal") { - "SELECT 10000" - } else { - "SELECT 10_000" - }; - let select = clickhouse().verified_only_select_with_canonical("SELECT 10_000", canonical); - - assert_eq!( - select.projection, - vec![SelectItem::UnnamedExpr(Expr::Value(number("10_000")))] - ) -} - fn clickhouse() -> TestedDialects { TestedDialects::new(vec![Box::new(ClickHouseDialect {})]) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 99660507d..6f6eccfdf 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -55,6 +55,24 @@ use sqlparser::ast::Expr::{Identifier, UnaryOp}; use sqlparser::ast::Value::Number; use sqlparser::test_utils::all_dialects_except; +#[test] +fn parse_numeric_literal_underscore() { + let dialects = all_dialects_where(|d| d.supports_numeric_literal_underscores()); + + let canonical = if cfg!(feature = "bigdecimal") { + "SELECT 10000" + } else { + "SELECT 10_000" + }; + + let select = dialects.verified_only_select_with_canonical("SELECT 10_000", canonical); + + assert_eq!( + select.projection, + vec![UnnamedExpr(Expr::Value(number("10_000")))] + ); +} + #[test] fn parse_insert_values() { let row = vec![ From aeaafbe6e4f44e21851e13a4d4a8de8518df2263 Mon Sep 17 00:00:00 2001 From: gstvg <28798827+gstvg@users.noreply.github.com> Date: Fri, 31 Jan 2025 02:59:16 -0300 Subject: [PATCH 028/286] Extend lambda support for ClickHouse and DuckDB dialects (#1686) --- src/ast/mod.rs | 4 ++- src/dialect/clickhouse.rs | 5 +++ src/dialect/duckdb.rs | 5 +++ tests/sqlparser_common.rs | 65 +++++++++++++++++++++++++++++++++++ tests/sqlparser_databricks.rs | 63 --------------------------------- 5 files changed, 78 insertions(+), 64 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d3a028b0b..5a1be7734 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -1045,7 +1045,9 @@ pub enum Expr { /// param -> expr | (param1, ...) -> expr /// ``` /// - /// See . + /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/functions#higher-order-functions---operator-and-lambdaparams-expr-function) + /// [Databricks](https://docs.databricks.com/en/sql/language-manual/sql-ref-lambda-functions.html) + /// [DuckDb](https://duckdb.org/docs/sql/functions/lambda.html) Lambda(LambdaFunction), } diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs index 830b3da90..9a0884a51 100644 --- a/src/dialect/clickhouse.rs +++ b/src/dialect/clickhouse.rs @@ -70,4 +70,9 @@ impl Dialect for ClickHouseDialect { fn supports_dictionary_syntax(&self) -> bool { true } + + /// See + fn supports_lambda_functions(&self) -> bool { + true + } } diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index a2699d850..c41aec81d 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -65,6 +65,11 @@ impl Dialect for DuckDbDialect { true } + /// See + fn supports_lambda_functions(&self) -> bool { + true + } + // DuckDB is compatible with PostgreSQL syntax for this statement, // although not all features may be implemented. fn supports_explain_with_utility_options(&self) -> bool { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 6f6eccfdf..a695204c2 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -13332,3 +13332,68 @@ fn test_trailing_commas_in_from() { "SELECT 1, 2 FROM (SELECT * FROM t1), (SELECT * FROM t2)", ); } + +#[test] +fn test_lambdas() { + let dialects = all_dialects_where(|d| d.supports_lambda_functions()); + + #[rustfmt::skip] + let sql = concat!( + "SELECT array_sort(array('Hello', 'World'), ", + "(p1, p2) -> CASE WHEN p1 = p2 THEN 0 ", + "WHEN reverse(p1) < reverse(p2) THEN -1 ", + "ELSE 1 END)", + ); + pretty_assertions::assert_eq!( + SelectItem::UnnamedExpr(call( + "array_sort", + [ + call( + "array", + [ + Expr::Value(Value::SingleQuotedString("Hello".to_owned())), + Expr::Value(Value::SingleQuotedString("World".to_owned())) + ] + ), + Expr::Lambda(LambdaFunction { + params: OneOrManyWithParens::Many(vec![Ident::new("p1"), Ident::new("p2")]), + body: Box::new(Expr::Case { + operand: None, + conditions: vec![ + Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("p1"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Identifier(Ident::new("p2"))) + }, + Expr::BinaryOp { + left: Box::new(call( + "reverse", + [Expr::Identifier(Ident::new("p1"))] + )), + op: BinaryOperator::Lt, + right: Box::new(call( + "reverse", + [Expr::Identifier(Ident::new("p2"))] + )) + } + ], + results: vec![ + Expr::Value(number("0")), + Expr::UnaryOp { + op: UnaryOperator::Minus, + expr: Box::new(Expr::Value(number("1"))) + } + ], + else_result: Some(Box::new(Expr::Value(number("1")))) + }) + }) + ] + )), + dialects.verified_only_select(sql).projection[0] + ); + + dialects.verified_expr( + "map_zip_with(map(1, 'a', 2, 'b'), map(1, 'x', 2, 'y'), (k, v1, v2) -> concat(v1, v2))", + ); + dialects.verified_expr("transform(array(1, 2, 3), x -> x + 1)"); +} diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index be7588de2..8338a0e71 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -83,69 +83,6 @@ fn test_databricks_exists() { ); } -#[test] -fn test_databricks_lambdas() { - #[rustfmt::skip] - let sql = concat!( - "SELECT array_sort(array('Hello', 'World'), ", - "(p1, p2) -> CASE WHEN p1 = p2 THEN 0 ", - "WHEN reverse(p1) < reverse(p2) THEN -1 ", - "ELSE 1 END)", - ); - pretty_assertions::assert_eq!( - SelectItem::UnnamedExpr(call( - "array_sort", - [ - call( - "array", - [ - Expr::Value(Value::SingleQuotedString("Hello".to_owned())), - Expr::Value(Value::SingleQuotedString("World".to_owned())) - ] - ), - Expr::Lambda(LambdaFunction { - params: OneOrManyWithParens::Many(vec![Ident::new("p1"), Ident::new("p2")]), - body: Box::new(Expr::Case { - operand: None, - conditions: vec![ - Expr::BinaryOp { - left: Box::new(Expr::Identifier(Ident::new("p1"))), - op: BinaryOperator::Eq, - right: Box::new(Expr::Identifier(Ident::new("p2"))) - }, - Expr::BinaryOp { - left: Box::new(call( - "reverse", - [Expr::Identifier(Ident::new("p1"))] - )), - op: BinaryOperator::Lt, - right: Box::new(call( - "reverse", - [Expr::Identifier(Ident::new("p2"))] - )) - } - ], - results: vec![ - Expr::Value(number("0")), - Expr::UnaryOp { - op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(number("1"))) - } - ], - else_result: Some(Box::new(Expr::Value(number("1")))) - }) - }) - ] - )), - databricks().verified_only_select(sql).projection[0] - ); - - databricks().verified_expr( - "map_zip_with(map(1, 'a', 2, 'b'), map(1, 'x', 2, 'y'), (k, v1, v2) -> concat(v1, v2))", - ); - databricks().verified_expr("transform(array(1, 2, 3), x -> x + 1)"); -} - #[test] fn test_values_clause() { let values = Values { From 447142c6d0e8e835b357833e2e9a4e9fee661f9d Mon Sep 17 00:00:00 2001 From: Paul Grau Date: Fri, 31 Jan 2025 08:04:41 +0200 Subject: [PATCH 029/286] Make TypedString preserve quote style (#1679) --- src/ast/mod.rs | 8 ++- src/ast/spans.rs | 4 +- src/ast/value.rs | 26 ++++++++++ src/parser/mod.rs | 2 +- tests/sqlparser_bigquery.rs | 100 +++++++++++++++++++++++------------- tests/sqlparser_common.rs | 45 +++++++++++----- tests/sqlparser_postgres.rs | 2 +- 7 files changed, 132 insertions(+), 55 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 5a1be7734..bccc580b3 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -898,6 +898,8 @@ pub enum Expr { /// IntroducedString { introducer: String, + /// The value of the constant. + /// Hint: you can unwrap the string value using `value.into_string()`. value: Value, }, /// A constant of form ` 'value'`. @@ -905,7 +907,9 @@ pub enum Expr { /// as well as constants of other types (a non-standard PostgreSQL extension). TypedString { data_type: DataType, - value: String, + /// The value of the constant. + /// Hint: you can unwrap the string value using `value.into_string()`. + value: Value, }, /// Scalar function call e.g. `LEFT(foo, 5)` Function(Function), @@ -1622,7 +1626,7 @@ impl fmt::Display for Expr { Expr::IntroducedString { introducer, value } => write!(f, "{introducer} {value}"), Expr::TypedString { data_type, value } => { write!(f, "{data_type}")?; - write!(f, " '{}'", &value::escape_single_quote_string(value)) + write!(f, " {value}") } Expr::Function(fun) => write!(f, "{fun}"), Expr::Method(method) => write!(f, "{method}"), diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 8c46fc073..f37c0194f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1266,7 +1266,7 @@ impl Spanned for AssignmentTarget { /// f.e. `IS NULL ` reports as `::span`. /// /// Missing spans: -/// - [Expr::TypedString] +/// - [Expr::TypedString] # missing span for data_type /// - [Expr::MatchAgainst] # MySQL specific /// - [Expr::RLike] # MySQL specific /// - [Expr::Struct] # BigQuery specific @@ -1362,7 +1362,7 @@ impl Spanned for Expr { .union(&union_spans(collation.0.iter().map(|i| i.span()))), Expr::Nested(expr) => expr.span(), Expr::Value(value) => value.span(), - Expr::TypedString { .. } => Span::empty(), + Expr::TypedString { value, .. } => value.span(), Expr::Function(function) => function.span(), Expr::GroupingSets(vec) => { union_spans(vec.iter().flat_map(|i| i.iter().map(|k| k.span()))) diff --git a/src/ast/value.rs b/src/ast/value.rs index 1b16646be..5798b5404 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -97,6 +97,32 @@ pub enum Value { Placeholder(String), } +impl Value { + /// If the underlying literal is a string, regardless of quote style, returns the associated string value + pub fn into_string(self) -> Option { + match self { + Value::SingleQuotedString(s) + | Value::DoubleQuotedString(s) + | Value::TripleSingleQuotedString(s) + | Value::TripleDoubleQuotedString(s) + | Value::SingleQuotedByteStringLiteral(s) + | Value::DoubleQuotedByteStringLiteral(s) + | Value::TripleSingleQuotedByteStringLiteral(s) + | Value::TripleDoubleQuotedByteStringLiteral(s) + | Value::SingleQuotedRawStringLiteral(s) + | Value::DoubleQuotedRawStringLiteral(s) + | Value::TripleSingleQuotedRawStringLiteral(s) + | Value::TripleDoubleQuotedRawStringLiteral(s) + | Value::EscapedStringLiteral(s) + | Value::UnicodeStringLiteral(s) + | Value::NationalStringLiteral(s) + | Value::HexStringLiteral(s) => Some(s), + Value::DollarQuotedString(s) => Some(s.value), + _ => None, + } + } +} + impl fmt::Display for Value { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index df5c19a38..ca858c42e 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1316,7 +1316,7 @@ impl<'a> Parser<'a> { DataType::Custom(..) => parser_err!("dummy", loc), data_type => Ok(Expr::TypedString { data_type, - value: parser.parse_literal_string()?, + value: parser.parse_value()?, }), } })?; diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 55de4801d..c5dfb27b5 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -39,43 +39,45 @@ fn parse_literal_string() { r#"'''triple-single'unescaped''', "#, r#""double\"escaped", "#, r#""""triple-double\"escaped""", "#, - r#""""triple-double"unescaped""""#, + r#""""triple-double"unescaped""", "#, + r#""""triple-double'unescaped""", "#, + r#"'''triple-single"unescaped'''"#, ); let dialect = TestedDialects::new_with_options( vec![Box::new(BigQueryDialect {})], ParserOptions::new().with_unescape(false), ); let select = dialect.verified_only_select(sql); - assert_eq!(10, select.projection.len()); + assert_eq!(12, select.projection.len()); assert_eq!( - &Expr::Value(Value::SingleQuotedString("single".to_string())), + &Expr::Value(Value::SingleQuotedString("single".into())), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedString("double".to_string())), + &Expr::Value(Value::DoubleQuotedString("double".into())), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedString("triple-single".to_string())), + &Expr::Value(Value::TripleSingleQuotedString("triple-single".into())), expr_from_projection(&select.projection[2]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedString("triple-double".to_string())), + &Expr::Value(Value::TripleDoubleQuotedString("triple-double".into())), expr_from_projection(&select.projection[3]) ); assert_eq!( - &Expr::Value(Value::SingleQuotedString(r#"single\'escaped"#.to_string())), + &Expr::Value(Value::SingleQuotedString(r#"single\'escaped"#.into())), expr_from_projection(&select.projection[4]) ); assert_eq!( &Expr::Value(Value::TripleSingleQuotedString( - r#"triple-single\'escaped"#.to_string() + r#"triple-single\'escaped"#.into() )), expr_from_projection(&select.projection[5]) ); assert_eq!( &Expr::Value(Value::TripleSingleQuotedString( - r#"triple-single'unescaped"#.to_string() + r#"triple-single'unescaped"#.into() )), expr_from_projection(&select.projection[6]) ); @@ -95,6 +97,18 @@ fn parse_literal_string() { )), expr_from_projection(&select.projection[9]) ); + assert_eq!( + &Expr::Value(Value::TripleDoubleQuotedString( + r#"triple-double'unescaped"#.to_string() + )), + expr_from_projection(&select.projection[10]) + ); + assert_eq!( + &Expr::Value(Value::TripleSingleQuotedString( + r#"triple-single"unescaped"#.to_string() + )), + expr_from_projection(&select.projection[11]) + ); } #[test] @@ -588,7 +602,7 @@ fn parse_tuple_struct_literal() { &Expr::Tuple(vec![ Expr::Value(number("1")), Expr::Value(number("1.0")), - Expr::Value(Value::SingleQuotedString("123".to_string())), + Expr::Value(Value::SingleQuotedString("123".into())), Expr::Value(Value::Boolean(true)) ]), expr_from_projection(&select.projection[1]) @@ -616,7 +630,7 @@ fn parse_typeless_struct_syntax() { assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("abc".to_string())),], + values: vec![Expr::Value(Value::SingleQuotedString("abc".into())),], fields: Default::default() }, expr_from_projection(&select.projection[1]) @@ -639,7 +653,7 @@ fn parse_typeless_struct_syntax() { name: Ident::from("a") }, Expr::Named { - expr: Expr::Value(Value::SingleQuotedString("abc".to_string())).into(), + expr: Expr::Value(Value::SingleQuotedString("abc".into())).into(), name: Ident::from("b") }, ], @@ -804,9 +818,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(4, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::DoubleQuotedString( - "2011-05-05".to_string() - )),], + values: vec![Expr::Value(Value::DoubleQuotedString("2011-05-05".into())),], fields: vec![StructField { field_name: None, field_type: DataType::Date @@ -818,7 +830,7 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), - value: "1999-01-01 01:23:34.45".to_string() + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) },], fields: vec![StructField { field_name: None, @@ -854,7 +866,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!( &Expr::Struct { values: vec![Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString("2".to_string()))), + value: Box::new(Expr::Value(Value::SingleQuotedString("2".into()))), leading_field: Some(DateTimeField::Hour), leading_precision: None, last_field: None, @@ -871,7 +883,9 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::JSON, - value: r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.to_string() + value: Value::SingleQuotedString( + r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() + ) },], fields: vec![StructField { field_name: None, @@ -886,7 +900,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(3, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::DoubleQuotedString("foo".to_string())),], + values: vec![Expr::Value(Value::DoubleQuotedString("foo".into())),], fields: vec![StructField { field_name: None, field_type: DataType::String(Some(42)) @@ -898,7 +912,7 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: "2008-12-25 15:30:00 America/Los_Angeles".to_string() + value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) },], fields: vec![StructField { field_name: None, @@ -912,7 +926,7 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: "15:30:00".to_string() + value: Value::SingleQuotedString("15:30:00".into()) },], fields: vec![StructField { field_name: None, @@ -929,7 +943,7 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), - value: "1".to_string() + value: Value::SingleQuotedString("1".into()) },], fields: vec![StructField { field_name: None, @@ -942,7 +956,7 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: "1".to_string() + value: Value::SingleQuotedString("1".into()) },], fields: vec![StructField { field_name: None, @@ -1119,9 +1133,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(4, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString( - "2011-05-05".to_string() - )),], + values: vec![Expr::Value(Value::SingleQuotedString("2011-05-05".into())),], fields: vec![StructField { field_name: None, field_type: DataType::Date @@ -1133,7 +1145,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), - value: "1999-01-01 01:23:34.45".to_string() + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) },], fields: vec![StructField { field_name: None, @@ -1169,7 +1181,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!( &Expr::Struct { values: vec![Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString("1".to_string()))), + value: Box::new(Expr::Value(Value::SingleQuotedString("1".into()))), leading_field: Some(DateTimeField::Month), leading_precision: None, last_field: None, @@ -1186,7 +1198,9 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::JSON, - value: r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.to_string() + value: Value::SingleQuotedString( + r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() + ) },], fields: vec![StructField { field_name: None, @@ -1201,7 +1215,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(3, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("foo".to_string())),], + values: vec![Expr::Value(Value::SingleQuotedString("foo".into())),], fields: vec![StructField { field_name: None, field_type: DataType::String(Some(42)) @@ -1213,7 +1227,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: "2008-12-25 15:30:00 America/Los_Angeles".to_string() + value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) },], fields: vec![StructField { field_name: None, @@ -1227,7 +1241,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: "15:30:00".to_string() + value: Value::SingleQuotedString("15:30:00".into()) },], fields: vec![StructField { field_name: None, @@ -1244,7 +1258,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), - value: "1".to_string() + value: Value::SingleQuotedString("1".into()) },], fields: vec![StructField { field_name: None, @@ -1257,7 +1271,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: "1".to_string() + value: Value::SingleQuotedString("1".into()) },], fields: vec![StructField { field_name: None, @@ -1285,7 +1299,7 @@ fn parse_typed_struct_with_field_name_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::DoubleQuotedString("foo".to_string())),], + values: vec![Expr::Value(Value::DoubleQuotedString("foo".into())),], fields: vec![StructField { field_name: Some(Ident::from("y")), field_type: DataType::String(None) @@ -1332,7 +1346,7 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("foo".to_string())),], + values: vec![Expr::Value(Value::SingleQuotedString("foo".into())),], fields: vec![StructField { field_name: Some(Ident::from("y")), field_type: DataType::String(None) @@ -2234,6 +2248,20 @@ fn test_select_as_value() { assert_eq!(Some(ValueTableMode::AsValue), select.value_table_mode); } +#[test] +fn test_triple_quote_typed_strings() { + bigquery().verified_expr(r#"JSON '''{"foo":"bar's"}'''"#); + + let expr = bigquery().verified_expr(r#"JSON """{"foo":"bar's"}""""#); + assert_eq!( + Expr::TypedString { + data_type: DataType::JSON, + value: Value::TripleDoubleQuotedString(r#"{"foo":"bar's"}"#.into()) + }, + expr + ); +} + #[test] fn test_array_agg() { bigquery_and_generic().verified_expr("ARRAY_AGG(state)"); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a695204c2..6113a3703 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -5409,7 +5409,7 @@ fn parse_literal_date() { assert_eq!( &Expr::TypedString { data_type: DataType::Date, - value: "1999-01-01".into(), + value: Value::SingleQuotedString("1999-01-01".into()), }, expr_from_projection(only(&select.projection)), ); @@ -5422,7 +5422,7 @@ fn parse_literal_time() { assert_eq!( &Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: "01:23:34".into(), + value: Value::SingleQuotedString("01:23:34".into()), }, expr_from_projection(only(&select.projection)), ); @@ -5435,7 +5435,7 @@ fn parse_literal_datetime() { assert_eq!( &Expr::TypedString { data_type: DataType::Datetime(None), - value: "1999-01-01 01:23:34.45".into(), + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), }, expr_from_projection(only(&select.projection)), ); @@ -5448,7 +5448,7 @@ fn parse_literal_timestamp_without_time_zone() { assert_eq!( &Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: "1999-01-01 01:23:34".into(), + value: Value::SingleQuotedString("1999-01-01 01:23:34".into()), }, expr_from_projection(only(&select.projection)), ); @@ -5463,7 +5463,7 @@ fn parse_literal_timestamp_with_time_zone() { assert_eq!( &Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::Tz), - value: "1999-01-01 01:23:34Z".into(), + value: Value::SingleQuotedString("1999-01-01 01:23:34Z".into()), }, expr_from_projection(only(&select.projection)), ); @@ -6015,7 +6015,8 @@ fn parse_json_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::JSON, - value: r#"{ + value: Value::SingleQuotedString( + r#"{ "id": 10, "type": "fruit", "name": "apple", @@ -6035,12 +6036,30 @@ fn parse_json_keyword() { ] } }"# - .into() + .to_string() + ) }, expr_from_projection(only(&select.projection)), ); } +#[test] +fn parse_typed_strings() { + let expr = verified_expr(r#"JSON '{"foo":"bar"}'"#); + assert_eq!( + Expr::TypedString { + data_type: DataType::JSON, + value: Value::SingleQuotedString(r#"{"foo":"bar"}"#.into()) + }, + expr + ); + + if let Expr::TypedString { data_type, value } = expr { + assert_eq!(DataType::JSON, data_type); + assert_eq!(r#"{"foo":"bar"}"#, value.into_string().unwrap()); + } +} + #[test] fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '0'"#; @@ -6048,7 +6067,7 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: r#"0"#.into() + value: Value::SingleQuotedString(r#"0"#.into()) }, expr_from_projection(only(&select.projection)), ); @@ -6059,7 +6078,7 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: r#"123456"#.into() + value: Value::SingleQuotedString(r#"123456"#.into()) }, expr_from_projection(only(&select.projection)), ); @@ -6070,7 +6089,7 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: r#"-3.14"#.into() + value: Value::SingleQuotedString(r#"-3.14"#.into()) }, expr_from_projection(only(&select.projection)), ); @@ -6081,7 +6100,7 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: r#"-0.54321"#.into() + value: Value::SingleQuotedString(r#"-0.54321"#.into()) }, expr_from_projection(only(&select.projection)), ); @@ -6092,7 +6111,7 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: r#"1.23456e05"#.into() + value: Value::SingleQuotedString(r#"1.23456e05"#.into()) }, expr_from_projection(only(&select.projection)), ); @@ -6103,7 +6122,7 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: r#"-9.876e-3"#.into() + value: Value::SingleQuotedString(r#"-9.876e-3"#.into()) }, expr_from_projection(only(&select.projection)), ); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 93af04980..ee4aa2a0a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -4638,7 +4638,7 @@ fn parse_at_time_zone() { left: Box::new(Expr::AtTimeZone { timestamp: Box::new(Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: "2001-09-28 01:00".to_owned(), + value: Value::SingleQuotedString("2001-09-28 01:00".to_string()), }), time_zone: Box::new(Expr::Cast { kind: CastKind::DoubleColon, From c3256a80d7b9708c6abf025f24a72779277f3a34 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sat, 1 Feb 2025 13:42:38 +0100 Subject: [PATCH 030/286] Do not parse ASOF and MATCH_CONDITION as table factor aliases (#1698) --- src/keywords.rs | 2 ++ tests/sqlparser_snowflake.rs | 14 ++++++++++++++ 2 files changed, 16 insertions(+) diff --git a/src/keywords.rs b/src/keywords.rs index 02ce04988..5937d7755 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -1000,6 +1000,8 @@ pub const RESERVED_FOR_TABLE_ALIAS: &[Keyword] = &[ Keyword::ANTI, Keyword::SEMI, Keyword::RETURNING, + Keyword::ASOF, + Keyword::MATCH_CONDITION, // for MSSQL-specific OUTER APPLY (seems reserved in most dialects) Keyword::OUTER, Keyword::SET, diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 1310b984f..1036ce1fa 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -2731,6 +2731,20 @@ fn asof_joins() { "ON s.state = p.state ", "ORDER BY s.observed", )); + + // Test without explicit aliases + #[rustfmt::skip] + snowflake_and_generic().verified_query(concat!( + "SELECT * ", + "FROM snowtime ", + "ASOF JOIN raintime ", + "MATCH_CONDITION (snowtime.observed >= raintime.observed) ", + "ON snowtime.state = raintime.state ", + "ASOF JOIN preciptime ", + "MATCH_CONDITION (showtime.observed >= preciptime.observed) ", + "ON showtime.state = preciptime.state ", + "ORDER BY showtime.observed", + )); } #[test] From 906f39534176f389b85e485044ca85d77caccaac Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 3 Feb 2025 08:21:17 +0100 Subject: [PATCH 031/286] Add support for GRANT on some common Snowflake objects (#1699) --- src/ast/mod.rs | 20 ++++++++++++++++++++ src/parser/mod.rs | 15 +++++++++++++-- tests/sqlparser_common.rs | 4 ++++ 3 files changed, 37 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index bccc580b3..c0d3ea574 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -5884,12 +5884,20 @@ pub enum GrantObjects { AllSequencesInSchema { schemas: Vec }, /// Grant privileges on `ALL TABLES IN SCHEMA [, ...]` AllTablesInSchema { schemas: Vec }, + /// Grant privileges on specific databases + Databases(Vec), /// Grant privileges on specific schemas Schemas(Vec), /// Grant privileges on specific sequences Sequences(Vec), /// Grant privileges on specific tables Tables(Vec), + /// Grant privileges on specific views + Views(Vec), + /// Grant privileges on specific warehouses + Warehouses(Vec), + /// Grant privileges on specific integrations + Integrations(Vec), } impl fmt::Display for GrantObjects { @@ -5898,12 +5906,24 @@ impl fmt::Display for GrantObjects { GrantObjects::Sequences(sequences) => { write!(f, "SEQUENCE {}", display_comma_separated(sequences)) } + GrantObjects::Databases(databases) => { + write!(f, "DATABASE {}", display_comma_separated(databases)) + } GrantObjects::Schemas(schemas) => { write!(f, "SCHEMA {}", display_comma_separated(schemas)) } GrantObjects::Tables(tables) => { write!(f, "{}", display_comma_separated(tables)) } + GrantObjects::Views(views) => { + write!(f, "VIEW {}", display_comma_separated(views)) + } + GrantObjects::Warehouses(warehouses) => { + write!(f, "WAREHOUSE {}", display_comma_separated(warehouses)) + } + GrantObjects::Integrations(integrations) => { + write!(f, "INTEGRATION {}", display_comma_separated(integrations)) + } GrantObjects::AllSequencesInSchema { schemas } => { write!( f, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ca858c42e..7d2d407d8 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12139,13 +12139,24 @@ impl<'a> Parser<'a> { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, } } else { - let object_type = - self.parse_one_of_keywords(&[Keyword::SEQUENCE, Keyword::SCHEMA, Keyword::TABLE]); + let object_type = self.parse_one_of_keywords(&[ + Keyword::SEQUENCE, + Keyword::DATABASE, + Keyword::SCHEMA, + Keyword::TABLE, + Keyword::VIEW, + Keyword::WAREHOUSE, + Keyword::INTEGRATION, + ]); let objects = self.parse_comma_separated(|p| p.parse_object_name_with_wildcards(false, true)); match object_type { + Some(Keyword::DATABASE) => GrantObjects::Databases(objects?), Some(Keyword::SCHEMA) => GrantObjects::Schemas(objects?), Some(Keyword::SEQUENCE) => GrantObjects::Sequences(objects?), + Some(Keyword::WAREHOUSE) => GrantObjects::Warehouses(objects?), + Some(Keyword::INTEGRATION) => GrantObjects::Integrations(objects?), + Some(Keyword::VIEW) => GrantObjects::Views(objects?), Some(Keyword::TABLE) | None => GrantObjects::Tables(objects?), _ => unreachable!(), } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 6113a3703..643ac357a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8779,6 +8779,10 @@ fn parse_grant() { verified_stmt("GRANT USAGE ON SCHEMA sc1 TO a:b"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO GROUP group1"); verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST"); + verified_stmt("GRANT USAGE ON DATABASE db1 TO ROLE role1"); + verified_stmt("GRANT USAGE ON WAREHOUSE wh1 TO ROLE role1"); + verified_stmt("GRANT OWNERSHIP ON INTEGRATION int1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON VIEW view1 TO ROLE role1"); } #[test] From 257da5a82c8a1e6c565abfa52bd490b211775069 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?R=C3=A9my=20SAISSY?= Date: Mon, 3 Feb 2025 08:27:37 +0100 Subject: [PATCH 032/286] Add RETURNS TABLE() support for CREATE FUNCTION in Postgresql (#1687) Co-authored-by: Ifeanyi Ubah --- src/ast/data_type.rs | 5 +++++ src/parser/mod.rs | 22 ++++++++++++++++++++++ tests/sqlparser_postgres.rs | 1 + 3 files changed, 28 insertions(+) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 02aa6cc9f..1f2b6be97 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -45,6 +45,10 @@ pub enum EnumMember { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum DataType { + /// Table type in [postgresql]. e.g. CREATE FUNCTION RETURNS TABLE(...) + /// + /// [postgresql]: https://www.postgresql.org/docs/15/sql-createfunction.html + Table(Vec), /// Fixed-length character type e.g. CHARACTER(10) Character(Option), /// Fixed-length char type e.g. CHAR(10) @@ -630,6 +634,7 @@ impl fmt::Display for DataType { DataType::Unspecified => Ok(()), DataType::Trigger => write!(f, "TRIGGER"), DataType::AnyType => write!(f, "ANY TYPE"), + DataType::Table(fields) => write!(f, "TABLE({})", display_comma_separated(fields)), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 7d2d407d8..931d97001 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8867,6 +8867,10 @@ impl<'a> Parser<'a> { let _ = self.parse_keyword(Keyword::TYPE); Ok(DataType::AnyType) } + Keyword::TABLE => { + let columns = self.parse_returns_table_columns()?; + Ok(DataType::Table(columns)) + } _ => { self.prev_token(); let type_name = self.parse_object_name(false)?; @@ -8894,6 +8898,24 @@ impl<'a> Parser<'a> { Ok((data, trailing_bracket)) } + fn parse_returns_table_column(&mut self) -> Result { + let name = self.parse_identifier()?; + let data_type = self.parse_data_type()?; + Ok(ColumnDef { + name, + data_type, + collation: None, + options: Vec::new(), // No constraints expected here + }) + } + + fn parse_returns_table_columns(&mut self) -> Result, ParserError> { + self.expect_token(&Token::LParen)?; + let columns = self.parse_comma_separated(Parser::parse_returns_table_column)?; + self.expect_token(&Token::RParen)?; + Ok(columns) + } + pub fn parse_string_values(&mut self) -> Result, ParserError> { self.expect_token(&Token::LParen)?; let mut values = Vec::new(); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index ee4aa2a0a..62da0f574 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -3803,6 +3803,7 @@ fn parse_create_function_detailed() { pg_and_generic().verified_stmt("CREATE OR REPLACE FUNCTION add(a INTEGER, IN b INTEGER = 1) RETURNS INTEGER LANGUAGE SQL STABLE CALLED ON NULL INPUT PARALLEL UNSAFE RETURN a + b"); pg_and_generic().verified_stmt(r#"CREATE OR REPLACE FUNCTION increment(i INTEGER) RETURNS INTEGER LANGUAGE plpgsql AS $$ BEGIN RETURN i + 1; END; $$"#); pg_and_generic().verified_stmt(r#"CREATE OR REPLACE FUNCTION no_arg() RETURNS VOID LANGUAGE plpgsql AS $$ BEGIN DELETE FROM my_table; END; $$"#); + pg_and_generic().verified_stmt(r#"CREATE OR REPLACE FUNCTION return_table(i INTEGER) RETURNS TABLE(id UUID, is_active BOOLEAN) LANGUAGE plpgsql AS $$ BEGIN RETURN QUERY SELECT NULL::UUID, NULL::BOOLEAN; END; $$"#); } #[test] fn parse_incorrect_create_function_parallel() { From ec948eaf6e9099451d69c525bb6987ebc721bec7 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 3 Feb 2025 20:17:47 +0100 Subject: [PATCH 033/286] Add parsing for GRANT ROLE and GRANT DATABASE ROLE in Snowflake dialect (#1689) --- src/ast/mod.rs | 20 ++++++-- src/parser/mod.rs | 89 +++++++++++++++++++----------------- tests/sqlparser_common.rs | 17 ++++--- tests/sqlparser_mysql.rs | 10 +++- tests/sqlparser_snowflake.rs | 12 +++++ 5 files changed, 94 insertions(+), 54 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index c0d3ea574..17b5276aa 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3230,7 +3230,7 @@ pub enum Statement { /// ``` Grant { privileges: Privileges, - objects: GrantObjects, + objects: Option, grantees: Vec, with_grant_option: bool, granted_by: Option, @@ -3240,7 +3240,7 @@ pub enum Statement { /// ``` Revoke { privileges: Privileges, - objects: GrantObjects, + objects: Option, grantees: Vec, granted_by: Option, cascade: Option, @@ -4785,7 +4785,9 @@ impl fmt::Display for Statement { granted_by, } => { write!(f, "GRANT {privileges} ")?; - write!(f, "ON {objects} ")?; + if let Some(objects) = objects { + write!(f, "ON {objects} ")?; + } write!(f, "TO {}", display_comma_separated(grantees))?; if *with_grant_option { write!(f, " WITH GRANT OPTION")?; @@ -4803,7 +4805,9 @@ impl fmt::Display for Statement { cascade, } => { write!(f, "REVOKE {privileges} ")?; - write!(f, "ON {objects} ")?; + if let Some(objects) = objects { + write!(f, "ON {objects} ")?; + } write!(f, "FROM {}", display_comma_separated(grantees))?; if let Some(grantor) = granted_by { write!(f, " GRANTED BY {grantor}")?; @@ -5503,6 +5507,9 @@ pub enum Action { Create { obj_type: Option, }, + DatabaseRole { + role: ObjectName, + }, Delete, EvolveSchema, Execute { @@ -5536,6 +5543,9 @@ pub enum Action { }, Replicate, ResolveAll, + Role { + role: Ident, + }, Select { columns: Option>, }, @@ -5565,6 +5575,7 @@ impl fmt::Display for Action { write!(f, " {obj_type}")? } } + Action::DatabaseRole { role } => write!(f, "DATABASE ROLE {role}")?, Action::Delete => f.write_str("DELETE")?, Action::EvolveSchema => f.write_str("EVOLVE SCHEMA")?, Action::Execute { obj_type } => { @@ -5591,6 +5602,7 @@ impl fmt::Display for Action { Action::References { .. } => f.write_str("REFERENCES")?, Action::Replicate => f.write_str("REPLICATE")?, Action::ResolveAll => f.write_str("RESOLVE ALL")?, + Action::Role { role } => write!(f, "ROLE {role}")?, Action::Select { .. } => f.write_str("SELECT")?, Action::Temporary => f.write_str("TEMPORARY")?, Action::Trigger => f.write_str("TRIGGER")?, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 931d97001..28cba0393 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12130,7 +12130,7 @@ impl<'a> Parser<'a> { pub fn parse_grant_revoke_privileges_objects( &mut self, - ) -> Result<(Privileges, GrantObjects), ParserError> { + ) -> Result<(Privileges, Option), ParserError> { let privileges = if self.parse_keyword(Keyword::ALL) { Privileges::All { with_privileges_keyword: self.parse_keyword(Keyword::PRIVILEGES), @@ -12140,48 +12140,49 @@ impl<'a> Parser<'a> { Privileges::Actions(actions) }; - self.expect_keyword_is(Keyword::ON)?; - - let objects = if self.parse_keywords(&[ - Keyword::ALL, - Keyword::TABLES, - Keyword::IN, - Keyword::SCHEMA, - ]) { - GrantObjects::AllTablesInSchema { - schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, - } - } else if self.parse_keywords(&[ - Keyword::ALL, - Keyword::SEQUENCES, - Keyword::IN, - Keyword::SCHEMA, - ]) { - GrantObjects::AllSequencesInSchema { - schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, - } - } else { - let object_type = self.parse_one_of_keywords(&[ - Keyword::SEQUENCE, - Keyword::DATABASE, + let objects = if self.parse_keyword(Keyword::ON) { + if self.parse_keywords(&[Keyword::ALL, Keyword::TABLES, Keyword::IN, Keyword::SCHEMA]) { + Some(GrantObjects::AllTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::SEQUENCES, + Keyword::IN, Keyword::SCHEMA, - Keyword::TABLE, - Keyword::VIEW, - Keyword::WAREHOUSE, - Keyword::INTEGRATION, - ]); - let objects = - self.parse_comma_separated(|p| p.parse_object_name_with_wildcards(false, true)); - match object_type { - Some(Keyword::DATABASE) => GrantObjects::Databases(objects?), - Some(Keyword::SCHEMA) => GrantObjects::Schemas(objects?), - Some(Keyword::SEQUENCE) => GrantObjects::Sequences(objects?), - Some(Keyword::WAREHOUSE) => GrantObjects::Warehouses(objects?), - Some(Keyword::INTEGRATION) => GrantObjects::Integrations(objects?), - Some(Keyword::VIEW) => GrantObjects::Views(objects?), - Some(Keyword::TABLE) | None => GrantObjects::Tables(objects?), - _ => unreachable!(), + ]) { + Some(GrantObjects::AllSequencesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else { + let object_type = self.parse_one_of_keywords(&[ + Keyword::SEQUENCE, + Keyword::DATABASE, + Keyword::DATABASE, + Keyword::SCHEMA, + Keyword::TABLE, + Keyword::VIEW, + Keyword::WAREHOUSE, + Keyword::INTEGRATION, + Keyword::VIEW, + Keyword::WAREHOUSE, + Keyword::INTEGRATION, + ]); + let objects = + self.parse_comma_separated(|p| p.parse_object_name_with_wildcards(false, true)); + match object_type { + Some(Keyword::DATABASE) => Some(GrantObjects::Databases(objects?)), + Some(Keyword::SCHEMA) => Some(GrantObjects::Schemas(objects?)), + Some(Keyword::SEQUENCE) => Some(GrantObjects::Sequences(objects?)), + Some(Keyword::WAREHOUSE) => Some(GrantObjects::Warehouses(objects?)), + Some(Keyword::INTEGRATION) => Some(GrantObjects::Integrations(objects?)), + Some(Keyword::VIEW) => Some(GrantObjects::Views(objects?)), + Some(Keyword::TABLE) | None => Some(GrantObjects::Tables(objects?)), + _ => unreachable!(), + } } + } else { + None }; Ok((privileges, objects)) @@ -12208,6 +12209,9 @@ impl<'a> Parser<'a> { Ok(Action::AttachPolicy) } else if self.parse_keywords(&[Keyword::BIND, Keyword::SERVICE, Keyword::ENDPOINT]) { Ok(Action::BindServiceEndpoint) + } else if self.parse_keywords(&[Keyword::DATABASE, Keyword::ROLE]) { + let role = self.parse_object_name(false)?; + Ok(Action::DatabaseRole { role }) } else if self.parse_keywords(&[Keyword::EVOLVE, Keyword::SCHEMA]) { Ok(Action::EvolveSchema) } else if self.parse_keywords(&[Keyword::IMPORT, Keyword::SHARE]) { @@ -12273,6 +12277,9 @@ impl<'a> Parser<'a> { Ok(Action::Read) } else if self.parse_keyword(Keyword::REPLICATE) { Ok(Action::Replicate) + } else if self.parse_keyword(Keyword::ROLE) { + let role = self.parse_identifier()?; + Ok(Action::Role { role }) } else if self.parse_keyword(Keyword::SELECT) { Ok(Action::Select { columns: parse_columns(self)?, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 643ac357a..4fcb37b6d 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8637,7 +8637,7 @@ fn parse_grant() { granted_by, .. } => match (privileges, objects) { - (Privileges::Actions(actions), GrantObjects::Tables(objects)) => { + (Privileges::Actions(actions), Some(GrantObjects::Tables(objects))) => { assert_eq!( vec![ Action::Select { columns: None }, @@ -8687,7 +8687,7 @@ fn parse_grant() { with_grant_option, .. } => match (privileges, objects) { - (Privileges::Actions(actions), GrantObjects::AllTablesInSchema { schemas }) => { + (Privileges::Actions(actions), Some(GrantObjects::AllTablesInSchema { schemas })) => { assert_eq!(vec![Action::Insert { columns: None }], actions); assert_eq_vec(&["public"], &schemas); assert_eq_vec(&["browser"], &grantees); @@ -8707,7 +8707,7 @@ fn parse_grant() { granted_by, .. } => match (privileges, objects, granted_by) { - (Privileges::Actions(actions), GrantObjects::Sequences(objects), None) => { + (Privileges::Actions(actions), Some(GrantObjects::Sequences(objects)), None) => { assert_eq!( vec![Action::Usage, Action::Select { columns: None }], actions @@ -8744,7 +8744,7 @@ fn parse_grant() { Privileges::All { with_privileges_keyword, }, - GrantObjects::Schemas(schemas), + Some(GrantObjects::Schemas(schemas)), ) => { assert!(!with_privileges_keyword); assert_eq_vec(&["aa", "b"], &schemas); @@ -8761,7 +8761,10 @@ fn parse_grant() { objects, .. } => match (privileges, objects) { - (Privileges::Actions(actions), GrantObjects::AllSequencesInSchema { schemas }) => { + ( + Privileges::Actions(actions), + Some(GrantObjects::AllSequencesInSchema { schemas }), + ) => { assert_eq!(vec![Action::Usage], actions); assert_eq_vec(&["bus"], &schemas); } @@ -8791,7 +8794,7 @@ fn test_revoke() { match verified_stmt(sql) { Statement::Revoke { privileges, - objects: GrantObjects::Tables(tables), + objects: Some(GrantObjects::Tables(tables)), grantees, granted_by, cascade, @@ -8817,7 +8820,7 @@ fn test_revoke_with_cascade() { match all_dialects_except(|d| d.is::()).verified_stmt(sql) { Statement::Revoke { privileges, - objects: GrantObjects::Tables(tables), + objects: Some(GrantObjects::Tables(tables)), grantees, granted_by, cascade, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 2e6dfc72b..9f00a9213 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3046,7 +3046,10 @@ fn parse_grant() { ); assert_eq!( objects, - GrantObjects::Tables(vec![ObjectName::from(vec!["*".into(), "*".into()])]) + Some(GrantObjects::Tables(vec![ObjectName::from(vec![ + "*".into(), + "*".into() + ])])) ); assert!(!with_grant_option); assert!(granted_by.is_none()); @@ -3087,7 +3090,10 @@ fn parse_revoke() { ); assert_eq!( objects, - GrantObjects::Tables(vec![ObjectName::from(vec!["db1".into(), "*".into()])]) + Some(GrantObjects::Tables(vec![ObjectName::from(vec![ + "db1".into(), + "*".into() + ])])) ); if let [Grantee { grantee_type: GranteesType::None, diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 1036ce1fa..c68ada3c9 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3263,3 +3263,15 @@ fn test_grant_account_privileges() { } } } + +#[test] +fn test_grant_role_to() { + snowflake_and_generic().verified_stmt("GRANT ROLE r1 TO ROLE r2"); + snowflake_and_generic().verified_stmt("GRANT ROLE r1 TO USER u1"); +} + +#[test] +fn test_grant_database_role_to() { + snowflake_and_generic().verified_stmt("GRANT DATABASE ROLE r1 TO ROLE r2"); + snowflake_and_generic().verified_stmt("GRANT DATABASE ROLE db1.sc1.r1 TO ROLE db1.sc1.r2"); +} From 486b29ffab6092d01807fee84701e397bcd73bd9 Mon Sep 17 00:00:00 2001 From: wugeer Date: Wed, 5 Feb 2025 01:33:12 +0800 Subject: [PATCH 034/286] Add support for `CREATE/ALTER/DROP CONNECTOR` syntax (#1701) --- src/ast/ddl.rs | 83 ++++++++++++++++++- src/ast/mod.rs | 71 ++++++++++++++-- src/ast/spans.rs | 3 + src/dialect/mod.rs | 1 + src/keywords.rs | 2 + src/parser/alter.rs | 45 +++++++++- src/parser/mod.rs | 61 +++++++++++++- tests/sqlparser_common.rs | 169 ++++++++++++++++++++++++++++++++++++++ 8 files changed, 420 insertions(+), 15 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 1b5ccda26..35e01e618 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -30,10 +30,10 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, CreateFunctionBody, CreateFunctionUsing, DataType, - Expr, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, - Ident, MySQLColumnPosition, ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, - SequenceOptions, SqlOption, Tag, Value, + display_comma_separated, display_separated, CommentDef, CreateFunctionBody, + CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, + FunctionDeterminismSpecifier, FunctionParallel, Ident, MySQLColumnPosition, ObjectName, + OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, Value, }; use crate::keywords::Keyword; use crate::tokenizer::Token; @@ -338,6 +338,23 @@ impl fmt::Display for Owner { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterConnectorOwner { + User(Ident), + Role(Ident), +} + +impl fmt::Display for AlterConnectorOwner { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + AlterConnectorOwner::User(ident) => write!(f, "USER {ident}"), + AlterConnectorOwner::Role(ident) => write!(f, "ROLE {ident}"), + } + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -2055,3 +2072,61 @@ impl fmt::Display for CreateFunction { Ok(()) } } + +/// ```sql +/// CREATE CONNECTOR [IF NOT EXISTS] connector_name +/// [TYPE datasource_type] +/// [URL datasource_url] +/// [COMMENT connector_comment] +/// [WITH DCPROPERTIES(property_name=property_value, ...)] +/// ``` +/// +/// [Hive](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27362034#LanguageManualDDL-CreateDataConnectorCreateConnector) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateConnector { + pub name: Ident, + pub if_not_exists: bool, + pub connector_type: Option, + pub url: Option, + pub comment: Option, + pub with_dcproperties: Option>, +} + +impl fmt::Display for CreateConnector { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "CREATE CONNECTOR {if_not_exists}{name}", + if_not_exists = if self.if_not_exists { + "IF NOT EXISTS " + } else { + "" + }, + name = self.name, + )?; + + if let Some(connector_type) = &self.connector_type { + write!(f, " TYPE '{connector_type}'")?; + } + + if let Some(url) = &self.url { + write!(f, " URL '{url}'")?; + } + + if let Some(comment) = &self.comment { + write!(f, " COMMENT = '{comment}'")?; + } + + if let Some(with_dcproperties) = &self.with_dcproperties { + write!( + f, + " WITH DCPROPERTIES({})", + display_comma_separated(with_dcproperties) + )?; + } + + Ok(()) + } +} diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 17b5276aa..35c6dcc1f 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -47,14 +47,14 @@ pub use self::dcl::{ AlterRoleOperation, ResetConfig, RoleOption, SecondaryRoles, SetConfigValue, Use, }; pub use self::ddl::{ - AlterColumnOperation, AlterIndexOperation, AlterPolicyOperation, AlterTableOperation, - ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, - ConstraintCharacteristics, CreateFunction, Deduplicate, DeferrableInitial, DropBehavior, - GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, - IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexOption, - IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, - ReferentialAction, TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, - UserDefinedTypeRepresentation, ViewColumnDef, + AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, + AlterTableOperation, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, + ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateFunction, Deduplicate, + DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, + IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, + IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, + ProcedureParam, ReferentialAction, TableConstraint, TagsColumnOption, + UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; @@ -2646,6 +2646,11 @@ pub enum Statement { with_check: Option, }, /// ```sql + /// CREATE CONNECTOR + /// ``` + /// See [Hive](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27362034#LanguageManualDDL-CreateDataConnectorCreateConnector) + CreateConnector(CreateConnector), + /// ```sql /// ALTER TABLE /// ``` AlterTable { @@ -2697,6 +2702,20 @@ pub enum Statement { operation: AlterPolicyOperation, }, /// ```sql + /// ALTER CONNECTOR connector_name SET DCPROPERTIES(property_name=property_value, ...); + /// or + /// ALTER CONNECTOR connector_name SET URL new_url; + /// or + /// ALTER CONNECTOR connector_name SET OWNER [USER|ROLE] user_or_role; + /// ``` + /// (Hive-specific) + AlterConnector { + name: Ident, + properties: Option>, + url: Option, + owner: Option, + }, + /// ```sql /// ATTACH DATABASE 'path/to/file' AS alias /// ``` /// (SQLite-specific) @@ -2795,6 +2814,11 @@ pub enum Statement { drop_behavior: Option, }, /// ```sql + /// DROP CONNECTOR + /// ``` + /// See [Hive](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27362034#LanguageManualDDL-DropConnector) + DropConnector { if_exists: bool, name: Ident }, + /// ```sql /// DECLARE /// ``` /// Declare Cursor Variables @@ -4354,6 +4378,7 @@ impl fmt::Display for Statement { Ok(()) } + Statement::CreateConnector(create_connector) => create_connector.fmt(f), Statement::AlterTable { name, if_exists, @@ -4411,6 +4436,28 @@ impl fmt::Display for Statement { } => { write!(f, "ALTER POLICY {name} ON {table_name}{operation}") } + Statement::AlterConnector { + name, + properties, + url, + owner, + } => { + write!(f, "ALTER CONNECTOR {name}")?; + if let Some(properties) = properties { + write!( + f, + " SET DCPROPERTIES({})", + display_comma_separated(properties) + )?; + } + if let Some(url) = url { + write!(f, " SET URL '{url}'")?; + } + if let Some(owner) = owner { + write!(f, " SET OWNER {owner}")?; + } + Ok(()) + } Statement::Drop { object_type, if_exists, @@ -4498,6 +4545,14 @@ impl fmt::Display for Statement { } Ok(()) } + Statement::DropConnector { if_exists, name } => { + write!( + f, + "DROP CONNECTOR {if_exists}{name}", + if_exists = if *if_exists { "IF EXISTS " } else { "" } + )?; + Ok(()) + } Statement::Discard { object_type } => { write!(f, "DISCARD {object_type}")?; Ok(()) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index f37c0194f..b26900857 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -398,6 +398,7 @@ impl Spanned for Statement { Statement::CreateIndex(create_index) => create_index.span(), Statement::CreateRole { .. } => Span::empty(), Statement::CreateSecret { .. } => Span::empty(), + Statement::CreateConnector { .. } => Span::empty(), Statement::AlterTable { name, if_exists: _, @@ -487,7 +488,9 @@ impl Spanned for Statement { Statement::OptimizeTable { .. } => Span::empty(), Statement::CreatePolicy { .. } => Span::empty(), Statement::AlterPolicy { .. } => Span::empty(), + Statement::AlterConnector { .. } => Span::empty(), Statement::DropPolicy { .. } => Span::empty(), + Statement::DropConnector { .. } => Span::empty(), Statement::ShowDatabases { .. } => Span::empty(), Statement::ShowSchemas { .. } => Span::empty(), Statement::ShowViews { .. } => Span::empty(), diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index b648869d2..205395f6c 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -876,6 +876,7 @@ pub trait Dialect: Debug + Any { fn supports_string_escape_constant(&self) -> bool { false } + /// Returns true if the dialect supports the table hints in the `FROM` clause. fn supports_table_hints(&self) -> bool { false diff --git a/src/keywords.rs b/src/keywords.rs index 5937d7755..5f36fa737 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -201,6 +201,7 @@ define_keywords!( CONFLICT, CONNECT, CONNECTION, + CONNECTOR, CONSTRAINT, CONTAINS, CONTINUE, @@ -246,6 +247,7 @@ define_keywords!( DAYOFWEEK, DAYOFYEAR, DAYS, + DCPROPERTIES, DEALLOCATE, DEC, DECADE, diff --git a/src/parser/alter.rs b/src/parser/alter.rs index bb6782c13..bff462ee0 100644 --- a/src/parser/alter.rs +++ b/src/parser/alter.rs @@ -18,8 +18,8 @@ use alloc::vec; use super::{Parser, ParserError}; use crate::{ ast::{ - AlterPolicyOperation, AlterRoleOperation, Expr, Password, ResetConfig, RoleOption, - SetConfigValue, Statement, + AlterConnectorOwner, AlterPolicyOperation, AlterRoleOperation, Expr, Password, ResetConfig, + RoleOption, SetConfigValue, Statement, }, dialect::{MsSqlDialect, PostgreSqlDialect}, keywords::Keyword, @@ -99,6 +99,47 @@ impl Parser<'_> { } } + /// Parse an `ALTER CONNECTOR` statement + /// ```sql + /// ALTER CONNECTOR connector_name SET DCPROPERTIES(property_name=property_value, ...); + /// + /// ALTER CONNECTOR connector_name SET URL new_url; + /// + /// ALTER CONNECTOR connector_name SET OWNER [USER|ROLE] user_or_role; + /// ``` + pub fn parse_alter_connector(&mut self) -> Result { + let name = self.parse_identifier()?; + self.expect_keyword_is(Keyword::SET)?; + + let properties = match self.parse_options_with_keywords(&[Keyword::DCPROPERTIES])? { + properties if !properties.is_empty() => Some(properties), + _ => None, + }; + + let url = if self.parse_keyword(Keyword::URL) { + Some(self.parse_literal_string()?) + } else { + None + }; + + let owner = if self.parse_keywords(&[Keyword::OWNER, Keyword::USER]) { + let owner = self.parse_identifier()?; + Some(AlterConnectorOwner::User(owner)) + } else if self.parse_keywords(&[Keyword::OWNER, Keyword::ROLE]) { + let owner = self.parse_identifier()?; + Some(AlterConnectorOwner::Role(owner)) + } else { + None + }; + + Ok(Statement::AlterConnector { + name, + properties, + url, + owner, + }) + } + fn parse_mssql_alter_role(&mut self) -> Result { let role_name = self.parse_identifier()?; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 28cba0393..6d84ff843 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4268,6 +4268,8 @@ impl<'a> Parser<'a> { self.parse_create_type() } else if self.parse_keyword(Keyword::PROCEDURE) { self.parse_create_procedure(or_alter) + } else if self.parse_keyword(Keyword::CONNECTOR) { + self.parse_create_connector() } else { self.expected("an object type after CREATE", self.peek_token()) } @@ -5580,6 +5582,49 @@ impl<'a> Parser<'a> { }) } + /// ```sql + /// CREATE CONNECTOR [IF NOT EXISTS] connector_name + /// [TYPE datasource_type] + /// [URL datasource_url] + /// [COMMENT connector_comment] + /// [WITH DCPROPERTIES(property_name=property_value, ...)] + /// ``` + /// + /// [Hive Documentation](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27362034#LanguageManualDDL-CreateDataConnectorCreateConnector) + pub fn parse_create_connector(&mut self) -> Result { + let if_not_exists = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let name = self.parse_identifier()?; + + let connector_type = if self.parse_keyword(Keyword::TYPE) { + Some(self.parse_literal_string()?) + } else { + None + }; + + let url = if self.parse_keyword(Keyword::URL) { + Some(self.parse_literal_string()?) + } else { + None + }; + + let comment = self.parse_optional_inline_comment()?; + + let with_dcproperties = + match self.parse_options_with_keywords(&[Keyword::WITH, Keyword::DCPROPERTIES])? { + properties if !properties.is_empty() => Some(properties), + _ => None, + }; + + Ok(Statement::CreateConnector(CreateConnector { + name, + if_not_exists, + connector_type, + url, + comment, + with_dcproperties, + })) + } + pub fn parse_drop(&mut self) -> Result { // MySQL dialect supports `TEMPORARY` let temporary = dialect_of!(self is MySqlDialect | GenericDialect | DuckDbDialect) @@ -5609,6 +5654,8 @@ impl<'a> Parser<'a> { return self.parse_drop_function(); } else if self.parse_keyword(Keyword::POLICY) { return self.parse_drop_policy(); + } else if self.parse_keyword(Keyword::CONNECTOR) { + return self.parse_drop_connector(); } else if self.parse_keyword(Keyword::PROCEDURE) { return self.parse_drop_procedure(); } else if self.parse_keyword(Keyword::SECRET) { @@ -5619,7 +5666,7 @@ impl<'a> Parser<'a> { return self.parse_drop_extension(); } else { return self.expected( - "DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, or VIEW after DROP", + "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, or VIEW after DROP", self.peek_token(), ); }; @@ -5693,6 +5740,16 @@ impl<'a> Parser<'a> { drop_behavior, }) } + /// ```sql + /// DROP CONNECTOR [IF EXISTS] name + /// ``` + /// + /// See [Hive](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27362034#LanguageManualDDL-DropConnector) + fn parse_drop_connector(&mut self) -> Result { + let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); + let name = self.parse_identifier()?; + Ok(Statement::DropConnector { if_exists, name }) + } /// ```sql /// DROP PROCEDURE [ IF EXISTS ] name [ ( [ [ argmode ] [ argname ] argtype [, ...] ] ) ] [, ...] @@ -7989,6 +8046,7 @@ impl<'a> Parser<'a> { Keyword::INDEX, Keyword::ROLE, Keyword::POLICY, + Keyword::CONNECTOR, ])?; match object_type { Keyword::VIEW => self.parse_alter_view(), @@ -8041,6 +8099,7 @@ impl<'a> Parser<'a> { } Keyword::ROLE => self.parse_alter_role(), Keyword::POLICY => self.parse_alter_policy(), + Keyword::CONNECTOR => self.parse_alter_connector(), // unreachable because expect_one_of_keywords used above _ => unreachable!(), } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 4fcb37b6d..3a6183a15 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -12289,6 +12289,175 @@ fn test_alter_policy() { ); } +#[test] +fn test_create_connector() { + let sql = "CREATE CONNECTOR my_connector \ + TYPE 'jdbc' \ + URL 'jdbc:mysql://localhost:3306/mydb' \ + WITH DCPROPERTIES('user' = 'root', 'password' = 'password')"; + let dialects = all_dialects(); + match dialects.verified_stmt(sql) { + Statement::CreateConnector(CreateConnector { + name, + connector_type, + url, + with_dcproperties, + .. + }) => { + assert_eq!(name.to_string(), "my_connector"); + assert_eq!(connector_type, Some("jdbc".to_string())); + assert_eq!(url, Some("jdbc:mysql://localhost:3306/mydb".to_string())); + assert_eq!( + with_dcproperties, + Some(vec![ + SqlOption::KeyValue { + key: Ident::with_quote('\'', "user"), + value: Expr::Value(Value::SingleQuotedString("root".to_string())) + }, + SqlOption::KeyValue { + key: Ident::with_quote('\'', "password"), + value: Expr::Value(Value::SingleQuotedString("password".to_string())) + } + ]) + ); + } + _ => unreachable!(), + } + + // omit IF NOT EXISTS/TYPE/URL/COMMENT/WITH DCPROPERTIES clauses is allowed + dialects.verified_stmt("CREATE CONNECTOR my_connector"); + + // missing connector name + assert_eq!( + dialects + .parse_sql_statements("CREATE CONNECTOR") + .unwrap_err() + .to_string(), + "sql parser error: Expected: identifier, found: EOF" + ); +} + +#[test] +fn test_drop_connector() { + let dialects = all_dialects(); + match dialects.verified_stmt("DROP CONNECTOR IF EXISTS my_connector") { + Statement::DropConnector { if_exists, name } => { + assert_eq!(if_exists, true); + assert_eq!(name.to_string(), "my_connector"); + } + _ => unreachable!(), + } + + // omit IF EXISTS is allowed + dialects.verified_stmt("DROP CONNECTOR my_connector"); + + // missing connector name + assert_eq!( + dialects + .parse_sql_statements("DROP CONNECTOR") + .unwrap_err() + .to_string(), + "sql parser error: Expected: identifier, found: EOF" + ); +} + +#[test] +fn test_alter_connector() { + let dialects = all_dialects(); + match dialects.verified_stmt( + "ALTER CONNECTOR my_connector SET DCPROPERTIES('user' = 'root', 'password' = 'password')", + ) { + Statement::AlterConnector { + name, + properties, + url, + owner, + } => { + assert_eq!(name.to_string(), "my_connector"); + assert_eq!( + properties, + Some(vec![ + SqlOption::KeyValue { + key: Ident::with_quote('\'', "user"), + value: Expr::Value(Value::SingleQuotedString("root".to_string())) + }, + SqlOption::KeyValue { + key: Ident::with_quote('\'', "password"), + value: Expr::Value(Value::SingleQuotedString("password".to_string())) + } + ]) + ); + assert_eq!(url, None); + assert_eq!(owner, None); + } + _ => unreachable!(), + } + + match dialects + .verified_stmt("ALTER CONNECTOR my_connector SET URL 'jdbc:mysql://localhost:3306/mydb'") + { + Statement::AlterConnector { + name, + properties, + url, + owner, + } => { + assert_eq!(name.to_string(), "my_connector"); + assert_eq!(properties, None); + assert_eq!(url, Some("jdbc:mysql://localhost:3306/mydb".to_string())); + assert_eq!(owner, None); + } + _ => unreachable!(), + } + + match dialects.verified_stmt("ALTER CONNECTOR my_connector SET OWNER USER 'root'") { + Statement::AlterConnector { + name, + properties, + url, + owner, + } => { + assert_eq!(name.to_string(), "my_connector"); + assert_eq!(properties, None); + assert_eq!(url, None); + assert_eq!( + owner, + Some(AlterConnectorOwner::User(Ident::with_quote('\'', "root"))) + ); + } + _ => unreachable!(), + } + + match dialects.verified_stmt("ALTER CONNECTOR my_connector SET OWNER ROLE 'admin'") { + Statement::AlterConnector { + name, + properties, + url, + owner, + } => { + assert_eq!(name.to_string(), "my_connector"); + assert_eq!(properties, None); + assert_eq!(url, None); + assert_eq!( + owner, + Some(AlterConnectorOwner::Role(Ident::with_quote('\'', "admin"))) + ); + } + _ => unreachable!(), + } + + // Wrong option name + assert_eq!( + dialects + .parse_sql_statements( + "ALTER CONNECTOR my_connector SET WRONG 'jdbc:mysql://localhost:3306/mydb'" + ) + .unwrap_err() + .to_string(), + "sql parser error: Expected: end of statement, found: WRONG" + ); +} + #[test] fn test_select_where_with_like_or_ilike_any() { verified_stmt(r#"SELECT * FROM x WHERE a ILIKE ANY '%abc%'"#); From 751dc5afce794398bcfc148aa6b2f9ecf3b6b5a9 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 5 Feb 2025 20:23:27 +0100 Subject: [PATCH 035/286] Parse Snowflake COPY INTO (#1669) --- src/ast/helpers/stmt_data_loading.rs | 10 +- src/ast/mod.rs | 102 +++++++---- src/ast/spans.rs | 7 +- src/dialect/snowflake.rs | 250 +++++++++++++++------------ tests/sqlparser_snowflake.rs | 197 +++++++++++++++++---- 5 files changed, 381 insertions(+), 185 deletions(-) diff --git a/src/ast/helpers/stmt_data_loading.rs b/src/ast/helpers/stmt_data_loading.rs index 42e1df06b..77de5d9ec 100644 --- a/src/ast/helpers/stmt_data_loading.rs +++ b/src/ast/helpers/stmt_data_loading.rs @@ -58,6 +58,7 @@ pub enum DataLoadingOptionType { STRING, BOOLEAN, ENUM, + NUMBER, } #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] @@ -128,12 +129,9 @@ impl fmt::Display for DataLoadingOption { DataLoadingOptionType::STRING => { write!(f, "{}='{}'", self.option_name, self.value)?; } - DataLoadingOptionType::ENUM => { - // single quote is omitted - write!(f, "{}={}", self.option_name, self.value)?; - } - DataLoadingOptionType::BOOLEAN => { - // single quote is omitted + DataLoadingOptionType::ENUM + | DataLoadingOptionType::BOOLEAN + | DataLoadingOptionType::NUMBER => { write!(f, "{}={}", self.option_name, self.value)?; } } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 35c6dcc1f..dc944c9e7 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2498,24 +2498,30 @@ pub enum Statement { values: Vec>, }, /// ```sql - /// COPY INTO + /// COPY INTO | /// ``` - /// See + /// See: + /// + /// + /// /// Copy Into syntax available for Snowflake is different than the one implemented in /// Postgres. Although they share common prefix, it is reasonable to implement them /// in different enums. This can be refactored later once custom dialects /// are allowed to have custom Statements. CopyIntoSnowflake { + kind: CopyIntoSnowflakeKind, into: ObjectName, - from_stage: ObjectName, - from_stage_alias: Option, + from_obj: Option, + from_obj_alias: Option, stage_params: StageParamsObject, from_transformations: Option>, + from_query: Option>, files: Option>, pattern: Option, file_format: DataLoadingOptions, copy_options: DataLoadingOptions, validation_mode: Option, + partition: Option>, }, /// ```sql /// CLOSE @@ -5048,60 +5054,69 @@ impl fmt::Display for Statement { Ok(()) } Statement::CopyIntoSnowflake { + kind, into, - from_stage, - from_stage_alias, + from_obj, + from_obj_alias, stage_params, from_transformations, + from_query, files, pattern, file_format, copy_options, validation_mode, + partition, } => { write!(f, "COPY INTO {}", into)?; - if from_transformations.is_none() { - // Standard data load - write!(f, " FROM {}{}", from_stage, stage_params)?; - if from_stage_alias.as_ref().is_some() { - write!(f, " AS {}", from_stage_alias.as_ref().unwrap())?; - } - } else { + if let Some(from_transformations) = from_transformations { // Data load with transformation - write!( - f, - " FROM (SELECT {} FROM {}{}", - display_separated(from_transformations.as_ref().unwrap(), ", "), - from_stage, - stage_params, - )?; - if from_stage_alias.as_ref().is_some() { - write!(f, " AS {}", from_stage_alias.as_ref().unwrap())?; + if let Some(from_stage) = from_obj { + write!( + f, + " FROM (SELECT {} FROM {}{}", + display_separated(from_transformations, ", "), + from_stage, + stage_params + )?; + } + if let Some(from_obj_alias) = from_obj_alias { + write!(f, " AS {}", from_obj_alias)?; } write!(f, ")")?; + } else if let Some(from_obj) = from_obj { + // Standard data load + write!(f, " FROM {}{}", from_obj, stage_params)?; + if let Some(from_obj_alias) = from_obj_alias { + write!(f, " AS {from_obj_alias}")?; + } + } else if let Some(from_query) = from_query { + // Data unload from query + write!(f, " FROM ({from_query})")?; } - if files.is_some() { - write!( - f, - " FILES = ('{}')", - display_separated(files.as_ref().unwrap(), "', '") - )?; + + if let Some(files) = files { + write!(f, " FILES = ('{}')", display_separated(files, "', '"))?; + } + if let Some(pattern) = pattern { + write!(f, " PATTERN = '{}'", pattern)?; } - if pattern.is_some() { - write!(f, " PATTERN = '{}'", pattern.as_ref().unwrap())?; + if let Some(partition) = partition { + write!(f, " PARTITION BY {partition}")?; } if !file_format.options.is_empty() { write!(f, " FILE_FORMAT=({})", file_format)?; } if !copy_options.options.is_empty() { - write!(f, " COPY_OPTIONS=({})", copy_options)?; + match kind { + CopyIntoSnowflakeKind::Table => { + write!(f, " COPY_OPTIONS=({})", copy_options)? + } + CopyIntoSnowflakeKind::Location => write!(f, " {copy_options}")?, + } } - if validation_mode.is_some() { - write!( - f, - " VALIDATION_MODE = {}", - validation_mode.as_ref().unwrap() - )?; + if let Some(validation_mode) = validation_mode { + write!(f, " VALIDATION_MODE = {}", validation_mode)?; } Ok(()) } @@ -8543,6 +8558,19 @@ impl Display for StorageSerializationPolicy { } } +/// Variants of the Snowflake `COPY INTO` statement +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum CopyIntoSnowflakeKind { + /// Loads data from files to a table + /// See: + Table, + /// Unloads data from a table or query to external files + /// See: + Location, +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index b26900857..f0c38942e 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -333,8 +333,8 @@ impl Spanned for Statement { } => source.span(), Statement::CopyIntoSnowflake { into: _, - from_stage: _, - from_stage_alias: _, + from_obj: _, + from_obj_alias: _, stage_params: _, from_transformations: _, files: _, @@ -342,6 +342,9 @@ impl Spanned for Statement { file_format: _, copy_options: _, validation_mode: _, + kind: _, + from_query: _, + partition: _, } => Span::empty(), Statement::Close { cursor } => match cursor { CloseCursor::All => Span::empty(), diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index d775ffc36..fc1926719 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -23,23 +23,26 @@ use crate::ast::helpers::stmt_data_loading::{ StageLoadSelectItem, StageParamsObject, }; use crate::ast::{ - ColumnOption, ColumnPolicy, ColumnPolicyProperty, Ident, IdentityParameters, IdentityProperty, - IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, ObjectName, - RowAccessPolicy, Statement, TagsColumnOption, WrappedCollection, + ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, + IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, + IdentityPropertyOrder, ObjectName, RowAccessPolicy, Statement, TagsColumnOption, + WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; -use crate::tokenizer::Token; +use crate::tokenizer::{Token, Word}; +#[cfg(not(feature = "std"))] +use alloc::boxed::Box; #[cfg(not(feature = "std"))] use alloc::string::String; #[cfg(not(feature = "std"))] use alloc::vec::Vec; #[cfg(not(feature = "std"))] use alloc::{format, vec}; -use sqlparser::ast::StorageSerializationPolicy; use super::keywords::RESERVED_FOR_IDENTIFIER; +use sqlparser::ast::StorageSerializationPolicy; /// A [`Dialect`] for [Snowflake](https://www.snowflake.com/) #[derive(Debug, Default)] @@ -665,24 +668,49 @@ pub fn parse_snowflake_stage_name(parser: &mut Parser) -> Result` +/// and `COPY INTO ` which have different syntax. pub fn parse_copy_into(parser: &mut Parser) -> Result { - let into: ObjectName = parse_snowflake_stage_name(parser)?; + let kind = match parser.peek_token().token { + // Indicates an internal stage + Token::AtSign => CopyIntoSnowflakeKind::Location, + // Indicates an external stage, i.e. s3://, gcs:// or azure:// + Token::SingleQuotedString(s) if s.contains("://") => CopyIntoSnowflakeKind::Location, + _ => CopyIntoSnowflakeKind::Table, + }; + let mut files: Vec = vec![]; let mut from_transformations: Option> = None; - let from_stage_alias; - let from_stage: ObjectName; - let stage_params: StageParamsObject; + let mut from_stage_alias = None; + let mut from_stage = None; + let mut stage_params = StageParamsObject { + url: None, + encryption: DataLoadingOptions { options: vec![] }, + endpoint: None, + storage_integration: None, + credentials: DataLoadingOptions { options: vec![] }, + }; + let mut from_query = None; + let mut partition = None; + let mut file_format = Vec::new(); + let mut pattern = None; + let mut validation_mode = None; + let mut copy_options = Vec::new(); + + let into: ObjectName = parse_snowflake_stage_name(parser)?; + if kind == CopyIntoSnowflakeKind::Location { + stage_params = parse_stage_params(parser)?; + } parser.expect_keyword_is(Keyword::FROM)?; - // check if data load transformations are present match parser.next_token().token { - Token::LParen => { - // data load with transformations + Token::LParen if kind == CopyIntoSnowflakeKind::Table => { + // Data load with transformations parser.expect_keyword_is(Keyword::SELECT)?; from_transformations = parse_select_items_for_data_load(parser)?; parser.expect_keyword_is(Keyword::FROM)?; - from_stage = parse_snowflake_stage_name(parser)?; + from_stage = Some(parse_snowflake_stage_name(parser)?); stage_params = parse_stage_params(parser)?; // as @@ -696,9 +724,14 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { }; parser.expect_token(&Token::RParen)?; } + Token::LParen if kind == CopyIntoSnowflakeKind::Location => { + // Data unload with a query + from_query = Some(parser.parse_query()?); + parser.expect_token(&Token::RParen)?; + } _ => { parser.prev_token(); - from_stage = parse_snowflake_stage_name(parser)?; + from_stage = Some(parse_snowflake_stage_name(parser)?); stage_params = parse_stage_params(parser)?; // as @@ -711,67 +744,71 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { None }; } - }; + } - // [ files ] - if parser.parse_keyword(Keyword::FILES) { - parser.expect_token(&Token::Eq)?; - parser.expect_token(&Token::LParen)?; - let mut continue_loop = true; - while continue_loop { - continue_loop = false; + loop { + // FILE_FORMAT + if parser.parse_keyword(Keyword::FILE_FORMAT) { + parser.expect_token(&Token::Eq)?; + file_format = parse_parentheses_options(parser)?; + // PARTITION BY + } else if parser.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) { + partition = Some(Box::new(parser.parse_expr()?)) + // FILES + } else if parser.parse_keyword(Keyword::FILES) { + parser.expect_token(&Token::Eq)?; + parser.expect_token(&Token::LParen)?; + let mut continue_loop = true; + while continue_loop { + continue_loop = false; + let next_token = parser.next_token(); + match next_token.token { + Token::SingleQuotedString(s) => files.push(s), + _ => parser.expected("file token", next_token)?, + }; + if parser.next_token().token.eq(&Token::Comma) { + continue_loop = true; + } else { + parser.prev_token(); // not a comma, need to go back + } + } + parser.expect_token(&Token::RParen)?; + // PATTERN + } else if parser.parse_keyword(Keyword::PATTERN) { + parser.expect_token(&Token::Eq)?; let next_token = parser.next_token(); - match next_token.token { - Token::SingleQuotedString(s) => files.push(s), - _ => parser.expected("file token", next_token)?, - }; - if parser.next_token().token.eq(&Token::Comma) { - continue_loop = true; - } else { - parser.prev_token(); // not a comma, need to go back + pattern = Some(match next_token.token { + Token::SingleQuotedString(s) => s, + _ => parser.expected("pattern", next_token)?, + }); + // VALIDATION MODE + } else if parser.parse_keyword(Keyword::VALIDATION_MODE) { + parser.expect_token(&Token::Eq)?; + validation_mode = Some(parser.next_token().token.to_string()); + // COPY OPTIONS + } else if parser.parse_keyword(Keyword::COPY_OPTIONS) { + parser.expect_token(&Token::Eq)?; + copy_options = parse_parentheses_options(parser)?; + } else { + match parser.next_token().token { + Token::SemiColon | Token::EOF => break, + Token::Comma => continue, + // In `COPY INTO ` the copy options do not have a shared key + // like in `COPY INTO
` + Token::Word(key) => copy_options.push(parse_copy_option(parser, key)?), + _ => return parser.expected("another copy option, ; or EOF'", parser.peek_token()), } } - parser.expect_token(&Token::RParen)?; - } - - // [ pattern ] - let mut pattern = None; - if parser.parse_keyword(Keyword::PATTERN) { - parser.expect_token(&Token::Eq)?; - let next_token = parser.next_token(); - pattern = Some(match next_token.token { - Token::SingleQuotedString(s) => s, - _ => parser.expected("pattern", next_token)?, - }); - } - - // [ file_format] - let mut file_format = Vec::new(); - if parser.parse_keyword(Keyword::FILE_FORMAT) { - parser.expect_token(&Token::Eq)?; - file_format = parse_parentheses_options(parser)?; - } - - // [ copy_options ] - let mut copy_options = Vec::new(); - if parser.parse_keyword(Keyword::COPY_OPTIONS) { - parser.expect_token(&Token::Eq)?; - copy_options = parse_parentheses_options(parser)?; - } - - // [ VALIDATION_MODE ] - let mut validation_mode = None; - if parser.parse_keyword(Keyword::VALIDATION_MODE) { - parser.expect_token(&Token::Eq)?; - validation_mode = Some(parser.next_token().token.to_string()); } Ok(Statement::CopyIntoSnowflake { + kind, into, - from_stage, - from_stage_alias, + from_obj: from_stage, + from_obj_alias: from_stage_alias, stage_params, from_transformations, + from_query, files: if files.is_empty() { None } else { Some(files) }, pattern, file_format: DataLoadingOptions { @@ -781,6 +818,7 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { options: copy_options, }, validation_mode, + partition, }) } @@ -930,55 +968,55 @@ fn parse_stage_params(parser: &mut Parser) -> Result Result, ParserError> { let mut options: Vec = Vec::new(); - parser.expect_token(&Token::LParen)?; loop { match parser.next_token().token { Token::RParen => break, - Token::Word(key) => { - parser.expect_token(&Token::Eq)?; - if parser.parse_keyword(Keyword::TRUE) { - options.push(DataLoadingOption { - option_name: key.value, - option_type: DataLoadingOptionType::BOOLEAN, - value: "TRUE".to_string(), - }); - Ok(()) - } else if parser.parse_keyword(Keyword::FALSE) { - options.push(DataLoadingOption { - option_name: key.value, - option_type: DataLoadingOptionType::BOOLEAN, - value: "FALSE".to_string(), - }); - Ok(()) - } else { - match parser.next_token().token { - Token::SingleQuotedString(value) => { - options.push(DataLoadingOption { - option_name: key.value, - option_type: DataLoadingOptionType::STRING, - value, - }); - Ok(()) - } - Token::Word(word) => { - options.push(DataLoadingOption { - option_name: key.value, - option_type: DataLoadingOptionType::ENUM, - value: word.value, - }); - Ok(()) - } - _ => parser.expected("expected option value", parser.peek_token()), - } - } - } - _ => parser.expected("another option or ')'", parser.peek_token()), - }?; + Token::Comma => continue, + Token::Word(key) => options.push(parse_copy_option(parser, key)?), + _ => return parser.expected("another option or ')'", parser.peek_token()), + }; } Ok(options) } +/// Parses a `KEY = VALUE` construct based on the specified key +fn parse_copy_option(parser: &mut Parser, key: Word) -> Result { + parser.expect_token(&Token::Eq)?; + if parser.parse_keyword(Keyword::TRUE) { + Ok(DataLoadingOption { + option_name: key.value, + option_type: DataLoadingOptionType::BOOLEAN, + value: "TRUE".to_string(), + }) + } else if parser.parse_keyword(Keyword::FALSE) { + Ok(DataLoadingOption { + option_name: key.value, + option_type: DataLoadingOptionType::BOOLEAN, + value: "FALSE".to_string(), + }) + } else { + match parser.next_token().token { + Token::SingleQuotedString(value) => Ok(DataLoadingOption { + option_name: key.value, + option_type: DataLoadingOptionType::STRING, + value, + }), + Token::Word(word) => Ok(DataLoadingOption { + option_name: key.value, + option_type: DataLoadingOptionType::ENUM, + value: word.value, + }), + Token::Number(n, _) => Ok(DataLoadingOption { + option_name: key.value, + option_type: DataLoadingOptionType::NUMBER, + value: n, + }), + _ => parser.expected("expected option value", parser.peek_token()), + } + } +} + /// Parsing a property of identity or autoincrement column option /// Syntax: /// ```sql diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index c68ada3c9..a18f1a4d8 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -2028,20 +2028,25 @@ fn test_copy_into() { ); match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { + kind, into, - from_stage, + from_obj, files, pattern, validation_mode, .. } => { + assert_eq!(kind, CopyIntoSnowflakeKind::Table); assert_eq!( into, ObjectName::from(vec![Ident::new("my_company"), Ident::new("emp_basic")]) ); assert_eq!( - from_stage, - ObjectName::from(vec![Ident::with_quote('\'', "gcs://mybucket/./../a.csv")]) + from_obj, + Some(ObjectName::from(vec![Ident::with_quote( + '\'', + "gcs://mybucket/./../a.csv" + )])) ); assert!(files.is_none()); assert!(pattern.is_none()); @@ -2050,6 +2055,60 @@ fn test_copy_into() { _ => unreachable!(), }; assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); + + let sql = concat!("COPY INTO 's3://a/b/c/data.parquet' ", "FROM db.sc.tbl ", "PARTITION BY ('date=' || to_varchar(dt, 'YYYY-MM-DD') || '/hour=' || to_varchar(date_part(hour, ts)))"); + match snowflake().verified_stmt(sql) { + Statement::CopyIntoSnowflake { + kind, + into, + from_obj, + from_query, + partition, + .. + } => { + assert_eq!(kind, CopyIntoSnowflakeKind::Location); + assert_eq!( + into, + ObjectName::from(vec![Ident::with_quote('\'', "s3://a/b/c/data.parquet")]) + ); + assert_eq!( + from_obj, + Some(ObjectName::from(vec![ + Ident::new("db"), + Ident::new("sc"), + Ident::new("tbl") + ])) + ); + assert!(from_query.is_none()); + assert!(partition.is_some()); + } + _ => unreachable!(), + }; + assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); + + let sql = concat!( + "COPY INTO 's3://a/b/c/data.parquet' ", + "FROM (SELECT * FROM tbl)" + ); + match snowflake().verified_stmt(sql) { + Statement::CopyIntoSnowflake { + kind, + into, + from_obj, + from_query, + .. + } => { + assert_eq!(kind, CopyIntoSnowflakeKind::Location); + assert_eq!( + into, + ObjectName::from(vec![Ident::with_quote('\'', "s3://a/b/c/data.parquet")]) + ); + assert!(from_query.is_some()); + assert!(from_obj.is_none()); + } + _ => unreachable!(), + }; + assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); } #[test] @@ -2065,14 +2124,17 @@ fn test_copy_into_with_stage_params() { match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { - from_stage, + from_obj, stage_params, .. } => { //assert_eq!("s3://load/files/", stage_params.url.unwrap()); assert_eq!( - from_stage, - ObjectName::from(vec![Ident::with_quote('\'', "s3://load/files/")]) + from_obj, + Some(ObjectName::from(vec![Ident::with_quote( + '\'', + "s3://load/files/" + )])) ); assert_eq!("myint", stage_params.storage_integration.unwrap()); assert_eq!( @@ -2125,13 +2187,16 @@ fn test_copy_into_with_stage_params() { match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { - from_stage, + from_obj, stage_params, .. } => { assert_eq!( - from_stage, - ObjectName::from(vec![Ident::with_quote('\'', "s3://load/files/")]) + from_obj, + Some(ObjectName::from(vec![Ident::with_quote( + '\'', + "s3://load/files/" + )])) ); assert_eq!("myint", stage_params.storage_integration.unwrap()); } @@ -2154,13 +2219,13 @@ fn test_copy_into_with_files_and_pattern_and_verification() { files, pattern, validation_mode, - from_stage_alias, + from_obj_alias, .. } => { assert_eq!(files.unwrap(), vec!["file1.json", "file2.json"]); assert_eq!(pattern.unwrap(), ".*employees0[1-5].csv.gz"); assert_eq!(validation_mode.unwrap(), "RETURN_7_ROWS"); - assert_eq!(from_stage_alias.unwrap(), Ident::new("some_alias")); + assert_eq!(from_obj_alias.unwrap(), Ident::new("some_alias")); } _ => unreachable!(), } @@ -2179,13 +2244,16 @@ fn test_copy_into_with_transformations() { match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { - from_stage, + from_obj, from_transformations, .. } => { assert_eq!( - from_stage, - ObjectName::from(vec![Ident::new("@schema"), Ident::new("general_finished")]) + from_obj, + Some(ObjectName::from(vec![ + Ident::new("@schema"), + Ident::new("general_finished") + ])) ); assert_eq!( from_transformations.as_ref().unwrap()[0], @@ -2254,6 +2322,41 @@ fn test_copy_into_file_format() { snowflake_without_unescape().verified_stmt(sql).to_string(), sql ); + + // Test commas in file format + let sql = concat!( + "COPY INTO my_company.emp_basic ", + "FROM 'gcs://mybucket/./../a.csv' ", + "FILES = ('file1.json', 'file2.json') ", + "PATTERN = '.*employees0[1-5].csv.gz' ", + r#"FILE_FORMAT=(COMPRESSION=AUTO, BINARY_FORMAT=HEX, ESCAPE='\\')"# + ); + + match snowflake_without_unescape() + .parse_sql_statements(sql) + .unwrap() + .first() + .unwrap() + { + Statement::CopyIntoSnowflake { file_format, .. } => { + assert!(file_format.options.contains(&DataLoadingOption { + option_name: "COMPRESSION".to_string(), + option_type: DataLoadingOptionType::ENUM, + value: "AUTO".to_string() + })); + assert!(file_format.options.contains(&DataLoadingOption { + option_name: "BINARY_FORMAT".to_string(), + option_type: DataLoadingOptionType::ENUM, + value: "HEX".to_string() + })); + assert!(file_format.options.contains(&DataLoadingOption { + option_name: "ESCAPE".to_string(), + option_type: DataLoadingOptionType::STRING, + value: r#"\\"#.to_string() + })); + } + _ => unreachable!(), + } } #[test] @@ -2285,16 +2388,8 @@ fn test_copy_into_copy_options() { } #[test] -fn test_snowflake_stage_object_names() { - let allowed_formatted_names = [ - "my_company.emp_basic", - "@namespace.%table_name", - "@namespace.%table_name/path", - "@namespace.stage_name/path", - "@~/path", - ]; +fn test_snowflake_stage_object_names_into_location() { let mut allowed_object_names = [ - ObjectName::from(vec![Ident::new("my_company"), Ident::new("emp_basic")]), ObjectName::from(vec![Ident::new("@namespace"), Ident::new("%table_name")]), ObjectName::from(vec![ Ident::new("@namespace"), @@ -2307,7 +2402,39 @@ fn test_snowflake_stage_object_names() { ObjectName::from(vec![Ident::new("@~/path")]), ]; - for it in allowed_formatted_names + let allowed_names_into_location = [ + "@namespace.%table_name", + "@namespace.%table_name/path", + "@namespace.stage_name/path", + "@~/path", + ]; + for it in allowed_names_into_location + .iter() + .zip(allowed_object_names.iter_mut()) + { + let (formatted_name, object_name) = it; + let sql = format!( + "COPY INTO {} FROM 'gcs://mybucket/./../a.csv'", + formatted_name + ); + match snowflake().verified_stmt(&sql) { + Statement::CopyIntoSnowflake { into, .. } => { + assert_eq!(into.0, object_name.0) + } + _ => unreachable!(), + } + } +} + +#[test] +fn test_snowflake_stage_object_names_into_table() { + let mut allowed_object_names = [ + ObjectName::from(vec![Ident::new("my_company"), Ident::new("emp_basic")]), + ObjectName::from(vec![Ident::new("emp_basic")]), + ]; + + let allowed_names_into_table = ["my_company.emp_basic", "emp_basic"]; + for it in allowed_names_into_table .iter() .zip(allowed_object_names.iter_mut()) { @@ -2330,16 +2457,17 @@ fn test_snowflake_copy_into() { let sql = "COPY INTO a.b FROM @namespace.stage_name"; assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); match snowflake().verified_stmt(sql) { - Statement::CopyIntoSnowflake { - into, from_stage, .. - } => { + Statement::CopyIntoSnowflake { into, from_obj, .. } => { assert_eq!( into, ObjectName::from(vec![Ident::new("a"), Ident::new("b")]) ); assert_eq!( - from_stage, - ObjectName::from(vec![Ident::new("@namespace"), Ident::new("stage_name")]) + from_obj, + Some(ObjectName::from(vec![ + Ident::new("@namespace"), + Ident::new("stage_name") + ])) ) } _ => unreachable!(), @@ -2351,9 +2479,7 @@ fn test_snowflake_copy_into_stage_name_ends_with_parens() { let sql = "COPY INTO SCHEMA.SOME_MONITORING_SYSTEM FROM (SELECT t.$1:st AS st FROM @schema.general_finished)"; assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); match snowflake().verified_stmt(sql) { - Statement::CopyIntoSnowflake { - into, from_stage, .. - } => { + Statement::CopyIntoSnowflake { into, from_obj, .. } => { assert_eq!( into, ObjectName::from(vec![ @@ -2362,8 +2488,11 @@ fn test_snowflake_copy_into_stage_name_ends_with_parens() { ]) ); assert_eq!( - from_stage, - ObjectName::from(vec![Ident::new("@schema"), Ident::new("general_finished")]) + from_obj, + Some(ObjectName::from(vec![ + Ident::new("@schema"), + Ident::new("general_finished") + ])) ) } _ => unreachable!(), From 443f492b4b9443d45266c77b073b50b91bcb95ed Mon Sep 17 00:00:00 2001 From: Hans Ott Date: Wed, 5 Feb 2025 20:47:17 +0100 Subject: [PATCH 036/286] Require space after -- to start single line comment in MySQL (#1705) --- src/dialect/mod.rs | 9 ++++ src/dialect/mysql.rs | 4 ++ src/tokenizer.rs | 105 ++++++++++++++++++++++++++++++++++++--- tests/sqlparser_mysql.rs | 15 ++++++ 4 files changed, 127 insertions(+), 6 deletions(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 205395f6c..965e6c77f 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -881,6 +881,15 @@ pub trait Dialect: Debug + Any { fn supports_table_hints(&self) -> bool { false } + + /// Returns true if this dialect requires a whitespace character after `--` to start a single line comment. + /// + /// MySQL: + /// e.g. UPDATE account SET balance=balance--1 + // WHERE account_id=5752 ^^^ will be interpreted as two minus signs instead of a comment + fn requires_single_line_comment_whitespace(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index a67fe67b0..55b91ad22 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -125,6 +125,10 @@ impl Dialect for MySqlDialect { fn supports_table_hints(&self) -> bool { true } + + fn requires_single_line_comment_whitespace(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 7742e8fae..d4e530c9d 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -1229,14 +1229,26 @@ impl<'a> Tokenizer<'a> { // operators '-' => { chars.next(); // consume the '-' + match chars.peek() { Some('-') => { - chars.next(); // consume the second '-', starting a single-line comment - let comment = self.tokenize_single_line_comment(chars); - Ok(Some(Token::Whitespace(Whitespace::SingleLineComment { - prefix: "--".to_owned(), - comment, - }))) + let mut is_comment = true; + if self.dialect.requires_single_line_comment_whitespace() { + is_comment = Some(' ') == chars.peekable.clone().nth(1); + } + + if is_comment { + chars.next(); // consume second '-' + let comment = self.tokenize_single_line_comment(chars); + return Ok(Some(Token::Whitespace( + Whitespace::SingleLineComment { + prefix: "--".to_owned(), + comment, + }, + ))); + } + + self.start_binop(chars, "-", Token::Minus) } Some('>') => { chars.next(); @@ -3685,4 +3697,85 @@ mod tests { ], ); } + + #[test] + fn test_whitespace_required_after_single_line_comment() { + all_dialects_where(|dialect| dialect.requires_single_line_comment_whitespace()) + .tokenizes_to( + "SELECT --'abc'", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Minus, + Token::Minus, + Token::SingleQuotedString("abc".to_string()), + ], + ); + + all_dialects_where(|dialect| dialect.requires_single_line_comment_whitespace()) + .tokenizes_to( + "SELECT -- 'abc'", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Whitespace(Whitespace::SingleLineComment { + prefix: "--".to_string(), + comment: " 'abc'".to_string(), + }), + ], + ); + + all_dialects_where(|dialect| dialect.requires_single_line_comment_whitespace()) + .tokenizes_to( + "SELECT --", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Minus, + Token::Minus, + ], + ); + } + + #[test] + fn test_whitespace_not_required_after_single_line_comment() { + all_dialects_where(|dialect| !dialect.requires_single_line_comment_whitespace()) + .tokenizes_to( + "SELECT --'abc'", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Whitespace(Whitespace::SingleLineComment { + prefix: "--".to_string(), + comment: "'abc'".to_string(), + }), + ], + ); + + all_dialects_where(|dialect| !dialect.requires_single_line_comment_whitespace()) + .tokenizes_to( + "SELECT -- 'abc'", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Whitespace(Whitespace::SingleLineComment { + prefix: "--".to_string(), + comment: " 'abc'".to_string(), + }), + ], + ); + + all_dialects_where(|dialect| !dialect.requires_single_line_comment_whitespace()) + .tokenizes_to( + "SELECT --", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Whitespace(Whitespace::SingleLineComment { + prefix: "--".to_string(), + comment: "".to_string(), + }), + ], + ); + } } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 9f00a9213..6bf9076d2 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3250,3 +3250,18 @@ fn parse_double_precision() { "CREATE TABLE foo (bar DOUBLE(11,0))", ); } + +#[test] +fn parse_looks_like_single_line_comment() { + mysql().one_statement_parses_to( + "UPDATE account SET balance=balance--1 WHERE account_id=5752", + "UPDATE account SET balance = balance - -1 WHERE account_id = 5752", + ); + mysql().one_statement_parses_to( + r#" + UPDATE account SET balance=balance-- 1 + WHERE account_id=5752 + "#, + "UPDATE account SET balance = balance WHERE account_id = 5752", + ); +} From 0b8ba91156a27ffd56a20339fa89a927611aa806 Mon Sep 17 00:00:00 2001 From: DanCodedThis <94703934+DanCodedThis@users.noreply.github.com> Date: Thu, 6 Feb 2025 19:11:02 +0200 Subject: [PATCH 037/286] Add suppport for Show Objects statement for the Snowflake parser (#1702) Co-authored-by: Denys Tsomenko Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 25 ++++++++++++++++++++ src/ast/spans.rs | 1 + src/dialect/snowflake.rs | 25 +++++++++++++++++++- src/keywords.rs | 1 + src/parser/mod.rs | 2 +- tests/sqlparser_snowflake.rs | 45 ++++++++++++++++++++++++++++++++++++ 6 files changed, 97 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index dc944c9e7..de2bbafc2 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3010,6 +3010,12 @@ pub enum Statement { show_options: ShowStatementOptions, }, /// ```sql + /// SHOW OBJECTS LIKE 'line%' IN mydb.public + /// ``` + /// Snowflake-specific statement + /// + ShowObjects(ShowObjects), + /// ```sql /// SHOW TABLES /// ``` ShowTables { @@ -4703,6 +4709,17 @@ impl fmt::Display for Statement { )?; Ok(()) } + Statement::ShowObjects(ShowObjects { + terse, + show_options, + }) => { + write!( + f, + "SHOW {terse}OBJECTS{show_options}", + terse = if *terse { "TERSE " } else { "" }, + )?; + Ok(()) + } Statement::ShowTables { terse, history, @@ -8343,6 +8360,14 @@ impl fmt::Display for ShowStatementIn { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ShowObjects { + pub terse: bool, + pub show_options: ShowStatementOptions, +} + /// MSSQL's json null clause /// /// ```plaintext diff --git a/src/ast/spans.rs b/src/ast/spans.rs index f0c38942e..1af5387c9 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -496,6 +496,7 @@ impl Spanned for Statement { Statement::DropConnector { .. } => Span::empty(), Statement::ShowDatabases { .. } => Span::empty(), Statement::ShowSchemas { .. } => Span::empty(), + Statement::ShowObjects { .. } => Span::empty(), Statement::ShowViews { .. } => Span::empty(), Statement::LISTEN { .. } => Span::empty(), Statement::NOTIFY { .. } => Span::empty(), diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index fc1926719..68166cbef 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -25,7 +25,7 @@ use crate::ast::helpers::stmt_data_loading::{ use crate::ast::{ ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, - IdentityPropertyOrder, ObjectName, RowAccessPolicy, Statement, TagsColumnOption, + IdentityPropertyOrder, ObjectName, RowAccessPolicy, ShowObjects, Statement, TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; @@ -185,6 +185,19 @@ impl Dialect for SnowflakeDialect { return Some(parse_file_staging_command(kw, parser)); } + if parser.parse_keyword(Keyword::SHOW) { + let terse = parser.parse_keyword(Keyword::TERSE); + if parser.parse_keyword(Keyword::OBJECTS) { + return Some(parse_show_objects(terse, parser)); + } + //Give back Keyword::TERSE + if terse { + parser.prev_token(); + } + //Give back Keyword::SHOW + parser.prev_token(); + } + None } @@ -1092,3 +1105,13 @@ fn parse_column_tags(parser: &mut Parser, with: bool) -> Result +fn parse_show_objects(terse: bool, parser: &mut Parser) -> Result { + let show_options = parser.parse_show_stmt_options()?; + Ok(Statement::ShowObjects(ShowObjects { + terse, + show_options, + })) +} diff --git a/src/keywords.rs b/src/keywords.rs index 5f36fa737..0178bf555 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -588,6 +588,7 @@ define_keywords!( NUMERIC, NVARCHAR, OBJECT, + OBJECTS, OCCURRENCES_REGEX, OCTETS, OCTET_LENGTH, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6d84ff843..ca7924155 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -14231,7 +14231,7 @@ impl<'a> Parser<'a> { false } - fn parse_show_stmt_options(&mut self) -> Result { + pub(crate) fn parse_show_stmt_options(&mut self) -> Result { let show_in; let mut filter_position = None; if self.dialect.supports_show_like_before_in() { diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index a18f1a4d8..ffcd4e69d 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3083,6 +3083,7 @@ fn test_parentheses_overflow() { #[test] fn test_show_databases() { snowflake().verified_stmt("SHOW DATABASES"); + snowflake().verified_stmt("SHOW TERSE DATABASES"); snowflake().verified_stmt("SHOW DATABASES HISTORY"); snowflake().verified_stmt("SHOW DATABASES LIKE '%abc%'"); snowflake().verified_stmt("SHOW DATABASES STARTS WITH 'demo_db'"); @@ -3095,6 +3096,7 @@ fn test_show_databases() { #[test] fn test_parse_show_schemas() { snowflake().verified_stmt("SHOW SCHEMAS"); + snowflake().verified_stmt("SHOW TERSE SCHEMAS"); snowflake().verified_stmt("SHOW SCHEMAS IN ACCOUNT"); snowflake().verified_stmt("SHOW SCHEMAS IN ACCOUNT abc"); snowflake().verified_stmt("SHOW SCHEMAS IN DATABASE"); @@ -3104,9 +3106,51 @@ fn test_parse_show_schemas() { snowflake().verified_stmt("SHOW SCHEMAS IN DATABASE STARTS WITH 'abc' LIMIT 20 FROM 'xyz'"); } +#[test] +fn test_parse_show_objects() { + snowflake().verified_stmt("SHOW OBJECTS"); + snowflake().verified_stmt("SHOW OBJECTS IN abc"); + snowflake().verified_stmt("SHOW OBJECTS LIKE '%test%' IN abc"); + snowflake().verified_stmt("SHOW OBJECTS IN ACCOUNT"); + snowflake().verified_stmt("SHOW OBJECTS IN DATABASE"); + snowflake().verified_stmt("SHOW OBJECTS IN DATABASE abc"); + snowflake().verified_stmt("SHOW OBJECTS IN SCHEMA"); + snowflake().verified_stmt("SHOW OBJECTS IN SCHEMA abc"); + snowflake().verified_stmt("SHOW TERSE OBJECTS"); + snowflake().verified_stmt("SHOW TERSE OBJECTS IN abc"); + snowflake().verified_stmt("SHOW TERSE OBJECTS LIKE '%test%' IN abc"); + snowflake().verified_stmt("SHOW TERSE OBJECTS LIKE '%test%' IN abc STARTS WITH 'b'"); + snowflake().verified_stmt("SHOW TERSE OBJECTS LIKE '%test%' IN abc STARTS WITH 'b' LIMIT 10"); + snowflake() + .verified_stmt("SHOW TERSE OBJECTS LIKE '%test%' IN abc STARTS WITH 'b' LIMIT 10 FROM 'x'"); + match snowflake().verified_stmt("SHOW TERSE OBJECTS LIKE '%test%' IN abc") { + Statement::ShowObjects(ShowObjects { + terse, + show_options, + }) => { + assert!(terse); + let name = match show_options.show_in { + Some(ShowStatementIn { + parent_name: Some(val), + .. + }) => val.to_string(), + _ => unreachable!(), + }; + assert_eq!("abc", name); + let like = match show_options.filter_position { + Some(ShowStatementFilterPosition::Infix(ShowStatementFilter::Like(val))) => val, + _ => unreachable!(), + }; + assert_eq!("%test%", like); + } + _ => unreachable!(), + } +} + #[test] fn test_parse_show_tables() { snowflake().verified_stmt("SHOW TABLES"); + snowflake().verified_stmt("SHOW TERSE TABLES"); snowflake().verified_stmt("SHOW TABLES IN ACCOUNT"); snowflake().verified_stmt("SHOW TABLES IN DATABASE"); snowflake().verified_stmt("SHOW TABLES IN DATABASE xyz"); @@ -3129,6 +3173,7 @@ fn test_parse_show_tables() { #[test] fn test_show_views() { snowflake().verified_stmt("SHOW VIEWS"); + snowflake().verified_stmt("SHOW TERSE VIEWS"); snowflake().verified_stmt("SHOW VIEWS IN ACCOUNT"); snowflake().verified_stmt("SHOW VIEWS IN DATABASE"); snowflake().verified_stmt("SHOW VIEWS IN DATABASE xyz"); From 86abbd6028d72cee5fa299bc171f492fb722157c Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 6 Feb 2025 18:14:03 +0100 Subject: [PATCH 038/286] Fix incorrect parsing of JsonAccess bracket notation after cast in Snowflake (#1708) --- src/dialect/duckdb.rs | 5 +++++ src/dialect/generic.rs | 4 ++++ src/dialect/mod.rs | 6 ++++++ src/dialect/postgresql.rs | 5 +++++ src/parser/mod.rs | 17 +++++++---------- tests/sqlparser_snowflake.rs | 26 ++++++++++++++++++++++++++ 6 files changed, 53 insertions(+), 10 deletions(-) diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index c41aec81d..e8dcf853c 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -80,4 +80,9 @@ impl Dialect for DuckDbDialect { fn supports_load_extension(&self) -> bool { true } + + // See DuckDB + fn supports_array_typedef_size(&self) -> bool { + true + } } diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 4021b5753..7aaf00742 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -143,4 +143,8 @@ impl Dialect for GenericDialect { fn supports_string_escape_constant(&self) -> bool { true } + + fn supports_array_typedef_size(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 965e6c77f..6b04bacc1 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -890,6 +890,12 @@ pub trait Dialect: Debug + Any { fn requires_single_line_comment_whitespace(&self) -> bool { false } + + /// Returns true if the dialect supports size definition for array types. + /// For example: ```CREATE TABLE my_table (my_array INT[3])```. + fn supports_array_typedef_size(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 5ce4250fb..74b963e82 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -253,6 +253,11 @@ impl Dialect for PostgreSqlDialect { fn supports_numeric_literal_underscores(&self) -> bool { true } + + /// See: + fn supports_array_typedef_size(&self) -> bool { + true + } } pub fn parse_create(parser: &mut Parser) -> Option> { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ca7924155..30124bbc2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8943,16 +8943,13 @@ impl<'a> Parser<'a> { _ => self.expected_at("a data type name", next_token_index), }?; - // Parse array data types. Note: this is postgresql-specific and different from - // Keyword::ARRAY syntax from above - while self.consume_token(&Token::LBracket) { - let size = if dialect_of!(self is GenericDialect | DuckDbDialect | PostgreSqlDialect) { - self.maybe_parse(|p| p.parse_literal_uint())? - } else { - None - }; - self.expect_token(&Token::RBracket)?; - data = DataType::Array(ArrayElemTypeDef::SquareBracket(Box::new(data), size)) + if self.dialect.supports_array_typedef_size() { + // Parse array data type size + while self.consume_token(&Token::LBracket) { + let size = self.maybe_parse(|p| p.parse_literal_uint())?; + self.expect_token(&Token::RBracket)?; + data = DataType::Array(ArrayElemTypeDef::SquareBracket(Box::new(data), size)) + } } Ok((data, trailing_bracket)) } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index ffcd4e69d..b0c215a5a 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -1256,6 +1256,32 @@ fn parse_semi_structured_data_traversal() { .to_string(), "sql parser error: Expected: variant object key name, found: 42" ); + + // casting a json access and accessing an array element + assert_eq!( + snowflake().verified_expr("a:b::ARRAY[1]"), + Expr::JsonAccess { + value: Box::new(Expr::Cast { + kind: CastKind::DoubleColon, + data_type: DataType::Array(ArrayElemTypeDef::None), + format: None, + expr: Box::new(Expr::JsonAccess { + value: Box::new(Expr::Identifier(Ident::new("a"))), + path: JsonPath { + path: vec![JsonPathElem::Dot { + key: "b".to_string(), + quoted: false + }] + } + }) + }), + path: JsonPath { + path: vec![JsonPathElem::Bracket { + key: Expr::Value(number("1")) + }] + } + } + ); } #[test] From cad49232c1247ed942de985ac958212071946528 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 7 Feb 2025 22:24:29 -0800 Subject: [PATCH 039/286] Parse Postgres VARBIT datatype (#1703) --- src/ast/data_type.rs | 8 +++++++- src/keywords.rs | 1 + src/parser/mod.rs | 1 + tests/sqlparser_postgres.rs | 26 ++++++++++++++++++++++++++ 4 files changed, 35 insertions(+), 1 deletion(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 1f2b6be97..4f1f21e0b 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -328,10 +328,15 @@ pub enum DataType { /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/bit-type.html /// [MSSQL]: https://learn.microsoft.com/en-us/sql/t-sql/data-types/bit-transact-sql?view=sql-server-ver16 Bit(Option), - /// Variable-length bit string e.g. [Postgres] + /// `BIT VARYING(n)`: Variable-length bit string e.g. [Postgres] /// /// [Postgres]: https://www.postgresql.org/docs/current/datatype-bit.html BitVarying(Option), + /// `VARBIT(n)`: Variable-length bit string. [Postgres] alias for `BIT VARYING` + /// + /// [Postgres]: https://www.postgresql.org/docs/current/datatype.html + VarBit(Option), + /// /// Custom type such as enums Custom(ObjectName, Vec), /// Arrays @@ -550,6 +555,7 @@ impl fmt::Display for DataType { DataType::BitVarying(size) => { format_type_with_optional_length(f, "BIT VARYING", size, false) } + DataType::VarBit(size) => format_type_with_optional_length(f, "VARBIT", size, false), DataType::Array(ty) => match ty { ArrayElemTypeDef::None => write!(f, "ARRAY"), ArrayElemTypeDef::SquareBracket(t, None) => write!(f, "{t}[]"), diff --git a/src/keywords.rs b/src/keywords.rs index 0178bf555..a67f6bc76 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -925,6 +925,7 @@ define_keywords!( VALUES, VALUE_OF, VARBINARY, + VARBIT, VARCHAR, VARIABLES, VARYING, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 30124bbc2..de2dc6602 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8779,6 +8779,7 @@ impl<'a> Parser<'a> { Ok(DataType::Bit(self.parse_optional_precision()?)) } } + Keyword::VARBIT => Ok(DataType::VarBit(self.parse_optional_precision()?)), Keyword::UUID => Ok(DataType::Uuid), Keyword::DATE => Ok(DataType::Date), Keyword::DATE32 => Ok(DataType::Date32), diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 62da0f574..d8d97b491 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5327,3 +5327,29 @@ fn parse_bitstring_literal() { ))] ); } + +#[test] +fn parse_varbit_datatype() { + match pg_and_generic().verified_stmt("CREATE TABLE foo (x VARBIT, y VARBIT(42))") { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ + ColumnDef { + name: "x".into(), + data_type: DataType::VarBit(None), + collation: None, + options: vec![], + }, + ColumnDef { + name: "y".into(), + data_type: DataType::VarBit(Some(42)), + collation: None, + options: vec![], + } + ] + ); + } + _ => unreachable!(), + } +} From 46cfcfe8f7afc63356930fd7e3ff1c549ebbf41e Mon Sep 17 00:00:00 2001 From: Armin Ronacher Date: Sun, 9 Feb 2025 06:10:58 +0100 Subject: [PATCH 040/286] Implement FROM-first selects (#1713) --- src/ast/mod.rs | 4 +- src/ast/query.rs | 29 +++++++++++++- src/ast/spans.rs | 1 + src/dialect/clickhouse.rs | 4 ++ src/dialect/duckdb.rs | 4 ++ src/dialect/mod.rs | 11 ++++++ src/parser/mod.rs | 52 ++++++++++++++++++++++--- tests/sqlparser_clickhouse.rs | 1 + tests/sqlparser_common.rs | 71 +++++++++++++++++++++++++++++++++++ tests/sqlparser_duckdb.rs | 2 + tests/sqlparser_mssql.rs | 3 ++ tests/sqlparser_mysql.rs | 8 ++++ tests/sqlparser_postgres.rs | 3 ++ 13 files changed, 184 insertions(+), 9 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index de2bbafc2..f693ab7d8 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -68,8 +68,8 @@ pub use self::query::{ NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, OffsetRows, OpenJsonTableColumn, OrderBy, OrderByExpr, PivotValueSource, ProjectionSelect, Query, RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, Select, - SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, SetOperator, SetQuantifier, - Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, + SelectFlavor, SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, SetOperator, + SetQuantifier, Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, TableFunctionArgs, TableIndexHintForClause, TableIndexHintType, TableIndexHints, TableIndexType, TableSample, TableSampleBucket, TableSampleKind, TableSampleMethod, TableSampleModifier, TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, diff --git a/src/ast/query.rs b/src/ast/query.rs index 239e14554..0446bcb75 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -275,6 +275,19 @@ impl fmt::Display for Table { } } +/// What did this select look like? +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum SelectFlavor { + /// `SELECT *` + Standard, + /// `FROM ... SELECT *` + FromFirst, + /// `FROM *` + FromFirstNoSelect, +} + /// A restricted variant of `SELECT` (without CTEs/`ORDER BY`), which may /// appear either as the only body item of a `Query`, or as an operand /// to a set operation like `UNION`. @@ -328,11 +341,23 @@ pub struct Select { pub value_table_mode: Option, /// STARTING WITH .. CONNECT BY pub connect_by: Option, + /// Was this a FROM-first query? + pub flavor: SelectFlavor, } impl fmt::Display for Select { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "SELECT")?; + match self.flavor { + SelectFlavor::Standard => { + write!(f, "SELECT")?; + } + SelectFlavor::FromFirst => { + write!(f, "FROM {} SELECT", display_comma_separated(&self.from))?; + } + SelectFlavor::FromFirstNoSelect => { + write!(f, "FROM {}", display_comma_separated(&self.from))?; + } + } if let Some(value_table_mode) = self.value_table_mode { write!(f, " {value_table_mode}")?; @@ -360,7 +385,7 @@ impl fmt::Display for Select { write!(f, " {into}")?; } - if !self.from.is_empty() { + if self.flavor == SelectFlavor::Standard && !self.from.is_empty() { write!(f, " FROM {}", display_comma_separated(&self.from))?; } if !self.lateral_views.is_empty() { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 1af5387c9..18f6f6e2f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2077,6 +2077,7 @@ impl Spanned for Select { value_table_mode: _, // todo, BigQuery specific connect_by, top_before_distinct: _, + flavor: _, } = self; union_spans( diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs index 9a0884a51..7d479219d 100644 --- a/src/dialect/clickhouse.rs +++ b/src/dialect/clickhouse.rs @@ -75,4 +75,8 @@ impl Dialect for ClickHouseDialect { fn supports_lambda_functions(&self) -> bool { true } + + fn supports_from_first_select(&self) -> bool { + true + } } diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index e8dcf853c..a2e7fb6c0 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -85,4 +85,8 @@ impl Dialect for DuckDbDialect { fn supports_array_typedef_size(&self) -> bool { true } + + fn supports_from_first_select(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 6b04bacc1..933203605 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -463,6 +463,17 @@ pub trait Dialect: Debug + Any { false } + /// Return true if the dialect supports "FROM-first" selects. + /// + /// Example: + /// ```sql + /// FROM table + /// SELECT * + /// ``` + fn supports_from_first_select(&self) -> bool { + false + } + /// Does the dialect support MySQL-style `'user'@'host'` grantee syntax? fn supports_user_host_grantee(&self) -> bool { false diff --git a/src/parser/mod.rs b/src/parser/mod.rs index de2dc6602..188e941b5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -528,7 +528,7 @@ impl<'a> Parser<'a> { Keyword::DESCRIBE => self.parse_explain(DescribeAlias::Describe), Keyword::EXPLAIN => self.parse_explain(DescribeAlias::Explain), Keyword::ANALYZE => self.parse_analyze(), - Keyword::SELECT | Keyword::WITH | Keyword::VALUES => { + Keyword::SELECT | Keyword::WITH | Keyword::VALUES | Keyword::FROM => { self.prev_token(); self.parse_query().map(Statement::Query) } @@ -10218,7 +10218,9 @@ impl<'a> Parser<'a> { pub fn parse_query_body(&mut self, precedence: u8) -> Result, ParserError> { // We parse the expression using a Pratt parser, as in `parse_expr()`. // Start by parsing a restricted SELECT or a `(subquery)`: - let expr = if self.peek_keyword(Keyword::SELECT) { + let expr = if self.peek_keyword(Keyword::SELECT) + || (self.peek_keyword(Keyword::FROM) && self.dialect.supports_from_first_select()) + { SetExpr::Select(self.parse_select().map(Box::new)?) } else if self.consume_token(&Token::LParen) { // CTEs are not allowed here, but the parser currently accepts them @@ -10317,6 +10319,39 @@ impl<'a> Parser<'a> { /// Parse a restricted `SELECT` statement (no CTEs / `UNION` / `ORDER BY`) pub fn parse_select(&mut self) -> Result { + let mut from_first = None; + + if self.dialect.supports_from_first_select() && self.peek_keyword(Keyword::FROM) { + let from_token = self.expect_keyword(Keyword::FROM)?; + let from = self.parse_table_with_joins()?; + if !self.peek_keyword(Keyword::SELECT) { + return Ok(Select { + select_token: AttachedToken(from_token), + distinct: None, + top: None, + top_before_distinct: false, + projection: vec![], + into: None, + from, + lateral_views: vec![], + prewhere: None, + selection: None, + group_by: GroupByExpr::Expressions(vec![], vec![]), + cluster_by: vec![], + distribute_by: vec![], + sort_by: vec![], + having: None, + named_window: vec![], + window_before_qualify: false, + qualify: None, + value_table_mode: None, + connect_by: None, + flavor: SelectFlavor::FromFirstNoSelect, + }); + } + from_first = Some(from); + } + let select_token = self.expect_keyword(Keyword::SELECT)?; let value_table_mode = if dialect_of!(self is BigQueryDialect) && self.parse_keyword(Keyword::AS) { @@ -10371,10 +10406,12 @@ impl<'a> Parser<'a> { // otherwise they may be parsed as an alias as part of the `projection` // or `from`. - let from = if self.parse_keyword(Keyword::FROM) { - self.parse_table_with_joins()? + let (from, from_first) = if let Some(from) = from_first.take() { + (from, true) + } else if self.parse_keyword(Keyword::FROM) { + (self.parse_table_with_joins()?, false) } else { - vec![] + (vec![], false) }; let mut lateral_views = vec![]; @@ -10506,6 +10543,11 @@ impl<'a> Parser<'a> { qualify, value_table_mode, connect_by, + flavor: if from_first { + SelectFlavor::FromFirst + } else { + SelectFlavor::Standard + }, }) } diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 0f22db389..76a56e709 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -101,6 +101,7 @@ fn parse_map_access_expr() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }, select ); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 3a6183a15..f8c0a26b8 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -461,6 +461,7 @@ fn parse_update_set_from() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -5289,6 +5290,7 @@ fn test_parse_named_window() { window_before_qualify: true, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }; assert_eq!(actual_select_only, expected); } @@ -5915,6 +5917,7 @@ fn parse_interval_and_or_xor() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -8022,6 +8025,7 @@ fn lateral_function() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }; assert_eq!(actual_select_only, expected); } @@ -8919,6 +8923,7 @@ fn parse_merge() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -10703,6 +10708,7 @@ fn parse_unload() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), with: None, limit: None, @@ -10913,6 +10919,7 @@ fn parse_connect_by() { ))))), }], }), + flavor: SelectFlavor::Standard, }; let connect_by_1 = concat!( @@ -10997,6 +11004,7 @@ fn parse_connect_by() { ))))), }], }), + flavor: SelectFlavor::Standard, } ); @@ -11860,6 +11868,7 @@ fn test_extract_seconds_ok() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -13592,3 +13601,65 @@ fn test_lambdas() { ); dialects.verified_expr("transform(array(1, 2, 3), x -> x + 1)"); } + +#[test] +fn test_select_from_first() { + let dialects = all_dialects_where(|d| d.supports_from_first_select()); + let q1 = "FROM capitals"; + let q2 = "FROM capitals SELECT *"; + + for (q, flavor, projection) in [ + (q1, SelectFlavor::FromFirstNoSelect, vec![]), + ( + q2, + SelectFlavor::FromFirst, + vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + ), + ] { + let ast = dialects.verified_query(q); + let expected = Query { + with: None, + body: Box::new(SetExpr::Select(Box::new(Select { + select_token: AttachedToken::empty(), + distinct: None, + top: None, + projection, + top_before_distinct: false, + into: None, + from: vec![TableWithJoins { + relation: table_from_name(ObjectName::from(vec![Ident { + value: "capitals".to_string(), + quote_style: None, + span: Span::empty(), + }])), + joins: vec![], + }], + lateral_views: vec![], + prewhere: None, + selection: None, + group_by: GroupByExpr::Expressions(vec![], vec![]), + cluster_by: vec![], + distribute_by: vec![], + sort_by: vec![], + having: None, + named_window: vec![], + window_before_qualify: false, + qualify: None, + value_table_mode: None, + connect_by: None, + flavor, + }))), + order_by: None, + limit: None, + offset: None, + fetch: None, + locks: vec![], + limit_by: vec![], + for_clause: None, + settings: None, + format_clause: None, + }; + assert_eq!(expected, ast); + assert_eq!(ast.to_string(), q); + } +} diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 4289ebd1f..43e12746c 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -288,6 +288,7 @@ fn test_select_union_by_name() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), right: Box::::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), @@ -317,6 +318,7 @@ fn test_select_union_by_name() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), }); assert_eq!(ast.body, expected); diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 9046e9e74..e6a12b9c7 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -137,6 +137,7 @@ fn parse_create_procedure() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))) }))], params: Some(vec![ @@ -1114,6 +1115,7 @@ fn parse_substring_in_select() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -1251,6 +1253,7 @@ fn parse_mssql_declare() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))) })) ], diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 6bf9076d2..48fcbbf9b 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1111,6 +1111,7 @@ fn parse_escaped_quote_identifiers_with_escape() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -1164,6 +1165,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -1211,6 +1213,7 @@ fn parse_escaped_backticks_with_escape() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -1262,6 +1265,7 @@ fn parse_escaped_backticks_with_no_escape() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -1931,6 +1935,7 @@ fn parse_select_with_numeric_prefix_column_name() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))) ); } @@ -1983,6 +1988,7 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))) ); } @@ -2503,6 +2509,7 @@ fn parse_substring_in_select() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -2799,6 +2806,7 @@ fn parse_hex_string_introducer() { value_table_mode: None, into: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index d8d97b491..3a4504ebb 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1315,6 +1315,7 @@ fn parse_copy_to() { qualify: None, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), order_by: None, limit: None, @@ -2666,6 +2667,7 @@ fn parse_array_subquery_expr() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), right: Box::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), @@ -2688,6 +2690,7 @@ fn parse_array_subquery_expr() { window_before_qualify: false, value_table_mode: None, connect_by: None, + flavor: SelectFlavor::Standard, }))), }), order_by: None, From 322209a9f34d447464e0eff46627785f02869fb3 Mon Sep 17 00:00:00 2001 From: Justin Joyce Date: Mon, 10 Feb 2025 05:51:22 +0000 Subject: [PATCH 041/286] Enable custom dialects to support `MATCH() AGAINST()` (#1719) --- src/dialect/generic.rs | 4 ++++ src/dialect/mod.rs | 5 +++++ src/dialect/mysql.rs | 4 ++++ src/parser/mod.rs | 2 +- 4 files changed, 14 insertions(+), 1 deletion(-) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 7aaf00742..dbd5cab28 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -147,4 +147,8 @@ impl Dialect for GenericDialect { fn supports_array_typedef_size(&self) -> bool { true } + + fn supports_match_against(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 933203605..a57c25d5e 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -479,6 +479,11 @@ pub trait Dialect: Debug + Any { false } + /// Does the dialect support the `MATCH() AGAINST()` syntax? + fn supports_match_against(&self) -> bool { + false + } + /// Dialect-specific infix parser override /// /// This method is called to parse the next infix expression. diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 55b91ad22..8a0da87e4 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -129,6 +129,10 @@ impl Dialect for MySqlDialect { fn requires_single_line_comment_whitespace(&self) -> bool { true } + + fn supports_match_against(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 188e941b5..c4f65a1a0 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1198,7 +1198,7 @@ impl<'a> Parser<'a> { }))) } Keyword::NOT => Ok(Some(self.parse_not()?)), - Keyword::MATCH if dialect_of!(self is MySqlDialect | GenericDialect) => { + Keyword::MATCH if self.dialect.supports_match_against() => { Ok(Some(self.parse_match_against()?)) } Keyword::STRUCT if self.dialect.supports_struct_literal() => { From 3e94877f03c5b47659093a2d55c3bc04e042013c Mon Sep 17 00:00:00 2001 From: Emil Date: Tue, 11 Feb 2025 16:22:50 +0100 Subject: [PATCH 042/286] Support group by cube/rollup etc in BigQuery (#1720) --- src/dialect/bigquery.rs | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index bb1a0d5ce..5354a645b 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -115,6 +115,11 @@ impl Dialect for BigQueryDialect { true } + // See + fn supports_group_by_expr(&self) -> bool { + true + } + fn is_column_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { !RESERVED_FOR_COLUMN_ALIAS.contains(kw) } From a5bbb5e8ac52c98fe0f3e85e3b25c24048deade9 Mon Sep 17 00:00:00 2001 From: Tyler Brinks Date: Thu, 13 Feb 2025 03:40:35 -0700 Subject: [PATCH 043/286] Add support for MS Varbinary(MAX) (#1714) (#1715) --- src/ast/data_type.rs | 44 ++++++++++++++++++++++++++++--- src/ast/mod.rs | 4 +-- src/parser/mod.rs | 20 +++++++++++++- tests/sqlparser_common.rs | 2 +- tests/sqlparser_mssql.rs | 55 ++++++++++++++++++++++++++++++++++++++- 5 files changed, 116 insertions(+), 9 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 4f1f21e0b..bd46f8bdb 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -85,7 +85,7 @@ pub enum DataType { /// /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-string-type /// [MS SQL Server]: https://learn.microsoft.com/pt-br/sql/t-sql/data-types/binary-and-varbinary-transact-sql?view=sql-server-ver16 - Varbinary(Option), + Varbinary(Option), /// Large binary object with optional length e.g. BLOB, BLOB(1000), [standard], [Oracle] /// /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-large-object-string-type @@ -408,9 +408,7 @@ impl fmt::Display for DataType { } DataType::Clob(size) => format_type_with_optional_length(f, "CLOB", size, false), DataType::Binary(size) => format_type_with_optional_length(f, "BINARY", size, false), - DataType::Varbinary(size) => { - format_type_with_optional_length(f, "VARBINARY", size, false) - } + DataType::Varbinary(size) => format_varbinary_type(f, "VARBINARY", size), DataType::Blob(size) => format_type_with_optional_length(f, "BLOB", size, false), DataType::TinyBlob => write!(f, "TINYBLOB"), DataType::MediumBlob => write!(f, "MEDIUMBLOB"), @@ -673,6 +671,18 @@ fn format_character_string_type( Ok(()) } +fn format_varbinary_type( + f: &mut fmt::Formatter, + sql_type: &str, + size: &Option, +) -> fmt::Result { + write!(f, "{sql_type}")?; + if let Some(size) = size { + write!(f, "({size})")?; + } + Ok(()) +} + fn format_datetime_precision_and_tz( f: &mut fmt::Formatter, sql_type: &'static str, @@ -862,6 +872,32 @@ impl fmt::Display for CharLengthUnits { } } +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum BinaryLength { + IntegerLength { + /// Default (if VARYING) + length: u64, + }, + /// VARBINARY(MAX) used in T-SQL (Microsoft SQL Server) + Max, +} + +impl fmt::Display for BinaryLength { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + BinaryLength::IntegerLength { length } => { + write!(f, "{}", length)?; + } + BinaryLength::Max => { + write!(f, "MAX")?; + } + } + Ok(()) + } +} + /// Represents the data type of the elements in an array (if any) as well as /// the syntax used to declare the array. /// diff --git a/src/ast/mod.rs b/src/ast/mod.rs index f693ab7d8..5835447db 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -40,8 +40,8 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::tokenizer::Span; pub use self::data_type::{ - ArrayElemTypeDef, CharLengthUnits, CharacterLength, DataType, EnumMember, ExactNumberInfo, - StructBracketKind, TimezoneInfo, + ArrayElemTypeDef, BinaryLength, CharLengthUnits, CharacterLength, DataType, EnumMember, + ExactNumberInfo, StructBracketKind, TimezoneInfo, }; pub use self::dcl::{ AlterRoleOperation, ResetConfig, RoleOption, SecondaryRoles, SetConfigValue, Use, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c4f65a1a0..6d3290d23 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8766,7 +8766,7 @@ impl<'a> Parser<'a> { } Keyword::CLOB => Ok(DataType::Clob(self.parse_optional_precision()?)), Keyword::BINARY => Ok(DataType::Binary(self.parse_optional_precision()?)), - Keyword::VARBINARY => Ok(DataType::Varbinary(self.parse_optional_precision()?)), + Keyword::VARBINARY => Ok(DataType::Varbinary(self.parse_optional_binary_length()?)), Keyword::BLOB => Ok(DataType::Blob(self.parse_optional_precision()?)), Keyword::TINYBLOB => Ok(DataType::TinyBlob), Keyword::MEDIUMBLOB => Ok(DataType::MediumBlob), @@ -9650,6 +9650,16 @@ impl<'a> Parser<'a> { } } + pub fn parse_optional_binary_length(&mut self) -> Result, ParserError> { + if self.consume_token(&Token::LParen) { + let binary_length = self.parse_binary_length()?; + self.expect_token(&Token::RParen)?; + Ok(Some(binary_length)) + } else { + Ok(None) + } + } + pub fn parse_character_length(&mut self) -> Result { if self.parse_keyword(Keyword::MAX) { return Ok(CharacterLength::Max); @@ -9665,6 +9675,14 @@ impl<'a> Parser<'a> { Ok(CharacterLength::IntegerLength { length, unit }) } + pub fn parse_binary_length(&mut self) -> Result { + if self.parse_keyword(Keyword::MAX) { + return Ok(BinaryLength::Max); + } + let length = self.parse_literal_uint()?; + Ok(BinaryLength::IntegerLength { length }) + } + pub fn parse_optional_precision_scale( &mut self, ) -> Result<(Option, Option), ParserError> { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index f8c0a26b8..aef4d0d72 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2633,7 +2633,7 @@ fn parse_cast() { &Expr::Cast { kind: CastKind::Cast, expr: Box::new(Expr::Identifier(Ident::new("id"))), - data_type: DataType::Varbinary(Some(50)), + data_type: DataType::Varbinary(Some(BinaryLength::IntegerLength { length: 50 })), format: None, }, expr_from_projection(only(&select.projection)) diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index e6a12b9c7..6865bdd4e 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -26,7 +26,7 @@ use helpers::attached_token::AttachedToken; use sqlparser::tokenizer::Span; use test_utils::*; -use sqlparser::ast::DataType::{Int, Text}; +use sqlparser::ast::DataType::{Int, Text, Varbinary}; use sqlparser::ast::DeclareAssignment::MsSqlAssignment; use sqlparser::ast::Value::SingleQuotedString; use sqlparser::ast::*; @@ -1796,6 +1796,59 @@ fn parse_mssql_set_session_value() { ms().verified_stmt("SET ANSI_NULLS, ANSI_PADDING ON"); } +#[test] +fn parse_mssql_varbinary_max_length() { + let sql = "CREATE TABLE example (var_binary_col VARBINARY(MAX))"; + + match ms_and_generic().verified_stmt(sql) { + Statement::CreateTable(CreateTable { name, columns, .. }) => { + assert_eq!( + name, + ObjectName::from(vec![Ident { + value: "example".to_string(), + quote_style: None, + span: Span::empty(), + }]) + ); + assert_eq!( + columns, + vec![ColumnDef { + name: Ident::new("var_binary_col"), + data_type: Varbinary(Some(BinaryLength::Max)), + collation: None, + options: vec![] + },], + ); + } + _ => unreachable!(), + } + + let sql = "CREATE TABLE example (var_binary_col VARBINARY(50))"; + + match ms_and_generic().verified_stmt(sql) { + Statement::CreateTable(CreateTable { name, columns, .. }) => { + assert_eq!( + name, + ObjectName::from(vec![Ident { + value: "example".to_string(), + quote_style: None, + span: Span::empty(), + }]) + ); + assert_eq!( + columns, + vec![ColumnDef { + name: Ident::new("var_binary_col"), + data_type: Varbinary(Some(BinaryLength::IntegerLength { length: 50 })), + collation: None, + options: vec![] + },], + ); + } + _ => unreachable!(), + } +} + fn ms() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {})]) } From 1c0e5d355438343915b86e75956e0103e4f937eb Mon Sep 17 00:00:00 2001 From: wugeer <1284057728@qq.com> Date: Sat, 15 Feb 2025 00:16:10 +0800 Subject: [PATCH 044/286] Add supports for Hive's `SELECT ... GROUP BY .. GROUPING SETS` syntax (#1653) Co-authored-by: Ifeanyi Ubah --- src/ast/query.rs | 10 +++- src/dialect/clickhouse.rs | 10 ++++ src/dialect/generic.rs | 4 ++ src/dialect/hive.rs | 11 +++-- src/dialect/mod.rs | 6 +++ src/parser/mod.rs | 10 +++- tests/sqlparser_clickhouse.rs | 55 ---------------------- tests/sqlparser_common.rs | 86 +++++++++++++++++++++++++++++++++++ 8 files changed, 132 insertions(+), 60 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 0446bcb75..097a4ad4c 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2547,13 +2547,18 @@ impl fmt::Display for SelectInto { /// e.g. GROUP BY year WITH ROLLUP WITH TOTALS /// /// [ClickHouse]: -#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum GroupByWithModifier { Rollup, Cube, Totals, + /// Hive supports GROUP BY GROUPING SETS syntax. + /// e.g. GROUP BY year , month GROUPING SETS((year,month),(year),(month)) + /// + /// [Hive]: + GroupingSets(Expr), } impl fmt::Display for GroupByWithModifier { @@ -2562,6 +2567,9 @@ impl fmt::Display for GroupByWithModifier { GroupByWithModifier::Rollup => write!(f, "WITH ROLLUP"), GroupByWithModifier::Cube => write!(f, "WITH CUBE"), GroupByWithModifier::Totals => write!(f, "WITH TOTALS"), + GroupByWithModifier::GroupingSets(expr) => { + write!(f, "{expr}") + } } } } diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs index 7d479219d..37da2ab5d 100644 --- a/src/dialect/clickhouse.rs +++ b/src/dialect/clickhouse.rs @@ -79,4 +79,14 @@ impl Dialect for ClickHouseDialect { fn supports_from_first_select(&self) -> bool { true } + + // See + fn supports_group_by_expr(&self) -> bool { + true + } + + /// See + fn supports_group_by_with_modifier(&self) -> bool { + true + } } diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index dbd5cab28..e04a288d6 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -48,6 +48,10 @@ impl Dialect for GenericDialect { true } + fn supports_group_by_with_modifier(&self) -> bool { + true + } + fn supports_connect_by(&self) -> bool { true } diff --git a/src/dialect/hive.rs b/src/dialect/hive.rs index 80f44cf7c..3e15d395b 100644 --- a/src/dialect/hive.rs +++ b/src/dialect/hive.rs @@ -52,18 +52,23 @@ impl Dialect for HiveDialect { true } - /// See Hive + /// See fn supports_bang_not_operator(&self) -> bool { true } - /// See Hive + /// See fn supports_load_data(&self) -> bool { true } - /// See Hive + /// See fn supports_table_sample_before_alias(&self) -> bool { true } + + /// See + fn supports_group_by_with_modifier(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index a57c25d5e..031fe9676 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -245,6 +245,12 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialects supports `GROUP BY` modifiers prefixed by a `WITH` keyword. + /// Example: `GROUP BY value WITH ROLLUP`. + fn supports_group_by_with_modifier(&self) -> bool { + false + } + /// Returns true if the dialect supports CONNECT BY. fn supports_connect_by(&self) -> bool { false diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6d3290d23..7c03c4e7e 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9148,7 +9148,7 @@ impl<'a> Parser<'a> { }; let mut modifiers = vec![]; - if dialect_of!(self is ClickHouseDialect | GenericDialect) { + if self.dialect.supports_group_by_with_modifier() { loop { if !self.parse_keyword(Keyword::WITH) { break; @@ -9171,6 +9171,14 @@ impl<'a> Parser<'a> { }); } } + if self.parse_keywords(&[Keyword::GROUPING, Keyword::SETS]) { + self.expect_token(&Token::LParen)?; + let result = self.parse_comma_separated(|p| p.parse_tuple(true, true))?; + self.expect_token(&Token::RParen)?; + modifiers.push(GroupByWithModifier::GroupingSets(Expr::GroupingSets( + result, + ))); + }; let group_by = match expressions { None => GroupByExpr::All(modifiers), Some(exprs) => GroupByExpr::Expressions(exprs, modifiers), diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 76a56e709..b94d6f698 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -1069,61 +1069,6 @@ fn parse_create_materialized_view() { clickhouse_and_generic().verified_stmt(sql); } -#[test] -fn parse_group_by_with_modifier() { - let clauses = ["x", "a, b", "ALL"]; - let modifiers = [ - "WITH ROLLUP", - "WITH CUBE", - "WITH TOTALS", - "WITH ROLLUP WITH CUBE", - ]; - let expected_modifiers = [ - vec![GroupByWithModifier::Rollup], - vec![GroupByWithModifier::Cube], - vec![GroupByWithModifier::Totals], - vec![GroupByWithModifier::Rollup, GroupByWithModifier::Cube], - ]; - for clause in &clauses { - for (modifier, expected_modifier) in modifiers.iter().zip(expected_modifiers.iter()) { - let sql = format!("SELECT * FROM t GROUP BY {clause} {modifier}"); - match clickhouse_and_generic().verified_stmt(&sql) { - Statement::Query(query) => { - let group_by = &query.body.as_select().unwrap().group_by; - if clause == &"ALL" { - assert_eq!(group_by, &GroupByExpr::All(expected_modifier.to_vec())); - } else { - assert_eq!( - group_by, - &GroupByExpr::Expressions( - clause - .split(", ") - .map(|c| Identifier(Ident::new(c))) - .collect(), - expected_modifier.to_vec() - ) - ); - } - } - _ => unreachable!(), - } - } - } - - // invalid cases - let invalid_cases = [ - "SELECT * FROM t GROUP BY x WITH", - "SELECT * FROM t GROUP BY x WITH ROLLUP CUBE", - "SELECT * FROM t GROUP BY x WITH WITH ROLLUP", - "SELECT * FROM t GROUP BY WITH ROLLUP", - ]; - for sql in invalid_cases { - clickhouse_and_generic() - .parse_sql_statements(sql) - .expect_err("Expected: one of ROLLUP or CUBE or TOTALS, found: WITH"); - } -} - #[test] fn parse_select_order_by_with_fill_interpolate() { let sql = "SELECT id, fname, lname FROM customer WHERE id < 5 \ diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index aef4d0d72..e4fa2e837 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2447,6 +2447,92 @@ fn parse_select_group_by_all() { ); } +#[test] +fn parse_group_by_with_modifier() { + let clauses = ["x", "a, b", "ALL"]; + let modifiers = [ + "WITH ROLLUP", + "WITH CUBE", + "WITH TOTALS", + "WITH ROLLUP WITH CUBE", + ]; + let expected_modifiers = [ + vec![GroupByWithModifier::Rollup], + vec![GroupByWithModifier::Cube], + vec![GroupByWithModifier::Totals], + vec![GroupByWithModifier::Rollup, GroupByWithModifier::Cube], + ]; + let dialects = all_dialects_where(|d| d.supports_group_by_with_modifier()); + + for clause in &clauses { + for (modifier, expected_modifier) in modifiers.iter().zip(expected_modifiers.iter()) { + let sql = format!("SELECT * FROM t GROUP BY {clause} {modifier}"); + match dialects.verified_stmt(&sql) { + Statement::Query(query) => { + let group_by = &query.body.as_select().unwrap().group_by; + if clause == &"ALL" { + assert_eq!(group_by, &GroupByExpr::All(expected_modifier.to_vec())); + } else { + assert_eq!( + group_by, + &GroupByExpr::Expressions( + clause + .split(", ") + .map(|c| Identifier(Ident::new(c))) + .collect(), + expected_modifier.to_vec() + ) + ); + } + } + _ => unreachable!(), + } + } + } + + // invalid cases + let invalid_cases = [ + "SELECT * FROM t GROUP BY x WITH", + "SELECT * FROM t GROUP BY x WITH ROLLUP CUBE", + "SELECT * FROM t GROUP BY x WITH WITH ROLLUP", + "SELECT * FROM t GROUP BY WITH ROLLUP", + ]; + for sql in invalid_cases { + dialects + .parse_sql_statements(sql) + .expect_err("Expected: one of ROLLUP or CUBE or TOTALS, found: WITH"); + } +} + +#[test] +fn parse_group_by_special_grouping_sets() { + let sql = "SELECT a, b, SUM(c) FROM tab1 GROUP BY a, b GROUPING SETS ((a, b), (a), (b), ())"; + match all_dialects().verified_stmt(sql) { + Statement::Query(query) => { + let group_by = &query.body.as_select().unwrap().group_by; + assert_eq!( + group_by, + &GroupByExpr::Expressions( + vec![ + Expr::Identifier(Ident::new("a")), + Expr::Identifier(Ident::new("b")) + ], + vec![GroupByWithModifier::GroupingSets(Expr::GroupingSets(vec![ + vec![ + Expr::Identifier(Ident::new("a")), + Expr::Identifier(Ident::new("b")) + ], + vec![Expr::Identifier(Ident::new("a")),], + vec![Expr::Identifier(Ident::new("b"))], + vec![] + ]))] + ) + ); + } + _ => unreachable!(), + } +} + #[test] fn parse_select_having() { let sql = "SELECT foo FROM bar GROUP BY foo HAVING COUNT(*) > 1"; From 68c41a9d5a99eea1f98b254d7c26fdbe0a1a4cf7 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 14 Feb 2025 08:23:00 -0800 Subject: [PATCH 045/286] Differentiate LEFT JOIN from LEFT OUTER JOIN (#1726) --- src/ast/query.rs | 20 +++++++++++-- src/ast/spans.rs | 2 ++ src/parser/mod.rs | 6 ++-- tests/sqlparser_common.rs | 60 +++++++++++++++++++++++++++------------ 4 files changed, 65 insertions(+), 23 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 097a4ad4c..c52a98b63 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2077,20 +2077,34 @@ impl fmt::Display for Join { self.relation, suffix(constraint) ), - JoinOperator::LeftOuter(constraint) => write!( + JoinOperator::Left(constraint) => write!( f, " {}LEFT JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) ), - JoinOperator::RightOuter(constraint) => write!( + JoinOperator::LeftOuter(constraint) => write!( + f, + " {}LEFT OUTER JOIN {}{}", + prefix(constraint), + self.relation, + suffix(constraint) + ), + JoinOperator::Right(constraint) => write!( f, " {}RIGHT JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) ), + JoinOperator::RightOuter(constraint) => write!( + f, + " {}RIGHT OUTER JOIN {}{}", + prefix(constraint), + self.relation, + suffix(constraint) + ), JoinOperator::FullOuter(constraint) => write!( f, " {}FULL JOIN {}{}", @@ -2162,7 +2176,9 @@ impl fmt::Display for Join { pub enum JoinOperator { Join(JoinConstraint), Inner(JoinConstraint), + Left(JoinConstraint), LeftOuter(JoinConstraint), + Right(JoinConstraint), RightOuter(JoinConstraint), FullOuter(JoinConstraint), CrossJoin, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 18f6f6e2f..574478692 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2010,7 +2010,9 @@ impl Spanned for JoinOperator { match self { JoinOperator::Join(join_constraint) => join_constraint.span(), JoinOperator::Inner(join_constraint) => join_constraint.span(), + JoinOperator::Left(join_constraint) => join_constraint.span(), JoinOperator::LeftOuter(join_constraint) => join_constraint.span(), + JoinOperator::Right(join_constraint) => join_constraint.span(), JoinOperator::RightOuter(join_constraint) => join_constraint.span(), JoinOperator::FullOuter(join_constraint) => join_constraint.span(), JoinOperator::CrossJoin => Span::empty(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 7c03c4e7e..903fadaa5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5740,7 +5740,7 @@ impl<'a> Parser<'a> { drop_behavior, }) } - /// ```sql + /// ```sql /// DROP CONNECTOR [IF EXISTS] name /// ``` /// @@ -11190,9 +11190,9 @@ impl<'a> Parser<'a> { } Some(Keyword::JOIN) => { if is_left { - JoinOperator::LeftOuter + JoinOperator::Left } else { - JoinOperator::RightOuter + JoinOperator::Right } } _ => { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index e4fa2e837..85845ea2a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -6661,8 +6661,16 @@ fn parse_joins_on() { only(&verified_only_select("SELECT * FROM t1 JOIN t2 ON c1 = c2").from).joins, vec![join_with_constraint("t2", None, false, JoinOperator::Join)] ); + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 INNER JOIN t2 ON c1 = c2").from).joins, + vec![join_with_constraint("t2", None, false, JoinOperator::Inner)] + ); assert_eq!( only(&verified_only_select("SELECT * FROM t1 LEFT JOIN t2 ON c1 = c2").from).joins, + vec![join_with_constraint("t2", None, false, JoinOperator::Left)] + ); + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 LEFT OUTER JOIN t2 ON c1 = c2").from).joins, vec![join_with_constraint( "t2", None, @@ -6672,6 +6680,10 @@ fn parse_joins_on() { ); assert_eq!( only(&verified_only_select("SELECT * FROM t1 RIGHT JOIN t2 ON c1 = c2").from).joins, + vec![join_with_constraint("t2", None, false, JoinOperator::Right)] + ); + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 RIGHT OUTER JOIN t2 ON c1 = c2").from).joins, vec![join_with_constraint( "t2", None, @@ -6794,10 +6806,18 @@ fn parse_joins_using() { ); assert_eq!( only(&verified_only_select("SELECT * FROM t1 LEFT JOIN t2 USING(c1)").from).joins, + vec![join_with_constraint("t2", None, JoinOperator::Left)] + ); + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 LEFT OUTER JOIN t2 USING(c1)").from).joins, vec![join_with_constraint("t2", None, JoinOperator::LeftOuter)] ); assert_eq!( only(&verified_only_select("SELECT * FROM t1 RIGHT JOIN t2 USING(c1)").from).joins, + vec![join_with_constraint("t2", None, JoinOperator::Right)] + ); + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 RIGHT OUTER JOIN t2 USING(c1)").from).joins, vec![join_with_constraint("t2", None, JoinOperator::RightOuter)] ); assert_eq!( @@ -6857,20 +6877,34 @@ fn parse_natural_join() { only(&verified_only_select("SELECT * FROM t1 NATURAL JOIN t2").from).joins, vec![natural_join(JoinOperator::Join, None)] ); + // inner join explicitly assert_eq!( only(&verified_only_select("SELECT * FROM t1 NATURAL INNER JOIN t2").from).joins, vec![natural_join(JoinOperator::Inner, None)] ); + // left join explicitly assert_eq!( only(&verified_only_select("SELECT * FROM t1 NATURAL LEFT JOIN t2").from).joins, + vec![natural_join(JoinOperator::Left, None)] + ); + + // left outer join explicitly + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 NATURAL LEFT OUTER JOIN t2").from).joins, vec![natural_join(JoinOperator::LeftOuter, None)] ); // right join explicitly assert_eq!( only(&verified_only_select("SELECT * FROM t1 NATURAL RIGHT JOIN t2").from).joins, + vec![natural_join(JoinOperator::Right, None)] + ); + + // right outer join explicitly + assert_eq!( + only(&verified_only_select("SELECT * FROM t1 NATURAL RIGHT OUTER JOIN t2").from).joins, vec![natural_join(JoinOperator::RightOuter, None)] ); @@ -6950,22 +6984,12 @@ fn parse_join_nesting() { #[test] fn parse_join_syntax_variants() { - one_statement_parses_to( - "SELECT c1 FROM t1 JOIN t2 USING(c1)", - "SELECT c1 FROM t1 JOIN t2 USING(c1)", - ); - one_statement_parses_to( - "SELECT c1 FROM t1 INNER JOIN t2 USING(c1)", - "SELECT c1 FROM t1 INNER JOIN t2 USING(c1)", - ); - one_statement_parses_to( - "SELECT c1 FROM t1 LEFT OUTER JOIN t2 USING(c1)", - "SELECT c1 FROM t1 LEFT JOIN t2 USING(c1)", - ); - one_statement_parses_to( - "SELECT c1 FROM t1 RIGHT OUTER JOIN t2 USING(c1)", - "SELECT c1 FROM t1 RIGHT JOIN t2 USING(c1)", - ); + verified_stmt("SELECT c1 FROM t1 JOIN t2 USING(c1)"); + verified_stmt("SELECT c1 FROM t1 INNER JOIN t2 USING(c1)"); + verified_stmt("SELECT c1 FROM t1 LEFT JOIN t2 USING(c1)"); + verified_stmt("SELECT c1 FROM t1 LEFT OUTER JOIN t2 USING(c1)"); + verified_stmt("SELECT c1 FROM t1 RIGHT JOIN t2 USING(c1)"); + verified_stmt("SELECT c1 FROM t1 RIGHT OUTER JOIN t2 USING(c1)"); one_statement_parses_to( "SELECT c1 FROM t1 FULL OUTER JOIN t2 USING(c1)", "SELECT c1 FROM t1 FULL JOIN t2 USING(c1)", @@ -8027,7 +8051,7 @@ fn lateral_derived() { let join = &from.joins[0]; assert_eq!( join.join_operator, - JoinOperator::LeftOuter(JoinConstraint::On(Expr::Value(test_utils::number("1")))) + JoinOperator::Left(JoinConstraint::On(Expr::Value(test_utils::number("1")))) ); if let TableFactor::Derived { lateral, @@ -8095,7 +8119,7 @@ fn lateral_function() { alias: None, }, global: false, - join_operator: JoinOperator::LeftOuter(JoinConstraint::None), + join_operator: JoinOperator::Left(JoinConstraint::None), }], }], lateral_views: vec![], From c75a99262102da1ac795c4272a640d7e36b0e157 Mon Sep 17 00:00:00 2001 From: Jesse Stuart Date: Mon, 17 Feb 2025 14:12:59 -0500 Subject: [PATCH 046/286] Add support for Postgres `ALTER TYPE` (#1727) --- src/ast/ddl.rs | 89 ++++++++++++++++++++++++++++++ src/ast/mod.rs | 24 ++++++--- src/ast/spans.rs | 2 + src/dialect/postgresql.rs | 44 --------------- src/parser/mod.rs | 69 ++++++++++++++++++++++++ tests/sqlparser_postgres.rs | 104 ++++++++++++++++++++++++++++++++---- 6 files changed, 271 insertions(+), 61 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 35e01e618..f90252000 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -640,6 +640,95 @@ impl fmt::Display for AlterIndexOperation { } } +/// An `ALTER TYPE` statement (`Statement::AlterType`) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterType { + pub name: ObjectName, + pub operation: AlterTypeOperation, +} + +/// An [AlterType] operation +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterTypeOperation { + Rename(AlterTypeRename), + AddValue(AlterTypeAddValue), + RenameValue(AlterTypeRenameValue), +} + +/// See [AlterTypeOperation::Rename] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterTypeRename { + pub new_name: Ident, +} + +/// See [AlterTypeOperation::AddValue] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterTypeAddValue { + pub if_not_exists: bool, + pub value: Ident, + pub position: Option, +} + +/// See [AlterTypeAddValue] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterTypeAddValuePosition { + Before(Ident), + After(Ident), +} + +/// See [AlterTypeOperation::RenameValue] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterTypeRenameValue { + pub from: Ident, + pub to: Ident, +} + +impl fmt::Display for AlterTypeOperation { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + Self::Rename(AlterTypeRename { new_name }) => { + write!(f, "RENAME TO {new_name}") + } + Self::AddValue(AlterTypeAddValue { + if_not_exists, + value, + position, + }) => { + write!(f, "ADD VALUE")?; + if *if_not_exists { + write!(f, " IF NOT EXISTS")?; + } + write!(f, " {value}")?; + match position { + Some(AlterTypeAddValuePosition::Before(neighbor_value)) => { + write!(f, " BEFORE {neighbor_value}")?; + } + Some(AlterTypeAddValuePosition::After(neighbor_value)) => { + write!(f, " AFTER {neighbor_value}")?; + } + None => {} + }; + Ok(()) + } + Self::RenameValue(AlterTypeRenameValue { from, to }) => { + write!(f, "RENAME VALUE {from} TO {to}") + } + } + } +} + /// An `ALTER COLUMN` (`Statement::AlterTable`) operation #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 5835447db..49c7b6fd7 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -48,13 +48,15 @@ pub use self::dcl::{ }; pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, - AlterTableOperation, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, - ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateFunction, Deduplicate, - DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, - IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, - IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, - ProcedureParam, ReferentialAction, TableConstraint, TagsColumnOption, - UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, + AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, + AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, + ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, + CreateConnector, CreateFunction, Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, + GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, + IdentityPropertyKind, IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, + NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, TableConstraint, + TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, + ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; @@ -2691,6 +2693,11 @@ pub enum Statement { with_options: Vec, }, /// ```sql + /// ALTER TYPE + /// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-altertype.html) + /// ``` + AlterType(AlterType), + /// ```sql /// ALTER ROLE /// ``` AlterRole { @@ -4438,6 +4445,9 @@ impl fmt::Display for Statement { } write!(f, " AS {query}") } + Statement::AlterType(AlterType { name, operation }) => { + write!(f, "ALTER TYPE {name} {operation}") + } Statement::AlterRole { name, operation } => { write!(f, "ALTER ROLE {name} {operation}") } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 574478692..191471675 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -215,6 +215,7 @@ impl Spanned for Values { /// - [Statement::CopyIntoSnowflake] /// - [Statement::CreateSecret] /// - [Statement::CreateRole] +/// - [Statement::AlterType] /// - [Statement::AlterRole] /// - [Statement::AttachDatabase] /// - [Statement::AttachDuckDBDatabase] @@ -427,6 +428,7 @@ impl Spanned for Statement { .chain(with_options.iter().map(|i| i.span())), ), // These statements need to be implemented + Statement::AlterType { .. } => Span::empty(), Statement::AlterRole { .. } => Span::empty(), Statement::AttachDatabase { .. } => Span::empty(), Statement::AttachDuckDBDatabase { .. } => Span::empty(), diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 74b963e82..60284364f 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -28,7 +28,6 @@ // limitations under the License. use log::debug; -use crate::ast::{ObjectName, Statement, UserDefinedTypeRepresentation}; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; @@ -135,15 +134,6 @@ impl Dialect for PostgreSqlDialect { } } - fn parse_statement(&self, parser: &mut Parser) -> Option> { - if parser.parse_keyword(Keyword::CREATE) { - parser.prev_token(); // unconsume the CREATE in case we don't end up parsing anything - parse_create(parser) - } else { - None - } - } - fn supports_filter_during_aggregation(&self) -> bool { true } @@ -259,37 +249,3 @@ impl Dialect for PostgreSqlDialect { true } } - -pub fn parse_create(parser: &mut Parser) -> Option> { - let name = parser.maybe_parse(|parser| -> Result { - parser.expect_keyword_is(Keyword::CREATE)?; - parser.expect_keyword_is(Keyword::TYPE)?; - let name = parser.parse_object_name(false)?; - parser.expect_keyword_is(Keyword::AS)?; - parser.expect_keyword_is(Keyword::ENUM)?; - Ok(name) - }); - - match name { - Ok(name) => name.map(|name| parse_create_type_as_enum(parser, name)), - Err(e) => Some(Err(e)), - } -} - -// https://www.postgresql.org/docs/current/sql-createtype.html -pub fn parse_create_type_as_enum( - parser: &mut Parser, - name: ObjectName, -) -> Result { - if !parser.consume_token(&Token::LParen) { - return parser.expected("'(' after CREATE TYPE AS ENUM", parser.peek_token()); - } - - let labels = parser.parse_comma_separated0(|p| p.parse_identifier(), Token::RParen)?; - parser.expect_token(&Token::RParen)?; - - Ok(Statement::CreateType { - name, - representation: UserDefinedTypeRepresentation::Enum { labels }, - }) -} diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 903fadaa5..88a9281fe 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8042,6 +8042,7 @@ impl<'a> Parser<'a> { pub fn parse_alter(&mut self) -> Result { let object_type = self.expect_one_of_keywords(&[ Keyword::VIEW, + Keyword::TYPE, Keyword::TABLE, Keyword::INDEX, Keyword::ROLE, @@ -8050,6 +8051,7 @@ impl<'a> Parser<'a> { ])?; match object_type { Keyword::VIEW => self.parse_alter_view(), + Keyword::TYPE => self.parse_alter_type(), Keyword::TABLE => { let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); let only = self.parse_keyword(Keyword::ONLY); // [ ONLY ] @@ -8122,6 +8124,55 @@ impl<'a> Parser<'a> { }) } + /// Parse a [Statement::AlterType] + pub fn parse_alter_type(&mut self) -> Result { + let name = self.parse_object_name(false)?; + + if self.parse_keywords(&[Keyword::RENAME, Keyword::TO]) { + let new_name = self.parse_identifier()?; + Ok(Statement::AlterType(AlterType { + name, + operation: AlterTypeOperation::Rename(AlterTypeRename { new_name }), + })) + } else if self.parse_keywords(&[Keyword::ADD, Keyword::VALUE]) { + let if_not_exists = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let new_enum_value = self.parse_identifier()?; + let position = if self.parse_keyword(Keyword::BEFORE) { + Some(AlterTypeAddValuePosition::Before(self.parse_identifier()?)) + } else if self.parse_keyword(Keyword::AFTER) { + Some(AlterTypeAddValuePosition::After(self.parse_identifier()?)) + } else { + None + }; + + Ok(Statement::AlterType(AlterType { + name, + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists, + value: new_enum_value, + position, + }), + })) + } else if self.parse_keywords(&[Keyword::RENAME, Keyword::VALUE]) { + let existing_enum_value = self.parse_identifier()?; + self.expect_keyword(Keyword::TO)?; + let new_enum_value = self.parse_identifier()?; + + Ok(Statement::AlterType(AlterType { + name, + operation: AlterTypeOperation::RenameValue(AlterTypeRenameValue { + from: existing_enum_value, + to: new_enum_value, + }), + })) + } else { + return self.expected_ref( + "{RENAME TO | { RENAME | ADD } VALUE}", + self.peek_token_ref(), + ); + } + } + /// Parse a `CALL procedure_name(arg1, arg2, ...)` /// or `CALL procedure_name` statement pub fn parse_call(&mut self) -> Result { @@ -14222,6 +14273,10 @@ impl<'a> Parser<'a> { let name = self.parse_object_name(false)?; self.expect_keyword_is(Keyword::AS)?; + if self.parse_keyword(Keyword::ENUM) { + return self.parse_create_type_enum(name); + } + let mut attributes = vec![]; if !self.consume_token(&Token::LParen) || self.consume_token(&Token::RParen) { return Ok(Statement::CreateType { @@ -14258,6 +14313,20 @@ impl<'a> Parser<'a> { }) } + /// Parse remainder of `CREATE TYPE AS ENUM` statement (see [Statement::CreateType] and [Self::parse_create_type]) + /// + /// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-createtype.html) + pub fn parse_create_type_enum(&mut self, name: ObjectName) -> Result { + self.expect_token(&Token::LParen)?; + let labels = self.parse_comma_separated0(|p| p.parse_identifier(), Token::RParen)?; + self.expect_token(&Token::RParen)?; + + Ok(Statement::CreateType { + name, + representation: UserDefinedTypeRepresentation::Enum { labels }, + }) + } + fn parse_parenthesized_identifiers(&mut self) -> Result, ParserError> { self.expect_token(&Token::LParen)?; let partitions = self.parse_comma_separated(|p| p.parse_identifier())?; diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 3a4504ebb..b9a5202f0 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5293,15 +5293,8 @@ fn arrow_cast_precedence() { #[test] fn parse_create_type_as_enum() { - let statement = pg().one_statement_parses_to( - r#"CREATE TYPE public.my_type AS ENUM ( - 'label1', - 'label2', - 'label3', - 'label4' - );"#, - "CREATE TYPE public.my_type AS ENUM ('label1', 'label2', 'label3', 'label4')", - ); + let sql = "CREATE TYPE public.my_type AS ENUM ('label1', 'label2', 'label3', 'label4')"; + let statement = pg_and_generic().verified_stmt(sql); match statement { Statement::CreateType { name, @@ -5316,10 +5309,101 @@ fn parse_create_type_as_enum() { labels ); } - _ => unreachable!(), + _ => unreachable!("{:?} should parse to Statement::CreateType", sql), } } +#[test] +fn parse_alter_type() { + struct TestCase { + sql: &'static str, + name: &'static str, + operation: AlterTypeOperation, + } + vec![ + TestCase { + sql: "ALTER TYPE public.my_type RENAME TO my_new_type", + name: "public.my_type", + operation: AlterTypeOperation::Rename(AlterTypeRename { + new_name: Ident::new("my_new_type"), + }), + }, + TestCase { + sql: "ALTER TYPE public.my_type ADD VALUE IF NOT EXISTS 'label3.5' BEFORE 'label4'", + name: "public.my_type", + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists: true, + value: Ident::with_quote('\'', "label3.5"), + position: Some(AlterTypeAddValuePosition::Before(Ident::with_quote( + '\'', "label4", + ))), + }), + }, + TestCase { + sql: "ALTER TYPE public.my_type ADD VALUE 'label3.5' BEFORE 'label4'", + name: "public.my_type", + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists: false, + value: Ident::with_quote('\'', "label3.5"), + position: Some(AlterTypeAddValuePosition::Before(Ident::with_quote( + '\'', "label4", + ))), + }), + }, + TestCase { + sql: "ALTER TYPE public.my_type ADD VALUE IF NOT EXISTS 'label3.5' AFTER 'label3'", + name: "public.my_type", + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists: true, + value: Ident::with_quote('\'', "label3.5"), + position: Some(AlterTypeAddValuePosition::After(Ident::with_quote( + '\'', "label3", + ))), + }), + }, + TestCase { + sql: "ALTER TYPE public.my_type ADD VALUE 'label3.5' AFTER 'label3'", + name: "public.my_type", + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists: false, + value: Ident::with_quote('\'', "label3.5"), + position: Some(AlterTypeAddValuePosition::After(Ident::with_quote( + '\'', "label3", + ))), + }), + }, + TestCase { + sql: "ALTER TYPE public.my_type ADD VALUE IF NOT EXISTS 'label5'", + name: "public.my_type", + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists: true, + value: Ident::with_quote('\'', "label5"), + position: None, + }), + }, + TestCase { + sql: "ALTER TYPE public.my_type ADD VALUE 'label5'", + name: "public.my_type", + operation: AlterTypeOperation::AddValue(AlterTypeAddValue { + if_not_exists: false, + value: Ident::with_quote('\'', "label5"), + position: None, + }), + }, + ] + .into_iter() + .enumerate() + .for_each(|(index, tc)| { + let statement = pg_and_generic().verified_stmt(tc.sql); + if let Statement::AlterType(AlterType { name, operation }) = statement { + assert_eq!(tc.name, name.to_string(), "TestCase[{index}].name"); + assert_eq!(tc.operation, operation, "TestCase[{index}].operation"); + } else { + unreachable!("{:?} should parse to Statement::AlterType", tc.sql); + } + }); +} + #[test] fn parse_bitstring_literal() { let select = pg_and_generic().verified_only_select("SELECT B'111'"); From 3e90a18f6d8cf6b5e240705c0f1fc2ad44330a95 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Wed, 19 Feb 2025 18:49:42 +0100 Subject: [PATCH 047/286] Replace `Method` and `CompositeAccess` with `CompoundFieldAccess` (#1716) --- src/ast/mod.rs | 26 --- src/ast/spans.rs | 2 - src/dialect/mod.rs | 17 +- src/dialect/mssql.rs | 8 +- src/dialect/postgresql.rs | 2 + src/dialect/snowflake.rs | 5 + src/parser/mod.rs | 377 ++++++++++++++++++++---------------- tests/sqlparser_common.rs | 123 +++++++++--- tests/sqlparser_postgres.rs | 50 +++-- 9 files changed, 347 insertions(+), 263 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 49c7b6fd7..1f05d590d 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -661,11 +661,6 @@ pub enum Expr { /// The path to the data to extract. path: JsonPath, }, - /// CompositeAccess eg: SELECT foo(bar).z, (information_schema._pg_expandarray(array['i','i'])).n - CompositeAccess { - expr: Box, - key: Ident, - }, /// `IS FALSE` operator IsFalse(Box), /// `IS NOT FALSE` operator @@ -915,23 +910,6 @@ pub enum Expr { }, /// Scalar function call e.g. `LEFT(foo, 5)` Function(Function), - /// Arbitrary expr method call - /// - /// Syntax: - /// - /// `.....` - /// - /// > `arbitrary-expr` can be any expression including a function call. - /// - /// Example: - /// - /// ```sql - /// SELECT (SELECT ',' + name FROM sys.objects FOR XML PATH(''), TYPE).value('.','NVARCHAR(MAX)') - /// SELECT CONVERT(XML,'abc').value('.','NVARCHAR(MAX)').value('.','NVARCHAR(MAX)') - /// ``` - /// - /// (mssql): - Method(Method), /// `CASE [] WHEN THEN ... [ELSE ] END` /// /// Note we only recognize a complete single expression as ``, @@ -1631,7 +1609,6 @@ impl fmt::Display for Expr { write!(f, " {value}") } Expr::Function(fun) => write!(f, "{fun}"), - Expr::Method(method) => write!(f, "{method}"), Expr::Case { operand, conditions, @@ -1789,9 +1766,6 @@ impl fmt::Display for Expr { Expr::JsonAccess { value, path } => { write!(f, "{value}{path}") } - Expr::CompositeAccess { expr, key } => { - write!(f, "{expr}.{key}") - } Expr::AtTimeZone { timestamp, time_zone, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 191471675..091a22f04 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1288,7 +1288,6 @@ impl Spanned for Expr { match self { Expr::Identifier(ident) => ident.span, Expr::CompoundIdentifier(vec) => union_spans(vec.iter().map(|i| i.span)), - Expr::CompositeAccess { expr, key } => expr.span().union(&key.span), Expr::CompoundFieldAccess { root, access_chain } => { union_spans(iter::once(root.span()).chain(access_chain.iter().map(|i| i.span()))) } @@ -1478,7 +1477,6 @@ impl Spanned for Expr { Expr::OuterJoin(expr) => expr.span(), Expr::Prior(expr) => expr.span(), Expr::Lambda(_) => Span::empty(), - Expr::Method(_) => Span::empty(), } } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 031fe9676..cb86cf7c6 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -251,6 +251,11 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports the `(+)` syntax for OUTER JOIN. + fn supports_outer_join_operator(&self) -> bool { + false + } + /// Returns true if the dialect supports CONNECT BY. fn supports_connect_by(&self) -> bool { false @@ -352,15 +357,6 @@ pub trait Dialect: Debug + Any { false } - /// Returns true if the dialect supports method calls, for example: - /// - /// ```sql - /// SELECT (SELECT ',' + name FROM sys.objects FOR XML PATH(''), TYPE).value('.','NVARCHAR(MAX)') - /// ``` - fn supports_methods(&self) -> bool { - false - } - /// Returns true if the dialect supports multiple variable assignment /// using parentheses in a `SET` variable declaration. /// @@ -581,6 +577,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::OPERATOR => Ok(p!(Between)), Token::Word(w) if w.keyword == Keyword::DIV => Ok(p!(MulDivModOp)), + Token::Period => Ok(p!(Period)), Token::Eq | Token::Lt | Token::LtEq @@ -654,6 +651,7 @@ pub trait Dialect: Debug + Any { /// Uses (APPROXIMATELY) as a reference fn prec_value(&self, prec: Precedence) -> u8 { match prec { + Precedence::Period => 100, Precedence::DoubleColon => 50, Precedence::AtTz => 41, Precedence::MulDivModOp => 40, @@ -925,6 +923,7 @@ pub trait Dialect: Debug + Any { /// higher number -> higher precedence #[derive(Debug, Clone, Copy)] pub enum Precedence { + Period, DoubleColon, AtTz, MulDivModOp, diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 7d8611cb0..980f5ec3f 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -46,6 +46,10 @@ impl Dialect for MsSqlDialect { true } + fn supports_outer_join_operator(&self) -> bool { + true + } + fn supports_connect_by(&self) -> bool { true } @@ -63,10 +67,6 @@ impl Dialect for MsSqlDialect { false } - fn supports_methods(&self) -> bool { - true - } - fn supports_named_fn_args_with_colon_operator(&self) -> bool { true } diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 60284364f..3a3f0e4c3 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -37,6 +37,7 @@ use crate::tokenizer::Token; #[derive(Debug)] pub struct PostgreSqlDialect {} +const PERIOD_PREC: u8 = 200; const DOUBLE_COLON_PREC: u8 = 140; const BRACKET_PREC: u8 = 130; const COLLATE_PREC: u8 = 120; @@ -144,6 +145,7 @@ impl Dialect for PostgreSqlDialect { fn prec_value(&self, prec: Precedence) -> u8 { match prec { + Precedence::Period => PERIOD_PREC, Precedence::DoubleColon => DOUBLE_COLON_PREC, Precedence::AtTz => AT_TZ_PREC, Precedence::MulDivModOp => MUL_DIV_MOD_OP_PREC, diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 68166cbef..bac9c49c7 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -87,6 +87,11 @@ impl Dialect for SnowflakeDialect { true } + /// See + fn supports_outer_join_operator(&self) -> bool { + true + } + fn supports_connect_by(&self) -> bool { true } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 88a9281fe..9d75bcdb5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1021,6 +1021,8 @@ impl<'a> Parser<'a> { debug!("parsing expr"); let mut expr = self.parse_prefix()?; + expr = self.parse_compound_expr(expr, vec![])?; + debug!("prefix: {:?}", expr); loop { let next_precedence = self.get_next_precedence()?; @@ -1030,6 +1032,12 @@ impl<'a> Parser<'a> { break; } + // The period operator is handled exclusively by the + // compound field access parsing. + if Token::Period == self.peek_token_ref().token { + break; + } + expr = self.parse_infix(expr, next_precedence)?; } Ok(expr) @@ -1105,8 +1113,8 @@ impl<'a> Parser<'a> { } } - // Tries to parse an expression by matching the specified word to known keywords that have a special meaning in the dialect. - // Returns `None if no match is found. + /// Tries to parse an expression by matching the specified word to known keywords that have a special meaning in the dialect. + /// Returns `None if no match is found. fn parse_expr_prefix_by_reserved_word( &mut self, w: &Word, @@ -1203,7 +1211,7 @@ impl<'a> Parser<'a> { } Keyword::STRUCT if self.dialect.supports_struct_literal() => { let struct_expr = self.parse_struct_literal()?; - Ok(Some(self.parse_compound_field_access(struct_expr, vec![])?)) + Ok(Some(struct_expr)) } Keyword::PRIOR if matches!(self.state, ParserState::ConnectBy) => { let expr = self.parse_subexpr(self.dialect.prec_value(Precedence::PlusMinus))?; @@ -1216,35 +1224,16 @@ impl<'a> Parser<'a> { } } - // Tries to parse an expression by a word that is not known to have a special meaning in the dialect. + /// Tries to parse an expression by a word that is not known to have a special meaning in the dialect. fn parse_expr_prefix_by_unreserved_word( &mut self, w: &Word, w_span: Span, ) -> Result { match self.peek_token().token { - Token::Period => self.parse_compound_field_access( - Expr::Identifier(w.clone().into_ident(w_span)), - vec![], - ), - Token::LParen => { + Token::LParen if !self.peek_outer_join_operator() => { let id_parts = vec![w.clone().into_ident(w_span)]; - if let Some(expr) = self.parse_outer_join_expr(&id_parts) { - Ok(expr) - } else { - let mut expr = self.parse_function(ObjectName::from(id_parts))?; - // consume all period if it's a method chain - expr = self.try_parse_method(expr)?; - let fields = vec![]; - self.parse_compound_field_access(expr, fields) - } - } - Token::LBracket if dialect_of!(self is PostgreSqlDialect | DuckDbDialect | GenericDialect | ClickHouseDialect | BigQueryDialect) => - { - let ident = Expr::Identifier(w.clone().into_ident(w_span)); - let mut fields = vec![]; - self.parse_multi_dim_subscript(&mut fields)?; - self.parse_compound_field_access(ident, fields) + self.parse_function(ObjectName::from(id_parts)) } // string introducer https://dev.mysql.com/doc/refman/8.0/en/charset-introducer.html Token::SingleQuotedString(_) @@ -1453,25 +1442,7 @@ impl<'a> Parser<'a> { } }; self.expect_token(&Token::RParen)?; - let expr = self.try_parse_method(expr)?; - if !self.consume_token(&Token::Period) { - Ok(expr) - } else { - let tok = self.next_token(); - let key = match tok.token { - Token::Word(word) => word.into_ident(tok.span), - _ => { - return parser_err!( - format!("Expected identifier, found: {tok}"), - tok.span.start - ) - } - }; - Ok(Expr::CompositeAccess { - expr: Box::new(expr), - key, - }) - } + Ok(expr) } Token::Placeholder(_) | Token::Colon | Token::AtSign => { self.prev_token(); @@ -1484,8 +1455,6 @@ impl<'a> Parser<'a> { _ => self.expected_at("an expression", next_token_index), }?; - let expr = self.try_parse_method(expr)?; - if self.parse_keyword(Keyword::COLLATE) { Ok(Expr::Collate { expr: Box::new(expr), @@ -1499,62 +1468,72 @@ impl<'a> Parser<'a> { /// Try to parse an [Expr::CompoundFieldAccess] like `a.b.c` or `a.b[1].c`. /// If all the fields are `Expr::Identifier`s, return an [Expr::CompoundIdentifier] instead. /// If only the root exists, return the root. - /// If self supports [Dialect::supports_partiql], it will fall back when occurs [Token::LBracket] for JsonAccess parsing. - pub fn parse_compound_field_access( + /// Parses compound expressions which may be delimited by period + /// or bracket notation. + /// For example: `a.b.c`, `a.b[1]`. + pub fn parse_compound_expr( &mut self, root: Expr, mut chain: Vec, ) -> Result { let mut ending_wildcard: Option = None; - let mut ending_lbracket = false; - while self.consume_token(&Token::Period) { - let next_token = self.next_token(); - match next_token.token { - Token::Word(w) => { - let expr = Expr::Identifier(w.into_ident(next_token.span)); - chain.push(AccessExpr::Dot(expr)); - if self.peek_token().token == Token::LBracket { - if self.dialect.supports_partiql() { - self.next_token(); - ending_lbracket = true; - break; + loop { + if self.consume_token(&Token::Period) { + let next_token = self.peek_token_ref(); + match &next_token.token { + Token::Mul => { + // Postgres explicitly allows funcnm(tablenm.*) and the + // function array_agg traverses this control flow + if dialect_of!(self is PostgreSqlDialect) { + ending_wildcard = Some(self.next_token()); } else { - self.parse_multi_dim_subscript(&mut chain)? + // Put back the consumed `.` tokens before exiting. + // If this expression is being parsed in the + // context of a projection, then the `.*` could imply + // a wildcard expansion. For example: + // `SELECT STRUCT('foo').* FROM T` + self.prev_token(); // . } + + break; } - } - Token::Mul => { - // Postgres explicitly allows funcnm(tablenm.*) and the - // function array_agg traverses this control flow - if dialect_of!(self is PostgreSqlDialect) { - ending_wildcard = Some(next_token); - } else { - // Put back the consumed .* tokens before exiting. - // If this expression is being parsed in the - // context of a projection, then this could imply - // a wildcard expansion. For example: - // `SELECT STRUCT('foo').* FROM T` - self.prev_token(); // * - self.prev_token(); // . + Token::SingleQuotedString(s) => { + let expr = + Expr::Identifier(Ident::with_quote_and_span('\'', next_token.span, s)); + chain.push(AccessExpr::Dot(expr)); + self.advance_token(); // The consumed string } - - break; - } - Token::SingleQuotedString(s) => { - let expr = Expr::Identifier(Ident::with_quote('\'', s)); - chain.push(AccessExpr::Dot(expr)); - } - _ => { - return self.expected("an identifier or a '*' after '.'", next_token); + // Fallback to parsing an arbitrary expression. + _ => match self.parse_subexpr(self.dialect.prec_value(Precedence::Period))? { + // If we get back a compound field access or identifier, + // we flatten the nested expression. + // For example if the current root is `foo` + // and we get back a compound identifier expression `bar.baz` + // The full expression should be `foo.bar.baz` (i.e. + // a root with an access chain with 2 entries) and not + // `foo.(bar.baz)` (i.e. a root with an access chain with + // 1 entry`). + Expr::CompoundFieldAccess { root, access_chain } => { + chain.push(AccessExpr::Dot(*root)); + chain.extend(access_chain); + } + Expr::CompoundIdentifier(parts) => chain + .extend(parts.into_iter().map(Expr::Identifier).map(AccessExpr::Dot)), + expr => { + chain.push(AccessExpr::Dot(expr)); + } + }, } + } else if !self.dialect.supports_partiql() + && self.peek_token_ref().token == Token::LBracket + { + self.parse_multi_dim_subscript(&mut chain)?; + } else { + break; } } - // if dialect supports partiql, we need to go back one Token::LBracket for the JsonAccess parsing - if self.dialect.supports_partiql() && ending_lbracket { - self.prev_token(); - } - + let tok_index = self.get_current_index(); if let Some(wildcard_token) = ending_wildcard { if !Self::is_all_ident(&root, &chain) { return self.expected("an identifier or a '*' after '.'", self.peek_token()); @@ -1563,32 +1542,112 @@ impl<'a> Parser<'a> { ObjectName::from(Self::exprs_to_idents(root, chain)?), AttachedToken(wildcard_token), )) - } else if self.peek_token().token == Token::LParen { + } else if self.maybe_parse_outer_join_operator() { if !Self::is_all_ident(&root, &chain) { - // consume LParen - self.next_token(); - return self.expected("an identifier or a '*' after '.'", self.peek_token()); + return self.expected_at("column identifier before (+)", tok_index); }; - let id_parts = Self::exprs_to_idents(root, chain)?; - if let Some(expr) = self.parse_outer_join_expr(&id_parts) { - Ok(expr) + let expr = if chain.is_empty() { + root } else { - self.parse_function(ObjectName::from(id_parts)) - } - } else if chain.is_empty() { - Ok(root) + Expr::CompoundIdentifier(Self::exprs_to_idents(root, chain)?) + }; + Ok(Expr::OuterJoin(expr.into())) } else { - if Self::is_all_ident(&root, &chain) { - return Ok(Expr::CompoundIdentifier(Self::exprs_to_idents( - root, chain, - )?)); + Self::build_compound_expr(root, chain) + } + } + + /// Combines a root expression and access chain to form + /// a compound expression. Which may be a [Expr::CompoundFieldAccess] + /// or other special cased expressions like [Expr::CompoundIdentifier], + /// [Expr::OuterJoin]. + fn build_compound_expr( + root: Expr, + mut access_chain: Vec, + ) -> Result { + if access_chain.is_empty() { + return Ok(root); + } + + if Self::is_all_ident(&root, &access_chain) { + return Ok(Expr::CompoundIdentifier(Self::exprs_to_idents( + root, + access_chain, + )?)); + } + + // Flatten qualified function calls. + // For example, the expression `a.b.c.foo(1,2,3)` should + // represent a function called `a.b.c.foo`, rather than + // a composite expression. + if matches!(root, Expr::Identifier(_)) + && matches!( + access_chain.last(), + Some(AccessExpr::Dot(Expr::Function(_))) + ) + && access_chain + .iter() + .rev() + .skip(1) // All except the Function + .all(|access| matches!(access, AccessExpr::Dot(Expr::Identifier(_)))) + { + let Some(AccessExpr::Dot(Expr::Function(mut func))) = access_chain.pop() else { + return parser_err!("expected function expression", root.span().start); + }; + + let compound_func_name = [root] + .into_iter() + .chain(access_chain.into_iter().flat_map(|access| match access { + AccessExpr::Dot(expr) => Some(expr), + _ => None, + })) + .flat_map(|expr| match expr { + Expr::Identifier(ident) => Some(ident), + _ => None, + }) + .map(ObjectNamePart::Identifier) + .chain(func.name.0) + .collect::>(); + func.name = ObjectName(compound_func_name); + + return Ok(Expr::Function(func)); + } + + // Flatten qualified outer join expressions. + // For example, the expression `T.foo(+)` should + // represent an outer join on the column name `T.foo` + // rather than a composite expression. + if access_chain.len() == 1 + && matches!( + access_chain.last(), + Some(AccessExpr::Dot(Expr::OuterJoin(_))) + ) + { + let Some(AccessExpr::Dot(Expr::OuterJoin(inner_expr))) = access_chain.pop() else { + return parser_err!("expected (+) expression", root.span().start); + }; + + if !Self::is_all_ident(&root, &[]) { + return parser_err!("column identifier before (+)", root.span().start); + }; + + let token_start = root.span().start; + let mut idents = Self::exprs_to_idents(root, vec![])?; + match *inner_expr { + Expr::CompoundIdentifier(suffix) => idents.extend(suffix), + Expr::Identifier(suffix) => idents.push(suffix), + _ => { + return parser_err!("column identifier before (+)", token_start); + } } - Ok(Expr::CompoundFieldAccess { - root: Box::new(root), - access_chain: chain, - }) + return Ok(Expr::OuterJoin(Expr::CompoundIdentifier(idents).into())); } + + Ok(Expr::CompoundFieldAccess { + root: Box::new(root), + access_chain, + }) } /// Check if the root is an identifier and all fields are identifiers. @@ -1625,20 +1684,23 @@ impl<'a> Parser<'a> { } } - /// Try to parse OuterJoin expression `(+)` - fn parse_outer_join_expr(&mut self, id_parts: &[Ident]) -> Option { - if dialect_of!(self is SnowflakeDialect | MsSqlDialect) - && self.consume_tokens(&[Token::LParen, Token::Plus, Token::RParen]) - { - Some(Expr::OuterJoin(Box::new( - match <[Ident; 1]>::try_from(id_parts.to_vec()) { - Ok([ident]) => Expr::Identifier(ident), - Err(parts) => Expr::CompoundIdentifier(parts), - }, - ))) - } else { - None + /// Returns true if the next tokens indicate the outer join operator `(+)`. + fn peek_outer_join_operator(&mut self) -> bool { + if !self.dialect.supports_outer_join_operator() { + return false; } + + let [maybe_lparen, maybe_plus, maybe_rparen] = self.peek_tokens_ref(); + Token::LParen == maybe_lparen.token + && Token::Plus == maybe_plus.token + && Token::RParen == maybe_rparen.token + } + + /// If the next tokens indicates the outer join operator `(+)`, consume + /// the tokens and return true. + fn maybe_parse_outer_join_operator(&mut self) -> bool { + self.dialect.supports_outer_join_operator() + && self.consume_tokens(&[Token::LParen, Token::Plus, Token::RParen]) } pub fn parse_utility_options(&mut self) -> Result, ParserError> { @@ -1688,41 +1750,6 @@ impl<'a> Parser<'a> { }) } - /// Parses method call expression - fn try_parse_method(&mut self, expr: Expr) -> Result { - if !self.dialect.supports_methods() { - return Ok(expr); - } - let method_chain = self.maybe_parse(|p| { - let mut method_chain = Vec::new(); - while p.consume_token(&Token::Period) { - let tok = p.next_token(); - let name = match tok.token { - Token::Word(word) => word.into_ident(tok.span), - _ => return p.expected("identifier", tok), - }; - let func = match p.parse_function(ObjectName::from(vec![name]))? { - Expr::Function(func) => func, - _ => return p.expected("function", p.peek_token()), - }; - method_chain.push(func); - } - if !method_chain.is_empty() { - Ok(method_chain) - } else { - p.expected("function", p.peek_token()) - } - })?; - if let Some(method_chain) = method_chain { - Ok(Expr::Method(Method { - expr: Box::new(expr), - method_chain, - })) - } else { - Ok(expr) - } - } - /// Tries to parse the body of an [ODBC function] call. /// i.e. without the enclosing braces /// @@ -3281,21 +3308,9 @@ impl<'a> Parser<'a> { op: UnaryOperator::PGPostfixFactorial, expr: Box::new(expr), }) - } else if Token::LBracket == *tok { - if dialect_of!(self is PostgreSqlDialect | DuckDbDialect | GenericDialect | ClickHouseDialect | BigQueryDialect) - { - let mut chain = vec![]; - // back to LBracket - self.prev_token(); - self.parse_multi_dim_subscript(&mut chain)?; - self.parse_compound_field_access(expr, chain) - } else if self.dialect.supports_partiql() { - self.prev_token(); - self.parse_json_access(expr) - } else { - parser_err!("Array subscripting is not supported", tok.span.start) - } - } else if dialect_of!(self is SnowflakeDialect | GenericDialect) && Token::Colon == *tok { + } else if Token::LBracket == *tok && self.dialect.supports_partiql() + || (dialect_of!(self is SnowflakeDialect | GenericDialect) && Token::Colon == *tok) + { self.prev_token(); self.parse_json_access(expr) } else { @@ -3605,6 +3620,26 @@ impl<'a> Parser<'a> { }) } + /// Returns references to the `N` next non-whitespace tokens + /// that have not yet been processed. + /// + /// See [`Self::peek_tokens`] for an example. + pub fn peek_tokens_ref(&self) -> [&TokenWithSpan; N] { + let mut index = self.index; + core::array::from_fn(|_| loop { + let token = self.tokens.get(index); + index += 1; + if let Some(TokenWithSpan { + token: Token::Whitespace(_), + span: _, + }) = token + { + continue; + } + break token.unwrap_or(&EOF_TOKEN); + }) + } + /// Return nth non-whitespace token that has not yet been processed pub fn peek_nth_token(&self, n: usize) -> TokenWithSpan { self.peek_nth_token_ref(n).clone() diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 85845ea2a..11e43daeb 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -73,6 +73,23 @@ fn parse_numeric_literal_underscore() { ); } +#[test] +fn parse_function_object_name() { + let select = verified_only_select("SELECT a.b.c.d(1, 2, 3) FROM T"); + let Expr::Function(func) = expr_from_projection(&select.projection[0]) else { + unreachable!() + }; + assert_eq!( + ObjectName::from( + ["a", "b", "c", "d"] + .into_iter() + .map(Ident::new) + .collect::>() + ), + func.name, + ); +} + #[test] fn parse_insert_values() { let row = vec![ @@ -936,6 +953,44 @@ fn parse_select_distinct_tuple() { ); } +#[test] +fn parse_outer_join_operator() { + let dialects = all_dialects_where(|d| d.supports_outer_join_operator()); + + let select = dialects.verified_only_select("SELECT 1 FROM T WHERE a = b (+)"); + assert_eq!( + select.selection, + Some(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("a"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::OuterJoin(Box::new(Expr::Identifier(Ident::new("b"))))) + }) + ); + + let select = dialects.verified_only_select("SELECT 1 FROM T WHERE t1.c1 = t2.c2.d3 (+)"); + assert_eq!( + select.selection, + Some(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("t1"), + Ident::new("c1") + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::OuterJoin(Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("t2"), + Ident::new("c2"), + Ident::new("d3"), + ])))) + }) + ); + + let res = dialects.parse_sql_statements("SELECT 1 FROM T WHERE 1 = 2 (+)"); + assert_eq!( + ParserError::ParserError("Expected: column identifier before (+), found: 2".to_string()), + res.unwrap_err() + ); +} + #[test] fn parse_select_distinct_on() { let sql = "SELECT DISTINCT ON (album_id) name FROM track ORDER BY album_id, milliseconds"; @@ -12623,68 +12678,76 @@ fn test_try_convert() { #[test] fn parse_method_select() { - let dialects = all_dialects_where(|d| d.supports_methods()); - let _ = dialects.verified_only_select( + let _ = verified_only_select( "SELECT LEFT('abc', 1).value('.', 'NVARCHAR(MAX)').value('.', 'NVARCHAR(MAX)') AS T", ); - let _ = dialects.verified_only_select("SELECT STUFF((SELECT ',' + name FROM sys.objects FOR XML PATH(''), TYPE).value('.', 'NVARCHAR(MAX)'), 1, 1, '') AS T"); - let _ = dialects - .verified_only_select("SELECT CAST(column AS XML).value('.', 'NVARCHAR(MAX)') AS T"); + let _ = verified_only_select("SELECT STUFF((SELECT ',' + name FROM sys.objects FOR XML PATH(''), TYPE).value('.', 'NVARCHAR(MAX)'), 1, 1, '') AS T"); + let _ = verified_only_select("SELECT CAST(column AS XML).value('.', 'NVARCHAR(MAX)') AS T"); // `CONVERT` support - let dialects = all_dialects_where(|d| { - d.supports_methods() && d.supports_try_convert() && d.convert_type_before_value() - }); + let dialects = + all_dialects_where(|d| d.supports_try_convert() && d.convert_type_before_value()); let _ = dialects.verified_only_select("SELECT CONVERT(XML, 'abc').value('.', 'NVARCHAR(MAX)').value('.', 'NVARCHAR(MAX)') AS T"); } #[test] fn parse_method_expr() { - let dialects = all_dialects_where(|d| d.supports_methods()); - let expr = dialects - .verified_expr("LEFT('abc', 1).value('.', 'NVARCHAR(MAX)').value('.', 'NVARCHAR(MAX)')"); + let expr = + verified_expr("LEFT('abc', 1).value('.', 'NVARCHAR(MAX)').value('.', 'NVARCHAR(MAX)')"); match expr { - Expr::Method(Method { expr, method_chain }) => { - assert!(matches!(*expr, Expr::Function(_))); + Expr::CompoundFieldAccess { root, access_chain } => { + assert!(matches!(*root, Expr::Function(_))); assert!(matches!( - method_chain[..], - [Function { .. }, Function { .. }] + access_chain[..], + [ + AccessExpr::Dot(Expr::Function(_)), + AccessExpr::Dot(Expr::Function(_)) + ] )); } _ => unreachable!(), } - let expr = dialects.verified_expr( + + let expr = verified_expr( "(SELECT ',' + name FROM sys.objects FOR XML PATH(''), TYPE).value('.', 'NVARCHAR(MAX)')", ); match expr { - Expr::Method(Method { expr, method_chain }) => { - assert!(matches!(*expr, Expr::Subquery(_))); - assert!(matches!(method_chain[..], [Function { .. }])); + Expr::CompoundFieldAccess { root, access_chain } => { + assert!(matches!(*root, Expr::Subquery(_))); + assert!(matches!( + access_chain[..], + [AccessExpr::Dot(Expr::Function(_))] + )); } _ => unreachable!(), } - let expr = dialects.verified_expr("CAST(column AS XML).value('.', 'NVARCHAR(MAX)')"); + let expr = verified_expr("CAST(column AS XML).value('.', 'NVARCHAR(MAX)')"); match expr { - Expr::Method(Method { expr, method_chain }) => { - assert!(matches!(*expr, Expr::Cast { .. })); - assert!(matches!(method_chain[..], [Function { .. }])); + Expr::CompoundFieldAccess { root, access_chain } => { + assert!(matches!(*root, Expr::Cast { .. })); + assert!(matches!( + access_chain[..], + [AccessExpr::Dot(Expr::Function(_))] + )); } _ => unreachable!(), } // `CONVERT` support - let dialects = all_dialects_where(|d| { - d.supports_methods() && d.supports_try_convert() && d.convert_type_before_value() - }); + let dialects = + all_dialects_where(|d| d.supports_try_convert() && d.convert_type_before_value()); let expr = dialects.verified_expr( "CONVERT(XML, 'abc').value('.', 'NVARCHAR(MAX)').value('.', 'NVARCHAR(MAX)')", ); match expr { - Expr::Method(Method { expr, method_chain }) => { - assert!(matches!(*expr, Expr::Convert { .. })); + Expr::CompoundFieldAccess { root, access_chain } => { + assert!(matches!(*root, Expr::Convert { .. })); assert!(matches!( - method_chain[..], - [Function { .. }, Function { .. }] + access_chain[..], + [ + AccessExpr::Dot(Expr::Function(_)), + AccessExpr::Dot(Expr::Function(_)) + ] )); } _ => unreachable!(), diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index b9a5202f0..25c152733 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2991,38 +2991,45 @@ fn parse_json_table_is_not_reserved() { fn test_composite_value() { let sql = "SELECT (on_hand.item).name FROM on_hand WHERE (on_hand.item).price > 9"; let select = pg().verified_only_select(sql); + + let Expr::CompoundFieldAccess { root, access_chain } = + expr_from_projection(&select.projection[0]) + else { + unreachable!("expected projection: got {:?}", &select.projection[0]); + }; assert_eq!( - SelectItem::UnnamedExpr(Expr::CompositeAccess { - key: Ident::new("name"), - expr: Box::new(Expr::Nested(Box::new(Expr::CompoundIdentifier(vec![ - Ident::new("on_hand"), - Ident::new("item") - ])))) - }), - select.projection[0] + root.as_ref(), + &Expr::Nested(Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("on_hand"), + Ident::new("item") + ]))) + ); + assert_eq!( + access_chain.as_slice(), + &[AccessExpr::Dot(Expr::Identifier(Ident::new("name")))] ); assert_eq!( - select.selection, - Some(Expr::BinaryOp { - left: Box::new(Expr::CompositeAccess { - key: Ident::new("price"), - expr: Box::new(Expr::Nested(Box::new(Expr::CompoundIdentifier(vec![ + select.selection.as_ref().unwrap(), + &Expr::BinaryOp { + left: Box::new(Expr::CompoundFieldAccess { + root: Expr::Nested(Box::new(Expr::CompoundIdentifier(vec![ Ident::new("on_hand"), Ident::new("item") - ])))) + ]))) + .into(), + access_chain: vec![AccessExpr::Dot(Expr::Identifier(Ident::new("price")))] }), op: BinaryOperator::Gt, right: Box::new(Expr::Value(number("9"))) - }) + } ); let sql = "SELECT (information_schema._pg_expandarray(ARRAY['i', 'i'])).n"; let select = pg().verified_only_select(sql); assert_eq!( - SelectItem::UnnamedExpr(Expr::CompositeAccess { - key: Ident::new("n"), - expr: Box::new(Expr::Nested(Box::new(Expr::Function(Function { + &Expr::CompoundFieldAccess { + root: Box::new(Expr::Nested(Box::new(Expr::Function(Function { name: ObjectName::from(vec![ Ident::new("information_schema"), Ident::new("_pg_expandarray") @@ -3046,9 +3053,10 @@ fn test_composite_value() { filter: None, over: None, within_group: vec![], - })))) - }), - select.projection[0] + })))), + access_chain: vec![AccessExpr::Dot(Expr::Identifier(Ident::new("n")))], + }, + expr_from_projection(&select.projection[0]) ); } From b482562618caa3efa89c2f42f87472b00a270926 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Wed, 19 Feb 2025 18:54:14 +0100 Subject: [PATCH 048/286] Add support for `EXECUTE IMMEDIATE` (#1717) --- src/ast/mod.rs | 31 +++++++++++++++++--------- src/dialect/bigquery.rs | 5 +++++ src/dialect/mod.rs | 5 +++++ src/dialect/snowflake.rs | 5 +++++ src/parser/mod.rs | 26 ++++++++++++++++------ tests/sqlparser_common.rs | 41 ++++++++++++++++++++++++++++++++-- tests/sqlparser_postgres.rs | 44 +++++++++++++++++++++++-------------- 7 files changed, 122 insertions(+), 35 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 1f05d590d..efdad164e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3269,18 +3269,21 @@ pub enum Statement { /// Note: this is a PostgreSQL-specific statement. Deallocate { name: Ident, prepare: bool }, /// ```sql - /// EXECUTE name [ ( parameter [, ...] ) ] [USING ] + /// An `EXECUTE` statement /// ``` /// - /// Note: this statement is supported by Postgres and MSSQL, with slight differences in syntax. - /// /// Postgres: /// MSSQL: + /// BigQuery: + /// Snowflake: Execute { - name: ObjectName, + name: Option, parameters: Vec, has_parentheses: bool, - using: Vec, + /// Is this an `EXECUTE IMMEDIATE` + immediate: bool, + into: Vec, + using: Vec, }, /// ```sql /// PREPARE name [ ( data_type [, ...] ) ] AS statement @@ -4889,6 +4892,8 @@ impl fmt::Display for Statement { name, parameters, has_parentheses, + immediate, + into, using, } => { let (open, close) = if *has_parentheses { @@ -4896,11 +4901,17 @@ impl fmt::Display for Statement { } else { (if parameters.is_empty() { "" } else { " " }, "") }; - write!( - f, - "EXECUTE {name}{open}{}{close}", - display_comma_separated(parameters), - )?; + write!(f, "EXECUTE")?; + if *immediate { + write!(f, " IMMEDIATE")?; + } + if let Some(name) = name { + write!(f, " {name}")?; + } + write!(f, "{open}{}{close}", display_comma_separated(parameters),)?; + if !into.is_empty() { + write!(f, " INTO {}", display_comma_separated(into))?; + } if !using.is_empty() { write!(f, " USING {}", display_comma_separated(using))?; }; diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index 5354a645b..b8e7e4cf0 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -110,6 +110,11 @@ impl Dialect for BigQueryDialect { true } + /// See + fn supports_execute_immediate(&self) -> bool { + true + } + // See fn supports_timestamp_versioning(&self) -> bool { true diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index cb86cf7c6..cf267a0f2 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -261,6 +261,11 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports `EXECUTE IMMEDIATE` statements. + fn supports_execute_immediate(&self) -> bool { + false + } + /// Returns true if the dialect supports the MATCH_RECOGNIZE operation. fn supports_match_recognize(&self) -> bool { false diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index bac9c49c7..6702b94b2 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -96,6 +96,11 @@ impl Dialect for SnowflakeDialect { true } + /// See + fn supports_execute_immediate(&self) -> bool { + true + } + fn supports_match_recognize(&self) -> bool { true } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9d75bcdb5..0ef0cc41a 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13849,7 +13849,14 @@ impl<'a> Parser<'a> { } pub fn parse_execute(&mut self) -> Result { - let name = self.parse_object_name(false)?; + let name = if self.dialect.supports_execute_immediate() + && self.parse_keyword(Keyword::IMMEDIATE) + { + None + } else { + let name = self.parse_object_name(false)?; + Some(name) + }; let has_parentheses = self.consume_token(&Token::LParen); @@ -13866,19 +13873,24 @@ impl<'a> Parser<'a> { self.expect_token(&Token::RParen)?; } - let mut using = vec![]; - if self.parse_keyword(Keyword::USING) { - using.push(self.parse_expr()?); + let into = if self.parse_keyword(Keyword::INTO) { + self.parse_comma_separated(Self::parse_identifier)? + } else { + vec![] + }; - while self.consume_token(&Token::Comma) { - using.push(self.parse_expr()?); - } + let using = if self.parse_keyword(Keyword::USING) { + self.parse_comma_separated(Self::parse_expr_with_alias)? + } else { + vec![] }; Ok(Statement::Execute { + immediate: name.is_none(), name, parameters, has_parentheses, + into, using, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 11e43daeb..3cc455b84 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -10745,7 +10745,7 @@ fn parse_call() { #[test] fn parse_execute_stored_procedure() { let expected = Statement::Execute { - name: ObjectName::from(vec![ + name: Some(ObjectName::from(vec![ Ident { value: "my_schema".to_string(), quote_style: None, @@ -10756,13 +10756,15 @@ fn parse_execute_stored_procedure() { quote_style: None, span: Span::empty(), }, - ]), + ])), parameters: vec![ Expr::Value(Value::NationalStringLiteral("param1".to_string())), Expr::Value(Value::NationalStringLiteral("param2".to_string())), ], has_parentheses: false, + immediate: false, using: vec![], + into: vec![], }; assert_eq!( // Microsoft SQL Server does not use parentheses around arguments for EXECUTE @@ -10779,6 +10781,41 @@ fn parse_execute_stored_procedure() { ); } +#[test] +fn parse_execute_immediate() { + let dialects = all_dialects_where(|d| d.supports_execute_immediate()); + + let expected = Statement::Execute { + parameters: vec![Expr::Value(Value::SingleQuotedString( + "SELECT 1".to_string(), + ))], + immediate: true, + using: vec![ExprWithAlias { + expr: Expr::Value(number("1")), + alias: Some(Ident::new("b")), + }], + into: vec![Ident::new("a")], + name: None, + has_parentheses: false, + }; + + let stmt = dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1' INTO a USING 1 AS b"); + assert_eq!(expected, stmt); + + dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1' INTO a, b USING 1 AS x, y"); + dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1' USING 1 AS x, y"); + dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1' INTO a, b"); + dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1'"); + dialects.verified_stmt("EXECUTE 'SELECT 1'"); + + assert_eq!( + ParserError::ParserError("Expected: identifier, found: ,".to_string()), + dialects + .parse_sql_statements("EXECUTE IMMEDIATE 'SELECT 1' USING 1 AS, y") + .unwrap_err() + ); +} + #[test] fn parse_create_table_collate() { pg_and_generic().verified_stmt("CREATE TABLE tbl (foo INT, bar TEXT COLLATE \"de_DE\")"); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 25c152733..7eb8086ea 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1659,10 +1659,12 @@ fn parse_execute() { assert_eq!( stmt, Statement::Execute { - name: ObjectName::from(vec!["a".into()]), + name: Some(ObjectName::from(vec!["a".into()])), parameters: vec![], has_parentheses: false, - using: vec![] + using: vec![], + immediate: false, + into: vec![] } ); @@ -1670,13 +1672,15 @@ fn parse_execute() { assert_eq!( stmt, Statement::Execute { - name: ObjectName::from(vec!["a".into()]), + name: Some(ObjectName::from(vec!["a".into()])), parameters: vec![ Expr::Value(number("1")), Expr::Value(Value::SingleQuotedString("t".to_string())) ], has_parentheses: true, - using: vec![] + using: vec![], + immediate: false, + into: vec![] } ); @@ -1685,23 +1689,31 @@ fn parse_execute() { assert_eq!( stmt, Statement::Execute { - name: ObjectName::from(vec!["a".into()]), + name: Some(ObjectName::from(vec!["a".into()])), parameters: vec![], has_parentheses: false, using: vec![ - Expr::Cast { - kind: CastKind::Cast, - expr: Box::new(Expr::Value(Value::Number("1337".parse().unwrap(), false))), - data_type: DataType::SmallInt(None), - format: None + ExprWithAlias { + expr: Expr::Cast { + kind: CastKind::Cast, + expr: Box::new(Expr::Value(Value::Number("1337".parse().unwrap(), false))), + data_type: DataType::SmallInt(None), + format: None + }, + alias: None }, - Expr::Cast { - kind: CastKind::Cast, - expr: Box::new(Expr::Value(Value::Number("7331".parse().unwrap(), false))), - data_type: DataType::SmallInt(None), - format: None + ExprWithAlias { + expr: Expr::Cast { + kind: CastKind::Cast, + expr: Box::new(Expr::Value(Value::Number("7331".parse().unwrap(), false))), + data_type: DataType::SmallInt(None), + format: None + }, + alias: None }, - ] + ], + immediate: false, + into: vec![] } ); } From 97f0be699120f2fffa57d9d6aaad0c53c484ea1b Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Wed, 19 Feb 2025 21:26:20 -0800 Subject: [PATCH 049/286] Treat COLLATE like any other column option (#1731) --- src/ast/ddl.rs | 6 ++-- src/ast/helpers/stmt_create_table.rs | 1 - src/ast/spans.rs | 8 ++--- src/parser/mod.rs | 15 +++------ tests/sqlparser_bigquery.rs | 5 --- tests/sqlparser_clickhouse.rs | 12 ------- tests/sqlparser_common.rs | 29 +++++------------ tests/sqlparser_duckdb.rs | 6 ---- tests/sqlparser_mssql.rs | 11 +++---- tests/sqlparser_mysql.rs | 48 +++++++++------------------- tests/sqlparser_postgres.rs | 42 +++++++----------------- tests/sqlparser_snowflake.rs | 24 +++++++------- tests/sqlparser_sqlite.rs | 4 --- 13 files changed, 60 insertions(+), 151 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index f90252000..1fbc45603 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1206,7 +1206,6 @@ impl fmt::Display for ProcedureParam { pub struct ColumnDef { pub name: Ident, pub data_type: DataType, - pub collation: Option, pub options: Vec, } @@ -1217,9 +1216,6 @@ impl fmt::Display for ColumnDef { } else { write!(f, "{} {}", self.name, self.data_type)?; } - if let Some(collation) = &self.collation { - write!(f, " COLLATE {collation}")?; - } for option in &self.options { write!(f, " {option}")?; } @@ -1566,6 +1562,7 @@ pub enum ColumnOption { /// - ... DialectSpecific(Vec), CharacterSet(ObjectName), + Collation(ObjectName), Comment(String), OnUpdate(Expr), /// `Generated`s are modifiers that follow a column definition in a `CREATE @@ -1665,6 +1662,7 @@ impl fmt::Display for ColumnOption { Check(expr) => write!(f, "CHECK ({expr})"), DialectSpecific(val) => write!(f, "{}", display_separated(val, " ")), CharacterSet(n) => write!(f, "CHARACTER SET {n}"), + Collation(n) => write!(f, "COLLATE {n}"), Comment(v) => write!(f, "COMMENT '{}'", escape_single_quote_string(v)), OnUpdate(expr) => write!(f, "ON UPDATE {expr}"), Generated { diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index 2a44cef3e..344e9dec9 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -47,7 +47,6 @@ use crate::parser::ParserError; /// .columns(vec![ColumnDef { /// name: Ident::new("c1"), /// data_type: DataType::Int(None), -/// collation: None, /// options: vec![], /// }]); /// // You can access internal elements with ease diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 091a22f04..de39e50d6 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -604,15 +604,10 @@ impl Spanned for ColumnDef { let ColumnDef { name, data_type: _, // enum - collation, options, } = self; - union_spans( - core::iter::once(name.span) - .chain(collation.iter().map(|i| i.span())) - .chain(options.iter().map(|i| i.span())), - ) + union_spans(core::iter::once(name.span).chain(options.iter().map(|i| i.span()))) } } @@ -767,6 +762,7 @@ impl Spanned for ColumnOption { ColumnOption::Check(expr) => expr.span(), ColumnOption::DialectSpecific(_) => Span::empty(), ColumnOption::CharacterSet(object_name) => object_name.span(), + ColumnOption::Collation(object_name) => object_name.span(), ColumnOption::Comment(_) => Span::empty(), ColumnOption::OnUpdate(expr) => expr.span(), ColumnOption::Generated { .. } => Span::empty(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0ef0cc41a..9c021d918 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6889,11 +6889,6 @@ impl<'a> Parser<'a> { } else { self.parse_data_type()? }; - let mut collation = if self.parse_keyword(Keyword::COLLATE) { - Some(self.parse_object_name(false)?) - } else { - None - }; let mut options = vec![]; loop { if self.parse_keyword(Keyword::CONSTRAINT) { @@ -6908,10 +6903,6 @@ impl<'a> Parser<'a> { } } else if let Some(option) = self.parse_optional_column_option()? { options.push(ColumnOptionDef { name: None, option }); - } else if dialect_of!(self is MySqlDialect | SnowflakeDialect | GenericDialect) - && self.parse_keyword(Keyword::COLLATE) - { - collation = Some(self.parse_object_name(false)?); } else { break; }; @@ -6919,7 +6910,6 @@ impl<'a> Parser<'a> { Ok(ColumnDef { name, data_type, - collation, options, }) } @@ -6956,6 +6946,10 @@ impl<'a> Parser<'a> { Ok(Some(ColumnOption::CharacterSet( self.parse_object_name(false)?, ))) + } else if self.parse_keywords(&[Keyword::COLLATE]) { + Ok(Some(ColumnOption::Collation( + self.parse_object_name(false)?, + ))) } else if self.parse_keywords(&[Keyword::NOT, Keyword::NULL]) { Ok(Some(ColumnOption::NotNull)) } else if self.parse_keywords(&[Keyword::COMMENT]) { @@ -9047,7 +9041,6 @@ impl<'a> Parser<'a> { Ok(ColumnDef { name, data_type, - collation: None, options: Vec::new(), // No constraints expected here }) } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index c5dfb27b5..52aa3b3b4 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -389,7 +389,6 @@ fn parse_create_table_with_unquoted_hyphen() { vec![ColumnDef { name: Ident::new("x"), data_type: DataType::Int64, - collation: None, options: vec![] },], columns @@ -427,7 +426,6 @@ fn parse_create_table_with_options() { ColumnDef { name: Ident::new("x"), data_type: DataType::Int64, - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -447,7 +445,6 @@ fn parse_create_table_with_options() { ColumnDef { name: Ident::new("y"), data_type: DataType::Bool, - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Options(vec![SqlOption::KeyValue { @@ -530,7 +527,6 @@ fn parse_nested_data_types() { ], StructBracketKind::AngleBrackets ), - collation: None, options: vec![], }, ColumnDef { @@ -544,7 +540,6 @@ fn parse_nested_data_types() { StructBracketKind::AngleBrackets ), ))), - collation: None, options: vec![], }, ] diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index b94d6f698..34f684c69 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -528,7 +528,6 @@ fn column_def(name: Ident, data_type: DataType) -> ColumnDef { ColumnDef { name, data_type, - collation: None, options: vec![], } } @@ -617,7 +616,6 @@ fn parse_create_table_with_nullable() { ColumnDef { name: "d".into(), data_type: DataType::Date32, - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Null @@ -661,7 +659,6 @@ fn parse_create_table_with_nested_data_types() { DataType::LowCardinality(Box::new(DataType::String(None))) ) ]), - collation: None, options: vec![], }, ColumnDef { @@ -678,7 +675,6 @@ fn parse_create_table_with_nested_data_types() { } ]) ))), - collation: None, options: vec![], }, ColumnDef { @@ -695,7 +691,6 @@ fn parse_create_table_with_nested_data_types() { )) }, ]), - collation: None, options: vec![], }, ColumnDef { @@ -704,7 +699,6 @@ fn parse_create_table_with_nested_data_types() { Box::new(DataType::String(None)), Box::new(DataType::UInt16) ), - collation: None, options: vec![], }, ] @@ -736,13 +730,11 @@ fn parse_create_table_with_primary_key() { ColumnDef { name: Ident::with_quote('`', "i"), data_type: DataType::Int(None), - collation: None, options: vec![], }, ColumnDef { name: Ident::with_quote('`', "k"), data_type: DataType::Int(None), - collation: None, options: vec![], }, ], @@ -814,7 +806,6 @@ fn parse_create_table_with_variant_default_expressions() { ColumnDef { name: Ident::new("a"), data_type: DataType::Datetime(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Materialized(Expr::Function(Function { @@ -836,7 +827,6 @@ fn parse_create_table_with_variant_default_expressions() { ColumnDef { name: Ident::new("b"), data_type: DataType::Datetime(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Ephemeral(Some(Expr::Function(Function { @@ -858,7 +848,6 @@ fn parse_create_table_with_variant_default_expressions() { ColumnDef { name: Ident::new("c"), data_type: DataType::Datetime(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Ephemeral(None) @@ -867,7 +856,6 @@ fn parse_create_table_with_variant_default_expressions() { ColumnDef { name: Ident::new("d"), data_type: DataType::String(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Alias(Expr::Function(Function { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 3cc455b84..a4e83be06 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3426,7 +3426,6 @@ fn parse_create_table() { length: 100, unit: None, })), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -3435,7 +3434,6 @@ fn parse_create_table() { ColumnDef { name: "lat".into(), data_type: DataType::Double(ExactNumberInfo::None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Null, @@ -3444,13 +3442,11 @@ fn parse_create_table() { ColumnDef { name: "lng".into(), data_type: DataType::Double(ExactNumberInfo::None), - collation: None, options: vec![], }, ColumnDef { name: "constrained".into(), data_type: DataType::Int(None), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -3483,7 +3479,6 @@ fn parse_create_table() { ColumnDef { name: "ref".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::ForeignKey { @@ -3498,7 +3493,6 @@ fn parse_create_table() { ColumnDef { name: "ref2".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::ForeignKey { @@ -3615,7 +3609,6 @@ fn parse_create_table_with_constraint_characteristics() { length: 100, unit: None, })), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -3624,7 +3617,6 @@ fn parse_create_table_with_constraint_characteristics() { ColumnDef { name: "lat".into(), data_type: DataType::Double(ExactNumberInfo::None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Null, @@ -3633,7 +3625,6 @@ fn parse_create_table_with_constraint_characteristics() { ColumnDef { name: "lng".into(), data_type: DataType::Double(ExactNumberInfo::None), - collation: None, options: vec![], }, ] @@ -3768,7 +3759,6 @@ fn parse_create_table_column_constraint_characteristics() { vec![ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Unique { @@ -3883,13 +3873,11 @@ fn parse_create_table_hive_array() { ColumnDef { name: Ident::new("name"), data_type: DataType::Int(None), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("val"), data_type: DataType::Array(expected), - collation: None, options: vec![], }, ], @@ -4255,7 +4243,6 @@ fn parse_create_external_table() { length: 100, unit: None, })), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -4264,7 +4251,6 @@ fn parse_create_external_table() { ColumnDef { name: "lat".into(), data_type: DataType::Double(ExactNumberInfo::None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Null, @@ -4273,7 +4259,6 @@ fn parse_create_external_table() { ColumnDef { name: "lng".into(), data_type: DataType::Double(ExactNumberInfo::None), - collation: None, options: vec![], }, ] @@ -4326,7 +4311,6 @@ fn parse_create_or_replace_external_table() { length: 100, unit: None, })), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -10818,7 +10802,14 @@ fn parse_execute_immediate() { #[test] fn parse_create_table_collate() { - pg_and_generic().verified_stmt("CREATE TABLE tbl (foo INT, bar TEXT COLLATE \"de_DE\")"); + all_dialects().verified_stmt("CREATE TABLE tbl (foo INT, bar TEXT COLLATE \"de_DE\")"); + // check ordering is preserved + all_dialects().verified_stmt( + "CREATE TABLE tbl (foo INT, bar TEXT CHARACTER SET utf8mb4 COLLATE \"de_DE\")", + ); + all_dialects().verified_stmt( + "CREATE TABLE tbl (foo INT, bar TEXT COLLATE \"de_DE\" CHARACTER SET utf8mb4)", + ); } #[test] @@ -10997,7 +10988,6 @@ fn test_parse_inline_comment() { vec![ColumnDef { name: Ident::new("id".to_string()), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: Comment("comment without equal".to_string()), @@ -13584,7 +13574,6 @@ fn parse_create_table_with_enum_types() { ], Some(8) ), - collation: None, options: vec![], }, ColumnDef { @@ -13602,7 +13591,6 @@ fn parse_create_table_with_enum_types() { ], Some(16) ), - collation: None, options: vec![], }, ColumnDef { @@ -13614,7 +13602,6 @@ fn parse_create_table_with_enum_types() { ], None ), - collation: None, options: vec![], } ], diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 43e12746c..05e60d556 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -60,7 +60,6 @@ fn test_struct() { vec![ColumnDef { name: "s".into(), data_type: struct_type1.clone(), - collation: None, options: vec![], }] ); @@ -75,7 +74,6 @@ fn test_struct() { Box::new(struct_type1), None )), - collation: None, options: vec![], }] ); @@ -120,7 +118,6 @@ fn test_struct() { Box::new(struct_type2), None )), - collation: None, options: vec![], }] ); @@ -671,7 +668,6 @@ fn test_duckdb_union_datatype() { field_name: "a".into(), field_type: DataType::Int(None) }]), - collation: Default::default(), options: Default::default() }, ColumnDef { @@ -686,7 +682,6 @@ fn test_duckdb_union_datatype() { field_type: DataType::Int(None) } ]), - collation: Default::default(), options: Default::default() }, ColumnDef { @@ -698,7 +693,6 @@ fn test_duckdb_union_datatype() { field_type: DataType::Int(None) }]) }]), - collation: Default::default(), options: Default::default() } ], diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 6865bdd4e..7b1277599 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1496,7 +1496,6 @@ fn parse_create_table_with_valid_options() { span: Span::empty(), }, data_type: Int(None,), - collation: None, options: vec![], }, ColumnDef { @@ -1506,7 +1505,7 @@ fn parse_create_table_with_valid_options() { span: Span::empty(), }, data_type: Int(None,), - collation: None, + options: vec![], }, ColumnDef { @@ -1516,7 +1515,7 @@ fn parse_create_table_with_valid_options() { span: Span::empty(), }, data_type: Int(None,), - collation: None, + options: vec![], }, ], @@ -1671,7 +1670,7 @@ fn parse_create_table_with_identity_column() { span: Span::empty(), }, data_type: Int(None,), - collation: None, + options: column_options, },], constraints: vec![], @@ -1815,7 +1814,7 @@ fn parse_mssql_varbinary_max_length() { vec![ColumnDef { name: Ident::new("var_binary_col"), data_type: Varbinary(Some(BinaryLength::Max)), - collation: None, + options: vec![] },], ); @@ -1840,7 +1839,7 @@ fn parse_mssql_varbinary_max_length() { vec![ColumnDef { name: Ident::new("var_binary_col"), data_type: Varbinary(Some(BinaryLength::IntegerLength { length: 50 })), - collation: None, + options: vec![] },], ); diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 48fcbbf9b..44c8350fa 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -636,7 +636,6 @@ fn parse_create_table_auto_increment() { vec![ColumnDef { name: Ident::new("bar"), data_type: DataType::Int(None), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -726,7 +725,6 @@ fn parse_create_table_primary_and_unique_key() { ColumnDef { name: Ident::new("id"), data_type: DataType::Int(None), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -746,7 +744,6 @@ fn parse_create_table_primary_and_unique_key() { ColumnDef { name: Ident::new("bar"), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -896,7 +893,6 @@ fn parse_create_table_set_enum() { ColumnDef { name: Ident::new("bar"), data_type: DataType::Set(vec!["a".to_string(), "b".to_string()]), - collation: None, options: vec![], }, ColumnDef { @@ -908,7 +904,6 @@ fn parse_create_table_set_enum() { ], None ), - collation: None, options: vec![], } ], @@ -935,7 +930,6 @@ fn parse_create_table_engine_default_charset() { vec![ColumnDef { name: Ident::new("id"), data_type: DataType::Int(Some(11)), - collation: None, options: vec![], },], columns @@ -968,7 +962,6 @@ fn parse_create_table_collate() { vec![ColumnDef { name: Ident::new("id"), data_type: DataType::Int(Some(11)), - collation: None, options: vec![], },], columns @@ -1016,7 +1009,6 @@ fn parse_create_table_comment_character_set() { vec![ColumnDef { name: Ident::new("s"), data_type: DataType::Text, - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -1063,7 +1055,6 @@ fn parse_quote_identifiers() { vec![ColumnDef { name: Ident::with_quote('`', "BEGIN"), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Unique { @@ -1326,31 +1317,26 @@ fn parse_create_table_with_minimum_display_width() { ColumnDef { name: Ident::new("bar_tinyint"), data_type: DataType::TinyInt(Some(3)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_smallint"), data_type: DataType::SmallInt(Some(5)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_mediumint"), data_type: DataType::MediumInt(Some(6)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_int"), data_type: DataType::Int(Some(11)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_bigint"), data_type: DataType::BigInt(Some(20)), - collation: None, options: vec![], } ], @@ -1372,31 +1358,26 @@ fn parse_create_table_unsigned() { ColumnDef { name: Ident::new("bar_tinyint"), data_type: DataType::UnsignedTinyInt(Some(3)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_smallint"), data_type: DataType::UnsignedSmallInt(Some(5)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_mediumint"), data_type: DataType::UnsignedMediumInt(Some(13)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_int"), data_type: DataType::UnsignedInt(Some(11)), - collation: None, options: vec![], }, ColumnDef { name: Ident::new("bar_bigint"), data_type: DataType::UnsignedBigInt(Some(20)), - collation: None, options: vec![], }, ], @@ -2159,7 +2140,6 @@ fn parse_alter_table_add_column() { column_def: ColumnDef { name: "b".into(), data_type: DataType::Int(None), - collation: None, options: vec![], }, column_position: Some(MySQLColumnPosition::First), @@ -2189,7 +2169,6 @@ fn parse_alter_table_add_column() { column_def: ColumnDef { name: "b".into(), data_type: DataType::Int(None), - collation: None, options: vec![], }, column_position: Some(MySQLColumnPosition::After(Ident { @@ -2229,7 +2208,6 @@ fn parse_alter_table_add_columns() { column_def: ColumnDef { name: "a".into(), data_type: DataType::Text, - collation: None, options: vec![], }, column_position: Some(MySQLColumnPosition::First), @@ -2240,7 +2218,6 @@ fn parse_alter_table_add_columns() { column_def: ColumnDef { name: "b".into(), data_type: DataType::Int(None), - collation: None, options: vec![], }, column_position: Some(MySQLColumnPosition::After(Ident { @@ -2593,7 +2570,6 @@ fn parse_table_column_option_on_update() { vec![ColumnDef { name: Ident::with_quote('`', "modification_time"), data_type: DataType::Datetime(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::OnUpdate(call("CURRENT_TIMESTAMP", [])), @@ -2878,21 +2854,27 @@ fn parse_convert_using() { #[test] fn parse_create_table_with_column_collate() { let sql = "CREATE TABLE tb (id TEXT CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci)"; - let canonical = "CREATE TABLE tb (id TEXT COLLATE utf8mb4_0900_ai_ci CHARACTER SET utf8mb4)"; - match mysql().one_statement_parses_to(sql, canonical) { + match mysql().verified_stmt(sql) { Statement::CreateTable(CreateTable { name, columns, .. }) => { assert_eq!(name.to_string(), "tb"); assert_eq!( vec![ColumnDef { name: Ident::new("id"), data_type: DataType::Text, - collation: Some(ObjectName::from(vec![Ident::new("utf8mb4_0900_ai_ci")])), - options: vec![ColumnOptionDef { - name: None, - option: ColumnOption::CharacterSet(ObjectName::from(vec![Ident::new( - "utf8mb4" - )])) - }], + options: vec![ + ColumnOptionDef { + name: None, + option: ColumnOption::CharacterSet(ObjectName::from(vec![Ident::new( + "utf8mb4" + )])) + }, + ColumnOptionDef { + name: None, + option: ColumnOption::Collation(ObjectName::from(vec![Ident::new( + "utf8mb4_0900_ai_ci" + )])) + } + ], },], columns ); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 7eb8086ea..d400792d9 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -363,7 +363,6 @@ fn parse_create_table_with_defaults() { ColumnDef { name: "customer_id".into(), data_type: DataType::Integer(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Default( @@ -374,7 +373,6 @@ fn parse_create_table_with_defaults() { ColumnDef { name: "store_id".into(), data_type: DataType::SmallInt(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -388,7 +386,6 @@ fn parse_create_table_with_defaults() { unit: None } )), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull, @@ -402,11 +399,18 @@ fn parse_create_table_with_defaults() { unit: None } )), - collation: Some(ObjectName::from(vec![Ident::with_quote('"', "es_ES")])), - options: vec![ColumnOptionDef { - name: None, - option: ColumnOption::NotNull, - }], + options: vec![ + ColumnOptionDef { + name: None, + option: ColumnOption::Collation(ObjectName::from(vec![ + Ident::with_quote('"', "es_ES") + ])), + }, + ColumnOptionDef { + name: None, + option: ColumnOption::NotNull, + } + ], }, ColumnDef { name: "email".into(), @@ -416,13 +420,11 @@ fn parse_create_table_with_defaults() { unit: None } )), - collation: None, options: vec![], }, ColumnDef { name: "address_id".into(), data_type: DataType::SmallInt(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull @@ -431,7 +433,6 @@ fn parse_create_table_with_defaults() { ColumnDef { name: "activebool".into(), data_type: DataType::Boolean, - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -446,7 +447,6 @@ fn parse_create_table_with_defaults() { ColumnDef { name: "create_date".into(), data_type: DataType::Date, - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -461,7 +461,6 @@ fn parse_create_table_with_defaults() { ColumnDef { name: "last_update".into(), data_type: DataType::Timestamp(None, TimezoneInfo::WithoutTimeZone), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -476,7 +475,6 @@ fn parse_create_table_with_defaults() { ColumnDef { name: "active".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::NotNull @@ -842,7 +840,6 @@ fn parse_alter_table_add_columns() { column_def: ColumnDef { name: "a".into(), data_type: DataType::Text, - collation: None, options: vec![], }, column_position: None, @@ -853,7 +850,6 @@ fn parse_alter_table_add_columns() { column_def: ColumnDef { name: "b".into(), data_type: DataType::Int(None), - collation: None, options: vec![], }, column_position: None, @@ -4291,37 +4287,31 @@ fn parse_create_table_with_alias() { ColumnDef { name: "int8_col".into(), data_type: DataType::Int8(None), - collation: None, options: vec![] }, ColumnDef { name: "int4_col".into(), data_type: DataType::Int4(None), - collation: None, options: vec![] }, ColumnDef { name: "int2_col".into(), data_type: DataType::Int2(None), - collation: None, options: vec![] }, ColumnDef { name: "float8_col".into(), data_type: DataType::Float8, - collation: None, options: vec![] }, ColumnDef { name: "float4_col".into(), data_type: DataType::Float4, - collation: None, options: vec![] }, ColumnDef { name: "bool_col".into(), data_type: DataType::Bool, - collation: None, options: vec![] }, ] @@ -4343,13 +4333,11 @@ fn parse_create_table_with_partition_by() { ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![] }, ColumnDef { name: "b".into(), data_type: DataType::Text, - collation: None, options: vec![] } ], @@ -5093,25 +5081,21 @@ fn parse_trigger_related_functions() { ColumnDef { name: "empname".into(), data_type: DataType::Text, - collation: None, options: vec![], }, ColumnDef { name: "salary".into(), data_type: DataType::Integer(None), - collation: None, options: vec![], }, ColumnDef { name: "last_date".into(), data_type: DataType::Timestamp(None, TimezoneInfo::None), - collation: None, options: vec![], }, ColumnDef { name: "last_user".into(), data_type: DataType::Text, - collation: None, options: vec![], }, ], @@ -5445,13 +5429,11 @@ fn parse_varbit_datatype() { ColumnDef { name: "x".into(), data_type: DataType::VarBit(None), - collation: None, options: vec![], }, ColumnDef { name: "y".into(), data_type: DataType::VarBit(Some(42)), - collation: None, options: vec![], } ] diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index b0c215a5a..76550b8e3 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -346,7 +346,6 @@ fn test_snowflake_create_table_column_comment() { name: None, option: ColumnOption::Comment("some comment".to_string()) }], - collation: None }], columns ) @@ -553,7 +552,6 @@ fn test_snowflake_create_table_with_autoincrement_columns() { ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Identity(IdentityPropertyKind::Autoincrement( @@ -567,7 +565,6 @@ fn test_snowflake_create_table_with_autoincrement_columns() { ColumnDef { name: "b".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Identity(IdentityPropertyKind::Autoincrement( @@ -586,7 +583,6 @@ fn test_snowflake_create_table_with_autoincrement_columns() { ColumnDef { name: "c".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Identity(IdentityPropertyKind::Identity( @@ -600,7 +596,6 @@ fn test_snowflake_create_table_with_autoincrement_columns() { ColumnDef { name: "d".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Identity(IdentityPropertyKind::Identity( @@ -634,8 +629,12 @@ fn test_snowflake_create_table_with_collated_column() { vec![ColumnDef { name: "a".into(), data_type: DataType::Text, - collation: Some(ObjectName::from(vec![Ident::with_quote('\'', "de_DE")])), - options: vec![] + options: vec![ColumnOptionDef { + name: None, + option: ColumnOption::Collation(ObjectName::from(vec![Ident::with_quote( + '\'', "de_DE" + )])), + }] },] ); } @@ -674,7 +673,6 @@ fn test_snowflake_create_table_with_columns_masking_policy() { vec![ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Policy(ColumnPolicy::MaskingPolicy( @@ -709,7 +707,6 @@ fn test_snowflake_create_table_with_columns_projection_policy() { vec![ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Policy(ColumnPolicy::ProjectionPolicy( @@ -747,7 +744,6 @@ fn test_snowflake_create_table_with_columns_tags() { vec![ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![ColumnOptionDef { name: None, option: ColumnOption::Tags(TagsColumnOption { @@ -782,7 +778,6 @@ fn test_snowflake_create_table_with_several_column_options() { ColumnDef { name: "a".into(), data_type: DataType::Int(None), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -818,8 +813,13 @@ fn test_snowflake_create_table_with_several_column_options() { ColumnDef { name: "b".into(), data_type: DataType::Text, - collation: Some(ObjectName::from(vec![Ident::with_quote('\'', "de_DE")])), options: vec![ + ColumnOptionDef { + name: None, + option: ColumnOption::Collation(ObjectName::from(vec![ + Ident::with_quote('\'', "de_DE") + ])), + }, ColumnOptionDef { name: None, option: ColumnOption::Policy(ColumnPolicy::ProjectionPolicy( diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 3a612f70a..c17743305 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -214,7 +214,6 @@ fn parse_create_table_auto_increment() { vec![ColumnDef { name: "bar".into(), data_type: DataType::Int(None), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -243,7 +242,6 @@ fn parse_create_table_primary_key_asc_desc() { let expected_column_def = |kind| ColumnDef { name: "bar".into(), data_type: DataType::Int(None), - collation: None, options: vec![ ColumnOptionDef { name: None, @@ -286,13 +284,11 @@ fn parse_create_sqlite_quote() { ColumnDef { name: Ident::with_quote('"', "KEY"), data_type: DataType::Int(None), - collation: None, options: vec![], }, ColumnDef { name: Ident::with_quote('[', "INDEX"), data_type: DataType::Int(None), - collation: None, options: vec![], }, ], From 339239d0c54fd0fe12e05d8de868340d07030ffa Mon Sep 17 00:00:00 2001 From: benrsatori Date: Thu, 20 Feb 2025 20:50:32 +0200 Subject: [PATCH 050/286] Add support for PostgreSQL/Redshift geometric operators (#1723) --- src/ast/data_type.rs | 36 ++++ src/ast/mod.rs | 12 +- src/ast/operator.rs | 84 +++++++++ src/dialect/mod.rs | 36 +++- src/dialect/postgresql.rs | 4 + src/dialect/redshift.rs | 4 + src/keywords.rs | 6 + src/parser/mod.rs | 107 ++++++++++- src/tokenizer.rs | 184 +++++++++++++++++- tests/sqlparser_common.rs | 358 ++++++++++++++++++++++++++++++++++++ tests/sqlparser_postgres.rs | 4 - 11 files changed, 809 insertions(+), 26 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index bd46f8bdb..cae8ca8f0 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -386,6 +386,10 @@ pub enum DataType { /// /// [bigquery]: https://cloud.google.com/bigquery/docs/user-defined-functions#templated-sql-udf-parameters AnyType, + /// geometric type + /// + /// [Postgres]: https://www.postgresql.org/docs/9.5/functions-geometry.html + GeometricType(GeometricTypeKind), } impl fmt::Display for DataType { @@ -639,6 +643,7 @@ impl fmt::Display for DataType { DataType::Trigger => write!(f, "TRIGGER"), DataType::AnyType => write!(f, "ANY TYPE"), DataType::Table(fields) => write!(f, "TABLE({})", display_comma_separated(fields)), + DataType::GeometricType(kind) => write!(f, "{}", kind), } } } @@ -915,3 +920,34 @@ pub enum ArrayElemTypeDef { /// `Array(Int64)` Parenthesis(Box), } + +/// Represents different types of geometric shapes which are commonly used in +/// PostgreSQL/Redshift for spatial operations and geometry-related computations. +/// +/// [Postgres]: https://www.postgresql.org/docs/9.5/functions-geometry.html +#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash, PartialOrd, Ord)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum GeometricTypeKind { + Point, + Line, + LineSegment, + GeometricBox, + GeometricPath, + Polygon, + Circle, +} + +impl fmt::Display for GeometricTypeKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + GeometricTypeKind::Point => write!(f, "point"), + GeometricTypeKind::Line => write!(f, "line"), + GeometricTypeKind::LineSegment => write!(f, "lseg"), + GeometricTypeKind::GeometricBox => write!(f, "box"), + GeometricTypeKind::GeometricPath => write!(f, "path"), + GeometricTypeKind::Polygon => write!(f, "polygon"), + GeometricTypeKind::Circle => write!(f, "circle"), + } + } +} diff --git a/src/ast/mod.rs b/src/ast/mod.rs index efdad164e..dc50871ce 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -95,6 +95,8 @@ use crate::ast::helpers::stmt_data_loading::{ #[cfg(feature = "visitor")] pub use visitor::*; +pub use self::data_type::GeometricTypeKind; + mod data_type; mod dcl; mod ddl; @@ -1513,7 +1515,15 @@ impl fmt::Display for Expr { Expr::UnaryOp { op, expr } => { if op == &UnaryOperator::PGPostfixFactorial { write!(f, "{expr}{op}") - } else if op == &UnaryOperator::Not { + } else if matches!( + op, + UnaryOperator::Not + | UnaryOperator::Hash + | UnaryOperator::AtDashAt + | UnaryOperator::DoubleAt + | UnaryOperator::QuestionDash + | UnaryOperator::QuestionPipe + ) { write!(f, "{op} {expr}") } else { write!(f, "{op}{expr}") diff --git a/src/ast/operator.rs b/src/ast/operator.rs index 1f9a6b82b..66a35fee5 100644 --- a/src/ast/operator.rs +++ b/src/ast/operator.rs @@ -53,6 +53,21 @@ pub enum UnaryOperator { PGAbs, /// Unary logical not operator: e.g. `! false` (Hive-specific) BangNot, + /// `#` Number of points in path or polygon (PostgreSQL/Redshift geometric operator) + /// see + Hash, + /// `@-@` Length or circumference (PostgreSQL/Redshift geometric operator) + /// see + AtDashAt, + /// `@@` Center (PostgreSQL/Redshift geometric operator) + /// see + DoubleAt, + /// `?-` Is horizontal? (PostgreSQL/Redshift geometric operator) + /// see + QuestionDash, + /// `?|` Is vertical? (PostgreSQL/Redshift geometric operator) + /// see + QuestionPipe, } impl fmt::Display for UnaryOperator { @@ -68,6 +83,11 @@ impl fmt::Display for UnaryOperator { UnaryOperator::PGPrefixFactorial => "!!", UnaryOperator::PGAbs => "@", UnaryOperator::BangNot => "!", + UnaryOperator::Hash => "#", + UnaryOperator::AtDashAt => "@-@", + UnaryOperator::DoubleAt => "@@", + UnaryOperator::QuestionDash => "?-", + UnaryOperator::QuestionPipe => "?|", }) } } @@ -253,6 +273,54 @@ pub enum BinaryOperator { /// Specifies a test for an overlap between two datetime periods: /// Overlaps, + /// `##` Point of closest proximity (PostgreSQL/Redshift geometric operator) + /// See + DoubleHash, + /// `<->` Distance between (PostgreSQL/Redshift geometric operator) + /// See + LtDashGt, + /// `&<` Overlaps to left? (PostgreSQL/Redshift geometric operator) + /// See + AndLt, + /// `&>` Overlaps to right? (PostgreSQL/Redshift geometric operator) + /// See + AndGt, + /// `<<|` Is strictly below? (PostgreSQL/Redshift geometric operator) + /// See + LtLtPipe, + /// `|>>` Is strictly above? (PostgreSQL/Redshift geometric operator) + /// See + PipeGtGt, + /// `&<|` Does not extend above? (PostgreSQL/Redshift geometric operator) + /// See + AndLtPipe, + /// `|&>` Does not extend below? (PostgreSQL/Redshift geometric operator) + /// See + PipeAndGt, + /// `<^` Is below? (PostgreSQL/Redshift geometric operator) + /// See + LtCaret, + /// `>^` Is above? (PostgreSQL/Redshift geometric operator) + /// See + GtCaret, + /// `?#` Intersects? (PostgreSQL/Redshift geometric operator) + /// See + QuestionHash, + /// `?-` Is horizontal? (PostgreSQL/Redshift geometric operator) + /// See + QuestionDash, + /// `?-|` Is perpendicular? (PostgreSQL/Redshift geometric operator) + /// See + QuestionDashPipe, + /// `?||` Are Parallel? (PostgreSQL/Redshift geometric operator) + /// See + QuestionDoublePipe, + /// `@` Contained or on? (PostgreSQL/Redshift geometric operator) + /// See + At, + /// `~=` Same as? (PostgreSQL/Redshift geometric operator) + /// See + TildeEq, } impl fmt::Display for BinaryOperator { @@ -310,6 +378,22 @@ impl fmt::Display for BinaryOperator { write!(f, "OPERATOR({})", display_separated(idents, ".")) } BinaryOperator::Overlaps => f.write_str("OVERLAPS"), + BinaryOperator::DoubleHash => f.write_str("##"), + BinaryOperator::LtDashGt => f.write_str("<->"), + BinaryOperator::AndLt => f.write_str("&<"), + BinaryOperator::AndGt => f.write_str("&>"), + BinaryOperator::LtLtPipe => f.write_str("<<|"), + BinaryOperator::PipeGtGt => f.write_str("|>>"), + BinaryOperator::AndLtPipe => f.write_str("&<|"), + BinaryOperator::PipeAndGt => f.write_str("|&>"), + BinaryOperator::LtCaret => f.write_str("<^"), + BinaryOperator::GtCaret => f.write_str(">^"), + BinaryOperator::QuestionHash => f.write_str("?#"), + BinaryOperator::QuestionDash => f.write_str("?-"), + BinaryOperator::QuestionDashPipe => f.write_str("?-|"), + BinaryOperator::QuestionDoublePipe => f.write_str("?||"), + BinaryOperator::At => f.write_str("@"), + BinaryOperator::TildeEq => f.write_str("~="), } } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index cf267a0f2..86e23c86f 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -599,18 +599,34 @@ pub trait Dialect: Debug + Any { | Token::ExclamationMarkDoubleTilde | Token::ExclamationMarkDoubleTildeAsterisk | Token::Spaceship => Ok(p!(Eq)), - Token::Pipe => Ok(p!(Pipe)), + Token::Pipe + | Token::QuestionMarkDash + | Token::DoubleSharp + | Token::Overlap + | Token::AmpersandLeftAngleBracket + | Token::AmpersandRightAngleBracket + | Token::QuestionMarkDashVerticalBar + | Token::AmpersandLeftAngleBracketVerticalBar + | Token::VerticalBarAmpersandRightAngleBracket + | Token::TwoWayArrow + | Token::LeftAngleBracketCaret + | Token::RightAngleBracketCaret + | Token::QuestionMarkSharp + | Token::QuestionMarkDoubleVerticalBar + | Token::QuestionPipe + | Token::TildeEqual + | Token::AtSign + | Token::ShiftLeftVerticalBar + | Token::VerticalBarShiftRight => Ok(p!(Pipe)), Token::Caret | Token::Sharp | Token::ShiftRight | Token::ShiftLeft => Ok(p!(Caret)), Token::Ampersand => Ok(p!(Ampersand)), Token::Plus | Token::Minus => Ok(p!(PlusMinus)), Token::Mul | Token::Div | Token::DuckIntDiv | Token::Mod | Token::StringConcat => { Ok(p!(MulDivModOp)) } - Token::DoubleColon - | Token::ExclamationMark - | Token::LBracket - | Token::Overlap - | Token::CaretAt => Ok(p!(DoubleColon)), + Token::DoubleColon | Token::ExclamationMark | Token::LBracket | Token::CaretAt => { + Ok(p!(DoubleColon)) + } Token::Arrow | Token::LongArrow | Token::HashArrow @@ -622,7 +638,6 @@ pub trait Dialect: Debug + Any { | Token::AtAt | Token::Question | Token::QuestionAnd - | Token::QuestionPipe | Token::CustomBinaryOperator(_) => Ok(p!(PgOther)), _ => Ok(self.prec_unknown()), } @@ -921,6 +936,13 @@ pub trait Dialect: Debug + Any { fn supports_array_typedef_size(&self) -> bool { false } + /// Returns true if the dialect supports geometric types. + /// + /// Postgres: + /// e.g. @@ circle '((0,0),10)' + fn supports_geometric_types(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 3a3f0e4c3..a20cfac4c 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -250,4 +250,8 @@ impl Dialect for PostgreSqlDialect { fn supports_array_typedef_size(&self) -> bool { true } + + fn supports_geometric_types(&self) -> bool { + true + } } diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index a4522bbf8..3dda762c3 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -113,4 +113,8 @@ impl Dialect for RedshiftSqlDialect { fn supports_string_escape_constant(&self) -> bool { true } + + fn supports_geometric_types(&self) -> bool { + true + } } diff --git a/src/keywords.rs b/src/keywords.rs index a67f6bc76..46c40fbaa 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -141,6 +141,7 @@ define_keywords!( BOOL, BOOLEAN, BOTH, + BOX, BROWSE, BTREE, BUCKET, @@ -175,6 +176,7 @@ define_keywords!( CHARSET, CHAR_LENGTH, CHECK, + CIRCLE, CLEAR, CLOB, CLONE, @@ -478,6 +480,7 @@ define_keywords!( LIKE, LIKE_REGEX, LIMIT, + LINE, LINES, LIST, LISTEN, @@ -499,6 +502,7 @@ define_keywords!( LOWER, LOW_PRIORITY, LS, + LSEG, MACRO, MANAGE, MANAGED, @@ -653,7 +657,9 @@ define_keywords!( PLACING, PLAN, PLANS, + POINT, POLICY, + POLYGON, POOL, PORTION, POSITION, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9c021d918..69268bc51 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1220,7 +1220,17 @@ impl<'a> Parser<'a> { Keyword::MAP if *self.peek_token_ref() == Token::LBrace && self.dialect.support_map_literal_syntax() => { Ok(Some(self.parse_duckdb_map_literal()?)) } - _ => Ok(None) + _ if self.dialect.supports_geometric_types() => match w.keyword { + Keyword::CIRCLE => Ok(Some(self.parse_geometric_type(GeometricTypeKind::Circle)?)), + Keyword::BOX => Ok(Some(self.parse_geometric_type(GeometricTypeKind::GeometricBox)?)), + Keyword::PATH => Ok(Some(self.parse_geometric_type(GeometricTypeKind::GeometricPath)?)), + Keyword::LINE => Ok(Some(self.parse_geometric_type(GeometricTypeKind::Line)?)), + Keyword::LSEG => Ok(Some(self.parse_geometric_type(GeometricTypeKind::LineSegment)?)), + Keyword::POINT => Ok(Some(self.parse_geometric_type(GeometricTypeKind::Point)?)), + Keyword::POLYGON => Ok(Some(self.parse_geometric_type(GeometricTypeKind::Polygon)?)), + _ => Ok(None), + }, + _ => Ok(None), } } @@ -1400,6 +1410,33 @@ impl<'a> Parser<'a> { ), }) } + tok @ Token::Sharp + | tok @ Token::AtDashAt + | tok @ Token::AtAt + | tok @ Token::QuestionMarkDash + | tok @ Token::QuestionPipe + if self.dialect.supports_geometric_types() => + { + let op = match tok { + Token::Sharp => UnaryOperator::Hash, + Token::AtDashAt => UnaryOperator::AtDashAt, + Token::AtAt => UnaryOperator::DoubleAt, + Token::QuestionMarkDash => UnaryOperator::QuestionDash, + Token::QuestionPipe => UnaryOperator::QuestionPipe, + _ => { + return Err(ParserError::ParserError(format!( + "Unexpected token in unary operator parsing: {:?}", + tok + ))) + } + }; + Ok(Expr::UnaryOp { + op, + expr: Box::new( + self.parse_subexpr(self.dialect.prec_value(Precedence::PlusMinus))?, + ), + }) + } Token::EscapedStringLiteral(_) if dialect_is!(dialect is PostgreSqlDialect | GenericDialect) => { self.prev_token(); @@ -1465,6 +1502,14 @@ impl<'a> Parser<'a> { } } + fn parse_geometric_type(&mut self, kind: GeometricTypeKind) -> Result { + let value: Value = self.parse_value()?; + Ok(Expr::TypedString { + data_type: DataType::GeometricType(kind), + value, + }) + } + /// Try to parse an [Expr::CompoundFieldAccess] like `a.b.c` or `a.b[1].c`. /// If all the fields are `Expr::Identifier`s, return an [Expr::CompoundIdentifier] instead. /// If only the root exists, return the root. @@ -3070,15 +3115,18 @@ impl<'a> Parser<'a> { Token::DuckIntDiv if dialect_is!(dialect is DuckDbDialect | GenericDialect) => { Some(BinaryOperator::DuckIntegerDivide) } - Token::ShiftLeft if dialect_is!(dialect is PostgreSqlDialect | DuckDbDialect | GenericDialect) => { + Token::ShiftLeft if dialect_is!(dialect is PostgreSqlDialect | DuckDbDialect | GenericDialect | RedshiftSqlDialect) => { Some(BinaryOperator::PGBitwiseShiftLeft) } - Token::ShiftRight if dialect_is!(dialect is PostgreSqlDialect | DuckDbDialect | GenericDialect) => { + Token::ShiftRight if dialect_is!(dialect is PostgreSqlDialect | DuckDbDialect | GenericDialect | RedshiftSqlDialect) => { Some(BinaryOperator::PGBitwiseShiftRight) } - Token::Sharp if dialect_is!(dialect is PostgreSqlDialect) => { + Token::Sharp if dialect_is!(dialect is PostgreSqlDialect | RedshiftSqlDialect) => { Some(BinaryOperator::PGBitwiseXor) } + Token::Overlap if dialect_is!(dialect is PostgreSqlDialect | RedshiftSqlDialect) => { + Some(BinaryOperator::PGOverlap) + } Token::Overlap if dialect_is!(dialect is PostgreSqlDialect | GenericDialect) => { Some(BinaryOperator::PGOverlap) } @@ -3106,6 +3154,57 @@ impl<'a> Parser<'a> { Token::QuestionAnd => Some(BinaryOperator::QuestionAnd), Token::QuestionPipe => Some(BinaryOperator::QuestionPipe), Token::CustomBinaryOperator(s) => Some(BinaryOperator::Custom(s.clone())), + Token::DoubleSharp if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::DoubleHash) + } + + Token::AmpersandLeftAngleBracket if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::AndLt) + } + Token::AmpersandRightAngleBracket if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::AndGt) + } + Token::QuestionMarkDash if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::QuestionDash) + } + Token::AmpersandLeftAngleBracketVerticalBar + if self.dialect.supports_geometric_types() => + { + Some(BinaryOperator::AndLtPipe) + } + Token::VerticalBarAmpersandRightAngleBracket + if self.dialect.supports_geometric_types() => + { + Some(BinaryOperator::PipeAndGt) + } + Token::TwoWayArrow if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::LtDashGt) + } + Token::LeftAngleBracketCaret if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::LtCaret) + } + Token::RightAngleBracketCaret if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::GtCaret) + } + Token::QuestionMarkSharp if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::QuestionHash) + } + Token::QuestionMarkDoubleVerticalBar if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::QuestionDoublePipe) + } + Token::QuestionMarkDashVerticalBar if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::QuestionDashPipe) + } + Token::TildeEqual if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::TildeEq) + } + Token::ShiftLeftVerticalBar if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::LtLtPipe) + } + Token::VerticalBarShiftRight if self.dialect.supports_geometric_types() => { + Some(BinaryOperator::PipeGtGt) + } + Token::AtSign if self.dialect.supports_geometric_types() => Some(BinaryOperator::At), Token::Word(w) => match w.keyword { Keyword::AND => Some(BinaryOperator::And), diff --git a/src/tokenizer.rs b/src/tokenizer.rs index d4e530c9d..bc0f0efeb 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -170,8 +170,10 @@ pub enum Token { RBrace, /// Right Arrow `=>` RArrow, - /// Sharp `#` used for PostgreSQL Bitwise XOR operator + /// Sharp `#` used for PostgreSQL Bitwise XOR operator, also PostgreSQL/Redshift geometrical unary/binary operator (Number of points in path or polygon/Intersection) Sharp, + /// `##` PostgreSQL/Redshift geometrical binary operator (Point of closest proximity) + DoubleSharp, /// Tilde `~` used for PostgreSQL Bitwise NOT operator or case sensitive match regular expression operator Tilde, /// `~*` , a case insensitive match regular expression operator in PostgreSQL @@ -198,7 +200,7 @@ pub enum Token { ExclamationMark, /// Double Exclamation Mark `!!` used for PostgreSQL prefix factorial operator DoubleExclamationMark, - /// AtSign `@` used for PostgreSQL abs operator + /// AtSign `@` used for PostgreSQL abs operator, also PostgreSQL/Redshift geometrical unary/binary operator (Center, Contained or on) AtSign, /// `^@`, a "starts with" string operator in PostgreSQL CaretAt, @@ -214,6 +216,36 @@ pub enum Token { LongArrow, /// `#>`, extracts JSON sub-object at the specified path HashArrow, + /// `@-@` PostgreSQL/Redshift geometrical unary operator (Length or circumference) + AtDashAt, + /// `?-` PostgreSQL/Redshift geometrical unary/binary operator (Is horizontal?/Are horizontally aligned?) + QuestionMarkDash, + /// `&<` PostgreSQL/Redshift geometrical binary operator (Overlaps to left?) + AmpersandLeftAngleBracket, + /// `&>` PostgreSQL/Redshift geometrical binary operator (Overlaps to right?)` + AmpersandRightAngleBracket, + /// `&<|` PostgreSQL/Redshift geometrical binary operator (Does not extend above?)` + AmpersandLeftAngleBracketVerticalBar, + /// `|&>` PostgreSQL/Redshift geometrical binary operator (Does not extend below?)` + VerticalBarAmpersandRightAngleBracket, + /// `<->` PostgreSQL/Redshift geometrical binary operator (Distance between) + TwoWayArrow, + /// `<^` PostgreSQL/Redshift geometrical binary operator (Is below?) + LeftAngleBracketCaret, + /// `>^` PostgreSQL/Redshift geometrical binary operator (Is above?) + RightAngleBracketCaret, + /// `?#` PostgreSQL/Redshift geometrical binary operator (Intersects or overlaps) + QuestionMarkSharp, + /// `?-|` PostgreSQL/Redshift geometrical binary operator (Is perpendicular?) + QuestionMarkDashVerticalBar, + /// `?||` PostgreSQL/Redshift geometrical binary operator (Are parallel?) + QuestionMarkDoubleVerticalBar, + /// `~=` PostgreSQL/Redshift geometrical binary operator (Same as) + TildeEqual, + /// `<<| PostgreSQL/Redshift geometrical binary operator (Is strictly below?) + ShiftLeftVerticalBar, + /// `|>> PostgreSQL/Redshift geometrical binary operator (Is strictly above?) + VerticalBarShiftRight, /// `#>>`, extracts JSON sub-object at the specified path as text HashLongArrow, /// jsonb @> jsonb -> boolean: Test whether left json contains the right json @@ -303,6 +335,7 @@ impl fmt::Display for Token { Token::RBrace => f.write_str("}"), Token::RArrow => f.write_str("=>"), Token::Sharp => f.write_str("#"), + Token::DoubleSharp => f.write_str("##"), Token::ExclamationMark => f.write_str("!"), Token::DoubleExclamationMark => f.write_str("!!"), Token::Tilde => f.write_str("~"), @@ -320,6 +353,21 @@ impl fmt::Display for Token { Token::Overlap => f.write_str("&&"), Token::PGSquareRoot => f.write_str("|/"), Token::PGCubeRoot => f.write_str("||/"), + Token::AtDashAt => f.write_str("@-@"), + Token::QuestionMarkDash => f.write_str("?-"), + Token::AmpersandLeftAngleBracket => f.write_str("&<"), + Token::AmpersandRightAngleBracket => f.write_str("&>"), + Token::AmpersandLeftAngleBracketVerticalBar => f.write_str("&<|"), + Token::VerticalBarAmpersandRightAngleBracket => f.write_str("|&>"), + Token::TwoWayArrow => f.write_str("<->"), + Token::LeftAngleBracketCaret => f.write_str("<^"), + Token::RightAngleBracketCaret => f.write_str(">^"), + Token::QuestionMarkSharp => f.write_str("?#"), + Token::QuestionMarkDashVerticalBar => f.write_str("?-|"), + Token::QuestionMarkDoubleVerticalBar => f.write_str("?||"), + Token::TildeEqual => f.write_str("~="), + Token::ShiftLeftVerticalBar => f.write_str("<<|"), + Token::VerticalBarShiftRight => f.write_str("|>>"), Token::Placeholder(ref s) => write!(f, "{s}"), Token::Arrow => write!(f, "->"), Token::LongArrow => write!(f, "->>"), @@ -1308,6 +1356,28 @@ impl<'a> Tokenizer<'a> { _ => self.start_binop(chars, "||", Token::StringConcat), } } + Some('&') if self.dialect.supports_geometric_types() => { + chars.next(); // consume + match chars.peek() { + Some('>') => self.consume_for_binop( + chars, + "|&>", + Token::VerticalBarAmpersandRightAngleBracket, + ), + _ => self.start_binop_opt(chars, "|&", None), + } + } + Some('>') if self.dialect.supports_geometric_types() => { + chars.next(); // consume + match chars.peek() { + Some('>') => self.consume_for_binop( + chars, + "|>>", + Token::VerticalBarShiftRight, + ), + _ => self.start_binop_opt(chars, "|>", None), + } + } // Bitshift '|' operator _ => self.start_binop(chars, "|", Token::Pipe), } @@ -1356,8 +1426,34 @@ impl<'a> Tokenizer<'a> { _ => self.start_binop(chars, "<=", Token::LtEq), } } + Some('|') if self.dialect.supports_geometric_types() => { + self.consume_for_binop(chars, "<<|", Token::ShiftLeftVerticalBar) + } Some('>') => self.consume_for_binop(chars, "<>", Token::Neq), + Some('<') if self.dialect.supports_geometric_types() => { + chars.next(); // consume + match chars.peek() { + Some('|') => self.consume_for_binop( + chars, + "<<|", + Token::ShiftLeftVerticalBar, + ), + _ => self.start_binop(chars, "<<", Token::ShiftLeft), + } + } Some('<') => self.consume_for_binop(chars, "<<", Token::ShiftLeft), + Some('-') if self.dialect.supports_geometric_types() => { + chars.next(); // consume + match chars.peek() { + Some('>') => { + self.consume_for_binop(chars, "<->", Token::TwoWayArrow) + } + _ => self.start_binop_opt(chars, "<-", None), + } + } + Some('^') if self.dialect.supports_geometric_types() => { + self.consume_for_binop(chars, "<^", Token::LeftAngleBracketCaret) + } Some('@') => self.consume_for_binop(chars, "<@", Token::ArrowAt), _ => self.start_binop(chars, "<", Token::Lt), } @@ -1367,6 +1463,9 @@ impl<'a> Tokenizer<'a> { match chars.peek() { Some('=') => self.consume_for_binop(chars, ">=", Token::GtEq), Some('>') => self.consume_for_binop(chars, ">>", Token::ShiftRight), + Some('^') if self.dialect.supports_geometric_types() => { + self.consume_for_binop(chars, ">^", Token::RightAngleBracketCaret) + } _ => self.start_binop(chars, ">", Token::Gt), } } @@ -1385,6 +1484,22 @@ impl<'a> Tokenizer<'a> { '&' => { chars.next(); // consume the '&' match chars.peek() { + Some('>') if self.dialect.supports_geometric_types() => { + chars.next(); + self.consume_and_return(chars, Token::AmpersandRightAngleBracket) + } + Some('<') if self.dialect.supports_geometric_types() => { + chars.next(); // consume + match chars.peek() { + Some('|') => self.consume_and_return( + chars, + Token::AmpersandLeftAngleBracketVerticalBar, + ), + _ => { + self.start_binop(chars, "&<", Token::AmpersandLeftAngleBracket) + } + } + } Some('&') => { chars.next(); // consume the second '&' self.start_binop(chars, "&&", Token::Overlap) @@ -1415,6 +1530,9 @@ impl<'a> Tokenizer<'a> { chars.next(); // consume match chars.peek() { Some('*') => self.consume_for_binop(chars, "~*", Token::TildeAsterisk), + Some('=') if self.dialect.supports_geometric_types() => { + self.consume_for_binop(chars, "~=", Token::TildeEqual) + } Some('~') => { chars.next(); match chars.peek() { @@ -1441,6 +1559,9 @@ impl<'a> Tokenizer<'a> { } } Some(' ') => Ok(Some(Token::Sharp)), + Some('#') if self.dialect.supports_geometric_types() => { + self.consume_for_binop(chars, "##", Token::DoubleSharp) + } Some(sch) if self.dialect.is_identifier_start('#') => { self.tokenize_identifier_or_keyword([ch, *sch], chars) } @@ -1450,6 +1571,16 @@ impl<'a> Tokenizer<'a> { '@' => { chars.next(); match chars.peek() { + Some('@') if self.dialect.supports_geometric_types() => { + self.consume_and_return(chars, Token::AtAt) + } + Some('-') if self.dialect.supports_geometric_types() => { + chars.next(); + match chars.peek() { + Some('@') => self.consume_and_return(chars, Token::AtDashAt), + _ => self.start_binop_opt(chars, "@-", None), + } + } Some('>') => self.consume_and_return(chars, Token::AtArrow), Some('?') => self.consume_and_return(chars, Token::AtQuestion), Some('@') => { @@ -1482,11 +1613,30 @@ impl<'a> Tokenizer<'a> { } } // Postgres uses ? for jsonb operators, not prepared statements - '?' if dialect_of!(self is PostgreSqlDialect) => { - chars.next(); + '?' if self.dialect.supports_geometric_types() => { + chars.next(); // consume match chars.peek() { - Some('|') => self.consume_and_return(chars, Token::QuestionPipe), + Some('|') => { + chars.next(); + match chars.peek() { + Some('|') => self.consume_and_return( + chars, + Token::QuestionMarkDoubleVerticalBar, + ), + _ => Ok(Some(Token::QuestionPipe)), + } + } + Some('&') => self.consume_and_return(chars, Token::QuestionAnd), + Some('-') => { + chars.next(); // consume + match chars.peek() { + Some('|') => self + .consume_and_return(chars, Token::QuestionMarkDashVerticalBar), + _ => Ok(Some(Token::QuestionMarkDash)), + } + } + Some('#') => self.consume_and_return(chars, Token::QuestionMarkSharp), _ => self.consume_and_return(chars, Token::Question), } } @@ -1520,7 +1670,7 @@ impl<'a> Tokenizer<'a> { default: Token, ) -> Result, TokenizerError> { chars.next(); // consume the first char - self.start_binop(chars, prefix, default) + self.start_binop_opt(chars, prefix, Some(default)) } /// parse a custom binary operator @@ -1529,6 +1679,16 @@ impl<'a> Tokenizer<'a> { chars: &mut State, prefix: &str, default: Token, + ) -> Result, TokenizerError> { + self.start_binop_opt(chars, prefix, Some(default)) + } + + /// parse a custom binary operator + fn start_binop_opt( + &self, + chars: &mut State, + prefix: &str, + default: Option, ) -> Result, TokenizerError> { let mut custom = None; while let Some(&ch) = chars.peek() { @@ -1539,10 +1699,14 @@ impl<'a> Tokenizer<'a> { custom.get_or_insert_with(|| prefix.to_string()).push(ch); chars.next(); } - - Ok(Some( - custom.map(Token::CustomBinaryOperator).unwrap_or(default), - )) + match (custom, default) { + (Some(custom), _) => Ok(Token::CustomBinaryOperator(custom).into()), + (None, Some(tok)) => Ok(Some(tok)), + (None, None) => self.tokenizer_error( + chars.location(), + format!("Expected a valid binary operator after '{}'", prefix), + ), + } } /// Tokenize dollar preceded value (i.e: a string/placeholder) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a4e83be06..653142dc6 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -13860,3 +13860,361 @@ fn test_select_from_first() { assert_eq!(ast.to_string(), q); } } + +#[test] +fn test_geometric_unary_operators() { + // Number of points in path or polygon + let sql = "# path '((1,0),(0,1),(-1,0))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::UnaryOp { + op: UnaryOperator::Hash, + .. + } + )); + + // Length or circumference + let sql = "@-@ path '((0,0),(1,0))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::UnaryOp { + op: UnaryOperator::AtDashAt, + .. + } + )); + + // Center + let sql = "@@ circle '((0,0),10)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::UnaryOp { + op: UnaryOperator::DoubleAt, + .. + } + )); + // Is horizontal? + let sql = "?- lseg '((-1,0),(1,0))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::UnaryOp { + op: UnaryOperator::QuestionDash, + .. + } + )); + + // Is vertical? + let sql = "?| lseg '((-1,0),(1,0))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::UnaryOp { + op: UnaryOperator::QuestionPipe, + .. + } + )); +} + +#[test] +fn test_geomtery_type() { + let sql = "point '1,2'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::Point), + value: Value::SingleQuotedString("1,2".to_string()), + } + ); + + let sql = "line '1,2,3,4'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::Line), + value: Value::SingleQuotedString("1,2,3,4".to_string()), + } + ); + + let sql = "path '1,2,3,4'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::GeometricPath), + value: Value::SingleQuotedString("1,2,3,4".to_string()), + } + ); + let sql = "box '1,2,3,4'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::GeometricBox), + value: Value::SingleQuotedString("1,2,3,4".to_string()), + } + ); + + let sql = "circle '1,2,3'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::Circle), + value: Value::SingleQuotedString("1,2,3".to_string()), + } + ); + + let sql = "polygon '1,2,3,4'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::Polygon), + value: Value::SingleQuotedString("1,2,3,4".to_string()), + } + ); + let sql = "lseg '1,2,3,4'"; + assert_eq!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::TypedString { + data_type: DataType::GeometricType(GeometricTypeKind::LineSegment), + value: Value::SingleQuotedString("1,2,3,4".to_string()), + } + ); +} +#[test] +fn test_geometric_binary_operators() { + // Translation plus + let sql = "box '((0,0),(1,1))' + point '(2.0,0)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::Plus, + .. + } + )); + // Translation minus + let sql = "box '((0,0),(1,1))' - point '(2.0,0)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::Minus, + .. + } + )); + + // Scaling multiply + let sql = "box '((0,0),(1,1))' * point '(2.0,0)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::Multiply, + .. + } + )); + + // Scaling divide + let sql = "box '((0,0),(1,1))' / point '(2.0,0)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::Divide, + .. + } + )); + + // Intersection + let sql = "'((1,-1),(-1,1))' # '((1,1),(-1,-1))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::PGBitwiseXor, + .. + } + )); + + //Point of closest proximity + let sql = "point '(0,0)' ## lseg '((2,0),(0,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::DoubleHash, + .. + } + )); + + // Point of closest proximity + let sql = "box '((0,0),(1,1))' && box '((0,0),(2,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::PGOverlap, + .. + } + )); + + // Overlaps to left? + let sql = "box '((0,0),(1,1))' &< box '((0,0),(2,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::AndLt, + .. + } + )); + + // Overlaps to right? + let sql = "box '((0,0),(3,3))' &> box '((0,0),(2,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::AndGt, + .. + } + )); + + // Distance between + let sql = "circle '((0,0),1)' <-> circle '((5,0),1)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::LtDashGt, + .. + } + )); + + // Is left of? + let sql = "circle '((0,0),1)' << circle '((5,0),1)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::PGBitwiseShiftLeft, + .. + } + )); + + // Is right of? + let sql = "circle '((5,0),1)' >> circle '((0,0),1)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::PGBitwiseShiftRight, + .. + } + )); + + // Is below? + let sql = "circle '((0,0),1)' <^ circle '((0,5),1)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::LtCaret, + .. + } + )); + + // Intersects or overlaps + let sql = "lseg '((-1,0),(1,0))' ?# box '((-2,-2),(2,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::QuestionHash, + .. + } + )); + + // Is horizontal? + let sql = "point '(1,0)' ?- point '(0,0)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::QuestionDash, + .. + } + )); + + // Is perpendicular? + let sql = "lseg '((0,0),(0,1))' ?-| lseg '((0,0),(1,0))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::QuestionDashPipe, + .. + } + )); + + // Is vertical? + let sql = "point '(0,1)' ?| point '(0,0)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::QuestionPipe, + .. + } + )); + + // Are parallel? + let sql = "lseg '((-1,0),(1,0))' ?|| lseg '((-1,2),(1,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::QuestionDoublePipe, + .. + } + )); + + // Contained or on? + let sql = "point '(1,1)' @ circle '((0,0),2)'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::At, + .. + } + )); + + // + // Same as? + let sql = "polygon '((0,0),(1,1))' ~= polygon '((1,1),(0,0))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::TildeEq, + .. + } + )); + + // Is strictly below? + let sql = "box '((0,0),(3,3))' <<| box '((3,4),(5,5))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::LtLtPipe, + .. + } + )); + + // Is strictly above? + let sql = "box '((3,4),(5,5))' |>> box '((0,0),(3,3))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::PipeGtGt, + .. + } + )); + + // Does not extend above? + let sql = "box '((0,0),(1,1))' &<| box '((0,0),(2,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::AndLtPipe, + .. + } + )); + + // Does not extend below? + let sql = "box '((0,0),(3,3))' |&> box '((0,0),(2,2))'"; + assert!(matches!( + all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), + Expr::BinaryOp { + op: BinaryOperator::PipeAndGt, + .. + } + )); +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index d400792d9..312ce1186 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2066,12 +2066,8 @@ fn parse_pg_custom_binary_ops() { let operators = [ // PostGIS "&&&", // n-D bounding boxes intersect - "&<", // (is strictly to the left of) - "&>", // (is strictly to the right of) "|=|", // distance between A and B trajectories at their closest point of approach "<<#>>", // n-D distance between A and B bounding boxes - "|>>", // A's bounding box is strictly above B's. - "~=", // bounding box is the same // PGroonga "&@", // Full text search by a keyword "&@~", // Full text search by easy to use query language From 28736da235174efa5e08a072a55f368c42271de7 Mon Sep 17 00:00:00 2001 From: Ian Alexander Joiner <14581281+iajoiner@users.noreply.github.com> Date: Fri, 21 Feb 2025 01:18:51 -0500 Subject: [PATCH 051/286] fix: make `serde` feature no_std (#1730) Co-authored-by: Jay White --- Cargo.toml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Cargo.toml b/Cargo.toml index 06fed2c68..9caff2512 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -49,7 +49,7 @@ bigdecimal = { version = "0.4.1", features = ["serde"], optional = true } log = "0.4" recursive = { version = "0.1.1", optional = true} -serde = { version = "1.0", features = ["derive"], optional = true } +serde = { version = "1.0", default-features = false, features = ["derive", "alloc"], optional = true } # serde_json is only used in examples/cli, but we have to put it outside # of dev-dependencies because of # https://github.com/rust-lang/cargo/issues/1596 From 3ace97c0efd46f8128590f88392ac30d3e99feae Mon Sep 17 00:00:00 2001 From: Artem Osipov <59066880+osipovartem@users.noreply.github.com> Date: Fri, 21 Feb 2025 09:22:24 +0300 Subject: [PATCH 052/286] Implement SnowFlake ALTER SESSION (#1712) --- src/ast/helpers/key_value_options.rs | 89 ++++++++++++ src/ast/helpers/mod.rs | 1 + src/ast/helpers/stmt_data_loading.rs | 65 +-------- src/ast/mod.rs | 49 +++++-- src/ast/spans.rs | 1 + src/dialect/snowflake.rs | 121 ++++++++++++---- src/keywords.rs | 1 + tests/sqlparser_snowflake.rs | 205 ++++++++++++++------------- 8 files changed, 335 insertions(+), 197 deletions(-) create mode 100644 src/ast/helpers/key_value_options.rs diff --git a/src/ast/helpers/key_value_options.rs b/src/ast/helpers/key_value_options.rs new file mode 100644 index 000000000..06f028dd2 --- /dev/null +++ b/src/ast/helpers/key_value_options.rs @@ -0,0 +1,89 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + +//! Key-value options for SQL statements. +//! See [this page](https://docs.snowflake.com/en/sql-reference/commands-data-loading) for more details. + +#[cfg(not(feature = "std"))] +use alloc::string::String; +#[cfg(not(feature = "std"))] +use alloc::vec::Vec; +use core::fmt; +use core::fmt::Formatter; + +#[cfg(feature = "serde")] +use serde::{Deserialize, Serialize}; + +#[cfg(feature = "visitor")] +use sqlparser_derive::{Visit, VisitMut}; + +#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct KeyValueOptions { + pub options: Vec, +} + +#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum KeyValueOptionType { + STRING, + BOOLEAN, + ENUM, + NUMBER, +} + +#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct KeyValueOption { + pub option_name: String, + pub option_type: KeyValueOptionType, + pub value: String, +} + +impl fmt::Display for KeyValueOptions { + fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result { + if !self.options.is_empty() { + let mut first = false; + for option in &self.options { + if !first { + first = true; + } else { + f.write_str(" ")?; + } + write!(f, "{}", option)?; + } + } + Ok(()) + } +} + +impl fmt::Display for KeyValueOption { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self.option_type { + KeyValueOptionType::STRING => { + write!(f, "{}='{}'", self.option_name, self.value)?; + } + KeyValueOptionType::ENUM | KeyValueOptionType::BOOLEAN | KeyValueOptionType::NUMBER => { + write!(f, "{}={}", self.option_name, self.value)?; + } + } + Ok(()) + } +} diff --git a/src/ast/helpers/mod.rs b/src/ast/helpers/mod.rs index a96bffc51..55831220d 100644 --- a/src/ast/helpers/mod.rs +++ b/src/ast/helpers/mod.rs @@ -15,5 +15,6 @@ // specific language governing permissions and limitations // under the License. pub mod attached_token; +pub mod key_value_options; pub mod stmt_create_table; pub mod stmt_data_loading; diff --git a/src/ast/helpers/stmt_data_loading.rs b/src/ast/helpers/stmt_data_loading.rs index 77de5d9ec..cc4fa12fa 100644 --- a/src/ast/helpers/stmt_data_loading.rs +++ b/src/ast/helpers/stmt_data_loading.rs @@ -24,11 +24,11 @@ use alloc::string::String; #[cfg(not(feature = "std"))] use alloc::vec::Vec; use core::fmt; -use core::fmt::Formatter; #[cfg(feature = "serde")] use serde::{Deserialize, Serialize}; +use crate::ast::helpers::key_value_options::KeyValueOptions; use crate::ast::{Ident, ObjectName}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; @@ -38,36 +38,10 @@ use sqlparser_derive::{Visit, VisitMut}; #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct StageParamsObject { pub url: Option, - pub encryption: DataLoadingOptions, + pub encryption: KeyValueOptions, pub endpoint: Option, pub storage_integration: Option, - pub credentials: DataLoadingOptions, -} - -#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct DataLoadingOptions { - pub options: Vec, -} - -#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub enum DataLoadingOptionType { - STRING, - BOOLEAN, - ENUM, - NUMBER, -} - -#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct DataLoadingOption { - pub option_name: String, - pub option_type: DataLoadingOptionType, - pub value: String, + pub credentials: KeyValueOptions, } #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] @@ -106,39 +80,6 @@ impl fmt::Display for StageParamsObject { } } -impl fmt::Display for DataLoadingOptions { - fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result { - if !self.options.is_empty() { - let mut first = false; - for option in &self.options { - if !first { - first = true; - } else { - f.write_str(" ")?; - } - write!(f, "{}", option)?; - } - } - Ok(()) - } -} - -impl fmt::Display for DataLoadingOption { - fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - match self.option_type { - DataLoadingOptionType::STRING => { - write!(f, "{}='{}'", self.option_name, self.value)?; - } - DataLoadingOptionType::ENUM - | DataLoadingOptionType::BOOLEAN - | DataLoadingOptionType::NUMBER => { - write!(f, "{}={}", self.option_name, self.value)?; - } - } - Ok(()) - } -} - impl fmt::Display for StageLoadSelectItem { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { if self.alias.is_some() { diff --git a/src/ast/mod.rs b/src/ast/mod.rs index dc50871ce..2b9016d9a 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -89,9 +89,8 @@ pub use self::value::{ NormalizationForm, TrimWhereField, Value, }; -use crate::ast::helpers::stmt_data_loading::{ - DataLoadingOptions, StageLoadSelectItem, StageParamsObject, -}; +use crate::ast::helpers::key_value_options::KeyValueOptions; +use crate::ast::helpers::stmt_data_loading::{StageLoadSelectItem, StageParamsObject}; #[cfg(feature = "visitor")] pub use visitor::*; @@ -2504,8 +2503,8 @@ pub enum Statement { from_query: Option>, files: Option>, pattern: Option, - file_format: DataLoadingOptions, - copy_options: DataLoadingOptions, + file_format: KeyValueOptions, + copy_options: KeyValueOptions, validation_mode: Option, partition: Option>, }, @@ -2713,6 +2712,17 @@ pub enum Statement { owner: Option, }, /// ```sql + /// ALTER SESSION SET sessionParam + /// ALTER SESSION UNSET [ , , ... ] + /// ``` + /// See + AlterSession { + /// true is to set for the session parameters, false is to unset + set: bool, + /// The session parameters to set or unset + session_params: KeyValueOptions, + }, + /// ```sql /// ATTACH DATABASE 'path/to/file' AS alias /// ``` /// (SQLite-specific) @@ -3240,9 +3250,9 @@ pub enum Statement { if_not_exists: bool, name: ObjectName, stage_params: StageParamsObject, - directory_table_params: DataLoadingOptions, - file_format: DataLoadingOptions, - copy_options: DataLoadingOptions, + directory_table_params: KeyValueOptions, + file_format: KeyValueOptions, + copy_options: KeyValueOptions, comment: Option, }, /// ```sql @@ -4467,6 +4477,29 @@ impl fmt::Display for Statement { } Ok(()) } + Statement::AlterSession { + set, + session_params, + } => { + write!( + f, + "ALTER SESSION {set}", + set = if *set { "SET" } else { "UNSET" } + )?; + if !session_params.options.is_empty() { + if *set { + write!(f, " {}", session_params)?; + } else { + let options = session_params + .options + .iter() + .map(|p| p.option_name.clone()) + .collect::>(); + write!(f, " {}", display_separated(&options, ", "))?; + } + } + Ok(()) + } Statement::Drop { object_type, if_exists, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index de39e50d6..3de41902d 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -430,6 +430,7 @@ impl Spanned for Statement { // These statements need to be implemented Statement::AlterType { .. } => Span::empty(), Statement::AlterRole { .. } => Span::empty(), + Statement::AlterSession { .. } => Span::empty(), Statement::AttachDatabase { .. } => Span::empty(), Statement::AttachDuckDBDatabase { .. } => Span::empty(), Statement::DetachDuckDBDatabase { .. } => Span::empty(), diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 6702b94b2..72252b277 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -17,10 +17,10 @@ #[cfg(not(feature = "std"))] use crate::alloc::string::ToString; +use crate::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionType, KeyValueOptions}; use crate::ast::helpers::stmt_create_table::CreateTableBuilder; use crate::ast::helpers::stmt_data_loading::{ - DataLoadingOption, DataLoadingOptionType, DataLoadingOptions, FileStagingCommand, - StageLoadSelectItem, StageParamsObject, + FileStagingCommand, StageLoadSelectItem, StageParamsObject, }; use crate::ast::{ ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, @@ -130,6 +130,16 @@ impl Dialect for SnowflakeDialect { } fn parse_statement(&self, parser: &mut Parser) -> Option> { + if parser.parse_keywords(&[Keyword::ALTER, Keyword::SESSION]) { + // ALTER SESSION + let set = match parser.parse_one_of_keywords(&[Keyword::SET, Keyword::UNSET]) { + Some(Keyword::SET) => true, + Some(Keyword::UNSET) => false, + _ => return Some(parser.expected("SET or UNSET", parser.peek_token())), + }; + return Some(parse_alter_session(parser, set)); + } + if parser.parse_keyword(Keyword::CREATE) { // possibly CREATE STAGE //[ OR REPLACE ] @@ -358,6 +368,18 @@ fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result +fn parse_alter_session(parser: &mut Parser, set: bool) -> Result { + let session_options = parse_session_options(parser, set)?; + Ok(Statement::AlterSession { + set, + session_params: KeyValueOptions { + options: session_options, + }, + }) +} + /// Parse snowflake create table statement. /// /// @@ -634,13 +656,13 @@ pub fn parse_create_stage( if_not_exists, name, stage_params, - directory_table_params: DataLoadingOptions { + directory_table_params: KeyValueOptions { options: directory_table_params, }, - file_format: DataLoadingOptions { + file_format: KeyValueOptions { options: file_format, }, - copy_options: DataLoadingOptions { + copy_options: KeyValueOptions { options: copy_options, }, comment, @@ -708,10 +730,10 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { let mut from_stage = None; let mut stage_params = StageParamsObject { url: None, - encryption: DataLoadingOptions { options: vec![] }, + encryption: KeyValueOptions { options: vec![] }, endpoint: None, storage_integration: None, - credentials: DataLoadingOptions { options: vec![] }, + credentials: KeyValueOptions { options: vec![] }, }; let mut from_query = None; let mut partition = None; @@ -818,7 +840,7 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { Token::Comma => continue, // In `COPY INTO ` the copy options do not have a shared key // like in `COPY INTO
` - Token::Word(key) => copy_options.push(parse_copy_option(parser, key)?), + Token::Word(key) => copy_options.push(parse_option(parser, key)?), _ => return parser.expected("another copy option, ; or EOF'", parser.peek_token()), } } @@ -834,10 +856,10 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { from_query, files: if files.is_empty() { None } else { Some(files) }, pattern, - file_format: DataLoadingOptions { + file_format: KeyValueOptions { options: file_format, }, - copy_options: DataLoadingOptions { + copy_options: KeyValueOptions { options: copy_options, }, validation_mode, @@ -931,8 +953,8 @@ fn parse_select_items_for_data_load( fn parse_stage_params(parser: &mut Parser) -> Result { let (mut url, mut storage_integration, mut endpoint) = (None, None, None); - let mut encryption: DataLoadingOptions = DataLoadingOptions { options: vec![] }; - let mut credentials: DataLoadingOptions = DataLoadingOptions { options: vec![] }; + let mut encryption: KeyValueOptions = KeyValueOptions { options: vec![] }; + let mut credentials: KeyValueOptions = KeyValueOptions { options: vec![] }; // URL if parser.parse_keyword(Keyword::URL) { @@ -961,7 +983,7 @@ fn parse_stage_params(parser: &mut Parser) -> Result Result Result Result, ParserError> { + let mut options: Vec = Vec::new(); + let empty = String::new; + loop { + match parser.next_token().token { + Token::Comma => continue, + Token::Word(key) => { + if set { + let option = parse_option(parser, key)?; + options.push(option); + } else { + options.push(KeyValueOption { + option_name: key.value, + option_type: KeyValueOptionType::STRING, + value: empty(), + }); + } + } + _ => { + if parser.peek_token().token == Token::EOF { + break; + } + return parser.expected("another option", parser.peek_token()); + } + } + } + options + .is_empty() + .then(|| { + Err(ParserError::ParserError( + "expected at least one option".to_string(), + )) + }) + .unwrap_or(Ok(options)) +} + /// Parses options provided within parentheses like: /// ( ENABLE = { TRUE | FALSE } /// [ AUTO_REFRESH = { TRUE | FALSE } ] /// [ REFRESH_ON_CREATE = { TRUE | FALSE } ] /// [ NOTIFICATION_INTEGRATION = '' ] ) /// -fn parse_parentheses_options(parser: &mut Parser) -> Result, ParserError> { - let mut options: Vec = Vec::new(); +fn parse_parentheses_options(parser: &mut Parser) -> Result, ParserError> { + let mut options: Vec = Vec::new(); parser.expect_token(&Token::LParen)?; loop { match parser.next_token().token { Token::RParen => break, Token::Comma => continue, - Token::Word(key) => options.push(parse_copy_option(parser, key)?), + Token::Word(key) => options.push(parse_option(parser, key)?), _ => return parser.expected("another option or ')'", parser.peek_token()), }; } @@ -1004,35 +1069,35 @@ fn parse_parentheses_options(parser: &mut Parser) -> Result Result { +fn parse_option(parser: &mut Parser, key: Word) -> Result { parser.expect_token(&Token::Eq)?; if parser.parse_keyword(Keyword::TRUE) { - Ok(DataLoadingOption { + Ok(KeyValueOption { option_name: key.value, - option_type: DataLoadingOptionType::BOOLEAN, + option_type: KeyValueOptionType::BOOLEAN, value: "TRUE".to_string(), }) } else if parser.parse_keyword(Keyword::FALSE) { - Ok(DataLoadingOption { + Ok(KeyValueOption { option_name: key.value, - option_type: DataLoadingOptionType::BOOLEAN, + option_type: KeyValueOptionType::BOOLEAN, value: "FALSE".to_string(), }) } else { match parser.next_token().token { - Token::SingleQuotedString(value) => Ok(DataLoadingOption { + Token::SingleQuotedString(value) => Ok(KeyValueOption { option_name: key.value, - option_type: DataLoadingOptionType::STRING, + option_type: KeyValueOptionType::STRING, value, }), - Token::Word(word) => Ok(DataLoadingOption { + Token::Word(word) => Ok(KeyValueOption { option_name: key.value, - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: word.value, }), - Token::Number(n, _) => Ok(DataLoadingOption { + Token::Number(n, _) => Ok(KeyValueOption { option_name: key.value, - option_type: DataLoadingOptionType::NUMBER, + option_type: KeyValueOptionType::NUMBER, value: n, }), _ => parser.expected("expected option value", parser.peek_token()), diff --git a/src/keywords.rs b/src/keywords.rs index 46c40fbaa..d62a038b8 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -913,6 +913,7 @@ define_keywords!( UNNEST, UNPIVOT, UNSAFE, + UNSET, UNSIGNED, UNTIL, UPDATE, diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 76550b8e3..12796bb65 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -19,9 +19,8 @@ //! Test SQL syntax specific to Snowflake. The parser based on the //! generic dialect is also tested (on the inputs it can handle). -use sqlparser::ast::helpers::stmt_data_loading::{ - DataLoadingOption, DataLoadingOptionType, StageLoadSelectItem, -}; +use sqlparser::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionType}; +use sqlparser::ast::helpers::stmt_data_loading::StageLoadSelectItem; use sqlparser::ast::*; use sqlparser::dialect::{Dialect, GenericDialect, SnowflakeDialect}; use sqlparser::parser::{ParserError, ParserOptions}; @@ -1914,38 +1913,26 @@ fn test_create_stage_with_stage_params() { "", stage_params.endpoint.unwrap() ); - assert!(stage_params - .credentials - .options - .contains(&DataLoadingOption { - option_name: "AWS_KEY_ID".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "1a2b3c".to_string() - })); - assert!(stage_params - .credentials - .options - .contains(&DataLoadingOption { - option_name: "AWS_SECRET_KEY".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "4x5y6z".to_string() - })); - assert!(stage_params - .encryption - .options - .contains(&DataLoadingOption { - option_name: "MASTER_KEY".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "key".to_string() - })); - assert!(stage_params - .encryption - .options - .contains(&DataLoadingOption { - option_name: "TYPE".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "AWS_SSE_KMS".to_string() - })); + assert!(stage_params.credentials.options.contains(&KeyValueOption { + option_name: "AWS_KEY_ID".to_string(), + option_type: KeyValueOptionType::STRING, + value: "1a2b3c".to_string() + })); + assert!(stage_params.credentials.options.contains(&KeyValueOption { + option_name: "AWS_SECRET_KEY".to_string(), + option_type: KeyValueOptionType::STRING, + value: "4x5y6z".to_string() + })); + assert!(stage_params.encryption.options.contains(&KeyValueOption { + option_name: "MASTER_KEY".to_string(), + option_type: KeyValueOptionType::STRING, + value: "key".to_string() + })); + assert!(stage_params.encryption.options.contains(&KeyValueOption { + option_name: "TYPE".to_string(), + option_type: KeyValueOptionType::STRING, + value: "AWS_SSE_KMS".to_string() + })); } _ => unreachable!(), }; @@ -1966,19 +1953,19 @@ fn test_create_stage_with_directory_table_params() { directory_table_params, .. } => { - assert!(directory_table_params.options.contains(&DataLoadingOption { + assert!(directory_table_params.options.contains(&KeyValueOption { option_name: "ENABLE".to_string(), - option_type: DataLoadingOptionType::BOOLEAN, + option_type: KeyValueOptionType::BOOLEAN, value: "TRUE".to_string() })); - assert!(directory_table_params.options.contains(&DataLoadingOption { + assert!(directory_table_params.options.contains(&KeyValueOption { option_name: "REFRESH_ON_CREATE".to_string(), - option_type: DataLoadingOptionType::BOOLEAN, + option_type: KeyValueOptionType::BOOLEAN, value: "FALSE".to_string() })); - assert!(directory_table_params.options.contains(&DataLoadingOption { + assert!(directory_table_params.options.contains(&KeyValueOption { option_name: "NOTIFICATION_INTEGRATION".to_string(), - option_type: DataLoadingOptionType::STRING, + option_type: KeyValueOptionType::STRING, value: "some-string".to_string() })); } @@ -1997,19 +1984,19 @@ fn test_create_stage_with_file_format() { match snowflake_without_unescape().verified_stmt(sql) { Statement::CreateStage { file_format, .. } => { - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "COMPRESSION".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "AUTO".to_string() })); - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "BINARY_FORMAT".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "HEX".to_string() })); - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "ESCAPE".to_string(), - option_type: DataLoadingOptionType::STRING, + option_type: KeyValueOptionType::STRING, value: r#"\\"#.to_string() })); } @@ -2030,14 +2017,14 @@ fn test_create_stage_with_copy_options() { ); match snowflake().verified_stmt(sql) { Statement::CreateStage { copy_options, .. } => { - assert!(copy_options.options.contains(&DataLoadingOption { + assert!(copy_options.options.contains(&KeyValueOption { option_name: "ON_ERROR".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "CONTINUE".to_string() })); - assert!(copy_options.options.contains(&DataLoadingOption { + assert!(copy_options.options.contains(&KeyValueOption { option_name: "FORCE".to_string(), - option_type: DataLoadingOptionType::BOOLEAN, + option_type: KeyValueOptionType::BOOLEAN, value: "TRUE".to_string() })); } @@ -2167,38 +2154,26 @@ fn test_copy_into_with_stage_params() { "", stage_params.endpoint.unwrap() ); - assert!(stage_params - .credentials - .options - .contains(&DataLoadingOption { - option_name: "AWS_KEY_ID".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "1a2b3c".to_string() - })); - assert!(stage_params - .credentials - .options - .contains(&DataLoadingOption { - option_name: "AWS_SECRET_KEY".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "4x5y6z".to_string() - })); - assert!(stage_params - .encryption - .options - .contains(&DataLoadingOption { - option_name: "MASTER_KEY".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "key".to_string() - })); - assert!(stage_params - .encryption - .options - .contains(&DataLoadingOption { - option_name: "TYPE".to_string(), - option_type: DataLoadingOptionType::STRING, - value: "AWS_SSE_KMS".to_string() - })); + assert!(stage_params.credentials.options.contains(&KeyValueOption { + option_name: "AWS_KEY_ID".to_string(), + option_type: KeyValueOptionType::STRING, + value: "1a2b3c".to_string() + })); + assert!(stage_params.credentials.options.contains(&KeyValueOption { + option_name: "AWS_SECRET_KEY".to_string(), + option_type: KeyValueOptionType::STRING, + value: "4x5y6z".to_string() + })); + assert!(stage_params.encryption.options.contains(&KeyValueOption { + option_name: "MASTER_KEY".to_string(), + option_type: KeyValueOptionType::STRING, + value: "key".to_string() + })); + assert!(stage_params.encryption.options.contains(&KeyValueOption { + option_name: "TYPE".to_string(), + option_type: KeyValueOptionType::STRING, + value: "AWS_SSE_KMS".to_string() + })); } _ => unreachable!(), }; @@ -2326,19 +2301,19 @@ fn test_copy_into_file_format() { match snowflake_without_unescape().verified_stmt(sql) { Statement::CopyIntoSnowflake { file_format, .. } => { - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "COMPRESSION".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "AUTO".to_string() })); - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "BINARY_FORMAT".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "HEX".to_string() })); - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "ESCAPE".to_string(), - option_type: DataLoadingOptionType::STRING, + option_type: KeyValueOptionType::STRING, value: r#"\\"#.to_string() })); } @@ -2365,19 +2340,19 @@ fn test_copy_into_file_format() { .unwrap() { Statement::CopyIntoSnowflake { file_format, .. } => { - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "COMPRESSION".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "AUTO".to_string() })); - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "BINARY_FORMAT".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "HEX".to_string() })); - assert!(file_format.options.contains(&DataLoadingOption { + assert!(file_format.options.contains(&KeyValueOption { option_name: "ESCAPE".to_string(), - option_type: DataLoadingOptionType::STRING, + option_type: KeyValueOptionType::STRING, value: r#"\\"#.to_string() })); } @@ -2397,14 +2372,14 @@ fn test_copy_into_copy_options() { match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { copy_options, .. } => { - assert!(copy_options.options.contains(&DataLoadingOption { + assert!(copy_options.options.contains(&KeyValueOption { option_name: "ON_ERROR".to_string(), - option_type: DataLoadingOptionType::ENUM, + option_type: KeyValueOptionType::ENUM, value: "CONTINUE".to_string() })); - assert!(copy_options.options.contains(&DataLoadingOption { + assert!(copy_options.options.contains(&KeyValueOption { option_name: "FORCE".to_string(), - option_type: DataLoadingOptionType::BOOLEAN, + option_type: KeyValueOptionType::BOOLEAN, value: "TRUE".to_string() })); } @@ -3475,3 +3450,35 @@ fn test_grant_database_role_to() { snowflake_and_generic().verified_stmt("GRANT DATABASE ROLE r1 TO ROLE r2"); snowflake_and_generic().verified_stmt("GRANT DATABASE ROLE db1.sc1.r1 TO ROLE db1.sc1.r2"); } + +#[test] +fn test_alter_session() { + assert_eq!( + snowflake() + .parse_sql_statements("ALTER SESSION SET") + .unwrap_err() + .to_string(), + "sql parser error: expected at least one option" + ); + assert_eq!( + snowflake() + .parse_sql_statements("ALTER SESSION UNSET") + .unwrap_err() + .to_string(), + "sql parser error: expected at least one option" + ); + + snowflake().verified_stmt("ALTER SESSION SET AUTOCOMMIT=TRUE"); + snowflake().verified_stmt("ALTER SESSION SET AUTOCOMMIT=FALSE QUERY_TAG='tag'"); + snowflake().verified_stmt("ALTER SESSION UNSET AUTOCOMMIT"); + snowflake().verified_stmt("ALTER SESSION UNSET AUTOCOMMIT, QUERY_TAG"); + snowflake().one_statement_parses_to( + "ALTER SESSION SET A=false, B='tag';", + "ALTER SESSION SET A=FALSE B='tag'", + ); + snowflake().one_statement_parses_to( + "ALTER SESSION SET A=true \nB='tag'", + "ALTER SESSION SET A=TRUE B='tag'", + ); + snowflake().one_statement_parses_to("ALTER SESSION UNSET a\nB", "ALTER SESSION UNSET a, B"); +} From 8fc8082e9a0c1c418b3510fae0cedd1e6e241785 Mon Sep 17 00:00:00 2001 From: tomershaniii <65544633+tomershaniii@users.noreply.github.com> Date: Sat, 22 Feb 2025 07:48:39 +0200 Subject: [PATCH 053/286] Extend Visitor trait for Value type (#1725) --- src/ast/value.rs | 7 +++- src/ast/visitor.rs | 99 ++++++++++++++++++++++++++++++++++++++++++---- 2 files changed, 98 insertions(+), 8 deletions(-) diff --git a/src/ast/value.rs b/src/ast/value.rs index 5798b5404..f2f02754b 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -33,7 +33,12 @@ use sqlparser_derive::{Visit, VisitMut}; /// Primitive SQL values such as number and string #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +#[cfg_attr( + feature = "visitor", + derive(Visit, VisitMut), + visit(with = "visit_value") +)] + pub enum Value { /// Numeric literal #[cfg(not(feature = "bigdecimal"))] diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index 457dbbaed..bb6246498 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -17,7 +17,7 @@ //! Recursive visitors for ast Nodes. See [`Visitor`] for more details. -use crate::ast::{Expr, ObjectName, Query, Statement, TableFactor}; +use crate::ast::{Expr, ObjectName, Query, Statement, TableFactor, Value}; use core::ops::ControlFlow; /// A type that can be visited by a [`Visitor`]. See [`Visitor`] for @@ -233,6 +233,16 @@ pub trait Visitor { fn post_visit_statement(&mut self, _statement: &Statement) -> ControlFlow { ControlFlow::Continue(()) } + + /// Invoked for any Value that appear in the AST before visiting children + fn pre_visit_value(&mut self, _value: &Value) -> ControlFlow { + ControlFlow::Continue(()) + } + + /// Invoked for any Value that appear in the AST after visiting children + fn post_visit_value(&mut self, _value: &Value) -> ControlFlow { + ControlFlow::Continue(()) + } } /// A visitor that can be used to mutate an AST tree. @@ -337,6 +347,16 @@ pub trait VisitorMut { fn post_visit_statement(&mut self, _statement: &mut Statement) -> ControlFlow { ControlFlow::Continue(()) } + + /// Invoked for any value that appear in the AST before visiting children + fn pre_visit_value(&mut self, _value: &mut Value) -> ControlFlow { + ControlFlow::Continue(()) + } + + /// Invoked for any statements that appear in the AST after visiting children + fn post_visit_value(&mut self, _value: &mut Value) -> ControlFlow { + ControlFlow::Continue(()) + } } struct RelationVisitor(F); @@ -647,6 +667,7 @@ where #[cfg(test)] mod tests { use super::*; + use crate::ast::Statement; use crate::dialect::GenericDialect; use crate::parser::Parser; use crate::tokenizer::Tokenizer; @@ -720,7 +741,7 @@ mod tests { } } - fn do_visit(sql: &str) -> Vec { + fn do_visit(sql: &str, visitor: &mut V) -> Statement { let dialect = GenericDialect {}; let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); let s = Parser::new(&dialect) @@ -728,9 +749,8 @@ mod tests { .parse_statement() .unwrap(); - let mut visitor = TestVisitor::default(); - s.visit(&mut visitor); - visitor.visited + s.visit(visitor); + s } #[test] @@ -889,8 +909,9 @@ mod tests { ), ]; for (sql, expected) in tests { - let actual = do_visit(sql); - let actual: Vec<_> = actual.iter().map(|x| x.as_str()).collect(); + let mut visitor = TestVisitor::default(); + let _ = do_visit(sql, &mut visitor); + let actual: Vec<_> = visitor.visited.iter().map(|x| x.as_str()).collect(); assert_eq!(actual, expected) } } @@ -920,3 +941,67 @@ mod tests { s.visit(&mut visitor); } } + +#[cfg(test)] +mod visit_mut_tests { + use crate::ast::{Statement, Value, VisitMut, VisitorMut}; + use crate::dialect::GenericDialect; + use crate::parser::Parser; + use crate::tokenizer::Tokenizer; + use core::ops::ControlFlow; + + #[derive(Default)] + struct MutatorVisitor { + index: u64, + } + + impl VisitorMut for MutatorVisitor { + type Break = (); + + fn pre_visit_value(&mut self, value: &mut Value) -> ControlFlow { + self.index += 1; + *value = Value::SingleQuotedString(format!("REDACTED_{}", self.index)); + ControlFlow::Continue(()) + } + + fn post_visit_value(&mut self, _value: &mut Value) -> ControlFlow { + ControlFlow::Continue(()) + } + } + + fn do_visit_mut(sql: &str, visitor: &mut V) -> Statement { + let dialect = GenericDialect {}; + let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); + let mut s = Parser::new(&dialect) + .with_tokens(tokens) + .parse_statement() + .unwrap(); + + s.visit(visitor); + s + } + + #[test] + fn test_value_redact() { + let tests = vec![ + ( + concat!( + "SELECT * FROM monthly_sales ", + "PIVOT(SUM(a.amount) FOR a.MONTH IN ('JAN', 'FEB', 'MAR', 'APR')) AS p (c, d) ", + "ORDER BY EMPID" + ), + concat!( + "SELECT * FROM monthly_sales ", + "PIVOT(SUM(a.amount) FOR a.MONTH IN ('REDACTED_1', 'REDACTED_2', 'REDACTED_3', 'REDACTED_4')) AS p (c, d) ", + "ORDER BY EMPID" + ), + ), + ]; + + for (sql, expected) in tests { + let mut visitor = MutatorVisitor::default(); + let mutated = do_visit_mut(sql, &mut visitor); + assert_eq!(mutated.to_string(), expected) + } + } +} From 1f1c0693daa031961a05f9a91cb154c2b3db5572 Mon Sep 17 00:00:00 2001 From: SiLe Zhou Date: Sat, 22 Feb 2025 13:52:07 +0800 Subject: [PATCH 054/286] Add support for `ORDER BY ALL` (#1724) --- src/ast/mod.rs | 10 +- src/ast/query.rs | 73 +++++++--- src/ast/spans.rs | 26 ++-- src/dialect/clickhouse.rs | 5 + src/dialect/duckdb.rs | 5 + src/dialect/mod.rs | 8 ++ src/parser/mod.rs | 58 +++++--- tests/sqlparser_clickhouse.rs | 60 +++++--- tests/sqlparser_common.rs | 258 ++++++++++++++++++++++++++++------ tests/sqlparser_hive.rs | 16 ++- tests/sqlparser_mysql.rs | 6 +- 11 files changed, 396 insertions(+), 129 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 2b9016d9a..ace752a86 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -68,11 +68,11 @@ pub use self::query::{ JsonTableColumn, JsonTableColumnErrorHandling, JsonTableNamedColumn, JsonTableNestedColumn, LateralView, LockClause, LockType, MatchRecognizePattern, MatchRecognizeSymbol, Measure, NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, OffsetRows, OpenJsonTableColumn, - OrderBy, OrderByExpr, PivotValueSource, ProjectionSelect, Query, RenameSelectItem, - RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, Select, - SelectFlavor, SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, SetOperator, - SetQuantifier, Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, - TableFunctionArgs, TableIndexHintForClause, TableIndexHintType, TableIndexHints, + OrderBy, OrderByExpr, OrderByKind, OrderByOptions, PivotValueSource, ProjectionSelect, Query, + RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, + Select, SelectFlavor, SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, + SetOperator, SetQuantifier, Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, + TableFactor, TableFunctionArgs, TableIndexHintForClause, TableIndexHintType, TableIndexHints, TableIndexType, TableSample, TableSampleBucket, TableSampleKind, TableSampleMethod, TableSampleModifier, TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, UpdateTableFromKind, diff --git a/src/ast/query.rs b/src/ast/query.rs index c52a98b63..bed991114 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2218,30 +2218,50 @@ pub enum JoinConstraint { None, } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum OrderByKind { + /// ALL syntax of [DuckDB] and [ClickHouse]. + /// + /// [DuckDB]: + /// [ClickHouse]: + All(OrderByOptions), + + /// Expressions + Expressions(Vec), +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct OrderBy { - pub exprs: Vec, + pub kind: OrderByKind, + /// Optional: `INTERPOLATE` /// Supported by [ClickHouse syntax] - /// - /// [ClickHouse syntax]: pub interpolate: Option, } impl fmt::Display for OrderBy { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "ORDER BY")?; - if !self.exprs.is_empty() { - write!(f, " {}", display_comma_separated(&self.exprs))?; + match &self.kind { + OrderByKind::Expressions(exprs) => { + write!(f, " {}", display_comma_separated(exprs))?; + } + OrderByKind::All(all) => { + write!(f, " ALL{}", all)?; + } } + if let Some(ref interpolate) = self.interpolate { match &interpolate.exprs { Some(exprs) => write!(f, " INTERPOLATE ({})", display_comma_separated(exprs))?, None => write!(f, " INTERPOLATE")?, } } + Ok(()) } } @@ -2252,10 +2272,7 @@ impl fmt::Display for OrderBy { #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct OrderByExpr { pub expr: Expr, - /// Optional `ASC` or `DESC` - pub asc: Option, - /// Optional `NULLS FIRST` or `NULLS LAST` - pub nulls_first: Option, + pub options: OrderByOptions, /// Optional: `WITH FILL` /// Supported by [ClickHouse syntax]: pub with_fill: Option, @@ -2263,17 +2280,7 @@ pub struct OrderByExpr { impl fmt::Display for OrderByExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{}", self.expr)?; - match self.asc { - Some(true) => write!(f, " ASC")?, - Some(false) => write!(f, " DESC")?, - None => (), - } - match self.nulls_first { - Some(true) => write!(f, " NULLS FIRST")?, - Some(false) => write!(f, " NULLS LAST")?, - None => (), - } + write!(f, "{}{}", self.expr, self.options)?; if let Some(ref with_fill) = self.with_fill { write!(f, " {}", with_fill)? } @@ -2339,6 +2346,32 @@ impl fmt::Display for InterpolateExpr { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct OrderByOptions { + /// Optional `ASC` or `DESC` + pub asc: Option, + /// Optional `NULLS FIRST` or `NULLS LAST` + pub nulls_first: Option, +} + +impl fmt::Display for OrderByOptions { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self.asc { + Some(true) => write!(f, " ASC")?, + Some(false) => write!(f, " DESC")?, + None => (), + } + match self.nulls_first { + Some(true) => write!(f, " NULLS FIRST")?, + Some(false) => write!(f, " NULLS LAST")?, + None => (), + } + Ok(()) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 3de41902d..88c3d8aea 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -30,7 +30,7 @@ use super::{ GroupByExpr, HavingBound, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, - OnConflictAction, OnInsert, OrderBy, OrderByExpr, Partition, PivotValueSource, + OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, @@ -1095,16 +1095,21 @@ impl Spanned for ProjectionSelect { } } +/// # partial span +/// +/// Missing spans: +/// - [OrderByKind::All] impl Spanned for OrderBy { fn span(&self) -> Span { - let OrderBy { exprs, interpolate } = self; - - union_spans( - exprs - .iter() - .map(|i| i.span()) - .chain(interpolate.iter().map(|i| i.span())), - ) + match &self.kind { + OrderByKind::All(_) => Span::empty(), + OrderByKind::Expressions(exprs) => union_spans( + exprs + .iter() + .map(|i| i.span()) + .chain(self.interpolate.iter().map(|i| i.span())), + ), + } } } @@ -1902,8 +1907,7 @@ impl Spanned for OrderByExpr { fn span(&self) -> Span { let OrderByExpr { expr, - asc: _, // bool - nulls_first: _, // bool + options: _, with_fill, } = self; diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs index 37da2ab5d..f5e70c309 100644 --- a/src/dialect/clickhouse.rs +++ b/src/dialect/clickhouse.rs @@ -80,6 +80,11 @@ impl Dialect for ClickHouseDialect { true } + /// See + fn supports_order_by_all(&self) -> bool { + true + } + // See fn supports_group_by_expr(&self) -> bool { true diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index a2e7fb6c0..3595aa713 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -89,4 +89,9 @@ impl Dialect for DuckDbDialect { fn supports_from_first_select(&self) -> bool { true } + + /// See DuckDB + fn supports_order_by_all(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 86e23c86f..b68adc170 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -943,6 +943,14 @@ pub trait Dialect: Debug + Any { fn supports_geometric_types(&self) -> bool { false } + + /// Returns true if the dialect supports `ORDER BY ALL`. + /// `ALL` which means all columns of the SELECT clause. + /// + /// For example: ```SELECT * FROM addresses ORDER BY ALL;```. + fn supports_order_by_all(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 69268bc51..f80754a3a 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9369,17 +9369,26 @@ impl<'a> Parser<'a> { pub fn parse_optional_order_by(&mut self) -> Result, ParserError> { if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) { - let order_by_exprs = self.parse_comma_separated(Parser::parse_order_by_expr)?; - let interpolate = if dialect_of!(self is ClickHouseDialect | GenericDialect) { - self.parse_interpolations()? - } else { - None - }; - - Ok(Some(OrderBy { - exprs: order_by_exprs, - interpolate, - })) + let order_by = + if self.dialect.supports_order_by_all() && self.parse_keyword(Keyword::ALL) { + let order_by_options = self.parse_order_by_options()?; + OrderBy { + kind: OrderByKind::All(order_by_options), + interpolate: None, + } + } else { + let exprs = self.parse_comma_separated(Parser::parse_order_by_expr)?; + let interpolate = if dialect_of!(self is ClickHouseDialect | GenericDialect) { + self.parse_interpolations()? + } else { + None + }; + OrderBy { + kind: OrderByKind::Expressions(exprs), + interpolate, + } + }; + Ok(Some(order_by)) } else { Ok(None) } @@ -13557,15 +13566,7 @@ impl<'a> Parser<'a> { pub fn parse_order_by_expr(&mut self) -> Result { let expr = self.parse_expr()?; - let asc = self.parse_asc_desc(); - - let nulls_first = if self.parse_keywords(&[Keyword::NULLS, Keyword::FIRST]) { - Some(true) - } else if self.parse_keywords(&[Keyword::NULLS, Keyword::LAST]) { - Some(false) - } else { - None - }; + let options = self.parse_order_by_options()?; let with_fill = if dialect_of!(self is ClickHouseDialect | GenericDialect) && self.parse_keywords(&[Keyword::WITH, Keyword::FILL]) @@ -13577,12 +13578,25 @@ impl<'a> Parser<'a> { Ok(OrderByExpr { expr, - asc, - nulls_first, + options, with_fill, }) } + fn parse_order_by_options(&mut self) -> Result { + let asc = self.parse_asc_desc(); + + let nulls_first = if self.parse_keywords(&[Keyword::NULLS, Keyword::FIRST]) { + Some(true) + } else if self.parse_keywords(&[Keyword::NULLS, Keyword::LAST]) { + Some(false) + } else { + None + }; + + Ok(OrderByOptions { asc, nulls_first }) + } + // Parse a WITH FILL clause (ClickHouse dialect) // that follow the WITH FILL keywords in a ORDER BY clause pub fn parse_with_fill(&mut self) -> Result { diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 34f684c69..99e76d45e 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -323,12 +323,14 @@ fn parse_alter_table_add_projection() { vec![] )), order_by: Some(OrderBy { - exprs: vec![OrderByExpr { + kind: OrderByKind::Expressions(vec![OrderByExpr { expr: Identifier(Ident::new("b")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, - }], + }]), interpolate: None, }), } @@ -1068,11 +1070,13 @@ fn parse_select_order_by_with_fill_interpolate() { let select = clickhouse().verified_query(sql); assert_eq!( OrderBy { - exprs: vec![ + kind: OrderByKind::Expressions(vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("fname")), - asc: Some(true), - nulls_first: Some(true), + options: OrderByOptions { + asc: Some(true), + nulls_first: Some(true), + }, with_fill: Some(WithFill { from: Some(Expr::Value(number("10"))), to: Some(Expr::Value(number("20"))), @@ -1081,15 +1085,17 @@ fn parse_select_order_by_with_fill_interpolate() { }, OrderByExpr { expr: Expr::Identifier(Ident::new("lname")), - asc: Some(false), - nulls_first: Some(false), + options: OrderByOptions { + asc: Some(false), + nulls_first: Some(false), + }, with_fill: Some(WithFill { from: Some(Expr::Value(number("30"))), to: Some(Expr::Value(number("40"))), step: Some(Expr::Value(number("3"))), }), }, - ], + ]), interpolate: Some(Interpolate { exprs: Some(vec![InterpolateExpr { column: Ident::new("col1"), @@ -1147,8 +1153,12 @@ fn parse_with_fill() { from: Some(Expr::Value(number("10"))), to: Some(Expr::Value(number("20"))), step: Some(Expr::Value(number("2"))), - }), - select.order_by.expect("ORDER BY expected").exprs[0].with_fill + }) + .as_ref(), + match select.order_by.expect("ORDER BY expected").kind { + OrderByKind::Expressions(ref exprs) => exprs[0].with_fill.as_ref(), + _ => None, + } ); } @@ -1199,8 +1209,13 @@ fn parse_interpolate_body_with_columns() { }), }, ]) - }), - select.order_by.expect("ORDER BY expected").interpolate + }) + .as_ref(), + select + .order_by + .expect("ORDER BY expected") + .interpolate + .as_ref() ); } @@ -1209,8 +1224,12 @@ fn parse_interpolate_without_body() { let sql = "SELECT fname FROM customer ORDER BY fname WITH FILL INTERPOLATE"; let select = clickhouse().verified_query(sql); assert_eq!( - Some(Interpolate { exprs: None }), - select.order_by.expect("ORDER BY expected").interpolate + Some(Interpolate { exprs: None }).as_ref(), + select + .order_by + .expect("ORDER BY expected") + .interpolate + .as_ref() ); } @@ -1221,8 +1240,13 @@ fn parse_interpolate_with_empty_body() { assert_eq!( Some(Interpolate { exprs: Some(vec![]) - }), - select.order_by.expect("ORDER BY expected").interpolate + }) + .as_ref(), + select + .order_by + .expect("ORDER BY expected") + .interpolate + .as_ref() ); } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 653142dc6..c46072d06 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2389,27 +2389,33 @@ fn parse_select_order_by() { fn chk(sql: &str) { let select = verified_query(sql); assert_eq!( - vec![ + OrderByKind::Expressions(vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("lname")), - asc: Some(true), - nulls_first: None, + options: OrderByOptions { + asc: Some(true), + nulls_first: None, + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("fname")), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("id")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }, - ], - select.order_by.expect("ORDER BY expected").exprs + ]), + select.order_by.expect("ORDER BY expected").kind ); } chk("SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY lname ASC, fname DESC, id"); @@ -2424,46 +2430,186 @@ fn parse_select_order_by_limit() { ORDER BY lname ASC, fname DESC LIMIT 2"; let select = verified_query(sql); assert_eq!( - vec![ + OrderByKind::Expressions(vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("lname")), - asc: Some(true), - nulls_first: None, + options: OrderByOptions { + asc: Some(true), + nulls_first: None, + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("fname")), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }, - ], - select.order_by.expect("ORDER BY expected").exprs + ]), + select.order_by.expect("ORDER BY expected").kind ); assert_eq!(Some(Expr::Value(number("2"))), select.limit); } +#[test] +fn parse_select_order_by_all() { + fn chk(sql: &str, except_order_by: OrderByKind) { + let dialects = all_dialects_where(|d| d.supports_order_by_all()); + let select = dialects.verified_query(sql); + assert_eq!( + except_order_by, + select.order_by.expect("ORDER BY expected").kind + ); + } + let test_cases = [ + ( + "SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY ALL", + OrderByKind::All(OrderByOptions { + asc: None, + nulls_first: None, + }), + ), + ( + "SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY ALL NULLS FIRST", + OrderByKind::All(OrderByOptions { + asc: None, + nulls_first: Some(true), + }), + ), + ( + "SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY ALL NULLS LAST", + OrderByKind::All(OrderByOptions { + asc: None, + nulls_first: Some(false), + }), + ), + ( + "SELECT id, fname, lname FROM customer ORDER BY ALL ASC", + OrderByKind::All(OrderByOptions { + asc: Some(true), + nulls_first: None, + }), + ), + ( + "SELECT id, fname, lname FROM customer ORDER BY ALL ASC NULLS FIRST", + OrderByKind::All(OrderByOptions { + asc: Some(true), + nulls_first: Some(true), + }), + ), + ( + "SELECT id, fname, lname FROM customer ORDER BY ALL ASC NULLS LAST", + OrderByKind::All(OrderByOptions { + asc: Some(true), + nulls_first: Some(false), + }), + ), + ( + "SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY ALL DESC", + OrderByKind::All(OrderByOptions { + asc: Some(false), + nulls_first: None, + }), + ), + ( + "SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY ALL DESC NULLS FIRST", + OrderByKind::All(OrderByOptions { + asc: Some(false), + nulls_first: Some(true), + }), + ), + ( + "SELECT id, fname, lname FROM customer WHERE id < 5 ORDER BY ALL DESC NULLS LAST", + OrderByKind::All(OrderByOptions { + asc: Some(false), + nulls_first: Some(false), + }), + ), + ]; + + for (sql, expected_order_by) in test_cases { + chk(sql, expected_order_by); + } +} + +#[test] +fn parse_select_order_by_not_support_all() { + fn chk(sql: &str, except_order_by: OrderByKind) { + let dialects = all_dialects_where(|d| !d.supports_order_by_all()); + let select = dialects.verified_query(sql); + assert_eq!( + except_order_by, + select.order_by.expect("ORDER BY expected").kind + ); + } + let test_cases = [ + ( + "SELECT id, ALL FROM customer WHERE id < 5 ORDER BY ALL", + OrderByKind::Expressions(vec![OrderByExpr { + expr: Expr::Identifier(Ident::new("ALL")), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }]), + ), + ( + "SELECT id, ALL FROM customer ORDER BY ALL ASC NULLS FIRST", + OrderByKind::Expressions(vec![OrderByExpr { + expr: Expr::Identifier(Ident::new("ALL")), + options: OrderByOptions { + asc: Some(true), + nulls_first: Some(true), + }, + with_fill: None, + }]), + ), + ( + "SELECT id, ALL FROM customer ORDER BY ALL DESC NULLS LAST", + OrderByKind::Expressions(vec![OrderByExpr { + expr: Expr::Identifier(Ident::new("ALL")), + options: OrderByOptions { + asc: Some(false), + nulls_first: Some(false), + }, + with_fill: None, + }]), + ), + ]; + + for (sql, expected_order_by) in test_cases { + chk(sql, expected_order_by); + } +} + #[test] fn parse_select_order_by_nulls_order() { let sql = "SELECT id, fname, lname FROM customer WHERE id < 5 \ ORDER BY lname ASC NULLS FIRST, fname DESC NULLS LAST LIMIT 2"; let select = verified_query(sql); assert_eq!( - vec![ + OrderByKind::Expressions(vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("lname")), - asc: Some(true), - nulls_first: Some(true), + options: OrderByOptions { + asc: Some(true), + nulls_first: Some(true), + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("fname")), - asc: Some(false), - nulls_first: Some(false), + options: OrderByOptions { + asc: Some(false), + nulls_first: Some(false), + }, with_fill: None, }, - ], - select.order_by.expect("ORDER BY expeccted").exprs + ]), + select.order_by.expect("ORDER BY expeccted").kind ); assert_eq!(Some(Expr::Value(number("2"))), select.limit); } @@ -2641,8 +2787,10 @@ fn parse_select_qualify() { partition_by: vec![Expr::Identifier(Ident::new("p"))], order_by: vec![OrderByExpr { expr: Expr::Identifier(Ident::new("o")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }], window_frame: None, @@ -3065,8 +3213,10 @@ fn parse_listagg() { quote_style: None, span: Span::empty(), }), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }, OrderByExpr { @@ -3075,8 +3225,10 @@ fn parse_listagg() { quote_style: None, span: Span::empty(), }), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }, ] @@ -5172,8 +5324,10 @@ fn parse_window_functions() { partition_by: vec![], order_by: vec![OrderByExpr { expr: Expr::Identifier(Ident::new("dt")), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }], window_frame: None, @@ -5386,8 +5540,10 @@ fn test_parse_named_window() { quote_style: None, span: Span::empty(), }), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }], window_frame: None, @@ -8584,14 +8740,18 @@ fn parse_create_index() { let indexed_columns = vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("name")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("age")), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }, ]; @@ -8620,14 +8780,18 @@ fn test_create_index_with_using_function() { let indexed_columns = vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("name")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("age")), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }, ]; @@ -8664,8 +8828,10 @@ fn test_create_index_with_with_clause() { let sql = "CREATE UNIQUE INDEX title_idx ON films(title) WITH (fillfactor = 70, single_param)"; let indexed_columns = vec![OrderByExpr { expr: Expr::Identifier(Ident::new("title")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }]; let with_parameters = vec![ @@ -11345,8 +11511,10 @@ fn test_match_recognize() { partition_by: vec![Expr::Identifier(Ident::new("company"))], order_by: vec![OrderByExpr { expr: Expr::Identifier(Ident::new("price_date")), - asc: None, - nulls_first: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, with_fill: None, }], measures: vec![ diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 5d710b17d..b2b300aec 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -23,8 +23,8 @@ use sqlparser::ast::{ ClusteredBy, CommentDef, CreateFunction, CreateFunctionBody, CreateFunctionUsing, CreateTable, Expr, Function, FunctionArgumentList, FunctionArguments, Ident, ObjectName, - OneOrManyWithParens, OrderByExpr, SelectItem, Statement, TableFactor, UnaryOperator, Use, - Value, + OneOrManyWithParens, OrderByExpr, OrderByOptions, SelectItem, Statement, TableFactor, + UnaryOperator, Use, Value, }; use sqlparser::dialect::{GenericDialect, HiveDialect, MsSqlDialect}; use sqlparser::parser::ParserError; @@ -171,14 +171,18 @@ fn create_table_with_clustered_by() { sorted_by: Some(vec![ OrderByExpr { expr: Expr::Identifier(Ident::new("a")), - asc: Some(true), - nulls_first: None, + options: OrderByOptions { + asc: Some(true), + nulls_first: None, + }, with_fill: None, }, OrderByExpr { expr: Expr::Identifier(Ident::new("b")), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }, ]), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 44c8350fa..030710743 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2096,8 +2096,10 @@ fn parse_delete_with_order_by() { quote_style: None, span: Span::empty(), }), - asc: Some(false), - nulls_first: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, with_fill: None, }], order_by From 7fc37a76e57aa677d8c3958ddda32e9d65a8bb74 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sat, 22 Feb 2025 07:21:45 +0100 Subject: [PATCH 055/286] Parse casting to array using double colon operator in Redshift (#1737) --- src/dialect/duckdb.rs | 2 +- src/dialect/generic.rs | 2 +- src/dialect/mod.rs | 8 +++++--- src/dialect/postgresql.rs | 2 +- src/dialect/redshift.rs | 4 ++++ src/parser/mod.rs | 4 ++-- tests/sqlparser_common.rs | 7 +++++++ 7 files changed, 21 insertions(+), 8 deletions(-) diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index 3595aa713..3366c6705 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -82,7 +82,7 @@ impl Dialect for DuckDbDialect { } // See DuckDB - fn supports_array_typedef_size(&self) -> bool { + fn supports_array_typedef_with_brackets(&self) -> bool { true } diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index e04a288d6..041d44bb2 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -148,7 +148,7 @@ impl Dialect for GenericDialect { true } - fn supports_array_typedef_size(&self) -> bool { + fn supports_array_typedef_with_brackets(&self) -> bool { true } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index b68adc170..1c32bc513 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -931,9 +931,11 @@ pub trait Dialect: Debug + Any { false } - /// Returns true if the dialect supports size definition for array types. - /// For example: ```CREATE TABLE my_table (my_array INT[3])```. - fn supports_array_typedef_size(&self) -> bool { + /// Returns true if the dialect supports array type definition with brackets with + /// an optional size. For example: + /// ```CREATE TABLE my_table (arr1 INT[], arr2 INT[3])``` + /// ```SELECT x::INT[]``` + fn supports_array_typedef_with_brackets(&self) -> bool { false } /// Returns true if the dialect supports geometric types. diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index a20cfac4c..57ed0b684 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -247,7 +247,7 @@ impl Dialect for PostgreSqlDialect { } /// See: - fn supports_array_typedef_size(&self) -> bool { + fn supports_array_typedef_with_brackets(&self) -> bool { true } diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index 3dda762c3..25b8f1644 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -117,4 +117,8 @@ impl Dialect for RedshiftSqlDialect { fn supports_geometric_types(&self) -> bool { true } + + fn supports_array_typedef_with_brackets(&self) -> bool { + true + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f80754a3a..176a7ca13 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9123,9 +9123,9 @@ impl<'a> Parser<'a> { _ => self.expected_at("a data type name", next_token_index), }?; - if self.dialect.supports_array_typedef_size() { - // Parse array data type size + if self.dialect.supports_array_typedef_with_brackets() { while self.consume_token(&Token::LBracket) { + // Parse optional array data type size let size = self.maybe_parse(|p| p.parse_literal_uint())?; self.expect_token(&Token::RBracket)?; data = DataType::Array(ArrayElemTypeDef::SquareBracket(Box::new(data), size)) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index c46072d06..d6b8824be 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14386,3 +14386,10 @@ fn test_geometric_binary_operators() { } )); } + +#[test] +fn parse_array_type_def_with_brackets() { + let dialects = all_dialects_where(|d| d.supports_array_typedef_with_brackets()); + dialects.verified_stmt("SELECT x::INT[]"); + dialects.verified_stmt("SELECT STRING_TO_ARRAY('1,2,3', ',')::INT[3]"); +} From 72312ba82af1886c80eb6ff48ea2ae6f9f75205d Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Sat, 22 Feb 2025 07:23:36 +0100 Subject: [PATCH 056/286] Replace parallel condition/result vectors with single CaseWhen vector in Expr::Case (#1733) --- src/ast/mod.rs | 25 ++++++-- src/ast/spans.rs | 6 +- src/parser/mod.rs | 7 +-- tests/sqlparser_common.rs | 107 ++++++++++++++++++++++------------ tests/sqlparser_databricks.rs | 65 +++++++++++++++++++++ 5 files changed, 160 insertions(+), 50 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index ace752a86..649b1f79e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -600,6 +600,22 @@ pub enum CeilFloorKind { Scale(Value), } +/// A WHEN clause in a CASE expression containing both +/// the condition and its corresponding result +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CaseWhen { + pub condition: Expr, + pub result: Expr, +} + +impl fmt::Display for CaseWhen { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "WHEN {} THEN {}", self.condition, self.result) + } +} + /// An SQL expression of any type. /// /// # Semantics / Type Checking @@ -918,8 +934,7 @@ pub enum Expr { /// Case { operand: Option>, - conditions: Vec, - results: Vec, + conditions: Vec, else_result: Option>, }, /// An exists expression `[ NOT ] EXISTS(SELECT ...)`, used in expressions like @@ -1621,17 +1636,15 @@ impl fmt::Display for Expr { Expr::Case { operand, conditions, - results, else_result, } => { write!(f, "CASE")?; if let Some(operand) = operand { write!(f, " {operand}")?; } - for (c, r) in conditions.iter().zip(results) { - write!(f, " WHEN {c} THEN {r}")?; + for when in conditions { + write!(f, " {when}")?; } - if let Some(else_result) = else_result { write!(f, " ELSE {else_result}")?; } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 88c3d8aea..a036271cb 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1450,15 +1450,15 @@ impl Spanned for Expr { Expr::Case { operand, conditions, - results, else_result, } => union_spans( operand .as_ref() .map(|i| i.span()) .into_iter() - .chain(conditions.iter().map(|i| i.span())) - .chain(results.iter().map(|i| i.span())) + .chain(conditions.iter().flat_map(|case_when| { + [case_when.condition.span(), case_when.result.span()] + })) .chain(else_result.as_ref().map(|i| i.span())), ), Expr::Exists { subquery, .. } => subquery.span(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 176a7ca13..e40f4d58a 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2065,11 +2065,11 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::WHEN)?; } let mut conditions = vec![]; - let mut results = vec![]; loop { - conditions.push(self.parse_expr()?); + let condition = self.parse_expr()?; self.expect_keyword_is(Keyword::THEN)?; - results.push(self.parse_expr()?); + let result = self.parse_expr()?; + conditions.push(CaseWhen { condition, result }); if !self.parse_keyword(Keyword::WHEN) { break; } @@ -2083,7 +2083,6 @@ impl<'a> Parser<'a> { Ok(Expr::Case { operand, conditions, - results, else_result, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index d6b8824be..578c42de1 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -6695,22 +6695,26 @@ fn parse_searched_case_expr() { &Case { operand: None, conditions: vec![ - IsNull(Box::new(Identifier(Ident::new("bar")))), - BinaryOp { - left: Box::new(Identifier(Ident::new("bar"))), - op: Eq, - right: Box::new(Expr::Value(number("0"))), + CaseWhen { + condition: IsNull(Box::new(Identifier(Ident::new("bar")))), + result: Expr::Value(Value::SingleQuotedString("null".to_string())), }, - BinaryOp { - left: Box::new(Identifier(Ident::new("bar"))), - op: GtEq, - right: Box::new(Expr::Value(number("0"))), + CaseWhen { + condition: BinaryOp { + left: Box::new(Identifier(Ident::new("bar"))), + op: Eq, + right: Box::new(Expr::Value(number("0"))), + }, + result: Expr::Value(Value::SingleQuotedString("=0".to_string())), + }, + CaseWhen { + condition: BinaryOp { + left: Box::new(Identifier(Ident::new("bar"))), + op: GtEq, + right: Box::new(Expr::Value(number("0"))), + }, + result: Expr::Value(Value::SingleQuotedString(">=0".to_string())), }, - ], - results: vec![ - Expr::Value(Value::SingleQuotedString("null".to_string())), - Expr::Value(Value::SingleQuotedString("=0".to_string())), - Expr::Value(Value::SingleQuotedString(">=0".to_string())), ], else_result: Some(Box::new(Expr::Value(Value::SingleQuotedString( "<0".to_string() @@ -6729,8 +6733,10 @@ fn parse_simple_case_expr() { assert_eq!( &Case { operand: Some(Box::new(Identifier(Ident::new("foo")))), - conditions: vec![Expr::Value(number("1"))], - results: vec![Expr::Value(Value::SingleQuotedString("Y".to_string()))], + conditions: vec![CaseWhen { + condition: Expr::Value(number("1")), + result: Expr::Value(Value::SingleQuotedString("Y".to_string())), + }], else_result: Some(Box::new(Expr::Value(Value::SingleQuotedString( "N".to_string() )))), @@ -13902,6 +13908,31 @@ fn test_trailing_commas_in_from() { ); } +#[test] +#[cfg(feature = "visitor")] +fn test_visit_order() { + let sql = "SELECT CASE a WHEN 1 THEN 2 WHEN 3 THEN 4 ELSE 5 END"; + let stmt = verified_stmt(sql); + let mut visited = vec![]; + sqlparser::ast::visit_expressions(&stmt, |expr| { + visited.push(expr.to_string()); + core::ops::ControlFlow::<()>::Continue(()) + }); + + assert_eq!( + visited, + [ + "CASE a WHEN 1 THEN 2 WHEN 3 THEN 4 ELSE 5 END", + "a", + "1", + "2", + "3", + "4", + "5" + ] + ); +} + #[test] fn test_lambdas() { let dialects = all_dialects_where(|d| d.supports_lambda_functions()); @@ -13929,28 +13960,30 @@ fn test_lambdas() { body: Box::new(Expr::Case { operand: None, conditions: vec![ - Expr::BinaryOp { - left: Box::new(Expr::Identifier(Ident::new("p1"))), - op: BinaryOperator::Eq, - right: Box::new(Expr::Identifier(Ident::new("p2"))) + CaseWhen { + condition: Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("p1"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Identifier(Ident::new("p2"))) + }, + result: Expr::Value(number("0")) }, - Expr::BinaryOp { - left: Box::new(call( - "reverse", - [Expr::Identifier(Ident::new("p1"))] - )), - op: BinaryOperator::Lt, - right: Box::new(call( - "reverse", - [Expr::Identifier(Ident::new("p2"))] - )) - } - ], - results: vec![ - Expr::Value(number("0")), - Expr::UnaryOp { - op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(number("1"))) + CaseWhen { + condition: Expr::BinaryOp { + left: Box::new(call( + "reverse", + [Expr::Identifier(Ident::new("p1"))] + )), + op: BinaryOperator::Lt, + right: Box::new(call( + "reverse", + [Expr::Identifier(Ident::new("p2"))] + )) + }, + result: Expr::UnaryOp { + op: UnaryOperator::Minus, + expr: Box::new(Expr::Value(number("1"))) + } } ], else_result: Some(Box::new(Expr::Value(number("1")))) diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 8338a0e71..724bedf47 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -83,6 +83,71 @@ fn test_databricks_exists() { ); } +#[test] +fn test_databricks_lambdas() { + #[rustfmt::skip] + let sql = concat!( + "SELECT array_sort(array('Hello', 'World'), ", + "(p1, p2) -> CASE WHEN p1 = p2 THEN 0 ", + "WHEN reverse(p1) < reverse(p2) THEN -1 ", + "ELSE 1 END)", + ); + pretty_assertions::assert_eq!( + SelectItem::UnnamedExpr(call( + "array_sort", + [ + call( + "array", + [ + Expr::Value(Value::SingleQuotedString("Hello".to_owned())), + Expr::Value(Value::SingleQuotedString("World".to_owned())) + ] + ), + Expr::Lambda(LambdaFunction { + params: OneOrManyWithParens::Many(vec![Ident::new("p1"), Ident::new("p2")]), + body: Box::new(Expr::Case { + operand: None, + conditions: vec![ + CaseWhen { + condition: Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("p1"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Identifier(Ident::new("p2"))) + }, + result: Expr::Value(number("0")) + }, + CaseWhen { + condition: Expr::BinaryOp { + left: Box::new(call( + "reverse", + [Expr::Identifier(Ident::new("p1"))] + )), + op: BinaryOperator::Lt, + right: Box::new(call( + "reverse", + [Expr::Identifier(Ident::new("p2"))] + )), + }, + result: Expr::UnaryOp { + op: UnaryOperator::Minus, + expr: Box::new(Expr::Value(number("1"))) + } + }, + ], + else_result: Some(Box::new(Expr::Value(number("1")))) + }) + }) + ] + )), + databricks().verified_only_select(sql).projection[0] + ); + + databricks().verified_expr( + "map_zip_with(map(1, 'a', 2, 'b'), map(1, 'x', 2, 'y'), (k, v1, v2) -> concat(v1, v2))", + ); + databricks().verified_expr("transform(array(1, 2, 3), x -> x + 1)"); +} + #[test] fn test_values_clause() { let values = Values { From aab12add36bfb4dfd60c2ba38682b503cc248199 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Mon, 24 Feb 2025 08:34:36 +0100 Subject: [PATCH 057/286] BigQuery: Add support for `BEGIN` (#1718) --- src/ast/mod.rs | 43 ++++++++++++++++++++++++++++ src/dialect/bigquery.rs | 57 +++++++++++++++++++++++++++++++++++-- src/parser/mod.rs | 27 ++++++++++++++++++ tests/sqlparser_bigquery.rs | 46 ++++++++++++++++++++++++++++++ tests/sqlparser_common.rs | 33 ++++++++++++--------- tests/sqlparser_sqlite.rs | 17 ----------- 6 files changed, 190 insertions(+), 33 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 649b1f79e..aad122fbf 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3072,6 +3072,28 @@ pub enum Statement { begin: bool, transaction: Option, modifier: Option, + /// List of statements belonging to the `BEGIN` block. + /// Example: + /// ```sql + /// BEGIN + /// SELECT 1; + /// SELECT 2; + /// END; + /// ``` + statements: Vec, + /// Statements of an exception clause. + /// Example: + /// ```sql + /// BEGIN + /// SELECT 1; + /// EXCEPTION WHEN ERROR THEN + /// SELECT 2; + /// SELECT 3; + /// END; + /// + exception_statements: Option>, + /// TRUE if the statement has an `END` keyword. + has_end_keyword: bool, }, /// ```sql /// SET TRANSACTION ... @@ -4815,6 +4837,9 @@ impl fmt::Display for Statement { begin: syntax_begin, transaction, modifier, + statements, + exception_statements, + has_end_keyword, } => { if *syntax_begin { if let Some(modifier) = *modifier { @@ -4831,6 +4856,24 @@ impl fmt::Display for Statement { if !modes.is_empty() { write!(f, " {}", display_comma_separated(modes))?; } + if !statements.is_empty() { + write!(f, " {}", display_separated(statements, "; "))?; + // We manually insert semicolon for the last statement, + // since display_separated doesn't handle that case. + write!(f, ";")?; + } + if let Some(exception_statements) = exception_statements { + write!(f, " EXCEPTION WHEN ERROR THEN")?; + if !exception_statements.is_empty() { + write!(f, " {}", display_separated(exception_statements, "; "))?; + // We manually insert semicolon for the last statement, + // since display_separated doesn't handle that case. + write!(f, ";")?; + } + } + if *has_end_keyword { + write!(f, " END")?; + } Ok(()) } Statement::SetTransaction { diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index b8e7e4cf0..49fb24f19 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -15,9 +15,10 @@ // specific language governing permissions and limitations // under the License. +use crate::ast::Statement; use crate::dialect::Dialect; use crate::keywords::Keyword; -use crate::parser::Parser; +use crate::parser::{Parser, ParserError}; /// These keywords are disallowed as column identifiers. Such that /// `SELECT 5 AS FROM T` is rejected by BigQuery. @@ -44,7 +45,11 @@ const RESERVED_FOR_COLUMN_ALIAS: &[Keyword] = &[ pub struct BigQueryDialect; impl Dialect for BigQueryDialect { - // See https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#identifiers + fn parse_statement(&self, parser: &mut Parser) -> Option> { + self.maybe_parse_statement(parser) + } + + /// See fn is_delimited_identifier_start(&self, ch: char) -> bool { ch == '`' } @@ -60,6 +65,9 @@ impl Dialect for BigQueryDialect { fn is_identifier_start(&self, ch: char) -> bool { ch.is_ascii_lowercase() || ch.is_ascii_uppercase() || ch == '_' + // BigQuery supports `@@foo.bar` variable syntax in its procedural language. + // https://cloud.google.com/bigquery/docs/reference/standard-sql/procedural-language#beginexceptionend + || ch == '@' } fn is_identifier_part(&self, ch: char) -> bool { @@ -129,3 +137,48 @@ impl Dialect for BigQueryDialect { !RESERVED_FOR_COLUMN_ALIAS.contains(kw) } } + +impl BigQueryDialect { + fn maybe_parse_statement(&self, parser: &mut Parser) -> Option> { + if parser.peek_keyword(Keyword::BEGIN) { + return Some(self.parse_begin(parser)); + } + None + } + + /// Parse a `BEGIN` statement. + /// + fn parse_begin(&self, parser: &mut Parser) -> Result { + parser.expect_keyword(Keyword::BEGIN)?; + + let statements = parser.parse_statement_list(&[Keyword::EXCEPTION, Keyword::END])?; + + let has_exception_when_clause = parser.parse_keywords(&[ + Keyword::EXCEPTION, + Keyword::WHEN, + Keyword::ERROR, + Keyword::THEN, + ]); + let exception_statements = if has_exception_when_clause { + if !parser.peek_keyword(Keyword::END) { + Some(parser.parse_statement_list(&[Keyword::END])?) + } else { + Some(Default::default()) + } + } else { + None + }; + + parser.expect_keyword(Keyword::END)?; + + Ok(Statement::StartTransaction { + begin: true, + statements, + exception_statements, + has_end_keyword: true, + transaction: None, + modifier: None, + modes: Default::default(), + }) + } +} diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e40f4d58a..c08c7049d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4273,6 +4273,27 @@ impl<'a> Parser<'a> { self.parse_comma_separated(f) } + /// Parses 0 or more statements, each followed by a semicolon. + /// If the next token is any of `terminal_keywords` then no more + /// statements will be parsed. + pub(crate) fn parse_statement_list( + &mut self, + terminal_keywords: &[Keyword], + ) -> Result, ParserError> { + let mut values = vec![]; + loop { + if let Token::Word(w) = &self.peek_nth_token_ref(0).token { + if w.quote_style.is_none() && terminal_keywords.contains(&w.keyword) { + break; + } + } + + values.push(self.parse_statement()?); + self.expect_token(&Token::SemiColon)?; + } + Ok(values) + } + /// Default implementation of a predicate that returns true if /// the specified keyword is reserved for column alias. /// See [Dialect::is_column_alias] @@ -13783,6 +13804,9 @@ impl<'a> Parser<'a> { begin: false, transaction: Some(BeginTransactionKind::Transaction), modifier: None, + statements: vec![], + exception_statements: None, + has_end_keyword: false, }) } @@ -13812,6 +13836,9 @@ impl<'a> Parser<'a> { begin: true, transaction, modifier, + statements: vec![], + exception_statements: None, + has_end_keyword: false, }) } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 52aa3b3b4..55e354221 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -236,6 +236,52 @@ fn parse_big_query_non_reserved_column_alias() { bigquery().verified_stmt(sql); } +#[test] +fn parse_at_at_identifier() { + bigquery().verified_stmt("SELECT @@error.stack_trace, @@error.message"); +} + +#[test] +fn parse_begin() { + let sql = r#"BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT 2; END"#; + let Statement::StartTransaction { + statements, + exception_statements, + has_end_keyword, + .. + } = bigquery().verified_stmt(sql) + else { + unreachable!(); + }; + assert_eq!(1, statements.len()); + assert_eq!(1, exception_statements.unwrap().len()); + assert!(has_end_keyword); + + bigquery().verified_stmt( + "BEGIN SELECT 1; SELECT 2; EXCEPTION WHEN ERROR THEN SELECT 2; SELECT 4; END", + ); + bigquery() + .verified_stmt("BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT @@error.stack_trace; END"); + bigquery().verified_stmt("BEGIN EXCEPTION WHEN ERROR THEN SELECT 2; END"); + bigquery().verified_stmt("BEGIN SELECT 1; SELECT 2; EXCEPTION WHEN ERROR THEN END"); + bigquery().verified_stmt("BEGIN EXCEPTION WHEN ERROR THEN END"); + bigquery().verified_stmt("BEGIN SELECT 1; SELECT 2; END"); + bigquery().verified_stmt("BEGIN END"); + + assert_eq!( + bigquery() + .parse_sql_statements("BEGIN SELECT 1; SELECT 2 END") + .unwrap_err(), + ParserError::ParserError("Expected: ;, found: END".to_string()) + ); + assert_eq!( + bigquery() + .parse_sql_statements("BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT 2 END") + .unwrap_err(), + ParserError::ParserError("Expected: ;, found: END".to_string()) + ); +} + #[test] fn parse_delete_statement() { let sql = "DELETE \"table\" WHERE 1"; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 578c42de1..0072baf75 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8343,7 +8343,12 @@ fn lateral_function() { #[test] fn parse_start_transaction() { - match verified_stmt("START TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE") { + let dialects = all_dialects_except(|d| + // BigQuery does not support this syntax + d.is::()); + match dialects + .verified_stmt("START TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE") + { Statement::StartTransaction { modes, .. } => assert_eq!( modes, vec![ @@ -8357,7 +8362,7 @@ fn parse_start_transaction() { // For historical reasons, PostgreSQL allows the commas between the modes to // be omitted. - match one_statement_parses_to( + match dialects.one_statement_parses_to( "START TRANSACTION READ ONLY READ WRITE ISOLATION LEVEL SERIALIZABLE", "START TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE", ) { @@ -8372,40 +8377,40 @@ fn parse_start_transaction() { _ => unreachable!(), } - verified_stmt("START TRANSACTION"); - verified_stmt("BEGIN"); - verified_stmt("BEGIN WORK"); - verified_stmt("BEGIN TRANSACTION"); + dialects.verified_stmt("START TRANSACTION"); + dialects.verified_stmt("BEGIN"); + dialects.verified_stmt("BEGIN WORK"); + dialects.verified_stmt("BEGIN TRANSACTION"); - verified_stmt("START TRANSACTION ISOLATION LEVEL READ UNCOMMITTED"); - verified_stmt("START TRANSACTION ISOLATION LEVEL READ COMMITTED"); - verified_stmt("START TRANSACTION ISOLATION LEVEL REPEATABLE READ"); - verified_stmt("START TRANSACTION ISOLATION LEVEL SERIALIZABLE"); + dialects.verified_stmt("START TRANSACTION ISOLATION LEVEL READ UNCOMMITTED"); + dialects.verified_stmt("START TRANSACTION ISOLATION LEVEL READ COMMITTED"); + dialects.verified_stmt("START TRANSACTION ISOLATION LEVEL REPEATABLE READ"); + dialects.verified_stmt("START TRANSACTION ISOLATION LEVEL SERIALIZABLE"); // Regression test for https://github.com/sqlparser-rs/sqlparser-rs/pull/139, // in which START TRANSACTION would fail to parse if followed by a statement // terminator. assert_eq!( - parse_sql_statements("START TRANSACTION; SELECT 1"), + dialects.parse_sql_statements("START TRANSACTION; SELECT 1"), Ok(vec![ verified_stmt("START TRANSACTION"), verified_stmt("SELECT 1"), ]) ); - let res = parse_sql_statements("START TRANSACTION ISOLATION LEVEL BAD"); + let res = dialects.parse_sql_statements("START TRANSACTION ISOLATION LEVEL BAD"); assert_eq!( ParserError::ParserError("Expected: isolation level, found: BAD".to_string()), res.unwrap_err() ); - let res = parse_sql_statements("START TRANSACTION BAD"); + let res = dialects.parse_sql_statements("START TRANSACTION BAD"); assert_eq!( ParserError::ParserError("Expected: end of statement, found: BAD".to_string()), res.unwrap_err() ); - let res = parse_sql_statements("START TRANSACTION READ ONLY,"); + let res = dialects.parse_sql_statements("START TRANSACTION READ ONLY,"); assert_eq!( ParserError::ParserError("Expected: transaction mode, found: EOF".to_string()), res.unwrap_err() diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index c17743305..17dcfed85 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -518,23 +518,6 @@ fn parse_start_transaction_with_modifier() { sqlite_and_generic().verified_stmt("BEGIN DEFERRED"); sqlite_and_generic().verified_stmt("BEGIN IMMEDIATE"); sqlite_and_generic().verified_stmt("BEGIN EXCLUSIVE"); - - let unsupported_dialects = all_dialects_except(|d| d.supports_start_transaction_modifier()); - let res = unsupported_dialects.parse_sql_statements("BEGIN DEFERRED"); - assert_eq!( - ParserError::ParserError("Expected: end of statement, found: DEFERRED".to_string()), - res.unwrap_err(), - ); - let res = unsupported_dialects.parse_sql_statements("BEGIN IMMEDIATE"); - assert_eq!( - ParserError::ParserError("Expected: end of statement, found: IMMEDIATE".to_string()), - res.unwrap_err(), - ); - let res = unsupported_dialects.parse_sql_statements("BEGIN EXCLUSIVE"); - assert_eq!( - ParserError::ParserError("Expected: end of statement, found: EXCLUSIVE".to_string()), - res.unwrap_err(), - ); } #[test] From c335c8883be2bf9407166a07c0b55ee9d8195a4d Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Tue, 25 Feb 2025 07:33:57 +0100 Subject: [PATCH 058/286] Store spans for Value expressions (#1738) --- src/ast/mod.rs | 21 +- src/ast/spans.rs | 41 +- src/ast/value.rs | 101 +++- src/ast/visitor.rs | 2 +- src/parser/mod.rs | 134 +++-- tests/sqlparser_bigquery.rs | 473 ++++++++-------- tests/sqlparser_clickhouse.rs | 97 ++-- tests/sqlparser_common.rs | 902 ++++++++++++++++++------------ tests/sqlparser_custom_dialect.rs | 2 +- tests/sqlparser_databricks.rs | 41 +- tests/sqlparser_duckdb.rs | 43 +- tests/sqlparser_hive.rs | 3 +- tests/sqlparser_mssql.rs | 137 +++-- tests/sqlparser_mysql.rs | 136 +++-- tests/sqlparser_postgres.rs | 431 ++++++++------ tests/sqlparser_redshift.rs | 24 +- tests/sqlparser_snowflake.rs | 46 +- tests/sqlparser_sqlite.rs | 28 +- 18 files changed, 1620 insertions(+), 1042 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index aad122fbf..afe3e77d2 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -86,7 +86,7 @@ pub use self::trigger::{ pub use self::value::{ escape_double_quote_string, escape_quoted_string, DateTimeField, DollarQuotedString, - NormalizationForm, TrimWhereField, Value, + NormalizationForm, TrimWhereField, Value, ValueWithSpan, }; use crate::ast::helpers::key_value_options::KeyValueOptions; @@ -908,7 +908,7 @@ pub enum Expr { /// Nested expression e.g. `(foo > bar)` or `(1)` Nested(Box), /// A literal value, such as string, number, date or NULL - Value(Value), + Value(ValueWithSpan), /// IntroducedString { introducer: String, @@ -1051,6 +1051,13 @@ pub enum Expr { Lambda(LambdaFunction), } +impl Expr { + /// Creates a new [`Expr::Value`] + pub fn value(value: impl Into) -> Self { + Expr::Value(value.into()) + } +} + /// The contents inside the `[` and `]` in a subscript expression. #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -8789,9 +8796,9 @@ mod tests { #[test] fn test_interval_display() { let interval = Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from( - "123:45.67", - )))), + value: Box::new(Expr::Value( + Value::SingleQuotedString(String::from("123:45.67")).with_empty_span(), + )), leading_field: Some(DateTimeField::Minute), leading_precision: Some(10), last_field: Some(DateTimeField::Second), @@ -8803,7 +8810,9 @@ mod tests { ); let interval = Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from("5")))), + value: Box::new(Expr::Value( + Value::SingleQuotedString(String::from("5")).with_empty_span(), + )), leading_field: Some(DateTimeField::Second), leading_precision: Some(1), last_field: None, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index a036271cb..cfc3eb633 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -21,21 +21,21 @@ use core::iter; use crate::tokenizer::Span; use super::{ - dcl::SecondaryRoles, AccessExpr, AlterColumnOperation, AlterIndexOperation, - AlterTableOperation, Array, Assignment, AssignmentTarget, CloseCursor, ClusteredIndex, - ColumnDef, ColumnOption, ColumnOptionDef, ConflictTarget, ConnectBy, ConstraintCharacteristics, - CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, Delete, DoUpdate, - ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, Function, - FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, FunctionArguments, - GroupByExpr, HavingBound, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, - JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, MatchRecognizePattern, - Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, - OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, - ProjectionSelect, Query, ReferentialAction, RenameSelectItem, ReplaceSelectElement, - ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, - SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, - TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, - WildcardAdditionalOptions, With, WithFill, + dcl::SecondaryRoles, value::ValueWithSpan, AccessExpr, AlterColumnOperation, + AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, CloseCursor, + ClusteredIndex, ColumnDef, ColumnOption, ColumnOptionDef, ConflictTarget, ConnectBy, + ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, + Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, + Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, + FunctionArguments, GroupByExpr, HavingBound, IlikeSelectItem, Insert, Interpolate, + InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, + MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, + OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, + PivotValueSource, ProjectionSelect, Query, ReferentialAction, RenameSelectItem, + ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, + Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, + TableFactor, TableObject, TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, + Value, Values, ViewColumnDef, WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -1978,10 +1978,13 @@ impl Spanned for TableAliasColumnDef { } } -/// # missing span -/// -/// The span of a `Value` is currently not implemented, as doing so -/// requires a breaking changes, which may be done in a future release. +impl Spanned for ValueWithSpan { + fn span(&self) -> Span { + self.span + } +} + +/// The span is stored in the `ValueWrapper` struct impl Spanned for Value { fn span(&self) -> Span { Span::empty() // # todo: Value needs to store spans before this is possible diff --git a/src/ast/value.rs b/src/ast/value.rs index f2f02754b..77e2e0e81 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -26,10 +26,88 @@ use bigdecimal::BigDecimal; #[cfg(feature = "serde")] use serde::{Deserialize, Serialize}; -use crate::ast::Ident; +use crate::{ast::Ident, tokenizer::Span}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; +/// Wraps a primitive SQL [`Value`] with its [`Span`] location +/// +/// # Example: create a `ValueWithSpan` from a `Value` +/// ``` +/// # use sqlparser::ast::{Value, ValueWithSpan}; +/// # use sqlparser::tokenizer::{Location, Span}; +/// let value = Value::SingleQuotedString(String::from("endpoint")); +/// // from line 1, column 1 to line 1, column 7 +/// let span = Span::new(Location::new(1, 1), Location::new(1, 7)); +/// let value_with_span = value.with_span(span); +/// ``` +/// +/// # Example: create a `ValueWithSpan` from a `Value` with an empty span +/// +/// You can call [`Value::with_empty_span`] to create a `ValueWithSpan` with an empty span +/// ``` +/// # use sqlparser::ast::{Value, ValueWithSpan}; +/// # use sqlparser::tokenizer::{Location, Span}; +/// let value = Value::SingleQuotedString(String::from("endpoint")); +/// let value_with_span = value.with_empty_span(); +/// assert_eq!(value_with_span.span, Span::empty()); +/// ``` +/// +/// You can also use the [`From`] trait to convert `ValueWithSpan` to/from `Value`s +/// ``` +/// # use sqlparser::ast::{Value, ValueWithSpan}; +/// # use sqlparser::tokenizer::{Location, Span}; +/// let value = Value::SingleQuotedString(String::from("endpoint")); +/// // converting `Value` to `ValueWithSpan` results in an empty span +/// let value_with_span: ValueWithSpan = value.into(); +/// assert_eq!(value_with_span.span, Span::empty()); +/// // convert back to `Value` +/// let value: Value = value_with_span.into(); +/// ``` +#[derive(Debug, Clone, Eq)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ValueWithSpan { + pub value: Value, + pub span: Span, +} + +impl PartialEq for ValueWithSpan { + fn eq(&self, other: &Self) -> bool { + self.value == other.value + } +} + +impl Ord for ValueWithSpan { + fn cmp(&self, other: &Self) -> core::cmp::Ordering { + self.value.cmp(&other.value) + } +} + +impl PartialOrd for ValueWithSpan { + fn partial_cmp(&self, other: &Self) -> Option { + Some(Ord::cmp(self, other)) + } +} + +impl core::hash::Hash for ValueWithSpan { + fn hash(&self, state: &mut H) { + self.value.hash(state); + } +} + +impl From for ValueWithSpan { + fn from(value: Value) -> Self { + value.with_empty_span() + } +} + +impl From for Value { + fn from(value: ValueWithSpan) -> Self { + value.value + } +} + /// Primitive SQL values such as number and string #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -102,6 +180,13 @@ pub enum Value { Placeholder(String), } +impl ValueWithSpan { + /// If the underlying literal is a string, regardless of quote style, returns the associated string value + pub fn into_string(self) -> Option { + self.value.into_string() + } +} + impl Value { /// If the underlying literal is a string, regardless of quote style, returns the associated string value pub fn into_string(self) -> Option { @@ -126,6 +211,20 @@ impl Value { _ => None, } } + + pub fn with_span(self, span: Span) -> ValueWithSpan { + ValueWithSpan { value: self, span } + } + + pub fn with_empty_span(self) -> ValueWithSpan { + self.with_span(Span::empty()) + } +} + +impl fmt::Display for ValueWithSpan { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}", self.value) + } } impl fmt::Display for Value { diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index bb6246498..a5d355fe4 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -547,7 +547,7 @@ where /// /// visit_expressions_mut(&mut statements, |expr| { /// if matches!(expr, Expr::Identifier(col_name) if col_name.value == "x") { -/// let old_expr = std::mem::replace(expr, Expr::Value(Value::Null)); +/// let old_expr = std::mem::replace(expr, Expr::value(Value::Null)); /// *expr = Expr::Function(Function { /// name: ObjectName::from(vec![Ident::new("f")]), /// uses_odbc_syntax: false, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c08c7049d..72e4567cb 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1315,7 +1315,7 @@ impl<'a> Parser<'a> { DataType::Custom(..) => parser_err!("dummy", loc), data_type => Ok(Expr::TypedString { data_type, - value: parser.parse_value()?, + value: parser.parse_value()?.value, }), } })?; @@ -1503,7 +1503,7 @@ impl<'a> Parser<'a> { } fn parse_geometric_type(&mut self, kind: GeometricTypeKind) -> Result { - let value: Value = self.parse_value()?; + let value: Value = self.parse_value()?.value; Ok(Expr::TypedString { data_type: DataType::GeometricType(kind), value, @@ -2089,7 +2089,7 @@ impl<'a> Parser<'a> { pub fn parse_optional_cast_format(&mut self) -> Result, ParserError> { if self.parse_keyword(Keyword::FORMAT) { - let value = self.parse_value()?; + let value = self.parse_value()?.value; match self.parse_optional_time_zone()? { Some(tz) => Ok(Some(CastFormat::ValueAtTimeZone(value, tz))), None => Ok(Some(CastFormat::Value(value))), @@ -2101,7 +2101,7 @@ impl<'a> Parser<'a> { pub fn parse_optional_time_zone(&mut self) -> Result, ParserError> { if self.parse_keywords(&[Keyword::AT, Keyword::TIME, Keyword::ZONE]) { - self.parse_value().map(Some) + self.parse_value().map(|v| Some(v.value)) } else { Ok(None) } @@ -2230,7 +2230,7 @@ impl<'a> Parser<'a> { CeilFloorKind::DateTimeField(self.parse_date_time_field()?) } else if self.consume_token(&Token::Comma) { // Parse `CEIL/FLOOR(expr, scale)` - match self.parse_value()? { + match self.parse_value()?.value { Value::Number(n, s) => CeilFloorKind::Scale(Value::Number(n, s)), _ => { return Err(ParserError::ParserError( @@ -2566,7 +2566,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; // MySQL is too permissive about the value, IMO we can't validate it perfectly on syntax level. - let match_value = self.parse_value()?; + let match_value = self.parse_value()?.value; let in_natural_language_mode_keywords = &[ Keyword::IN, @@ -6324,11 +6324,11 @@ impl<'a> Parser<'a> { FetchDirection::Last } else if self.parse_keyword(Keyword::ABSOLUTE) { FetchDirection::Absolute { - limit: self.parse_number_value()?, + limit: self.parse_number_value()?.value, } } else if self.parse_keyword(Keyword::RELATIVE) { FetchDirection::Relative { - limit: self.parse_number_value()?, + limit: self.parse_number_value()?.value, } } else if self.parse_keyword(Keyword::FORWARD) { if self.parse_keyword(Keyword::ALL) { @@ -6336,7 +6336,7 @@ impl<'a> Parser<'a> { } else { FetchDirection::Forward { // TODO: Support optional - limit: Some(self.parse_number_value()?), + limit: Some(self.parse_number_value()?.value), } } } else if self.parse_keyword(Keyword::BACKWARD) { @@ -6345,14 +6345,14 @@ impl<'a> Parser<'a> { } else { FetchDirection::Backward { // TODO: Support optional - limit: Some(self.parse_number_value()?), + limit: Some(self.parse_number_value()?.value), } } } else if self.parse_keyword(Keyword::ALL) { FetchDirection::All } else { FetchDirection::Count { - limit: self.parse_number_value()?, + limit: self.parse_number_value()?.value, } }; @@ -7345,7 +7345,7 @@ impl<'a> Parser<'a> { }; self.expect_keyword_is(Keyword::INTO)?; - let num_buckets = self.parse_number_value()?; + let num_buckets = self.parse_number_value()?.value; self.expect_keyword_is(Keyword::BUCKETS)?; Some(ClusteredBy { columns, @@ -8579,21 +8579,22 @@ impl<'a> Parser<'a> { } /// Parse a literal value (numbers, strings, date/time, booleans) - pub fn parse_value(&mut self) -> Result { + pub fn parse_value(&mut self) -> Result { let next_token = self.next_token(); let span = next_token.span; + let ok_value = |value: Value| Ok(value.with_span(span)); match next_token.token { Token::Word(w) => match w.keyword { Keyword::TRUE if self.dialect.supports_boolean_literals() => { - Ok(Value::Boolean(true)) + ok_value(Value::Boolean(true)) } Keyword::FALSE if self.dialect.supports_boolean_literals() => { - Ok(Value::Boolean(false)) + ok_value(Value::Boolean(false)) } - Keyword::NULL => Ok(Value::Null), + Keyword::NULL => ok_value(Value::Null), Keyword::NoKeyword if w.quote_style.is_some() => match w.quote_style { - Some('"') => Ok(Value::DoubleQuotedString(w.value)), - Some('\'') => Ok(Value::SingleQuotedString(w.value)), + Some('"') => ok_value(Value::DoubleQuotedString(w.value)), + Some('\'') => ok_value(Value::SingleQuotedString(w.value)), _ => self.expected( "A value?", TokenWithSpan { @@ -8613,45 +8614,51 @@ impl<'a> Parser<'a> { // The call to n.parse() returns a bigdecimal when the // bigdecimal feature is enabled, and is otherwise a no-op // (i.e., it returns the input string). - Token::Number(n, l) => Ok(Value::Number(Self::parse(n, span.start)?, l)), - Token::SingleQuotedString(ref s) => Ok(Value::SingleQuotedString(s.to_string())), - Token::DoubleQuotedString(ref s) => Ok(Value::DoubleQuotedString(s.to_string())), + Token::Number(n, l) => ok_value(Value::Number(Self::parse(n, span.start)?, l)), + Token::SingleQuotedString(ref s) => ok_value(Value::SingleQuotedString(s.to_string())), + Token::DoubleQuotedString(ref s) => ok_value(Value::DoubleQuotedString(s.to_string())), Token::TripleSingleQuotedString(ref s) => { - Ok(Value::TripleSingleQuotedString(s.to_string())) + ok_value(Value::TripleSingleQuotedString(s.to_string())) } Token::TripleDoubleQuotedString(ref s) => { - Ok(Value::TripleDoubleQuotedString(s.to_string())) + ok_value(Value::TripleDoubleQuotedString(s.to_string())) } - Token::DollarQuotedString(ref s) => Ok(Value::DollarQuotedString(s.clone())), + Token::DollarQuotedString(ref s) => ok_value(Value::DollarQuotedString(s.clone())), Token::SingleQuotedByteStringLiteral(ref s) => { - Ok(Value::SingleQuotedByteStringLiteral(s.clone())) + ok_value(Value::SingleQuotedByteStringLiteral(s.clone())) } Token::DoubleQuotedByteStringLiteral(ref s) => { - Ok(Value::DoubleQuotedByteStringLiteral(s.clone())) + ok_value(Value::DoubleQuotedByteStringLiteral(s.clone())) } Token::TripleSingleQuotedByteStringLiteral(ref s) => { - Ok(Value::TripleSingleQuotedByteStringLiteral(s.clone())) + ok_value(Value::TripleSingleQuotedByteStringLiteral(s.clone())) } Token::TripleDoubleQuotedByteStringLiteral(ref s) => { - Ok(Value::TripleDoubleQuotedByteStringLiteral(s.clone())) + ok_value(Value::TripleDoubleQuotedByteStringLiteral(s.clone())) } Token::SingleQuotedRawStringLiteral(ref s) => { - Ok(Value::SingleQuotedRawStringLiteral(s.clone())) + ok_value(Value::SingleQuotedRawStringLiteral(s.clone())) } Token::DoubleQuotedRawStringLiteral(ref s) => { - Ok(Value::DoubleQuotedRawStringLiteral(s.clone())) + ok_value(Value::DoubleQuotedRawStringLiteral(s.clone())) } Token::TripleSingleQuotedRawStringLiteral(ref s) => { - Ok(Value::TripleSingleQuotedRawStringLiteral(s.clone())) + ok_value(Value::TripleSingleQuotedRawStringLiteral(s.clone())) } Token::TripleDoubleQuotedRawStringLiteral(ref s) => { - Ok(Value::TripleDoubleQuotedRawStringLiteral(s.clone())) + ok_value(Value::TripleDoubleQuotedRawStringLiteral(s.clone())) } - Token::NationalStringLiteral(ref s) => Ok(Value::NationalStringLiteral(s.to_string())), - Token::EscapedStringLiteral(ref s) => Ok(Value::EscapedStringLiteral(s.to_string())), - Token::UnicodeStringLiteral(ref s) => Ok(Value::UnicodeStringLiteral(s.to_string())), - Token::HexStringLiteral(ref s) => Ok(Value::HexStringLiteral(s.to_string())), - Token::Placeholder(ref s) => Ok(Value::Placeholder(s.to_string())), + Token::NationalStringLiteral(ref s) => { + ok_value(Value::NationalStringLiteral(s.to_string())) + } + Token::EscapedStringLiteral(ref s) => { + ok_value(Value::EscapedStringLiteral(s.to_string())) + } + Token::UnicodeStringLiteral(ref s) => { + ok_value(Value::UnicodeStringLiteral(s.to_string())) + } + Token::HexStringLiteral(ref s) => ok_value(Value::HexStringLiteral(s.to_string())), + Token::Placeholder(ref s) => ok_value(Value::Placeholder(s.to_string())), tok @ Token::Colon | tok @ Token::AtSign => { // Not calling self.parse_identifier(false)? because only in placeholder we want to check numbers as idfentifies // This because snowflake allows numbers as placeholders @@ -8662,7 +8669,7 @@ impl<'a> Parser<'a> { _ => self.expected("placeholder", next_token), }?; let placeholder = tok.to_string() + &ident.value; - Ok(Value::Placeholder(placeholder)) + ok_value(Value::Placeholder(placeholder)) } unexpected => self.expected( "a value", @@ -8675,10 +8682,11 @@ impl<'a> Parser<'a> { } /// Parse an unsigned numeric literal - pub fn parse_number_value(&mut self) -> Result { - match self.parse_value()? { - v @ Value::Number(_, _) => Ok(v), - v @ Value::Placeholder(_) => Ok(v), + pub fn parse_number_value(&mut self) -> Result { + let value_wrapper = self.parse_value()?; + match &value_wrapper.value { + Value::Number(_, _) => Ok(value_wrapper), + Value::Placeholder(_) => Ok(value_wrapper), _ => { self.prev_token(); self.expected("literal number", self.peek_token()) @@ -8736,15 +8744,16 @@ impl<'a> Parser<'a> { /// e.g. `CREATE FUNCTION ... AS $$ body $$`. fn parse_create_function_body_string(&mut self) -> Result { let peek_token = self.peek_token(); + let span = peek_token.span; match peek_token.token { Token::DollarQuotedString(s) if dialect_of!(self is PostgreSqlDialect | GenericDialect) => { self.next_token(); - Ok(Expr::Value(Value::DollarQuotedString(s))) + Ok(Expr::Value(Value::DollarQuotedString(s).with_span(span))) } - _ => Ok(Expr::Value(Value::SingleQuotedString( - self.parse_literal_string()?, - ))), + _ => Ok(Expr::Value( + Value::SingleQuotedString(self.parse_literal_string()?).with_span(span), + )), } } @@ -10268,7 +10277,7 @@ impl<'a> Parser<'a> { let key_values = self.parse_comma_separated(|p| { let key = p.parse_identifier()?; p.expect_token(&Token::Eq)?; - let value = p.parse_value()?; + let value = p.parse_value()?.value; Ok(Setting { key, value }) })?; Some(key_values) @@ -10990,7 +10999,7 @@ impl<'a> Parser<'a> { }) } else if variable.to_string() == "TRANSACTION" && modifier.is_none() { if self.parse_keyword(Keyword::SNAPSHOT) { - let snapshot_id = self.parse_value()?; + let snapshot_id = self.parse_value()?.value; return Ok(Statement::SetTransaction { modes: vec![], snapshot: Some(snapshot_id), @@ -11655,7 +11664,7 @@ impl<'a> Parser<'a> { } else if self.parse_keyword_with_tokens(Keyword::JSON_TABLE, &[Token::LParen]) { let json_expr = self.parse_expr()?; self.expect_token(&Token::Comma)?; - let json_path = self.parse_value()?; + let json_path = self.parse_value()?.value; self.expect_keyword_is(Keyword::COLUMNS)?; self.expect_token(&Token::LParen)?; let columns = self.parse_comma_separated(Parser::parse_json_table_column_def)?; @@ -11790,9 +11799,9 @@ impl<'a> Parser<'a> { let parenthesized = self.consume_token(&Token::LParen); let (quantity, bucket) = if parenthesized && self.parse_keyword(Keyword::BUCKET) { - let selected_bucket = self.parse_number_value()?; + let selected_bucket = self.parse_number_value()?.value; self.expect_keywords(&[Keyword::OUT, Keyword::OF])?; - let total = self.parse_number_value()?; + let total = self.parse_number_value()?.value; let on = if self.parse_keyword(Keyword::ON) { Some(self.parse_expr()?) } else { @@ -11810,8 +11819,9 @@ impl<'a> Parser<'a> { let value = match self.maybe_parse(|p| p.parse_expr())? { Some(num) => num, None => { - if let Token::Word(w) = self.next_token().token { - Expr::Value(Value::Placeholder(w.value)) + let next_token = self.next_token(); + if let Token::Word(w) = next_token.token { + Expr::Value(Value::Placeholder(w.value).with_span(next_token.span)) } else { return parser_err!( "Expecting number or byte length e.g. 100M", @@ -11869,7 +11879,7 @@ impl<'a> Parser<'a> { modifier: TableSampleSeedModifier, ) -> Result { self.expect_token(&Token::LParen)?; - let value = self.parse_number_value()?; + let value = self.parse_number_value()?.value; self.expect_token(&Token::RParen)?; Ok(TableSampleSeed { modifier, value }) } @@ -11880,7 +11890,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; let json_expr = self.parse_expr()?; let json_path = if self.consume_token(&Token::Comma) { - Some(self.parse_value()?) + Some(self.parse_value()?.value) } else { None }; @@ -12149,7 +12159,7 @@ impl<'a> Parser<'a> { pub fn parse_json_table_column_def(&mut self) -> Result { if self.parse_keyword(Keyword::NESTED) { let _has_path_keyword = self.parse_keyword(Keyword::PATH); - let path = self.parse_value()?; + let path = self.parse_value()?.value; self.expect_keyword_is(Keyword::COLUMNS)?; let columns = self.parse_parenthesized(|p| { p.parse_comma_separated(Self::parse_json_table_column_def) @@ -12167,7 +12177,7 @@ impl<'a> Parser<'a> { let r#type = self.parse_data_type()?; let exists = self.parse_keyword(Keyword::EXISTS); self.expect_keyword_is(Keyword::PATH)?; - let path = self.parse_value()?; + let path = self.parse_value()?.value; let mut on_empty = None; let mut on_error = None; while let Some(error_handling) = self.parse_json_table_column_error_handling()? { @@ -12224,7 +12234,7 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::ERROR) { JsonTableColumnErrorHandling::Error } else if self.parse_keyword(Keyword::DEFAULT) { - JsonTableColumnErrorHandling::Default(self.parse_value()?) + JsonTableColumnErrorHandling::Default(self.parse_value()?.value) } else { return Ok(None); }; @@ -13299,7 +13309,7 @@ impl<'a> Parser<'a> { if dialect_of!(self is GenericDialect | MySqlDialect) && self.parse_keyword(Keyword::SEPARATOR) { - clauses.push(FunctionArgumentClause::Separator(self.parse_value()?)); + clauses.push(FunctionArgumentClause::Separator(self.parse_value()?.value)); } if let Some(on_overflow) = self.parse_listagg_on_overflow()? { @@ -14183,7 +14193,7 @@ impl<'a> Parser<'a> { } fn parse_pragma_value(&mut self) -> Result { - match self.parse_value()? { + match self.parse_value()?.value { v @ Value::SingleQuotedString(_) => Ok(v), v @ Value::DoubleQuotedString(_) => Ok(v), v @ Value::Number(_, _) => Ok(v), @@ -14643,7 +14653,7 @@ impl<'a> Parser<'a> { fn maybe_parse_show_stmt_starts_with(&mut self) -> Result, ParserError> { if self.parse_keywords(&[Keyword::STARTS, Keyword::WITH]) { - Ok(Some(self.parse_value()?)) + Ok(Some(self.parse_value()?.value)) } else { Ok(None) } @@ -14659,7 +14669,7 @@ impl<'a> Parser<'a> { fn maybe_parse_show_stmt_from(&mut self) -> Result, ParserError> { if self.parse_keyword(Keyword::FROM) { - Ok(Some(self.parse_value()?)) + Ok(Some(self.parse_value()?.value)) } else { Ok(None) } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 55e354221..3037d4ae5 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -29,19 +29,19 @@ use test_utils::*; #[test] fn parse_literal_string() { let sql = concat!( - "SELECT ", - "'single', ", - r#""double", "#, - "'''triple-single''', ", - r#""""triple-double""", "#, - r#"'single\'escaped', "#, - r#"'''triple-single\'escaped''', "#, - r#"'''triple-single'unescaped''', "#, - r#""double\"escaped", "#, - r#""""triple-double\"escaped""", "#, - r#""""triple-double"unescaped""", "#, - r#""""triple-double'unescaped""", "#, - r#"'''triple-single"unescaped'''"#, + "SELECT ", // line 1, column 1 + "'single', ", // line 1, column 7 + r#""double", "#, // line 1, column 14 + "'''triple-single''', ", // line 1, column 22 + r#""""triple-double""", "#, // line 1, column 33 + r#"'single\'escaped', "#, // line 1, column 43 + r#"'''triple-single\'escaped''', "#, // line 1, column 55 + r#"'''triple-single'unescaped''', "#, // line 1, column 68 + r#""double\"escaped", "#, // line 1, column 83 + r#""""triple-double\"escaped""", "#, // line 1, column 92 + r#""""triple-double"unescaped""", "#, // line 1, column 105 + r#""""triple-double'unescaped""", "#, // line 1, column 118 + r#"'''triple-single"unescaped'''"#, // line 1, column 131 ); let dialect = TestedDialects::new_with_options( vec![Box::new(BigQueryDialect {})], @@ -50,63 +50,67 @@ fn parse_literal_string() { let select = dialect.verified_only_select(sql); assert_eq!(12, select.projection.len()); assert_eq!( - &Expr::Value(Value::SingleQuotedString("single".into())), + &Expr::Value(Value::SingleQuotedString("single".into()).with_empty_span()), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedString("double".into())), + &Expr::Value(Value::DoubleQuotedString("double".into()).with_empty_span()), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedString("triple-single".into())), + &Expr::Value(Value::TripleSingleQuotedString("triple-single".into()).with_empty_span()), expr_from_projection(&select.projection[2]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedString("triple-double".into())), + &Expr::Value(Value::TripleDoubleQuotedString("triple-double".into()).with_empty_span()), expr_from_projection(&select.projection[3]) ); assert_eq!( - &Expr::Value(Value::SingleQuotedString(r#"single\'escaped"#.into())), + &Expr::Value(Value::SingleQuotedString(r#"single\'escaped"#.into()).with_empty_span()), expr_from_projection(&select.projection[4]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedString( - r#"triple-single\'escaped"#.into() - )), + &Expr::Value( + Value::TripleSingleQuotedString(r#"triple-single\'escaped"#.into()).with_empty_span() + ), expr_from_projection(&select.projection[5]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedString( - r#"triple-single'unescaped"#.into() - )), + &Expr::Value( + Value::TripleSingleQuotedString(r#"triple-single'unescaped"#.into()).with_empty_span() + ), expr_from_projection(&select.projection[6]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedString(r#"double\"escaped"#.to_string())), + &Expr::Value(Value::DoubleQuotedString(r#"double\"escaped"#.to_string()).with_empty_span()), expr_from_projection(&select.projection[7]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedString( - r#"triple-double\"escaped"#.to_string() - )), + &Expr::Value( + Value::TripleDoubleQuotedString(r#"triple-double\"escaped"#.to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[8]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedString( - r#"triple-double"unescaped"#.to_string() - )), + &Expr::Value( + Value::TripleDoubleQuotedString(r#"triple-double"unescaped"#.to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[9]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedString( - r#"triple-double'unescaped"#.to_string() - )), + &Expr::Value( + Value::TripleDoubleQuotedString(r#"triple-double'unescaped"#.to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[10]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedString( - r#"triple-single"unescaped"#.to_string() - )), + &Expr::Value( + Value::TripleSingleQuotedString(r#"triple-single"unescaped"#.to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[11]) ); } @@ -114,48 +118,56 @@ fn parse_literal_string() { #[test] fn parse_byte_literal() { let sql = concat!( - "SELECT ", - "B'abc', ", - r#"B"abc", "#, - r#"B'f\(abc,(.*),def\)', "#, - r#"B"f\(abc,(.*),def\)", "#, - r#"B'''abc''', "#, - r#"B"""abc""""#, + "SELECT ", // line 1, column 1 + "B'abc', ", // line 1, column 8 + r#"B"abc", "#, // line 1, column 15 + r#"B'f\(abc,(.*),def\)', "#, // line 1, column 22 + r#"B"f\(abc,(.*),def\)", "#, // line 1, column 42 + r#"B'''abc''', "#, // line 1, column 62 + r#"B"""abc""""#, // line 1, column 74 ); let stmt = bigquery().verified_stmt(sql); if let Statement::Query(query) = stmt { if let SetExpr::Select(select) = *query.body { assert_eq!(6, select.projection.len()); assert_eq!( - &Expr::Value(Value::SingleQuotedByteStringLiteral("abc".to_string())), + &Expr::Value( + Value::SingleQuotedByteStringLiteral("abc".to_string()).with_empty_span() + ), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedByteStringLiteral("abc".to_string())), + &Expr::Value( + Value::DoubleQuotedByteStringLiteral("abc".to_string()).with_empty_span() + ), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::SingleQuotedByteStringLiteral( - r"f\(abc,(.*),def\)".to_string() - )), + &Expr::Value( + Value::SingleQuotedByteStringLiteral(r"f\(abc,(.*),def\)".to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[2]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedByteStringLiteral( - r"f\(abc,(.*),def\)".to_string() - )), + &Expr::Value( + Value::DoubleQuotedByteStringLiteral(r"f\(abc,(.*),def\)".to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[3]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedByteStringLiteral( - r"abc".to_string() - )), + &Expr::Value( + Value::TripleSingleQuotedByteStringLiteral(r"abc".to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[4]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedByteStringLiteral( - r"abc".to_string() - )), + &Expr::Value( + Value::TripleDoubleQuotedByteStringLiteral(r"abc".to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[5]) ); } @@ -172,48 +184,54 @@ fn parse_byte_literal() { #[test] fn parse_raw_literal() { let sql = concat!( - "SELECT ", - "R'abc', ", - r#"R"abc", "#, - r#"R'f\(abc,(.*),def\)', "#, - r#"R"f\(abc,(.*),def\)", "#, - r#"R'''abc''', "#, - r#"R"""abc""""#, + "SELECT ", // line 1, column 1 + "R'abc', ", // line 1, column 8 + r#"R"abc", "#, // line 1, column 15 + r#"R'f\(abc,(.*),def\)', "#, // line 1, column 22 + r#"R"f\(abc,(.*),def\)", "#, // line 1, column 42 + r#"R'''abc''', "#, // line 1, column 62 + r#"R"""abc""""#, // line 1, column 74 ); let stmt = bigquery().verified_stmt(sql); if let Statement::Query(query) = stmt { if let SetExpr::Select(select) = *query.body { assert_eq!(6, select.projection.len()); assert_eq!( - &Expr::Value(Value::SingleQuotedRawStringLiteral("abc".to_string())), + &Expr::Value( + Value::SingleQuotedRawStringLiteral("abc".to_string()).with_empty_span() + ), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedRawStringLiteral("abc".to_string())), + &Expr::Value( + Value::DoubleQuotedRawStringLiteral("abc".to_string()).with_empty_span() + ), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::SingleQuotedRawStringLiteral( - r"f\(abc,(.*),def\)".to_string() - )), + &Expr::Value( + Value::SingleQuotedRawStringLiteral(r"f\(abc,(.*),def\)".to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[2]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedRawStringLiteral( - r"f\(abc,(.*),def\)".to_string() - )), + &Expr::Value( + Value::DoubleQuotedRawStringLiteral(r"f\(abc,(.*),def\)".to_string()) + .with_empty_span() + ), expr_from_projection(&select.projection[3]) ); assert_eq!( - &Expr::Value(Value::TripleSingleQuotedRawStringLiteral( - r"abc".to_string() - )), + &Expr::Value( + Value::TripleSingleQuotedRawStringLiteral(r"abc".to_string()).with_empty_span() + ), expr_from_projection(&select.projection[4]) ); assert_eq!( - &Expr::Value(Value::TripleDoubleQuotedRawStringLiteral( - r"abc".to_string() - )), + &Expr::Value( + Value::TripleDoubleQuotedRawStringLiteral(r"abc".to_string()).with_empty_span() + ), expr_from_projection(&select.projection[5]) ); } @@ -336,7 +354,11 @@ fn parse_create_view_with_options() { data_type: None, options: Some(vec![ColumnOption::Options(vec![SqlOption::KeyValue { key: Ident::new("description"), - value: Expr::Value(Value::DoubleQuotedString("field age".to_string())), + value: Expr::Value( + Value::DoubleQuotedString("field age".to_string()).with_span( + Span::new(Location::new(1, 42), Location::new(1, 52)) + ) + ), }])]), }, ], @@ -356,9 +378,10 @@ fn parse_create_view_with_options() { assert_eq!( &SqlOption::KeyValue { key: Ident::new("description"), - value: Expr::Value(Value::DoubleQuotedString( - "a view that expires in 2 days".to_string() - )), + value: Expr::Value( + Value::DoubleQuotedString("a view that expires in 2 days".to_string()) + .with_empty_span() + ), }, &options[2], ); @@ -366,6 +389,7 @@ fn parse_create_view_with_options() { _ => unreachable!(), } } + #[test] fn parse_create_view_if_not_exists() { let sql = "CREATE VIEW IF NOT EXISTS mydataset.newview AS SELECT foo FROM bar"; @@ -481,9 +505,11 @@ fn parse_create_table_with_options() { name: None, option: ColumnOption::Options(vec![SqlOption::KeyValue { key: Ident::new("description"), - value: Expr::Value(Value::DoubleQuotedString( - "field x".to_string() - )), + value: Expr::Value( + Value::DoubleQuotedString("field x".to_string()).with_span( + Span::new(Location::new(1, 42), Location::new(1, 52)) + ) + ), },]) }, ] @@ -495,9 +521,11 @@ fn parse_create_table_with_options() { name: None, option: ColumnOption::Options(vec![SqlOption::KeyValue { key: Ident::new("description"), - value: Expr::Value(Value::DoubleQuotedString( - "field y".to_string() - )), + value: Expr::Value( + Value::DoubleQuotedString("field y".to_string()).with_span( + Span::new(Location::new(1, 42), Location::new(1, 52)) + ) + ), },]) }] }, @@ -514,13 +542,22 @@ fn parse_create_table_with_options() { Some(vec![ SqlOption::KeyValue { key: Ident::new("partition_expiration_days"), - value: Expr::Value(number("1")), + value: Expr::Value( + number("1").with_span(Span::new( + Location::new(1, 42), + Location::new(1, 43) + )) + ), }, SqlOption::KeyValue { key: Ident::new("description"), - value: Expr::Value(Value::DoubleQuotedString( - "table option description".to_string() - )), + value: Expr::Value( + Value::DoubleQuotedString("table option description".to_string()) + .with_span(Span::new( + Location::new(1, 42), + Location::new(1, 52) + )) + ), }, ]) ), @@ -628,23 +665,23 @@ fn parse_invalid_brackets() { fn parse_tuple_struct_literal() { // tuple syntax: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#tuple_syntax // syntax: (expr1, expr2 [, ... ]) - let sql = "SELECT (1, 2, 3), (1, 1.0, '123', true)"; + let sql = "SELECT (1, 2, 3), (1, 1.0, '123', true)"; // line 1, column 1 let select = bigquery().verified_only_select(sql); assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Tuple(vec![ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")), + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")), ]), expr_from_projection(&select.projection[0]) ); assert_eq!( &Expr::Tuple(vec![ - Expr::Value(number("1")), - Expr::Value(number("1.0")), - Expr::Value(Value::SingleQuotedString("123".into())), - Expr::Value(Value::Boolean(true)) + Expr::value(number("1")), + Expr::value(number("1.0")), + Expr::Value(Value::SingleQuotedString("123".into()).with_empty_span()), + Expr::Value(Value::Boolean(true).with_empty_span()) ]), expr_from_projection(&select.projection[1]) ); @@ -660,9 +697,9 @@ fn parse_typeless_struct_syntax() { assert_eq!( &Expr::Struct { values: vec![ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")), + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")), ], fields: Default::default() }, @@ -671,30 +708,35 @@ fn parse_typeless_struct_syntax() { assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("abc".into())),], + values: vec![Expr::Value( + Value::SingleQuotedString("abc".into()).with_empty_span() + )], fields: Default::default() }, expr_from_projection(&select.projection[1]) ); + assert_eq!( &Expr::Struct { values: vec![ - Expr::Value(number("1")), + Expr::value(number("1")), Expr::CompoundIdentifier(vec![Ident::from("t"), Ident::from("str_col")]), ], fields: Default::default() }, expr_from_projection(&select.projection[2]) ); + assert_eq!( &Expr::Struct { values: vec![ Expr::Named { - expr: Expr::Value(number("1")).into(), + expr: Expr::value(number("1")).into(), name: Ident::from("a") }, Expr::Named { - expr: Expr::Value(Value::SingleQuotedString("abc".into())).into(), + expr: Expr::Value(Value::SingleQuotedString("abc".into()).with_empty_span()) + .into(), name: Ident::from("b") }, ], @@ -702,6 +744,7 @@ fn parse_typeless_struct_syntax() { }, expr_from_projection(&select.projection[3]) ); + assert_eq!( &Expr::Struct { values: vec![Expr::Named { @@ -724,7 +767,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(3, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5")),], + values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: None, field_type: DataType::Int64, @@ -735,7 +778,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!( &Expr::Struct { values: vec![ - Expr::Value(number("1")), + Expr::value(number("1")), Expr::CompoundIdentifier(vec![ Ident { value: "t".into(), @@ -776,7 +819,7 @@ fn parse_typed_struct_syntax_bigquery() { value: "nested_col".into(), quote_style: None, span: Span::empty(), - }),], + })], fields: vec![ StructField { field_name: Some("arr".into()), @@ -808,7 +851,7 @@ fn parse_typed_struct_syntax_bigquery() { value: "nested_col".into(), quote_style: None, span: Span::empty(), - }),], + })], fields: vec![ StructField { field_name: Some("x".into()), @@ -833,7 +876,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::Boolean(true)),], + values: vec![Expr::Value(Value::Boolean(true).with_empty_span())], fields: vec![StructField { field_name: None, field_type: DataType::Bool @@ -843,9 +886,9 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedByteStringLiteral( - "abc".into() - )),], + values: vec![Expr::Value( + Value::SingleQuotedByteStringLiteral("abc".into()).with_empty_span() + )], fields: vec![StructField { field_name: None, field_type: DataType::Bytes(Some(42)) @@ -859,7 +902,9 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(4, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::DoubleQuotedString("2011-05-05".into())),], + values: vec![Expr::Value( + Value::DoubleQuotedString("2011-05-05".into()).with_empty_span() + )], fields: vec![StructField { field_name: None, field_type: DataType::Date @@ -872,7 +917,7 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Datetime(None) @@ -882,7 +927,7 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5.0")),], + values: vec![Expr::value(number("5.0"))], fields: vec![StructField { field_name: None, field_type: DataType::Float64 @@ -892,7 +937,7 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("1")),], + values: vec![Expr::value(number("1"))], fields: vec![StructField { field_name: None, field_type: DataType::Int64 @@ -907,12 +952,14 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!( &Expr::Struct { values: vec![Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString("2".into()))), + value: Box::new(Expr::Value( + Value::SingleQuotedString("2".into()).with_empty_span() + )), leading_field: Some(DateTimeField::Hour), leading_precision: None, last_field: None, fractional_seconds_precision: None - }),], + })], fields: vec![StructField { field_name: None, field_type: DataType::Interval @@ -927,7 +974,7 @@ fn parse_typed_struct_syntax_bigquery() { value: Value::SingleQuotedString( r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() ) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::JSON @@ -941,7 +988,9 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(3, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::DoubleQuotedString("foo".into())),], + values: vec![Expr::Value( + Value::DoubleQuotedString("foo".into()).with_empty_span() + )], fields: vec![StructField { field_name: None, field_type: DataType::String(Some(42)) @@ -954,7 +1003,7 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Timestamp(None, TimezoneInfo::None) @@ -968,7 +1017,7 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), value: Value::SingleQuotedString("15:30:00".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Time(None, TimezoneInfo::None) @@ -985,7 +1034,7 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), value: Value::SingleQuotedString("1".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Numeric(ExactNumberInfo::None) @@ -998,7 +1047,7 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: Value::SingleQuotedString("1".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::BigNumeric(ExactNumberInfo::None) @@ -1013,7 +1062,7 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(1, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("1")), Expr::Value(number("2")),], + values: vec![Expr::value(number("1")), Expr::value(number("2")),], fields: vec![ StructField { field_name: Some("key".into()), @@ -1039,7 +1088,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(3, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5")),], + values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: None, field_type: DataType::Int64, @@ -1050,7 +1099,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!( &Expr::Struct { values: vec![ - Expr::Value(number("1")), + Expr::value(number("1")), Expr::CompoundIdentifier(vec![ Ident { value: "t".into(), @@ -1085,34 +1134,6 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }, expr_from_projection(&select.projection[1]) ); - assert_eq!( - &Expr::Struct { - values: vec![Expr::Identifier(Ident { - value: "nested_col".into(), - quote_style: None, - span: Span::empty(), - }),], - fields: vec![ - StructField { - field_name: Some("arr".into()), - field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( - DataType::Float64 - ))) - }, - StructField { - field_name: Some("str".into()), - field_type: DataType::Struct( - vec![StructField { - field_name: None, - field_type: DataType::Bool - }], - StructBracketKind::AngleBrackets - ) - }, - ] - }, - expr_from_projection(&select.projection[2]) - ); let sql = r#"SELECT STRUCT>(nested_col)"#; let select = bigquery_and_generic().verified_only_select(sql); @@ -1123,7 +1144,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { value: "nested_col".into(), quote_style: None, span: Span::empty(), - }),], + })], fields: vec![ StructField { field_name: Some("x".into()), @@ -1148,7 +1169,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::Boolean(true)),], + values: vec![Expr::Value(Value::Boolean(true).with_empty_span())], fields: vec![StructField { field_name: None, field_type: DataType::Bool @@ -1158,9 +1179,9 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedByteStringLiteral( - "abc".into() - )),], + values: vec![Expr::Value( + Value::SingleQuotedByteStringLiteral("abc".into()).with_empty_span() + )], fields: vec![StructField { field_name: None, field_type: DataType::Bytes(Some(42)) @@ -1174,7 +1195,9 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(4, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("2011-05-05".into())),], + values: vec![Expr::Value( + Value::SingleQuotedString("2011-05-05".into()).with_empty_span() + )], fields: vec![StructField { field_name: None, field_type: DataType::Date @@ -1187,7 +1210,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Datetime(None) @@ -1197,7 +1220,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5.0")),], + values: vec![Expr::value(number("5.0"))], fields: vec![StructField { field_name: None, field_type: DataType::Float64 @@ -1207,7 +1230,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("1")),], + values: vec![Expr::value(number("1"))], fields: vec![StructField { field_name: None, field_type: DataType::Int64 @@ -1222,12 +1245,14 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!( &Expr::Struct { values: vec![Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString("1".into()))), + value: Box::new(Expr::Value( + Value::SingleQuotedString("1".into()).with_empty_span() + )), leading_field: Some(DateTimeField::Month), leading_precision: None, last_field: None, fractional_seconds_precision: None - }),], + })], fields: vec![StructField { field_name: None, field_type: DataType::Interval @@ -1242,7 +1267,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { value: Value::SingleQuotedString( r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() ) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::JSON @@ -1256,7 +1281,9 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(3, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("foo".into())),], + values: vec![Expr::Value( + Value::SingleQuotedString("foo".into()).with_empty_span() + )], fields: vec![StructField { field_name: None, field_type: DataType::String(Some(42)) @@ -1269,7 +1296,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Timestamp(None, TimezoneInfo::None) @@ -1283,7 +1310,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), value: Value::SingleQuotedString("15:30:00".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Time(None, TimezoneInfo::None) @@ -1300,7 +1327,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), value: Value::SingleQuotedString("1".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::Numeric(ExactNumberInfo::None) @@ -1313,7 +1340,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: Value::SingleQuotedString("1".into()) - },], + }], fields: vec![StructField { field_name: None, field_type: DataType::BigNumeric(ExactNumberInfo::None) @@ -1325,12 +1352,12 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { #[test] fn parse_typed_struct_with_field_name_bigquery() { - let sql = r#"SELECT STRUCT(5), STRUCT("foo")"#; + let sql = r#"SELECT STRUCT(5), STRUCT("foo")"#; // line 1, column 1 let select = bigquery().verified_only_select(sql); assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5")),], + values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: Some(Ident::from("x")), field_type: DataType::Int64 @@ -1340,7 +1367,9 @@ fn parse_typed_struct_with_field_name_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::DoubleQuotedString("foo".into())),], + values: vec![Expr::Value( + Value::DoubleQuotedString("foo".into()).with_empty_span() + )], fields: vec![StructField { field_name: Some(Ident::from("y")), field_type: DataType::String(None) @@ -1349,12 +1378,12 @@ fn parse_typed_struct_with_field_name_bigquery() { expr_from_projection(&select.projection[1]) ); - let sql = r#"SELECT STRUCT(5, 5)"#; + let sql = r#"SELECT STRUCT(5, 5)"#; // line 1, column 1 let select = bigquery().verified_only_select(sql); assert_eq!(1, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5")), Expr::Value(number("5")),], + values: vec![Expr::value(number("5")), Expr::value(number("5")),], fields: vec![ StructField { field_name: Some(Ident::from("x")), @@ -1372,12 +1401,12 @@ fn parse_typed_struct_with_field_name_bigquery() { #[test] fn parse_typed_struct_with_field_name_bigquery_and_generic() { - let sql = r#"SELECT STRUCT(5), STRUCT('foo')"#; + let sql = r#"SELECT STRUCT(5), STRUCT('foo')"#; // line 1, column 1 let select = bigquery().verified_only_select(sql); assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5")),], + values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: Some(Ident::from("x")), field_type: DataType::Int64 @@ -1387,7 +1416,9 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(Value::SingleQuotedString("foo".into())),], + values: vec![Expr::Value( + Value::SingleQuotedString("foo".into()).with_empty_span() + )], fields: vec![StructField { field_name: Some(Ident::from("y")), field_type: DataType::String(None) @@ -1396,12 +1427,12 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { expr_from_projection(&select.projection[1]) ); - let sql = r#"SELECT STRUCT(5, 5)"#; + let sql = r#"SELECT STRUCT(5, 5)"#; // line 1, column 1 let select = bigquery_and_generic().verified_only_select(sql); assert_eq!(1, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::Value(number("5")), Expr::Value(number("5")),], + values: vec![Expr::value(number("5")), Expr::value(number("5")),], fields: vec![ StructField { field_name: Some(Ident::from("x")), @@ -1609,7 +1640,7 @@ fn parse_hyphenated_table_identifiers() { #[test] fn parse_table_time_travel() { let version = "2023-08-18 23:08:18".to_string(); - let sql = format!("SELECT 1 FROM t1 FOR SYSTEM_TIME AS OF '{version}'"); + let sql = format!("SELECT 1 FROM t1 FOR SYSTEM_TIME AS OF '{version}'"); // line 1, column 1 let select = bigquery().verified_only_select(&sql); assert_eq!( select.from, @@ -1620,7 +1651,7 @@ fn parse_table_time_travel() { args: None, with_hints: vec![], version: Some(TableVersion::ForSystemTimeAsOf(Expr::Value( - Value::SingleQuotedString(version) + Value::SingleQuotedString(version).with_empty_span() ))), partitions: vec![], with_ordinality: false, @@ -1689,18 +1720,18 @@ fn parse_merge() { columns: vec![Ident::new("product"), Ident::new("quantity")], kind: MergeInsertKind::Values(Values { explicit_row: false, - rows: vec![vec![Expr::Value(number("1")), Expr::Value(number("2"))]], + rows: vec![vec![Expr::value(number("1")), Expr::value(number("2"))]], }), }); let update_action = MergeAction::Update { assignments: vec![ Assignment { target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new("a")])), - value: Expr::Value(number("1")), + value: Expr::value(number("1")), }, Assignment { target: AssignmentTarget::ColumnName(ObjectName::from(vec![Ident::new("b")])), - value: Expr::Value(number("2")), + value: Expr::value(number("2")), }, ], }; @@ -1749,17 +1780,17 @@ fn parse_merge() { }, source ); - assert_eq!(Expr::Value(Value::Boolean(false)), *on); + assert_eq!(Expr::Value(Value::Boolean(false).with_empty_span()), *on); assert_eq!( vec![ MergeClause { clause_kind: MergeClauseKind::NotMatched, - predicate: Some(Expr::Value(number("1"))), + predicate: Some(Expr::value(number("1"))), action: insert_action.clone(), }, MergeClause { clause_kind: MergeClauseKind::NotMatchedByTarget, - predicate: Some(Expr::Value(number("1"))), + predicate: Some(Expr::value(number("1"))), action: insert_action.clone(), }, MergeClause { @@ -1769,7 +1800,7 @@ fn parse_merge() { }, MergeClause { clause_kind: MergeClauseKind::NotMatchedBySource, - predicate: Some(Expr::Value(number("2"))), + predicate: Some(Expr::value(number("2"))), action: MergeAction::Delete }, MergeClause { @@ -1779,12 +1810,12 @@ fn parse_merge() { }, MergeClause { clause_kind: MergeClauseKind::NotMatchedBySource, - predicate: Some(Expr::Value(number("1"))), + predicate: Some(Expr::value(number("1"))), action: update_action.clone(), }, MergeClause { clause_kind: MergeClauseKind::NotMatched, - predicate: Some(Expr::Value(number("1"))), + predicate: Some(Expr::value(number("1"))), action: MergeAction::Insert(MergeInsertExpr { columns: vec![Ident::new("product"), Ident::new("quantity"),], kind: MergeInsertKind::Row, @@ -1800,7 +1831,7 @@ fn parse_merge() { }, MergeClause { clause_kind: MergeClauseKind::NotMatched, - predicate: Some(Expr::Value(number("1"))), + predicate: Some(Expr::value(number("1"))), action: MergeAction::Insert(MergeInsertExpr { columns: vec![], kind: MergeInsertKind::Row @@ -1816,7 +1847,7 @@ fn parse_merge() { }, MergeClause { clause_kind: MergeClauseKind::Matched, - predicate: Some(Expr::Value(number("1"))), + predicate: Some(Expr::value(number("1"))), action: MergeAction::Delete, }, MergeClause { @@ -1832,7 +1863,7 @@ fn parse_merge() { kind: MergeInsertKind::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(number("1")), + Expr::value(number("1")), Expr::Identifier(Ident::new("DEFAULT")), ]] }) @@ -1846,7 +1877,7 @@ fn parse_merge() { kind: MergeInsertKind::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(number("1")), + Expr::value(number("1")), Expr::Identifier(Ident::new("DEFAULT")), ]] }) @@ -1962,7 +1993,7 @@ fn parse_array_agg_func() { fn parse_big_query_declare() { for (sql, expected_names, expected_data_type, expected_assigned_expr) in [ ( - "DECLARE x INT64", + "DECLARE x INT64", // line 1, column 1 vec![Ident::new("x")], Some(DataType::Int64), None, @@ -1971,25 +2002,25 @@ fn parse_big_query_declare() { "DECLARE x INT64 DEFAULT 42", vec![Ident::new("x")], Some(DataType::Int64), - Some(DeclareAssignment::Default(Box::new(Expr::Value(number( - "42", - ))))), + Some(DeclareAssignment::Default(Box::new(Expr::Value( + number("42").with_empty_span(), + )))), ), ( "DECLARE x, y, z INT64 DEFAULT 42", vec![Ident::new("x"), Ident::new("y"), Ident::new("z")], Some(DataType::Int64), - Some(DeclareAssignment::Default(Box::new(Expr::Value(number( - "42", - ))))), + Some(DeclareAssignment::Default(Box::new(Expr::Value( + number("42").with_empty_span(), + )))), ), ( "DECLARE x DEFAULT 42", vec![Ident::new("x")], None, - Some(DeclareAssignment::Default(Box::new(Expr::Value(number( - "42", - ))))), + Some(DeclareAssignment::Default(Box::new(Expr::Value( + number("42").with_empty_span(), + )))), ), ] { match bigquery().verified_stmt(sql) { @@ -2047,7 +2078,7 @@ fn parse_map_access_expr() { AccessExpr::Subscript(Subscript::Index { index: Expr::UnaryOp { op: UnaryOperator::Minus, - expr: Expr::Value(number("1")).into(), + expr: Expr::value(number("1")).into(), }, }), AccessExpr::Subscript(Subscript::Index { @@ -2060,7 +2091,7 @@ fn parse_map_access_expr() { args: FunctionArguments::List(FunctionArgumentList { duplicate_treatment: None, args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - number("2"), + number("2").with_empty_span(), )))], clauses: vec![], }), @@ -2111,12 +2142,12 @@ fn test_bigquery_create_function() { ]), args: Some(vec![OperateFunctionArg::with_name("x", DataType::Float64),]), return_type: Some(DataType::Float64), - function_body: Some(CreateFunctionBody::AsAfterOptions(Expr::Value(number( - "42" - )))), + function_body: Some(CreateFunctionBody::AsAfterOptions(Expr::Value( + number("42").with_empty_span() + ))), options: Some(vec![SqlOption::KeyValue { key: Ident::new("x"), - value: Expr::Value(Value::SingleQuotedString("y".into())), + value: Expr::Value(Value::SingleQuotedString("y".into()).with_empty_span()), }]), behavior: None, using: None, @@ -2235,10 +2266,14 @@ fn test_bigquery_trim() { let select = bigquery().verified_only_select(sql_only_select); assert_eq!( &Expr::Trim { - expr: Box::new(Expr::Value(Value::SingleQuotedString("xyz".to_owned()))), + expr: Box::new(Expr::Value( + Value::SingleQuotedString("xyz".to_owned()).with_empty_span() + )), trim_where: None, trim_what: None, - trim_characters: Some(vec![Expr::Value(Value::SingleQuotedString("a".to_owned()))]), + trim_characters: Some(vec![Expr::Value( + Value::SingleQuotedString("a".to_owned()).with_empty_span() + )]), }, expr_from_projection(only(&select.projection)) ); diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 99e76d45e..98a1aef62 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -55,7 +55,10 @@ fn parse_map_access_expr() { "indexOf", [ Expr::Identifier(Ident::new("string_names")), - Expr::Value(Value::SingleQuotedString("endpoint".to_string())) + Expr::Value( + (Value::SingleQuotedString("endpoint".to_string())) + .with_empty_span() + ) ] ), })], @@ -71,7 +74,9 @@ fn parse_map_access_expr() { left: Box::new(BinaryOp { left: Box::new(Identifier(Ident::new("id"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString("test".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("test".to_string())).with_empty_span() + )), }), op: BinaryOperator::And, right: Box::new(BinaryOp { @@ -82,13 +87,18 @@ fn parse_map_access_expr() { "indexOf", [ Expr::Identifier(Ident::new("string_name")), - Expr::Value(Value::SingleQuotedString("app".to_string())) + Expr::Value( + (Value::SingleQuotedString("app".to_string())) + .with_empty_span() + ) ] ), })], }), op: BinaryOperator::NotEq, - right: Box::new(Expr::Value(Value::SingleQuotedString("foo".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("foo".to_string())).with_empty_span() + )), }), }), group_by: GroupByExpr::Expressions(vec![], vec![]), @@ -114,8 +124,8 @@ fn parse_array_expr() { assert_eq!( &Expr::Array(Array { elem: vec![ - Expr::Value(Value::SingleQuotedString("1".to_string())), - Expr::Value(Value::SingleQuotedString("2".to_string())), + Expr::Value((Value::SingleQuotedString("1".to_string())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("2".to_string())).with_empty_span()), ], named: false, }), @@ -1016,17 +1026,15 @@ fn parse_select_parametric_function() { assert_eq!(parameters.args.len(), 2); assert_eq!( parameters.args[0], - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(Value::Number( - "0.5".parse().unwrap(), - false - )))) + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (Value::Number("0.5".parse().unwrap(), false)).with_empty_span() + ))) ); assert_eq!( parameters.args[1], - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(Value::Number( - "0.6".parse().unwrap(), - false - )))) + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (Value::Number("0.6".parse().unwrap(), false)).with_empty_span() + ))) ); } _ => unreachable!(), @@ -1078,9 +1086,9 @@ fn parse_select_order_by_with_fill_interpolate() { nulls_first: Some(true), }, with_fill: Some(WithFill { - from: Some(Expr::Value(number("10"))), - to: Some(Expr::Value(number("20"))), - step: Some(Expr::Value(number("2"))), + from: Some(Expr::value(number("10"))), + to: Some(Expr::value(number("20"))), + step: Some(Expr::value(number("2"))), }), }, OrderByExpr { @@ -1090,9 +1098,9 @@ fn parse_select_order_by_with_fill_interpolate() { nulls_first: Some(false), }, with_fill: Some(WithFill { - from: Some(Expr::Value(number("30"))), - to: Some(Expr::Value(number("40"))), - step: Some(Expr::Value(number("3"))), + from: Some(Expr::value(number("30"))), + to: Some(Expr::value(number("40"))), + step: Some(Expr::value(number("3"))), }), }, ]), @@ -1102,14 +1110,14 @@ fn parse_select_order_by_with_fill_interpolate() { expr: Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("col1"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), }]) }) }, select.order_by.expect("ORDER BY expected") ); - assert_eq!(Some(Expr::Value(number("2"))), select.limit); + assert_eq!(Some(Expr::value(number("2"))), select.limit); } #[test] @@ -1150,9 +1158,9 @@ fn parse_with_fill() { let select = clickhouse().verified_query(sql); assert_eq!( Some(WithFill { - from: Some(Expr::Value(number("10"))), - to: Some(Expr::Value(number("20"))), - step: Some(Expr::Value(number("2"))), + from: Some(Expr::value(number("10"))), + to: Some(Expr::value(number("20"))), + step: Some(Expr::value(number("2"))), }) .as_ref(), match select.order_by.expect("ORDER BY expected").kind { @@ -1193,7 +1201,7 @@ fn parse_interpolate_body_with_columns() { expr: Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("col1"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), }, InterpolateExpr { @@ -1205,7 +1213,7 @@ fn parse_interpolate_body_with_columns() { expr: Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("col4"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("4"))), + right: Box::new(Expr::value(number("4"))), }), }, ]) @@ -1260,7 +1268,9 @@ fn test_prewhere() { Some(&BinaryOp { left: Box::new(Identifier(Ident::new("x"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), }) ); let selection = query.as_ref().body.as_select().unwrap().selection.as_ref(); @@ -1269,7 +1279,9 @@ fn test_prewhere() { Some(&BinaryOp { left: Box::new(Identifier(Ident::new("y"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("2".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("2".parse().unwrap(), false)).with_empty_span() + )), }) ); } @@ -1285,13 +1297,17 @@ fn test_prewhere() { left: Box::new(BinaryOp { left: Box::new(Identifier(Ident::new("x"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), }), op: BinaryOperator::And, right: Box::new(BinaryOp { left: Box::new(Identifier(Ident::new("y"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("2".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("2".parse().unwrap(), false)).with_empty_span() + )), }), }) ); @@ -1399,10 +1415,9 @@ fn parse_create_table_on_commit_and_as_query() { assert_eq!(on_commit, Some(OnCommit::PreserveRows)); assert_eq!( query.unwrap().body.as_select().unwrap().projection, - vec![UnnamedExpr(Expr::Value(Value::Number( - "1".parse().unwrap(), - false - )))] + vec![UnnamedExpr(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + ))] ); } _ => unreachable!(), @@ -1415,9 +1430,9 @@ fn parse_freeze_and_unfreeze_partition() { for operation_name in &["FREEZE", "UNFREEZE"] { let sql = format!("ALTER TABLE t {operation_name} PARTITION '2024-08-14'"); - let expected_partition = Partition::Expr(Expr::Value(Value::SingleQuotedString( - "2024-08-14".to_string(), - ))); + let expected_partition = Partition::Expr(Expr::Value( + Value::SingleQuotedString("2024-08-14".to_string()).with_empty_span(), + )); match clickhouse_and_generic().verified_stmt(&sql) { Statement::AlterTable { operations, .. } => { assert_eq!(operations.len(), 1); @@ -1445,9 +1460,9 @@ fn parse_freeze_and_unfreeze_partition() { match clickhouse_and_generic().verified_stmt(&sql) { Statement::AlterTable { operations, .. } => { assert_eq!(operations.len(), 1); - let expected_partition = Partition::Expr(Expr::Value(Value::SingleQuotedString( - "2024-08-14".to_string(), - ))); + let expected_partition = Partition::Expr(Expr::Value( + Value::SingleQuotedString("2024-08-14".to_string()).with_empty_span(), + )); let expected_operation = if operation_name == &"FREEZE" { AlterTableOperation::FreezePartition { partition: expected_partition, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 0072baf75..0a68d31e8 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -69,7 +69,9 @@ fn parse_numeric_literal_underscore() { assert_eq!( select.projection, - vec![UnnamedExpr(Expr::Value(number("10_000")))] + vec![UnnamedExpr(Expr::Value( + (number("10_000")).with_empty_span() + ))] ); } @@ -93,9 +95,9 @@ fn parse_function_object_name() { #[test] fn parse_insert_values() { let row = vec![ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")), + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")), ]; let rows1 = vec![row.clone()]; let rows2 = vec![row.clone(), row]; @@ -384,15 +386,15 @@ fn parse_update() { vec![ Assignment { target: AssignmentTarget::ColumnName(ObjectName::from(vec!["a".into()])), - value: Expr::Value(number("1")), + value: Expr::value(number("1")), }, Assignment { target: AssignmentTarget::ColumnName(ObjectName::from(vec!["b".into()])), - value: Expr::Value(number("2")), + value: Expr::value(number("2")), }, Assignment { target: AssignmentTarget::ColumnName(ObjectName::from(vec!["c".into()])), - value: Expr::Value(number("3")), + value: Expr::value(number("3")), }, ] ); @@ -556,7 +558,9 @@ fn parse_update_with_table_alias() { Ident::new("u"), Ident::new("username") ])), - value: Expr::Value(Value::SingleQuotedString("new_user".to_string())), + value: Expr::Value( + (Value::SingleQuotedString("new_user".to_string())).with_empty_span() + ), }], assignments ); @@ -567,9 +571,9 @@ fn parse_update_with_table_alias() { Ident::new("username"), ])), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "old_user".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("old_user".to_string())).with_empty_span() + )), }), selection ); @@ -797,7 +801,7 @@ fn parse_where_delete_statement() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("name"))), op: Eq, - right: Box::new(Expr::Value(number("5"))), + right: Box::new(Expr::value(number("5"))), }, selection.unwrap(), ); @@ -896,7 +900,7 @@ fn parse_simple_select() { assert!(select.distinct.is_none()); assert_eq!(3, select.projection.len()); let select = verified_query(sql); - assert_eq!(Some(Expr::Value(number("5"))), select.limit); + assert_eq!(Some(Expr::value(number("5"))), select.limit); } #[test] @@ -908,10 +912,10 @@ fn parse_limit() { fn parse_limit_is_not_an_alias() { // In dialects supporting LIMIT it shouldn't be parsed as a table alias let ast = verified_query("SELECT id FROM customer LIMIT 1"); - assert_eq!(Some(Expr::Value(number("1"))), ast.limit); + assert_eq!(Some(Expr::value(number("1"))), ast.limit); let ast = verified_query("SELECT 1 LIMIT 5"); - assert_eq!(Some(Expr::Value(number("5"))), ast.limit); + assert_eq!(Some(Expr::value(number("5"))), ast.limit); } #[test] @@ -1129,7 +1133,7 @@ fn parse_column_aliases() { } = only(&select.projection) { assert_eq!(&BinaryOperator::Plus, op); - assert_eq!(&Expr::Value(number("1")), right.as_ref()); + assert_eq!(&Expr::value(number("1")), right.as_ref()); assert_eq!(&Ident::new("newname"), alias); } else { panic!("Expected: ExprWithAlias") @@ -1356,7 +1360,7 @@ fn parse_null_in_select() { let sql = "SELECT NULL"; let select = verified_only_select(sql); assert_eq!( - &Expr::Value(Value::Null), + &Expr::Value((Value::Null).with_empty_span()), expr_from_projection(only(&select.projection)), ); } @@ -1392,18 +1396,18 @@ fn parse_exponent_in_select() -> Result<(), ParserError> { assert_eq!( &vec![ - SelectItem::UnnamedExpr(Expr::Value(number("10e-20"))), - SelectItem::UnnamedExpr(Expr::Value(number("1e3"))), - SelectItem::UnnamedExpr(Expr::Value(number("1e+3"))), + SelectItem::UnnamedExpr(Expr::Value((number("10e-20")).with_empty_span())), + SelectItem::UnnamedExpr(Expr::value(number("1e3"))), + SelectItem::UnnamedExpr(Expr::Value((number("1e+3")).with_empty_span())), SelectItem::ExprWithAlias { - expr: Expr::Value(number("1e3")), + expr: Expr::value(number("1e3")), alias: Ident::new("a") }, SelectItem::ExprWithAlias { - expr: Expr::Value(number("1")), + expr: Expr::value(number("1")), alias: Ident::new("e") }, - SelectItem::UnnamedExpr(Expr::Value(number("0.5e2"))), + SelectItem::UnnamedExpr(Expr::value(number("0.5e2"))), ], &select.projection ); @@ -1437,9 +1441,9 @@ fn parse_escaped_single_quote_string_predicate_with_escape() { Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("salary"))), op: NotEq, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "Jim's salary".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("Jim's salary".to_string())).with_empty_span() + )), }), ast.selection, ); @@ -1463,9 +1467,9 @@ fn parse_escaped_single_quote_string_predicate_with_no_escape() { Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("salary"))), op: NotEq, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "Jim''s salary".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("Jim''s salary".to_string())).with_empty_span() + )), }), ast.selection, ); @@ -1478,11 +1482,14 @@ fn parse_number() { #[cfg(feature = "bigdecimal")] assert_eq!( expr, - Expr::Value(Value::Number(bigdecimal::BigDecimal::from(1), false)) + Expr::Value((Value::Number(bigdecimal::BigDecimal::from(1), false)).with_empty_span()) ); #[cfg(not(feature = "bigdecimal"))] - assert_eq!(expr, Expr::Value(Value::Number("1.0".into(), false))); + assert_eq!( + expr, + Expr::Value((Value::Number("1.0".into(), false)).with_empty_span()) + ); } #[test] @@ -1634,15 +1641,15 @@ fn parse_json_object() { }) => assert_eq!( &[ FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString("name".into())), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "value".into() - ))), + name: Expr::Value((Value::SingleQuotedString("name".into())).with_empty_span()), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("value".into())).with_empty_span() + )), operator: FunctionArgOperator::Colon }, FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString("type".into())), - arg: FunctionArgExpr::Expr(Expr::Value(number("1"))), + name: Expr::Value((Value::SingleQuotedString("type".into())).with_empty_span()), + arg: FunctionArgExpr::Expr(Expr::value(number("1"))), operator: FunctionArgOperator::Colon } ], @@ -1660,15 +1667,19 @@ fn parse_json_object() { assert_eq!( &[ FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString("name".into())), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "value".into() - ))), + name: Expr::Value( + (Value::SingleQuotedString("name".into())).with_empty_span() + ), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("value".into())).with_empty_span() + )), operator: FunctionArgOperator::Colon }, FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString("type".into())), - arg: FunctionArgExpr::Expr(Expr::Value(Value::Null)), + name: Expr::Value( + (Value::SingleQuotedString("type".into())).with_empty_span() + ), + arg: FunctionArgExpr::Expr(Expr::Value((Value::Null).with_empty_span())), operator: FunctionArgOperator::Colon } ], @@ -1725,10 +1736,10 @@ fn parse_json_object() { }) => { assert_eq!( &FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString("name".into())), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "value".into() - ))), + name: Expr::Value((Value::SingleQuotedString("name".into())).with_empty_span()), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("value".into())).with_empty_span() + )), operator: FunctionArgOperator::Colon }, &args[0] @@ -1736,7 +1747,10 @@ fn parse_json_object() { assert!(matches!( args[1], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::Function(_)), operator: FunctionArgOperator::Colon } @@ -1760,10 +1774,10 @@ fn parse_json_object() { }) => { assert_eq!( &FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString("name".into())), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "value".into() - ))), + name: Expr::Value((Value::SingleQuotedString("name".into())).with_empty_span()), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("value".into())).with_empty_span() + )), operator: FunctionArgOperator::Colon }, &args[0] @@ -1771,7 +1785,10 @@ fn parse_json_object() { assert!(matches!( args[1], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::Function(_)), operator: FunctionArgOperator::Colon } @@ -1880,9 +1897,9 @@ fn parse_not_precedence() { Expr::UnaryOp { op: UnaryOperator::Not, expr: Box::new(Expr::Between { - expr: Box::new(Expr::Value(number("1"))), - low: Box::new(Expr::Value(number("1"))), - high: Box::new(Expr::Value(number("2"))), + expr: Box::new(Expr::value(number("1"))), + low: Box::new(Expr::value(number("1"))), + high: Box::new(Expr::value(number("2"))), negated: true, }), }, @@ -1895,9 +1912,13 @@ fn parse_not_precedence() { Expr::UnaryOp { op: UnaryOperator::Not, expr: Box::new(Expr::Like { - expr: Box::new(Expr::Value(Value::SingleQuotedString("a".into()))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("a".into())).with_empty_span() + )), negated: true, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("b".into()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("b".into())).with_empty_span() + )), escape_char: None, any: false, }), @@ -1912,7 +1933,9 @@ fn parse_not_precedence() { op: UnaryOperator::Not, expr: Box::new(Expr::InList { expr: Box::new(Expr::Identifier("a".into())), - list: vec![Expr::Value(Value::SingleQuotedString("a".into()))], + list: vec![Expr::Value( + (Value::SingleQuotedString("a".into())).with_empty_span() + )], negated: true, }), }, @@ -1932,7 +1955,7 @@ fn parse_null_like() { expr: Box::new(Expr::Identifier(Ident::new("column1"))), any: false, negated: false, - pattern: Box::new(Expr::Value(Value::Null)), + pattern: Box::new(Expr::Value((Value::Null).with_empty_span())), escape_char: None, }, alias: Ident { @@ -1946,7 +1969,7 @@ fn parse_null_like() { assert_eq!( SelectItem::ExprWithAlias { expr: Expr::Like { - expr: Box::new(Expr::Value(Value::Null)), + expr: Box::new(Expr::Value((Value::Null).with_empty_span())), any: false, negated: false, pattern: Box::new(Expr::Identifier(Ident::new("column1"))), @@ -1974,7 +1997,9 @@ fn parse_ilike() { Expr::ILike { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: None, any: false, }, @@ -1991,7 +2016,9 @@ fn parse_ilike() { Expr::ILike { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: Some('^'.to_string()), any: false, }, @@ -2009,7 +2036,9 @@ fn parse_ilike() { Expr::IsNull(Box::new(Expr::ILike { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: None, any: false, })), @@ -2032,7 +2061,9 @@ fn parse_like() { Expr::Like { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: None, any: false, }, @@ -2049,7 +2080,9 @@ fn parse_like() { Expr::Like { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: Some('^'.to_string()), any: false, }, @@ -2067,7 +2100,9 @@ fn parse_like() { Expr::IsNull(Box::new(Expr::Like { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: None, any: false, })), @@ -2090,7 +2125,9 @@ fn parse_similar_to() { Expr::SimilarTo { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: None, }, select.selection.unwrap() @@ -2106,7 +2143,9 @@ fn parse_similar_to() { Expr::SimilarTo { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: Some('^'.to_string()), }, select.selection.unwrap() @@ -2122,7 +2161,9 @@ fn parse_similar_to() { Expr::IsNull(Box::new(Expr::SimilarTo { expr: Box::new(Expr::Identifier(Ident::new("name"))), negated, - pattern: Box::new(Expr::Value(Value::SingleQuotedString("%a".to_string()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), escape_char: Some('^'.to_string()), })), select.selection.unwrap() @@ -2144,8 +2185,8 @@ fn parse_in_list() { Expr::InList { expr: Box::new(Expr::Identifier(Ident::new("segment"))), list: vec![ - Expr::Value(Value::SingleQuotedString("HIGH".to_string())), - Expr::Value(Value::SingleQuotedString("MED".to_string())), + Expr::Value((Value::SingleQuotedString("HIGH".to_string())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("MED".to_string())).with_empty_span()), ], negated, }, @@ -2287,8 +2328,8 @@ fn parse_between() { assert_eq!( Expr::Between { expr: Box::new(Expr::Identifier(Ident::new("age"))), - low: Box::new(Expr::Value(number("25"))), - high: Box::new(Expr::Value(number("32"))), + low: Box::new(Expr::value(number("25"))), + high: Box::new(Expr::value(number("32"))), negated, }, select.selection.unwrap() @@ -2305,16 +2346,16 @@ fn parse_between_with_expr() { let select = verified_only_select(sql); assert_eq!( Expr::IsNull(Box::new(Expr::Between { - expr: Box::new(Expr::Value(number("1"))), + expr: Box::new(Expr::value(number("1"))), low: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(number("1"))), + left: Box::new(Expr::value(number("1"))), op: Plus, - right: Box::new(Expr::Value(number("2"))), + right: Box::new(Expr::value(number("2"))), }), high: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(number("3"))), + left: Box::new(Expr::value(number("3"))), op: Plus, - right: Box::new(Expr::Value(number("4"))), + right: Box::new(Expr::value(number("4"))), }), negated: false, })), @@ -2326,19 +2367,19 @@ fn parse_between_with_expr() { assert_eq!( Expr::BinaryOp { left: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(number("1"))), + left: Box::new(Expr::value(number("1"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), op: BinaryOperator::And, right: Box::new(Expr::Between { expr: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(number("1"))), + left: Box::new(Expr::value(number("1"))), op: BinaryOperator::Plus, right: Box::new(Expr::Identifier(Ident::new("x"))), }), - low: Box::new(Expr::Value(number("1"))), - high: Box::new(Expr::Value(number("2"))), + low: Box::new(Expr::value(number("1"))), + high: Box::new(Expr::value(number("2"))), negated: false, }), }, @@ -2353,13 +2394,15 @@ fn parse_tuples() { assert_eq!( vec![ SelectItem::UnnamedExpr(Expr::Tuple(vec![ - Expr::Value(number("1")), - Expr::Value(number("2")), + Expr::value(number("1")), + Expr::value(number("2")), ])), - SelectItem::UnnamedExpr(Expr::Nested(Box::new(Expr::Value(number("1"))))), + SelectItem::UnnamedExpr(Expr::Nested(Box::new(Expr::Value( + (number("1")).with_empty_span() + )))), SelectItem::UnnamedExpr(Expr::Tuple(vec![ - Expr::Value(Value::SingleQuotedString("foo".into())), - Expr::Value(number("3")), + Expr::Value((Value::SingleQuotedString("foo".into())).with_empty_span()), + Expr::value(number("3")), Expr::Identifier(Ident::new("baz")), ])), ], @@ -2450,7 +2493,7 @@ fn parse_select_order_by_limit() { ]), select.order_by.expect("ORDER BY expected").kind ); - assert_eq!(Some(Expr::Value(number("2"))), select.limit); + assert_eq!(Some(Expr::value(number("2"))), select.limit); } #[test] @@ -2611,7 +2654,7 @@ fn parse_select_order_by_nulls_order() { ]), select.order_by.expect("ORDER BY expeccted").kind ); - assert_eq!(Some(Expr::Value(number("2"))), select.limit); + assert_eq!(Some(Expr::value(number("2"))), select.limit); } #[test] @@ -2755,7 +2798,7 @@ fn parse_select_having() { within_group: vec![] })), op: BinaryOperator::Gt, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), select.having ); @@ -2798,7 +2841,7 @@ fn parse_select_qualify() { within_group: vec![] })), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), select.qualify ); @@ -2809,7 +2852,7 @@ fn parse_select_qualify() { Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("row_num"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), select.qualify ); @@ -3191,14 +3234,14 @@ fn parse_listagg() { "dateid" )))), FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - Value::SingleQuotedString(", ".to_owned()) + (Value::SingleQuotedString(", ".to_owned())).with_empty_span() ))) ], clauses: vec![FunctionArgumentClause::OnOverflow( ListAggOnOverflow::Truncate { - filler: Some(Box::new(Expr::Value(Value::SingleQuotedString( - "%".to_string(), - )))), + filler: Some(Box::new(Expr::Value( + (Value::SingleQuotedString("%".to_string(),)).with_empty_span() + ))), with_count: false, } )], @@ -3454,13 +3497,13 @@ fn gen_number_case(value: &str) -> (Vec, Vec) { format!("{} AS col_alias", value), ]; let expected = vec![ - SelectItem::UnnamedExpr(Expr::Value(number(value))), + SelectItem::UnnamedExpr(Expr::value(number(value))), SelectItem::ExprWithAlias { - expr: Expr::Value(number(value)), + expr: Expr::value(number(value)), alias: Ident::new("col_alias"), }, SelectItem::ExprWithAlias { - expr: Expr::Value(number(value)), + expr: Expr::value(number(value)), alias: Ident::new("col_alias"), }, ]; @@ -3481,19 +3524,19 @@ fn gen_sign_number_case(value: &str, op: UnaryOperator) -> (Vec, Vec { match message { - Expr::Value(Value::SingleQuotedString(s)) => assert_eq!(s, "No rows in my_table"), + Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(s), + span: _, + }) => assert_eq!(s, "No rows in my_table"), _ => unreachable!(), }; } @@ -4324,11 +4370,13 @@ fn parse_create_table_with_options() { vec![ SqlOption::KeyValue { key: "foo".into(), - value: Expr::Value(Value::SingleQuotedString("bar".into())), + value: Expr::Value( + (Value::SingleQuotedString("bar".into())).with_empty_span() + ), }, SqlOption::KeyValue { key: "a".into(), - value: Expr::Value(number("123")), + value: Expr::value(number("123")), }, ], with_options @@ -4554,7 +4602,9 @@ fn parse_alter_table() { quote_style: Some('\''), span: Span::empty(), }, - value: Expr::Value(Value::SingleQuotedString("parquet".to_string())), + value: Expr::Value( + (Value::SingleQuotedString("parquet".to_string())).with_empty_span() + ), }], ); } @@ -4698,11 +4748,13 @@ fn parse_alter_view_with_options() { vec![ SqlOption::KeyValue { key: "foo".into(), - value: Expr::Value(Value::SingleQuotedString("bar".into())), + value: Expr::Value( + (Value::SingleQuotedString("bar".into())).with_empty_span() + ), }, SqlOption::KeyValue { key: "a".into(), - value: Expr::Value(number("123")), + value: Expr::value(number("123")), }, ], with_options @@ -4868,7 +4920,7 @@ fn parse_alter_table_alter_column() { assert_eq!( op, AlterColumnOperation::SetDefault { - value: Expr::Value(test_utils::number("0")) + value: Expr::Value((test_utils::number("0")).with_empty_span()) } ); } @@ -5202,16 +5254,16 @@ fn parse_named_argument_function() { args: vec![ FunctionArg::Named { name: Ident::new("a"), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "1".to_owned() - ))), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("1".to_owned())).with_empty_span() + )), operator: FunctionArgOperator::RightArrow }, FunctionArg::Named { name: Ident::new("b"), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "2".to_owned() - ))), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("2".to_owned())).with_empty_span() + )), operator: FunctionArgOperator::RightArrow }, ], @@ -5242,16 +5294,16 @@ fn parse_named_argument_function_with_eq_operator() { args: vec![ FunctionArg::Named { name: Ident::new("a"), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "1".to_owned() - ))), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("1".to_owned())).with_empty_span() + )), operator: FunctionArgOperator::Equals }, FunctionArg::Named { name: Ident::new("b"), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "2".to_owned() - ))), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("2".to_owned())).with_empty_span() + )), operator: FunctionArgOperator::Equals }, ], @@ -5275,7 +5327,7 @@ fn parse_named_argument_function_with_eq_operator() { [Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("bar"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(number("42"))), + right: Box::new(Expr::value(number("42"))), }] ), ); @@ -5625,14 +5677,14 @@ fn parse_literal_integer() { let select = verified_only_select(sql); assert_eq!(3, select.projection.len()); assert_eq!( - &Expr::Value(number("1")), + &Expr::value(number("1")), expr_from_projection(&select.projection[0]), ); // negative literal is parsed as a - and expr assert_eq!( &UnaryOp { op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(number("10"))) + expr: Box::new(Expr::value(number("10"))) }, expr_from_projection(&select.projection[1]), ); @@ -5640,7 +5692,7 @@ fn parse_literal_integer() { assert_eq!( &UnaryOp { op: UnaryOperator::Plus, - expr: Box::new(Expr::Value(number("20"))) + expr: Box::new(Expr::value(number("20"))) }, expr_from_projection(&select.projection[2]), ) @@ -5654,11 +5706,11 @@ fn parse_literal_decimal() { let select = verified_only_select(sql); assert_eq!(2, select.projection.len()); assert_eq!( - &Expr::Value(number("0.300000000000000004")), + &Expr::value(number("0.300000000000000004")), expr_from_projection(&select.projection[0]), ); assert_eq!( - &Expr::Value(number("9007199254740993.0")), + &Expr::value(number("9007199254740993.0")), expr_from_projection(&select.projection[1]), ) } @@ -5669,15 +5721,17 @@ fn parse_literal_string() { let select = verified_only_select(sql); assert_eq!(3, select.projection.len()); assert_eq!( - &Expr::Value(Value::SingleQuotedString("one".to_string())), + &Expr::Value((Value::SingleQuotedString("one".to_string())).with_empty_span()), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::NationalStringLiteral("national string".to_string())), + &Expr::Value( + (Value::NationalStringLiteral("national string".to_string())).with_empty_span() + ), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::HexStringLiteral("deadBEEF".to_string())), + &Expr::Value((Value::HexStringLiteral("deadBEEF".to_string())).with_empty_span()), expr_from_projection(&select.projection[2]) ); @@ -5762,7 +5816,9 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from("1-1")))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("1-1"))).with_empty_span() + )), leading_field: Some(DateTimeField::Year), leading_precision: None, last_field: Some(DateTimeField::Month), @@ -5775,9 +5831,9 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from( - "01:01.01" - )))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("01:01.01"))).with_empty_span() + )), leading_field: Some(DateTimeField::Minute), leading_precision: Some(5), last_field: Some(DateTimeField::Second), @@ -5790,7 +5846,9 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from("1")))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("1"))).with_empty_span() + )), leading_field: Some(DateTimeField::Second), leading_precision: Some(5), last_field: None, @@ -5803,7 +5861,9 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from("10")))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("10"))).with_empty_span() + )), leading_field: Some(DateTimeField::Hour), leading_precision: None, last_field: None, @@ -5816,7 +5876,7 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(number("5"))), + value: Box::new(Expr::value(number("5"))), leading_field: Some(DateTimeField::Day), leading_precision: None, last_field: None, @@ -5829,7 +5889,7 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(number("5"))), + value: Box::new(Expr::value(number("5"))), leading_field: Some(DateTimeField::Days), leading_precision: None, last_field: None, @@ -5842,7 +5902,9 @@ fn parse_interval_all() { let select = verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from("10")))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("10"))).with_empty_span() + )), leading_field: Some(DateTimeField::Hour), leading_precision: Some(1), last_field: None, @@ -5911,9 +5973,9 @@ fn parse_interval_dont_require_unit() { let select = dialects.verified_only_select(sql); assert_eq!( &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from( - "1 DAY" - )))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("1 DAY"))).with_empty_span() + )), leading_field: None, leading_precision: None, last_field: None, @@ -5950,9 +6012,9 @@ fn parse_interval_require_qualifier() { expr_from_projection(only(&select.projection)), &Expr::Interval(Interval { value: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(number("1"))), + left: Box::new(Expr::value(number("1"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), leading_field: Some(DateTimeField::Day), leading_precision: None, @@ -5967,9 +6029,13 @@ fn parse_interval_require_qualifier() { expr_from_projection(only(&select.projection)), &Expr::Interval(Interval { value: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::SingleQuotedString("1".to_string()))), + left: Box::new(Expr::Value( + (Value::SingleQuotedString("1".to_string())).with_empty_span() + )), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(Value::SingleQuotedString("1".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("1".to_string())).with_empty_span() + )), }), leading_field: Some(DateTimeField::Day), leading_precision: None, @@ -5985,12 +6051,18 @@ fn parse_interval_require_qualifier() { &Expr::Interval(Interval { value: Box::new(Expr::BinaryOp { left: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::SingleQuotedString("1".to_string()))), + left: Box::new(Expr::Value( + (Value::SingleQuotedString("1".to_string())).with_empty_span() + )), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(Value::SingleQuotedString("2".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("2".to_string())).with_empty_span() + )), }), op: BinaryOperator::Minus, - right: Box::new(Expr::Value(Value::SingleQuotedString("3".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("3".to_string())).with_empty_span() + )), }), leading_field: Some(DateTimeField::Day), leading_precision: None, @@ -6009,9 +6081,9 @@ fn parse_interval_disallow_interval_expr() { assert_eq!( expr_from_projection(only(&select.projection)), &Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from( - "1 DAY" - )))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("1 DAY"))).with_empty_span() + )), leading_field: None, leading_precision: None, last_field: None, @@ -6032,9 +6104,9 @@ fn parse_interval_disallow_interval_expr() { expr_from_projection(only(&select.projection)), &Expr::BinaryOp { left: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from( - "1 DAY" - )))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("1 DAY"))).with_empty_span() + )), leading_field: None, leading_precision: None, last_field: None, @@ -6042,9 +6114,9 @@ fn parse_interval_disallow_interval_expr() { })), op: BinaryOperator::Gt, right: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString(String::from( - "1 SECOND" - )))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString(String::from("1 SECOND"))).with_empty_span() + )), leading_field: None, leading_precision: None, last_field: None, @@ -6062,9 +6134,9 @@ fn interval_disallow_interval_expr_gt() { expr, Expr::BinaryOp { left: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString( - "1 second".to_string() - ))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("1 second".to_string())).with_empty_span() + )), leading_field: None, leading_precision: None, last_field: None, @@ -6089,9 +6161,9 @@ fn interval_disallow_interval_expr_double_colon() { Expr::Cast { kind: CastKind::DoubleColon, expr: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString( - "1 second".to_string() - ))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("1 second".to_string())).with_empty_span() + )), leading_field: None, leading_precision: None, last_field: None, @@ -6151,9 +6223,9 @@ fn parse_interval_and_or_xor() { })), op: BinaryOperator::Plus, right: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString( - "5 days".to_string(), - ))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("5 days".to_string())).with_empty_span(), + )), leading_field: None, leading_precision: None, last_field: None, @@ -6177,9 +6249,9 @@ fn parse_interval_and_or_xor() { })), op: BinaryOperator::Plus, right: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString( - "3 days".to_string(), - ))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("3 days".to_string())).with_empty_span(), + )), leading_field: None, leading_precision: None, last_field: None, @@ -6234,15 +6306,15 @@ fn parse_interval_and_or_xor() { #[test] fn parse_at_timezone() { - let zero = Expr::Value(number("0")); + let zero = Expr::value(number("0")); let sql = "SELECT FROM_UNIXTIME(0) AT TIME ZONE 'UTC-06:00' FROM t"; let select = verified_only_select(sql); assert_eq!( &Expr::AtTimeZone { timestamp: Box::new(call("FROM_UNIXTIME", [zero.clone()])), - time_zone: Box::new(Expr::Value(Value::SingleQuotedString( - "UTC-06:00".to_string() - ))), + time_zone: Box::new(Expr::Value( + (Value::SingleQuotedString("UTC-06:00".to_string())).with_empty_span() + )), }, expr_from_projection(only(&select.projection)), ); @@ -6256,11 +6328,13 @@ fn parse_at_timezone() { [ Expr::AtTimeZone { timestamp: Box::new(call("FROM_UNIXTIME", [zero])), - time_zone: Box::new(Expr::Value(Value::SingleQuotedString( - "UTC-06:00".to_string() - ))), + time_zone: Box::new(Expr::Value( + (Value::SingleQuotedString("UTC-06:00".to_string())).with_empty_span() + )), }, - Expr::Value(Value::SingleQuotedString("%Y-%m-%dT%H".to_string()),) + Expr::Value( + (Value::SingleQuotedString("%Y-%m-%dT%H".to_string())).with_empty_span() + ) ] ), alias: Ident { @@ -6434,7 +6508,9 @@ fn parse_table_function() { assert_eq!( call( "FUN", - [Expr::Value(Value::SingleQuotedString("1".to_owned()))], + [Expr::Value( + (Value::SingleQuotedString("1".to_owned())).with_empty_span() + )], ), expr ); @@ -6617,9 +6693,9 @@ fn parse_unnest_in_from_clause() { array_exprs: vec![call( "make_array", [ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")), + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")), ], )], with_offset: false, @@ -6643,14 +6719,14 @@ fn parse_unnest_in_from_clause() { call( "make_array", [ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")), + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")), ], ), call( "make_array", - [Expr::Value(number("5")), Expr::Value(number("6"))], + [Expr::value(number("5")), Expr::value(number("6"))], ), ], with_offset: false, @@ -6697,26 +6773,26 @@ fn parse_searched_case_expr() { conditions: vec![ CaseWhen { condition: IsNull(Box::new(Identifier(Ident::new("bar")))), - result: Expr::Value(Value::SingleQuotedString("null".to_string())), + result: Expr::value(Value::SingleQuotedString("null".to_string())), }, CaseWhen { condition: BinaryOp { left: Box::new(Identifier(Ident::new("bar"))), op: Eq, - right: Box::new(Expr::Value(number("0"))), + right: Box::new(Expr::value(number("0"))), }, - result: Expr::Value(Value::SingleQuotedString("=0".to_string())), + result: Expr::value(Value::SingleQuotedString("=0".to_string())), }, CaseWhen { condition: BinaryOp { left: Box::new(Identifier(Ident::new("bar"))), op: GtEq, - right: Box::new(Expr::Value(number("0"))), + right: Box::new(Expr::value(number("0"))), }, - result: Expr::Value(Value::SingleQuotedString(">=0".to_string())), + result: Expr::value(Value::SingleQuotedString(">=0".to_string())), }, ], - else_result: Some(Box::new(Expr::Value(Value::SingleQuotedString( + else_result: Some(Box::new(Expr::value(Value::SingleQuotedString( "<0".to_string() )))), }, @@ -6734,10 +6810,10 @@ fn parse_simple_case_expr() { &Case { operand: Some(Box::new(Identifier(Ident::new("foo")))), conditions: vec![CaseWhen { - condition: Expr::Value(number("1")), - result: Expr::Value(Value::SingleQuotedString("Y".to_string())), + condition: Expr::value(number("1")), + result: Expr::value(Value::SingleQuotedString("Y".to_string())), }], - else_result: Some(Box::new(Expr::Value(Value::SingleQuotedString( + else_result: Some(Box::new(Expr::value(Value::SingleQuotedString( "N".to_string() )))), }, @@ -7488,13 +7564,15 @@ fn parse_overlay() { let select = verified_only_select(sql); assert_eq!( &Expr::Overlay { - expr: Box::new(Expr::Value(Value::SingleQuotedString("abcdef".to_string()))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("abcdef".to_string())).with_empty_span() + )), overlay_what: Box::new(Expr::Identifier(Ident::new("name"))), - overlay_from: Box::new(Expr::Value(number("3"))), + overlay_from: Box::new(Expr::value(number("3"))), overlay_for: Some(Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("id"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), })), }, expr_from_projection(only(&select.projection)) @@ -7722,11 +7800,13 @@ fn parse_create_view_with_options() { CreateTableOptions::With(vec![ SqlOption::KeyValue { key: "foo".into(), - value: Expr::Value(Value::SingleQuotedString("bar".into())), + value: Expr::Value( + (Value::SingleQuotedString("bar".into())).with_empty_span() + ), }, SqlOption::KeyValue { key: "a".into(), - value: Expr::Value(number("123")), + value: Expr::value(number("123")), }, ]), options @@ -8069,7 +8149,7 @@ fn parse_offset() { all_dialects_where(|d| !d.is_column_alias(&Keyword::OFFSET, &mut Parser::new(d))); let expect = Some(Offset { - value: Expr::Value(number("2")), + value: Expr::value(number("2")), rows: OffsetRows::Rows, }); let ast = dialects.verified_query("SELECT foo FROM bar OFFSET 2 ROWS"); @@ -8097,7 +8177,7 @@ fn parse_offset() { assert_eq!( ast.offset, Some(Offset { - value: Expr::Value(number("0")), + value: Expr::value(number("0")), rows: OffsetRows::Rows, }) ); @@ -8105,7 +8185,7 @@ fn parse_offset() { assert_eq!( ast.offset, Some(Offset { - value: Expr::Value(number("1")), + value: Expr::value(number("1")), rows: OffsetRows::Row, }) ); @@ -8113,7 +8193,7 @@ fn parse_offset() { assert_eq!( ast.offset, Some(Offset { - value: Expr::Value(number("1")), + value: Expr::value(number("1")), rows: OffsetRows::None, }) ); @@ -8124,7 +8204,7 @@ fn parse_fetch() { let fetch_first_two_rows_only = Some(Fetch { with_ties: false, percent: false, - quantity: Some(Expr::Value(number("2"))), + quantity: Some(Expr::value(number("2"))), }); let ast = verified_query("SELECT foo FROM bar FETCH FIRST 2 ROWS ONLY"); assert_eq!(ast.fetch, fetch_first_two_rows_only); @@ -8151,7 +8231,7 @@ fn parse_fetch() { Some(Fetch { with_ties: true, percent: false, - quantity: Some(Expr::Value(number("2"))), + quantity: Some(Expr::value(number("2"))), }) ); let ast = verified_query("SELECT foo FROM bar FETCH FIRST 50 PERCENT ROWS ONLY"); @@ -8160,7 +8240,7 @@ fn parse_fetch() { Some(Fetch { with_ties: false, percent: true, - quantity: Some(Expr::Value(number("50"))), + quantity: Some(Expr::value(number("50"))), }) ); let ast = verified_query( @@ -8169,7 +8249,7 @@ fn parse_fetch() { assert_eq!( ast.offset, Some(Offset { - value: Expr::Value(number("2")), + value: Expr::value(number("2")), rows: OffsetRows::Rows, }) ); @@ -8191,7 +8271,7 @@ fn parse_fetch() { assert_eq!( ast.offset, Some(Offset { - value: Expr::Value(number("2")), + value: Expr::value(number("2")), rows: OffsetRows::Rows, }) ); @@ -8202,7 +8282,7 @@ fn parse_fetch() { assert_eq!( subquery.offset, Some(Offset { - value: Expr::Value(number("2")), + value: Expr::value(number("2")), rows: OffsetRows::Rows, }) ); @@ -8252,7 +8332,9 @@ fn lateral_derived() { let join = &from.joins[0]; assert_eq!( join.join_operator, - JoinOperator::Left(JoinConstraint::On(Expr::Value(test_utils::number("1")))) + JoinOperator::Left(JoinConstraint::On(Expr::Value( + (test_utils::number("1")).with_empty_span() + ))) ); if let TableFactor::Derived { lateral, @@ -8312,7 +8394,9 @@ fn lateral_function() { lateral: true, name: ObjectName::from(vec!["generate_series".into()]), args: vec![ - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("1")))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (number("1")).with_empty_span(), + ))), FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::CompoundIdentifier( vec![Ident::new("customer"), Ident::new("id")], ))), @@ -8481,7 +8565,9 @@ fn parse_set_variable() { ); assert_eq!( value, - vec![Expr::Value(Value::SingleQuotedString("1".into()))] + vec![Expr::Value( + (Value::SingleQuotedString("1".into())).with_empty_span() + )] ); } _ => unreachable!(), @@ -8509,9 +8595,9 @@ fn parse_set_variable() { assert_eq!( value, vec![ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")), + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")), ] ); } @@ -8581,7 +8667,9 @@ fn parse_set_role_as_variable() { ); assert_eq!( value, - vec![Expr::Value(Value::SingleQuotedString("foobar".into()))] + vec![Expr::Value( + (Value::SingleQuotedString("foobar".into())).with_empty_span() + )] ); } _ => unreachable!(), @@ -8597,15 +8685,16 @@ fn parse_double_colon_cast_at_timezone() { &Expr::AtTimeZone { timestamp: Box::new(Expr::Cast { kind: CastKind::DoubleColon, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "2001-01-01T00:00:00.000Z".to_string() - ),)), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("2001-01-01T00:00:00.000Z".to_string())) + .with_empty_span() + )), data_type: DataType::Timestamp(None, TimezoneInfo::None), format: None }), - time_zone: Box::new(Expr::Value(Value::SingleQuotedString( - "Europe/Brussels".to_string() - ))), + time_zone: Box::new(Expr::Value( + (Value::SingleQuotedString("Europe/Brussels".to_string())).with_empty_span() + )), }, expr_from_projection(only(&select.projection)), ); @@ -8628,7 +8717,9 @@ fn parse_set_time_zone() { ); assert_eq!( value, - vec![Expr::Value(Value::SingleQuotedString("UTC".into()))] + vec![Expr::Value( + (Value::SingleQuotedString("UTC".into())).with_empty_span() + )] ); } _ => unreachable!(), @@ -8642,7 +8733,10 @@ fn parse_set_time_zone_alias() { match verified_stmt("SET TIME ZONE 'UTC'") { Statement::SetTimeZone { local, value } => { assert!(!local); - assert_eq!(value, Expr::Value(Value::SingleQuotedString("UTC".into()))); + assert_eq!( + value, + Expr::Value((Value::SingleQuotedString("UTC".into())).with_empty_span()) + ); } _ => unreachable!(), } @@ -8849,7 +8943,7 @@ fn test_create_index_with_with_clause() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("fillfactor"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(number("70"))), + right: Box::new(Expr::value(number("70"))), }, Expr::Identifier(Ident::new("single_param")), ]; @@ -9338,9 +9432,9 @@ fn parse_merge() { Ident::new("A"), ])), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "a".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("a".to_string())).with_empty_span() + )), }), action: MergeAction::Update { assignments: vec![ @@ -9566,7 +9660,9 @@ fn test_placeholder() { Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("id"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Placeholder("$Id1".into()))), + right: Box::new(Expr::Value( + (Value::Placeholder("$Id1".into())).with_empty_span() + )), }) ); @@ -9574,12 +9670,14 @@ fn test_placeholder() { let ast = dialects.verified_query(sql); assert_eq!( ast.limit, - Some(Expr::Value(Value::Placeholder("$1".into()))) + Some(Expr::Value( + (Value::Placeholder("$1".into())).with_empty_span() + )) ); assert_eq!( ast.offset, Some(Offset { - value: Expr::Value(Value::Placeholder("$2".into())), + value: Expr::Value((Value::Placeholder("$2".into())).with_empty_span()), rows: OffsetRows::None, }), ); @@ -9603,7 +9701,9 @@ fn test_placeholder() { Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("id"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Placeholder("?".into()))), + right: Box::new(Expr::Value( + (Value::Placeholder("?".into())).with_empty_span() + )), }) ); @@ -9612,9 +9712,15 @@ fn test_placeholder() { assert_eq!( ast.projection, vec![ - UnnamedExpr(Expr::Value(Value::Placeholder("$fromage_français".into()))), - UnnamedExpr(Expr::Value(Value::Placeholder(":x".into()))), - UnnamedExpr(Expr::Value(Value::Placeholder("?123".into()))), + UnnamedExpr(Expr::Value( + (Value::Placeholder("$fromage_français".into())).with_empty_span() + )), + UnnamedExpr(Expr::Value( + (Value::Placeholder(":x".into())).with_empty_span() + )), + UnnamedExpr(Expr::Value( + (Value::Placeholder("?123".into())).with_empty_span() + )), ] ); } @@ -9655,12 +9761,12 @@ fn verified_expr(query: &str) -> Expr { fn parse_offset_and_limit() { let sql = "SELECT foo FROM bar LIMIT 1 OFFSET 2"; let expect = Some(Offset { - value: Expr::Value(number("2")), + value: Expr::value(number("2")), rows: OffsetRows::None, }); let ast = verified_query(sql); assert_eq!(ast.offset, expect); - assert_eq!(ast.limit, Some(Expr::Value(number("1")))); + assert_eq!(ast.limit, Some(Expr::value(number("1")))); // different order is OK one_statement_parses_to("SELECT foo FROM bar OFFSET 2 LIMIT 1", sql); @@ -9680,18 +9786,18 @@ fn parse_offset_and_limit() { assert_eq!( ast.limit, Some(Expr::BinaryOp { - left: Box::new(Expr::Value(number("1"))), + left: Box::new(Expr::value(number("1"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("2"))), + right: Box::new(Expr::value(number("2"))), }), ); assert_eq!( ast.offset, Some(Offset { value: Expr::BinaryOp { - left: Box::new(Expr::Value(number("3"))), + left: Box::new(Expr::value(number("3"))), op: BinaryOperator::Multiply, - right: Box::new(Expr::Value(number("4"))), + right: Box::new(Expr::value(number("4"))), }, rows: OffsetRows::None, }), @@ -9762,7 +9868,9 @@ fn parse_time_functions() { fn parse_position() { assert_eq!( Expr::Position { - expr: Box::new(Expr::Value(Value::SingleQuotedString("@".to_string()))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("@".to_string())).with_empty_span() + )), r#in: Box::new(Expr::Identifier(Ident::new("field"))), }, verified_expr("POSITION('@' IN field)"), @@ -9773,9 +9881,9 @@ fn parse_position() { call( "position", [ - Expr::Value(Value::SingleQuotedString("an".to_owned())), - Expr::Value(Value::SingleQuotedString("banana".to_owned())), - Expr::Value(number("1")), + Expr::Value((Value::SingleQuotedString("an".to_owned())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("banana".to_owned())).with_empty_span()), + Expr::value(number("1")), ] ), verified_expr("position('an', 'banana', 1)") @@ -10028,11 +10136,11 @@ fn parse_cache_table() { options: vec![ SqlOption::KeyValue { key: Ident::with_quote('\'', "K1"), - value: Expr::Value(Value::SingleQuotedString("V1".into())), + value: Expr::Value((Value::SingleQuotedString("V1".into())).with_empty_span()), }, SqlOption::KeyValue { key: Ident::with_quote('\'', "K2"), - value: Expr::Value(number("0.88")), + value: Expr::value(number("0.88")), }, ], query: None, @@ -10053,11 +10161,11 @@ fn parse_cache_table() { options: vec![ SqlOption::KeyValue { key: Ident::with_quote('\'', "K1"), - value: Expr::Value(Value::SingleQuotedString("V1".into())), + value: Expr::Value((Value::SingleQuotedString("V1".into())).with_empty_span()), }, SqlOption::KeyValue { key: Ident::with_quote('\'', "K2"), - value: Expr::Value(number("0.88")), + value: Expr::value(number("0.88")), }, ], query: Some(query.clone().into()), @@ -10078,11 +10186,11 @@ fn parse_cache_table() { options: vec![ SqlOption::KeyValue { key: Ident::with_quote('\'', "K1"), - value: Expr::Value(Value::SingleQuotedString("V1".into())), + value: Expr::Value((Value::SingleQuotedString("V1".into())).with_empty_span()), }, SqlOption::KeyValue { key: Ident::with_quote('\'', "K2"), - value: Expr::Value(number("0.88")), + value: Expr::value(number("0.88")), }, ], query: Some(query.clone().into()), @@ -10293,7 +10401,9 @@ fn parse_escaped_string_with_unescape() { let expr = expr_from_projection(only(&value.projection)); assert_eq!( *expr, - Expr::Value(Value::SingleQuotedString(quoted.to_string())) + Expr::Value( + (Value::SingleQuotedString(quoted.to_string())).with_empty_span() + ) ); } _ => unreachable!(), @@ -10333,7 +10443,9 @@ fn parse_escaped_string_without_unescape() { let expr = expr_from_projection(only(&value.projection)); assert_eq!( *expr, - Expr::Value(Value::SingleQuotedString(quoted.to_string())) + Expr::Value( + (Value::SingleQuotedString(quoted.to_string())).with_empty_span() + ) ); } _ => unreachable!(), @@ -10404,11 +10516,13 @@ fn parse_pivot_table() { value_column: vec![Ident::new("a"), Ident::new("MONTH")], value_source: PivotValueSource::List(vec![ ExprWithAlias { - expr: Expr::Value(number("1")), + expr: Expr::value(number("1")), alias: Some(Ident::new("x")) }, ExprWithAlias { - expr: Expr::Value(Value::SingleQuotedString("two".to_string())), + expr: Expr::Value( + (Value::SingleQuotedString("two".to_string())).with_empty_span() + ), alias: None }, ExprWithAlias { @@ -10639,11 +10753,17 @@ fn parse_pivot_unpivot_table() { value_column: vec![Ident::new("year")], value_source: PivotValueSource::List(vec![ ExprWithAlias { - expr: Expr::Value(Value::SingleQuotedString("population_2000".to_string())), + expr: Expr::Value( + (Value::SingleQuotedString("population_2000".to_string())) + .with_empty_span() + ), alias: None }, ExprWithAlias { - expr: Expr::Value(Value::SingleQuotedString("population_2010".to_string())), + expr: Expr::Value( + (Value::SingleQuotedString("population_2010".to_string())) + .with_empty_span() + ), alias: None }, ]), @@ -10890,7 +11010,7 @@ fn parse_call() { args: FunctionArguments::List(FunctionArgumentList { duplicate_treatment: None, args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - Value::SingleQuotedString("a".to_string()) + (Value::SingleQuotedString("a".to_string())).with_empty_span() )))], clauses: vec![], }), @@ -10919,8 +11039,8 @@ fn parse_execute_stored_procedure() { }, ])), parameters: vec![ - Expr::Value(Value::NationalStringLiteral("param1".to_string())), - Expr::Value(Value::NationalStringLiteral("param2".to_string())), + Expr::Value((Value::NationalStringLiteral("param1".to_string())).with_empty_span()), + Expr::Value((Value::NationalStringLiteral("param2".to_string())).with_empty_span()), ], has_parentheses: false, immediate: false, @@ -10947,12 +11067,12 @@ fn parse_execute_immediate() { let dialects = all_dialects_where(|d| d.supports_execute_immediate()); let expected = Statement::Execute { - parameters: vec![Expr::Value(Value::SingleQuotedString( - "SELECT 1".to_string(), - ))], + parameters: vec![Expr::Value( + (Value::SingleQuotedString("SELECT 1".to_string())).with_empty_span(), + )], immediate: true, using: vec![ExprWithAlias { - expr: Expr::Value(number("1")), + expr: Expr::value(number("1")), alias: Some(Ident::new("b")), }], into: vec![Ident::new("a")], @@ -11102,7 +11222,9 @@ fn parse_unload() { quote_style: None, span: Span::empty(), }, - value: Expr::Value(Value::SingleQuotedString("AVRO".to_string())) + value: Expr::Value( + (Value::SingleQuotedString("AVRO".to_string())).with_empty_span() + ) }] } ); @@ -11210,7 +11332,7 @@ fn parse_map_access_expr() { AccessExpr::Subscript(Subscript::Index { index: Expr::UnaryOp { op: UnaryOperator::Minus, - expr: Expr::Value(number("1")).into(), + expr: Expr::value(number("1")).into(), }, }), AccessExpr::Subscript(Subscript::Index { @@ -11223,7 +11345,7 @@ fn parse_map_access_expr() { args: FunctionArguments::List(FunctionArgumentList { duplicate_treatment: None, args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - number("2"), + (number("2")).with_empty_span(), )))], clauses: vec![], }), @@ -11276,9 +11398,9 @@ fn parse_connect_by() { condition: Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("title"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "president".to_owned(), - ))), + right: Box::new(Expr::Value( + Value::SingleQuotedString("president".to_owned()).with_empty_span(), + )), }, relationships: vec![Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("manager_id"))), @@ -11346,7 +11468,7 @@ fn parse_connect_by() { selection: Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("employee_id"))), op: BinaryOperator::NotEq, - right: Box::new(Expr::Value(number("42"))), + right: Box::new(Expr::value(number("42"))), }), group_by: GroupByExpr::Expressions(vec![], vec![]), cluster_by: vec![], @@ -11361,9 +11483,9 @@ fn parse_connect_by() { condition: Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("title"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "president".to_owned(), - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("president".to_owned(),)).with_empty_span() + )), }, relationships: vec![Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("manager_id"))), @@ -11446,7 +11568,9 @@ fn test_selective_aggregation() { filter: Some(Box::new(Expr::Like { negated: false, expr: Box::new(Expr::Identifier(Ident::new("name"))), - pattern: Box::new(Expr::Value(Value::SingleQuotedString("a%".to_owned()))), + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("a%".to_owned())).with_empty_span() + )), escape_char: None, any: false, })), @@ -11804,7 +11928,9 @@ fn test_select_wildcard_with_replace() { let expected = SelectItem::Wildcard(WildcardAdditionalOptions { opt_replace: Some(ReplaceSelectItem { items: vec![Box::new(ReplaceSelectElement { - expr: Expr::Value(Value::SingleQuotedString("widget".to_owned())), + expr: Expr::Value( + (Value::SingleQuotedString("widget".to_owned())).with_empty_span(), + ), column_name: Ident::new("item_name"), as_keyword: true, })], @@ -11823,13 +11949,13 @@ fn test_select_wildcard_with_replace() { expr: Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("quantity"))), op: BinaryOperator::Divide, - right: Box::new(Expr::Value(number("2"))), + right: Box::new(Expr::value(number("2"))), }, column_name: Ident::new("quantity"), as_keyword: true, }), Box::new(ReplaceSelectElement { - expr: Expr::Value(number("3")), + expr: Expr::value(number("3")), column_name: Ident::new("order_id"), as_keyword: true, }), @@ -11912,15 +12038,15 @@ fn test_dictionary_syntax() { Expr::Dictionary(vec![ DictionaryField { key: Ident::with_quote('\'', "Alberta"), - value: Box::new(Expr::Value(Value::SingleQuotedString( - "Edmonton".to_owned(), - ))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("Edmonton".to_owned())).with_empty_span(), + )), }, DictionaryField { key: Ident::with_quote('\'', "Manitoba"), - value: Box::new(Expr::Value(Value::SingleQuotedString( - "Winnipeg".to_owned(), - ))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("Winnipeg".to_owned())).with_empty_span(), + )), }, ]), ); @@ -11932,9 +12058,9 @@ fn test_dictionary_syntax() { key: Ident::with_quote('\'', "start"), value: Box::new(Expr::Cast { kind: CastKind::Cast, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "2023-04-01".to_owned(), - ))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("2023-04-01".to_owned())).with_empty_span(), + )), data_type: DataType::Timestamp(None, TimezoneInfo::None), format: None, }), @@ -11943,9 +12069,9 @@ fn test_dictionary_syntax() { key: Ident::with_quote('\'', "end"), value: Box::new(Expr::Cast { kind: CastKind::Cast, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "2023-04-05".to_owned(), - ))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("2023-04-05".to_owned())).with_empty_span(), + )), data_type: DataType::Timestamp(None, TimezoneInfo::None), format: None, }), @@ -11968,25 +12094,27 @@ fn test_map_syntax() { Expr::Map(Map { entries: vec![ MapEntry { - key: Box::new(Expr::Value(Value::SingleQuotedString("Alberta".to_owned()))), - value: Box::new(Expr::Value(Value::SingleQuotedString( - "Edmonton".to_owned(), - ))), + key: Box::new(Expr::Value( + (Value::SingleQuotedString("Alberta".to_owned())).with_empty_span(), + )), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("Edmonton".to_owned())).with_empty_span(), + )), }, MapEntry { - key: Box::new(Expr::Value(Value::SingleQuotedString( - "Manitoba".to_owned(), - ))), - value: Box::new(Expr::Value(Value::SingleQuotedString( - "Winnipeg".to_owned(), - ))), + key: Box::new(Expr::Value( + (Value::SingleQuotedString("Manitoba".to_owned())).with_empty_span(), + )), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("Winnipeg".to_owned())).with_empty_span(), + )), }, ], }), ); fn number_expr(s: &str) -> Expr { - Expr::Value(number(s)) + Expr::value(number(s)) } check( @@ -12014,14 +12142,14 @@ fn test_map_syntax() { elem: vec![number_expr("1"), number_expr("2"), number_expr("3")], named: false, })), - value: Box::new(Expr::Value(number("10.0"))), + value: Box::new(Expr::value(number("10.0"))), }, MapEntry { key: Box::new(Expr::Array(Array { elem: vec![number_expr("4"), number_expr("5"), number_expr("6")], named: false, })), - value: Box::new(Expr::Value(number("20.0"))), + value: Box::new(Expr::value(number("20.0"))), }, ], }), @@ -12033,17 +12161,21 @@ fn test_map_syntax() { root: Box::new(Expr::Map(Map { entries: vec![ MapEntry { - key: Box::new(Expr::Value(Value::SingleQuotedString("a".to_owned()))), + key: Box::new(Expr::Value( + (Value::SingleQuotedString("a".to_owned())).with_empty_span(), + )), value: Box::new(number_expr("10")), }, MapEntry { - key: Box::new(Expr::Value(Value::SingleQuotedString("b".to_owned()))), + key: Box::new(Expr::Value( + (Value::SingleQuotedString("b".to_owned())).with_empty_span(), + )), value: Box::new(number_expr("20")), }, ], })), access_chain: vec![AccessExpr::Subscript(Subscript::Index { - index: Expr::Value(Value::SingleQuotedString("a".to_owned())), + index: Expr::Value((Value::SingleQuotedString("a".to_owned())).with_empty_span()), })], }, ); @@ -12192,9 +12324,9 @@ fn test_extract_seconds_ok() { syntax: ExtractSyntax::From, expr: Box::new(Expr::Cast { kind: CastKind::DoubleColon, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "2 seconds".to_string() - ))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("2 seconds".to_string())).with_empty_span() + )), data_type: DataType::Interval, format: None, }), @@ -12217,9 +12349,9 @@ fn test_extract_seconds_ok() { syntax: ExtractSyntax::From, expr: Box::new(Expr::Cast { kind: CastKind::DoubleColon, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "2 seconds".to_string(), - ))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("2 seconds".to_string())).with_empty_span(), + )), data_type: DataType::Interval, format: None, }), @@ -12271,9 +12403,9 @@ fn test_extract_seconds_single_quote_ok() { syntax: ExtractSyntax::From, expr: Box::new(Expr::Cast { kind: CastKind::DoubleColon, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "2 seconds".to_string() - ))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("2 seconds".to_string())).with_empty_span() + )), data_type: DataType::Interval, format: None, }), @@ -12324,11 +12456,11 @@ fn parse_explain_with_option_list() { Some(vec![ UtilityOption { name: Ident::new("ANALYZE"), - arg: Some(Expr::Value(Value::Boolean(false))), + arg: Some(Expr::Value((Value::Boolean(false)).with_empty_span())), }, UtilityOption { name: Ident::new("VERBOSE"), - arg: Some(Expr::Value(Value::Boolean(true))), + arg: Some(Expr::Value((Value::Boolean(true)).with_empty_span())), }, ]), ); @@ -12364,7 +12496,9 @@ fn parse_explain_with_option_list() { }, UtilityOption { name: Ident::new("FORMAT2"), - arg: Some(Expr::Value(Value::SingleQuotedString("JSON".to_string()))), + arg: Some(Expr::Value( + (Value::SingleQuotedString("JSON".to_string())).with_empty_span(), + )), }, UtilityOption { name: Ident::new("FORMAT3"), @@ -12386,20 +12520,26 @@ fn parse_explain_with_option_list() { Some(vec![ UtilityOption { name: Ident::new("NUM1"), - arg: Some(Expr::Value(Value::Number("10".parse().unwrap(), false))), + arg: Some(Expr::Value( + (Value::Number("10".parse().unwrap(), false)).with_empty_span(), + )), }, UtilityOption { name: Ident::new("NUM2"), arg: Some(Expr::UnaryOp { op: UnaryOperator::Plus, - expr: Box::new(Expr::Value(Value::Number("10.1".parse().unwrap(), false))), + expr: Box::new(Expr::Value( + (Value::Number("10.1".parse().unwrap(), false)).with_empty_span(), + )), }), }, UtilityOption { name: Ident::new("NUM3"), arg: Some(Expr::UnaryOp { op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(Value::Number("10.2".parse().unwrap(), false))), + expr: Box::new(Expr::Value( + (Value::Number("10.2".parse().unwrap(), false)).with_empty_span(), + )), }), }, ]), @@ -12412,7 +12552,7 @@ fn parse_explain_with_option_list() { }, UtilityOption { name: Ident::new("VERBOSE"), - arg: Some(Expr::Value(Value::Boolean(true))), + arg: Some(Expr::Value((Value::Boolean(true)).with_empty_span())), }, UtilityOption { name: Ident::new("WAL"), @@ -12426,7 +12566,9 @@ fn parse_explain_with_option_list() { name: Ident::new("USER_DEF_NUM"), arg: Some(Expr::UnaryOp { op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(Value::Number("100.1".parse().unwrap(), false))), + expr: Box::new(Expr::Value( + (Value::Number("100.1".parse().unwrap(), false)).with_empty_span(), + )), }), }, ]; @@ -12471,15 +12613,21 @@ fn test_create_policy() { Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("c0"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), }) ); assert_eq!( with_check, Some(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + left: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), }) ); } @@ -12692,11 +12840,15 @@ fn test_create_connector() { Some(vec![ SqlOption::KeyValue { key: Ident::with_quote('\'', "user"), - value: Expr::Value(Value::SingleQuotedString("root".to_string())) + value: Expr::Value( + (Value::SingleQuotedString("root".to_string())).with_empty_span() + ) }, SqlOption::KeyValue { key: Ident::with_quote('\'', "password"), - value: Expr::Value(Value::SingleQuotedString("password".to_string())) + value: Expr::Value( + (Value::SingleQuotedString("password".to_string())).with_empty_span() + ) } ]) ); @@ -12759,11 +12911,15 @@ fn test_alter_connector() { Some(vec![ SqlOption::KeyValue { key: Ident::with_quote('\'', "user"), - value: Expr::Value(Value::SingleQuotedString("root".to_string())) + value: Expr::Value( + (Value::SingleQuotedString("root".to_string())).with_empty_span() + ) }, SqlOption::KeyValue { key: Ident::with_quote('\'', "password"), - value: Expr::Value(Value::SingleQuotedString("password".to_string())) + value: Expr::Value( + (Value::SingleQuotedString("password".to_string())).with_empty_span() + ) } ]) ); @@ -13230,12 +13386,16 @@ fn parse_load_data() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("year"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("2024".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("2024".parse().unwrap(), false)).with_empty_span() + )), }, Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("month"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("11".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("11".parse().unwrap(), false)).with_empty_span() + )), } ]), partitioned @@ -13268,24 +13428,34 @@ fn parse_load_data() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("year"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("2024".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("2024".parse().unwrap(), false)).with_empty_span() + )), }, Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("month"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::Number("11".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("11".parse().unwrap(), false)).with_empty_span() + )), } ]), partitioned ); assert_eq!( Some(HiveLoadDataFormat { - serde: Expr::Value(Value::SingleQuotedString( - "org.apache.hadoop.hive.serde2.OpenCSVSerde".to_string() - )), - input_format: Expr::Value(Value::SingleQuotedString( - "org.apache.hadoop.mapred.TextInputFormat".to_string() - )) + serde: Expr::Value( + (Value::SingleQuotedString( + "org.apache.hadoop.hive.serde2.OpenCSVSerde".to_string() + )) + .with_empty_span() + ), + input_format: Expr::Value( + (Value::SingleQuotedString( + "org.apache.hadoop.mapred.TextInputFormat".to_string() + )) + .with_empty_span() + ) }), table_format ); @@ -13363,7 +13533,9 @@ fn parse_bang_not() { Box::new(Expr::Nested(Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("b"))), op: BinaryOperator::Gt, - right: Box::new(Expr::Value(Value::Number("3".parse().unwrap(), false))), + right: Box::new(Expr::Value( + Value::Number("3".parse().unwrap(), false).with_empty_span(), + )), }))), ] .into_iter() @@ -13705,11 +13877,13 @@ fn parse_composite_access_expr() { values: vec![ Expr::Named { name: Ident::new("a"), - expr: Box::new(Expr::Value(Number("1".parse().unwrap(), false))), + expr: Box::new(Expr::Value( + (Number("1".parse().unwrap(), false)).with_empty_span(), + )), }, Expr::Named { name: Ident::new("b"), - expr: Box::new(Expr::Value(Value::Null)), + expr: Box::new(Expr::Value((Value::Null).with_empty_span())), }, ], fields: vec![], @@ -13717,7 +13891,7 @@ fn parse_composite_access_expr() { }, Expr::Named { name: Ident::new("d"), - expr: Box::new(Expr::Value(Value::Null)), + expr: Box::new(Expr::Value((Value::Null).with_empty_span())), }, ], fields: vec![], @@ -13744,11 +13918,15 @@ fn parse_create_table_with_enum_types() { vec![ EnumMember::NamedValue( "a".to_string(), - Expr::Value(Number("1".parse().unwrap(), false)) + Expr::Value( + (Number("1".parse().unwrap(), false)).with_empty_span() + ) ), EnumMember::NamedValue( "b".to_string(), - Expr::Value(Number("2".parse().unwrap(), false)) + Expr::Value( + (Number("2".parse().unwrap(), false)).with_empty_span() + ) ) ], Some(8) @@ -13761,11 +13939,15 @@ fn parse_create_table_with_enum_types() { vec![ EnumMember::NamedValue( "a".to_string(), - Expr::Value(Number("1".parse().unwrap(), false)) + Expr::Value( + (Number("1".parse().unwrap(), false)).with_empty_span() + ) ), EnumMember::NamedValue( "b".to_string(), - Expr::Value(Number("2".parse().unwrap(), false)) + Expr::Value( + (Number("2".parse().unwrap(), false)).with_empty_span() + ) ) ], Some(16) @@ -13956,8 +14138,12 @@ fn test_lambdas() { call( "array", [ - Expr::Value(Value::SingleQuotedString("Hello".to_owned())), - Expr::Value(Value::SingleQuotedString("World".to_owned())) + Expr::Value( + (Value::SingleQuotedString("Hello".to_owned())).with_empty_span() + ), + Expr::Value( + (Value::SingleQuotedString("World".to_owned())).with_empty_span() + ) ] ), Expr::Lambda(LambdaFunction { @@ -13971,7 +14157,7 @@ fn test_lambdas() { op: BinaryOperator::Eq, right: Box::new(Expr::Identifier(Ident::new("p2"))) }, - result: Expr::Value(number("0")) + result: Expr::value(number("0")), }, CaseWhen { condition: Expr::BinaryOp { @@ -13983,15 +14169,15 @@ fn test_lambdas() { right: Box::new(call( "reverse", [Expr::Identifier(Ident::new("p2"))] - )) + )), }, result: Expr::UnaryOp { op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(number("1"))) + expr: Box::new(Expr::value(number("1"))) } - } + }, ], - else_result: Some(Box::new(Expr::Value(number("1")))) + else_result: Some(Box::new(Expr::value(number("1")))), }) }) ] diff --git a/tests/sqlparser_custom_dialect.rs b/tests/sqlparser_custom_dialect.rs index 61874fc27..cee604aca 100644 --- a/tests/sqlparser_custom_dialect.rs +++ b/tests/sqlparser_custom_dialect.rs @@ -41,7 +41,7 @@ fn custom_prefix_parser() -> Result<(), ParserError> { fn parse_prefix(&self, parser: &mut Parser) -> Option> { if parser.consume_token(&Token::Number("1".to_string(), false)) { - Some(Ok(Expr::Value(Value::Null))) + Some(Ok(Expr::Value(Value::Null.with_empty_span()))) } else { None } diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 724bedf47..3b36d7a13 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -47,7 +47,9 @@ fn test_databricks_identifiers() { databricks() .verified_only_select(r#"SELECT "Ä""#) .projection[0], - SelectItem::UnnamedExpr(Expr::Value(Value::DoubleQuotedString("Ä".to_owned()))) + SelectItem::UnnamedExpr(Expr::Value( + (Value::DoubleQuotedString("Ä".to_owned())).with_empty_span() + )) ); } @@ -62,9 +64,9 @@ fn test_databricks_exists() { call( "array", [ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")) + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")) ] ), Expr::Lambda(LambdaFunction { @@ -99,8 +101,8 @@ fn test_databricks_lambdas() { call( "array", [ - Expr::Value(Value::SingleQuotedString("Hello".to_owned())), - Expr::Value(Value::SingleQuotedString("World".to_owned())) + Expr::value(Value::SingleQuotedString("Hello".to_owned())), + Expr::value(Value::SingleQuotedString("World".to_owned())) ] ), Expr::Lambda(LambdaFunction { @@ -114,7 +116,7 @@ fn test_databricks_lambdas() { op: BinaryOperator::Eq, right: Box::new(Expr::Identifier(Ident::new("p2"))) }, - result: Expr::Value(number("0")) + result: Expr::value(number("0")) }, CaseWhen { condition: Expr::BinaryOp { @@ -130,11 +132,11 @@ fn test_databricks_lambdas() { }, result: Expr::UnaryOp { op: UnaryOperator::Minus, - expr: Box::new(Expr::Value(number("1"))) + expr: Box::new(Expr::value(number("1"))) } }, ], - else_result: Some(Box::new(Expr::Value(number("1")))) + else_result: Some(Box::new(Expr::value(number("1")))) }) }) ] @@ -154,12 +156,12 @@ fn test_values_clause() { explicit_row: false, rows: vec![ vec![ - Expr::Value(Value::DoubleQuotedString("one".to_owned())), - Expr::Value(number("1")), + Expr::Value((Value::DoubleQuotedString("one".to_owned())).with_empty_span()), + Expr::value(number("1")), ], vec![ - Expr::Value(Value::SingleQuotedString("two".to_owned())), - Expr::Value(number("2")), + Expr::Value((Value::SingleQuotedString("two".to_owned())).with_empty_span()), + Expr::value(number("2")), ], ], }; @@ -286,8 +288,8 @@ fn parse_databricks_struct_function() { .projection[0], SelectItem::UnnamedExpr(Expr::Struct { values: vec![ - Expr::Value(number("1")), - Expr::Value(Value::SingleQuotedString("foo".to_string())) + Expr::value(number("1")), + Expr::Value((Value::SingleQuotedString("foo".to_string())).with_empty_span()) ], fields: vec![] }) @@ -299,14 +301,17 @@ fn parse_databricks_struct_function() { SelectItem::UnnamedExpr(Expr::Struct { values: vec![ Expr::Named { - expr: Expr::Value(number("1")).into(), + expr: Expr::value(number("1")).into(), name: Ident::new("one") }, Expr::Named { - expr: Expr::Value(Value::SingleQuotedString("foo".to_string())).into(), + expr: Expr::Value( + (Value::SingleQuotedString("foo".to_string())).with_empty_span() + ) + .into(), name: Ident::new("foo") }, - Expr::Value(Value::Boolean(false)) + Expr::Value((Value::Boolean(false)).with_empty_span()) ], fields: vec![] }) diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 05e60d556..bed02428d 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -210,7 +210,7 @@ fn test_create_macro_default_args() { MacroArg::new("a"), MacroArg { name: Ident::new("b"), - default_expr: Some(Expr::Value(number("5"))), + default_expr: Some(Expr::value(number("5"))), }, ]), definition: MacroDefinition::Expr(Expr::BinaryOp { @@ -363,15 +363,15 @@ fn test_duckdb_struct_literal() { &Expr::Dictionary(vec![ DictionaryField { key: Ident::with_quote('\'', "a"), - value: Box::new(Expr::Value(number("1"))), + value: Box::new(Expr::value(number("1"))), }, DictionaryField { key: Ident::with_quote('\'', "b"), - value: Box::new(Expr::Value(number("2"))), + value: Box::new(Expr::value(number("2"))), }, DictionaryField { key: Ident::with_quote('\'', "c"), - value: Box::new(Expr::Value(number("3"))), + value: Box::new(Expr::value(number("3"))), }, ],), expr_from_projection(&select.projection[0]) @@ -381,7 +381,9 @@ fn test_duckdb_struct_literal() { &Expr::Array(Array { elem: vec![Expr::Dictionary(vec![DictionaryField { key: Ident::with_quote('\'', "a"), - value: Box::new(Expr::Value(Value::SingleQuotedString("abc".to_string()))), + value: Box::new(Expr::Value( + (Value::SingleQuotedString("abc".to_string())).with_empty_span() + )), },],)], named: false }), @@ -391,7 +393,7 @@ fn test_duckdb_struct_literal() { &Expr::Dictionary(vec![ DictionaryField { key: Ident::with_quote('\'', "a"), - value: Box::new(Expr::Value(number("1"))), + value: Box::new(Expr::value(number("1"))), }, DictionaryField { key: Ident::with_quote('\'', "b"), @@ -410,11 +412,14 @@ fn test_duckdb_struct_literal() { &Expr::Dictionary(vec![ DictionaryField { key: Ident::with_quote('\'', "a"), - value: Expr::Value(number("1")).into(), + value: Expr::value(number("1")).into(), }, DictionaryField { key: Ident::with_quote('\'', "b"), - value: Expr::Value(Value::SingleQuotedString("abc".to_string())).into(), + value: Expr::Value( + (Value::SingleQuotedString("abc".to_string())).with_empty_span() + ) + .into(), }, ],), expr_from_projection(&select.projection[3]) @@ -431,7 +436,7 @@ fn test_duckdb_struct_literal() { key: Ident::with_quote('\'', "a"), value: Expr::Dictionary(vec![DictionaryField { key: Ident::with_quote('\'', "aa"), - value: Expr::Value(number("1")).into(), + value: Expr::value(number("1")).into(), }],) .into(), }],), @@ -594,16 +599,16 @@ fn test_duckdb_named_argument_function_with_assignment_operator() { args: vec![ FunctionArg::Named { name: Ident::new("a"), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "1".to_owned() - ))), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("1".to_owned())).with_empty_span() + )), operator: FunctionArgOperator::Assignment }, FunctionArg::Named { name: Ident::new("b"), - arg: FunctionArgExpr::Expr(Expr::Value(Value::SingleQuotedString( - "2".to_owned() - ))), + arg: FunctionArgExpr::Expr(Expr::Value( + (Value::SingleQuotedString("2".to_owned())).with_empty_span() + )), operator: FunctionArgOperator::Assignment }, ], @@ -632,14 +637,14 @@ fn test_array_index() { &Expr::CompoundFieldAccess { root: Box::new(Expr::Array(Array { elem: vec![ - Expr::Value(Value::SingleQuotedString("a".to_owned())), - Expr::Value(Value::SingleQuotedString("b".to_owned())), - Expr::Value(Value::SingleQuotedString("c".to_owned())) + Expr::Value((Value::SingleQuotedString("a".to_owned())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("b".to_owned())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("c".to_owned())).with_empty_span()) ], named: false })), access_chain: vec![AccessExpr::Subscript(Subscript::Index { - index: Expr::Value(number("3")) + index: Expr::value(number("3")) })] }, expr diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index b2b300aec..d7f3c014b 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -409,7 +409,8 @@ fn parse_create_function() { assert_eq!( function_body, Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( - Value::SingleQuotedString("org.random.class.Name".to_string()) + (Value::SingleQuotedString("org.random.class.Name".to_string())) + .with_empty_span() ))) ); assert_eq!( diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 7b1277599..ec565e50e 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -68,7 +68,7 @@ fn parse_table_time_travel() { args: None, with_hints: vec![], version: Some(TableVersion::ForSystemTimeAsOf(Expr::Value( - Value::SingleQuotedString(version) + (Value::SingleQuotedString(version)).with_empty_span() ))), partitions: vec![], with_ordinality: false, @@ -121,7 +121,9 @@ fn parse_create_procedure() { distinct: None, top: None, top_before_distinct: false, - projection: vec![SelectItem::UnnamedExpr(Expr::Value(number("1")))], + projection: vec![SelectItem::UnnamedExpr(Expr::Value( + (number("1")).with_empty_span() + ))], into: None, from: vec![], lateral_views: vec![], @@ -473,7 +475,9 @@ fn parse_mssql_top_paren() { let select = ms_and_generic().verified_only_select(sql); let top = select.top.unwrap(); assert_eq!( - Some(TopQuantity::Expr(Expr::Value(number("5")))), + Some(TopQuantity::Expr(Expr::Value( + (number("5")).with_empty_span() + ))), top.quantity ); assert!(!top.percent); @@ -485,7 +489,9 @@ fn parse_mssql_top_percent() { let select = ms_and_generic().verified_only_select(sql); let top = select.top.unwrap(); assert_eq!( - Some(TopQuantity::Expr(Expr::Value(number("5")))), + Some(TopQuantity::Expr(Expr::Value( + (number("5")).with_empty_span() + ))), top.quantity ); assert!(top.percent); @@ -497,7 +503,9 @@ fn parse_mssql_top_with_ties() { let select = ms_and_generic().verified_only_select(sql); let top = select.top.unwrap(); assert_eq!( - Some(TopQuantity::Expr(Expr::Value(number("5")))), + Some(TopQuantity::Expr(Expr::Value( + (number("5")).with_empty_span() + ))), top.quantity ); assert!(top.with_ties); @@ -509,7 +517,9 @@ fn parse_mssql_top_percent_with_ties() { let select = ms_and_generic().verified_only_select(sql); let top = select.top.unwrap(); assert_eq!( - Some(TopQuantity::Expr(Expr::Value(number("10")))), + Some(TopQuantity::Expr(Expr::Value( + (number("10")).with_empty_span() + ))), top.quantity ); assert!(top.percent); @@ -746,7 +756,10 @@ fn parse_mssql_json_object() { assert!(matches!( args[0], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::Function(_)), operator: FunctionArgOperator::Colon } @@ -762,7 +775,10 @@ fn parse_mssql_json_object() { assert!(matches!( args[2], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::Subquery(_)), operator: FunctionArgOperator::Colon } @@ -793,7 +809,10 @@ fn parse_mssql_json_object() { assert!(matches!( args[0], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::CompoundIdentifier(_)), operator: FunctionArgOperator::Colon } @@ -801,7 +820,10 @@ fn parse_mssql_json_object() { assert!(matches!( args[1], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::CompoundIdentifier(_)), operator: FunctionArgOperator::Colon } @@ -809,7 +831,10 @@ fn parse_mssql_json_object() { assert!(matches!( args[2], FunctionArg::ExprNamed { - name: Expr::Value(Value::SingleQuotedString(_)), + name: Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(_), + span: _ + }), arg: FunctionArgExpr::Expr(Expr::CompoundIdentifier(_)), operator: FunctionArgOperator::Colon } @@ -830,11 +855,17 @@ fn parse_mssql_json_array() { assert_eq!( &[ FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - Value::SingleQuotedString("a".into()) + (Value::SingleQuotedString("a".into())).with_empty_span() + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (number("1")).with_empty_span() + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (Value::Null).with_empty_span() + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (number("2")).with_empty_span() ))), - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("1")))), - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(Value::Null))), - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("2")))), ], &args[..] ); @@ -856,11 +887,17 @@ fn parse_mssql_json_array() { assert_eq!( &[ FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - Value::SingleQuotedString("a".into()) + (Value::SingleQuotedString("a".into())).with_empty_span() + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (number("1")).with_empty_span() + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (Value::Null).with_empty_span() + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (number("2")).with_empty_span() ))), - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("1")))), - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(Value::Null))), - FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("2")))), ], &args[..] ); @@ -915,7 +952,7 @@ fn parse_mssql_json_array() { }) => { assert_eq!( &FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - Value::SingleQuotedString("a".into()) + (Value::SingleQuotedString("a".into())).with_empty_span() ))), &args[0] ); @@ -942,7 +979,7 @@ fn parse_mssql_json_array() { }) => { assert_eq!( &FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( - Value::SingleQuotedString("a".into()) + (Value::SingleQuotedString("a".into())).with_empty_span() ))), &args[0] ); @@ -964,7 +1001,9 @@ fn parse_mssql_json_array() { .. }) => { assert_eq!( - &FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("1")))), + &FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + (number("1")).with_empty_span() + ))), &args[0] ); assert!(matches!( @@ -1042,15 +1081,15 @@ fn parse_convert() { unreachable!() }; assert!(!is_try); - assert_eq!(Expr::Value(number("1")), *expr); + assert_eq!(Expr::value(number("1")), *expr); assert_eq!(Some(DataType::Int(None)), data_type); assert!(charset.is_none()); assert!(target_before_value); assert_eq!( vec![ - Expr::Value(number("2")), - Expr::Value(number("3")), - Expr::Value(Value::Null), + Expr::value(number("2")), + Expr::value(number("3")), + Expr::Value((Value::Null).with_empty_span()), ], styles ); @@ -1089,8 +1128,12 @@ fn parse_substring_in_select() { quote_style: None, span: Span::empty(), })), - substring_from: Some(Box::new(Expr::Value(number("0")))), - substring_for: Some(Box::new(Expr::Value(number("1")))), + substring_from: Some(Box::new(Expr::Value( + (number("0")).with_empty_span() + ))), + substring_for: Some(Box::new(Expr::Value( + (number("1")).with_empty_span() + ))), special: true, })], into: None, @@ -1179,9 +1222,9 @@ fn parse_mssql_declare() { span: Span::empty(), }], data_type: Some(Text), - assignment: Some(MsSqlAssignment(Box::new(Expr::Value(SingleQuotedString( - "foobar".to_string() - ))))), + assignment: Some(MsSqlAssignment(Box::new(Expr::Value( + (SingleQuotedString("foobar".to_string())).with_empty_span() + )))), declare_type: None, binary: None, sensitive: None, @@ -1215,7 +1258,9 @@ fn parse_mssql_declare() { local: false, hivevar: false, variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("@bar")])), - value: vec![Expr::Value(Value::Number("2".parse().unwrap(), false))], + value: vec![Expr::Value( + (Value::Number("2".parse().unwrap(), false)).with_empty_span() + )], }, Statement::Query(Box::new(Query { with: None, @@ -1236,7 +1281,9 @@ fn parse_mssql_declare() { projection: vec![SelectItem::UnnamedExpr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("@bar"))), op: BinaryOperator::Multiply, - right: Box::new(Expr::Value(Value::Number("4".parse().unwrap(), false))), + right: Box::new(Expr::Value( + (Value::Number("4".parse().unwrap(), false)).with_empty_span() + )), })], into: None, from: vec![], @@ -1268,11 +1315,15 @@ fn test_parse_raiserror() { assert_eq!( s, Statement::RaisError { - message: Box::new(Expr::Value(Value::SingleQuotedString( - "This is a test".to_string() - ))), - severity: Box::new(Expr::Value(Value::Number("16".parse().unwrap(), false))), - state: Box::new(Expr::Value(Value::Number("1".parse().unwrap(), false))), + message: Box::new(Expr::Value( + (Value::SingleQuotedString("This is a test".to_string())).with_empty_span() + )), + severity: Box::new(Expr::Value( + (Value::Number("16".parse().unwrap(), false)).with_empty_span() + )), + state: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), arguments: vec![], options: vec![], } @@ -1347,7 +1398,7 @@ fn parse_create_table_with_valid_options() { SqlOption::Partition { column_name: "column_a".into(), range_direction: None, - for_values: vec![Expr::Value(test_utils::number("10")), Expr::Value(test_utils::number("11"))] , + for_values: vec![Expr::Value((test_utils::number("10")).with_empty_span()), Expr::Value((test_utils::number("11")).with_empty_span())] , }, ], ), @@ -1358,8 +1409,8 @@ fn parse_create_table_with_valid_options() { column_name: "column_a".into(), range_direction: Some(PartitionRangeDirection::Left), for_values: vec![ - Expr::Value(test_utils::number("10")), - Expr::Value(test_utils::number("11")), + Expr::Value((test_utils::number("10")).with_empty_span()), + Expr::Value((test_utils::number("11")).with_empty_span()), ], } ], @@ -1630,8 +1681,8 @@ fn parse_create_table_with_identity_column() { IdentityProperty { parameters: Some(IdentityPropertyFormatKind::FunctionCall( IdentityParameters { - seed: Expr::Value(number("1")), - increment: Expr::Value(number("1")), + seed: Expr::value(number("1")), + increment: Expr::value(number("1")), }, )), order: None, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 030710743..ad2987b28 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -52,11 +52,11 @@ fn parse_literal_string() { let select = mysql().verified_only_select(sql); assert_eq!(2, select.projection.len()); assert_eq!( - &Expr::Value(Value::SingleQuotedString("single".to_string())), + &Expr::Value((Value::SingleQuotedString("single".to_string())).with_empty_span()), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::DoubleQuotedString("double".to_string())), + &Expr::Value((Value::DoubleQuotedString("double".to_string())).with_empty_span()), expr_from_projection(&select.projection[1]) ); } @@ -621,7 +621,7 @@ fn parse_set_variables() { local: true, hivevar: false, variables: OneOrManyWithParens::One(ObjectName::from(vec!["autocommit".into()])), - value: vec![Expr::Value(number("1"))], + value: vec![Expr::value(number("1"))], } ); } @@ -986,7 +986,9 @@ fn parse_create_table_both_options_and_as_query() { assert_eq!(collation, Some("utf8mb4_0900_ai_ci".to_string())); assert_eq!( query.unwrap().body.as_select().unwrap().projection, - vec![SelectItem::UnnamedExpr(Expr::Value(number("1")))] + vec![SelectItem::UnnamedExpr(Expr::Value( + (number("1")).with_empty_span() + ))] ); } _ => unreachable!(), @@ -1413,18 +1415,25 @@ fn parse_simple_insert() { explicit_row: false, rows: vec![ vec![ - Expr::Value(Value::SingleQuotedString( - "Test Some Inserts".to_string() - )), - Expr::Value(number("1")) + Expr::Value( + (Value::SingleQuotedString("Test Some Inserts".to_string())) + .with_empty_span() + ), + Expr::value(number("1")) ], vec![ - Expr::Value(Value::SingleQuotedString("Test Entry 2".to_string())), - Expr::Value(number("2")) + Expr::Value( + (Value::SingleQuotedString("Test Entry 2".to_string())) + .with_empty_span() + ), + Expr::value(number("2")) ], vec![ - Expr::Value(Value::SingleQuotedString("Test Entry 3".to_string())), - Expr::Value(number("3")) + Expr::Value( + (Value::SingleQuotedString("Test Entry 3".to_string())) + .with_empty_span() + ), + Expr::value(number("3")) ] ] })), @@ -1471,8 +1480,11 @@ fn parse_ignore_insert() { body: Box::new(SetExpr::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(Value::SingleQuotedString("Test Some Inserts".to_string())), - Expr::Value(number("1")) + Expr::Value( + (Value::SingleQuotedString("Test Some Inserts".to_string())) + .with_empty_span() + ), + Expr::value(number("1")) ]] })), order_by: None, @@ -1518,8 +1530,11 @@ fn parse_priority_insert() { body: Box::new(SetExpr::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(Value::SingleQuotedString("Test Some Inserts".to_string())), - Expr::Value(number("1")) + Expr::Value( + (Value::SingleQuotedString("Test Some Inserts".to_string())) + .with_empty_span() + ), + Expr::value(number("1")) ]] })), order_by: None, @@ -1562,8 +1577,11 @@ fn parse_priority_insert() { body: Box::new(SetExpr::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(Value::SingleQuotedString("Test Some Inserts".to_string())), - Expr::Value(number("1")) + Expr::Value( + (Value::SingleQuotedString("Test Some Inserts".to_string())) + .with_empty_span() + ), + Expr::value(number("1")) ]] })), order_by: None, @@ -1611,9 +1629,9 @@ fn parse_insert_as() { with: None, body: Box::new(SetExpr::Values(Values { explicit_row: false, - rows: vec![vec![Expr::Value(Value::SingleQuotedString( - "2024-01-01".to_string() - ))]] + rows: vec![vec![Expr::Value( + (Value::SingleQuotedString("2024-01-01".to_string())).with_empty_span() + )]] })), order_by: None, limit: None, @@ -1672,8 +1690,11 @@ fn parse_insert_as() { body: Box::new(SetExpr::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(number("1")), - Expr::Value(Value::SingleQuotedString("2024-01-01".to_string())) + Expr::value(number("1")), + Expr::Value( + (Value::SingleQuotedString("2024-01-01".to_string())) + .with_empty_span() + ) ]] })), order_by: None, @@ -1720,8 +1741,11 @@ fn parse_replace_insert() { body: Box::new(SetExpr::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(Value::SingleQuotedString("Test Some Inserts".to_string())), - Expr::Value(number("1")) + Expr::Value( + (Value::SingleQuotedString("Test Some Inserts".to_string())) + .with_empty_span() + ), + Expr::value(number("1")) ]] })), order_by: None, @@ -1816,16 +1840,20 @@ fn parse_insert_with_on_duplicate_update() { body: Box::new(SetExpr::Values(Values { explicit_row: false, rows: vec![vec![ - Expr::Value(Value::SingleQuotedString( - "accounting_manager".to_string() - )), - Expr::Value(Value::SingleQuotedString( - "Some description about the group".to_string() - )), - Expr::Value(Value::Boolean(true)), - Expr::Value(Value::Boolean(true)), - Expr::Value(Value::Boolean(true)), - Expr::Value(Value::Boolean(true)), + Expr::Value( + (Value::SingleQuotedString("accounting_manager".to_string())) + .with_empty_span() + ), + Expr::Value( + (Value::SingleQuotedString( + "Some description about the group".to_string() + )) + .with_empty_span() + ), + Expr::Value((Value::Boolean(true)).with_empty_span()), + Expr::Value((Value::Boolean(true)).with_empty_span()), + Expr::Value((Value::Boolean(true)).with_empty_span()), + Expr::Value((Value::Boolean(true)).with_empty_span()), ]] })), order_by: None, @@ -1946,7 +1974,7 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { top: None, top_before_distinct: false, projection: vec![ - SelectItem::UnnamedExpr(Expr::Value(number("123e4"))), + SelectItem::UnnamedExpr(Expr::value(number("123e4"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("123col_$@123abc"))) ], into: None, @@ -2063,7 +2091,7 @@ fn parse_update_with_joins() { Ident::new("o"), Ident::new("completed") ])), - value: Expr::Value(Value::Boolean(true)) + value: Expr::Value((Value::Boolean(true)).with_empty_span()) }], assignments ); @@ -2074,7 +2102,9 @@ fn parse_update_with_joins() { Ident::new("firstname") ])), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString("Peter".to_string()))) + right: Box::new(Expr::Value( + (Value::SingleQuotedString("Peter".to_string())).with_empty_span() + )) }), selection ); @@ -2114,7 +2144,7 @@ fn parse_delete_with_limit() { let sql = "DELETE FROM customers LIMIT 100"; match mysql().verified_stmt(sql) { Statement::Delete(Delete { limit, .. }) => { - assert_eq!(Some(Expr::Value(number("100"))), limit); + assert_eq!(Some(Expr::value(number("100"))), limit); } _ => unreachable!(), } @@ -2462,8 +2492,12 @@ fn parse_substring_in_select() { quote_style: None, span: Span::empty(), })), - substring_from: Some(Box::new(Expr::Value(number("0")))), - substring_for: Some(Box::new(Expr::Value(number("1")))), + substring_from: Some(Box::new(Expr::Value( + (number("0")).with_empty_span() + ))), + substring_for: Some(Box::new(Expr::Value( + (number("1")).with_empty_span() + ))), special: true, })], into: None, @@ -2937,7 +2971,7 @@ fn parse_json_table() { .from[0] .relation, TableFactor::JsonTable { - json_expr: Expr::Value(Value::SingleQuotedString("[1,2]".to_string())), + json_expr: Expr::Value((Value::SingleQuotedString("[1,2]".to_string())).with_empty_span()), json_path: Value::SingleQuotedString("$[*]".to_string()), columns: vec![ JsonTableColumn::Named(JsonTableNamedColumn { @@ -2975,33 +3009,33 @@ fn parse_logical_xor() { let select = mysql_and_generic().verified_only_select(sql); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::Boolean(true))), + left: Box::new(Expr::Value((Value::Boolean(true)).with_empty_span())), op: BinaryOperator::Xor, - right: Box::new(Expr::Value(Value::Boolean(true))), + right: Box::new(Expr::Value((Value::Boolean(true)).with_empty_span())), }), select.projection[0] ); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::Boolean(false))), + left: Box::new(Expr::Value((Value::Boolean(false)).with_empty_span())), op: BinaryOperator::Xor, - right: Box::new(Expr::Value(Value::Boolean(false))), + right: Box::new(Expr::Value((Value::Boolean(false)).with_empty_span())), }), select.projection[1] ); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::Boolean(true))), + left: Box::new(Expr::Value((Value::Boolean(true)).with_empty_span())), op: BinaryOperator::Xor, - right: Box::new(Expr::Value(Value::Boolean(false))), + right: Box::new(Expr::Value((Value::Boolean(false)).with_empty_span())), }), select.projection[2] ); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::Boolean(false))), + left: Box::new(Expr::Value((Value::Boolean(false)).with_empty_span())), op: BinaryOperator::Xor, - right: Box::new(Expr::Value(Value::Boolean(true))), + right: Box::new(Expr::Value((Value::Boolean(true)).with_empty_span())), }), select.projection[3] ); @@ -3013,7 +3047,7 @@ fn parse_bitstring_literal() { assert_eq!( select.projection, vec![SelectItem::UnnamedExpr(Expr::Value( - Value::SingleQuotedByteStringLiteral("111".to_string()) + (Value::SingleQuotedByteStringLiteral("111".to_string())).with_empty_span() ))] ); } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 312ce1186..22558329d 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -436,7 +436,9 @@ fn parse_create_table_with_defaults() { options: vec![ ColumnOptionDef { name: None, - option: ColumnOption::Default(Expr::Value(Value::Boolean(true))), + option: ColumnOption::Default(Expr::Value( + (Value::Boolean(true)).with_empty_span() + )), }, ColumnOptionDef { name: None, @@ -488,15 +490,15 @@ fn parse_create_table_with_defaults() { vec![ SqlOption::KeyValue { key: "fillfactor".into(), - value: Expr::Value(number("20")) + value: Expr::value(number("20")) }, SqlOption::KeyValue { key: "user_catalog_table".into(), - value: Expr::Value(Value::Boolean(true)) + value: Expr::Value((Value::Boolean(true)).with_empty_span()) }, SqlOption::KeyValue { key: "autovacuum_vacuum_threshold".into(), - value: Expr::Value(number("100")) + value: Expr::value(number("100")) }, ] ); @@ -768,7 +770,8 @@ fn parse_alter_table_alter_column() { ) { AlterTableOperation::AlterColumn { column_name, op } => { assert_eq!("is_active", column_name.to_string()); - let using_expr = Expr::Value(Value::SingleQuotedString("text".to_string())); + let using_expr = + Expr::Value(Value::SingleQuotedString("text".to_string()).with_empty_span()); assert_eq!( op, AlterColumnOperation::SetDataType { @@ -1280,7 +1283,7 @@ fn parse_copy_to() { top_before_distinct: false, projection: vec![ SelectItem::ExprWithAlias { - expr: Expr::Value(number("42")), + expr: Expr::value(number("42")), alias: Ident { value: "a".into(), quote_style: None, @@ -1288,7 +1291,9 @@ fn parse_copy_to() { }, }, SelectItem::ExprWithAlias { - expr: Expr::Value(Value::SingleQuotedString("hello".into())), + expr: Expr::Value( + (Value::SingleQuotedString("hello".into())).with_empty_span() + ), alias: Ident { value: "b".into(), quote_style: None, @@ -1446,7 +1451,9 @@ fn parse_set() { local: false, hivevar: false, variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::Value(Value::SingleQuotedString("b".into()))], + value: vec![Expr::Value( + (Value::SingleQuotedString("b".into())).with_empty_span() + )], } ); @@ -1457,7 +1464,7 @@ fn parse_set() { local: false, hivevar: false, variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::Value(number("0"))], + value: vec![Expr::value(number("0"))], } ); @@ -1518,7 +1525,7 @@ fn parse_set() { Ident::new("reducer"), Ident::new("parallelism") ])), - value: vec![Expr::Value(Value::Boolean(false))], + value: vec![Expr::Value((Value::Boolean(false)).with_empty_span())], } ); @@ -1670,8 +1677,8 @@ fn parse_execute() { Statement::Execute { name: Some(ObjectName::from(vec!["a".into()])), parameters: vec![ - Expr::Value(number("1")), - Expr::Value(Value::SingleQuotedString("t".to_string())) + Expr::value(number("1")), + Expr::Value((Value::SingleQuotedString("t".to_string())).with_empty_span()) ], has_parentheses: true, using: vec![], @@ -1692,7 +1699,9 @@ fn parse_execute() { ExprWithAlias { expr: Expr::Cast { kind: CastKind::Cast, - expr: Box::new(Expr::Value(Value::Number("1337".parse().unwrap(), false))), + expr: Box::new(Expr::Value( + (Value::Number("1337".parse().unwrap(), false)).with_empty_span() + )), data_type: DataType::SmallInt(None), format: None }, @@ -1701,7 +1710,9 @@ fn parse_execute() { ExprWithAlias { expr: Expr::Cast { kind: CastKind::Cast, - expr: Box::new(Expr::Value(Value::Number("7331".parse().unwrap(), false))), + expr: Box::new(Expr::Value( + (Value::Number("7331".parse().unwrap(), false)).with_empty_span() + )), data_type: DataType::SmallInt(None), format: None }, @@ -1899,7 +1910,9 @@ fn parse_pg_on_conflict() { target: AssignmentTarget::ColumnName(ObjectName::from( vec!["dname".into()] )), - value: Expr::Value(Value::Placeholder("$1".to_string())) + value: Expr::Value( + (Value::Placeholder("$1".to_string())).with_empty_span() + ) },], selection: Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident { @@ -1908,7 +1921,9 @@ fn parse_pg_on_conflict() { span: Span::empty(), })), op: BinaryOperator::Gt, - right: Box::new(Expr::Value(Value::Placeholder("$2".to_string()))) + right: Box::new(Expr::Value( + (Value::Placeholder("$2".to_string())).with_empty_span() + )) }) }), action @@ -1942,7 +1957,9 @@ fn parse_pg_on_conflict() { target: AssignmentTarget::ColumnName(ObjectName::from( vec!["dname".into()] )), - value: Expr::Value(Value::Placeholder("$1".to_string())) + value: Expr::Value( + (Value::Placeholder("$1".to_string())).with_empty_span() + ) },], selection: Some(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident { @@ -1951,7 +1968,9 @@ fn parse_pg_on_conflict() { span: Span::empty(), })), op: BinaryOperator::Gt, - right: Box::new(Expr::Value(Value::Placeholder("$2".to_string()))) + right: Box::new(Expr::Value( + (Value::Placeholder("$2".to_string())).with_empty_span() + )) }) }), action @@ -2167,9 +2186,13 @@ fn parse_pg_regex_match_ops() { let select = pg().verified_only_select(&format!("SELECT 'abc' {} '^a'", &str_op)); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::SingleQuotedString("abc".into()))), + left: Box::new(Expr::Value( + (Value::SingleQuotedString("abc".into())).with_empty_span() + )), op: op.clone(), - right: Box::new(Expr::Value(Value::SingleQuotedString("^a".into()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("^a".into())).with_empty_span() + )), }), select.projection[0] ); @@ -2189,9 +2212,13 @@ fn parse_pg_like_match_ops() { let select = pg().verified_only_select(&format!("SELECT 'abc' {} 'a_c%'", &str_op)); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value(Value::SingleQuotedString("abc".into()))), + left: Box::new(Expr::Value( + (Value::SingleQuotedString("abc".into())).with_empty_span() + )), op: op.clone(), - right: Box::new(Expr::Value(Value::SingleQuotedString("a_c%".into()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("a_c%".into())).with_empty_span() + )), }), select.projection[0] ); @@ -2201,7 +2228,7 @@ fn parse_pg_like_match_ops() { #[test] fn parse_array_index_expr() { let num: Vec = (0..=10) - .map(|s| Expr::Value(number(&s.to_string()))) + .map(|s| Expr::Value(number(&s.to_string()).with_empty_span())) .collect(); let sql = "SELECT foo[0] FROM foos"; @@ -2312,7 +2339,7 @@ fn parse_array_subscript() { ( "(ARRAY[1, 2, 3, 4, 5, 6])[2]", Subscript::Index { - index: Expr::Value(number("2")), + index: Expr::value(number("2")), }, ), ( @@ -2324,17 +2351,17 @@ fn parse_array_subscript() { ( "(ARRAY[1, 2, 3, 4, 5, 6])[2:5]", Subscript::Slice { - lower_bound: Some(Expr::Value(number("2"))), - upper_bound: Some(Expr::Value(number("5"))), + lower_bound: Some(Expr::value(number("2"))), + upper_bound: Some(Expr::value(number("5"))), stride: None, }, ), ( "(ARRAY[1, 2, 3, 4, 5, 6])[2:5:3]", Subscript::Slice { - lower_bound: Some(Expr::Value(number("2"))), - upper_bound: Some(Expr::Value(number("5"))), - stride: Some(Expr::Value(number("3"))), + lower_bound: Some(Expr::value(number("2"))), + upper_bound: Some(Expr::value(number("5"))), + stride: Some(Expr::value(number("3"))), }, ), ( @@ -2343,12 +2370,12 @@ fn parse_array_subscript() { lower_bound: Some(Expr::BinaryOp { left: Box::new(call("array_length", [Expr::Identifier(Ident::new("arr"))])), op: BinaryOperator::Minus, - right: Box::new(Expr::Value(number("3"))), + right: Box::new(Expr::value(number("3"))), }), upper_bound: Some(Expr::BinaryOp { left: Box::new(call("array_length", [Expr::Identifier(Ident::new("arr"))])), op: BinaryOperator::Minus, - right: Box::new(Expr::Value(number("1"))), + right: Box::new(Expr::value(number("1"))), }), stride: None, }, @@ -2357,14 +2384,14 @@ fn parse_array_subscript() { "(ARRAY[1, 2, 3, 4, 5, 6])[:5]", Subscript::Slice { lower_bound: None, - upper_bound: Some(Expr::Value(number("5"))), + upper_bound: Some(Expr::value(number("5"))), stride: None, }, ), ( "(ARRAY[1, 2, 3, 4, 5, 6])[2:]", Subscript::Slice { - lower_bound: Some(Expr::Value(number("2"))), + lower_bound: Some(Expr::value(number("2"))), upper_bound: None, stride: None, }, @@ -2400,19 +2427,19 @@ fn parse_array_multi_subscript() { root: Box::new(call( "make_array", vec![ - Expr::Value(number("1")), - Expr::Value(number("2")), - Expr::Value(number("3")) + Expr::value(number("1")), + Expr::value(number("2")), + Expr::value(number("3")) ] )), access_chain: vec![ AccessExpr::Subscript(Subscript::Slice { - lower_bound: Some(Expr::Value(number("1"))), - upper_bound: Some(Expr::Value(number("2"))), + lower_bound: Some(Expr::value(number("1"))), + upper_bound: Some(Expr::value(number("2"))), stride: None, }), AccessExpr::Subscript(Subscript::Index { - index: Expr::Value(number("2")), + index: Expr::value(number("2")), }), ], }, @@ -2655,7 +2682,9 @@ fn parse_array_subquery_expr() { distinct: None, top: None, top_before_distinct: false, - projection: vec![SelectItem::UnnamedExpr(Expr::Value(number("1")))], + projection: vec![SelectItem::UnnamedExpr(Expr::Value( + (number("1")).with_empty_span() + ))], into: None, from: vec![], lateral_views: vec![], @@ -2678,7 +2707,9 @@ fn parse_array_subquery_expr() { distinct: None, top: None, top_before_distinct: false, - projection: vec![SelectItem::UnnamedExpr(Expr::Value(number("2")))], + projection: vec![SelectItem::UnnamedExpr(Expr::Value( + (number("2")).with_empty_span() + ))], into: None, from: vec![], lateral_views: vec![], @@ -2750,7 +2781,9 @@ fn test_json() { SelectItem::UnnamedExpr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("params"))), op: BinaryOperator::LongArrow, - right: Box::new(Expr::Value(Value::SingleQuotedString("name".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("name".to_string())).with_empty_span() + )), }), select.projection[0] ); @@ -2761,7 +2794,9 @@ fn test_json() { SelectItem::UnnamedExpr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("params"))), op: BinaryOperator::Arrow, - right: Box::new(Expr::Value(Value::SingleQuotedString("name".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("name".to_string())).with_empty_span() + )), }), select.projection[0] ); @@ -2773,12 +2808,14 @@ fn test_json() { left: Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("info"))), op: BinaryOperator::Arrow, - right: Box::new(Expr::Value(Value::SingleQuotedString("items".to_string()))) + right: Box::new(Expr::Value( + (Value::SingleQuotedString("items".to_string())).with_empty_span() + )) }), op: BinaryOperator::LongArrow, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "product".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("product".to_string())).with_empty_span() + )), }), select.projection[0] ); @@ -2790,7 +2827,7 @@ fn test_json() { SelectItem::UnnamedExpr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("obj"))), op: BinaryOperator::Arrow, - right: Box::new(Expr::Value(number("42"))), + right: Box::new(Expr::value(number("42"))), }), select.projection[0] ); @@ -2815,9 +2852,9 @@ fn test_json() { left: Box::new(Expr::Identifier(Ident::new("obj"))), op: BinaryOperator::Arrow, right: Box::new(Expr::BinaryOp { - left: Box::new(Expr::Value(number("3"))), + left: Box::new(Expr::value(number("3"))), op: BinaryOperator::Multiply, - right: Box::new(Expr::Value(number("2"))), + right: Box::new(Expr::value(number("2"))), }), }), select.projection[0] @@ -2829,9 +2866,9 @@ fn test_json() { SelectItem::UnnamedExpr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("info"))), op: BinaryOperator::HashArrow, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "{a,b,c}".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("{a,b,c}".to_string())).with_empty_span() + )), }), select.projection[0] ); @@ -2842,9 +2879,9 @@ fn test_json() { SelectItem::UnnamedExpr(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("info"))), op: BinaryOperator::HashLongArrow, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "{a,b,c}".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("{a,b,c}".to_string())).with_empty_span() + )), }), select.projection[0] ); @@ -2855,9 +2892,9 @@ fn test_json() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("info"))), op: BinaryOperator::AtArrow, - right: Box::new(Expr::Value(Value::SingleQuotedString( - "{\"a\": 1}".to_string() - ))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("{\"a\": 1}".to_string())).with_empty_span() + )), }, select.selection.unwrap(), ); @@ -2866,9 +2903,9 @@ fn test_json() { let select = pg().verified_only_select(sql); assert_eq!( Expr::BinaryOp { - left: Box::new(Expr::Value(Value::SingleQuotedString( - "{\"a\": 1}".to_string() - ))), + left: Box::new(Expr::Value( + (Value::SingleQuotedString("{\"a\": 1}".to_string())).with_empty_span() + )), op: BinaryOperator::ArrowAt, right: Box::new(Expr::Identifier(Ident::new("info"))), }, @@ -2883,8 +2920,8 @@ fn test_json() { op: BinaryOperator::HashMinus, right: Box::new(Expr::Array(Array { elem: vec![ - Expr::Value(Value::SingleQuotedString("a".to_string())), - Expr::Value(Value::SingleQuotedString("b".to_string())), + Expr::Value((Value::SingleQuotedString("a".to_string())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("b".to_string())).with_empty_span()), ], named: true, })), @@ -2898,7 +2935,9 @@ fn test_json() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::from("info"))), op: BinaryOperator::AtQuestion, - right: Box::new(Expr::Value(Value::SingleQuotedString("$.a".to_string())),), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("$.a".to_string())).with_empty_span() + ),), }, select.selection.unwrap(), ); @@ -2909,7 +2948,9 @@ fn test_json() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::from("info"))), op: BinaryOperator::AtAt, - right: Box::new(Expr::Value(Value::SingleQuotedString("$.a".to_string())),), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("$.a".to_string())).with_empty_span() + ),), }, select.selection.unwrap(), ); @@ -2920,7 +2961,9 @@ fn test_json() { Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("info"))), op: BinaryOperator::Question, - right: Box::new(Expr::Value(Value::SingleQuotedString("b".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("b".to_string())).with_empty_span() + )), }, select.selection.unwrap(), ); @@ -2933,8 +2976,8 @@ fn test_json() { op: BinaryOperator::QuestionAnd, right: Box::new(Expr::Array(Array { elem: vec![ - Expr::Value(Value::SingleQuotedString("b".to_string())), - Expr::Value(Value::SingleQuotedString("c".to_string())) + Expr::Value((Value::SingleQuotedString("b".to_string())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("c".to_string())).with_empty_span()) ], named: true })) @@ -2950,8 +2993,8 @@ fn test_json() { op: BinaryOperator::QuestionPipe, right: Box::new(Expr::Array(Array { elem: vec![ - Expr::Value(Value::SingleQuotedString("b".to_string())), - Expr::Value(Value::SingleQuotedString("c".to_string())) + Expr::Value((Value::SingleQuotedString("b".to_string())).with_empty_span()), + Expr::Value((Value::SingleQuotedString("c".to_string())).with_empty_span()) ], named: true })) @@ -3025,7 +3068,7 @@ fn test_composite_value() { access_chain: vec![AccessExpr::Dot(Expr::Identifier(Ident::new("price")))] }), op: BinaryOperator::Gt, - right: Box::new(Expr::Value(number("9"))) + right: Box::new(Expr::value(number("9"))) } ); @@ -3045,8 +3088,12 @@ fn test_composite_value() { args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Array( Array { elem: vec![ - Expr::Value(Value::SingleQuotedString("i".to_string())), - Expr::Value(Value::SingleQuotedString("i".to_string())), + Expr::Value( + (Value::SingleQuotedString("i".to_string())).with_empty_span() + ), + Expr::Value( + (Value::SingleQuotedString("i".to_string())).with_empty_span() + ), ], named: true } @@ -3106,27 +3153,27 @@ fn parse_escaped_literal_string() { let select = pg_and_generic().verified_only_select(sql); assert_eq!(6, select.projection.len()); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("s1 \n s1".to_string())), + &Expr::Value((Value::EscapedStringLiteral("s1 \n s1".to_string())).with_empty_span()), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("s2 \\n s2".to_string())), + &Expr::Value((Value::EscapedStringLiteral("s2 \\n s2".to_string())).with_empty_span()), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("s3 \\\n s3".to_string())), + &Expr::Value((Value::EscapedStringLiteral("s3 \\\n s3".to_string())).with_empty_span()), expr_from_projection(&select.projection[2]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("s4 \\\\n s4".to_string())), + &Expr::Value((Value::EscapedStringLiteral("s4 \\\\n s4".to_string())).with_empty_span()), expr_from_projection(&select.projection[3]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("'".to_string())), + &Expr::Value((Value::EscapedStringLiteral("'".to_string())).with_empty_span()), expr_from_projection(&select.projection[4]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("foo \\".to_string())), + &Expr::Value((Value::EscapedStringLiteral("foo \\".to_string())).with_empty_span()), expr_from_projection(&select.projection[5]) ); @@ -3144,31 +3191,31 @@ fn parse_escaped_literal_string() { let select = pg_and_generic().verified_only_select_with_canonical(sql, canonical); assert_eq!(7, select.projection.len()); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("\u{0001}".to_string())), + &Expr::Value((Value::EscapedStringLiteral("\u{0001}".to_string())).with_empty_span()), expr_from_projection(&select.projection[0]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("\u{10ffff}".to_string())), + &Expr::Value((Value::EscapedStringLiteral("\u{10ffff}".to_string())).with_empty_span()), expr_from_projection(&select.projection[1]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("\u{000c}".to_string())), + &Expr::Value((Value::EscapedStringLiteral("\u{000c}".to_string())).with_empty_span()), expr_from_projection(&select.projection[2]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("%".to_string())), + &Expr::Value((Value::EscapedStringLiteral("%".to_string())).with_empty_span()), expr_from_projection(&select.projection[3]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("\u{0002}".to_string())), + &Expr::Value((Value::EscapedStringLiteral("\u{0002}".to_string())).with_empty_span()), expr_from_projection(&select.projection[4]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("%".to_string())), + &Expr::Value((Value::EscapedStringLiteral("%".to_string())).with_empty_span()), expr_from_projection(&select.projection[5]) ); assert_eq!( - &Expr::Value(Value::EscapedStringLiteral("%".to_string())), + &Expr::Value((Value::EscapedStringLiteral("%".to_string())).with_empty_span()), expr_from_projection(&select.projection[6]) ); @@ -3310,7 +3357,9 @@ fn parse_custom_operator() { "pg_catalog".into(), "~".into() ]), - right: Box::new(Expr::Value(Value::SingleQuotedString("^(table)$".into()))) + right: Box::new(Expr::Value( + (Value::SingleQuotedString("^(table)$".into())).with_empty_span() + )) }) ); @@ -3326,7 +3375,9 @@ fn parse_custom_operator() { span: Span::empty(), })), op: BinaryOperator::PGCustomBinaryOperator(vec!["pg_catalog".into(), "~".into()]), - right: Box::new(Expr::Value(Value::SingleQuotedString("^(table)$".into()))) + right: Box::new(Expr::Value( + (Value::SingleQuotedString("^(table)$".into())).with_empty_span() + )) }) ); @@ -3342,7 +3393,9 @@ fn parse_custom_operator() { span: Span::empty(), })), op: BinaryOperator::PGCustomBinaryOperator(vec!["~".into()]), - right: Box::new(Expr::Value(Value::SingleQuotedString("^(table)$".into()))) + right: Box::new(Expr::Value( + (Value::SingleQuotedString("^(table)$".into())).with_empty_span() + )) }) ); } @@ -3428,9 +3481,9 @@ fn parse_create_role() { assert_eq!(*bypassrls, Some(true)); assert_eq!( *password, - Some(Password::Password(Expr::Value(Value::SingleQuotedString( - "abcdef".into() - )))) + Some(Password::Password(Expr::Value( + (Value::SingleQuotedString("abcdef".into())).with_empty_span() + ))) ); assert_eq!(*superuser, Some(true)); assert_eq!(*create_db, Some(false)); @@ -3439,7 +3492,9 @@ fn parse_create_role() { assert_eq!(*connection_limit, None); assert_eq!( *valid_until, - Some(Expr::Value(Value::SingleQuotedString("2025-01-01".into()))) + Some(Expr::Value( + (Value::SingleQuotedString("2025-01-01".into())).with_empty_span() + )) ); assert_eq_vec(&["role1", "role2"], in_role); assert!(in_group.is_empty()); @@ -3521,13 +3576,15 @@ fn parse_alter_role() { RoleOption::Login(true), RoleOption::Replication(true), RoleOption::BypassRLS(true), - RoleOption::ConnectionLimit(Expr::Value(number("100"))), + RoleOption::ConnectionLimit(Expr::value(number("100"))), RoleOption::Password({ - Password::Password(Expr::Value(Value::SingleQuotedString("abcdef".into()))) + Password::Password(Expr::Value( + (Value::SingleQuotedString("abcdef".into())).with_empty_span(), + )) }), - RoleOption::ValidUntil(Expr::Value(Value::SingleQuotedString( - "2025-01-01".into(), - ))) + RoleOption::ValidUntil(Expr::Value( + (Value::SingleQuotedString("2025-01-01".into(),)).with_empty_span() + )) ] }, } @@ -3593,7 +3650,9 @@ fn parse_alter_role() { quote_style: None, span: Span::empty(), }]), - config_value: SetConfigValue::Value(Expr::Value(number("100000"))), + config_value: SetConfigValue::Value(Expr::Value( + (number("100000")).with_empty_span() + )), in_database: Some(ObjectName::from(vec![Ident { value: "database_name".into(), quote_style: None, @@ -3618,7 +3677,9 @@ fn parse_alter_role() { quote_style: None, span: Span::empty(), }]), - config_value: SetConfigValue::Value(Expr::Value(number("100000"))), + config_value: SetConfigValue::Value(Expr::Value( + (number("100000")).with_empty_span() + )), in_database: Some(ObjectName::from(vec![Ident { value: "database_name".into(), quote_style: None, @@ -3799,7 +3860,7 @@ fn parse_create_function() { called_on_null: Some(FunctionCalledOnNull::Strict), parallel: Some(FunctionParallel::Safe), function_body: Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( - Value::SingleQuotedString("select $1 + $2;".into()) + (Value::SingleQuotedString("select $1 + $2;".into())).with_empty_span() ))), if_not_exists: false, using: None, @@ -3862,7 +3923,9 @@ fn parse_drop_function() { mode: Some(ArgMode::In), name: Some("b".into()), data_type: DataType::Integer(None), - default_expr: Some(Expr::Value(Value::Number("1".parse().unwrap(), false))), + default_expr: Some(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), } ]), }], @@ -3888,10 +3951,9 @@ fn parse_drop_function() { mode: Some(ArgMode::In), name: Some("b".into()), data_type: DataType::Integer(None), - default_expr: Some(Expr::Value(Value::Number( - "1".parse().unwrap(), - false - ))), + default_expr: Some(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), } ]), }, @@ -3907,10 +3969,9 @@ fn parse_drop_function() { mode: Some(ArgMode::In), name: Some("b".into()), data_type: DataType::Integer(None), - default_expr: Some(Expr::Value(Value::Number( - "1".parse().unwrap(), - false - ))), + default_expr: Some(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), } ]), } @@ -3956,7 +4017,9 @@ fn parse_drop_procedure() { mode: Some(ArgMode::In), name: Some("b".into()), data_type: DataType::Integer(None), - default_expr: Some(Expr::Value(Value::Number("1".parse().unwrap(), false))), + default_expr: Some(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), } ]), }], @@ -3982,10 +4045,9 @@ fn parse_drop_procedure() { mode: Some(ArgMode::In), name: Some("b".into()), data_type: DataType::Integer(None), - default_expr: Some(Expr::Value(Value::Number( - "1".parse().unwrap(), - false - ))), + default_expr: Some(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), } ]), }, @@ -4001,10 +4063,9 @@ fn parse_drop_procedure() { mode: Some(ArgMode::In), name: Some("b".into()), data_type: DataType::Integer(None), - default_expr: Some(Expr::Value(Value::Number( - "1".parse().unwrap(), - false - ))), + default_expr: Some(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), } ]), } @@ -4041,36 +4102,48 @@ fn parse_dollar_quoted_string() { }; assert_eq!( - &Expr::Value(Value::DollarQuotedString(DollarQuotedString { - tag: None, - value: "hello".into() - })), + &Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + tag: None, + value: "hello".into() + })) + .with_empty_span() + ), expr_from_projection(&projection[0]) ); assert_eq!( - &Expr::Value(Value::DollarQuotedString(DollarQuotedString { - tag: Some("tag_name".into()), - value: "world".into() - })), + &Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + tag: Some("tag_name".into()), + value: "world".into() + })) + .with_empty_span() + ), expr_from_projection(&projection[1]) ); assert_eq!( - &Expr::Value(Value::DollarQuotedString(DollarQuotedString { - tag: None, - value: "Foo$Bar".into() - })), + &Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + tag: None, + value: "Foo$Bar".into() + })) + .with_empty_span() + ), expr_from_projection(&projection[2]) ); assert_eq!( projection[3], SelectItem::ExprWithAlias { - expr: Expr::Value(Value::DollarQuotedString(DollarQuotedString { - tag: None, - value: "Foo$Bar".into(), - })), + expr: Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + tag: None, + value: "Foo$Bar".into(), + })) + .with_empty_span() + ), alias: Ident { value: "col_name".into(), quote_style: None, @@ -4081,18 +4154,24 @@ fn parse_dollar_quoted_string() { assert_eq!( expr_from_projection(&projection[4]), - &Expr::Value(Value::DollarQuotedString(DollarQuotedString { - tag: None, - value: "".into() - })), + &Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + tag: None, + value: "".into() + })) + .with_empty_span() + ), ); assert_eq!( expr_from_projection(&projection[5]), - &Expr::Value(Value::DollarQuotedString(DollarQuotedString { - tag: Some("tag_name".into()), - value: "".into() - })), + &Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + tag: Some("tag_name".into()), + value: "".into() + })) + .with_empty_span() + ), ); } @@ -4438,7 +4517,7 @@ fn test_simple_postgres_insert_with_alias() { explicit_row: false, rows: vec![vec![ Expr::Identifier(Ident::new("DEFAULT")), - Expr::Value(Value::Number("123".to_string(), false)) + Expr::Value((Value::Number("123".to_string(), false)).with_empty_span()) ]] })), order_by: None, @@ -4508,10 +4587,10 @@ fn test_simple_postgres_insert_with_alias() { explicit_row: false, rows: vec![vec![ Expr::Identifier(Ident::new("DEFAULT")), - Expr::Value(Value::Number( - bigdecimal::BigDecimal::new(123.into(), 0), - false - )) + Expr::Value( + (Value::Number(bigdecimal::BigDecimal::new(123.into(), 0), false)) + .with_empty_span() + ) ]] })), order_by: None, @@ -4580,7 +4659,9 @@ fn test_simple_insert_with_quoted_alias() { explicit_row: false, rows: vec![vec![ Expr::Identifier(Ident::new("DEFAULT")), - Expr::Value(Value::SingleQuotedString("0123".to_string())) + Expr::Value( + (Value::SingleQuotedString("0123".to_string())).with_empty_span() + ) ]] })), order_by: None, @@ -4650,18 +4731,18 @@ fn parse_at_time_zone() { }), time_zone: Box::new(Expr::Cast { kind: CastKind::DoubleColon, - expr: Box::new(Expr::Value(Value::SingleQuotedString( - "America/Los_Angeles".to_owned(), - ))), + expr: Box::new(Expr::Value( + Value::SingleQuotedString("America/Los_Angeles".to_owned()).with_empty_span(), + )), data_type: DataType::Text, format: None, }), }), op: BinaryOperator::Plus, right: Box::new(Expr::Interval(Interval { - value: Box::new(Expr::Value(Value::SingleQuotedString( - "23 hours".to_owned(), - ))), + value: Box::new(Expr::Value( + Value::SingleQuotedString("23 hours".to_owned()).with_empty_span(), + )), leading_field: None, leading_precision: None, last_field: None, @@ -4685,11 +4766,13 @@ fn parse_create_table_with_options() { vec![ SqlOption::KeyValue { key: "foo".into(), - value: Expr::Value(Value::SingleQuotedString("bar".into())), + value: Expr::Value( + (Value::SingleQuotedString("bar".into())).with_empty_span() + ), }, SqlOption::KeyValue { key: "a".into(), - value: Expr::Value(number("123")), + value: Expr::value(number("123")), }, ], with_options @@ -4735,7 +4818,10 @@ fn test_table_unnest_with_ordinality() { #[test] fn test_escaped_string_literal() { match pg().verified_expr(r#"E'\n'"#) { - Expr::Value(Value::EscapedStringLiteral(s)) => { + Expr::Value(ValueWithSpan { + value: Value::EscapedStringLiteral(s), + span: _, + }) => { assert_eq!("\n", s); } _ => unreachable!(), @@ -4790,7 +4876,7 @@ fn parse_create_after_update_trigger_with_condition() { Ident::new("balance"), ])), op: BinaryOperator::Gt, - right: Box::new(Expr::Value(number("10000"))), + right: Box::new(Expr::value(number("10000"))), }))), exec_body: TriggerExecBody { exec_type: TriggerExecBodyType::Function, @@ -5155,7 +5241,7 @@ fn parse_trigger_related_functions() { return_type: Some(DataType::Trigger), function_body: Some( CreateFunctionBody::AsBeforeOptions( - Expr::Value( + Expr::Value(( Value::DollarQuotedString( DollarQuotedString { value: "\n BEGIN\n -- Check that empname and salary are given\n IF NEW.empname IS NULL THEN\n RAISE EXCEPTION 'empname cannot be null';\n END IF;\n IF NEW.salary IS NULL THEN\n RAISE EXCEPTION '% cannot have null salary', NEW.empname;\n END IF;\n\n -- Who works for us when they must pay for it?\n IF NEW.salary < 0 THEN\n RAISE EXCEPTION '% cannot have a negative salary', NEW.empname;\n END IF;\n\n -- Remember who changed the payroll when\n NEW.last_date := current_timestamp;\n NEW.last_user := current_user;\n RETURN NEW;\n END;\n ".to_owned(), @@ -5163,8 +5249,8 @@ fn parse_trigger_related_functions() { "emp_stamp".to_owned(), ), }, - ), - ), + ) + ).with_empty_span()), ), ), behavior: None, @@ -5231,7 +5317,10 @@ fn test_unicode_string_literal() { ]; for (input, expected) in pairs { match pg_and_generic().verified_expr(input) { - Expr::Value(Value::UnicodeStringLiteral(s)) => { + Expr::Value(ValueWithSpan { + value: Value::UnicodeStringLiteral(s), + span: _, + }) => { assert_eq!(expected, s); } _ => unreachable!(), @@ -5250,10 +5339,14 @@ fn check_arrow_precedence(sql: &str, arrow_operator: BinaryOperator) { span: Span::empty(), })), op: arrow_operator, - right: Box::new(Expr::Value(Value::SingleQuotedString("bar".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("bar".to_string())).with_empty_span() + )), }), op: BinaryOperator::Eq, - right: Box::new(Expr::Value(Value::SingleQuotedString("spam".to_string()))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("spam".to_string())).with_empty_span() + )), } ) } @@ -5283,7 +5376,9 @@ fn arrow_cast_precedence() { op: BinaryOperator::Arrow, right: Box::new(Expr::Cast { kind: CastKind::DoubleColon, - expr: Box::new(Expr::Value(Value::SingleQuotedString("bar".to_string()))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("bar".to_string())).with_empty_span() + )), data_type: DataType::Text, format: None, }), @@ -5410,7 +5505,7 @@ fn parse_bitstring_literal() { assert_eq!( select.projection, vec![SelectItem::UnnamedExpr(Expr::Value( - Value::SingleQuotedByteStringLiteral("111".to_string()) + (Value::SingleQuotedByteStringLiteral("111".to_string())).with_empty_span() ))] ); } diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index c4b897f01..7736735cb 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -208,7 +208,7 @@ fn test_redshift_json_path() { path: JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")) + key: Expr::value(number("0")) }, JsonPathElem::Dot { key: "o_orderkey".to_string(), @@ -231,10 +231,12 @@ fn test_redshift_json_path() { path: JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")) + key: Expr::value(number("0")) }, JsonPathElem::Bracket { - key: Expr::Value(Value::SingleQuotedString("id".to_owned())) + key: Expr::Value( + (Value::SingleQuotedString("id".to_owned())).with_empty_span() + ) } ] } @@ -255,10 +257,12 @@ fn test_redshift_json_path() { path: JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")) + key: Expr::value(number("0")) }, JsonPathElem::Bracket { - key: Expr::Value(Value::SingleQuotedString("id".to_owned())) + key: Expr::Value( + (Value::SingleQuotedString("id".to_owned())).with_empty_span() + ) } ] } @@ -279,7 +283,7 @@ fn test_redshift_json_path() { path: JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")) + key: Expr::value(number("0")) }, JsonPathElem::Dot { key: "id".to_string(), @@ -306,7 +310,7 @@ fn test_parse_json_path_from() { &Some(JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")) + key: Expr::value(number("0")) }, JsonPathElem::Dot { key: "a".to_string(), @@ -330,14 +334,16 @@ fn test_parse_json_path_from() { &Some(JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")) + key: Expr::value(number("0")) }, JsonPathElem::Dot { key: "a".to_string(), quoted: false }, JsonPathElem::Bracket { - key: Expr::Value(Value::Number("1".parse().unwrap(), false)) + key: Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + ) }, JsonPathElem::Dot { key: "b".to_string(), diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 12796bb65..b1d31e6dd 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -570,8 +570,8 @@ fn test_snowflake_create_table_with_autoincrement_columns() { IdentityProperty { parameters: Some(IdentityPropertyFormatKind::FunctionCall( IdentityParameters { - seed: Expr::Value(number("100")), - increment: Expr::Value(number("1")), + seed: Expr::value(number("100")), + increment: Expr::value(number("1")), } )), order: Some(IdentityPropertyOrder::NoOrder), @@ -602,8 +602,12 @@ fn test_snowflake_create_table_with_autoincrement_columns() { parameters: Some( IdentityPropertyFormatKind::StartAndIncrement( IdentityParameters { - seed: Expr::Value(number("100")), - increment: Expr::Value(number("1")), + seed: Expr::Value( + (number("100")).with_empty_span() + ), + increment: Expr::Value( + (number("1")).with_empty_span() + ), } ) ), @@ -1108,9 +1112,9 @@ fn parse_semi_structured_data_traversal() { path: JsonPath { path: vec![JsonPathElem::Bracket { key: Expr::BinaryOp { - left: Box::new(Expr::Value(number("2"))), + left: Box::new(Expr::value(number("2"))), op: BinaryOperator::Plus, - right: Box::new(Expr::Value(number("2"))) + right: Box::new(Expr::value(number("2"))) }, }] }, @@ -1188,7 +1192,7 @@ fn parse_semi_structured_data_traversal() { quoted: false, }, JsonPathElem::Bracket { - key: Expr::Value(number("0")), + key: Expr::value(number("0")), }, JsonPathElem::Dot { key: "bar".to_owned(), @@ -1210,7 +1214,7 @@ fn parse_semi_structured_data_traversal() { path: JsonPath { path: vec![ JsonPathElem::Bracket { - key: Expr::Value(number("0")), + key: Expr::value(number("0")), }, JsonPathElem::Dot { key: "foo".to_owned(), @@ -1276,7 +1280,7 @@ fn parse_semi_structured_data_traversal() { }), path: JsonPath { path: vec![JsonPathElem::Bracket { - key: Expr::Value(number("1")) + key: Expr::value(number("1")) }] } } @@ -1661,13 +1665,13 @@ fn parse_snowflake_declare_result_set() { ( "DECLARE res RESULTSET DEFAULT 42", "res", - Some(DeclareAssignment::Default(Expr::Value(number("42")).into())), + Some(DeclareAssignment::Default(Expr::value(number("42")).into())), ), ( "DECLARE res RESULTSET := 42", "res", Some(DeclareAssignment::DuckAssignment( - Expr::Value(number("42")).into(), + Expr::value(number("42")).into(), )), ), ("DECLARE res RESULTSET", "res", None), @@ -1717,8 +1721,8 @@ fn parse_snowflake_declare_exception() { "ex", Some(DeclareAssignment::Expr( Expr::Tuple(vec![ - Expr::Value(number("42")), - Expr::Value(Value::SingleQuotedString("ERROR".to_string())), + Expr::value(number("42")), + Expr::Value((Value::SingleQuotedString("ERROR".to_string())).with_empty_span()), ]) .into(), )), @@ -1754,13 +1758,13 @@ fn parse_snowflake_declare_variable() { "DECLARE profit TEXT DEFAULT 42", "profit", Some(DataType::Text), - Some(DeclareAssignment::Default(Expr::Value(number("42")).into())), + Some(DeclareAssignment::Default(Expr::value(number("42")).into())), ), ( "DECLARE profit DEFAULT 42", "profit", None, - Some(DeclareAssignment::Default(Expr::Value(number("42")).into())), + Some(DeclareAssignment::Default(Expr::value(number("42")).into())), ), ("DECLARE profit TEXT", "profit", Some(DataType::Text), None), ("DECLARE profit", "profit", None, None), @@ -2509,10 +2513,14 @@ fn test_snowflake_trim() { let select = snowflake().verified_only_select(sql_only_select); assert_eq!( &Expr::Trim { - expr: Box::new(Expr::Value(Value::SingleQuotedString("xyz".to_owned()))), + expr: Box::new(Expr::Value( + (Value::SingleQuotedString("xyz".to_owned())).with_empty_span() + )), trim_where: None, trim_what: None, - trim_characters: Some(vec![Expr::Value(Value::SingleQuotedString("a".to_owned()))]), + trim_characters: Some(vec![Expr::Value( + (Value::SingleQuotedString("a".to_owned())).with_empty_span() + )]), }, expr_from_projection(only(&select.projection)) ); @@ -2530,7 +2538,7 @@ fn test_number_placeholder() { let sql_only_select = "SELECT :1"; let select = snowflake().verified_only_select(sql_only_select); assert_eq!( - &Expr::Value(Value::Placeholder(":1".into())), + &Expr::Value((Value::Placeholder(":1".into())).with_empty_span()), expr_from_projection(only(&select.projection)) ); @@ -2676,7 +2684,7 @@ fn parse_comma_outer_join() { "myudf", [Expr::UnaryOp { op: UnaryOperator::Plus, - expr: Box::new(Expr::Value(number("42"))) + expr: Box::new(Expr::value(number("42"))) }] )), }) diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 17dcfed85..361c9b051 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -369,7 +369,9 @@ fn test_placeholder() { let ast = sqlite().verified_only_select(sql); assert_eq!( ast.projection[0], - UnnamedExpr(Expr::Value(Value::Placeholder("@xxx".into()))), + UnnamedExpr(Expr::Value( + (Value::Placeholder("@xxx".into())).with_empty_span() + )), ); } @@ -446,7 +448,11 @@ fn parse_attach_database() { match verified_stmt { Statement::AttachDatabase { schema_name, - database_file_name: Expr::Value(Value::SingleQuotedString(literal_name)), + database_file_name: + Expr::Value(ValueWithSpan { + value: Value::SingleQuotedString(literal_name), + span: _, + }), database: true, } => { assert_eq!(schema_name.value, "test"); @@ -469,8 +475,8 @@ fn parse_update_tuple_row_values() { ObjectName::from(vec![Ident::new("b"),]), ]), value: Expr::Tuple(vec![ - Expr::Value(Value::Number("1".parse().unwrap(), false)), - Expr::Value(Value::Number("2".parse().unwrap(), false)) + Expr::Value((Value::Number("1".parse().unwrap(), false)).with_empty_span()), + Expr::Value((Value::Number("2".parse().unwrap(), false)).with_empty_span()) ]) }], selection: None, @@ -530,7 +536,12 @@ fn test_dollar_identifier_as_placeholder() { Expr::BinaryOp { op, left, right } => { assert_eq!(op, BinaryOperator::Eq); assert_eq!(left, Box::new(Expr::Identifier(Ident::new("id")))); - assert_eq!(right, Box::new(Expr::Value(Placeholder("$id".to_string())))); + assert_eq!( + right, + Box::new(Expr::Value( + (Placeholder("$id".to_string())).with_empty_span() + )) + ); } _ => unreachable!(), } @@ -540,7 +551,12 @@ fn test_dollar_identifier_as_placeholder() { Expr::BinaryOp { op, left, right } => { assert_eq!(op, BinaryOperator::Eq); assert_eq!(left, Box::new(Expr::Identifier(Ident::new("id")))); - assert_eq!(right, Box::new(Expr::Value(Placeholder("$$".to_string())))); + assert_eq!( + right, + Box::new(Expr::Value( + (Placeholder("$$".to_string())).with_empty_span() + )) + ); } _ => unreachable!(), } From 648efd7057d63c65b53eddc3d05cc89d5697d85c Mon Sep 17 00:00:00 2001 From: Michael Date: Tue, 25 Feb 2025 08:50:29 +0200 Subject: [PATCH 059/286] feat: adjust create and drop trigger for mysql dialect (#1734) --- src/ast/mod.rs | 7 ++++-- src/parser/mod.rs | 11 +++++---- tests/sqlparser_mysql.rs | 48 +++++++++++++++++++++++++++++++++++++ tests/sqlparser_postgres.rs | 4 ++-- 4 files changed, 62 insertions(+), 8 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index afe3e77d2..5263bfc3c 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3256,7 +3256,7 @@ pub enum Statement { DropTrigger { if_exists: bool, trigger_name: ObjectName, - table_name: ObjectName, + table_name: Option, /// `CASCADE` or `RESTRICT` option: Option, }, @@ -4062,7 +4062,10 @@ impl fmt::Display for Statement { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {trigger_name} ON {table_name}")?; + match &table_name { + Some(table_name) => write!(f, " {trigger_name} ON {table_name}")?, + None => write!(f, " {trigger_name}")?, + }; if let Some(option) = option { write!(f, " {option}")?; } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 72e4567cb..2d64ff3e1 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4990,14 +4990,17 @@ impl<'a> Parser<'a> { /// DROP TRIGGER [ IF EXISTS ] name ON table_name [ CASCADE | RESTRICT ] /// ``` pub fn parse_drop_trigger(&mut self) -> Result { - if !dialect_of!(self is PostgreSqlDialect | GenericDialect) { + if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect) { self.prev_token(); return self.expected("an object type after DROP", self.peek_token()); } let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); let trigger_name = self.parse_object_name(false)?; - self.expect_keyword_is(Keyword::ON)?; - let table_name = self.parse_object_name(false)?; + let table_name = if self.parse_keyword(Keyword::ON) { + Some(self.parse_object_name(false)?) + } else { + None + }; let option = self .parse_one_of_keywords(&[Keyword::CASCADE, Keyword::RESTRICT]) .map(|keyword| match keyword { @@ -5018,7 +5021,7 @@ impl<'a> Parser<'a> { or_replace: bool, is_constraint: bool, ) -> Result { - if !dialect_of!(self is PostgreSqlDialect | GenericDialect) { + if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect) { self.prev_token(); return self.expected("an object type after CREATE", self.peek_token()); } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index ad2987b28..861f782c6 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3291,3 +3291,51 @@ fn parse_looks_like_single_line_comment() { "UPDATE account SET balance = balance WHERE account_id = 5752", ); } + +#[test] +fn parse_create_trigger() { + let sql_create_trigger = r#" + CREATE TRIGGER emp_stamp BEFORE INSERT ON emp + FOR EACH ROW EXECUTE FUNCTION emp_stamp(); + "#; + let create_stmt = mysql().one_statement_parses_to(sql_create_trigger, ""); + assert_eq!( + create_stmt, + Statement::CreateTrigger { + or_replace: false, + is_constraint: false, + name: ObjectName::from(vec![Ident::new("emp_stamp")]), + period: TriggerPeriod::Before, + events: vec![TriggerEvent::Insert], + table_name: ObjectName::from(vec![Ident::new("emp")]), + referenced_table_name: None, + referencing: vec![], + trigger_object: TriggerObject::Row, + include_each: true, + condition: None, + exec_body: TriggerExecBody { + exec_type: TriggerExecBodyType::Function, + func_desc: FunctionDesc { + name: ObjectName::from(vec![Ident::new("emp_stamp")]), + args: None, + } + }, + characteristics: None, + } + ); +} + +#[test] +fn parse_drop_trigger() { + let sql_drop_trigger = "DROP TRIGGER emp_stamp;"; + let drop_stmt = mysql().one_statement_parses_to(sql_drop_trigger, ""); + assert_eq!( + drop_stmt, + Statement::DropTrigger { + if_exists: false, + trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), + table_name: None, + option: None, + } + ); +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 22558329d..7508218f9 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5044,7 +5044,7 @@ fn parse_drop_trigger() { Statement::DropTrigger { if_exists, trigger_name: ObjectName::from(vec![Ident::new("check_update")]), - table_name: ObjectName::from(vec![Ident::new("table_name")]), + table_name: Some(ObjectName::from(vec![Ident::new("table_name")])), option } ); @@ -5297,7 +5297,7 @@ fn parse_trigger_related_functions() { Statement::DropTrigger { if_exists: false, trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), - table_name: ObjectName::from(vec![Ident::new("emp")]), + table_name: Some(ObjectName::from(vec![Ident::new("emp")])), option: None } ); From de4dbc5b1d5bb91f3887478c8e32e739ab701e84 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Tue, 25 Feb 2025 22:00:08 -0800 Subject: [PATCH 060/286] Parse SIGNED INTEGER type in MySQL CAST (#1739) --- src/ast/data_type.rs | 70 +++++++++++++++++++++++++++++----------- src/ast/mod.rs | 5 +-- src/keywords.rs | 1 + src/parser/mod.rs | 32 ++++++++++++------ tests/sqlparser_mysql.rs | 28 +++++++++++++--- 5 files changed, 101 insertions(+), 35 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index cae8ca8f0..57bc67441 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -132,19 +132,19 @@ pub enum DataType { /// Tiny integer with optional display width e.g. TINYINT or TINYINT(3) TinyInt(Option), /// Unsigned tiny integer with optional display width e.g. TINYINT UNSIGNED or TINYINT(3) UNSIGNED - UnsignedTinyInt(Option), + TinyIntUnsigned(Option), /// Int2 as alias for SmallInt in [postgresql] /// Note: Int2 mean 2 bytes in postgres (not 2 bits) /// Int2 with optional display width e.g. INT2 or INT2(5) /// /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html Int2(Option), - /// Unsigned Int2 with optional display width e.g. INT2 Unsigned or INT2(5) Unsigned - UnsignedInt2(Option), + /// Unsigned Int2 with optional display width e.g. INT2 UNSIGNED or INT2(5) UNSIGNED + Int2Unsigned(Option), /// Small integer with optional display width e.g. SMALLINT or SMALLINT(5) SmallInt(Option), /// Unsigned small integer with optional display width e.g. SMALLINT UNSIGNED or SMALLINT(5) UNSIGNED - UnsignedSmallInt(Option), + SmallIntUnsigned(Option), /// MySQL medium integer ([1]) with optional display width e.g. MEDIUMINT or MEDIUMINT(5) /// /// [1]: https://dev.mysql.com/doc/refman/8.0/en/integer-types.html @@ -152,7 +152,7 @@ pub enum DataType { /// Unsigned medium integer ([1]) with optional display width e.g. MEDIUMINT UNSIGNED or MEDIUMINT(5) UNSIGNED /// /// [1]: https://dev.mysql.com/doc/refman/8.0/en/integer-types.html - UnsignedMediumInt(Option), + MediumIntUnsigned(Option), /// Int with optional display width e.g. INT or INT(11) Int(Option), /// Int4 as alias for Integer in [postgresql] @@ -197,11 +197,11 @@ pub enum DataType { /// Integer with optional display width e.g. INTEGER or INTEGER(11) Integer(Option), /// Unsigned int with optional display width e.g. INT UNSIGNED or INT(11) UNSIGNED - UnsignedInt(Option), + IntUnsigned(Option), /// Unsigned int4 with optional display width e.g. INT4 UNSIGNED or INT4(11) UNSIGNED - UnsignedInt4(Option), + Int4Unsigned(Option), /// Unsigned integer with optional display width e.g. INTEGER UNSIGNED or INTEGER(11) UNSIGNED - UnsignedInteger(Option), + IntegerUnsigned(Option), /// Unsigned integer type in [clickhouse] /// Note: UInt8 mean 8 bits in [clickhouse] /// @@ -235,9 +235,29 @@ pub enum DataType { /// Big integer with optional display width e.g. BIGINT or BIGINT(20) BigInt(Option), /// Unsigned big integer with optional display width e.g. BIGINT UNSIGNED or BIGINT(20) UNSIGNED - UnsignedBigInt(Option), + BigIntUnsigned(Option), /// Unsigned Int8 with optional display width e.g. INT8 UNSIGNED or INT8(11) UNSIGNED - UnsignedInt8(Option), + Int8Unsigned(Option), + /// Signed integer as used in [MySQL CAST] target types, without optional `INTEGER` suffix: + /// `SIGNED` + /// + /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html + Signed, + /// Signed integer as used in [MySQL CAST] target types, with optional `INTEGER` suffix: + /// `SIGNED INTEGER` + /// + /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html + SignedInteger, + /// Signed integer as used in [MySQL CAST] target types, without optional `INTEGER` suffix: + /// `SIGNED` + /// + /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html + Unsigned, + /// Unsigned integer as used in [MySQL CAST] target types, with optional `INTEGER` suffix: + /// `UNSIGNED INTEGER` + /// + /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html + UnsignedInteger, /// Float4 as alias for Real in [postgresql] /// /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html @@ -433,29 +453,29 @@ impl fmt::Display for DataType { DataType::TinyInt(zerofill) => { format_type_with_optional_length(f, "TINYINT", zerofill, false) } - DataType::UnsignedTinyInt(zerofill) => { + DataType::TinyIntUnsigned(zerofill) => { format_type_with_optional_length(f, "TINYINT", zerofill, true) } DataType::Int2(zerofill) => { format_type_with_optional_length(f, "INT2", zerofill, false) } - DataType::UnsignedInt2(zerofill) => { + DataType::Int2Unsigned(zerofill) => { format_type_with_optional_length(f, "INT2", zerofill, true) } DataType::SmallInt(zerofill) => { format_type_with_optional_length(f, "SMALLINT", zerofill, false) } - DataType::UnsignedSmallInt(zerofill) => { + DataType::SmallIntUnsigned(zerofill) => { format_type_with_optional_length(f, "SMALLINT", zerofill, true) } DataType::MediumInt(zerofill) => { format_type_with_optional_length(f, "MEDIUMINT", zerofill, false) } - DataType::UnsignedMediumInt(zerofill) => { + DataType::MediumIntUnsigned(zerofill) => { format_type_with_optional_length(f, "MEDIUMINT", zerofill, true) } DataType::Int(zerofill) => format_type_with_optional_length(f, "INT", zerofill, false), - DataType::UnsignedInt(zerofill) => { + DataType::IntUnsigned(zerofill) => { format_type_with_optional_length(f, "INT", zerofill, true) } DataType::Int4(zerofill) => { @@ -479,22 +499,22 @@ impl fmt::Display for DataType { DataType::Int256 => { write!(f, "Int256") } - DataType::UnsignedInt4(zerofill) => { + DataType::Int4Unsigned(zerofill) => { format_type_with_optional_length(f, "INT4", zerofill, true) } DataType::Integer(zerofill) => { format_type_with_optional_length(f, "INTEGER", zerofill, false) } - DataType::UnsignedInteger(zerofill) => { + DataType::IntegerUnsigned(zerofill) => { format_type_with_optional_length(f, "INTEGER", zerofill, true) } DataType::BigInt(zerofill) => { format_type_with_optional_length(f, "BIGINT", zerofill, false) } - DataType::UnsignedBigInt(zerofill) => { + DataType::BigIntUnsigned(zerofill) => { format_type_with_optional_length(f, "BIGINT", zerofill, true) } - DataType::UnsignedInt8(zerofill) => { + DataType::Int8Unsigned(zerofill) => { format_type_with_optional_length(f, "INT8", zerofill, true) } DataType::UInt8 => { @@ -515,6 +535,18 @@ impl fmt::Display for DataType { DataType::UInt256 => { write!(f, "UInt256") } + DataType::Signed => { + write!(f, "SIGNED") + } + DataType::SignedInteger => { + write!(f, "SIGNED INTEGER") + } + DataType::Unsigned => { + write!(f, "UNSIGNED") + } + DataType::UnsignedInteger => { + write!(f, "UNSIGNED INTEGER") + } DataType::Real => write!(f, "REAL"), DataType::Float4 => write!(f, "FLOAT4"), DataType::Float32 => write!(f, "Float32"), diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 5263bfc3c..85d3ed91c 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -814,8 +814,9 @@ pub enum Expr { kind: CastKind, expr: Box, data_type: DataType, - // Optional CAST(string_expression AS type FORMAT format_string_expression) as used by BigQuery - // https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#formatting_syntax + /// Optional CAST(string_expression AS type FORMAT format_string_expression) as used by [BigQuery] + /// + /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#formatting_syntax format: Option, }, /// AT a timestamp to a different timezone e.g. `FROM_UNIXTIME(0) AT TIME ZONE 'UTC-06:00'` diff --git a/src/keywords.rs b/src/keywords.rs index d62a038b8..020b404ed 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -790,6 +790,7 @@ define_keywords!( SHARE, SHARING, SHOW, + SIGNED, SIMILAR, SKIP, SLOW, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2d64ff3e1..ddcb6055f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8867,7 +8867,7 @@ impl<'a> Parser<'a> { Keyword::TINYINT => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedTinyInt(optional_precision?)) + Ok(DataType::TinyIntUnsigned(optional_precision?)) } else { Ok(DataType::TinyInt(optional_precision?)) } @@ -8875,7 +8875,7 @@ impl<'a> Parser<'a> { Keyword::INT2 => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedInt2(optional_precision?)) + Ok(DataType::Int2Unsigned(optional_precision?)) } else { Ok(DataType::Int2(optional_precision?)) } @@ -8883,7 +8883,7 @@ impl<'a> Parser<'a> { Keyword::SMALLINT => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedSmallInt(optional_precision?)) + Ok(DataType::SmallIntUnsigned(optional_precision?)) } else { Ok(DataType::SmallInt(optional_precision?)) } @@ -8891,7 +8891,7 @@ impl<'a> Parser<'a> { Keyword::MEDIUMINT => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedMediumInt(optional_precision?)) + Ok(DataType::MediumIntUnsigned(optional_precision?)) } else { Ok(DataType::MediumInt(optional_precision?)) } @@ -8899,7 +8899,7 @@ impl<'a> Parser<'a> { Keyword::INT => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedInt(optional_precision?)) + Ok(DataType::IntUnsigned(optional_precision?)) } else { Ok(DataType::Int(optional_precision?)) } @@ -8907,7 +8907,7 @@ impl<'a> Parser<'a> { Keyword::INT4 => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedInt4(optional_precision?)) + Ok(DataType::Int4Unsigned(optional_precision?)) } else { Ok(DataType::Int4(optional_precision?)) } @@ -8915,7 +8915,7 @@ impl<'a> Parser<'a> { Keyword::INT8 => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedInt8(optional_precision?)) + Ok(DataType::Int8Unsigned(optional_precision?)) } else { Ok(DataType::Int8(optional_precision?)) } @@ -8928,7 +8928,7 @@ impl<'a> Parser<'a> { Keyword::INTEGER => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedInteger(optional_precision?)) + Ok(DataType::IntegerUnsigned(optional_precision?)) } else { Ok(DataType::Integer(optional_precision?)) } @@ -8936,7 +8936,7 @@ impl<'a> Parser<'a> { Keyword::BIGINT => { let optional_precision = self.parse_optional_precision(); if self.parse_keyword(Keyword::UNSIGNED) { - Ok(DataType::UnsignedBigInt(optional_precision?)) + Ok(DataType::BigIntUnsigned(optional_precision?)) } else { Ok(DataType::BigInt(optional_precision?)) } @@ -9142,6 +9142,20 @@ impl<'a> Parser<'a> { let columns = self.parse_returns_table_columns()?; Ok(DataType::Table(columns)) } + Keyword::SIGNED => { + if self.parse_keyword(Keyword::INTEGER) { + Ok(DataType::SignedInteger) + } else { + Ok(DataType::Signed) + } + } + Keyword::UNSIGNED => { + if self.parse_keyword(Keyword::INTEGER) { + Ok(DataType::UnsignedInteger) + } else { + Ok(DataType::Unsigned) + } + } _ => { self.prev_token(); let type_name = self.parse_object_name(false)?; diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 861f782c6..4856bd894 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1359,27 +1359,27 @@ fn parse_create_table_unsigned() { vec![ ColumnDef { name: Ident::new("bar_tinyint"), - data_type: DataType::UnsignedTinyInt(Some(3)), + data_type: DataType::TinyIntUnsigned(Some(3)), options: vec![], }, ColumnDef { name: Ident::new("bar_smallint"), - data_type: DataType::UnsignedSmallInt(Some(5)), + data_type: DataType::SmallIntUnsigned(Some(5)), options: vec![], }, ColumnDef { name: Ident::new("bar_mediumint"), - data_type: DataType::UnsignedMediumInt(Some(13)), + data_type: DataType::MediumIntUnsigned(Some(13)), options: vec![], }, ColumnDef { name: Ident::new("bar_int"), - data_type: DataType::UnsignedInt(Some(11)), + data_type: DataType::IntUnsigned(Some(11)), options: vec![], }, ColumnDef { name: Ident::new("bar_bigint"), - data_type: DataType::UnsignedBigInt(Some(20)), + data_type: DataType::BigIntUnsigned(Some(20)), options: vec![], }, ], @@ -3339,3 +3339,21 @@ fn parse_drop_trigger() { } ); } + +#[test] +fn parse_cast_integers() { + mysql().verified_expr("CAST(foo AS UNSIGNED)"); + mysql().verified_expr("CAST(foo AS SIGNED)"); + mysql().verified_expr("CAST(foo AS UNSIGNED INTEGER)"); + mysql().verified_expr("CAST(foo AS SIGNED INTEGER)"); + + mysql() + .run_parser_method("CAST(foo AS UNSIGNED(3))", |p| p.parse_expr()) + .expect_err("CAST doesn't allow display width"); + mysql() + .run_parser_method("CAST(foo AS UNSIGNED(3) INTEGER)", |p| p.parse_expr()) + .expect_err("CAST doesn't allow display width"); + mysql() + .run_parser_method("CAST(foo AS UNSIGNED INTEGER(3))", |p| p.parse_expr()) + .expect_err("CAST doesn't allow display width"); +} From 3adc746b11411746f78f5c2506e1bf72e1caf583 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Tue, 25 Feb 2025 22:03:38 -0800 Subject: [PATCH 061/286] Parse MySQL ALTER TABLE ALGORITHM option (#1745) --- src/ast/ddl.rs | 45 +++++++++++++++++++++++++++++++++---- src/ast/mod.rs | 18 +++++++-------- src/ast/spans.rs | 1 + src/keywords.rs | 2 ++ src/parser/mod.rs | 18 +++++++++++++++ tests/sqlparser_mysql.rs | 48 ++++++++++++++++++++++++++++++++++++++++ 6 files changed, 119 insertions(+), 13 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 1fbc45603..372a75adb 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -65,7 +65,6 @@ pub enum AlterTableOperation { name: Ident, select: ProjectionSelect, }, - /// `DROP PROJECTION [IF EXISTS] name` /// /// Note: this is a ClickHouse-specific operation. @@ -74,7 +73,6 @@ pub enum AlterTableOperation { if_exists: bool, name: Ident, }, - /// `MATERIALIZE PROJECTION [IF EXISTS] name [IN PARTITION partition_name]` /// /// Note: this is a ClickHouse-specific operation. @@ -84,7 +82,6 @@ pub enum AlterTableOperation { name: Ident, partition: Option, }, - /// `CLEAR PROJECTION [IF EXISTS] name [IN PARTITION partition_name]` /// /// Note: this is a ClickHouse-specific operation. @@ -94,7 +91,6 @@ pub enum AlterTableOperation { name: Ident, partition: Option, }, - /// `DISABLE ROW LEVEL SECURITY` /// /// Note: this is a PostgreSQL-specific operation. @@ -272,6 +268,15 @@ pub enum AlterTableOperation { DropClusteringKey, SuspendRecluster, ResumeRecluster, + /// `ALGORITHM [=] { DEFAULT | INSTANT | INPLACE | COPY }` + /// + /// [MySQL]-specific table alter algorithm. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html + Algorithm { + equals: bool, + algorithm: AlterTableAlgorithm, + }, } /// An `ALTER Policy` (`Statement::AlterPolicy`) operation @@ -317,6 +322,30 @@ impl fmt::Display for AlterPolicyOperation { } } +/// [MySQL] `ALTER TABLE` algorithm. +/// +/// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterTableAlgorithm { + Default, + Instant, + Inplace, + Copy, +} + +impl fmt::Display for AlterTableAlgorithm { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.write_str(match self { + Self::Default => "DEFAULT", + Self::Instant => "INSTANT", + Self::Inplace => "INPLACE", + Self::Copy => "COPY", + }) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -407,6 +436,14 @@ impl fmt::Display for AlterTableOperation { } write!(f, " {} ({})", name, query) } + AlterTableOperation::Algorithm { equals, algorithm } => { + write!( + f, + "ALGORITHM {}{}", + if *equals { "= " } else { "" }, + algorithm + ) + } AlterTableOperation::DropProjection { if_exists, name } => { write!(f, "DROP PROJECTION")?; if *if_exists { diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 85d3ed91c..72be3ff6c 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -48,15 +48,15 @@ pub use self::dcl::{ }; pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, - AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, - AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, - ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, - CreateConnector, CreateFunction, Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, - GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, - IdentityPropertyKind, IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, - NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, TableConstraint, - TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, - ViewColumnDef, + AlterTableAlgorithm, AlterTableOperation, AlterType, AlterTypeAddValue, + AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, + ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, + ConstraintCharacteristics, CreateConnector, CreateFunction, Deduplicate, DeferrableInitial, + DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, + IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexOption, + IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, + ReferentialAction, TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, + UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index cfc3eb633..7cb5ddfed 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1062,6 +1062,7 @@ impl Spanned for AlterTableOperation { AlterTableOperation::DropClusteringKey => Span::empty(), AlterTableOperation::SuspendRecluster => Span::empty(), AlterTableOperation::ResumeRecluster => Span::empty(), + AlterTableOperation::Algorithm { .. } => Span::empty(), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 020b404ed..a6854f073 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -427,11 +427,13 @@ define_keywords!( INNER, INOUT, INPATH, + INPLACE, INPUT, INPUTFORMAT, INSENSITIVE, INSERT, INSTALL, + INSTANT, INSTEAD, INT, INT128, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ddcb6055f..86a86824d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8163,6 +8163,24 @@ impl<'a> Parser<'a> { AlterTableOperation::SuspendRecluster } else if self.parse_keywords(&[Keyword::RESUME, Keyword::RECLUSTER]) { AlterTableOperation::ResumeRecluster + } else if self.parse_keyword(Keyword::ALGORITHM) { + let equals = self.consume_token(&Token::Eq); + let algorithm = match self.parse_one_of_keywords(&[ + Keyword::DEFAULT, + Keyword::INSTANT, + Keyword::INPLACE, + Keyword::COPY, + ]) { + Some(Keyword::DEFAULT) => AlterTableAlgorithm::Default, + Some(Keyword::INSTANT) => AlterTableAlgorithm::Instant, + Some(Keyword::INPLACE) => AlterTableAlgorithm::Inplace, + Some(Keyword::COPY) => AlterTableAlgorithm::Copy, + _ => self.expected( + "DEFAULT, INSTANT, INPLACE, or COPY after ALGORITHM [=]", + self.peek_token(), + )?, + }; + AlterTableOperation::Algorithm { equals, algorithm } } else { let options: Vec = self.parse_options_with_keywords(&[Keyword::SET, Keyword::TBLPROPERTIES])?; diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 4856bd894..5e98d3f41 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2422,6 +2422,54 @@ fn parse_alter_table_modify_column() { assert_eq!(expected_operation, operation); } +#[test] +fn parse_alter_table_with_algorithm() { + let sql = "ALTER TABLE tab ALGORITHM = COPY"; + let expected_operation = AlterTableOperation::Algorithm { + equals: true, + algorithm: AlterTableAlgorithm::Copy, + }; + let operation = alter_table_op(mysql_and_generic().verified_stmt(sql)); + assert_eq!(expected_operation, operation); + + // Check order doesn't matter + let sql = + "ALTER TABLE users DROP COLUMN password_digest, ALGORITHM = COPY, RENAME COLUMN name TO username"; + let stmt = mysql_and_generic().verified_stmt(sql); + match stmt { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![ + AlterTableOperation::DropColumn { + column_name: Ident::new("password_digest"), + if_exists: false, + drop_behavior: None, + }, + AlterTableOperation::Algorithm { + equals: true, + algorithm: AlterTableAlgorithm::Copy, + }, + AlterTableOperation::RenameColumn { + old_column_name: Ident::new("name"), + new_column_name: Ident::new("username") + }, + ] + ) + } + _ => panic!("Unexpected statement {stmt}"), + } + + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM DEFAULT"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM INSTANT"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM INPLACE"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM COPY"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM = DEFAULT"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM = INSTANT"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM = INPLACE"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM = COPY"); +} + #[test] fn parse_alter_table_modify_column_with_column_position() { let expected_name = ObjectName::from(vec![Ident::new("orders")]); From 5b3500139a9bb76959e2c880b2c60fbd7fb9738e Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Thu, 27 Feb 2025 00:33:08 -0500 Subject: [PATCH 062/286] Random test cleanups use Expr::value (#1749) --- tests/sqlparser_clickhouse.rs | 22 ++++++---------------- 1 file changed, 6 insertions(+), 16 deletions(-) diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 98a1aef62..72a64a488 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -55,10 +55,7 @@ fn parse_map_access_expr() { "indexOf", [ Expr::Identifier(Ident::new("string_names")), - Expr::Value( - (Value::SingleQuotedString("endpoint".to_string())) - .with_empty_span() - ) + Expr::value(Value::SingleQuotedString("endpoint".to_string())) ] ), })], @@ -74,9 +71,7 @@ fn parse_map_access_expr() { left: Box::new(BinaryOp { left: Box::new(Identifier(Ident::new("id"))), op: BinaryOperator::Eq, - right: Box::new(Expr::Value( - (Value::SingleQuotedString("test".to_string())).with_empty_span() - )), + right: Box::new(Expr::value(Value::SingleQuotedString("test".to_string()))), }), op: BinaryOperator::And, right: Box::new(BinaryOp { @@ -87,18 +82,13 @@ fn parse_map_access_expr() { "indexOf", [ Expr::Identifier(Ident::new("string_name")), - Expr::Value( - (Value::SingleQuotedString("app".to_string())) - .with_empty_span() - ) + Expr::value(Value::SingleQuotedString("app".to_string())) ] ), })], }), op: BinaryOperator::NotEq, - right: Box::new(Expr::Value( - (Value::SingleQuotedString("foo".to_string())).with_empty_span() - )), + right: Box::new(Expr::value(Value::SingleQuotedString("foo".to_string()))), }), }), group_by: GroupByExpr::Expressions(vec![], vec![]), @@ -124,8 +114,8 @@ fn parse_array_expr() { assert_eq!( &Expr::Array(Array { elem: vec![ - Expr::Value((Value::SingleQuotedString("1".to_string())).with_empty_span()), - Expr::Value((Value::SingleQuotedString("2".to_string())).with_empty_span()), + Expr::value(Value::SingleQuotedString("1".to_string())), + Expr::value(Value::SingleQuotedString("2".to_string())), ], named: false, }), From c2914f82e10b0e45d83ca5e218ad5a1a760edde6 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Wed, 26 Feb 2025 21:40:30 -0800 Subject: [PATCH 063/286] Parse ALTER TABLE AUTO_INCREMENT operation for MySQL (#1748) --- src/ast/ddl.rs | 18 ++++++++++++++++++ src/ast/spans.rs | 1 + src/parser/mod.rs | 4 ++++ tests/sqlparser_mysql.rs | 13 +++++++++++++ 4 files changed, 36 insertions(+) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 372a75adb..bb85eb06c 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -34,6 +34,7 @@ use crate::ast::{ CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, Ident, MySQLColumnPosition, ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, Value, + ValueWithSpan, }; use crate::keywords::Keyword; use crate::tokenizer::Token; @@ -277,6 +278,15 @@ pub enum AlterTableOperation { equals: bool, algorithm: AlterTableAlgorithm, }, + /// `AUTO_INCREMENT [=] ` + /// + /// [MySQL]-specific table option for raising current auto increment value. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html + AutoIncrement { + equals: bool, + value: ValueWithSpan, + }, } /// An `ALTER Policy` (`Statement::AlterPolicy`) operation @@ -663,6 +673,14 @@ impl fmt::Display for AlterTableOperation { write!(f, "RESUME RECLUSTER")?; Ok(()) } + AlterTableOperation::AutoIncrement { equals, value } => { + write!( + f, + "AUTO_INCREMENT {}{}", + if *equals { "= " } else { "" }, + value + ) + } } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 7cb5ddfed..38e9e258e 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1063,6 +1063,7 @@ impl Spanned for AlterTableOperation { AlterTableOperation::SuspendRecluster => Span::empty(), AlterTableOperation::ResumeRecluster => Span::empty(), AlterTableOperation::Algorithm { .. } => Span::empty(), + AlterTableOperation::AutoIncrement { value, .. } => value.span(), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 86a86824d..f234fcc07 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8181,6 +8181,10 @@ impl<'a> Parser<'a> { )?, }; AlterTableOperation::Algorithm { equals, algorithm } + } else if self.parse_keyword(Keyword::AUTO_INCREMENT) { + let equals = self.consume_token(&Token::Eq); + let value = self.parse_number_value()?; + AlterTableOperation::AutoIncrement { equals, value } } else { let options: Vec = self.parse_options_with_keywords(&[Keyword::SET, Keyword::TBLPROPERTIES])?; diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 5e98d3f41..15f79b4c2 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2470,6 +2470,19 @@ fn parse_alter_table_with_algorithm() { mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM = COPY"); } +#[test] +fn parse_alter_table_auto_increment() { + let sql = "ALTER TABLE tab AUTO_INCREMENT = 42"; + let expected_operation = AlterTableOperation::AutoIncrement { + equals: true, + value: number("42").with_empty_span(), + }; + let operation = alter_table_op(mysql().verified_stmt(sql)); + assert_eq!(expected_operation, operation); + + mysql_and_generic().verified_stmt("ALTER TABLE `users` AUTO_INCREMENT 42"); +} + #[test] fn parse_alter_table_modify_column_with_column_position() { let expected_name = ObjectName::from(vec![Ident::new("orders")]); From ed416548dcfe4a73a3240bbf625fb9010a4925c8 Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Thu, 27 Feb 2025 12:29:59 -0500 Subject: [PATCH 064/286] Prepare for 55.0.0 release: Version and CHANGELOG (#1750) Co-authored-by: Ophir LOJKINE --- CHANGELOG.md | 1 + Cargo.toml | 2 +- changelog/0.55.0.md | 173 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 175 insertions(+), 1 deletion(-) create mode 100644 changelog/0.55.0.md diff --git a/CHANGELOG.md b/CHANGELOG.md index d1c55b285..362a637d9 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -28,6 +28,7 @@ technically be breaking and thus will result in a `0.(N+1)` version. - Unreleased: Check https://github.com/sqlparser-rs/sqlparser-rs/commits/main for undocumented changes. +- `0.55.0`: [changelog/0.55.0.md](changelog/0.55.0.md) - `0.54.0`: [changelog/0.54.0.md](changelog/0.54.0.md) - `0.53.0`: [changelog/0.53.0.md](changelog/0.53.0.md) - `0.52.0`: [changelog/0.52.0.md](changelog/0.52.0.md) diff --git a/Cargo.toml b/Cargo.toml index 9caff2512..99bfdc241 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.54.0" +version = "0.55.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/changelog/0.55.0.md b/changelog/0.55.0.md new file mode 100644 index 000000000..046bf22bc --- /dev/null +++ b/changelog/0.55.0.md @@ -0,0 +1,173 @@ + + +# sqlparser-rs 0.55.0 Changelog + +This release consists of 55 commits from 25 contributors. See credits at the end of this changelog for more information. + +## Migrating usages of `Expr::Value` + +In v0.55 of sqlparser the `Expr::Value` enum variant contains a `ValueWithSpan` instead of a `Value`. Here is how to migrate. + +### When pattern matching + +```diff +- Expr::Value(Value::SingleQuotedString(my_string)) => { ... } ++ Expr::Value(ValueWithSpan{ value: Value::SingleQuotedString(my_string), span: _ }) => { ... } +``` + +### When creating an `Expr` + +Use the new `Expr::value` method (notice the lowercase `v`), which will create a `ValueWithSpan` containing an empty span: + +```diff +- Expr::Value(Value::SingleQuotedString(my_string)) ++ Expr::value(Value::SingleQuotedString(my_string)) +``` + +## Migrating usages of `ObjectName` + +In v0.55 of sqlparser, the `ObjectName` structure has been changed as shown below. Here is now to migrate. + +```diff +- pub struct ObjectName(pub Vec); ++ pub struct ObjectName(pub Vec) +``` + +### When constructing `ObjectName` + +Use the `From` impl: + +```diff +- name: ObjectName(vec![Ident::new("f")]), ++ name: ObjectName::from(vec![Ident::new("f")]), +``` + +### Accessing Spans + +Use the `span()` function + +```diff +- name.span ++ name.span() +``` + + + +**Breaking changes:** + +- Enhance object name path segments [#1539](https://github.com/apache/datafusion-sqlparser-rs/pull/1539) (ayman-sigma) +- Store spans for Value expressions [#1738](https://github.com/apache/datafusion-sqlparser-rs/pull/1738) (lovasoa) + +**Implemented enhancements:** + +- feat: adjust create and drop trigger for mysql dialect [#1734](https://github.com/apache/datafusion-sqlparser-rs/pull/1734) (invm) + +**Fixed bugs:** + +- fix: make `serde` feature no_std [#1730](https://github.com/apache/datafusion-sqlparser-rs/pull/1730) (iajoiner) + +**Other:** + +- Update rat_exclude_file.txt [#1670](https://github.com/apache/datafusion-sqlparser-rs/pull/1670) (alamb) +- Add support for Snowflake account privileges [#1666](https://github.com/apache/datafusion-sqlparser-rs/pull/1666) (yoavcloud) +- Add support for Create Iceberg Table statement for Snowflake parser [#1664](https://github.com/apache/datafusion-sqlparser-rs/pull/1664) (Vedin) +- National strings: check if dialect supports backslash escape [#1672](https://github.com/apache/datafusion-sqlparser-rs/pull/1672) (hansott) +- Only support escape literals for Postgres, Redshift and generic dialect [#1674](https://github.com/apache/datafusion-sqlparser-rs/pull/1674) (hansott) +- BigQuery: Support trailing commas in column definitions list [#1682](https://github.com/apache/datafusion-sqlparser-rs/pull/1682) (iffyio) +- Enable GROUP BY exp for Snowflake dialect [#1683](https://github.com/apache/datafusion-sqlparser-rs/pull/1683) (yoavcloud) +- Add support for parsing empty dictionary expressions [#1684](https://github.com/apache/datafusion-sqlparser-rs/pull/1684) (yoavcloud) +- Support multiple tables in `UPDATE FROM` clause [#1681](https://github.com/apache/datafusion-sqlparser-rs/pull/1681) (iffyio) +- Add support for mysql table hints [#1675](https://github.com/apache/datafusion-sqlparser-rs/pull/1675) (AvivDavid-Satori) +- BigQuery: Add support for select expr star [#1680](https://github.com/apache/datafusion-sqlparser-rs/pull/1680) (iffyio) +- Support underscore separators in numbers for Clickhouse. Fixes #1659 [#1677](https://github.com/apache/datafusion-sqlparser-rs/pull/1677) (graup) +- BigQuery: Fix column identifier reserved keywords list [#1678](https://github.com/apache/datafusion-sqlparser-rs/pull/1678) (iffyio) +- Fix bug when parsing a Snowflake stage with `;` suffix [#1688](https://github.com/apache/datafusion-sqlparser-rs/pull/1688) (yoavcloud) +- Allow plain JOIN without turning it into INNER [#1692](https://github.com/apache/datafusion-sqlparser-rs/pull/1692) (mvzink) +- Fix DDL generation in case of an empty arguments function. [#1690](https://github.com/apache/datafusion-sqlparser-rs/pull/1690) (remysaissy) +- Fix `CREATE FUNCTION` round trip for Hive dialect [#1693](https://github.com/apache/datafusion-sqlparser-rs/pull/1693) (iffyio) +- Make numeric literal underscore test dialect agnostic [#1685](https://github.com/apache/datafusion-sqlparser-rs/pull/1685) (iffyio) +- Extend lambda support for ClickHouse and DuckDB dialects [#1686](https://github.com/apache/datafusion-sqlparser-rs/pull/1686) (gstvg) +- Make TypedString preserve quote style [#1679](https://github.com/apache/datafusion-sqlparser-rs/pull/1679) (graup) +- Do not parse ASOF and MATCH_CONDITION as table factor aliases [#1698](https://github.com/apache/datafusion-sqlparser-rs/pull/1698) (yoavcloud) +- Add support for GRANT on some common Snowflake objects [#1699](https://github.com/apache/datafusion-sqlparser-rs/pull/1699) (yoavcloud) +- Add RETURNS TABLE() support for CREATE FUNCTION in Postgresql [#1687](https://github.com/apache/datafusion-sqlparser-rs/pull/1687) (remysaissy) +- Add parsing for GRANT ROLE and GRANT DATABASE ROLE in Snowflake dialect [#1689](https://github.com/apache/datafusion-sqlparser-rs/pull/1689) (yoavcloud) +- Add support for `CREATE/ALTER/DROP CONNECTOR` syntax [#1701](https://github.com/apache/datafusion-sqlparser-rs/pull/1701) (wugeer) +- Parse Snowflake COPY INTO [#1669](https://github.com/apache/datafusion-sqlparser-rs/pull/1669) (yoavcloud) +- Require space after -- to start single line comment in MySQL [#1705](https://github.com/apache/datafusion-sqlparser-rs/pull/1705) (hansott) +- Add suppport for Show Objects statement for the Snowflake parser [#1702](https://github.com/apache/datafusion-sqlparser-rs/pull/1702) (DanCodedThis) +- Fix incorrect parsing of JsonAccess bracket notation after cast in Snowflake [#1708](https://github.com/apache/datafusion-sqlparser-rs/pull/1708) (yoavcloud) +- Parse Postgres VARBIT datatype [#1703](https://github.com/apache/datafusion-sqlparser-rs/pull/1703) (mvzink) +- Implement FROM-first selects [#1713](https://github.com/apache/datafusion-sqlparser-rs/pull/1713) (mitsuhiko) +- Enable custom dialects to support `MATCH() AGAINST()` [#1719](https://github.com/apache/datafusion-sqlparser-rs/pull/1719) (joocer) +- Support group by cube/rollup etc in BigQuery [#1720](https://github.com/apache/datafusion-sqlparser-rs/pull/1720) (Groennbeck) +- Add support for MS Varbinary(MAX) (#1714) [#1715](https://github.com/apache/datafusion-sqlparser-rs/pull/1715) (TylerBrinks) +- Add supports for Hive's `SELECT ... GROUP BY .. GROUPING SETS` syntax [#1653](https://github.com/apache/datafusion-sqlparser-rs/pull/1653) (wugeer) +- Differentiate LEFT JOIN from LEFT OUTER JOIN [#1726](https://github.com/apache/datafusion-sqlparser-rs/pull/1726) (mvzink) +- Add support for Postgres `ALTER TYPE` [#1727](https://github.com/apache/datafusion-sqlparser-rs/pull/1727) (jvatic) +- Replace `Method` and `CompositeAccess` with `CompoundFieldAccess` [#1716](https://github.com/apache/datafusion-sqlparser-rs/pull/1716) (iffyio) +- Add support for `EXECUTE IMMEDIATE` [#1717](https://github.com/apache/datafusion-sqlparser-rs/pull/1717) (iffyio) +- Treat COLLATE like any other column option [#1731](https://github.com/apache/datafusion-sqlparser-rs/pull/1731) (mvzink) +- Add support for PostgreSQL/Redshift geometric operators [#1723](https://github.com/apache/datafusion-sqlparser-rs/pull/1723) (benrsatori) +- Implement SnowFlake ALTER SESSION [#1712](https://github.com/apache/datafusion-sqlparser-rs/pull/1712) (osipovartem) +- Extend Visitor trait for Value type [#1725](https://github.com/apache/datafusion-sqlparser-rs/pull/1725) (tomershaniii) +- Add support for `ORDER BY ALL` [#1724](https://github.com/apache/datafusion-sqlparser-rs/pull/1724) (PokIsemaine) +- Parse casting to array using double colon operator in Redshift [#1737](https://github.com/apache/datafusion-sqlparser-rs/pull/1737) (yoavcloud) +- Replace parallel condition/result vectors with single CaseWhen vector in Expr::Case. This fixes the iteration order when using the `Visitor` trait. Expressions are now visited in the same order as they appear in the sql source. [#1733](https://github.com/apache/datafusion-sqlparser-rs/pull/1733) (lovasoa) +- BigQuery: Add support for `BEGIN` [#1718](https://github.com/apache/datafusion-sqlparser-rs/pull/1718) (iffyio) +- Parse SIGNED INTEGER type in MySQL CAST [#1739](https://github.com/apache/datafusion-sqlparser-rs/pull/1739) (mvzink) +- Parse MySQL ALTER TABLE ALGORITHM option [#1745](https://github.com/apache/datafusion-sqlparser-rs/pull/1745) (mvzink) +- Random test cleanups use Expr::value [#1749](https://github.com/apache/datafusion-sqlparser-rs/pull/1749) (alamb) +- Parse ALTER TABLE AUTO_INCREMENT operation for MySQL [#1748](https://github.com/apache/datafusion-sqlparser-rs/pull/1748) (mvzink) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 10 Yoav Cohen + 9 Ifeanyi Ubah + 7 Michael Victor Zink + 3 Hans Ott + 2 Andrew Lamb + 2 Ophir LOJKINE + 2 Paul Grau + 2 Rémy SAISSY + 2 wugeer + 1 Armin Ronacher + 1 Artem Osipov + 1 AvivDavid-Satori + 1 Ayman Elkfrawy + 1 DanCodedThis + 1 Denys Tsomenko + 1 Emil + 1 Ian Alexander Joiner + 1 Jesse Stuart + 1 Justin Joyce + 1 Michael + 1 SiLe Zhou + 1 Tyler Brinks + 1 benrsatori + 1 gstvg + 1 tomershaniii +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + From a629ddf89b0fbfb189d07b3130bc34aad9b9faf3 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 28 Feb 2025 22:07:39 -0800 Subject: [PATCH 065/286] Ignore escaped LIKE wildcards in MySQL (#1735) --- src/dialect/mod.rs | 27 ++++++++++++++++++++++++++ src/dialect/mysql.rs | 4 ++++ src/tokenizer.rs | 22 +++++++++++++++++++-- tests/sqlparser_common.rs | 40 ++++++++++++++++++++++++++------------- tests/sqlparser_mysql.rs | 11 +++++++++++ 5 files changed, 89 insertions(+), 15 deletions(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 1c32bc513..1cea6bc2b 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -201,6 +201,33 @@ pub trait Dialect: Debug + Any { false } + /// Determine whether the dialect strips the backslash when escaping LIKE wildcards (%, _). + /// + /// [MySQL] has a special case when escaping single quoted strings which leaves these unescaped + /// so they can be used in LIKE patterns without double-escaping (as is necessary in other + /// escaping dialects, such as [Snowflake]). Generally, special characters have escaping rules + /// causing them to be replaced with a different byte sequences (e.g. `'\0'` becoming the zero + /// byte), and the default if an escaped character does not have a specific escaping rule is to + /// strip the backslash (e.g. there is no rule for `h`, so `'\h' = 'h'`). MySQL's special case + /// for ignoring LIKE wildcard escapes is to *not* strip the backslash, so that `'\%' = '\\%'`. + /// This applies to all string literals though, not just those used in LIKE patterns. + /// + /// ```text + /// mysql> select '\_', hex('\\'), hex('_'), hex('\_'); + /// +----+-----------+----------+-----------+ + /// | \_ | hex('\\') | hex('_') | hex('\_') | + /// +----+-----------+----------+-----------+ + /// | \_ | 5C | 5F | 5C5F | + /// +----+-----------+----------+-----------+ + /// 1 row in set (0.00 sec) + /// ``` + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/string-literals.html + /// [Snowflake]: https://docs.snowflake.com/en/sql-reference/functions/like#usage-notes + fn ignores_wildcard_escapes(&self) -> bool { + false + } + /// Determine if the dialect supports string literals with `U&` prefix. /// This is used to specify Unicode code points in string literals. /// For example, in PostgreSQL, the following is a valid string literal: diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 8a0da87e4..cb86f2b47 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -62,6 +62,10 @@ impl Dialect for MySqlDialect { true } + fn ignores_wildcard_escapes(&self) -> bool { + true + } + fn supports_numeric_prefix(&self) -> bool { true } diff --git a/src/tokenizer.rs b/src/tokenizer.rs index bc0f0efeb..d33a7d8af 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -2011,8 +2011,13 @@ impl<'a> Tokenizer<'a> { num_consecutive_quotes = 0; if let Some(next) = chars.peek() { - if !self.unescape { - // In no-escape mode, the given query has to be saved completely including backslashes. + if !self.unescape + || (self.dialect.ignores_wildcard_escapes() + && (*next == '%' || *next == '_')) + { + // In no-escape mode, the given query has to be saved completely + // including backslashes. Similarly, with ignore_like_wildcard_escapes, + // the backslash is not stripped. s.push(ch); s.push(*next); chars.next(); // consume next @@ -3585,6 +3590,9 @@ mod tests { (r#"'\\a\\b\'c'"#, r#"\\a\\b\'c"#, r#"\a\b'c"#), (r#"'\'abcd'"#, r#"\'abcd"#, r#"'abcd"#), (r#"'''a''b'"#, r#"''a''b"#, r#"'a'b"#), + (r#"'\q'"#, r#"\q"#, r#"q"#), + (r#"'\%\_'"#, r#"\%\_"#, r#"%_"#), + (r#"'\\%\\_'"#, r#"\\%\\_"#, r#"\%\_"#), ] { let tokens = Tokenizer::new(&dialect, sql) .with_unescape(false) @@ -3618,6 +3626,16 @@ mod tests { compare(expected, tokens); } + + // MySQL special case for LIKE escapes + for (sql, expected) in [(r#"'\%'"#, r#"\%"#), (r#"'\_'"#, r#"\_"#)] { + let dialect = MySqlDialect {}; + let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); + + let expected = vec![Token::SingleQuotedString(expected.to_string())]; + + compare(expected, tokens); + } } #[test] diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 0a68d31e8..3c43ed61f 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -10387,15 +10387,8 @@ fn parse_with_recursion_limit() { #[test] fn parse_escaped_string_with_unescape() { - fn assert_mysql_query_value(sql: &str, quoted: &str) { - let stmt = TestedDialects::new(vec![ - Box::new(MySqlDialect {}), - Box::new(BigQueryDialect {}), - Box::new(SnowflakeDialect {}), - ]) - .one_statement_parses_to(sql, ""); - - match stmt { + fn assert_mysql_query_value(dialects: &TestedDialects, sql: &str, quoted: &str) { + match dialects.one_statement_parses_to(sql, "") { Statement::Query(query) => match *query.body { SetExpr::Select(value) => { let expr = expr_from_projection(only(&value.projection)); @@ -10411,17 +10404,38 @@ fn parse_escaped_string_with_unescape() { _ => unreachable!(), }; } + + let escaping_dialects = + &all_dialects_where(|dialect| dialect.supports_string_literal_backslash_escape()); + let no_wildcard_exception = &all_dialects_where(|dialect| { + dialect.supports_string_literal_backslash_escape() && !dialect.ignores_wildcard_escapes() + }); + let with_wildcard_exception = &all_dialects_where(|dialect| { + dialect.supports_string_literal_backslash_escape() && dialect.ignores_wildcard_escapes() + }); + let sql = r"SELECT 'I\'m fine'"; - assert_mysql_query_value(sql, "I'm fine"); + assert_mysql_query_value(escaping_dialects, sql, "I'm fine"); let sql = r#"SELECT 'I''m fine'"#; - assert_mysql_query_value(sql, "I'm fine"); + assert_mysql_query_value(escaping_dialects, sql, "I'm fine"); let sql = r#"SELECT 'I\"m fine'"#; - assert_mysql_query_value(sql, "I\"m fine"); + assert_mysql_query_value(escaping_dialects, sql, "I\"m fine"); let sql = r"SELECT 'Testing: \0 \\ \% \_ \b \n \r \t \Z \a \h \ '"; - assert_mysql_query_value(sql, "Testing: \0 \\ % _ \u{8} \n \r \t \u{1a} \u{7} h "); + assert_mysql_query_value( + no_wildcard_exception, + sql, + "Testing: \0 \\ % _ \u{8} \n \r \t \u{1a} \u{7} h ", + ); + + // check MySQL doesn't remove backslash from escaped LIKE wildcards + assert_mysql_query_value( + with_wildcard_exception, + sql, + "Testing: \0 \\ \\% \\_ \u{8} \n \r \t \u{1a} \u{7} h ", + ); } #[test] diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 15f79b4c2..f0774fcf5 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2627,6 +2627,17 @@ fn parse_rlike_and_regexp() { } } +#[test] +fn parse_like_with_escape() { + // verify backslash is not stripped for escaped wildcards + mysql().verified_only_select(r#"SELECT 'a\%c' LIKE 'a\%c'"#); + mysql().verified_only_select(r#"SELECT 'a\_c' LIKE 'a\_c'"#); + mysql().verified_only_select(r#"SELECT '%\_\%' LIKE '%\_\%'"#); + mysql().verified_only_select(r#"SELECT '\_\%' LIKE CONCAT('\_', '\%')"#); + mysql().verified_only_select(r#"SELECT 'a%c' LIKE 'a$%c' ESCAPE '$'"#); + mysql().verified_only_select(r#"SELECT 'a_c' LIKE 'a#_c' ESCAPE '#'"#); +} + #[test] fn parse_kill() { let stmt = mysql_and_generic().verified_stmt("KILL CONNECTION 5"); From 9e09b617e89b20b8c1f3ddccee35e1b04ac77188 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 28 Feb 2025 22:12:25 -0800 Subject: [PATCH 066/286] Parse SET NAMES syntax in Postgres (#1752) --- src/ast/mod.rs | 7 ++----- src/dialect/generic.rs | 4 ++++ src/dialect/mod.rs | 10 ++++++++++ src/dialect/mysql.rs | 4 ++++ src/dialect/postgresql.rs | 4 ++++ src/parser/mod.rs | 8 ++++---- tests/sqlparser_common.rs | 8 ++++++++ tests/sqlparser_mysql.rs | 6 +++--- 8 files changed, 39 insertions(+), 12 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 72be3ff6c..554ec19b7 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2956,10 +2956,8 @@ pub enum Statement { /// ```sql /// SET NAMES 'charset_name' [COLLATE 'collation_name'] /// ``` - /// - /// Note: this is a MySQL-specific statement. SetNames { - charset_name: String, + charset_name: Ident, collation_name: Option, }, /// ```sql @@ -4684,8 +4682,7 @@ impl fmt::Display for Statement { charset_name, collation_name, } => { - f.write_str("SET NAMES ")?; - f.write_str(charset_name)?; + write!(f, "SET NAMES {}", charset_name)?; if let Some(collation) = collation_name { f.write_str(" COLLATE ")?; diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 041d44bb2..c13d5aa69 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -155,4 +155,8 @@ impl Dialect for GenericDialect { fn supports_match_against(&self) -> bool { true } + + fn supports_set_names(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 1cea6bc2b..aeb097cfd 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -980,6 +980,16 @@ pub trait Dialect: Debug + Any { fn supports_order_by_all(&self) -> bool { false } + + /// Returns true if the dialect supports `SET NAMES [COLLATE ]`. + /// + /// - [MySQL](https://dev.mysql.com/doc/refman/8.4/en/set-names.html) + /// - [Postgres](https://www.postgresql.org/docs/17/sql-set.html) + /// + /// Note: Postgres doesn't support the `COLLATE` clause, but we permissively parse it anyway. + fn supports_set_names(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index cb86f2b47..0bdfc9bf3 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -137,6 +137,10 @@ impl Dialect for MySqlDialect { fn supports_match_against(&self) -> bool { true } + + fn supports_set_names(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 57ed0b684..9b08b8f32 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -254,4 +254,8 @@ impl Dialect for PostgreSqlDialect { fn supports_geometric_types(&self) -> bool { true } + + fn supports_set_names(&self) -> bool { + true + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f234fcc07..b11e57791 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10962,14 +10962,14 @@ impl<'a> Parser<'a> { OneOrManyWithParens::One(self.parse_object_name(false)?) }; - if matches!(&variables, OneOrManyWithParens::One(variable) if variable.to_string().eq_ignore_ascii_case("NAMES") - && dialect_of!(self is MySqlDialect | GenericDialect)) - { + let names = matches!(&variables, OneOrManyWithParens::One(variable) if variable.to_string().eq_ignore_ascii_case("NAMES")); + + if names && self.dialect.supports_set_names() { if self.parse_keyword(Keyword::DEFAULT) { return Ok(Statement::SetNamesDefault {}); } - let charset_name = self.parse_literal_string()?; + let charset_name = self.parse_identifier()?; let collation_name = if self.parse_one_of_keywords(&[Keyword::COLLATE]).is_some() { Some(self.parse_literal_string()?) } else { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 3c43ed61f..2c35c2438 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14631,3 +14631,11 @@ fn parse_array_type_def_with_brackets() { dialects.verified_stmt("SELECT x::INT[]"); dialects.verified_stmt("SELECT STRING_TO_ARRAY('1,2,3', ',')::INT[3]"); } + +#[test] +fn parse_set_names() { + let dialects = all_dialects_where(|d| d.supports_set_names()); + dialects.verified_stmt("SET NAMES 'UTF8'"); + dialects.verified_stmt("SET NAMES 'utf8'"); + dialects.verified_stmt("SET NAMES UTF8 COLLATE bogus"); +} diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index f0774fcf5..8d89ce4eb 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2696,7 +2696,7 @@ fn parse_set_names() { assert_eq!( stmt, Statement::SetNames { - charset_name: "utf8mb4".to_string(), + charset_name: "utf8mb4".into(), collation_name: None, } ); @@ -2705,7 +2705,7 @@ fn parse_set_names() { assert_eq!( stmt, Statement::SetNames { - charset_name: "utf8mb4".to_string(), + charset_name: "utf8mb4".into(), collation_name: Some("bogus".to_string()), } ); @@ -2716,7 +2716,7 @@ fn parse_set_names() { assert_eq!( stmt, vec![Statement::SetNames { - charset_name: "utf8mb4".to_string(), + charset_name: "utf8mb4".into(), collation_name: Some("bogus".to_string()), }] ); From d5dbe86da9c0591d5c2bf573b641d539f04e3028 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Sat, 1 Mar 2025 07:13:33 +0100 Subject: [PATCH 067/286] re-add support for nested comments in mssql (#1754) --- src/dialect/mssql.rs | 5 +++++ tests/sqlparser_mssql.rs | 16 ++++++++++++++++ 2 files changed, 21 insertions(+) diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 980f5ec3f..aeed1eb79 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -95,4 +95,9 @@ impl Dialect for MsSqlDialect { fn supports_timestamp_versioning(&self) -> bool { true } + + /// See + fn supports_nested_comments(&self) -> bool { + true + } } diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index ec565e50e..3f313af4f 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1620,6 +1620,22 @@ fn parse_create_table_with_valid_options() { } } +#[test] +fn parse_nested_slash_star_comment() { + let sql = r#" + select + /* + comment level 1 + /* + comment level 2 + */ + */ + 1; + "#; + let canonical = "SELECT 1"; + ms().one_statement_parses_to(sql, canonical); +} + #[test] fn parse_create_table_with_invalid_options() { let invalid_cases = vec![ From 6ec5223f50e4056b8f07a61141e7f3e67b25d5e3 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Tue, 4 Mar 2025 06:59:39 +0100 Subject: [PATCH 068/286] Extend support for INDEX parsing (#1707) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 17 ++- src/ast/dml.rs | 31 ++++- src/ast/mod.rs | 3 +- src/ast/spans.rs | 5 +- src/keywords.rs | 5 + src/parser/mod.rs | 92 ++++++++++++--- tests/sqlparser_common.rs | 83 +++++++------ tests/sqlparser_postgres.rs | 230 ++++++++++++++++++++++++++++++++++++ 8 files changed, 405 insertions(+), 61 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index bb85eb06c..619631432 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1174,13 +1174,20 @@ impl fmt::Display for KeyOrIndexDisplay { /// [1]: https://dev.mysql.com/doc/refman/8.0/en/create-table.html /// [2]: https://dev.mysql.com/doc/refman/8.0/en/create-index.html /// [3]: https://www.postgresql.org/docs/14/sql-createindex.html -#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum IndexType { BTree, Hash, - // TODO add Postgresql's possible indexes + GIN, + GiST, + SPGiST, + BRIN, + Bloom, + /// Users may define their own index types, which would + /// not be covered by the above variants. + Custom(Ident), } impl fmt::Display for IndexType { @@ -1188,6 +1195,12 @@ impl fmt::Display for IndexType { match self { Self::BTree => write!(f, "BTREE"), Self::Hash => write!(f, "HASH"), + Self::GIN => write!(f, "GIN"), + Self::GiST => write!(f, "GIST"), + Self::SPGiST => write!(f, "SPGIST"), + Self::BRIN => write!(f, "BRIN"), + Self::Bloom => write!(f, "BLOOM"), + Self::Custom(name) => write!(f, "{}", name), } } } diff --git a/src/ast/dml.rs b/src/ast/dml.rs index 8cfc67414..ccea7fbcb 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -34,12 +34,31 @@ pub use super::ddl::{ColumnDef, TableConstraint}; use super::{ display_comma_separated, display_separated, query::InputFormatClause, Assignment, ClusteredBy, CommentDef, Expr, FileFormat, FromTable, HiveDistributionStyle, HiveFormat, HiveIOFormat, - HiveRowFormat, Ident, InsertAliases, MysqlInsertPriority, ObjectName, OnCommit, OnInsert, - OneOrManyWithParens, OrderByExpr, Query, RowAccessPolicy, SelectItem, Setting, SqlOption, - SqliteOnConflict, StorageSerializationPolicy, TableEngine, TableObject, TableWithJoins, Tag, - WrappedCollection, + HiveRowFormat, Ident, IndexType, InsertAliases, MysqlInsertPriority, ObjectName, OnCommit, + OnInsert, OneOrManyWithParens, OrderByExpr, Query, RowAccessPolicy, SelectItem, Setting, + SqlOption, SqliteOnConflict, StorageSerializationPolicy, TableEngine, TableObject, + TableWithJoins, Tag, WrappedCollection, }; +/// Index column type. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct IndexColumn { + pub column: OrderByExpr, + pub operator_class: Option, +} + +impl Display for IndexColumn { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}", self.column)?; + if let Some(operator_class) = &self.operator_class { + write!(f, " {}", operator_class)?; + } + Ok(()) + } +} + /// CREATE INDEX statement. #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -49,8 +68,8 @@ pub struct CreateIndex { pub name: Option, #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] pub table_name: ObjectName, - pub using: Option, - pub columns: Vec, + pub using: Option, + pub columns: Vec, pub unique: bool, pub concurrently: bool, pub if_not_exists: bool, diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 554ec19b7..e5e4aef05 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -58,7 +58,7 @@ pub use self::ddl::{ ReferentialAction, TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; -pub use self::dml::{CreateIndex, CreateTable, Delete, Insert}; +pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; pub use self::query::{ AfterMatchSkip, ConnectBy, Cte, CteAsMaterialized, Distinct, EmptyMatchesMode, @@ -91,6 +91,7 @@ pub use self::value::{ use crate::ast::helpers::key_value_options::KeyValueOptions; use crate::ast::helpers::stmt_data_loading::{StageLoadSelectItem, StageParamsObject}; + #[cfg(feature = "visitor")] pub use visitor::*; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 38e9e258e..0a64fb8ea 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -704,7 +704,7 @@ impl Spanned for CreateIndex { let CreateIndex { name, table_name, - using, + using: _, columns, unique: _, // bool concurrently: _, // bool @@ -719,8 +719,7 @@ impl Spanned for CreateIndex { name.iter() .map(|i| i.span()) .chain(core::iter::once(table_name.span())) - .chain(using.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span())) + .chain(columns.iter().map(|i| i.column.span())) .chain(include.iter().map(|i| i.span)) .chain(with.iter().map(|i| i.span())) .chain(predicate.iter().map(|i| i.span())), diff --git a/src/keywords.rs b/src/keywords.rs index a6854f073..bda817df9 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -137,11 +137,13 @@ define_keywords!( BIT, BLOB, BLOCK, + BLOOM, BLOOMFILTER, BOOL, BOOLEAN, BOTH, BOX, + BRIN, BROWSE, BTREE, BUCKET, @@ -386,6 +388,8 @@ define_keywords!( GENERATED, GEOGRAPHY, GET, + GIN, + GIST, GLOBAL, GRANT, GRANTED, @@ -805,6 +809,7 @@ define_keywords!( SPATIAL, SPECIFIC, SPECIFICTYPE, + SPGIST, SQL, SQLEXCEPTION, SQLSTATE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b11e57791..b34415388 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3955,6 +3955,18 @@ impl<'a> Parser<'a> { true } + /// If the current token is one of the given `keywords`, returns the keyword + /// that matches, without consuming the token. Otherwise, returns [`None`]. + #[must_use] + pub fn peek_one_of_keywords(&self, keywords: &[Keyword]) -> Option { + for keyword in keywords { + if self.peek_keyword(*keyword) { + return Some(*keyword); + } + } + None + } + /// If the current token is one of the given `keywords`, consume the token /// and return the keyword that matches. Otherwise, no tokens are consumed /// and returns [`None`]. @@ -6406,12 +6418,13 @@ impl<'a> Parser<'a> { }; let table_name = self.parse_object_name(false)?; let using = if self.parse_keyword(Keyword::USING) { - Some(self.parse_identifier()?) + Some(self.parse_index_type()?) } else { None }; + self.expect_token(&Token::LParen)?; - let columns = self.parse_comma_separated(Parser::parse_order_by_expr)?; + let columns = self.parse_comma_separated(Parser::parse_create_index_expr)?; self.expect_token(&Token::RParen)?; let include = if self.parse_keyword(Keyword::INCLUDE) { @@ -7629,16 +7642,30 @@ impl<'a> Parser<'a> { } pub fn parse_index_type(&mut self) -> Result { - if self.parse_keyword(Keyword::BTREE) { - Ok(IndexType::BTree) + Ok(if self.parse_keyword(Keyword::BTREE) { + IndexType::BTree } else if self.parse_keyword(Keyword::HASH) { - Ok(IndexType::Hash) - } else { - self.expected("index type {BTREE | HASH}", self.peek_token()) - } + IndexType::Hash + } else if self.parse_keyword(Keyword::GIN) { + IndexType::GIN + } else if self.parse_keyword(Keyword::GIST) { + IndexType::GiST + } else if self.parse_keyword(Keyword::SPGIST) { + IndexType::SPGiST + } else if self.parse_keyword(Keyword::BRIN) { + IndexType::BRIN + } else if self.parse_keyword(Keyword::BLOOM) { + IndexType::Bloom + } else { + IndexType::Custom(self.parse_identifier()?) + }) } - /// Parse [USING {BTREE | HASH}] + /// Optionally parse the `USING` keyword, followed by an [IndexType] + /// Example: + /// ```sql + //// USING BTREE (name, age DESC) + /// ``` pub fn parse_optional_using_then_index_type( &mut self, ) -> Result, ParserError> { @@ -13631,10 +13658,42 @@ impl<'a> Parser<'a> { } } - /// Parse an expression, optionally followed by ASC or DESC (used in ORDER BY) + /// Parse an [OrderByExpr] expression. pub fn parse_order_by_expr(&mut self) -> Result { + self.parse_order_by_expr_inner(false) + .map(|(order_by, _)| order_by) + } + + /// Parse an [IndexColumn]. + pub fn parse_create_index_expr(&mut self) -> Result { + self.parse_order_by_expr_inner(true) + .map(|(column, operator_class)| IndexColumn { + column, + operator_class, + }) + } + + fn parse_order_by_expr_inner( + &mut self, + with_operator_class: bool, + ) -> Result<(OrderByExpr, Option), ParserError> { let expr = self.parse_expr()?; + let operator_class: Option = if with_operator_class { + // We check that if non of the following keywords are present, then we parse an + // identifier as operator class. + if self + .peek_one_of_keywords(&[Keyword::ASC, Keyword::DESC, Keyword::NULLS, Keyword::WITH]) + .is_some() + { + None + } else { + self.maybe_parse(|parser| parser.parse_identifier())? + } + } else { + None + }; + let options = self.parse_order_by_options()?; let with_fill = if dialect_of!(self is ClickHouseDialect | GenericDialect) @@ -13645,11 +13704,14 @@ impl<'a> Parser<'a> { None }; - Ok(OrderByExpr { - expr, - options, - with_fill, - }) + Ok(( + OrderByExpr { + expr, + options, + with_fill, + }, + operator_class, + )) } fn parse_order_by_options(&mut self) -> Result { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 2c35c2438..a8ccd70a7 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8842,22 +8842,28 @@ fn ensure_multiple_dialects_are_tested() { #[test] fn parse_create_index() { let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test(name,age DESC)"; - let indexed_columns = vec![ - OrderByExpr { - expr: Expr::Identifier(Ident::new("name")), - options: OrderByOptions { - asc: None, - nulls_first: None, + let indexed_columns: Vec = vec![ + IndexColumn { + operator_class: None, + column: OrderByExpr { + expr: Expr::Identifier(Ident::new("name")), + with_fill: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, }, - with_fill: None, }, - OrderByExpr { - expr: Expr::Identifier(Ident::new("age")), - options: OrderByOptions { - asc: Some(false), - nulls_first: None, + IndexColumn { + operator_class: None, + column: OrderByExpr { + expr: Expr::Identifier(Ident::new("age")), + with_fill: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, }, - with_fill: None, }, ]; match verified_stmt(sql) { @@ -8881,23 +8887,29 @@ fn parse_create_index() { #[test] fn test_create_index_with_using_function() { - let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test USING btree (name,age DESC)"; - let indexed_columns = vec![ - OrderByExpr { - expr: Expr::Identifier(Ident::new("name")), - options: OrderByOptions { - asc: None, - nulls_first: None, + let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test USING BTREE (name,age DESC)"; + let indexed_columns: Vec = vec![ + IndexColumn { + operator_class: None, + column: OrderByExpr { + expr: Expr::Identifier(Ident::new("name")), + with_fill: None, + options: OrderByOptions { + asc: None, + nulls_first: None, + }, }, - with_fill: None, }, - OrderByExpr { - expr: Expr::Identifier(Ident::new("age")), - options: OrderByOptions { - asc: Some(false), - nulls_first: None, + IndexColumn { + operator_class: None, + column: OrderByExpr { + expr: Expr::Identifier(Ident::new("age")), + with_fill: None, + options: OrderByOptions { + asc: Some(false), + nulls_first: None, + }, }, - with_fill: None, }, ]; match verified_stmt(sql) { @@ -8916,7 +8928,7 @@ fn test_create_index_with_using_function() { }) => { assert_eq!("idx_name", name.to_string()); assert_eq!("test", table_name.to_string()); - assert_eq!("btree", using.unwrap().to_string()); + assert_eq!("BTREE", using.unwrap().to_string()); assert_eq!(indexed_columns, columns); assert!(unique); assert!(!concurrently); @@ -8931,13 +8943,16 @@ fn test_create_index_with_using_function() { #[test] fn test_create_index_with_with_clause() { let sql = "CREATE UNIQUE INDEX title_idx ON films(title) WITH (fillfactor = 70, single_param)"; - let indexed_columns = vec![OrderByExpr { - expr: Expr::Identifier(Ident::new("title")), - options: OrderByOptions { - asc: None, - nulls_first: None, + let indexed_columns: Vec = vec![IndexColumn { + column: OrderByExpr { + expr: Expr::Identifier(Ident::new("title")), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, }, - with_fill: None, + operator_class: None, }]; let with_parameters = vec![ Expr::BinaryOp { diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 7508218f9..0dfcc24ea 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2509,6 +2509,236 @@ fn parse_create_anonymous_index() { } } +#[test] +/// Test to verify the correctness of parsing the `CREATE INDEX` statement with optional operator classes. +/// +/// # Implementative details +/// +/// At this time, since the parser library is not intended to take care of the semantics of the SQL statements, +/// there is no way to verify the correctness of the operator classes, nor whether they are valid for the given +/// index type. This test is only intended to verify that the parser can correctly parse the statement. For this +/// reason, the test includes a `totally_not_valid` operator class. +fn parse_create_indices_with_operator_classes() { + let indices = [ + IndexType::GIN, + IndexType::GiST, + IndexType::SPGiST, + IndexType::Custom("CustomIndexType".into()), + ]; + let operator_classes: [Option; 4] = [ + None, + Some("gin_trgm_ops".into()), + Some("gist_trgm_ops".into()), + Some("totally_not_valid".into()), + ]; + + for expected_index_type in indices { + for expected_operator_class in &operator_classes { + let single_column_sql_statement = format!( + "CREATE INDEX the_index_name ON users USING {expected_index_type} (concat_users_name(first_name, last_name){})", + expected_operator_class.as_ref().map(|oc| format!(" {}", oc)) + .unwrap_or_default() + ); + let multi_column_sql_statement = format!( + "CREATE INDEX the_index_name ON users USING {expected_index_type} (column_name,concat_users_name(first_name, last_name){})", + expected_operator_class.as_ref().map(|oc| format!(" {}", oc)) + .unwrap_or_default() + ); + + let expected_function_column = IndexColumn { + column: OrderByExpr { + expr: Expr::Function(Function { + name: ObjectName(vec![ObjectNamePart::Identifier(Ident { + value: "concat_users_name".to_owned(), + quote_style: None, + span: Span::empty(), + })]), + uses_odbc_syntax: false, + parameters: FunctionArguments::None, + args: FunctionArguments::List(FunctionArgumentList { + duplicate_treatment: None, + args: vec![ + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Identifier( + Ident { + value: "first_name".to_owned(), + quote_style: None, + span: Span::empty(), + }, + ))), + FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Identifier( + Ident { + value: "last_name".to_owned(), + quote_style: None, + span: Span::empty(), + }, + ))), + ], + clauses: vec![], + }), + filter: None, + null_treatment: None, + over: None, + within_group: vec![], + }), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }, + operator_class: expected_operator_class.clone(), + }; + + match pg().verified_stmt(&single_column_sql_statement) { + Statement::CreateIndex(CreateIndex { + name: Some(ObjectName(name)), + table_name: ObjectName(table_name), + using: Some(using), + columns, + unique: false, + concurrently: false, + if_not_exists: false, + include, + nulls_distinct: None, + with, + predicate: None, + }) => { + assert_eq_vec(&["the_index_name"], &name); + assert_eq_vec(&["users"], &table_name); + assert_eq!(expected_index_type, using); + assert_eq!(expected_function_column, columns[0],); + assert!(include.is_empty()); + assert!(with.is_empty()); + } + _ => unreachable!(), + } + + match pg().verified_stmt(&multi_column_sql_statement) { + Statement::CreateIndex(CreateIndex { + name: Some(ObjectName(name)), + table_name: ObjectName(table_name), + using: Some(using), + columns, + unique: false, + concurrently: false, + if_not_exists: false, + include, + nulls_distinct: None, + with, + predicate: None, + }) => { + assert_eq_vec(&["the_index_name"], &name); + assert_eq_vec(&["users"], &table_name); + assert_eq!(expected_index_type, using); + assert_eq!( + IndexColumn { + column: OrderByExpr { + expr: Expr::Identifier(Ident { + value: "column_name".to_owned(), + quote_style: None, + span: Span::empty() + }), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }, + operator_class: None + }, + columns[0], + ); + assert_eq!(expected_function_column, columns[1],); + assert!(include.is_empty()); + assert!(with.is_empty()); + } + _ => unreachable!(), + } + } + } +} + +#[test] +fn parse_create_bloom() { + let sql = + "CREATE INDEX bloomidx ON tbloom USING BLOOM (i1,i2,i3) WITH (length = 80, col1 = 2, col2 = 2, col3 = 4)"; + match pg().verified_stmt(sql) { + Statement::CreateIndex(CreateIndex { + name: Some(ObjectName(name)), + table_name: ObjectName(table_name), + using: Some(using), + columns, + unique: false, + concurrently: false, + if_not_exists: false, + include, + nulls_distinct: None, + with, + predicate: None, + }) => { + assert_eq_vec(&["bloomidx"], &name); + assert_eq_vec(&["tbloom"], &table_name); + assert_eq!(IndexType::Bloom, using); + assert_eq_vec(&["i1", "i2", "i3"], &columns); + assert!(include.is_empty()); + assert_eq!( + vec![ + Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("length"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Value(number("80").into())), + }, + Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("col1"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Value(number("2").into())), + }, + Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("col2"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Value(number("2").into())), + }, + Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("col3"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Value(number("4").into())), + }, + ], + with + ); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_create_brin() { + let sql = "CREATE INDEX brin_sensor_data_recorded_at ON sensor_data USING BRIN (recorded_at)"; + match pg().verified_stmt(sql) { + Statement::CreateIndex(CreateIndex { + name: Some(ObjectName(name)), + table_name: ObjectName(table_name), + using: Some(using), + columns, + unique: false, + concurrently: false, + if_not_exists: false, + include, + nulls_distinct: None, + with, + predicate: None, + }) => { + assert_eq_vec(&["brin_sensor_data_recorded_at"], &name); + assert_eq_vec(&["sensor_data"], &table_name); + assert_eq!(IndexType::BRIN, using); + assert_eq_vec(&["recorded_at"], &columns); + assert!(include.is_empty()); + assert!(with.is_empty()); + } + _ => unreachable!(), + } +} + #[test] fn parse_create_index_concurrently() { let sql = "CREATE INDEX CONCURRENTLY IF NOT EXISTS my_index ON my_table(col1,col2)"; From 1e54a34acdea192c3d67330e604e0bf9ce8bf866 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Tue, 11 Mar 2025 23:34:18 -0700 Subject: [PATCH 069/286] Parse MySQL `ALTER TABLE DROP FOREIGN KEY` syntax (#1762) --- src/ast/ddl.rs | 13 ++++++++++++- src/ast/spans.rs | 1 + src/parser/mod.rs | 7 ++++--- tests/sqlparser_mysql.rs | 10 ++++++++++ 4 files changed, 27 insertions(+), 4 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 619631432..99d8521cd 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -151,8 +151,18 @@ pub enum AlterTableOperation { }, /// `DROP PRIMARY KEY` /// - /// Note: this is a MySQL-specific operation. + /// Note: this is a [MySQL]-specific operation. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html DropPrimaryKey, + /// `DROP FOREIGN KEY ` + /// + /// Note: this is a [MySQL]-specific operation. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html + DropForeignKey { + name: Ident, + }, /// `ENABLE ALWAYS RULE rewrite_rule_name` /// /// Note: this is a PostgreSQL-specific operation. @@ -530,6 +540,7 @@ impl fmt::Display for AlterTableOperation { ) } AlterTableOperation::DropPrimaryKey => write!(f, "DROP PRIMARY KEY"), + AlterTableOperation::DropForeignKey { name } => write!(f, "DROP FOREIGN KEY {name}"), AlterTableOperation::DropColumn { column_name, if_exists, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 0a64fb8ea..62ca9dc0a 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -998,6 +998,7 @@ impl Spanned for AlterTableOperation { .span() .union_opt(&with_name.as_ref().map(|n| n.span)), AlterTableOperation::DropPrimaryKey => Span::empty(), + AlterTableOperation::DropForeignKey { name } => name.span, AlterTableOperation::EnableAlwaysRule { name } => name.span, AlterTableOperation::EnableAlwaysTrigger { name } => name.span, AlterTableOperation::EnableReplicaRule { name } => name.span, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b34415388..2b95d6740 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7998,10 +7998,11 @@ impl<'a> Parser<'a> { name, drop_behavior, } - } else if self.parse_keywords(&[Keyword::PRIMARY, Keyword::KEY]) - && dialect_of!(self is MySqlDialect | GenericDialect) - { + } else if self.parse_keywords(&[Keyword::PRIMARY, Keyword::KEY]) { AlterTableOperation::DropPrimaryKey + } else if self.parse_keywords(&[Keyword::FOREIGN, Keyword::KEY]) { + let name = self.parse_identifier()?; + AlterTableOperation::DropForeignKey { name } } else if self.parse_keyword(Keyword::PROJECTION) && dialect_of!(self is ClickHouseDialect|GenericDialect) { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 8d89ce4eb..560ea9dae 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2273,6 +2273,16 @@ fn parse_alter_table_drop_primary_key() { ); } +#[test] +fn parse_alter_table_drop_foreign_key() { + assert_matches!( + alter_table_op( + mysql_and_generic().verified_stmt("ALTER TABLE tab DROP FOREIGN KEY foo_ibfk_1") + ), + AlterTableOperation::DropForeignKey { name } if name.value == "foo_ibfk_1" + ); +} + #[test] fn parse_alter_table_change_column() { let expected_name = ObjectName::from(vec![Ident::new("orders")]); From 3392623b00718be68268fa65298ae98f74d2eda5 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Wed, 12 Mar 2025 11:42:51 +0100 Subject: [PATCH 070/286] add support for `with` clauses (CTEs) in `delete` statements (#1764) --- src/ast/query.rs | 2 ++ src/ast/spans.rs | 1 + src/parser/mod.rs | 22 ++++++++++++++++++++++ tests/sqlparser_common.rs | 27 +++++++++++++++++++++++++++ 4 files changed, 52 insertions(+) diff --git a/src/ast/query.rs b/src/ast/query.rs index bed991114..12f72932f 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -156,6 +156,7 @@ pub enum SetExpr { Values(Values), Insert(Statement), Update(Statement), + Delete(Statement), Table(Box
), } @@ -178,6 +179,7 @@ impl fmt::Display for SetExpr { SetExpr::Values(v) => write!(f, "{v}"), SetExpr::Insert(v) => write!(f, "{v}"), SetExpr::Update(v) => write!(f, "{v}"), + SetExpr::Delete(v) => write!(f, "{v}"), SetExpr::Table(t) => write!(f, "{t}"), SetExpr::SetOperation { left, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 62ca9dc0a..8c3eff3c3 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -191,6 +191,7 @@ impl Spanned for SetExpr { SetExpr::Insert(statement) => statement.span(), SetExpr::Table(_) => Span::empty(), SetExpr::Update(statement) => statement.span(), + SetExpr::Delete(statement) => statement.span(), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2b95d6740..400a9480f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10050,6 +10050,13 @@ impl<'a> Parser<'a> { Ok(parent_type(inside_type.into())) } + /// Parse a DELETE statement, returning a `Box`ed SetExpr + /// + /// This is used to reduce the size of the stack frames in debug builds + fn parse_delete_setexpr_boxed(&mut self) -> Result, ParserError> { + Ok(Box::new(SetExpr::Delete(self.parse_delete()?))) + } + pub fn parse_delete(&mut self) -> Result { let (tables, with_from_keyword) = if !self.parse_keyword(Keyword::FROM) { // `FROM` keyword is optional in BigQuery SQL. @@ -10249,6 +10256,21 @@ impl<'a> Parser<'a> { format_clause: None, } .into()) + } else if self.parse_keyword(Keyword::DELETE) { + Ok(Query { + with, + body: self.parse_delete_setexpr_boxed()?, + limit: None, + limit_by: vec![], + order_by: None, + offset: None, + fetch: None, + locks: vec![], + for_clause: None, + settings: None, + format_clause: None, + } + .into()) } else { let body = self.parse_query_body(self.dialect.prec_unknown())?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a8ccd70a7..8225d3672 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7383,6 +7383,33 @@ fn parse_recursive_cte() { assert_eq!(with.cte_tables.first().unwrap(), &expected); } +#[test] +fn parse_cte_in_data_modification_statements() { + match verified_stmt("WITH x AS (SELECT 1) UPDATE t SET bar = (SELECT * FROM x)") { + Statement::Query(query) => { + assert_eq!(query.with.unwrap().to_string(), "WITH x AS (SELECT 1)"); + assert!(matches!(*query.body, SetExpr::Update(_))); + } + other => panic!("Expected: UPDATE, got: {:?}", other), + } + + match verified_stmt("WITH t (x) AS (SELECT 9) DELETE FROM q WHERE id IN (SELECT x FROM t)") { + Statement::Query(query) => { + assert_eq!(query.with.unwrap().to_string(), "WITH t (x) AS (SELECT 9)"); + assert!(matches!(*query.body, SetExpr::Delete(_))); + } + other => panic!("Expected: DELETE, got: {:?}", other), + } + + match verified_stmt("WITH x AS (SELECT 42) INSERT INTO t SELECT foo FROM x") { + Statement::Query(query) => { + assert_eq!(query.with.unwrap().to_string(), "WITH x AS (SELECT 42)"); + assert!(matches!(*query.body, SetExpr::Insert(_))); + } + other => panic!("Expected: INSERT, got: {:?}", other), + } +} + #[test] fn parse_derived_tables() { let sql = "SELECT a.x, b.y FROM (SELECT x FROM foo) AS a CROSS JOIN (SELECT y FROM bar) AS b"; From 85f855150fddf9326b0c2de0a5808fb46a1f2527 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Wed, 12 Mar 2025 22:02:39 +0200 Subject: [PATCH 071/286] SET with a list of comma separated assignments (#1757) --- src/ast/mod.rs | 361 +++++++++++++++++++++--------------- src/ast/spans.rs | 15 +- src/dialect/mod.rs | 10 + src/dialect/mssql.rs | 1 + src/dialect/mysql.rs | 4 + src/keywords.rs | 2 + src/parser/mod.rs | 294 +++++++++++++++++++---------- tests/sqlparser_common.rs | 119 ++++++------ tests/sqlparser_hive.rs | 17 +- tests/sqlparser_mssql.rs | 8 +- tests/sqlparser_mysql.rs | 22 +-- tests/sqlparser_postgres.rs | 82 ++++---- 12 files changed, 562 insertions(+), 373 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index e5e4aef05..8ab3fc0f1 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2394,6 +2394,168 @@ pub enum CreatePolicyCommand { Delete, } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum Set { + /// SQL Standard-style + /// SET a = 1; + SingleAssignment { + local: bool, + hivevar: bool, + variable: ObjectName, + values: Vec, + }, + /// Snowflake-style + /// SET (a, b, ..) = (1, 2, ..); + ParenthesizedAssignments { + variables: Vec, + values: Vec, + }, + /// MySQL-style + /// SET a = 1, b = 2, ..; + MultipleAssignments { assignments: Vec }, + /// MS-SQL session + /// + /// See + SetSessionParam(SetSessionParamKind), + /// ```sql + /// SET [ SESSION | LOCAL ] ROLE role_name + /// ``` + /// + /// Sets session state. Examples: [ANSI][1], [Postgresql][2], [MySQL][3], and [Oracle][4] + /// + /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#set-role-statement + /// [2]: https://www.postgresql.org/docs/14/sql-set-role.html + /// [3]: https://dev.mysql.com/doc/refman/8.0/en/set-role.html + /// [4]: https://docs.oracle.com/cd/B19306_01/server.102/b14200/statements_10004.htm + SetRole { + /// Non-ANSI optional identifier to inform if the role is defined inside the current session (`SESSION`) or transaction (`LOCAL`). + context_modifier: ContextModifier, + /// Role name. If NONE is specified, then the current role name is removed. + role_name: Option, + }, + /// ```sql + /// SET TIME ZONE + /// ``` + /// + /// Note: this is a PostgreSQL-specific statements + /// `SET TIME ZONE ` is an alias for `SET timezone TO ` in PostgreSQL + /// However, we allow it for all dialects. + SetTimeZone { local: bool, value: Expr }, + /// ```sql + /// SET NAMES 'charset_name' [COLLATE 'collation_name'] + /// ``` + SetNames { + charset_name: Ident, + collation_name: Option, + }, + /// ```sql + /// SET NAMES DEFAULT + /// ``` + /// + /// Note: this is a MySQL-specific statement. + SetNamesDefault {}, + /// ```sql + /// SET TRANSACTION ... + /// ``` + SetTransaction { + modes: Vec, + snapshot: Option, + session: bool, + }, +} + +impl Display for Set { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + Self::ParenthesizedAssignments { variables, values } => write!( + f, + "SET ({}) = ({})", + display_comma_separated(variables), + display_comma_separated(values) + ), + Self::MultipleAssignments { assignments } => { + write!(f, "SET {}", display_comma_separated(assignments)) + } + Self::SetRole { + context_modifier, + role_name, + } => { + let role_name = role_name.clone().unwrap_or_else(|| Ident::new("NONE")); + write!(f, "SET{context_modifier} ROLE {role_name}") + } + Self::SetSessionParam(kind) => write!(f, "SET {kind}"), + Self::SetTransaction { + modes, + snapshot, + session, + } => { + if *session { + write!(f, "SET SESSION CHARACTERISTICS AS TRANSACTION")?; + } else { + write!(f, "SET TRANSACTION")?; + } + if !modes.is_empty() { + write!(f, " {}", display_comma_separated(modes))?; + } + if let Some(snapshot_id) = snapshot { + write!(f, " SNAPSHOT {snapshot_id}")?; + } + Ok(()) + } + Self::SetTimeZone { local, value } => { + f.write_str("SET ")?; + if *local { + f.write_str("LOCAL ")?; + } + write!(f, "TIME ZONE {value}") + } + Self::SetNames { + charset_name, + collation_name, + } => { + write!(f, "SET NAMES {}", charset_name)?; + + if let Some(collation) = collation_name { + f.write_str(" COLLATE ")?; + f.write_str(collation)?; + }; + + Ok(()) + } + Self::SetNamesDefault {} => { + f.write_str("SET NAMES DEFAULT")?; + + Ok(()) + } + Set::SingleAssignment { + local, + hivevar, + variable, + values, + } => { + write!( + f, + "SET {}{}{} = {}", + if *local { "LOCAL " } else { "" }, + if *hivevar { "HIVEVAR:" } else { "" }, + variable, + display_comma_separated(values) + ) + } + } + } +} + +/// Convert a `Set` into a `Statement`. +/// Convenience function, instead of writing `Statement::Set(Set::Set...{...})` +impl From for Statement { + fn from(set: Set) -> Self { + Statement::Set(set) + } +} + /// A top-level statement (SELECT, INSERT, CREATE, etc.) #[allow(clippy::large_enum_variant)] #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -2419,6 +2581,7 @@ pub enum Statement { compute_statistics: bool, has_table_keyword: bool, }, + Set(Set), /// ```sql /// TRUNCATE /// ``` @@ -2846,7 +3009,10 @@ pub enum Statement { /// DROP CONNECTOR /// ``` /// See [Hive](https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=27362034#LanguageManualDDL-DropConnector) - DropConnector { if_exists: bool, name: Ident }, + DropConnector { + if_exists: bool, + name: Ident, + }, /// ```sql /// DECLARE /// ``` @@ -2854,7 +3020,9 @@ pub enum Statement { /// /// Note: this is a PostgreSQL-specific statement, /// but may also compatible with other SQL. - Declare { stmts: Vec }, + Declare { + stmts: Vec, + }, /// ```sql /// CREATE EXTENSION [ IF NOT EXISTS ] extension_name /// [ WITH ] [ SCHEMA schema_name ] @@ -2916,67 +3084,23 @@ pub enum Statement { /// /// Note: this is a PostgreSQL-specific statement, /// but may also compatible with other SQL. - Discard { object_type: DiscardObject }, - /// ```sql - /// SET [ SESSION | LOCAL ] ROLE role_name - /// ``` - /// - /// Sets session state. Examples: [ANSI][1], [Postgresql][2], [MySQL][3], and [Oracle][4] - /// - /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#set-role-statement - /// [2]: https://www.postgresql.org/docs/14/sql-set-role.html - /// [3]: https://dev.mysql.com/doc/refman/8.0/en/set-role.html - /// [4]: https://docs.oracle.com/cd/B19306_01/server.102/b14200/statements_10004.htm - SetRole { - /// Non-ANSI optional identifier to inform if the role is defined inside the current session (`SESSION`) or transaction (`LOCAL`). - context_modifier: ContextModifier, - /// Role name. If NONE is specified, then the current role name is removed. - role_name: Option, - }, - /// ```sql - /// SET = expression; - /// SET (variable[, ...]) = (expression[, ...]); - /// ``` - /// - /// Note: this is not a standard SQL statement, but it is supported by at - /// least MySQL and PostgreSQL. Not all MySQL-specific syntactic forms are - /// supported yet. - SetVariable { - local: bool, - hivevar: bool, - variables: OneOrManyWithParens, - value: Vec, + Discard { + object_type: DiscardObject, }, - /// ```sql - /// SET TIME ZONE - /// ``` - /// - /// Note: this is a PostgreSQL-specific statements - /// `SET TIME ZONE ` is an alias for `SET timezone TO ` in PostgreSQL - SetTimeZone { local: bool, value: Expr }, - /// ```sql - /// SET NAMES 'charset_name' [COLLATE 'collation_name'] - /// ``` - SetNames { - charset_name: Ident, - collation_name: Option, - }, - /// ```sql - /// SET NAMES DEFAULT - /// ``` - /// - /// Note: this is a MySQL-specific statement. - SetNamesDefault {}, /// `SHOW FUNCTIONS` /// /// Note: this is a Presto-specific statement. - ShowFunctions { filter: Option }, + ShowFunctions { + filter: Option, + }, /// ```sql /// SHOW /// ``` /// /// Note: this is a PostgreSQL-specific statement. - ShowVariable { variable: Vec }, + ShowVariable { + variable: Vec, + }, /// ```sql /// SHOW [GLOBAL | SESSION] STATUS [LIKE 'pattern' | WHERE expr] /// ``` @@ -3060,7 +3184,9 @@ pub enum Statement { /// ``` /// /// Note: this is a MySQL-specific statement. - ShowCollation { filter: Option }, + ShowCollation { + filter: Option, + }, /// ```sql /// `USE ...` /// ``` @@ -3103,14 +3229,6 @@ pub enum Statement { has_end_keyword: bool, }, /// ```sql - /// SET TRANSACTION ... - /// ``` - SetTransaction { - modes: Vec, - snapshot: Option, - session: bool, - }, - /// ```sql /// COMMENT ON ... /// ``` /// @@ -3329,7 +3447,10 @@ pub enum Statement { /// ``` /// /// Note: this is a PostgreSQL-specific statement. - Deallocate { name: Ident, prepare: bool }, + Deallocate { + name: Ident, + prepare: bool, + }, /// ```sql /// An `EXECUTE` statement /// ``` @@ -3415,11 +3536,15 @@ pub enum Statement { /// SAVEPOINT /// ``` /// Define a new savepoint within the current transaction - Savepoint { name: Ident }, + Savepoint { + name: Ident, + }, /// ```sql /// RELEASE [ SAVEPOINT ] savepoint_name /// ``` - ReleaseSavepoint { name: Ident }, + ReleaseSavepoint { + name: Ident, + }, /// A `MERGE` statement. /// /// ```sql @@ -3499,7 +3624,9 @@ pub enum Statement { /// LOCK TABLES [READ [LOCAL] | [LOW_PRIORITY] WRITE] /// ``` /// Note: this is a MySQL-specific statement. See - LockTables { tables: Vec }, + LockTables { + tables: Vec, + }, /// ```sql /// UNLOCK TABLES /// ``` @@ -3533,14 +3660,18 @@ pub enum Statement { /// listen for a notification channel /// /// See Postgres - LISTEN { channel: Ident }, + LISTEN { + channel: Ident, + }, /// ```sql /// UNLISTEN /// ``` /// stop listening for a notification /// /// See Postgres - UNLISTEN { channel: Ident }, + UNLISTEN { + channel: Ident, + }, /// ```sql /// NOTIFY channel [ , payload ] /// ``` @@ -3580,10 +3711,6 @@ pub enum Statement { /// Snowflake `REMOVE` /// See: Remove(FileStagingCommand), - /// MS-SQL session - /// - /// See - SetSessionParam(SetSessionParamKind), /// RaiseError (MSSQL) /// RAISERROR ( { msg_id | msg_str | @local_variable } /// { , severity , state } @@ -4644,59 +4771,7 @@ impl fmt::Display for Statement { write!(f, "DISCARD {object_type}")?; Ok(()) } - Self::SetRole { - context_modifier, - role_name, - } => { - let role_name = role_name.clone().unwrap_or_else(|| Ident::new("NONE")); - write!(f, "SET{context_modifier} ROLE {role_name}") - } - Statement::SetVariable { - local, - variables, - hivevar, - value, - } => { - f.write_str("SET ")?; - if *local { - f.write_str("LOCAL ")?; - } - let parenthesized = matches!(variables, OneOrManyWithParens::Many(_)); - write!( - f, - "{hivevar}{name} = {l_paren}{value}{r_paren}", - hivevar = if *hivevar { "HIVEVAR:" } else { "" }, - name = variables, - l_paren = parenthesized.then_some("(").unwrap_or_default(), - value = display_comma_separated(value), - r_paren = parenthesized.then_some(")").unwrap_or_default(), - ) - } - Statement::SetTimeZone { local, value } => { - f.write_str("SET ")?; - if *local { - f.write_str("LOCAL ")?; - } - write!(f, "TIME ZONE {value}") - } - Statement::SetNames { - charset_name, - collation_name, - } => { - write!(f, "SET NAMES {}", charset_name)?; - - if let Some(collation) = collation_name { - f.write_str(" COLLATE ")?; - f.write_str(collation)?; - }; - - Ok(()) - } - Statement::SetNamesDefault {} => { - f.write_str("SET NAMES DEFAULT")?; - - Ok(()) - } + Self::Set(set) => write!(f, "{set}"), Statement::ShowVariable { variable } => { write!(f, "SHOW")?; if !variable.is_empty() { @@ -4885,24 +4960,6 @@ impl fmt::Display for Statement { } Ok(()) } - Statement::SetTransaction { - modes, - snapshot, - session, - } => { - if *session { - write!(f, "SET SESSION CHARACTERISTICS AS TRANSACTION")?; - } else { - write!(f, "SET TRANSACTION")?; - } - if !modes.is_empty() { - write!(f, " {}", display_comma_separated(modes))?; - } - if let Some(snapshot_id) = snapshot { - write!(f, " SNAPSHOT {snapshot_id}")?; - } - Ok(()) - } Statement::Commit { chain, end: end_syntax, @@ -5333,7 +5390,6 @@ impl fmt::Display for Statement { Statement::List(command) => write!(f, "LIST {command}"), Statement::Remove(command) => write!(f, "REMOVE {command}"), - Statement::SetSessionParam(kind) => write!(f, "SET {kind}"), } } } @@ -5397,6 +5453,21 @@ impl fmt::Display for SequenceOptions { } } +/// Assignment for a `SET` statement (name [=|TO] value) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct SetAssignment { + pub name: ObjectName, + pub value: Expr, +} + +impl fmt::Display for SetAssignment { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{} = {}", self.name, self.value) + } +} + /// Target of a `TRUNCATE TABLE` command /// /// Note this is its own struct because `visit_relation` requires an `ObjectName` (not a `Vec`) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 8c3eff3c3..fb0fc3f33 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -230,11 +230,7 @@ impl Spanned for Values { /// - [Statement::Fetch] /// - [Statement::Flush] /// - [Statement::Discard] -/// - [Statement::SetRole] -/// - [Statement::SetVariable] -/// - [Statement::SetTimeZone] -/// - [Statement::SetNames] -/// - [Statement::SetNamesDefault] +/// - [Statement::Set] /// - [Statement::ShowFunctions] /// - [Statement::ShowVariable] /// - [Statement::ShowStatus] @@ -244,7 +240,6 @@ impl Spanned for Values { /// - [Statement::ShowTables] /// - [Statement::ShowCollation] /// - [Statement::StartTransaction] -/// - [Statement::SetTransaction] /// - [Statement::Comment] /// - [Statement::Commit] /// - [Statement::Rollback] @@ -445,11 +440,7 @@ impl Spanned for Statement { Statement::Fetch { .. } => Span::empty(), Statement::Flush { .. } => Span::empty(), Statement::Discard { .. } => Span::empty(), - Statement::SetRole { .. } => Span::empty(), - Statement::SetVariable { .. } => Span::empty(), - Statement::SetTimeZone { .. } => Span::empty(), - Statement::SetNames { .. } => Span::empty(), - Statement::SetNamesDefault {} => Span::empty(), + Statement::Set(_) => Span::empty(), Statement::ShowFunctions { .. } => Span::empty(), Statement::ShowVariable { .. } => Span::empty(), Statement::ShowStatus { .. } => Span::empty(), @@ -460,7 +451,6 @@ impl Spanned for Statement { Statement::ShowCollation { .. } => Span::empty(), Statement::Use(u) => u.span(), Statement::StartTransaction { .. } => Span::empty(), - Statement::SetTransaction { .. } => Span::empty(), Statement::Comment { .. } => Span::empty(), Statement::Commit { .. } => Span::empty(), Statement::Rollback { .. } => Span::empty(), @@ -509,7 +499,6 @@ impl Spanned for Statement { Statement::RenameTable { .. } => Span::empty(), Statement::RaisError { .. } => Span::empty(), Statement::List(..) | Statement::Remove(..) => Span::empty(), - Statement::SetSessionParam { .. } => Span::empty(), } } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index aeb097cfd..8d4557e2f 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -399,6 +399,16 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports multiple `SET` statements + /// in a single statement. + /// + /// ```sql + /// SET variable = expression [, variable = expression]; + /// ``` + fn supports_comma_separated_set_assignments(&self) -> bool { + false + } + /// Returns true if the dialect supports an `EXCEPT` clause following a /// wildcard in a select list. /// diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index aeed1eb79..3db34748e 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -82,6 +82,7 @@ impl Dialect for MsSqlDialect { fn supports_start_transaction_modifier(&self) -> bool { true } + fn supports_end_transaction_modifier(&self) -> bool { true } diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 0bdfc9bf3..2077ea195 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -141,6 +141,10 @@ impl Dialect for MySqlDialect { fn supports_set_names(&self) -> bool { true } + + fn supports_comma_separated_set_assignments(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/keywords.rs b/src/keywords.rs index bda817df9..195bbb172 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -173,6 +173,7 @@ define_keywords!( CHANNEL, CHAR, CHARACTER, + CHARACTERISTICS, CHARACTERS, CHARACTER_LENGTH, CHARSET, @@ -557,6 +558,7 @@ define_keywords!( MULTISET, MUTATION, NAME, + NAMES, NANOSECOND, NANOSECONDS, NATIONAL, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 400a9480f..32a7bccd2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4314,7 +4314,8 @@ impl<'a> Parser<'a> { } /// Run a parser method `f`, reverting back to the current position if unsuccessful. - /// Returns `None` if `f` returns an error + /// Returns `ParserError::RecursionLimitExceeded` if `f` returns a `RecursionLimitExceeded`. + /// Returns `Ok(None)` if `f` returns any other error. pub fn maybe_parse(&mut self, f: F) -> Result, ParserError> where F: FnMut(&mut Parser) -> Result, @@ -10978,47 +10979,108 @@ impl<'a> Parser<'a> { } else { Some(self.parse_identifier()?) }; - Ok(Statement::SetRole { + Ok(Statement::Set(Set::SetRole { context_modifier, role_name, - }) + })) } - pub fn parse_set(&mut self) -> Result { - let modifier = - self.parse_one_of_keywords(&[Keyword::SESSION, Keyword::LOCAL, Keyword::HIVEVAR]); - if let Some(Keyword::HIVEVAR) = modifier { - self.expect_token(&Token::Colon)?; - } else if let Some(set_role_stmt) = - self.maybe_parse(|parser| parser.parse_set_role(modifier))? - { - return Ok(set_role_stmt); + fn parse_set_values( + &mut self, + parenthesized_assignment: bool, + ) -> Result, ParserError> { + let mut values = vec![]; + + if parenthesized_assignment { + self.expect_token(&Token::LParen)?; + } + + loop { + let value = if let Some(expr) = self.try_parse_expr_sub_query()? { + expr + } else if let Ok(expr) = self.parse_expr() { + expr + } else { + self.expected("variable value", self.peek_token())? + }; + + values.push(value); + if self.consume_token(&Token::Comma) { + continue; + } + + if parenthesized_assignment { + self.expect_token(&Token::RParen)?; + } + return Ok(values); } + } - let variables = if self.parse_keywords(&[Keyword::TIME, Keyword::ZONE]) { - OneOrManyWithParens::One(ObjectName::from(vec!["TIMEZONE".into()])) - } else if self.dialect.supports_parenthesized_set_variables() + fn parse_set_assignment( + &mut self, + ) -> Result<(OneOrManyWithParens, Expr), ParserError> { + let variables = if self.dialect.supports_parenthesized_set_variables() && self.consume_token(&Token::LParen) { - let variables = OneOrManyWithParens::Many( + let vars = OneOrManyWithParens::Many( self.parse_comma_separated(|parser: &mut Parser<'a>| parser.parse_identifier())? .into_iter() .map(|ident| ObjectName::from(vec![ident])) .collect(), ); self.expect_token(&Token::RParen)?; - variables + vars } else { OneOrManyWithParens::One(self.parse_object_name(false)?) }; - let names = matches!(&variables, OneOrManyWithParens::One(variable) if variable.to_string().eq_ignore_ascii_case("NAMES")); + if !(self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO)) { + return self.expected("assignment operator", self.peek_token()); + } + + let values = self.parse_expr()?; + + Ok((variables, values)) + } + + fn parse_set(&mut self) -> Result { + let modifier = + self.parse_one_of_keywords(&[Keyword::SESSION, Keyword::LOCAL, Keyword::HIVEVAR]); - if names && self.dialect.supports_set_names() { + if let Some(Keyword::HIVEVAR) = modifier { + self.expect_token(&Token::Colon)?; + } + + if let Some(set_role_stmt) = self.maybe_parse(|parser| parser.parse_set_role(modifier))? { + return Ok(set_role_stmt); + } + + // Handle special cases first + if self.parse_keywords(&[Keyword::TIME, Keyword::ZONE]) + || self.parse_keyword(Keyword::TIMEZONE) + { + if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { + return Ok(Set::SingleAssignment { + local: modifier == Some(Keyword::LOCAL), + hivevar: modifier == Some(Keyword::HIVEVAR), + variable: ObjectName::from(vec!["TIMEZONE".into()]), + values: self.parse_set_values(false)?, + } + .into()); + } else { + // A shorthand alias for SET TIME ZONE that doesn't require + // the assignment operator. It's originally PostgreSQL specific, + // but we allow it for all the dialects + return Ok(Set::SetTimeZone { + local: modifier == Some(Keyword::LOCAL), + value: self.parse_expr()?, + } + .into()); + } + } else if self.dialect.supports_set_names() && self.parse_keyword(Keyword::NAMES) { if self.parse_keyword(Keyword::DEFAULT) { - return Ok(Statement::SetNamesDefault {}); + return Ok(Set::SetNamesDefault {}.into()); } - let charset_name = self.parse_identifier()?; let collation_name = if self.parse_one_of_keywords(&[Keyword::COLLATE]).is_some() { Some(self.parse_literal_string()?) @@ -11026,86 +11088,117 @@ impl<'a> Parser<'a> { None }; - return Ok(Statement::SetNames { + return Ok(Set::SetNames { charset_name, collation_name, - }); - } - - let parenthesized_assignment = matches!(&variables, OneOrManyWithParens::Many(_)); - - if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { - if parenthesized_assignment { - self.expect_token(&Token::LParen)?; - } - - let mut values = vec![]; - loop { - let value = if let Some(expr) = self.try_parse_expr_sub_query()? { - expr - } else if let Ok(expr) = self.parse_expr() { - expr - } else { - self.expected("variable value", self.peek_token())? - }; - - values.push(value); - if self.consume_token(&Token::Comma) { - continue; - } - - if parenthesized_assignment { - self.expect_token(&Token::RParen)?; - } - return Ok(Statement::SetVariable { - local: modifier == Some(Keyword::LOCAL), - hivevar: Some(Keyword::HIVEVAR) == modifier, - variables, - value: values, - }); } - } - - let OneOrManyWithParens::One(variable) = variables else { - return self.expected("set variable", self.peek_token()); - }; - - if variable.to_string().eq_ignore_ascii_case("TIMEZONE") { - // for some db (e.g. postgresql), SET TIME ZONE is an alias for SET TIMEZONE [TO|=] - match self.parse_expr() { - Ok(expr) => Ok(Statement::SetTimeZone { - local: modifier == Some(Keyword::LOCAL), - value: expr, - }), - _ => self.expected("timezone value", self.peek_token())?, - } - } else if variable.to_string() == "CHARACTERISTICS" { + .into()); + } else if self.parse_keyword(Keyword::CHARACTERISTICS) { self.expect_keywords(&[Keyword::AS, Keyword::TRANSACTION])?; - Ok(Statement::SetTransaction { + return Ok(Set::SetTransaction { modes: self.parse_transaction_modes()?, snapshot: None, session: true, - }) - } else if variable.to_string() == "TRANSACTION" && modifier.is_none() { + } + .into()); + } else if self.parse_keyword(Keyword::TRANSACTION) { if self.parse_keyword(Keyword::SNAPSHOT) { let snapshot_id = self.parse_value()?.value; - return Ok(Statement::SetTransaction { + return Ok(Set::SetTransaction { modes: vec![], snapshot: Some(snapshot_id), session: false, - }); + } + .into()); } - Ok(Statement::SetTransaction { + return Ok(Set::SetTransaction { modes: self.parse_transaction_modes()?, snapshot: None, session: false, - }) - } else if self.dialect.supports_set_stmt_without_operator() { - self.prev_token(); - self.parse_set_session_params() + } + .into()); + } + + if self.dialect.supports_comma_separated_set_assignments() { + if let Some(assignments) = self + .maybe_parse(|parser| parser.parse_comma_separated(Parser::parse_set_assignment))? + { + return if assignments.len() > 1 { + let assignments = assignments + .into_iter() + .map(|(var, val)| match var { + OneOrManyWithParens::One(v) => Ok(SetAssignment { + name: v, + value: val, + }), + OneOrManyWithParens::Many(_) => { + self.expected("List of single identifiers", self.peek_token()) + } + }) + .collect::>()?; + + Ok(Set::MultipleAssignments { assignments }.into()) + } else { + let (vars, values): (Vec<_>, Vec<_>) = assignments.into_iter().unzip(); + + let variable = match vars.into_iter().next() { + Some(OneOrManyWithParens::One(v)) => Ok(v), + Some(OneOrManyWithParens::Many(_)) => self.expected( + "Single assignment or list of assignments", + self.peek_token(), + ), + None => self.expected("At least one identifier", self.peek_token()), + }?; + + Ok(Set::SingleAssignment { + local: modifier == Some(Keyword::LOCAL), + hivevar: modifier == Some(Keyword::HIVEVAR), + variable, + values, + } + .into()) + }; + } + } + + let variables = if self.dialect.supports_parenthesized_set_variables() + && self.consume_token(&Token::LParen) + { + let vars = OneOrManyWithParens::Many( + self.parse_comma_separated(|parser: &mut Parser<'a>| parser.parse_identifier())? + .into_iter() + .map(|ident| ObjectName::from(vec![ident])) + .collect(), + ); + self.expect_token(&Token::RParen)?; + vars } else { - self.expected("equals sign or TO", self.peek_token()) + OneOrManyWithParens::One(self.parse_object_name(false)?) + }; + + if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { + let stmt = match variables { + OneOrManyWithParens::One(var) => Set::SingleAssignment { + local: modifier == Some(Keyword::LOCAL), + hivevar: modifier == Some(Keyword::HIVEVAR), + variable: var, + values: self.parse_set_values(false)?, + }, + OneOrManyWithParens::Many(vars) => Set::ParenthesizedAssignments { + variables: vars, + values: self.parse_set_values(true)?, + }, + }; + + return Ok(stmt.into()); } + + if self.dialect.supports_set_stmt_without_operator() { + self.prev_token(); + return self.parse_set_session_params(); + }; + + self.expected("equals sign or TO", self.peek_token()) } pub fn parse_set_session_params(&mut self) -> Result { @@ -11123,15 +11216,20 @@ impl<'a> Parser<'a> { _ => return self.expected("IO, PROFILE, TIME or XML", self.peek_token()), }; let value = self.parse_session_param_value()?; - Ok(Statement::SetSessionParam(SetSessionParamKind::Statistics( - SetSessionParamStatistics { topic, value }, - ))) + Ok( + Set::SetSessionParam(SetSessionParamKind::Statistics(SetSessionParamStatistics { + topic, + value, + })) + .into(), + ) } else if self.parse_keyword(Keyword::IDENTITY_INSERT) { let obj = self.parse_object_name(false)?; let value = self.parse_session_param_value()?; - Ok(Statement::SetSessionParam( - SetSessionParamKind::IdentityInsert(SetSessionParamIdentityInsert { obj, value }), + Ok(Set::SetSessionParam(SetSessionParamKind::IdentityInsert( + SetSessionParamIdentityInsert { obj, value }, )) + .into()) } else if self.parse_keyword(Keyword::OFFSETS) { let keywords = self.parse_comma_separated(|parser| { let next_token = parser.next_token(); @@ -11141,9 +11239,13 @@ impl<'a> Parser<'a> { } })?; let value = self.parse_session_param_value()?; - Ok(Statement::SetSessionParam(SetSessionParamKind::Offsets( - SetSessionParamOffsets { keywords, value }, - ))) + Ok( + Set::SetSessionParam(SetSessionParamKind::Offsets(SetSessionParamOffsets { + keywords, + value, + })) + .into(), + ) } else { let names = self.parse_comma_separated(|parser| { let next_token = parser.next_token(); @@ -11153,9 +11255,13 @@ impl<'a> Parser<'a> { } })?; let value = self.parse_expr()?.to_string(); - Ok(Statement::SetSessionParam(SetSessionParamKind::Generic( - SetSessionParamGeneric { names, value }, - ))) + Ok( + Set::SetSessionParam(SetSessionParamKind::Generic(SetSessionParamGeneric { + names, + value, + })) + .into(), + ) } } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 8225d3672..c7bf287cd 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8555,11 +8555,11 @@ fn parse_set_transaction() { // TRANSACTION, so no need to duplicate the tests here. We just do a quick // sanity check. match verified_stmt("SET TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE") { - Statement::SetTransaction { + Statement::Set(Set::SetTransaction { modes, session, snapshot, - } => { + }) => { assert_eq!( modes, vec![ @@ -8578,20 +8578,17 @@ fn parse_set_transaction() { #[test] fn parse_set_variable() { match verified_stmt("SET SOMETHING = '1'") { - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local, hivevar, - variables, - value, - } => { + variable, + values, + }) => { assert!(!local); assert!(!hivevar); + assert_eq!(variable, ObjectName::from(vec!["SOMETHING".into()])); assert_eq!( - variables, - OneOrManyWithParens::One(ObjectName::from(vec!["SOMETHING".into()])) - ); - assert_eq!( - value, + values, vec![Expr::Value( (Value::SingleQuotedString("1".into())).with_empty_span() )] @@ -8603,24 +8600,17 @@ fn parse_set_variable() { let multi_variable_dialects = all_dialects_where(|d| d.supports_parenthesized_set_variables()); let sql = r#"SET (a, b, c) = (1, 2, 3)"#; match multi_variable_dialects.verified_stmt(sql) { - Statement::SetVariable { - local, - hivevar, - variables, - value, - } => { - assert!(!local); - assert!(!hivevar); + Statement::Set(Set::ParenthesizedAssignments { variables, values }) => { assert_eq!( variables, - OneOrManyWithParens::Many(vec![ + vec![ ObjectName::from(vec!["a".into()]), ObjectName::from(vec!["b".into()]), ObjectName::from(vec!["c".into()]), - ]) + ] ); assert_eq!( - value, + values, vec![ Expr::value(number("1")), Expr::value(number("2")), @@ -8680,20 +8670,17 @@ fn parse_set_variable() { #[test] fn parse_set_role_as_variable() { match verified_stmt("SET role = 'foobar'") { - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local, hivevar, - variables, - value, - } => { + variable, + values, + }) => { assert!(!local); assert!(!hivevar); + assert_eq!(variable, ObjectName::from(vec!["role".into()])); assert_eq!( - variables, - OneOrManyWithParens::One(ObjectName::from(vec!["role".into()])) - ); - assert_eq!( - value, + values, vec![Expr::Value( (Value::SingleQuotedString("foobar".into())).with_empty_span() )] @@ -8730,20 +8717,17 @@ fn parse_double_colon_cast_at_timezone() { #[test] fn parse_set_time_zone() { match verified_stmt("SET TIMEZONE = 'UTC'") { - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local, hivevar, - variables: variable, - value, - } => { + variable, + values, + }) => { assert!(!local); assert!(!hivevar); + assert_eq!(variable, ObjectName::from(vec!["TIMEZONE".into()])); assert_eq!( - variable, - OneOrManyWithParens::One(ObjectName::from(vec!["TIMEZONE".into()])) - ); - assert_eq!( - value, + values, vec![Expr::Value( (Value::SingleQuotedString("UTC".into())).with_empty_span() )] @@ -8755,20 +8739,6 @@ fn parse_set_time_zone() { one_statement_parses_to("SET TIME ZONE TO 'UTC'", "SET TIMEZONE = 'UTC'"); } -#[test] -fn parse_set_time_zone_alias() { - match verified_stmt("SET TIME ZONE 'UTC'") { - Statement::SetTimeZone { local, value } => { - assert!(!local); - assert_eq!( - value, - Expr::Value((Value::SingleQuotedString("UTC".into())).with_empty_span()) - ); - } - _ => unreachable!(), - } -} - #[test] fn parse_commit() { match verified_stmt("COMMIT") { @@ -14681,3 +14651,44 @@ fn parse_set_names() { dialects.verified_stmt("SET NAMES 'utf8'"); dialects.verified_stmt("SET NAMES UTF8 COLLATE bogus"); } + +#[test] +fn parse_multiple_set_statements() -> Result<(), ParserError> { + let dialects = all_dialects_where(|d| d.supports_comma_separated_set_assignments()); + let stmt = dialects.verified_stmt("SET @a = 1, b = 2"); + + match stmt { + Statement::Set(Set::MultipleAssignments { assignments }) => { + assert_eq!( + assignments, + vec![ + SetAssignment { + name: ObjectName::from(vec!["@a".into()]), + value: Expr::value(number("1")) + }, + SetAssignment { + name: ObjectName::from(vec!["b".into()]), + value: Expr::value(number("2")) + } + ] + ); + } + _ => panic!("Expected SetVariable with 2 variables and 2 values"), + }; + + Ok(()) +} + +#[test] +fn parse_set_time_zone_alias() { + match all_dialects().verified_stmt("SET TIME ZONE 'UTC'") { + Statement::Set(Set::SetTimeZone { local, value }) => { + assert!(!local); + assert_eq!( + value, + Expr::Value((Value::SingleQuotedString("UTC".into())).with_empty_span()) + ); + } + _ => unreachable!(), + } +} diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index d7f3c014b..56fe22a0d 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -22,9 +22,8 @@ use sqlparser::ast::{ ClusteredBy, CommentDef, CreateFunction, CreateFunctionBody, CreateFunctionUsing, CreateTable, - Expr, Function, FunctionArgumentList, FunctionArguments, Ident, ObjectName, - OneOrManyWithParens, OrderByExpr, OrderByOptions, SelectItem, Statement, TableFactor, - UnaryOperator, Use, Value, + Expr, Function, FunctionArgumentList, FunctionArguments, Ident, ObjectName, OrderByExpr, + OrderByOptions, SelectItem, Set, Statement, TableFactor, UnaryOperator, Use, Value, }; use sqlparser::dialect::{GenericDialect, HiveDialect, MsSqlDialect}; use sqlparser::parser::ParserError; @@ -92,7 +91,7 @@ fn parse_msck() { } #[test] -fn parse_set() { +fn parse_set_hivevar() { let set = "SET HIVEVAR:name = a, b, c_d"; hive().verified_stmt(set); } @@ -369,20 +368,20 @@ fn from_cte() { fn set_statement_with_minus() { assert_eq!( hive().verified_stmt("SET hive.tez.java.opts = -Xmx4g"), - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![ + variable: ObjectName::from(vec![ Ident::new("hive"), Ident::new("tez"), Ident::new("java"), Ident::new("opts") - ])), - value: vec![Expr::UnaryOp { + ]), + values: vec![Expr::UnaryOp { op: UnaryOperator::Minus, expr: Box::new(Expr::Identifier(Ident::new("Xmx4g"))) }], - } + }) ); assert_eq!( diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 3f313af4f..386bd1788 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1254,14 +1254,14 @@ fn parse_mssql_declare() { for_query: None }] }, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("@bar")])), - value: vec![Expr::Value( + variable: ObjectName::from(vec![Ident::new("@bar")]), + values: vec![Expr::Value( (Value::Number("2".parse().unwrap(), false)).with_empty_span() )], - }, + }), Statement::Query(Box::new(Query { with: None, limit: None, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 560ea9dae..13a8a6cc1 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -617,12 +617,12 @@ fn parse_set_variables() { mysql_and_generic().verified_stmt("SET sql_mode = CONCAT(@@sql_mode, ',STRICT_TRANS_TABLES')"); assert_eq!( mysql_and_generic().verified_stmt("SET LOCAL autocommit = 1"), - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: true, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec!["autocommit".into()])), - value: vec![Expr::value(number("1"))], - } + variable: ObjectName::from(vec!["autocommit".into()]), + values: vec![Expr::value(number("1"))], + }) ); } @@ -2705,19 +2705,19 @@ fn parse_set_names() { let stmt = mysql_and_generic().verified_stmt("SET NAMES utf8mb4"); assert_eq!( stmt, - Statement::SetNames { + Statement::Set(Set::SetNames { charset_name: "utf8mb4".into(), collation_name: None, - } + }) ); let stmt = mysql_and_generic().verified_stmt("SET NAMES utf8mb4 COLLATE bogus"); assert_eq!( stmt, - Statement::SetNames { + Statement::Set(Set::SetNames { charset_name: "utf8mb4".into(), collation_name: Some("bogus".to_string()), - } + }) ); let stmt = mysql_and_generic() @@ -2725,14 +2725,14 @@ fn parse_set_names() { .unwrap(); assert_eq!( stmt, - vec![Statement::SetNames { + vec![Statement::Set(Set::SetNames { charset_name: "utf8mb4".into(), collation_name: Some("bogus".to_string()), - }] + })] ); let stmt = mysql_and_generic().verified_stmt("SET NAMES DEFAULT"); - assert_eq!(stmt, Statement::SetNamesDefault {}); + assert_eq!(stmt, Statement::Set(Set::SetNamesDefault {})); } #[test] diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 0dfcc24ea..a65c4fa38 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1432,81 +1432,77 @@ fn parse_set() { let stmt = pg_and_generic().verified_stmt("SET a = b"); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::Identifier(Ident { + variable: ObjectName::from(vec![Ident::new("a")]), + values: vec![Expr::Identifier(Ident { value: "b".into(), quote_style: None, span: Span::empty(), })], - } + }) ); let stmt = pg_and_generic().verified_stmt("SET a = 'b'"); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::Value( + variable: ObjectName::from(vec![Ident::new("a")]), + values: vec![Expr::Value( (Value::SingleQuotedString("b".into())).with_empty_span() )], - } + }) ); let stmt = pg_and_generic().verified_stmt("SET a = 0"); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::value(number("0"))], - } + variable: ObjectName::from(vec![Ident::new("a")]), + values: vec![Expr::value(number("0"))], + }) ); let stmt = pg_and_generic().verified_stmt("SET a = DEFAULT"); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::Identifier(Ident::new("DEFAULT"))], - } + variable: ObjectName::from(vec![Ident::new("a")]), + values: vec![Expr::Identifier(Ident::new("DEFAULT"))], + }) ); let stmt = pg_and_generic().verified_stmt("SET LOCAL a = b"); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: true, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![Ident::new("a")])), - value: vec![Expr::Identifier("b".into())], - } + variable: ObjectName::from(vec![Ident::new("a")]), + values: vec![Expr::Identifier("b".into())], + }) ); let stmt = pg_and_generic().verified_stmt("SET a.b.c = b"); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![ - Ident::new("a"), - Ident::new("b"), - Ident::new("c") - ])), - value: vec![Expr::Identifier(Ident { + variable: ObjectName::from(vec![Ident::new("a"), Ident::new("b"), Ident::new("c")]), + values: vec![Expr::Identifier(Ident { value: "b".into(), quote_style: None, span: Span::empty(), })], - } + }) ); let stmt = pg_and_generic().one_statement_parses_to( @@ -1515,18 +1511,18 @@ fn parse_set() { ); assert_eq!( stmt, - Statement::SetVariable { + Statement::Set(Set::SingleAssignment { local: false, hivevar: false, - variables: OneOrManyWithParens::One(ObjectName::from(vec![ + variable: ObjectName::from(vec![ Ident::new("hive"), Ident::new("tez"), Ident::new("auto"), Ident::new("reducer"), Ident::new("parallelism") - ])), - value: vec![Expr::Value((Value::Boolean(false)).with_empty_span())], - } + ]), + values: vec![Expr::Value((Value::Boolean(false)).with_empty_span())], + }) ); pg_and_generic().one_statement_parses_to("SET a TO b", "SET a = b"); @@ -1560,10 +1556,10 @@ fn parse_set_role() { let stmt = pg_and_generic().verified_stmt(query); assert_eq!( stmt, - Statement::SetRole { + Statement::Set(Set::SetRole { context_modifier: ContextModifier::Session, role_name: None, - } + }) ); assert_eq!(query, stmt.to_string()); @@ -1571,14 +1567,14 @@ fn parse_set_role() { let stmt = pg_and_generic().verified_stmt(query); assert_eq!( stmt, - Statement::SetRole { + Statement::Set(Set::SetRole { context_modifier: ContextModifier::Local, role_name: Some(Ident { value: "rolename".to_string(), quote_style: Some('\"'), span: Span::empty(), }), - } + }) ); assert_eq!(query, stmt.to_string()); @@ -1586,14 +1582,14 @@ fn parse_set_role() { let stmt = pg_and_generic().verified_stmt(query); assert_eq!( stmt, - Statement::SetRole { + Statement::Set(Set::SetRole { context_modifier: ContextModifier::None, role_name: Some(Ident { value: "rolename".to_string(), quote_style: Some('\''), span: Span::empty(), }), - } + }) ); assert_eq!(query, stmt.to_string()); } @@ -2982,16 +2978,16 @@ fn test_transaction_statement() { let statement = pg().verified_stmt("SET TRANSACTION SNAPSHOT '000003A1-1'"); assert_eq!( statement, - Statement::SetTransaction { + Statement::Set(Set::SetTransaction { modes: vec![], snapshot: Some(Value::SingleQuotedString(String::from("000003A1-1"))), session: false - } + }) ); let statement = pg().verified_stmt("SET SESSION CHARACTERISTICS AS TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE"); assert_eq!( statement, - Statement::SetTransaction { + Statement::Set(Set::SetTransaction { modes: vec![ TransactionMode::AccessMode(TransactionAccessMode::ReadOnly), TransactionMode::AccessMode(TransactionAccessMode::ReadWrite), @@ -2999,7 +2995,7 @@ fn test_transaction_statement() { ], snapshot: None, session: true - } + }) ); } From fb578bb419d08d2d5b49fb75a61f1ddd6df77ba4 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Wed, 12 Mar 2025 13:24:06 -0700 Subject: [PATCH 072/286] Preserve MySQL-style `LIMIT , ` syntax (#1765) --- src/ast/mod.rs | 23 ++-- src/ast/query.rs | 73 ++++++++--- src/ast/spans.rs | 31 +++-- src/ast/visitor.rs | 4 +- src/parser/mod.rs | 109 ++++++++------- tests/sqlparser_clickhouse.rs | 15 ++- tests/sqlparser_common.rs | 240 +++++++++++++++++++--------------- tests/sqlparser_mssql.rs | 12 +- tests/sqlparser_mysql.rs | 66 +++------- tests/sqlparser_postgres.rs | 20 +-- 10 files changed, 327 insertions(+), 266 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8ab3fc0f1..139850e86 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -66,17 +66,18 @@ pub use self::query::{ FormatClause, GroupByExpr, GroupByWithModifier, IdentWithAlias, IlikeSelectItem, InputFormatClause, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonTableColumn, JsonTableColumnErrorHandling, JsonTableNamedColumn, JsonTableNestedColumn, - LateralView, LockClause, LockType, MatchRecognizePattern, MatchRecognizeSymbol, Measure, - NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, OffsetRows, OpenJsonTableColumn, - OrderBy, OrderByExpr, OrderByKind, OrderByOptions, PivotValueSource, ProjectionSelect, Query, - RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, - Select, SelectFlavor, SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, - SetOperator, SetQuantifier, Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, - TableFactor, TableFunctionArgs, TableIndexHintForClause, TableIndexHintType, TableIndexHints, - TableIndexType, TableSample, TableSampleBucket, TableSampleKind, TableSampleMethod, - TableSampleModifier, TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, - TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, UpdateTableFromKind, - ValueTableMode, Values, WildcardAdditionalOptions, With, WithFill, + LateralView, LimitClause, LockClause, LockType, MatchRecognizePattern, MatchRecognizeSymbol, + Measure, NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, OffsetRows, + OpenJsonTableColumn, OrderBy, OrderByExpr, OrderByKind, OrderByOptions, PivotValueSource, + ProjectionSelect, Query, RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, + ReplaceSelectItem, RowsPerMatch, Select, SelectFlavor, SelectInto, SelectItem, + SelectItemQualifiedWildcardKind, SetExpr, SetOperator, SetQuantifier, Setting, + SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, TableFunctionArgs, + TableIndexHintForClause, TableIndexHintType, TableIndexHints, TableIndexType, TableSample, + TableSampleBucket, TableSampleKind, TableSampleMethod, TableSampleModifier, + TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, TableSampleUnit, TableVersion, + TableWithJoins, Top, TopQuantity, UpdateTableFromKind, ValueTableMode, Values, + WildcardAdditionalOptions, With, WithFill, }; pub use self::trigger::{ diff --git a/src/ast/query.rs b/src/ast/query.rs index 12f72932f..1b30dcf17 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -43,14 +43,8 @@ pub struct Query { pub body: Box, /// ORDER BY pub order_by: Option, - /// `LIMIT { | ALL }` - pub limit: Option, - - /// `LIMIT { } BY { ,,... } }` - pub limit_by: Vec, - - /// `OFFSET [ { ROW | ROWS } ]` - pub offset: Option, + /// `LIMIT ... OFFSET ... | LIMIT , ` + pub limit_clause: Option, /// `FETCH { FIRST | NEXT } [ PERCENT ] { ROW | ROWS } | { ONLY | WITH TIES }` pub fetch: Option, /// `FOR { UPDATE | SHARE } [ OF table_name ] [ SKIP LOCKED | NOWAIT ]` @@ -79,14 +73,9 @@ impl fmt::Display for Query { if let Some(ref order_by) = self.order_by { write!(f, " {order_by}")?; } - if let Some(ref limit) = self.limit { - write!(f, " LIMIT {limit}")?; - } - if let Some(ref offset) = self.offset { - write!(f, " {offset}")?; - } - if !self.limit_by.is_empty() { - write!(f, " BY {}", display_separated(&self.limit_by, ", "))?; + + if let Some(ref limit_clause) = self.limit_clause { + limit_clause.fmt(f)?; } if let Some(ref settings) = self.settings { write!(f, " SETTINGS {}", display_comma_separated(settings))?; @@ -2374,6 +2363,58 @@ impl fmt::Display for OrderByOptions { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum LimitClause { + /// Standard SQL syntax + /// + /// `LIMIT [BY ,,...] [OFFSET ]` + LimitOffset { + /// `LIMIT { | ALL }` + limit: Option, + /// `OFFSET [ { ROW | ROWS } ]` + offset: Option, + /// `BY { ,,... } }` + /// + /// [ClickHouse](https://clickhouse.com/docs/sql-reference/statements/select/limit-by) + limit_by: Vec, + }, + /// [MySQL]-specific syntax; the order of expressions is reversed. + /// + /// `LIMIT , ` + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/select.html + OffsetCommaLimit { offset: Expr, limit: Expr }, +} + +impl fmt::Display for LimitClause { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + LimitClause::LimitOffset { + limit, + limit_by, + offset, + } => { + if let Some(ref limit) = limit { + write!(f, " LIMIT {limit}")?; + } + if let Some(ref offset) = offset { + write!(f, " {offset}")?; + } + if !limit_by.is_empty() { + debug_assert!(limit.is_some()); + write!(f, " BY {}", display_separated(limit_by, ", "))?; + } + Ok(()) + } + LimitClause::OffsetCommaLimit { offset, limit } => { + write!(f, " LIMIT {}, {}", offset, limit) + } + } + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index fb0fc3f33..a4f5eb46c 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -29,8 +29,8 @@ use super::{ Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, FunctionArguments, GroupByExpr, HavingBound, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, - MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, - OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, + LimitClause, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, + Offset, OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, @@ -94,9 +94,7 @@ impl Spanned for Query { with, body, order_by, - limit, - limit_by, - offset, + limit_clause, fetch, locks: _, // todo for_clause: _, // todo, mssql specific @@ -109,14 +107,31 @@ impl Spanned for Query { .map(|i| i.span()) .chain(core::iter::once(body.span())) .chain(order_by.as_ref().map(|i| i.span())) - .chain(limit.as_ref().map(|i| i.span())) - .chain(limit_by.iter().map(|i| i.span())) - .chain(offset.as_ref().map(|i| i.span())) + .chain(limit_clause.as_ref().map(|i| i.span())) .chain(fetch.as_ref().map(|i| i.span())), ) } } +impl Spanned for LimitClause { + fn span(&self) -> Span { + match self { + LimitClause::LimitOffset { + limit, + offset, + limit_by, + } => union_spans( + limit + .iter() + .map(|i| i.span()) + .chain(offset.as_ref().map(|i| i.span())) + .chain(limit_by.iter().map(|i| i.span())), + ), + LimitClause::OffsetCommaLimit { offset, limit } => offset.span().union(&limit.span()), + } + } +} + impl Spanned for Offset { fn span(&self) -> Span { let Offset { diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index a5d355fe4..50985a3e7 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -523,7 +523,7 @@ where /// // Remove all select limits in sub-queries /// visit_expressions_mut(&mut statements, |expr| { /// if let Expr::Subquery(q) = expr { -/// q.limit = None +/// q.limit_clause = None; /// } /// ControlFlow::<()>::Continue(()) /// }); @@ -647,7 +647,7 @@ where /// // Remove all select limits in outer statements (not in sub-queries) /// visit_statements_mut(&mut statements, |stmt| { /// if let Statement::Query(q) = stmt { -/// q.limit = None +/// q.limit_clause = None; /// } /// ControlFlow::<()>::Continue(()) /// }); diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 32a7bccd2..d3c48a6e7 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9491,6 +9491,60 @@ impl<'a> Parser<'a> { } } + fn parse_optional_limit_clause(&mut self) -> Result, ParserError> { + let mut offset = if self.parse_keyword(Keyword::OFFSET) { + Some(self.parse_offset()?) + } else { + None + }; + + let (limit, limit_by) = if self.parse_keyword(Keyword::LIMIT) { + let expr = self.parse_limit()?; + + if self.dialect.supports_limit_comma() + && offset.is_none() + && expr.is_some() // ALL not supported with comma + && self.consume_token(&Token::Comma) + { + let offset = expr.ok_or_else(|| { + ParserError::ParserError( + "Missing offset for LIMIT , ".to_string(), + ) + })?; + return Ok(Some(LimitClause::OffsetCommaLimit { + offset, + limit: self.parse_expr()?, + })); + } + + let limit_by = if dialect_of!(self is ClickHouseDialect | GenericDialect) + && self.parse_keyword(Keyword::BY) + { + Some(self.parse_comma_separated(Parser::parse_expr)?) + } else { + None + }; + + (Some(expr), limit_by) + } else { + (None, None) + }; + + if offset.is_none() && limit.is_some() && self.parse_keyword(Keyword::OFFSET) { + offset = Some(self.parse_offset()?); + } + + if offset.is_some() || (limit.is_some() && limit != Some(None)) || limit_by.is_some() { + Ok(Some(LimitClause::LimitOffset { + limit: limit.unwrap_or_default(), + offset, + limit_by: limit_by.unwrap_or_default(), + })) + } else { + Ok(None) + } + } + /// Parse a table object for insertion /// e.g. `some_database.some_table` or `FUNCTION some_table_func(...)` pub fn parse_table_object(&mut self) -> Result { @@ -10231,10 +10285,8 @@ impl<'a> Parser<'a> { Ok(Query { with, body: self.parse_insert_setexpr_boxed()?, - limit: None, - limit_by: vec![], order_by: None, - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -10246,10 +10298,8 @@ impl<'a> Parser<'a> { Ok(Query { with, body: self.parse_update_setexpr_boxed()?, - limit: None, - limit_by: vec![], order_by: None, - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -10261,10 +10311,8 @@ impl<'a> Parser<'a> { Ok(Query { with, body: self.parse_delete_setexpr_boxed()?, - limit: None, - limit_by: vec![], + limit_clause: None, order_by: None, - offset: None, fetch: None, locks: vec![], for_clause: None, @@ -10277,40 +10325,7 @@ impl<'a> Parser<'a> { let order_by = self.parse_optional_order_by()?; - let mut limit = None; - let mut offset = None; - - for _x in 0..2 { - if limit.is_none() && self.parse_keyword(Keyword::LIMIT) { - limit = self.parse_limit()? - } - - if offset.is_none() && self.parse_keyword(Keyword::OFFSET) { - offset = Some(self.parse_offset()?) - } - - if self.dialect.supports_limit_comma() - && limit.is_some() - && offset.is_none() - && self.consume_token(&Token::Comma) - { - // MySQL style LIMIT x,y => LIMIT y OFFSET x. - // Check for more details. - offset = Some(Offset { - value: limit.unwrap(), - rows: OffsetRows::None, - }); - limit = Some(self.parse_expr()?); - } - } - - let limit_by = if dialect_of!(self is ClickHouseDialect | GenericDialect) - && self.parse_keyword(Keyword::BY) - { - self.parse_comma_separated(Parser::parse_expr)? - } else { - vec![] - }; + let limit_clause = self.parse_optional_limit_clause()?; let settings = self.parse_settings()?; @@ -10347,9 +10362,7 @@ impl<'a> Parser<'a> { with, body, order_by, - limit, - limit_by, - offset, + limit_clause, fetch, locks, for_clause, @@ -11809,9 +11822,7 @@ impl<'a> Parser<'a> { with: None, body: Box::new(values), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 72a64a488..c56f98860 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -944,6 +944,12 @@ fn parse_limit_by() { clickhouse_and_generic().verified_stmt( r#"SELECT * FROM default.last_asset_runs_mv ORDER BY created_at DESC LIMIT 1 BY asset, toStartOfDay(created_at)"#, ); + clickhouse_and_generic().parse_sql_statements( + r#"SELECT * FROM default.last_asset_runs_mv ORDER BY created_at DESC BY asset, toStartOfDay(created_at)"#, + ).expect_err("BY without LIMIT"); + clickhouse_and_generic() + .parse_sql_statements("SELECT * FROM T OFFSET 5 BY foo") + .expect_err("BY with OFFSET but without LIMIT"); } #[test] @@ -1107,7 +1113,14 @@ fn parse_select_order_by_with_fill_interpolate() { }, select.order_by.expect("ORDER BY expected") ); - assert_eq!(Some(Expr::value(number("2"))), select.limit); + assert_eq!( + select.limit_clause, + Some(LimitClause::LimitOffset { + limit: Some(Expr::value(number("2"))), + offset: None, + limit_by: vec![] + }) + ); } #[test] diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index c7bf287cd..b5d42ea6c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -483,9 +483,7 @@ fn parse_update_set_from() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -900,7 +898,12 @@ fn parse_simple_select() { assert!(select.distinct.is_none()); assert_eq!(3, select.projection.len()); let select = verified_query(sql); - assert_eq!(Some(Expr::value(number("5"))), select.limit); + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::value(number("5"))), + offset: None, + limit_by: vec![], + }; + assert_eq!(Some(expected_limit_clause), select.limit_clause); } #[test] @@ -908,14 +911,31 @@ fn parse_limit() { verified_stmt("SELECT * FROM user LIMIT 1"); } +#[test] +fn parse_invalid_limit_by() { + all_dialects() + .parse_sql_statements("SELECT * FROM user BY name") + .expect_err("BY without LIMIT"); +} + #[test] fn parse_limit_is_not_an_alias() { // In dialects supporting LIMIT it shouldn't be parsed as a table alias let ast = verified_query("SELECT id FROM customer LIMIT 1"); - assert_eq!(Some(Expr::value(number("1"))), ast.limit); + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::value(number("1"))), + offset: None, + limit_by: vec![], + }; + assert_eq!(Some(expected_limit_clause), ast.limit_clause); let ast = verified_query("SELECT 1 LIMIT 5"); - assert_eq!(Some(Expr::value(number("5"))), ast.limit); + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::value(number("5"))), + offset: None, + limit_by: vec![], + }; + assert_eq!(Some(expected_limit_clause), ast.limit_clause); } #[test] @@ -2493,7 +2513,12 @@ fn parse_select_order_by_limit() { ]), select.order_by.expect("ORDER BY expected").kind ); - assert_eq!(Some(Expr::value(number("2"))), select.limit); + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::value(number("2"))), + offset: None, + limit_by: vec![], + }; + assert_eq!(Some(expected_limit_clause), select.limit_clause); } #[test] @@ -2654,7 +2679,12 @@ fn parse_select_order_by_nulls_order() { ]), select.order_by.expect("ORDER BY expeccted").kind ); - assert_eq!(Some(Expr::value(number("2"))), select.limit); + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::value(number("2"))), + offset: None, + limit_by: vec![], + }; + assert_eq!(Some(expected_limit_clause), select.limit_clause); } #[test] @@ -2864,6 +2894,14 @@ fn parse_limit_accepts_all() { "SELECT id, fname, lname FROM customer WHERE id = 1 LIMIT ALL", "SELECT id, fname, lname FROM customer WHERE id = 1", ); + one_statement_parses_to( + "SELECT id, fname, lname FROM customer WHERE id = 1 LIMIT ALL OFFSET 1", + "SELECT id, fname, lname FROM customer WHERE id = 1 OFFSET 1", + ); + one_statement_parses_to( + "SELECT id, fname, lname FROM customer WHERE id = 1 OFFSET 1 LIMIT ALL", + "SELECT id, fname, lname FROM customer WHERE id = 1 OFFSET 1", + ); } #[test] @@ -4247,9 +4285,7 @@ fn parse_create_table_as_table() { schema_name: None, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -4274,9 +4310,7 @@ fn parse_create_table_as_table() { schema_name: Some("schema_name".to_string()), }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -6273,9 +6307,7 @@ fn parse_interval_and_or_xor() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -8175,55 +8207,65 @@ fn parse_offset() { let dialects = all_dialects_where(|d| !d.is_column_alias(&Keyword::OFFSET, &mut Parser::new(d))); - let expect = Some(Offset { - value: Expr::value(number("2")), - rows: OffsetRows::Rows, + let expected_limit_clause = &Some(LimitClause::LimitOffset { + limit: None, + offset: Some(Offset { + value: Expr::value(number("2")), + rows: OffsetRows::Rows, + }), + limit_by: vec![], }); let ast = dialects.verified_query("SELECT foo FROM bar OFFSET 2 ROWS"); - assert_eq!(ast.offset, expect); + assert_eq!(&ast.limit_clause, expected_limit_clause); let ast = dialects.verified_query("SELECT foo FROM bar WHERE foo = 4 OFFSET 2 ROWS"); - assert_eq!(ast.offset, expect); + assert_eq!(&ast.limit_clause, expected_limit_clause); let ast = dialects.verified_query("SELECT foo FROM bar ORDER BY baz OFFSET 2 ROWS"); - assert_eq!(ast.offset, expect); + assert_eq!(&ast.limit_clause, expected_limit_clause); let ast = dialects.verified_query("SELECT foo FROM bar WHERE foo = 4 ORDER BY baz OFFSET 2 ROWS"); - assert_eq!(ast.offset, expect); + assert_eq!(&ast.limit_clause, expected_limit_clause); let ast = dialects.verified_query("SELECT foo FROM (SELECT * FROM bar OFFSET 2 ROWS) OFFSET 2 ROWS"); - assert_eq!(ast.offset, expect); + assert_eq!(&ast.limit_clause, expected_limit_clause); match *ast.body { SetExpr::Select(s) => match only(s.from).relation { TableFactor::Derived { subquery, .. } => { - assert_eq!(subquery.offset, expect); + assert_eq!(&subquery.limit_clause, expected_limit_clause); } _ => panic!("Test broke"), }, _ => panic!("Test broke"), } - let ast = dialects.verified_query("SELECT 'foo' OFFSET 0 ROWS"); - assert_eq!( - ast.offset, - Some(Offset { + let expected_limit_clause = LimitClause::LimitOffset { + limit: None, + offset: Some(Offset { value: Expr::value(number("0")), rows: OffsetRows::Rows, - }) - ); - let ast = dialects.verified_query("SELECT 'foo' OFFSET 1 ROW"); - assert_eq!( - ast.offset, - Some(Offset { + }), + limit_by: vec![], + }; + let ast = dialects.verified_query("SELECT 'foo' OFFSET 0 ROWS"); + assert_eq!(ast.limit_clause, Some(expected_limit_clause)); + let expected_limit_clause = LimitClause::LimitOffset { + limit: None, + offset: Some(Offset { value: Expr::value(number("1")), rows: OffsetRows::Row, - }) - ); - let ast = dialects.verified_query("SELECT 'foo' OFFSET 1"); - assert_eq!( - ast.offset, - Some(Offset { - value: Expr::value(number("1")), + }), + limit_by: vec![], + }; + let ast = dialects.verified_query("SELECT 'foo' OFFSET 1 ROW"); + assert_eq!(ast.limit_clause, Some(expected_limit_clause)); + let expected_limit_clause = LimitClause::LimitOffset { + limit: None, + offset: Some(Offset { + value: Expr::value(number("2")), rows: OffsetRows::None, - }) - ); + }), + limit_by: vec![], + }; + let ast = dialects.verified_query("SELECT 'foo' OFFSET 2"); + assert_eq!(ast.limit_clause, Some(expected_limit_clause)); } #[test] @@ -8273,13 +8315,15 @@ fn parse_fetch() { let ast = verified_query( "SELECT foo FROM bar WHERE foo = 4 ORDER BY baz OFFSET 2 ROWS FETCH FIRST 2 ROWS ONLY", ); - assert_eq!( - ast.offset, - Some(Offset { + let expected_limit_clause = Some(LimitClause::LimitOffset { + limit: None, + offset: Some(Offset { value: Expr::value(number("2")), rows: OffsetRows::Rows, - }) - ); + }), + limit_by: vec![], + }); + assert_eq!(ast.limit_clause, expected_limit_clause); assert_eq!(ast.fetch, fetch_first_two_rows_only); let ast = verified_query( "SELECT foo FROM (SELECT * FROM bar FETCH FIRST 2 ROWS ONLY) FETCH FIRST 2 ROWS ONLY", @@ -8295,24 +8339,20 @@ fn parse_fetch() { _ => panic!("Test broke"), } let ast = verified_query("SELECT foo FROM (SELECT * FROM bar OFFSET 2 ROWS FETCH FIRST 2 ROWS ONLY) OFFSET 2 ROWS FETCH FIRST 2 ROWS ONLY"); - assert_eq!( - ast.offset, - Some(Offset { + let expected_limit_clause = &Some(LimitClause::LimitOffset { + limit: None, + offset: Some(Offset { value: Expr::value(number("2")), rows: OffsetRows::Rows, - }) - ); + }), + limit_by: vec![], + }); + assert_eq!(&ast.limit_clause, expected_limit_clause); assert_eq!(ast.fetch, fetch_first_two_rows_only); match *ast.body { SetExpr::Select(s) => match only(s.from).relation { TableFactor::Derived { subquery, .. } => { - assert_eq!( - subquery.offset, - Some(Offset { - value: Expr::value(number("2")), - rows: OffsetRows::Rows, - }) - ); + assert_eq!(&subquery.limit_clause, expected_limit_clause); assert_eq!(subquery.fetch, fetch_first_two_rows_only); } _ => panic!("Test broke"), @@ -9358,9 +9398,7 @@ fn parse_merge() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -9678,21 +9716,18 @@ fn test_placeholder() { }) ); - let sql = "SELECT * FROM student LIMIT $1 OFFSET $2"; - let ast = dialects.verified_query(sql); - assert_eq!( - ast.limit, - Some(Expr::Value( - (Value::Placeholder("$1".into())).with_empty_span() - )) - ); - assert_eq!( - ast.offset, - Some(Offset { + let ast = dialects.verified_query("SELECT * FROM student LIMIT $1 OFFSET $2"); + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::Value( + (Value::Placeholder("$1".into())).with_empty_span(), + )), + offset: Some(Offset { value: Expr::Value((Value::Placeholder("$2".into())).with_empty_span()), rows: OffsetRows::None, }), - ); + limit_by: vec![], + }; + assert_eq!(ast.limit_clause, Some(expected_limit_clause)); let dialects = TestedDialects::new(vec![ Box::new(GenericDialect {}), @@ -9772,40 +9807,34 @@ fn verified_expr(query: &str) -> Expr { #[test] fn parse_offset_and_limit() { let sql = "SELECT foo FROM bar LIMIT 1 OFFSET 2"; - let expect = Some(Offset { - value: Expr::value(number("2")), - rows: OffsetRows::None, + let expected_limit_clause = Some(LimitClause::LimitOffset { + limit: Some(Expr::value(number("1"))), + offset: Some(Offset { + value: Expr::value(number("2")), + rows: OffsetRows::None, + }), + limit_by: vec![], }); let ast = verified_query(sql); - assert_eq!(ast.offset, expect); - assert_eq!(ast.limit, Some(Expr::value(number("1")))); + assert_eq!(ast.limit_clause, expected_limit_clause); // different order is OK one_statement_parses_to("SELECT foo FROM bar OFFSET 2 LIMIT 1", sql); // mysql syntax is ok for some dialects - TestedDialects::new(vec![ - Box::new(GenericDialect {}), - Box::new(MySqlDialect {}), - Box::new(SQLiteDialect {}), - Box::new(ClickHouseDialect {}), - ]) - .one_statement_parses_to("SELECT foo FROM bar LIMIT 2, 1", sql); + all_dialects_where(|d| d.supports_limit_comma()) + .verified_query("SELECT foo FROM bar LIMIT 2, 1"); // expressions are allowed let sql = "SELECT foo FROM bar LIMIT 1 + 2 OFFSET 3 * 4"; let ast = verified_query(sql); - assert_eq!( - ast.limit, - Some(Expr::BinaryOp { + let expected_limit_clause = LimitClause::LimitOffset { + limit: Some(Expr::BinaryOp { left: Box::new(Expr::value(number("1"))), op: BinaryOperator::Plus, right: Box::new(Expr::value(number("2"))), }), - ); - assert_eq!( - ast.offset, - Some(Offset { + offset: Some(Offset { value: Expr::BinaryOp { left: Box::new(Expr::value(number("3"))), op: BinaryOperator::Multiply, @@ -9813,7 +9842,12 @@ fn parse_offset_and_limit() { }, rows: OffsetRows::None, }), - ); + limit_by: vec![], + }; + assert_eq!(ast.limit_clause, Some(expected_limit_clause),); + + // OFFSET without LIMIT + verified_stmt("SELECT foo FROM bar OFFSET 2"); // Can't repeat OFFSET / LIMIT let res = parse_sql_statements("SELECT foo FROM bar OFFSET 2 OFFSET 2"); @@ -11227,9 +11261,7 @@ fn parse_unload() { flavor: SelectFlavor::Standard, }))), with: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -12400,9 +12432,7 @@ fn test_extract_seconds_ok() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -14265,11 +14295,9 @@ fn test_select_from_first() { flavor, }))), order_by: None, - limit: None, - offset: None, + limit_clause: None, fetch: None, locks: vec![], - limit_by: vec![], for_clause: None, settings: None, format_clause: None, diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 386bd1788..af71d2523 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -107,9 +107,7 @@ fn parse_create_procedure() { or_alter: true, body: vec![Statement::Query(Box::new(Query { with: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1161,9 +1159,7 @@ fn parse_substring_in_select() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1264,9 +1260,7 @@ fn parse_mssql_declare() { }), Statement::Query(Box::new(Query { with: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 13a8a6cc1..a56335934 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1107,9 +1107,7 @@ fn parse_escaped_quote_identifiers_with_escape() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1161,9 +1159,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1209,9 +1205,7 @@ fn parse_escaped_backticks_with_escape() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1261,9 +1255,7 @@ fn parse_escaped_backticks_with_no_escape() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1438,9 +1430,7 @@ fn parse_simple_insert() { ] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1488,9 +1478,7 @@ fn parse_ignore_insert() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1538,9 +1526,7 @@ fn parse_priority_insert() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1585,9 +1571,7 @@ fn parse_priority_insert() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1634,9 +1618,7 @@ fn parse_insert_as() { )]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1698,9 +1680,7 @@ fn parse_insert_as() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1749,9 +1729,7 @@ fn parse_replace_insert() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1791,9 +1769,7 @@ fn parse_empty_row_insert() { rows: vec![vec![], vec![]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -1857,9 +1833,7 @@ fn parse_insert_with_on_duplicate_update() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -2596,9 +2570,7 @@ fn parse_substring_in_select() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -2737,10 +2709,8 @@ fn parse_set_names() { #[test] fn parse_limit_my_sql_syntax() { - mysql_and_generic().one_statement_parses_to( - "SELECT id, fname, lname FROM customer LIMIT 5, 10", - "SELECT id, fname, lname FROM customer LIMIT 10 OFFSET 5", - ); + mysql_and_generic().verified_stmt("SELECT id, fname, lname FROM customer LIMIT 10 OFFSET 5"); + mysql_and_generic().verified_stmt("SELECT id, fname, lname FROM customer LIMIT 5, 10"); mysql_and_generic().verified_stmt("SELECT * FROM user LIMIT ? OFFSET ?"); } @@ -2903,9 +2873,7 @@ fn parse_hex_string_introducer() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index a65c4fa38..1a98870f1 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1319,9 +1319,7 @@ fn parse_copy_to() { flavor: SelectFlavor::Standard, }))), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -2955,9 +2953,7 @@ fn parse_array_subquery_expr() { }))), }), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -4747,9 +4743,7 @@ fn test_simple_postgres_insert_with_alias() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -4820,9 +4814,7 @@ fn test_simple_postgres_insert_with_alias() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, @@ -4891,9 +4883,7 @@ fn test_simple_insert_with_quoted_alias() { ]] })), order_by: None, - limit: None, - limit_by: vec![], - offset: None, + limit_clause: None, fetch: None, locks: vec![], for_clause: None, From cf4ab7f9ab031d168e10f8dbaaa3fa07f15acc63 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Thu, 13 Mar 2025 20:51:29 +0100 Subject: [PATCH 073/286] Add support for `DROP MATERIALIZED VIEW` (#1743) --- src/ast/mod.rs | 2 ++ src/parser/mod.rs | 4 +++- tests/sqlparser_common.rs | 3 +++ 3 files changed, 8 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 139850e86..4c0ffea9e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6733,6 +6733,7 @@ impl fmt::Display for HavingBoundKind { pub enum ObjectType { Table, View, + MaterializedView, Index, Schema, Database, @@ -6747,6 +6748,7 @@ impl fmt::Display for ObjectType { f.write_str(match self { ObjectType::Table => "TABLE", ObjectType::View => "VIEW", + ObjectType::MaterializedView => "MATERIALIZED VIEW", ObjectType::Index => "INDEX", ObjectType::Schema => "SCHEMA", ObjectType::Database => "DATABASE", diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d3c48a6e7..60e1c1463 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5806,6 +5806,8 @@ impl<'a> Parser<'a> { ObjectType::Table } else if self.parse_keyword(Keyword::VIEW) { ObjectType::View + } else if self.parse_keywords(&[Keyword::MATERIALIZED, Keyword::VIEW]) { + ObjectType::MaterializedView } else if self.parse_keyword(Keyword::INDEX) { ObjectType::Index } else if self.parse_keyword(Keyword::ROLE) { @@ -5836,7 +5838,7 @@ impl<'a> Parser<'a> { return self.parse_drop_extension(); } else { return self.expected( - "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, or VIEW after DROP", + "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, VIEW, or MATERIALIZED VIEW after DROP", self.peek_token(), ); }; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index b5d42ea6c..d0edafb7c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8189,6 +8189,9 @@ fn parse_drop_view() { } _ => unreachable!(), } + + verified_stmt("DROP MATERIALIZED VIEW a.b.c"); + verified_stmt("DROP MATERIALIZED VIEW IF EXISTS a.b.c"); } #[test] From 862e887a66cf8ede2dc3a641db7cdcf52b061b76 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Fri, 14 Mar 2025 07:49:25 +0100 Subject: [PATCH 074/286] Add `CASE` and `IF` statement support (#1741) --- src/ast/mod.rs | 198 ++++++++++++++++++++++++++++++++++++-- src/ast/spans.rs | 78 ++++++++++++--- src/keywords.rs | 1 + src/parser/mod.rs | 104 ++++++++++++++++++++ tests/sqlparser_common.rs | 114 ++++++++++++++++++++++ 5 files changed, 473 insertions(+), 22 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 4c0ffea9e..66fd4c6fb 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -151,6 +151,15 @@ where DisplaySeparated { slice, sep: ", " } } +/// Writes the given statements to the formatter, each ending with +/// a semicolon and space separated. +fn format_statement_list(f: &mut fmt::Formatter, statements: &[Statement]) -> fmt::Result { + write!(f, "{}", display_separated(statements, "; "))?; + // We manually insert semicolon for the last statement, + // since display_separated doesn't handle that case. + write!(f, ";") +} + /// An identifier, decomposed into its value or character data and the quote style. #[derive(Debug, Clone, PartialOrd, Ord)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -2080,6 +2089,173 @@ pub enum Password { NullPassword, } +/// A `CASE` statement. +/// +/// Examples: +/// ```sql +/// CASE +/// WHEN EXISTS(SELECT 1) +/// THEN SELECT 1 FROM T; +/// WHEN EXISTS(SELECT 2) +/// THEN SELECT 1 FROM U; +/// ELSE +/// SELECT 1 FROM V; +/// END CASE; +/// ``` +/// +/// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/procedural-language#case_search_expression) +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/snowflake-scripting/case) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CaseStatement { + pub match_expr: Option, + pub when_blocks: Vec, + pub else_block: Option>, + /// TRUE if the statement ends with `END CASE` (vs `END`). + pub has_end_case: bool, +} + +impl fmt::Display for CaseStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let CaseStatement { + match_expr, + when_blocks, + else_block, + has_end_case, + } = self; + + write!(f, "CASE")?; + + if let Some(expr) = match_expr { + write!(f, " {expr}")?; + } + + if !when_blocks.is_empty() { + write!(f, " {}", display_separated(when_blocks, " "))?; + } + + if let Some(else_block) = else_block { + write!(f, " ELSE ")?; + format_statement_list(f, else_block)?; + } + + write!(f, " END")?; + if *has_end_case { + write!(f, " CASE")?; + } + + Ok(()) + } +} + +/// An `IF` statement. +/// +/// Examples: +/// ```sql +/// IF TRUE THEN +/// SELECT 1; +/// SELECT 2; +/// ELSEIF TRUE THEN +/// SELECT 3; +/// ELSE +/// SELECT 4; +/// END IF +/// ``` +/// +/// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/procedural-language#if) +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/snowflake-scripting/if) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct IfStatement { + pub if_block: ConditionalStatements, + pub elseif_blocks: Vec, + pub else_block: Option>, +} + +impl fmt::Display for IfStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let IfStatement { + if_block, + elseif_blocks, + else_block, + } = self; + + write!(f, "{if_block}")?; + + if !elseif_blocks.is_empty() { + write!(f, " {}", display_separated(elseif_blocks, " "))?; + } + + if let Some(else_block) = else_block { + write!(f, " ELSE ")?; + format_statement_list(f, else_block)?; + } + + write!(f, " END IF")?; + + Ok(()) + } +} + +/// Represents a type of [ConditionalStatements] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ConditionalStatementKind { + /// `WHEN THEN ` + When, + /// `IF THEN ` + If, + /// `ELSEIF THEN ` + ElseIf, +} + +/// A block within a [Statement::Case] or [Statement::If]-like statement +/// +/// Examples: +/// ```sql +/// WHEN EXISTS(SELECT 1) THEN SELECT 1; +/// +/// IF TRUE THEN SELECT 1; SELECT 2; +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ConditionalStatements { + /// The condition expression. + pub condition: Expr, + /// Statement list of the `THEN` clause. + pub statements: Vec, + pub kind: ConditionalStatementKind, +} + +impl fmt::Display for ConditionalStatements { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let ConditionalStatements { + condition: expr, + statements, + kind, + } = self; + + let kind = match kind { + ConditionalStatementKind::When => "WHEN", + ConditionalStatementKind::If => "IF", + ConditionalStatementKind::ElseIf => "ELSEIF", + }; + + write!(f, "{kind} {expr} THEN")?; + + if !statements.is_empty() { + write!(f, " ")?; + format_statement_list(f, statements)?; + } + + Ok(()) + } +} + /// Represents an expression assignment within a variable `DECLARE` statement. /// /// Examples: @@ -2647,6 +2823,10 @@ pub enum Statement { file_format: Option, source: Box, }, + /// A `CASE` statement. + Case(CaseStatement), + /// An `IF` statement. + If(IfStatement), /// ```sql /// CALL /// ``` @@ -3940,6 +4120,12 @@ impl fmt::Display for Statement { } Ok(()) } + Statement::Case(stmt) => { + write!(f, "{stmt}") + } + Statement::If(stmt) => { + write!(f, "{stmt}") + } Statement::AttachDatabase { schema_name, database_file_name, @@ -4942,18 +5128,14 @@ impl fmt::Display for Statement { write!(f, " {}", display_comma_separated(modes))?; } if !statements.is_empty() { - write!(f, " {}", display_separated(statements, "; "))?; - // We manually insert semicolon for the last statement, - // since display_separated doesn't handle that case. - write!(f, ";")?; + write!(f, " ")?; + format_statement_list(f, statements)?; } if let Some(exception_statements) = exception_statements { write!(f, " EXCEPTION WHEN ERROR THEN")?; if !exception_statements.is_empty() { - write!(f, " {}", display_separated(exception_statements, "; "))?; - // We manually insert semicolon for the last statement, - // since display_separated doesn't handle that case. - write!(f, ";")?; + write!(f, " ")?; + format_statement_list(f, exception_statements)?; } } if *has_end_keyword { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index a4f5eb46c..0ee11f23f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -22,20 +22,21 @@ use crate::tokenizer::Span; use super::{ dcl::SecondaryRoles, value::ValueWithSpan, AccessExpr, AlterColumnOperation, - AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, CloseCursor, - ClusteredIndex, ColumnDef, ColumnOption, ColumnOptionDef, ConflictTarget, ConnectBy, - ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, - Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, - Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, - FunctionArguments, GroupByExpr, HavingBound, IlikeSelectItem, Insert, Interpolate, - InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, - LimitClause, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, - Offset, OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, - PivotValueSource, ProjectionSelect, Query, ReferentialAction, RenameSelectItem, - ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, - Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, - TableFactor, TableObject, TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, - Value, Values, ViewColumnDef, WildcardAdditionalOptions, With, WithFill, + AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, CaseStatement, + CloseCursor, ClusteredIndex, ColumnDef, ColumnOption, ColumnOptionDef, ConditionalStatements, + ConflictTarget, ConnectBy, ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, + CreateTableOptions, Cte, Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, + ExprWithAlias, Fetch, FromTable, Function, FunctionArg, FunctionArgExpr, + FunctionArgumentClause, FunctionArgumentList, FunctionArguments, GroupByExpr, HavingBound, + IfStatement, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, + JoinOperator, JsonPath, JsonPathElem, LateralView, LimitClause, MatchRecognizePattern, Measure, + NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, + OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, + Query, ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, + SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, + TableAliasColumnDef, TableConstraint, TableFactor, TableObject, TableOptionsClustered, + TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, + WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -334,6 +335,8 @@ impl Spanned for Statement { file_format: _, source, } => source.span(), + Statement::Case(stmt) => stmt.span(), + Statement::If(stmt) => stmt.span(), Statement::Call(function) => function.span(), Statement::Copy { source, @@ -732,6 +735,53 @@ impl Spanned for CreateIndex { } } +impl Spanned for CaseStatement { + fn span(&self) -> Span { + let CaseStatement { + match_expr, + when_blocks, + else_block, + has_end_case: _, + } = self; + + union_spans( + match_expr + .iter() + .map(|e| e.span()) + .chain(when_blocks.iter().map(|b| b.span())) + .chain(else_block.iter().flat_map(|e| e.iter().map(|s| s.span()))), + ) + } +} + +impl Spanned for IfStatement { + fn span(&self) -> Span { + let IfStatement { + if_block, + elseif_blocks, + else_block, + } = self; + + union_spans( + iter::once(if_block.span()) + .chain(elseif_blocks.iter().map(|b| b.span())) + .chain(else_block.iter().flat_map(|e| e.iter().map(|s| s.span()))), + ) + } +} + +impl Spanned for ConditionalStatements { + fn span(&self) -> Span { + let ConditionalStatements { + condition, + statements, + kind: _, + } = self; + + union_spans(iter::once(condition.span()).chain(statements.iter().map(|s| s.span()))) + } +} + /// # partial span /// /// Missing spans: diff --git a/src/keywords.rs b/src/keywords.rs index 195bbb172..47da10096 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -297,6 +297,7 @@ define_keywords!( ELEMENT, ELEMENTS, ELSE, + ELSEIF, EMPTY, ENABLE, ENABLE_SCHEMA_EVOLUTION, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 60e1c1463..3adfe55e4 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -528,6 +528,14 @@ impl<'a> Parser<'a> { Keyword::DESCRIBE => self.parse_explain(DescribeAlias::Describe), Keyword::EXPLAIN => self.parse_explain(DescribeAlias::Explain), Keyword::ANALYZE => self.parse_analyze(), + Keyword::CASE => { + self.prev_token(); + self.parse_case_stmt() + } + Keyword::IF => { + self.prev_token(); + self.parse_if_stmt() + } Keyword::SELECT | Keyword::WITH | Keyword::VALUES | Keyword::FROM => { self.prev_token(); self.parse_query().map(Statement::Query) @@ -615,6 +623,102 @@ impl<'a> Parser<'a> { } } + /// Parse a `CASE` statement. + /// + /// See [Statement::Case] + pub fn parse_case_stmt(&mut self) -> Result { + self.expect_keyword_is(Keyword::CASE)?; + + let match_expr = if self.peek_keyword(Keyword::WHEN) { + None + } else { + Some(self.parse_expr()?) + }; + + self.expect_keyword_is(Keyword::WHEN)?; + let when_blocks = self.parse_keyword_separated(Keyword::WHEN, |parser| { + parser.parse_conditional_statements( + ConditionalStatementKind::When, + &[Keyword::WHEN, Keyword::ELSE, Keyword::END], + ) + })?; + + let else_block = if self.parse_keyword(Keyword::ELSE) { + Some(self.parse_statement_list(&[Keyword::END])?) + } else { + None + }; + + self.expect_keyword_is(Keyword::END)?; + let has_end_case = self.parse_keyword(Keyword::CASE); + + Ok(Statement::Case(CaseStatement { + match_expr, + when_blocks, + else_block, + has_end_case, + })) + } + + /// Parse an `IF` statement. + /// + /// See [Statement::If] + pub fn parse_if_stmt(&mut self) -> Result { + self.expect_keyword_is(Keyword::IF)?; + let if_block = self.parse_conditional_statements( + ConditionalStatementKind::If, + &[Keyword::ELSE, Keyword::ELSEIF, Keyword::END], + )?; + + let elseif_blocks = if self.parse_keyword(Keyword::ELSEIF) { + self.parse_keyword_separated(Keyword::ELSEIF, |parser| { + parser.parse_conditional_statements( + ConditionalStatementKind::ElseIf, + &[Keyword::ELSEIF, Keyword::ELSE, Keyword::END], + ) + })? + } else { + vec![] + }; + + let else_block = if self.parse_keyword(Keyword::ELSE) { + Some(self.parse_statement_list(&[Keyword::END])?) + } else { + None + }; + + self.expect_keywords(&[Keyword::END, Keyword::IF])?; + + Ok(Statement::If(IfStatement { + if_block, + elseif_blocks, + else_block, + })) + } + + /// Parses an expression and associated list of statements + /// belonging to a conditional statement like `IF` or `WHEN`. + /// + /// Example: + /// ```sql + /// IF condition THEN statement1; statement2; + /// ``` + fn parse_conditional_statements( + &mut self, + kind: ConditionalStatementKind, + terminal_keywords: &[Keyword], + ) -> Result { + let condition = self.parse_expr()?; + self.expect_keyword_is(Keyword::THEN)?; + let statements = self.parse_statement_list(terminal_keywords)?; + + Ok(ConditionalStatements { + condition, + statements, + kind, + }) + } + pub fn parse_comment(&mut self) -> Result { let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index d0edafb7c..8c9cae831 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14179,6 +14179,120 @@ fn test_visit_order() { ); } +#[test] +fn parse_case_statement() { + let sql = "CASE 1 WHEN 2 THEN SELECT 1; SELECT 2; ELSE SELECT 3; END CASE"; + let Statement::Case(stmt) = verified_stmt(sql) else { + unreachable!() + }; + + assert_eq!(Some(Expr::value(number("1"))), stmt.match_expr); + assert_eq!(Expr::value(number("2")), stmt.when_blocks[0].condition); + assert_eq!(2, stmt.when_blocks[0].statements.len()); + assert_eq!(1, stmt.else_block.unwrap().len()); + + verified_stmt(concat!( + "CASE 1", + " WHEN a THEN", + " SELECT 1; SELECT 2; SELECT 3;", + " WHEN b THEN", + " SELECT 4; SELECT 5;", + " ELSE", + " SELECT 7; SELECT 8;", + " END CASE" + )); + verified_stmt(concat!( + "CASE 1", + " WHEN a THEN", + " SELECT 1; SELECT 2; SELECT 3;", + " WHEN b THEN", + " SELECT 4; SELECT 5;", + " END CASE" + )); + verified_stmt(concat!( + "CASE 1", + " WHEN a THEN", + " SELECT 1; SELECT 2; SELECT 3;", + " END CASE" + )); + verified_stmt(concat!( + "CASE 1", + " WHEN a THEN", + " SELECT 1; SELECT 2; SELECT 3;", + " END" + )); + + assert_eq!( + ParserError::ParserError("Expected: THEN, found: END".to_string()), + parse_sql_statements("CASE 1 WHEN a END").unwrap_err() + ); + assert_eq!( + ParserError::ParserError("Expected: WHEN, found: ELSE".to_string()), + parse_sql_statements("CASE 1 ELSE SELECT 1; END").unwrap_err() + ); +} + +#[test] +fn parse_if_statement() { + let sql = "IF 1 THEN SELECT 1; ELSEIF 2 THEN SELECT 2; ELSE SELECT 3; END IF"; + let Statement::If(stmt) = verified_stmt(sql) else { + unreachable!() + }; + assert_eq!(Expr::value(number("1")), stmt.if_block.condition); + assert_eq!(Expr::value(number("2")), stmt.elseif_blocks[0].condition); + assert_eq!(1, stmt.else_block.unwrap().len()); + + verified_stmt(concat!( + "IF 1 THEN", + " SELECT 1;", + " SELECT 2;", + " SELECT 3;", + " ELSEIF 2 THEN", + " SELECT 4;", + " SELECT 5;", + " ELSEIF 3 THEN", + " SELECT 6;", + " SELECT 7;", + " ELSE", + " SELECT 8;", + " SELECT 9;", + " END IF" + )); + verified_stmt(concat!( + "IF 1 THEN", + " SELECT 1;", + " SELECT 2;", + " ELSE", + " SELECT 3;", + " SELECT 4;", + " END IF" + )); + verified_stmt(concat!( + "IF 1 THEN", + " SELECT 1;", + " SELECT 2;", + " SELECT 3;", + " ELSEIF 2 THEN", + " SELECT 3;", + " SELECT 4;", + " END IF" + )); + verified_stmt(concat!("IF 1 THEN", " SELECT 1;", " SELECT 2;", " END IF")); + verified_stmt(concat!( + "IF (1) THEN", + " SELECT 1;", + " SELECT 2;", + " END IF" + )); + verified_stmt("IF 1 THEN END IF"); + verified_stmt("IF 1 THEN SELECT 1; ELSEIF 1 THEN END IF"); + + assert_eq!( + ParserError::ParserError("Expected: IF, found: EOF".to_string()), + parse_sql_statements("IF 1 THEN SELECT 1; ELSEIF 1 THEN SELECT 2; END").unwrap_err() + ); +} + #[test] fn test_lambdas() { let dialects = all_dialects_where(|d| d.supports_lambda_functions()); From f81aed6359d9da35df434a6b4b0df07b6cffd5c5 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Fri, 14 Mar 2025 08:00:19 +0100 Subject: [PATCH 075/286] BigQuery: Add support for `CREATE SCHEMA` options (#1742) --- src/ast/mod.rs | 42 +++++++++++++++++++++++++++++++------ src/parser/mod.rs | 14 +++++++++++++ tests/sqlparser_common.rs | 5 +++++ tests/sqlparser_postgres.rs | 2 ++ 4 files changed, 57 insertions(+), 6 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 66fd4c6fb..8c4079213 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3450,6 +3450,22 @@ pub enum Statement { /// ` | AUTHORIZATION | AUTHORIZATION ` schema_name: SchemaName, if_not_exists: bool, + /// Schema options. + /// + /// ```sql + /// CREATE SCHEMA myschema OPTIONS(key1='value1'); + /// ``` + /// + /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_schema_statement) + options: Option>, + /// Default collation specification for the schema. + /// + /// ```sql + /// CREATE SCHEMA myschema DEFAULT COLLATE 'und:ci'; + /// ``` + /// + /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_schema_statement) + default_collate_spec: Option, }, /// ```sql /// CREATE DATABASE @@ -5177,12 +5193,26 @@ impl fmt::Display for Statement { Statement::CreateSchema { schema_name, if_not_exists, - } => write!( - f, - "CREATE SCHEMA {if_not_exists}{name}", - if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, - name = schema_name - ), + options, + default_collate_spec, + } => { + write!( + f, + "CREATE SCHEMA {if_not_exists}{name}", + if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, + name = schema_name + )?; + + if let Some(collate) = default_collate_spec { + write!(f, " DEFAULT COLLATE {collate}")?; + } + + if let Some(options) = options { + write!(f, " OPTIONS({})", display_comma_separated(options))?; + } + + Ok(()) + } Statement::Assert { condition, message } => { write!(f, "ASSERT {condition}")?; if let Some(m) = message { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 3adfe55e4..864fd579f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4731,9 +4731,23 @@ impl<'a> Parser<'a> { let schema_name = self.parse_schema_name()?; + let default_collate_spec = if self.parse_keywords(&[Keyword::DEFAULT, Keyword::COLLATE]) { + Some(self.parse_expr()?) + } else { + None + }; + + let options = if self.peek_keyword(Keyword::OPTIONS) { + Some(self.parse_options(Keyword::OPTIONS)?) + } else { + None + }; + Ok(Statement::CreateSchema { schema_name, if_not_exists, + options, + default_collate_spec, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 8c9cae831..c65cc6b53 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4208,6 +4208,11 @@ fn parse_create_schema() { } _ => unreachable!(), } + + verified_stmt(r#"CREATE SCHEMA a.b.c OPTIONS(key1 = 'value1', key2 = 'value2')"#); + verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a OPTIONS(key1 = 'value1')"#); + verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a OPTIONS()"#); + verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a DEFAULT COLLATE 'und:ci' OPTIONS()"#); } #[test] diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 1a98870f1..e62f23597 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -988,6 +988,8 @@ fn parse_create_schema_if_not_exists() { Statement::CreateSchema { if_not_exists: true, schema_name, + options: _, + default_collate_spec: _, } => assert_eq!("schema_name", schema_name.to_string()), _ => unreachable!(), } From 10cf7c164ee0bae8a71e1d8f0af5851b96465692 Mon Sep 17 00:00:00 2001 From: Aleksei Piianin Date: Sat, 15 Mar 2025 07:07:07 +0100 Subject: [PATCH 076/286] Snowflake: Support dollar quoted comments (#1755) --- src/dialect/snowflake.rs | 5 +---- src/parser/mod.rs | 37 ++++++++++++++++++------------------ tests/sqlparser_snowflake.rs | 15 +++++++++++++++ 3 files changed, 34 insertions(+), 23 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 72252b277..09a0e57ce 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -644,10 +644,7 @@ pub fn parse_create_stage( // [ comment ] if parser.parse_keyword(Keyword::COMMENT) { parser.expect_token(&Token::Eq)?; - comment = Some(match parser.next_token().token { - Token::SingleQuotedString(word) => Ok(word), - _ => parser.expected("a comment statement", parser.peek_token()), - }?) + comment = Some(parser.parse_comment_value()?); } Ok(Statement::CreateStage { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 864fd579f..e4c170edd 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5451,11 +5451,7 @@ impl<'a> Parser<'a> { && self.parse_keyword(Keyword::COMMENT) { self.expect_token(&Token::Eq)?; - let next_token = self.next_token(); - match next_token.token { - Token::SingleQuotedString(str) => Some(str), - _ => self.expected("string literal", next_token)?, - } + Some(self.parse_comment_value()?) } else { None }; @@ -7059,21 +7055,28 @@ impl<'a> Parser<'a> { pub fn parse_optional_inline_comment(&mut self) -> Result, ParserError> { let comment = if self.parse_keyword(Keyword::COMMENT) { let has_eq = self.consume_token(&Token::Eq); - let next_token = self.next_token(); - match next_token.token { - Token::SingleQuotedString(str) => Some(if has_eq { - CommentDef::WithEq(str) - } else { - CommentDef::WithoutEq(str) - }), - _ => self.expected("comment", next_token)?, - } + let comment = self.parse_comment_value()?; + Some(if has_eq { + CommentDef::WithEq(comment) + } else { + CommentDef::WithoutEq(comment) + }) } else { None }; Ok(comment) } + pub fn parse_comment_value(&mut self) -> Result { + let next_token = self.next_token(); + let value = match next_token.token { + Token::SingleQuotedString(str) => str, + Token::DollarQuotedString(str) => str.value, + _ => self.expected("string literal", next_token)?, + }; + Ok(value) + } + pub fn parse_optional_procedure_parameters( &mut self, ) -> Result>, ParserError> { @@ -7209,11 +7212,7 @@ impl<'a> Parser<'a> { } else if self.parse_keywords(&[Keyword::NOT, Keyword::NULL]) { Ok(Some(ColumnOption::NotNull)) } else if self.parse_keywords(&[Keyword::COMMENT]) { - let next_token = self.next_token(); - match next_token.token { - Token::SingleQuotedString(value, ..) => Ok(Some(ColumnOption::Comment(value))), - _ => self.expected("string", next_token), - } + Ok(Some(ColumnOption::Comment(self.parse_comment_value()?))) } else if self.parse_keyword(Keyword::NULL) { Ok(Some(ColumnOption::Null)) } else if self.parse_keyword(Keyword::DEFAULT) { diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index b1d31e6dd..f37b657e5 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -976,6 +976,21 @@ fn parse_sf_create_or_replace_with_comment_for_snowflake() { } } +#[test] +fn parse_sf_create_table_or_view_with_dollar_quoted_comment() { + // Snowflake transforms dollar quoted comments into a common comment in DDL representation of creation + snowflake() + .one_statement_parses_to( + r#"CREATE OR REPLACE TEMPORARY VIEW foo.bar.baz ("COL_1" COMMENT $$comment 1$$) COMMENT = $$view comment$$ AS (SELECT 1)"#, + r#"CREATE OR REPLACE TEMPORARY VIEW foo.bar.baz ("COL_1" COMMENT 'comment 1') COMMENT = 'view comment' AS (SELECT 1)"# + ); + + snowflake().one_statement_parses_to( + r#"CREATE TABLE my_table (a STRING COMMENT $$comment 1$$) COMMENT = $$table comment$$"#, + r#"CREATE TABLE my_table (a STRING COMMENT 'comment 1') COMMENT = 'table comment'"#, + ); +} + #[test] fn test_sf_derived_table_in_parenthesis() { // Nesting a subquery in an extra set of parentheses is non-standard, From da5892802f181f1df3609104b238fa2ab0baadaf Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Tue, 18 Mar 2025 08:22:37 +0200 Subject: [PATCH 077/286] Add LOCK operation for ALTER TABLE (#1768) --- src/ast/ddl.rs | 37 ++++++++++++++++++++++++++++++++ src/ast/mod.rs | 2 +- src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 18 ++++++++++++++++ tests/sqlparser_mysql.rs | 46 ++++++++++++++++++++++++++++++++++++++++ 6 files changed, 104 insertions(+), 1 deletion(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 99d8521cd..39e43ef15 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -288,6 +288,16 @@ pub enum AlterTableOperation { equals: bool, algorithm: AlterTableAlgorithm, }, + + /// `LOCK [=] { DEFAULT | NONE | SHARED | EXCLUSIVE }` + /// + /// [MySQL]-specific table alter lock. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html + Lock { + equals: bool, + lock: AlterTableLock, + }, /// `AUTO_INCREMENT [=] ` /// /// [MySQL]-specific table option for raising current auto increment value. @@ -366,6 +376,30 @@ impl fmt::Display for AlterTableAlgorithm { } } +/// [MySQL] `ALTER TABLE` lock. +/// +/// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterTableLock { + Default, + None, + Shared, + Exclusive, +} + +impl fmt::Display for AlterTableLock { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.write_str(match self { + Self::Default => "DEFAULT", + Self::None => "NONE", + Self::Shared => "SHARED", + Self::Exclusive => "EXCLUSIVE", + }) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -692,6 +726,9 @@ impl fmt::Display for AlterTableOperation { value ) } + AlterTableOperation::Lock { equals, lock } => { + write!(f, "LOCK {}{}", if *equals { "= " } else { "" }, lock) + } } } } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8c4079213..69f2e57f1 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -48,7 +48,7 @@ pub use self::dcl::{ }; pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, - AlterTableAlgorithm, AlterTableOperation, AlterType, AlterTypeAddValue, + AlterTableAlgorithm, AlterTableLock, AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateFunction, Deduplicate, DeferrableInitial, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 0ee11f23f..783248a78 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1119,6 +1119,7 @@ impl Spanned for AlterTableOperation { AlterTableOperation::ResumeRecluster => Span::empty(), AlterTableOperation::Algorithm { .. } => Span::empty(), AlterTableOperation::AutoIncrement { value, .. } => value.span(), + AlterTableOperation::Lock { .. } => Span::empty(), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 47da10096..974230f19 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -797,6 +797,7 @@ define_keywords!( SETS, SETTINGS, SHARE, + SHARED, SHARING, SHOW, SIGNED, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e4c170edd..0d8edc05c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8311,6 +8311,24 @@ impl<'a> Parser<'a> { AlterTableOperation::SuspendRecluster } else if self.parse_keywords(&[Keyword::RESUME, Keyword::RECLUSTER]) { AlterTableOperation::ResumeRecluster + } else if self.parse_keyword(Keyword::LOCK) { + let equals = self.consume_token(&Token::Eq); + let lock = match self.parse_one_of_keywords(&[ + Keyword::DEFAULT, + Keyword::EXCLUSIVE, + Keyword::NONE, + Keyword::SHARED, + ]) { + Some(Keyword::DEFAULT) => AlterTableLock::Default, + Some(Keyword::EXCLUSIVE) => AlterTableLock::Exclusive, + Some(Keyword::NONE) => AlterTableLock::None, + Some(Keyword::SHARED) => AlterTableLock::Shared, + _ => self.expected( + "DEFAULT, EXCLUSIVE, NONE or SHARED after LOCK [=]", + self.peek_token(), + )?, + }; + AlterTableOperation::Lock { equals, lock } } else if self.parse_keyword(Keyword::ALGORITHM) { let equals = self.consume_token(&Token::Eq); let algorithm = match self.parse_one_of_keywords(&[ diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index a56335934..b6287d92d 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2454,6 +2454,52 @@ fn parse_alter_table_with_algorithm() { mysql_and_generic().verified_stmt("ALTER TABLE `users` ALGORITHM = COPY"); } +#[test] +fn parse_alter_table_with_lock() { + let sql = "ALTER TABLE tab LOCK = SHARED"; + let expected_operation = AlterTableOperation::Lock { + equals: true, + lock: AlterTableLock::Shared, + }; + let operation = alter_table_op(mysql_and_generic().verified_stmt(sql)); + assert_eq!(expected_operation, operation); + + let sql = + "ALTER TABLE users DROP COLUMN password_digest, LOCK = EXCLUSIVE, RENAME COLUMN name TO username"; + let stmt = mysql_and_generic().verified_stmt(sql); + match stmt { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![ + AlterTableOperation::DropColumn { + column_name: Ident::new("password_digest"), + if_exists: false, + drop_behavior: None, + }, + AlterTableOperation::Lock { + equals: true, + lock: AlterTableLock::Exclusive, + }, + AlterTableOperation::RenameColumn { + old_column_name: Ident::new("name"), + new_column_name: Ident::new("username") + }, + ] + ) + } + _ => panic!("Unexpected statement {stmt}"), + } + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK DEFAULT"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK SHARED"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK NONE"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK EXCLUSIVE"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK = DEFAULT"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK = SHARED"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK = NONE"); + mysql_and_generic().verified_stmt("ALTER TABLE `users` LOCK = EXCLUSIVE"); +} + #[test] fn parse_alter_table_auto_increment() { let sql = "ALTER TABLE tab AUTO_INCREMENT = 42"; From e3e88290cd44df48e9bde2e931985193218449fc Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Tue, 18 Mar 2025 15:19:51 +0100 Subject: [PATCH 078/286] Add support for `RAISE` statement (#1766) --- src/ast/mod.rs | 56 +++++++++++++++++++++++++++++++++++++++ src/ast/spans.rs | 28 ++++++++++++++++---- src/keywords.rs | 2 ++ src/parser/mod.rs | 20 ++++++++++++++ tests/sqlparser_common.rs | 23 ++++++++++++++++ 5 files changed, 124 insertions(+), 5 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 69f2e57f1..6dc4f5b20 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2256,6 +2256,57 @@ impl fmt::Display for ConditionalStatements { } } +/// A `RAISE` statement. +/// +/// Examples: +/// ```sql +/// RAISE USING MESSAGE = 'error'; +/// +/// RAISE myerror; +/// ``` +/// +/// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/procedural-language#raise) +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/snowflake-scripting/raise) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct RaiseStatement { + pub value: Option, +} + +impl fmt::Display for RaiseStatement { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + let RaiseStatement { value } = self; + + write!(f, "RAISE")?; + if let Some(value) = value { + write!(f, " {value}")?; + } + + Ok(()) + } +} + +/// Represents the error value of a [RaiseStatement]. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum RaiseStatementValue { + /// `RAISE USING MESSAGE = 'error'` + UsingMessage(Expr), + /// `RAISE myerror` + Expr(Expr), +} + +impl fmt::Display for RaiseStatementValue { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + RaiseStatementValue::Expr(expr) => write!(f, "{expr}"), + RaiseStatementValue::UsingMessage(expr) => write!(f, "USING MESSAGE = {expr}"), + } + } +} + /// Represents an expression assignment within a variable `DECLARE` statement. /// /// Examples: @@ -2827,6 +2878,8 @@ pub enum Statement { Case(CaseStatement), /// An `IF` statement. If(IfStatement), + /// A `RAISE` statement. + Raise(RaiseStatement), /// ```sql /// CALL /// ``` @@ -4142,6 +4195,9 @@ impl fmt::Display for Statement { Statement::If(stmt) => { write!(f, "{stmt}") } + Statement::Raise(stmt) => { + write!(f, "{stmt}") + } Statement::AttachDatabase { schema_name, database_file_name, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 783248a78..65d43c108 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -32,11 +32,11 @@ use super::{ JoinOperator, JsonPath, JsonPathElem, LateralView, LimitClause, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, - Query, ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, - SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, - TableAliasColumnDef, TableConstraint, TableFactor, TableObject, TableOptionsClustered, - TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, - WildcardAdditionalOptions, With, WithFill, + Query, RaiseStatement, RaiseStatementValue, ReferentialAction, RenameSelectItem, + ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, + Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, + TableFactor, TableObject, TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, + Value, Values, ViewColumnDef, WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -337,6 +337,7 @@ impl Spanned for Statement { } => source.span(), Statement::Case(stmt) => stmt.span(), Statement::If(stmt) => stmt.span(), + Statement::Raise(stmt) => stmt.span(), Statement::Call(function) => function.span(), Statement::Copy { source, @@ -782,6 +783,23 @@ impl Spanned for ConditionalStatements { } } +impl Spanned for RaiseStatement { + fn span(&self) -> Span { + let RaiseStatement { value } = self; + + union_spans(value.iter().map(|value| value.span())) + } +} + +impl Spanned for RaiseStatementValue { + fn span(&self) -> Span { + match self { + RaiseStatementValue::UsingMessage(expr) => expr.span(), + RaiseStatementValue::Expr(expr) => expr.span(), + } + } +} + /// # partial span /// /// Missing spans: diff --git a/src/keywords.rs b/src/keywords.rs index 974230f19..7b9c8bf29 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -533,6 +533,7 @@ define_keywords!( MEDIUMTEXT, MEMBER, MERGE, + MESSAGE, METADATA, METHOD, METRIC, @@ -695,6 +696,7 @@ define_keywords!( QUARTER, QUERY, QUOTE, + RAISE, RAISERROR, RANGE, RANK, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0d8edc05c..ee50cd04c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -536,6 +536,10 @@ impl<'a> Parser<'a> { self.prev_token(); self.parse_if_stmt() } + Keyword::RAISE => { + self.prev_token(); + self.parse_raise_stmt() + } Keyword::SELECT | Keyword::WITH | Keyword::VALUES | Keyword::FROM => { self.prev_token(); self.parse_query().map(Statement::Query) @@ -719,6 +723,22 @@ impl<'a> Parser<'a> { }) } + /// Parse a `RAISE` statement. + /// + /// See [Statement::Raise] + pub fn parse_raise_stmt(&mut self) -> Result { + self.expect_keyword_is(Keyword::RAISE)?; + + let value = if self.parse_keywords(&[Keyword::USING, Keyword::MESSAGE]) { + self.expect_token(&Token::Eq)?; + Some(RaiseStatementValue::UsingMessage(self.parse_expr()?)) + } else { + self.maybe_parse(|parser| parser.parse_expr().map(RaiseStatementValue::Expr))? + }; + + Ok(Statement::Raise(RaiseStatement { value })) + } + pub fn parse_comment(&mut self) -> Result { let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index c65cc6b53..c8df7cab3 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14298,6 +14298,29 @@ fn parse_if_statement() { ); } +#[test] +fn parse_raise_statement() { + let sql = "RAISE USING MESSAGE = 42"; + let Statement::Raise(stmt) = verified_stmt(sql) else { + unreachable!() + }; + assert_eq!( + Some(RaiseStatementValue::UsingMessage(Expr::value(number("42")))), + stmt.value + ); + + verified_stmt("RAISE USING MESSAGE = 'error'"); + verified_stmt("RAISE myerror"); + verified_stmt("RAISE 42"); + verified_stmt("RAISE using"); + verified_stmt("RAISE"); + + assert_eq!( + ParserError::ParserError("Expected: =, found: error".to_string()), + parse_sql_statements("RAISE USING MESSAGE error").unwrap_err() + ); +} + #[test] fn test_lambdas() { let dialects = all_dialects_where(|d| d.supports_lambda_functions()); From f487cbe00404454ce88a297bbe27dd98122f6fab Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Thu, 20 Mar 2025 07:52:56 +0200 Subject: [PATCH 079/286] Add GLOBAL context/modifier to SET statements (#1767) --- src/ast/mod.rs | 19 ++++++++++++------- src/parser/mod.rs | 29 +++++++++++++++++++---------- tests/sqlparser_common.rs | 32 ++++++++++++++++++++++++++------ tests/sqlparser_hive.rs | 9 +++++---- tests/sqlparser_mssql.rs | 2 +- tests/sqlparser_mysql.rs | 2 +- tests/sqlparser_postgres.rs | 18 ++++++++---------- 7 files changed, 72 insertions(+), 39 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 6dc4f5b20..9f895ee64 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2629,7 +2629,7 @@ pub enum Set { /// SQL Standard-style /// SET a = 1; SingleAssignment { - local: bool, + scope: ContextModifier, hivevar: bool, variable: ObjectName, values: Vec, @@ -2711,7 +2711,7 @@ impl Display for Set { role_name, } => { let role_name = role_name.clone().unwrap_or_else(|| Ident::new("NONE")); - write!(f, "SET{context_modifier} ROLE {role_name}") + write!(f, "SET {context_modifier}ROLE {role_name}") } Self::SetSessionParam(kind) => write!(f, "SET {kind}"), Self::SetTransaction { @@ -2758,7 +2758,7 @@ impl Display for Set { Ok(()) } Set::SingleAssignment { - local, + scope, hivevar, variable, values, @@ -2766,7 +2766,7 @@ impl Display for Set { write!( f, "SET {}{}{} = {}", - if *local { "LOCAL " } else { "" }, + scope, if *hivevar { "HIVEVAR:" } else { "" }, variable, display_comma_separated(values) @@ -7955,7 +7955,7 @@ impl fmt::Display for FlushLocation { } } -/// Optional context modifier for statements that can be or `LOCAL`, or `SESSION`. +/// Optional context modifier for statements that can be or `LOCAL`, `GLOBAL`, or `SESSION`. #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -7966,6 +7966,8 @@ pub enum ContextModifier { Local, /// `SESSION` identifier Session, + /// `GLOBAL` identifier + Global, } impl fmt::Display for ContextModifier { @@ -7975,10 +7977,13 @@ impl fmt::Display for ContextModifier { write!(f, "") } Self::Local => { - write!(f, " LOCAL") + write!(f, "LOCAL ") } Self::Session => { - write!(f, " SESSION") + write!(f, "SESSION ") + } + Self::Global => { + write!(f, "GLOBAL ") } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ee50cd04c..dcf7a4a8f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1819,6 +1819,15 @@ impl<'a> Parser<'a> { }) } + fn keyword_to_modifier(k: Option) -> ContextModifier { + match k { + Some(Keyword::LOCAL) => ContextModifier::Local, + Some(Keyword::GLOBAL) => ContextModifier::Global, + Some(Keyword::SESSION) => ContextModifier::Session, + _ => ContextModifier::None, + } + } + /// Check if the root is an identifier and all fields are identifiers. fn is_all_ident(root: &Expr, fields: &[AccessExpr]) -> bool { if !matches!(root, Expr::Identifier(_)) { @@ -11138,11 +11147,7 @@ impl<'a> Parser<'a> { /// Parse a `SET ROLE` statement. Expects SET to be consumed already. fn parse_set_role(&mut self, modifier: Option) -> Result { self.expect_keyword_is(Keyword::ROLE)?; - let context_modifier = match modifier { - Some(Keyword::LOCAL) => ContextModifier::Local, - Some(Keyword::SESSION) => ContextModifier::Session, - _ => ContextModifier::None, - }; + let context_modifier = Self::keyword_to_modifier(modifier); let role_name = if self.parse_keyword(Keyword::NONE) { None @@ -11214,8 +11219,12 @@ impl<'a> Parser<'a> { } fn parse_set(&mut self) -> Result { - let modifier = - self.parse_one_of_keywords(&[Keyword::SESSION, Keyword::LOCAL, Keyword::HIVEVAR]); + let modifier = self.parse_one_of_keywords(&[ + Keyword::SESSION, + Keyword::LOCAL, + Keyword::HIVEVAR, + Keyword::GLOBAL, + ]); if let Some(Keyword::HIVEVAR) = modifier { self.expect_token(&Token::Colon)?; @@ -11231,7 +11240,7 @@ impl<'a> Parser<'a> { { if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { return Ok(Set::SingleAssignment { - local: modifier == Some(Keyword::LOCAL), + scope: Self::keyword_to_modifier(modifier), hivevar: modifier == Some(Keyword::HIVEVAR), variable: ObjectName::from(vec!["TIMEZONE".into()]), values: self.parse_set_values(false)?, @@ -11321,7 +11330,7 @@ impl<'a> Parser<'a> { }?; Ok(Set::SingleAssignment { - local: modifier == Some(Keyword::LOCAL), + scope: Self::keyword_to_modifier(modifier), hivevar: modifier == Some(Keyword::HIVEVAR), variable, values, @@ -11349,7 +11358,7 @@ impl<'a> Parser<'a> { if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { let stmt = match variables { OneOrManyWithParens::One(var) => Set::SingleAssignment { - local: modifier == Some(Keyword::LOCAL), + scope: Self::keyword_to_modifier(modifier), hivevar: modifier == Some(Keyword::HIVEVAR), variable: var, values: self.parse_set_values(false)?, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index c8df7cab3..4ba8df7fb 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8627,12 +8627,12 @@ fn parse_set_transaction() { fn parse_set_variable() { match verified_stmt("SET SOMETHING = '1'") { Statement::Set(Set::SingleAssignment { - local, + scope, hivevar, variable, values, }) => { - assert!(!local); + assert_eq!(scope, ContextModifier::None); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["SOMETHING".into()])); assert_eq!( @@ -8645,6 +8645,26 @@ fn parse_set_variable() { _ => unreachable!(), } + match verified_stmt("SET GLOBAL VARIABLE = 'Value'") { + Statement::Set(Set::SingleAssignment { + scope, + hivevar, + variable, + values, + }) => { + assert_eq!(scope, ContextModifier::Global); + assert!(!hivevar); + assert_eq!(variable, ObjectName::from(vec!["VARIABLE".into()])); + assert_eq!( + values, + vec![Expr::Value( + (Value::SingleQuotedString("Value".into())).with_empty_span() + )] + ); + } + _ => unreachable!(), + } + let multi_variable_dialects = all_dialects_where(|d| d.supports_parenthesized_set_variables()); let sql = r#"SET (a, b, c) = (1, 2, 3)"#; match multi_variable_dialects.verified_stmt(sql) { @@ -8719,12 +8739,12 @@ fn parse_set_variable() { fn parse_set_role_as_variable() { match verified_stmt("SET role = 'foobar'") { Statement::Set(Set::SingleAssignment { - local, + scope, hivevar, variable, values, }) => { - assert!(!local); + assert_eq!(scope, ContextModifier::None); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["role".into()])); assert_eq!( @@ -8766,12 +8786,12 @@ fn parse_double_colon_cast_at_timezone() { fn parse_set_time_zone() { match verified_stmt("SET TIMEZONE = 'UTC'") { Statement::Set(Set::SingleAssignment { - local, + scope, hivevar, variable, values, }) => { - assert!(!local); + assert_eq!(scope, ContextModifier::None); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["TIMEZONE".into()])); assert_eq!( diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 56fe22a0d..a9549cb60 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -21,9 +21,10 @@ //! is also tested (on the inputs it can handle). use sqlparser::ast::{ - ClusteredBy, CommentDef, CreateFunction, CreateFunctionBody, CreateFunctionUsing, CreateTable, - Expr, Function, FunctionArgumentList, FunctionArguments, Ident, ObjectName, OrderByExpr, - OrderByOptions, SelectItem, Set, Statement, TableFactor, UnaryOperator, Use, Value, + ClusteredBy, CommentDef, ContextModifier, CreateFunction, CreateFunctionBody, + CreateFunctionUsing, CreateTable, Expr, Function, FunctionArgumentList, FunctionArguments, + Ident, ObjectName, OrderByExpr, OrderByOptions, SelectItem, Set, Statement, TableFactor, + UnaryOperator, Use, Value, }; use sqlparser::dialect::{GenericDialect, HiveDialect, MsSqlDialect}; use sqlparser::parser::ParserError; @@ -369,7 +370,7 @@ fn set_statement_with_minus() { assert_eq!( hive().verified_stmt("SET hive.tez.java.opts = -Xmx4g"), Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![ Ident::new("hive"), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index af71d2523..d4e5fa719 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1251,7 +1251,7 @@ fn parse_mssql_declare() { }] }, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![Ident::new("@bar")]), values: vec![Expr::Value( diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index b6287d92d..884351491 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -618,7 +618,7 @@ fn parse_set_variables() { assert_eq!( mysql_and_generic().verified_stmt("SET LOCAL autocommit = 1"), Statement::Set(Set::SingleAssignment { - local: true, + scope: ContextModifier::Local, hivevar: false, variable: ObjectName::from(vec!["autocommit".into()]), values: vec![Expr::value(number("1"))], diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index e62f23597..cf66af74e 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -988,8 +988,7 @@ fn parse_create_schema_if_not_exists() { Statement::CreateSchema { if_not_exists: true, schema_name, - options: _, - default_collate_spec: _, + .. } => assert_eq!("schema_name", schema_name.to_string()), _ => unreachable!(), } @@ -1433,7 +1432,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Identifier(Ident { @@ -1448,7 +1447,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Value( @@ -1461,7 +1460,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::value(number("0"))], @@ -1472,7 +1471,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Identifier(Ident::new("DEFAULT"))], @@ -1483,7 +1482,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: true, + scope: ContextModifier::Local, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Identifier("b".into())], @@ -1494,7 +1493,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a"), Ident::new("b"), Ident::new("c")]), values: vec![Expr::Identifier(Ident { @@ -1512,7 +1511,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - local: false, + scope: ContextModifier::None, hivevar: false, variable: ObjectName::from(vec![ Ident::new("hive"), @@ -1526,7 +1525,6 @@ fn parse_set() { ); pg_and_generic().one_statement_parses_to("SET a TO b", "SET a = b"); - pg_and_generic().one_statement_parses_to("SET SESSION a = b", "SET a = b"); assert_eq!( pg_and_generic().parse_sql_statements("SET"), From 939fbdd4f62014d478b624b4d0429fcd06d775a6 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 21 Mar 2025 22:34:43 -0700 Subject: [PATCH 080/286] Parse `SUBSTR` as alias for `SUBSTRING` (#1769) --- src/ast/mod.rs | 11 ++++++++++- src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 18 +++++++++++++++--- tests/sqlparser_common.rs | 3 +++ tests/sqlparser_mssql.rs | 1 + tests/sqlparser_mysql.rs | 1 + 7 files changed, 32 insertions(+), 4 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 9f895ee64..b1bce3e97 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -890,6 +890,10 @@ pub enum Expr { /// true if the expression is represented using the `SUBSTRING(expr, start, len)` syntax /// This flag is used for formatting. special: bool, + + /// true if the expression is represented using the `SUBSTR` shorthand + /// This flag is used for formatting. + shorthand: bool, }, /// ```sql /// TRIM([BOTH | LEADING | TRAILING] [ FROM] ) @@ -1719,8 +1723,13 @@ impl fmt::Display for Expr { substring_from, substring_for, special, + shorthand, } => { - write!(f, "SUBSTRING({expr}")?; + f.write_str("SUBSTR")?; + if !*shorthand { + f.write_str("ING")?; + } + write!(f, "({expr}")?; if let Some(from_part) = substring_from { if *special { write!(f, ", {from_part}")?; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 65d43c108..11770d1bc 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1503,6 +1503,7 @@ impl Spanned for Expr { substring_from, substring_for, special: _, + shorthand: _, } => union_spans( core::iter::once(expr.span()) .chain(substring_from.as_ref().map(|i| i.span())) diff --git a/src/keywords.rs b/src/keywords.rs index 7b9c8bf29..349c9ffb0 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -841,6 +841,7 @@ define_keywords!( STRING, STRUCT, SUBMULTISET, + SUBSTR, SUBSTRING, SUBSTRING_REGEX, SUCCEEDS, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index dcf7a4a8f..a43be5c68 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1302,7 +1302,10 @@ impl<'a> Parser<'a> { Keyword::POSITION if self.peek_token_ref().token == Token::LParen => { Ok(Some(self.parse_position_expr(w.clone().into_ident(w_span))?)) } - Keyword::SUBSTRING => Ok(Some(self.parse_substring_expr()?)), + Keyword::SUBSTR | Keyword::SUBSTRING => { + self.prev_token(); + Ok(Some(self.parse_substring()?)) + } Keyword::OVERLAY => Ok(Some(self.parse_overlay_expr()?)), Keyword::TRIM => Ok(Some(self.parse_trim_expr()?)), Keyword::INTERVAL => Ok(Some(self.parse_interval()?)), @@ -2412,8 +2415,16 @@ impl<'a> Parser<'a> { } } - pub fn parse_substring_expr(&mut self) -> Result { - // PARSE SUBSTRING (EXPR [FROM 1] [FOR 3]) + // { SUBSTRING | SUBSTR } ( [FROM 1] [FOR 3]) + pub fn parse_substring(&mut self) -> Result { + let shorthand = match self.expect_one_of_keywords(&[Keyword::SUBSTR, Keyword::SUBSTRING])? { + Keyword::SUBSTR => true, + Keyword::SUBSTRING => false, + _ => { + self.prev_token(); + return self.expected("SUBSTR or SUBSTRING", self.peek_token()); + } + }; self.expect_token(&Token::LParen)?; let expr = self.parse_expr()?; let mut from_expr = None; @@ -2433,6 +2444,7 @@ impl<'a> Parser<'a> { substring_from: from_expr.map(Box::new), substring_for: to_expr.map(Box::new), special, + shorthand, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 4ba8df7fb..cf95e2ae9 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7607,6 +7607,9 @@ fn parse_substring() { verified_stmt("SELECT SUBSTRING('1', 1, 3)"); verified_stmt("SELECT SUBSTRING('1', 1)"); verified_stmt("SELECT SUBSTRING('1' FOR 3)"); + verified_stmt("SELECT SUBSTRING('foo' FROM 1 FOR 2) FROM t"); + verified_stmt("SELECT SUBSTR('foo' FROM 1 FOR 2) FROM t"); + verified_stmt("SELECT SUBSTR('foo', 1, 2) FROM t"); } #[test] diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index d4e5fa719..4ea42efc7 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1133,6 +1133,7 @@ fn parse_substring_in_select() { (number("1")).with_empty_span() ))), special: true, + shorthand: false, })], into: None, from: vec![TableWithJoins { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 884351491..2767a78c8 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2590,6 +2590,7 @@ fn parse_substring_in_select() { (number("1")).with_empty_span() ))), special: true, + shorthand: false, })], into: None, from: vec![TableWithJoins { From 3a8a3bb7a52c2a855e40ccbf88c3073abda1f1e7 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Sat, 22 Mar 2025 07:38:00 +0200 Subject: [PATCH 081/286] SET statements: scope modifier for multiple assignments (#1772) --- src/ast/mod.rs | 28 ++++++--- src/dialect/generic.rs | 4 ++ src/parser/mod.rs | 121 +++++++++++++++++------------------- tests/sqlparser_common.rs | 43 +++++++++++-- tests/sqlparser_hive.rs | 9 ++- tests/sqlparser_mssql.rs | 2 +- tests/sqlparser_mysql.rs | 2 +- tests/sqlparser_postgres.rs | 20 +++--- 8 files changed, 134 insertions(+), 95 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index b1bce3e97..3264cf038 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2638,7 +2638,7 @@ pub enum Set { /// SQL Standard-style /// SET a = 1; SingleAssignment { - scope: ContextModifier, + scope: Option, hivevar: bool, variable: ObjectName, values: Vec, @@ -2668,7 +2668,7 @@ pub enum Set { /// [4]: https://docs.oracle.com/cd/B19306_01/server.102/b14200/statements_10004.htm SetRole { /// Non-ANSI optional identifier to inform if the role is defined inside the current session (`SESSION`) or transaction (`LOCAL`). - context_modifier: ContextModifier, + context_modifier: Option, /// Role name. If NONE is specified, then the current role name is removed. role_name: Option, }, @@ -2720,7 +2720,13 @@ impl Display for Set { role_name, } => { let role_name = role_name.clone().unwrap_or_else(|| Ident::new("NONE")); - write!(f, "SET {context_modifier}ROLE {role_name}") + write!( + f, + "SET {modifier}ROLE {role_name}", + modifier = context_modifier + .map(|m| format!("{}", m)) + .unwrap_or_default() + ) } Self::SetSessionParam(kind) => write!(f, "SET {kind}"), Self::SetTransaction { @@ -2775,7 +2781,7 @@ impl Display for Set { write!( f, "SET {}{}{} = {}", - scope, + scope.map(|s| format!("{}", s)).unwrap_or_default(), if *hivevar { "HIVEVAR:" } else { "" }, variable, display_comma_separated(values) @@ -5736,13 +5742,20 @@ impl fmt::Display for SequenceOptions { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct SetAssignment { + pub scope: Option, pub name: ObjectName, pub value: Expr, } impl fmt::Display for SetAssignment { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{} = {}", self.name, self.value) + write!( + f, + "{}{} = {}", + self.scope.map(|s| format!("{}", s)).unwrap_or_default(), + self.name, + self.value + ) } } @@ -7969,8 +7982,6 @@ impl fmt::Display for FlushLocation { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum ContextModifier { - /// No context defined. Each dialect defines the default in this scenario. - None, /// `LOCAL` identifier, usually related to transactional states. Local, /// `SESSION` identifier @@ -7982,9 +7993,6 @@ pub enum ContextModifier { impl fmt::Display for ContextModifier { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - Self::None => { - write!(f, "") - } Self::Local => { write!(f, "LOCAL ") } diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index c13d5aa69..92cfca8fd 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -159,4 +159,8 @@ impl Dialect for GenericDialect { fn supports_set_names(&self) -> bool { true } + + fn supports_comma_separated_set_assignments(&self) -> bool { + true + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a43be5c68..7d8417f36 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1822,12 +1822,12 @@ impl<'a> Parser<'a> { }) } - fn keyword_to_modifier(k: Option) -> ContextModifier { + fn keyword_to_modifier(k: Keyword) -> Option { match k { - Some(Keyword::LOCAL) => ContextModifier::Local, - Some(Keyword::GLOBAL) => ContextModifier::Global, - Some(Keyword::SESSION) => ContextModifier::Session, - _ => ContextModifier::None, + Keyword::LOCAL => Some(ContextModifier::Local), + Keyword::GLOBAL => Some(ContextModifier::Global), + Keyword::SESSION => Some(ContextModifier::Session), + _ => None, } } @@ -11157,9 +11157,11 @@ impl<'a> Parser<'a> { } /// Parse a `SET ROLE` statement. Expects SET to be consumed already. - fn parse_set_role(&mut self, modifier: Option) -> Result { + fn parse_set_role( + &mut self, + modifier: Option, + ) -> Result { self.expect_keyword_is(Keyword::ROLE)?; - let context_modifier = Self::keyword_to_modifier(modifier); let role_name = if self.parse_keyword(Keyword::NONE) { None @@ -11167,7 +11169,7 @@ impl<'a> Parser<'a> { Some(self.parse_identifier()?) }; Ok(Statement::Set(Set::SetRole { - context_modifier, + context_modifier: modifier, role_name, })) } @@ -11203,46 +11205,52 @@ impl<'a> Parser<'a> { } } - fn parse_set_assignment( - &mut self, - ) -> Result<(OneOrManyWithParens, Expr), ParserError> { - let variables = if self.dialect.supports_parenthesized_set_variables() + fn parse_context_modifier(&mut self) -> Option { + let modifier = + self.parse_one_of_keywords(&[Keyword::SESSION, Keyword::LOCAL, Keyword::GLOBAL])?; + + Self::keyword_to_modifier(modifier) + } + + /// Parse a single SET statement assignment `var = expr`. + fn parse_set_assignment(&mut self) -> Result { + let scope = self.parse_context_modifier(); + + let name = if self.dialect.supports_parenthesized_set_variables() && self.consume_token(&Token::LParen) { - let vars = OneOrManyWithParens::Many( - self.parse_comma_separated(|parser: &mut Parser<'a>| parser.parse_identifier())? - .into_iter() - .map(|ident| ObjectName::from(vec![ident])) - .collect(), - ); - self.expect_token(&Token::RParen)?; - vars + // Parenthesized assignments are handled in the `parse_set` function after + // trying to parse list of assignments using this function. + // If a dialect supports both, and we find a LParen, we early exit from this function. + self.expected("Unparenthesized assignment", self.peek_token())? } else { - OneOrManyWithParens::One(self.parse_object_name(false)?) + self.parse_object_name(false)? }; if !(self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO)) { return self.expected("assignment operator", self.peek_token()); } - let values = self.parse_expr()?; + let value = self.parse_expr()?; - Ok((variables, values)) + Ok(SetAssignment { scope, name, value }) } fn parse_set(&mut self) -> Result { - let modifier = self.parse_one_of_keywords(&[ - Keyword::SESSION, - Keyword::LOCAL, - Keyword::HIVEVAR, - Keyword::GLOBAL, - ]); - - if let Some(Keyword::HIVEVAR) = modifier { + let hivevar = self.parse_keyword(Keyword::HIVEVAR); + + // Modifier is either HIVEVAR: or a ContextModifier (LOCAL, SESSION, etc), not both + let scope = if !hivevar { + self.parse_context_modifier() + } else { + None + }; + + if hivevar { self.expect_token(&Token::Colon)?; } - if let Some(set_role_stmt) = self.maybe_parse(|parser| parser.parse_set_role(modifier))? { + if let Some(set_role_stmt) = self.maybe_parse(|parser| parser.parse_set_role(scope))? { return Ok(set_role_stmt); } @@ -11252,8 +11260,8 @@ impl<'a> Parser<'a> { { if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { return Ok(Set::SingleAssignment { - scope: Self::keyword_to_modifier(modifier), - hivevar: modifier == Some(Keyword::HIVEVAR), + scope, + hivevar, variable: ObjectName::from(vec!["TIMEZONE".into()]), values: self.parse_set_values(false)?, } @@ -11263,7 +11271,7 @@ impl<'a> Parser<'a> { // the assignment operator. It's originally PostgreSQL specific, // but we allow it for all the dialects return Ok(Set::SetTimeZone { - local: modifier == Some(Keyword::LOCAL), + local: scope == Some(ContextModifier::Local), value: self.parse_expr()?, } .into()); @@ -11311,41 +11319,26 @@ impl<'a> Parser<'a> { } if self.dialect.supports_comma_separated_set_assignments() { + if scope.is_some() { + self.prev_token(); + } + if let Some(assignments) = self .maybe_parse(|parser| parser.parse_comma_separated(Parser::parse_set_assignment))? { return if assignments.len() > 1 { - let assignments = assignments - .into_iter() - .map(|(var, val)| match var { - OneOrManyWithParens::One(v) => Ok(SetAssignment { - name: v, - value: val, - }), - OneOrManyWithParens::Many(_) => { - self.expected("List of single identifiers", self.peek_token()) - } - }) - .collect::>()?; - Ok(Set::MultipleAssignments { assignments }.into()) } else { - let (vars, values): (Vec<_>, Vec<_>) = assignments.into_iter().unzip(); - - let variable = match vars.into_iter().next() { - Some(OneOrManyWithParens::One(v)) => Ok(v), - Some(OneOrManyWithParens::Many(_)) => self.expected( - "Single assignment or list of assignments", - self.peek_token(), - ), - None => self.expected("At least one identifier", self.peek_token()), - }?; + let SetAssignment { scope, name, value } = + assignments.into_iter().next().ok_or_else(|| { + ParserError::ParserError("Expected at least one assignment".to_string()) + })?; Ok(Set::SingleAssignment { - scope: Self::keyword_to_modifier(modifier), - hivevar: modifier == Some(Keyword::HIVEVAR), - variable, - values, + scope, + hivevar, + variable: name, + values: vec![value], } .into()) }; @@ -11370,8 +11363,8 @@ impl<'a> Parser<'a> { if self.consume_token(&Token::Eq) || self.parse_keyword(Keyword::TO) { let stmt = match variables { OneOrManyWithParens::One(var) => Set::SingleAssignment { - scope: Self::keyword_to_modifier(modifier), - hivevar: modifier == Some(Keyword::HIVEVAR), + scope, + hivevar, variable: var, values: self.parse_set_values(false)?, }, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index cf95e2ae9..a137291ac 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8635,7 +8635,7 @@ fn parse_set_variable() { variable, values, }) => { - assert_eq!(scope, ContextModifier::None); + assert_eq!(scope, None); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["SOMETHING".into()])); assert_eq!( @@ -8655,7 +8655,7 @@ fn parse_set_variable() { variable, values, }) => { - assert_eq!(scope, ContextModifier::Global); + assert_eq!(scope, Some(ContextModifier::Global)); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["VARIABLE".into()])); assert_eq!( @@ -8747,7 +8747,7 @@ fn parse_set_role_as_variable() { variable, values, }) => { - assert_eq!(scope, ContextModifier::None); + assert_eq!(scope, None); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["role".into()])); assert_eq!( @@ -8794,7 +8794,7 @@ fn parse_set_time_zone() { variable, values, }) => { - assert_eq!(scope, ContextModifier::None); + assert_eq!(scope, None); assert!(!hivevar); assert_eq!(variable, ObjectName::from(vec!["TIMEZONE".into()])); assert_eq!( @@ -14859,10 +14859,12 @@ fn parse_multiple_set_statements() -> Result<(), ParserError> { assignments, vec![ SetAssignment { + scope: None, name: ObjectName::from(vec!["@a".into()]), value: Expr::value(number("1")) }, SetAssignment { + scope: None, name: ObjectName::from(vec!["b".into()]), value: Expr::value(number("2")) } @@ -14872,6 +14874,39 @@ fn parse_multiple_set_statements() -> Result<(), ParserError> { _ => panic!("Expected SetVariable with 2 variables and 2 values"), }; + let stmt = dialects.verified_stmt("SET GLOBAL @a = 1, SESSION b = 2, LOCAL c = 3, d = 4"); + + match stmt { + Statement::Set(Set::MultipleAssignments { assignments }) => { + assert_eq!( + assignments, + vec![ + SetAssignment { + scope: Some(ContextModifier::Global), + name: ObjectName::from(vec!["@a".into()]), + value: Expr::value(number("1")) + }, + SetAssignment { + scope: Some(ContextModifier::Session), + name: ObjectName::from(vec!["b".into()]), + value: Expr::value(number("2")) + }, + SetAssignment { + scope: Some(ContextModifier::Local), + name: ObjectName::from(vec!["c".into()]), + value: Expr::value(number("3")) + }, + SetAssignment { + scope: None, + name: ObjectName::from(vec!["d".into()]), + value: Expr::value(number("4")) + } + ] + ); + } + _ => panic!("Expected MultipleAssignments with 4 scoped variables and 4 values"), + }; + Ok(()) } diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index a9549cb60..2af93db7d 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -21,10 +21,9 @@ //! is also tested (on the inputs it can handle). use sqlparser::ast::{ - ClusteredBy, CommentDef, ContextModifier, CreateFunction, CreateFunctionBody, - CreateFunctionUsing, CreateTable, Expr, Function, FunctionArgumentList, FunctionArguments, - Ident, ObjectName, OrderByExpr, OrderByOptions, SelectItem, Set, Statement, TableFactor, - UnaryOperator, Use, Value, + ClusteredBy, CommentDef, CreateFunction, CreateFunctionBody, CreateFunctionUsing, CreateTable, + Expr, Function, FunctionArgumentList, FunctionArguments, Ident, ObjectName, OrderByExpr, + OrderByOptions, SelectItem, Set, Statement, TableFactor, UnaryOperator, Use, Value, }; use sqlparser::dialect::{GenericDialect, HiveDialect, MsSqlDialect}; use sqlparser::parser::ParserError; @@ -370,7 +369,7 @@ fn set_statement_with_minus() { assert_eq!( hive().verified_stmt("SET hive.tez.java.opts = -Xmx4g"), Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![ Ident::new("hive"), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 4ea42efc7..ca59b164e 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1252,7 +1252,7 @@ fn parse_mssql_declare() { }] }, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![Ident::new("@bar")]), values: vec![Expr::Value( diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 2767a78c8..d52619d54 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -618,7 +618,7 @@ fn parse_set_variables() { assert_eq!( mysql_and_generic().verified_stmt("SET LOCAL autocommit = 1"), Statement::Set(Set::SingleAssignment { - scope: ContextModifier::Local, + scope: Some(ContextModifier::Local), hivevar: false, variable: ObjectName::from(vec!["autocommit".into()]), values: vec![Expr::value(number("1"))], diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index cf66af74e..a6d65ec75 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1432,7 +1432,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Identifier(Ident { @@ -1447,7 +1447,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Value( @@ -1460,7 +1460,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::value(number("0"))], @@ -1471,7 +1471,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Identifier(Ident::new("DEFAULT"))], @@ -1482,7 +1482,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::Local, + scope: Some(ContextModifier::Local), hivevar: false, variable: ObjectName::from(vec![Ident::new("a")]), values: vec![Expr::Identifier("b".into())], @@ -1493,7 +1493,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![Ident::new("a"), Ident::new("b"), Ident::new("c")]), values: vec![Expr::Identifier(Ident { @@ -1511,7 +1511,7 @@ fn parse_set() { assert_eq!( stmt, Statement::Set(Set::SingleAssignment { - scope: ContextModifier::None, + scope: None, hivevar: false, variable: ObjectName::from(vec![ Ident::new("hive"), @@ -1555,7 +1555,7 @@ fn parse_set_role() { assert_eq!( stmt, Statement::Set(Set::SetRole { - context_modifier: ContextModifier::Session, + context_modifier: Some(ContextModifier::Session), role_name: None, }) ); @@ -1566,7 +1566,7 @@ fn parse_set_role() { assert_eq!( stmt, Statement::Set(Set::SetRole { - context_modifier: ContextModifier::Local, + context_modifier: Some(ContextModifier::Local), role_name: Some(Ident { value: "rolename".to_string(), quote_style: Some('\"'), @@ -1581,7 +1581,7 @@ fn parse_set_role() { assert_eq!( stmt, Statement::Set(Set::SetRole { - context_modifier: ContextModifier::None, + context_modifier: None, role_name: Some(Ident { value: "rolename".to_string(), quote_style: Some('\''), From 53aba68e2dc758e591292115b1dd5faf71374346 Mon Sep 17 00:00:00 2001 From: tomershaniii <65544633+tomershaniii@users.noreply.github.com> Date: Tue, 25 Mar 2025 11:13:11 +0200 Subject: [PATCH 082/286] Support qualified column names in `MATCH AGAINST` clause (#1774) --- src/ast/mod.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_mysql.rs | 29 +++++++++++++++++++++++++++++ 3 files changed, 31 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 3264cf038..f187df995 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -1027,7 +1027,7 @@ pub enum Expr { /// [(1)]: https://dev.mysql.com/doc/refman/8.0/en/fulltext-search.html#function_match MatchAgainst { /// `(, , ...)`. - columns: Vec, + columns: Vec, /// ``. match_value: Value, /// `` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 7d8417f36..65d536f7e 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2704,7 +2704,7 @@ impl<'a> Parser<'a> { /// This method will raise an error if the column list is empty or with invalid identifiers, /// the match expression is not a literal string, or if the search modifier is not valid. pub fn parse_match_against(&mut self) -> Result { - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_qualified_column_list(Mandatory, false)?; self.expect_keyword_is(Keyword::AGAINST)?; diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index d52619d54..3d318f701 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3454,3 +3454,32 @@ fn parse_cast_integers() { .run_parser_method("CAST(foo AS UNSIGNED INTEGER(3))", |p| p.parse_expr()) .expect_err("CAST doesn't allow display width"); } + +#[test] +fn parse_match_against_with_alias() { + let sql = "SELECT tbl.ProjectID FROM surveys.tbl1 AS tbl WHERE MATCH (tbl.ReferenceID) AGAINST ('AAA' IN BOOLEAN MODE)"; + match mysql().verified_stmt(sql) { + Statement::Query(query) => match *query.body { + SetExpr::Select(select) => match select.selection { + Some(Expr::MatchAgainst { + columns, + match_value, + opt_search_modifier, + }) => { + assert_eq!( + columns, + vec![ObjectName::from(vec![ + Ident::new("tbl"), + Ident::new("ReferenceID") + ])] + ); + assert_eq!(match_value, Value::SingleQuotedString("AAA".to_owned())); + assert_eq!(opt_search_modifier, Some(SearchModifier::InBooleanMode)); + } + _ => unreachable!(), + }, + _ => unreachable!(), + }, + _ => unreachable!(), + } +} From 62495f2f0d998690495f111063a0b0a3466ffa69 Mon Sep 17 00:00:00 2001 From: bar sela Date: Thu, 27 Mar 2025 23:48:57 +0200 Subject: [PATCH 083/286] Mysql: Add support for := operator (#1779) --- src/ast/operator.rs | 4 ++ src/dialect/mod.rs | 3 +- src/parser/mod.rs | 1 + tests/sqlparser_mysql.rs | 104 +++++++++++++++++++++++++++++++++++++++ 4 files changed, 111 insertions(+), 1 deletion(-) diff --git a/src/ast/operator.rs b/src/ast/operator.rs index 66a35fee5..73fe9cf42 100644 --- a/src/ast/operator.rs +++ b/src/ast/operator.rs @@ -321,6 +321,9 @@ pub enum BinaryOperator { /// `~=` Same as? (PostgreSQL/Redshift geometric operator) /// See TildeEq, + /// ':=' Assignment Operator + /// See + Assignment, } impl fmt::Display for BinaryOperator { @@ -394,6 +397,7 @@ impl fmt::Display for BinaryOperator { BinaryOperator::QuestionDoublePipe => f.write_str("?||"), BinaryOperator::At => f.write_str("@"), BinaryOperator::TildeEq => f.write_str("~="), + BinaryOperator::Assignment => f.write_str(":="), } } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 8d4557e2f..79184dd70 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -620,7 +620,8 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::OPERATOR => Ok(p!(Between)), Token::Word(w) if w.keyword == Keyword::DIV => Ok(p!(MulDivModOp)), Token::Period => Ok(p!(Period)), - Token::Eq + Token::Assignment + | Token::Eq | Token::Lt | Token::LtEq | Token::Neq diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 65d536f7e..adaae2862 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3233,6 +3233,7 @@ impl<'a> Parser<'a> { let regular_binary_operator = match &tok.token { Token::Spaceship => Some(BinaryOperator::Spaceship), Token::DoubleEq => Some(BinaryOperator::Eq), + Token::Assignment => Some(BinaryOperator::Assignment), Token::Eq => Some(BinaryOperator::Eq), Token::Neq => Some(BinaryOperator::NotEq), Token::Gt => Some(BinaryOperator::Gt), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 3d318f701..1d4fd6a0d 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3483,3 +3483,107 @@ fn parse_match_against_with_alias() { _ => unreachable!(), } } + +#[test] +fn test_variable_assignment_using_colon_equal() { + let sql_select = "SELECT @price := price, @tax := price * 0.1 FROM products WHERE id = 1"; + let stmt = mysql().verified_stmt(sql_select); + match stmt { + Statement::Query(query) => { + let select = query.body.as_select().unwrap(); + + assert_eq!( + select.projection, + vec![ + SelectItem::UnnamedExpr(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident { + value: "@price".to_string(), + quote_style: None, + span: Span::empty(), + })), + op: BinaryOperator::Assignment, + right: Box::new(Expr::Identifier(Ident { + value: "price".to_string(), + quote_style: None, + span: Span::empty(), + })), + }), + SelectItem::UnnamedExpr(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident { + value: "@tax".to_string(), + quote_style: None, + span: Span::empty(), + })), + op: BinaryOperator::Assignment, + right: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident { + value: "price".to_string(), + quote_style: None, + span: Span::empty(), + })), + op: BinaryOperator::Multiply, + right: Box::new(Expr::Value( + (test_utils::number("0.1")).with_empty_span() + )), + }), + }), + ] + ); + + assert_eq!( + select.selection, + Some(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident { + value: "id".to_string(), + quote_style: None, + span: Span::empty(), + })), + op: BinaryOperator::Eq, + right: Box::new(Expr::Value((test_utils::number("1")).with_empty_span())), + }) + ); + } + _ => panic!("Unexpected statement {stmt}"), + } + + let sql_update = + "UPDATE products SET price = @new_price := price * 1.1 WHERE category = 'Books'"; + let stmt = mysql().verified_stmt(sql_update); + + match stmt { + Statement::Update { assignments, .. } => { + assert_eq!( + assignments, + vec![Assignment { + target: AssignmentTarget::ColumnName(ObjectName(vec![ + ObjectNamePart::Identifier(Ident { + value: "price".to_string(), + quote_style: None, + span: Span::empty(), + }) + ])), + value: Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident { + value: "@new_price".to_string(), + quote_style: None, + span: Span::empty(), + })), + op: BinaryOperator::Assignment, + right: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident { + value: "price".to_string(), + quote_style: None, + span: Span::empty(), + })), + op: BinaryOperator::Multiply, + right: Box::new(Expr::Value( + (test_utils::number("1.1")).with_empty_span() + )), + }), + }, + }] + ) + } + _ => panic!("Unexpected statement {stmt}"), + } +} From be98b30eb3afc811e971856cbdf9809d32509af1 Mon Sep 17 00:00:00 2001 From: Dan Draper Date: Mon, 31 Mar 2025 21:46:59 +1100 Subject: [PATCH 084/286] Add cipherstash-proxy to list of users in README.md (#1782) --- README.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index d18a76b50..5e8e460f6 100644 --- a/README.md +++ b/README.md @@ -156,7 +156,7 @@ $ cargo run --features json_example --example cli FILENAME.sql [--dialectname] ## Users This parser is currently being used by the [DataFusion] query engine, [LocustDB], -[Ballista], [GlueSQL], [Opteryx], [Polars], [PRQL], [Qrlew], [JumpWire], and [ParadeDB]. +[Ballista], [GlueSQL], [Opteryx], [Polars], [PRQL], [Qrlew], [JumpWire], [ParadeDB] and [CipherStash Proxy]. If your project is using sqlparser-rs feel free to make a PR to add it to this list. @@ -275,3 +275,4 @@ licensed as above, without any additional terms or conditions. [sql-standard]: https://en.wikipedia.org/wiki/ISO/IEC_9075 [`Dialect`]: https://docs.rs/sqlparser/latest/sqlparser/dialect/trait.Dialect.html [`GenericDialect`]: https://docs.rs/sqlparser/latest/sqlparser/dialect/struct.GenericDialect.html +[CipherStash Proxy]: https://github.com/cipherstash/proxy From 95d7b86da54c9cc4c75c47adffe73bf2cc38dd71 Mon Sep 17 00:00:00 2001 From: John Vandenberg Date: Mon, 31 Mar 2025 21:47:53 +1100 Subject: [PATCH 085/286] Fix typos (#1785) --- src/dialect/snowflake.rs | 2 +- src/keywords.rs | 2 +- tests/sqlparser_common.rs | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 09a0e57ce..2b04f9e9a 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -310,7 +310,7 @@ impl Dialect for SnowflakeDialect { } // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` - // which would give it a different meanins, for example: `SELECT 1 FETCH FIRST 10 ROWS` - not an alias + // which would give it a different meanings, for example: `SELECT 1 FETCH FIRST 10 ROWS` - not an alias Keyword::FETCH if parser.peek_keyword(Keyword::FIRST) || parser.peek_keyword(Keyword::NEXT) => { diff --git a/src/keywords.rs b/src/keywords.rs index 349c9ffb0..a0c556d2d 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -1084,7 +1084,7 @@ pub const RESERVED_FOR_COLUMN_ALIAS: &[Keyword] = &[ Keyword::END, ]; -// Global list of reserved keywords alloweed after FROM. +// Global list of reserved keywords allowed after FROM. // Parser should call Dialect::get_reserved_keyword_after_from // to allow for each dialect to customize the list. pub const RESERVED_FOR_TABLE_FACTOR: &[Keyword] = &[ diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a137291ac..795dae4b3 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14528,7 +14528,7 @@ fn test_geometric_unary_operators() { } #[test] -fn test_geomtery_type() { +fn test_geometry_type() { let sql = "point '1,2'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), From 91327bb0c02a09e0b5c5322c813c4b2a3b564439 Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Mon, 31 Mar 2025 17:51:55 +0200 Subject: [PATCH 086/286] Add support for Databricks TIMESTAMP_NTZ. (#1781) Co-authored-by: Roman Borschel --- src/ast/data_type.rs | 5 +++++ src/keywords.rs | 1 + src/parser/mod.rs | 1 + tests/sqlparser_databricks.rs | 40 +++++++++++++++++++++++++++++++++++ 4 files changed, 47 insertions(+) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 57bc67441..dc523696a 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -312,6 +312,10 @@ pub enum DataType { /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type Timestamp(Option, TimezoneInfo), + /// Databricks timestamp without time zone. See [1]. + /// + /// [1]: https://docs.databricks.com/aws/en/sql/language-manual/data-types/timestamp-ntz-type + TimestampNtz, /// Interval Interval, /// JSON type @@ -567,6 +571,7 @@ impl fmt::Display for DataType { DataType::Timestamp(precision, timezone_info) => { format_datetime_precision_and_tz(f, "TIMESTAMP", precision, timezone_info) } + DataType::TimestampNtz => write!(f, "TIMESTAMP_NTZ"), DataType::Datetime64(precision, timezone) => { format_clickhouse_datetime_precision_and_timezone( f, diff --git a/src/keywords.rs b/src/keywords.rs index a0c556d2d..8609ec43d 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -875,6 +875,7 @@ define_keywords!( TIME, TIMESTAMP, TIMESTAMPTZ, + TIMESTAMP_NTZ, TIMETZ, TIMEZONE, TIMEZONE_ABBR, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index adaae2862..2b61529ff 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9247,6 +9247,7 @@ impl<'a> Parser<'a> { self.parse_optional_precision()?, TimezoneInfo::Tz, )), + Keyword::TIMESTAMP_NTZ => Ok(DataType::TimestampNtz), Keyword::TIME => { let precision = self.parse_optional_precision()?; let tz = if self.parse_keyword(Keyword::WITH) { diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 3b36d7a13..88aae499a 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -317,3 +317,43 @@ fn parse_databricks_struct_function() { }) ); } + +#[test] +fn data_type_timestamp_ntz() { + // Literal + assert_eq!( + databricks().verified_expr("TIMESTAMP_NTZ '2025-03-29T18:52:00'"), + Expr::TypedString { + data_type: DataType::TimestampNtz, + value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()) + } + ); + + // Cast + assert_eq!( + databricks().verified_expr("(created_at)::TIMESTAMP_NTZ"), + Expr::Cast { + kind: CastKind::DoubleColon, + expr: Box::new(Expr::Nested(Box::new(Expr::Identifier( + "created_at".into() + )))), + data_type: DataType::TimestampNtz, + format: None + } + ); + + // Column definition + match databricks().verified_stmt("CREATE TABLE foo (x TIMESTAMP_NTZ)") { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ColumnDef { + name: "x".into(), + data_type: DataType::TimestampNtz, + options: vec![], + }] + ); + } + s => panic!("Unexpected statement: {:?}", s), + } +} From 25bb87117583162069e8effd94ddc1b1c45bc476 Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Mon, 31 Mar 2025 17:53:56 +0200 Subject: [PATCH 087/286] Enable double-dot-notation for mssql. (#1787) Co-authored-by: Roman Borschel --- src/dialect/mssql.rs | 5 +++++ tests/sqlparser_mssql.rs | 5 +++++ 2 files changed, 10 insertions(+) diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 3db34748e..18a963a4b 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -101,4 +101,9 @@ impl Dialect for MsSqlDialect { fn supports_nested_comments(&self) -> bool { true } + + /// See + fn supports_object_name_double_dot_notation(&self) -> bool { + true + } } diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index ca59b164e..2bfc38a6a 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1910,6 +1910,11 @@ fn parse_mssql_varbinary_max_length() { } } +#[test] +fn parse_mssql_table_identifier_with_default_schema() { + ms().verified_stmt("SELECT * FROM mydatabase..MyTable"); +} + fn ms() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {})]) } From 45420cedd60f64769a8fe9c8f960b9af4e0091fa Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Mon, 31 Mar 2025 18:09:58 +0200 Subject: [PATCH 088/286] Fix: Snowflake ALTER SESSION cannot be followed by other statements. (#1786) Co-authored-by: Roman Borschel --- src/dialect/snowflake.rs | 15 +++++++++------ tests/sqlparser_snowflake.rs | 11 +++++++++++ 2 files changed, 20 insertions(+), 6 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 2b04f9e9a..d1a696a0b 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -1013,9 +1013,15 @@ fn parse_session_options( let mut options: Vec = Vec::new(); let empty = String::new; loop { - match parser.next_token().token { - Token::Comma => continue, + let next_token = parser.peek_token(); + match next_token.token { + Token::SemiColon | Token::EOF => break, + Token::Comma => { + parser.advance_token(); + continue; + } Token::Word(key) => { + parser.advance_token(); if set { let option = parse_option(parser, key)?; options.push(option); @@ -1028,10 +1034,7 @@ fn parse_session_options( } } _ => { - if parser.peek_token().token == Token::EOF { - break; - } - return parser.expected("another option", parser.peek_token()); + return parser.expected("another option or end of statement", next_token); } } } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index f37b657e5..097af3466 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3505,3 +3505,14 @@ fn test_alter_session() { ); snowflake().one_statement_parses_to("ALTER SESSION UNSET a\nB", "ALTER SESSION UNSET a, B"); } + +#[test] +fn test_alter_session_followed_by_statement() { + let stmts = snowflake() + .parse_sql_statements("ALTER SESSION SET QUERY_TAG='hello'; SELECT 42") + .unwrap(); + match stmts[..] { + [Statement::AlterSession { .. }, Statement::Query { .. }] => {} + _ => panic!("Unexpected statements: {:?}", stmts), + } +} From 776b10afe608a88811b807ab795831d55f186ee3 Mon Sep 17 00:00:00 2001 From: LFC <990479+MichaelScofield@users.noreply.github.com> Date: Wed, 2 Apr 2025 01:06:19 +0800 Subject: [PATCH 089/286] Add GreptimeDB to the "Users" in README (#1788) --- README.md | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index 5e8e460f6..6acfbcef7 100644 --- a/README.md +++ b/README.md @@ -156,7 +156,8 @@ $ cargo run --features json_example --example cli FILENAME.sql [--dialectname] ## Users This parser is currently being used by the [DataFusion] query engine, [LocustDB], -[Ballista], [GlueSQL], [Opteryx], [Polars], [PRQL], [Qrlew], [JumpWire], [ParadeDB] and [CipherStash Proxy]. +[Ballista], [GlueSQL], [Opteryx], [Polars], [PRQL], [Qrlew], [JumpWire], [ParadeDB], [CipherStash Proxy], +and [GreptimeDB]. If your project is using sqlparser-rs feel free to make a PR to add it to this list. @@ -276,3 +277,4 @@ licensed as above, without any additional terms or conditions. [`Dialect`]: https://docs.rs/sqlparser/latest/sqlparser/dialect/trait.Dialect.html [`GenericDialect`]: https://docs.rs/sqlparser/latest/sqlparser/dialect/struct.GenericDialect.html [CipherStash Proxy]: https://github.com/cipherstash/proxy +[GreptimeDB]: https://github.com/GreptimeTeam/greptimedb From 7efa686d7800dbe5a09a25c12980a7086a58d9aa Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 3 Apr 2025 19:52:23 +0200 Subject: [PATCH 090/286] Extend snowflake grant options support (#1794) --- src/ast/mod.rs | 55 ++++++++++++++++++++++++++++++++---- src/keywords.rs | 1 + src/parser/mod.rs | 50 +++++++++++++++++++++++--------- tests/sqlparser_snowflake.rs | 39 ++++++++++++++++++++++++- 4 files changed, 126 insertions(+), 19 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index f187df995..9456991ed 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6079,10 +6079,10 @@ pub enum Action { ManageReleases, ManageVersions, Modify { - modify_type: ActionModifyType, + modify_type: Option, }, Monitor { - monitor_type: ActionMonitorType, + monitor_type: Option, }, Operate, OverrideShareRestrictions, @@ -6115,7 +6115,7 @@ impl fmt::Display for Action { match self { Action::AddSearchOptimization => f.write_str("ADD SEARCH OPTIMIZATION")?, Action::Apply { apply_type } => write!(f, "APPLY {apply_type}")?, - Action::ApplyBudget => f.write_str("APPLY BUDGET")?, + Action::ApplyBudget => f.write_str("APPLYBUDGET")?, Action::AttachListing => f.write_str("ATTACH LISTING")?, Action::AttachPolicy => f.write_str("ATTACH POLICY")?, Action::Audit => f.write_str("AUDIT")?, @@ -6143,8 +6143,18 @@ impl fmt::Display for Action { Action::Manage { manage_type } => write!(f, "MANAGE {manage_type}")?, Action::ManageReleases => f.write_str("MANAGE RELEASES")?, Action::ManageVersions => f.write_str("MANAGE VERSIONS")?, - Action::Modify { modify_type } => write!(f, "MODIFY {modify_type}")?, - Action::Monitor { monitor_type } => write!(f, "MONITOR {monitor_type}")?, + Action::Modify { modify_type } => { + write!(f, "MODIFY")?; + if let Some(modify_type) = modify_type { + write!(f, " {modify_type}")?; + } + } + Action::Monitor { monitor_type } => { + write!(f, "MONITOR")?; + if let Some(monitor_type) = monitor_type { + write!(f, " {monitor_type}")? + } + } Action::Operate => f.write_str("OPERATE")?, Action::OverrideShareRestrictions => f.write_str("OVERRIDE SHARE RESTRICTIONS")?, Action::Ownership => f.write_str("OWNERSHIP")?, @@ -6462,6 +6472,20 @@ pub enum GrantObjects { Warehouses(Vec), /// Grant privileges on specific integrations Integrations(Vec), + /// Grant privileges on resource monitors + ResourceMonitors(Vec), + /// Grant privileges on users + Users(Vec), + /// Grant privileges on compute pools + ComputePools(Vec), + /// Grant privileges on connections + Connections(Vec), + /// Grant privileges on failover groups + FailoverGroup(Vec), + /// Grant privileges on replication group + ReplicationGroup(Vec), + /// Grant privileges on external volumes + ExternalVolumes(Vec), } impl fmt::Display for GrantObjects { @@ -6502,6 +6526,27 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::ResourceMonitors(objects) => { + write!(f, "RESOURCE MONITOR {}", display_comma_separated(objects)) + } + GrantObjects::Users(objects) => { + write!(f, "USER {}", display_comma_separated(objects)) + } + GrantObjects::ComputePools(objects) => { + write!(f, "COMPUTE POOL {}", display_comma_separated(objects)) + } + GrantObjects::Connections(objects) => { + write!(f, "CONNECTION {}", display_comma_separated(objects)) + } + GrantObjects::FailoverGroup(objects) => { + write!(f, "FAILOVER GROUP {}", display_comma_separated(objects)) + } + GrantObjects::ReplicationGroup(objects) => { + write!(f, "REPLICATION GROUP {}", display_comma_separated(objects)) + } + GrantObjects::ExternalVolumes(objects) => { + write!(f, "EXTERNAL VOLUME {}", display_comma_separated(objects)) + } } } } diff --git a/src/keywords.rs b/src/keywords.rs index 8609ec43d..1aa2190c0 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -738,6 +738,7 @@ define_keywords!( REPLICATION, RESET, RESOLVE, + RESOURCE, RESPECT, RESTART, RESTRICT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2b61529ff..40d6b0acd 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12875,6 +12875,26 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllSequencesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[Keyword::RESOURCE, Keyword::MONITOR]) { + Some(GrantObjects::ResourceMonitors(self.parse_comma_separated( + |p| p.parse_object_name_with_wildcards(false, true), + )?)) + } else if self.parse_keywords(&[Keyword::COMPUTE, Keyword::POOL]) { + Some(GrantObjects::ComputePools(self.parse_comma_separated( + |p| p.parse_object_name_with_wildcards(false, true), + )?)) + } else if self.parse_keywords(&[Keyword::FAILOVER, Keyword::GROUP]) { + Some(GrantObjects::FailoverGroup(self.parse_comma_separated( + |p| p.parse_object_name_with_wildcards(false, true), + )?)) + } else if self.parse_keywords(&[Keyword::REPLICATION, Keyword::GROUP]) { + Some(GrantObjects::ReplicationGroup(self.parse_comma_separated( + |p| p.parse_object_name_with_wildcards(false, true), + )?)) + } else if self.parse_keywords(&[Keyword::EXTERNAL, Keyword::VOLUME]) { + Some(GrantObjects::ExternalVolumes(self.parse_comma_separated( + |p| p.parse_object_name_with_wildcards(false, true), + )?)) } else { let object_type = self.parse_one_of_keywords(&[ Keyword::SEQUENCE, @@ -12888,6 +12908,8 @@ impl<'a> Parser<'a> { Keyword::VIEW, Keyword::WAREHOUSE, Keyword::INTEGRATION, + Keyword::USER, + Keyword::CONNECTION, ]); let objects = self.parse_comma_separated(|p| p.parse_object_name_with_wildcards(false, true)); @@ -12898,6 +12920,8 @@ impl<'a> Parser<'a> { Some(Keyword::WAREHOUSE) => Some(GrantObjects::Warehouses(objects?)), Some(Keyword::INTEGRATION) => Some(GrantObjects::Integrations(objects?)), Some(Keyword::VIEW) => Some(GrantObjects::Views(objects?)), + Some(Keyword::USER) => Some(GrantObjects::Users(objects?)), + Some(Keyword::CONNECTION) => Some(GrantObjects::Connections(objects?)), Some(Keyword::TABLE) | None => Some(GrantObjects::Tables(objects?)), _ => unreachable!(), } @@ -12983,10 +13007,10 @@ impl<'a> Parser<'a> { let manage_type = self.parse_action_manage_type()?; Ok(Action::Manage { manage_type }) } else if self.parse_keyword(Keyword::MODIFY) { - let modify_type = self.parse_action_modify_type()?; + let modify_type = self.parse_action_modify_type(); Ok(Action::Modify { modify_type }) } else if self.parse_keyword(Keyword::MONITOR) { - let monitor_type = self.parse_action_monitor_type()?; + let monitor_type = self.parse_action_monitor_type(); Ok(Action::Monitor { monitor_type }) } else if self.parse_keyword(Keyword::OPERATE) { Ok(Action::Operate) @@ -13127,29 +13151,29 @@ impl<'a> Parser<'a> { } } - fn parse_action_modify_type(&mut self) -> Result { + fn parse_action_modify_type(&mut self) -> Option { if self.parse_keywords(&[Keyword::LOG, Keyword::LEVEL]) { - Ok(ActionModifyType::LogLevel) + Some(ActionModifyType::LogLevel) } else if self.parse_keywords(&[Keyword::TRACE, Keyword::LEVEL]) { - Ok(ActionModifyType::TraceLevel) + Some(ActionModifyType::TraceLevel) } else if self.parse_keywords(&[Keyword::SESSION, Keyword::LOG, Keyword::LEVEL]) { - Ok(ActionModifyType::SessionLogLevel) + Some(ActionModifyType::SessionLogLevel) } else if self.parse_keywords(&[Keyword::SESSION, Keyword::TRACE, Keyword::LEVEL]) { - Ok(ActionModifyType::SessionTraceLevel) + Some(ActionModifyType::SessionTraceLevel) } else { - self.expected("GRANT MODIFY type", self.peek_token()) + None } } - fn parse_action_monitor_type(&mut self) -> Result { + fn parse_action_monitor_type(&mut self) -> Option { if self.parse_keyword(Keyword::EXECUTION) { - Ok(ActionMonitorType::Execution) + Some(ActionMonitorType::Execution) } else if self.parse_keyword(Keyword::SECURITY) { - Ok(ActionMonitorType::Security) + Some(ActionMonitorType::Security) } else if self.parse_keyword(Keyword::USAGE) { - Ok(ActionMonitorType::Usage) + Some(ActionMonitorType::Usage) } else { - self.expected("GRANT MONITOR type", self.peek_token()) + None } } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 097af3466..62e52e2d1 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3357,7 +3357,7 @@ fn test_timetravel_at_before() { } #[test] -fn test_grant_account_privileges() { +fn test_grant_account_global_privileges() { let privileges = vec![ "ALL", "ALL PRIVILEGES", @@ -3462,6 +3462,43 @@ fn test_grant_account_privileges() { } } +#[test] +fn test_grant_account_object_privileges() { + let privileges = vec![ + "ALL", + "ALL PRIVILEGES", + "APPLYBUDGET", + "MODIFY", + "MONITOR", + "USAGE", + "OPERATE", + ]; + + let objects_types = vec![ + "USER", + "RESOURCE MONITOR", + "WAREHOUSE", + "COMPUTE POOL", + "DATABASE", + "INTEGRATION", + "CONNECTION", + "FAILOVER GROUP", + "REPLICATION GROUP", + "EXTERNAL VOLUME", + ]; + + let with_grant_options = vec!["", " WITH GRANT OPTION"]; + + for t in &objects_types { + for p in &privileges { + for wgo in &with_grant_options { + let sql = format!("GRANT {p} ON {t} obj1 TO ROLE role1{wgo}"); + snowflake_and_generic().verified_stmt(&sql); + } + } + } +} + #[test] fn test_grant_role_to() { snowflake_and_generic().verified_stmt("GRANT ROLE r1 TO ROLE r2"); From a847e4410572d55a1ff6f6bb01ba34b615f6f043 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Fri, 4 Apr 2025 12:34:18 +0200 Subject: [PATCH 091/286] Fix clippy lint on rust 1.86 (#1796) --- src/ast/ddl.rs | 2 +- src/ast/mod.rs | 12 ++++++------ src/dialect/snowflake.rs | 15 +++++++-------- src/keywords.rs | 12 ++++++------ tests/sqlparser_common.rs | 3 +-- 5 files changed, 21 insertions(+), 23 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 39e43ef15..6a649b73b 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -868,7 +868,7 @@ impl fmt::Display for AlterColumnOperation { AlterColumnOperation::SetDefault { value } => { write!(f, "SET DEFAULT {value}") } - AlterColumnOperation::DropDefault {} => { + AlterColumnOperation::DropDefault => { write!(f, "DROP DEFAULT") } AlterColumnOperation::SetDataType { data_type, using } => { diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 9456991ed..07a22798e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -662,17 +662,17 @@ pub enum Expr { /// such as maps, arrays, and lists: /// - Array /// - A 1-dim array `a[1]` will be represented like: - /// `CompoundFieldAccess(Ident('a'), vec![Subscript(1)]` + /// `CompoundFieldAccess(Ident('a'), vec![Subscript(1)]` /// - A 2-dim array `a[1][2]` will be represented like: - /// `CompoundFieldAccess(Ident('a'), vec![Subscript(1), Subscript(2)]` + /// `CompoundFieldAccess(Ident('a'), vec![Subscript(1), Subscript(2)]` /// - Map or Struct (Bracket-style) /// - A map `a['field1']` will be represented like: - /// `CompoundFieldAccess(Ident('a'), vec![Subscript('field')]` + /// `CompoundFieldAccess(Ident('a'), vec![Subscript('field')]` /// - A 2-dim map `a['field1']['field2']` will be represented like: - /// `CompoundFieldAccess(Ident('a'), vec![Subscript('field1'), Subscript('field2')]` + /// `CompoundFieldAccess(Ident('a'), vec![Subscript('field1'), Subscript('field2')]` /// - Struct (Dot-style) (only effect when the chain contains both subscript and expr) /// - A struct access `a[field1].field2` will be represented like: - /// `CompoundFieldAccess(Ident('a'), vec![Subscript('field1'), Ident('field2')]` + /// `CompoundFieldAccess(Ident('a'), vec![Subscript('field1'), Ident('field2')]` /// - If a struct access likes `a.field1.field2`, it will be represented by CompoundIdentifier([a, field1, field2]) CompoundFieldAccess { root: Box, @@ -7617,7 +7617,7 @@ impl fmt::Display for CopyTarget { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { use CopyTarget::*; match self { - Stdin { .. } => write!(f, "STDIN"), + Stdin => write!(f, "STDIN"), Stdout => write!(f, "STDOUT"), File { filename } => write!(f, "'{}'", value::escape_single_quote_string(filename)), Program { command } => write!( diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index d1a696a0b..f303f8218 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -1038,14 +1038,13 @@ fn parse_session_options( } } } - options - .is_empty() - .then(|| { - Err(ParserError::ParserError( - "expected at least one option".to_string(), - )) - }) - .unwrap_or(Ok(options)) + if options.is_empty() { + Err(ParserError::ParserError( + "expected at least one option".to_string(), + )) + } else { + Ok(options) + } } /// Parses options provided within parentheses like: diff --git a/src/keywords.rs b/src/keywords.rs index 1aa2190c0..bf1206f6a 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -18,14 +18,14 @@ //! This module defines //! 1) a list of constants for every keyword //! 2) an `ALL_KEYWORDS` array with every keyword in it -//! This is not a list of *reserved* keywords: some of these can be -//! parsed as identifiers if the parser decides so. This means that -//! new keywords can be added here without affecting the parse result. +//! This is not a list of *reserved* keywords: some of these can be +//! parsed as identifiers if the parser decides so. This means that +//! new keywords can be added here without affecting the parse result. //! -//! As a matter of fact, most of these keywords are not used at all -//! and could be removed. +//! As a matter of fact, most of these keywords are not used at all +//! and could be removed. //! 3) a `RESERVED_FOR_TABLE_ALIAS` array with keywords reserved in a -//! "table alias" context. +//! "table alias" context. #[cfg(feature = "serde")] use serde::{Deserialize, Serialize}; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 795dae4b3..9fe6eae7a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14091,8 +14091,7 @@ fn test_table_sample() { #[test] fn overflow() { - let expr = std::iter::repeat("1") - .take(1000) + let expr = std::iter::repeat_n("1", 1000) .collect::>() .join(" + "); let sql = format!("SELECT {}", expr); From 3ed4ad9c66eca16b116be63afaa2d9e383bc2bc1 Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Fri, 4 Apr 2025 22:18:31 +0200 Subject: [PATCH 092/286] Allow single quotes in EXTRACT() for Redshift. (#1795) Co-authored-by: Roman Borschel --- src/dialect/redshift.rs | 4 ++++ tests/sqlparser_redshift.rs | 6 ++++++ 2 files changed, 10 insertions(+) diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index 25b8f1644..d90eb6e7d 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -121,4 +121,8 @@ impl Dialect for RedshiftSqlDialect { fn supports_array_typedef_with_brackets(&self) -> bool { true } + + fn allow_extract_single_quotes(&self) -> bool { + true + } } diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index 7736735cb..c75abe16f 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -391,3 +391,9 @@ fn test_parse_nested_quoted_identifier() { .parse_sql_statements(r#"SELECT 1 AS ["1]"#) .is_err()); } + +#[test] +fn parse_extract_single_quotes() { + let sql = "SELECT EXTRACT('month' FROM my_timestamp) FROM my_table"; + redshift().verified_stmt(&sql); +} From 610096cad8bdec90d86dc1f290dbac85cdc4e8ea Mon Sep 17 00:00:00 2001 From: DilovanCelik Date: Sat, 5 Apr 2025 18:37:28 +0000 Subject: [PATCH 093/286] MSSQL: Add support for functionality `MERGE` output clause (#1790) --- src/ast/mod.rs | 39 ++++++++++++++++++++++++++++- src/keywords.rs | 1 + src/parser/mod.rs | 50 ++++++++++++++++++++++++++----------- tests/sqlparser_bigquery.rs | 2 ++ tests/sqlparser_common.rs | 15 +++++++++++ tests/sqlparser_mssql.rs | 16 ++++++++++++ tests/sqlparser_redshift.rs | 2 +- 7 files changed, 109 insertions(+), 16 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 07a22798e..8537bd858 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3817,6 +3817,7 @@ pub enum Statement { /// ``` /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/merge) /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/dml-syntax#merge_statement) + /// [MSSQL](https://learn.microsoft.com/en-us/sql/t-sql/statements/merge-transact-sql?view=sql-server-ver16) Merge { /// optional INTO keyword into: bool, @@ -3828,6 +3829,8 @@ pub enum Statement { on: Box, /// Specifies the actions to perform when values match or do not match. clauses: Vec, + // Specifies the output to save changes in MSSQL + output: Option, }, /// ```sql /// CACHE [ FLAG ] TABLE [ OPTIONS('K1' = 'V1', 'K2' = V2) ] [ AS ] [ ] @@ -5407,6 +5410,7 @@ impl fmt::Display for Statement { source, on, clauses, + output, } => { write!( f, @@ -5414,7 +5418,11 @@ impl fmt::Display for Statement { int = if *into { " INTO" } else { "" } )?; write!(f, "ON {on} ")?; - write!(f, "{}", display_separated(clauses, " ")) + write!(f, "{}", display_separated(clauses, " "))?; + if let Some(output) = output { + write!(f, " {output}")?; + } + Ok(()) } Statement::Cache { table_name, @@ -7945,6 +7953,35 @@ impl Display for MergeClause { } } +/// A Output Clause in the end of a 'MERGE' Statement +/// +/// Example: +/// OUTPUT $action, deleted.* INTO dbo.temp_products; +/// [mssql](https://learn.microsoft.com/en-us/sql/t-sql/queries/output-clause-transact-sql) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct OutputClause { + pub select_items: Vec, + pub into_table: SelectInto, +} + +impl fmt::Display for OutputClause { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let OutputClause { + select_items, + into_table, + } = self; + + write!( + f, + "OUTPUT {} {}", + display_comma_separated(select_items), + into_table + ) + } +} + #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/keywords.rs b/src/keywords.rs index bf1206f6a..a1b4c0c3a 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -632,6 +632,7 @@ define_keywords!( ORGANIZATION, OUT, OUTER, + OUTPUT, OUTPUTFORMAT, OVER, OVERFLOW, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 40d6b0acd..50a9aae84 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10910,18 +10910,7 @@ impl<'a> Parser<'a> { }; let into = if self.parse_keyword(Keyword::INTO) { - let temporary = self - .parse_one_of_keywords(&[Keyword::TEMP, Keyword::TEMPORARY]) - .is_some(); - let unlogged = self.parse_keyword(Keyword::UNLOGGED); - let table = self.parse_keyword(Keyword::TABLE); - let name = self.parse_object_name(false)?; - Some(SelectInto { - temporary, - unlogged, - table, - name, - }) + Some(self.parse_select_into()?) } else { None }; @@ -14513,10 +14502,9 @@ impl<'a> Parser<'a> { pub fn parse_merge_clauses(&mut self) -> Result, ParserError> { let mut clauses = vec![]; loop { - if self.peek_token() == Token::EOF || self.peek_token() == Token::SemiColon { + if !(self.parse_keyword(Keyword::WHEN)) { break; } - self.expect_keyword_is(Keyword::WHEN)?; let mut clause_kind = MergeClauseKind::Matched; if self.parse_keyword(Keyword::NOT) { @@ -14610,6 +14598,34 @@ impl<'a> Parser<'a> { Ok(clauses) } + fn parse_output(&mut self) -> Result { + self.expect_keyword_is(Keyword::OUTPUT)?; + let select_items = self.parse_projection()?; + self.expect_keyword_is(Keyword::INTO)?; + let into_table = self.parse_select_into()?; + + Ok(OutputClause { + select_items, + into_table, + }) + } + + fn parse_select_into(&mut self) -> Result { + let temporary = self + .parse_one_of_keywords(&[Keyword::TEMP, Keyword::TEMPORARY]) + .is_some(); + let unlogged = self.parse_keyword(Keyword::UNLOGGED); + let table = self.parse_keyword(Keyword::TABLE); + let name = self.parse_object_name(false)?; + + Ok(SelectInto { + temporary, + unlogged, + table, + name, + }) + } + pub fn parse_merge(&mut self) -> Result { let into = self.parse_keyword(Keyword::INTO); @@ -14620,6 +14636,11 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::ON)?; let on = self.parse_expr()?; let clauses = self.parse_merge_clauses()?; + let output = if self.peek_keyword(Keyword::OUTPUT) { + Some(self.parse_output()?) + } else { + None + }; Ok(Statement::Merge { into, @@ -14627,6 +14648,7 @@ impl<'a> Parser<'a> { source, on: Box::new(on), clauses, + output, }) } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 3037d4ae5..5eb30d15c 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -1735,6 +1735,7 @@ fn parse_merge() { }, ], }; + match bigquery_and_generic().verified_stmt(sql) { Statement::Merge { into, @@ -1742,6 +1743,7 @@ fn parse_merge() { source, on, clauses, + .. } => { assert!(!into); assert_eq!( diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 9fe6eae7a..365332170 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9359,6 +9359,7 @@ fn parse_merge() { source, on, clauses, + .. }, Statement::Merge { into: no_into, @@ -9366,6 +9367,7 @@ fn parse_merge() { source: source_no_into, on: on_no_into, clauses: clauses_no_into, + .. }, ) => { assert!(into); @@ -9558,6 +9560,19 @@ fn parse_merge() { verified_stmt(sql); } +#[test] +fn test_merge_with_output() { + let sql = "MERGE INTO target_table USING source_table \ + ON target_table.id = source_table.oooid \ + WHEN MATCHED THEN \ + UPDATE SET target_table.description = source_table.description \ + WHEN NOT MATCHED THEN \ + INSERT (ID, description) VALUES (source_table.id, source_table.description) \ + OUTPUT inserted.* INTO log_target"; + + verified_stmt(sql); +} + #[test] fn test_merge_into_using_table() { let sql = "MERGE INTO target_table USING source_table \ diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 2bfc38a6a..5d76fd019 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1921,3 +1921,19 @@ fn ms() -> TestedDialects { fn ms_and_generic() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {}), Box::new(GenericDialect {})]) } + +#[test] +fn parse_mssql_merge_with_output() { + let stmt = "MERGE dso.products AS t \ + USING dsi.products AS \ + s ON s.ProductID = t.ProductID \ + WHEN MATCHED AND \ + NOT (t.ProductName = s.ProductName OR (ISNULL(t.ProductName, s.ProductName) IS NULL)) \ + THEN UPDATE SET t.ProductName = s.ProductName \ + WHEN NOT MATCHED BY TARGET \ + THEN INSERT (ProductID, ProductName) \ + VALUES (s.ProductID, s.ProductName) \ + WHEN NOT MATCHED BY SOURCE THEN DELETE \ + OUTPUT $action, deleted.ProductID INTO dsi.temp_products"; + ms_and_generic().verified_stmt(stmt); +} diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index c75abe16f..060e3853d 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -395,5 +395,5 @@ fn test_parse_nested_quoted_identifier() { #[test] fn parse_extract_single_quotes() { let sql = "SELECT EXTRACT('month' FROM my_timestamp) FROM my_table"; - redshift().verified_stmt(&sql); + redshift().verified_stmt(sql); } From 4deed260061c625d97cde1ac9986697ad2f53cb5 Mon Sep 17 00:00:00 2001 From: Alexander Beedie Date: Sat, 5 Apr 2025 22:41:39 +0400 Subject: [PATCH 094/286] Support additional DuckDB integer types such as HUGEINT, UHUGEINT, etc (#1797) Co-authored-by: Alexander Beedie --- src/ast/data_type.rs | 391 +++++++++++++++++++++----------------- src/ast/ddl.rs | 10 +- src/ast/mod.rs | 30 +-- src/dialect/mod.rs | 2 +- src/keywords.rs | 5 + src/parser/mod.rs | 11 +- tests/sqlparser_duckdb.rs | 26 +++ 7 files changed, 274 insertions(+), 201 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index dc523696a..52919de8a 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -36,7 +36,7 @@ pub enum EnumMember { Name(String), /// ClickHouse allows to specify an integer value for each enum value. /// - /// [clickhouse](https://clickhouse.com/docs/en/sql-reference/data-types/enum) + /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/data-types/enum) NamedValue(String, Expr), } @@ -45,270 +45,289 @@ pub enum EnumMember { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum DataType { - /// Table type in [postgresql]. e.g. CREATE FUNCTION RETURNS TABLE(...) + /// Table type in [PostgreSQL], e.g. CREATE FUNCTION RETURNS TABLE(...). /// - /// [postgresql]: https://www.postgresql.org/docs/15/sql-createfunction.html + /// [PostgreSQL]: https://www.postgresql.org/docs/15/sql-createfunction.html Table(Vec), - /// Fixed-length character type e.g. CHARACTER(10) + /// Fixed-length character type, e.g. CHARACTER(10). Character(Option), - /// Fixed-length char type e.g. CHAR(10) + /// Fixed-length char type, e.g. CHAR(10). Char(Option), - /// Character varying type e.g. CHARACTER VARYING(10) + /// Character varying type, e.g. CHARACTER VARYING(10). CharacterVarying(Option), - /// Char varying type e.g. CHAR VARYING(10) + /// Char varying type, e.g. CHAR VARYING(10). CharVarying(Option), - /// Variable-length character type e.g. VARCHAR(10) + /// Variable-length character type, e.g. VARCHAR(10). Varchar(Option), - /// Variable-length character type e.g. NVARCHAR(10) + /// Variable-length character type, e.g. NVARCHAR(10). Nvarchar(Option), - /// Uuid type + /// Uuid type. Uuid, - /// Large character object with optional length e.g. CHARACTER LARGE OBJECT, CHARACTER LARGE OBJECT(1000), [standard] + /// Large character object with optional length, + /// e.g. CHARACTER LARGE OBJECT, CHARACTER LARGE OBJECT(1000), [SQL Standard]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#character-large-object-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#character-large-object-type CharacterLargeObject(Option), - /// Large character object with optional length e.g. CHAR LARGE OBJECT, CHAR LARGE OBJECT(1000), [standard] + /// Large character object with optional length, + /// e.g. CHAR LARGE OBJECT, CHAR LARGE OBJECT(1000), [SQL Standard]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#character-large-object-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#character-large-object-type CharLargeObject(Option), - /// Large character object with optional length e.g. CLOB, CLOB(1000), [standard] + /// Large character object with optional length, + /// e.g. CLOB, CLOB(1000), [SQL Standard]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#character-large-object-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#character-large-object-type /// [Oracle]: https://docs.oracle.com/javadb/10.10.1.2/ref/rrefclob.html Clob(Option), - /// Fixed-length binary type with optional length e.g. [standard], [MS SQL Server] + /// Fixed-length binary type with optional length, + /// see [SQL Standard], [MS SQL Server]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-string-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-string-type /// [MS SQL Server]: https://learn.microsoft.com/pt-br/sql/t-sql/data-types/binary-and-varbinary-transact-sql?view=sql-server-ver16 Binary(Option), - /// Variable-length binary with optional length type e.g. [standard], [MS SQL Server] + /// Variable-length binary with optional length type, + /// see [SQL Standard], [MS SQL Server]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-string-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-string-type /// [MS SQL Server]: https://learn.microsoft.com/pt-br/sql/t-sql/data-types/binary-and-varbinary-transact-sql?view=sql-server-ver16 Varbinary(Option), - /// Large binary object with optional length e.g. BLOB, BLOB(1000), [standard], [Oracle] + /// Large binary object with optional length, + /// see [SQL Standard], [Oracle]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-large-object-string-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#binary-large-object-string-type /// [Oracle]: https://docs.oracle.com/javadb/10.8.3.0/ref/rrefblob.html Blob(Option), - /// [MySQL] blob with up to 2**8 bytes + /// [MySQL] blob with up to 2**8 bytes. /// /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/blob.html TinyBlob, - /// [MySQL] blob with up to 2**24 bytes + /// [MySQL] blob with up to 2**24 bytes. /// /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/blob.html MediumBlob, - /// [MySQL] blob with up to 2**32 bytes + /// [MySQL] blob with up to 2**32 bytes. /// /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/blob.html LongBlob, /// Variable-length binary data with optional length. /// - /// [bigquery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#bytes_type + /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#bytes_type Bytes(Option), - /// Numeric type with optional precision and scale e.g. NUMERIC(10,2), [standard][1] + /// Numeric type with optional precision and scale, e.g. NUMERIC(10,2), [SQL Standard][1]. /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type Numeric(ExactNumberInfo), - /// Decimal type with optional precision and scale e.g. DECIMAL(10,2), [standard][1] + /// Decimal type with optional precision and scale, e.g. DECIMAL(10,2), [SQL Standard][1]. /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type Decimal(ExactNumberInfo), - /// [BigNumeric] type used in BigQuery + /// [BigNumeric] type used in BigQuery. /// /// [BigNumeric]: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#bignumeric_literals BigNumeric(ExactNumberInfo), - /// This is alias for `BigNumeric` type used in BigQuery + /// This is alias for `BigNumeric` type used in BigQuery. /// /// [BigDecimal]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#decimal_types BigDecimal(ExactNumberInfo), - /// Dec type with optional precision and scale e.g. DEC(10,2), [standard][1] + /// Dec type with optional precision and scale, e.g. DEC(10,2), [SQL Standard][1]. /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type Dec(ExactNumberInfo), - /// Floating point with optional precision e.g. FLOAT(8) + /// Floating point with optional precision, e.g. FLOAT(8). Float(Option), - /// Tiny integer with optional display width e.g. TINYINT or TINYINT(3) + /// Tiny integer with optional display width, e.g. TINYINT or TINYINT(3). TinyInt(Option), - /// Unsigned tiny integer with optional display width e.g. TINYINT UNSIGNED or TINYINT(3) UNSIGNED + /// Unsigned tiny integer with optional display width, + /// e.g. TINYINT UNSIGNED or TINYINT(3) UNSIGNED. TinyIntUnsigned(Option), - /// Int2 as alias for SmallInt in [postgresql] - /// Note: Int2 mean 2 bytes in postgres (not 2 bits) - /// Int2 with optional display width e.g. INT2 or INT2(5) + /// Unsigned tiny integer, e.g. UTINYINT + UTinyInt, + /// Int2 is an alias for SmallInt in [PostgreSQL]. + /// Note: Int2 means 2 bytes in PostgreSQL (not 2 bits). + /// Int2 with optional display width, e.g. INT2 or INT2(5). /// - /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Int2(Option), - /// Unsigned Int2 with optional display width e.g. INT2 UNSIGNED or INT2(5) UNSIGNED + /// Unsigned Int2 with optional display width, e.g. INT2 UNSIGNED or INT2(5) UNSIGNED. Int2Unsigned(Option), - /// Small integer with optional display width e.g. SMALLINT or SMALLINT(5) + /// Small integer with optional display width, e.g. SMALLINT or SMALLINT(5). SmallInt(Option), - /// Unsigned small integer with optional display width e.g. SMALLINT UNSIGNED or SMALLINT(5) UNSIGNED + /// Unsigned small integer with optional display width, + /// e.g. SMALLINT UNSIGNED or SMALLINT(5) UNSIGNED. SmallIntUnsigned(Option), - /// MySQL medium integer ([1]) with optional display width e.g. MEDIUMINT or MEDIUMINT(5) + /// Unsigned small integer, e.g. USMALLINT. + USmallInt, + /// MySQL medium integer ([1]) with optional display width, + /// e.g. MEDIUMINT or MEDIUMINT(5). /// /// [1]: https://dev.mysql.com/doc/refman/8.0/en/integer-types.html MediumInt(Option), - /// Unsigned medium integer ([1]) with optional display width e.g. MEDIUMINT UNSIGNED or MEDIUMINT(5) UNSIGNED + /// Unsigned medium integer ([1]) with optional display width, + /// e.g. MEDIUMINT UNSIGNED or MEDIUMINT(5) UNSIGNED. /// /// [1]: https://dev.mysql.com/doc/refman/8.0/en/integer-types.html MediumIntUnsigned(Option), - /// Int with optional display width e.g. INT or INT(11) + /// Int with optional display width, e.g. INT or INT(11). Int(Option), - /// Int4 as alias for Integer in [postgresql] - /// Note: Int4 mean 4 bytes in postgres (not 4 bits) - /// Int4 with optional display width e.g. Int4 or Int4(11) + /// Int4 is an alias for Integer in [PostgreSQL]. + /// Note: Int4 means 4 bytes in PostgreSQL (not 4 bits). + /// Int4 with optional display width, e.g. Int4 or Int4(11). /// - /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Int4(Option), - /// Int8 as alias for Bigint in [postgresql] and integer type in [clickhouse] - /// Note: Int8 mean 8 bytes in [postgresql] (not 8 bits) - /// Int8 with optional display width e.g. INT8 or INT8(11) - /// Note: Int8 mean 8 bits in [clickhouse] + /// Int8 is an alias for BigInt in [PostgreSQL] and Integer type in [ClickHouse]. + /// Int8 with optional display width, e.g. INT8 or INT8(11). + /// Note: Int8 means 8 bytes in [PostgreSQL], but 8 bits in [ClickHouse]. /// - /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint Int8(Option), - /// Integer type in [clickhouse] - /// Note: Int16 mean 16 bits in [clickhouse] + /// Integer type in [ClickHouse]. + /// Note: Int16 means 16 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint Int16, - /// Integer type in [clickhouse] - /// Note: Int16 mean 32 bits in [clickhouse] + /// Integer type in [ClickHouse]. + /// Note: Int32 means 32 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint Int32, - /// Integer type in [bigquery], [clickhouse] + /// Integer type in [BigQuery], [ClickHouse]. /// - /// [bigquery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#integer_types - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#integer_types + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint Int64, - /// Integer type in [clickhouse] - /// Note: Int128 mean 128 bits in [clickhouse] + /// Integer type in [ClickHouse]. + /// Note: Int128 means 128 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint Int128, - /// Integer type in [clickhouse] - /// Note: Int256 mean 256 bits in [clickhouse] + /// Integer type in [ClickHouse]. + /// Note: Int256 means 256 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint Int256, - /// Integer with optional display width e.g. INTEGER or INTEGER(11) + /// Integer with optional display width, e.g. INTEGER or INTEGER(11). Integer(Option), - /// Unsigned int with optional display width e.g. INT UNSIGNED or INT(11) UNSIGNED + /// Unsigned int with optional display width, e.g. INT UNSIGNED or INT(11) UNSIGNED. IntUnsigned(Option), - /// Unsigned int4 with optional display width e.g. INT4 UNSIGNED or INT4(11) UNSIGNED + /// Unsigned int4 with optional display width, e.g. INT4 UNSIGNED or INT4(11) UNSIGNED. Int4Unsigned(Option), - /// Unsigned integer with optional display width e.g. INTEGER UNSIGNED or INTEGER(11) UNSIGNED + /// Unsigned integer with optional display width, e.g. INTEGER UNSIGNED or INTEGER(11) UNSIGNED. IntegerUnsigned(Option), - /// Unsigned integer type in [clickhouse] - /// Note: UInt8 mean 8 bits in [clickhouse] - /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// 128-bit integer type, e.g. HUGEINT. + HugeInt, + /// Unsigned 128-bit integer type, e.g. UHUGEINT. + UHugeInt, + /// Unsigned integer type in [ClickHouse]. + /// Note: UInt8 means 8 bits in [ClickHouse]. + /// + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint UInt8, - /// Unsigned integer type in [clickhouse] - /// Note: UInt16 mean 16 bits in [clickhouse] + /// Unsigned integer type in [ClickHouse]. + /// Note: UInt16 means 16 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint UInt16, - /// Unsigned integer type in [clickhouse] - /// Note: UInt32 mean 32 bits in [clickhouse] + /// Unsigned integer type in [ClickHouse]. + /// Note: UInt32 means 32 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint UInt32, - /// Unsigned integer type in [clickhouse] - /// Note: UInt64 mean 64 bits in [clickhouse] + /// Unsigned integer type in [ClickHouse]. + /// Note: UInt64 means 64 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint UInt64, - /// Unsigned integer type in [clickhouse] - /// Note: UInt128 mean 128 bits in [clickhouse] + /// Unsigned integer type in [ClickHouse]. + /// Note: UInt128 means 128 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint UInt128, - /// Unsigned integer type in [clickhouse] - /// Note: UInt256 mean 256 bits in [clickhouse] + /// Unsigned integer type in [ClickHouse]. + /// Note: UInt256 means 256 bits in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/int-uint UInt256, - /// Big integer with optional display width e.g. BIGINT or BIGINT(20) + /// Big integer with optional display width, e.g. BIGINT or BIGINT(20). BigInt(Option), - /// Unsigned big integer with optional display width e.g. BIGINT UNSIGNED or BIGINT(20) UNSIGNED + /// Unsigned big integer with optional display width, e.g. BIGINT UNSIGNED or BIGINT(20) UNSIGNED. BigIntUnsigned(Option), - /// Unsigned Int8 with optional display width e.g. INT8 UNSIGNED or INT8(11) UNSIGNED + /// Unsigned big integer, e.g. UBIGINT. + UBigInt, + /// Unsigned Int8 with optional display width, e.g. INT8 UNSIGNED or INT8(11) UNSIGNED. Int8Unsigned(Option), - /// Signed integer as used in [MySQL CAST] target types, without optional `INTEGER` suffix: - /// `SIGNED` + /// Signed integer as used in [MySQL CAST] target types, without optional `INTEGER` suffix, + /// e.g. `SIGNED` /// /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html Signed, - /// Signed integer as used in [MySQL CAST] target types, with optional `INTEGER` suffix: - /// `SIGNED INTEGER` + /// Signed integer as used in [MySQL CAST] target types, with optional `INTEGER` suffix, + /// e.g. `SIGNED INTEGER` /// /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html SignedInteger, - /// Signed integer as used in [MySQL CAST] target types, without optional `INTEGER` suffix: - /// `SIGNED` + /// Signed integer as used in [MySQL CAST] target types, without optional `INTEGER` suffix, + /// e.g. `SIGNED` /// /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html Unsigned, - /// Unsigned integer as used in [MySQL CAST] target types, with optional `INTEGER` suffix: - /// `UNSIGNED INTEGER` + /// Unsigned integer as used in [MySQL CAST] target types, with optional `INTEGER` suffix, + /// e.g. `UNSIGNED INTEGER`. /// /// [MySQL CAST]: https://dev.mysql.com/doc/refman/8.4/en/cast-functions.html UnsignedInteger, - /// Float4 as alias for Real in [postgresql] + /// Float4 is an alias for Real in [PostgreSQL]. /// - /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Float4, - /// Floating point in [clickhouse] + /// Floating point in [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/float + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/float Float32, - /// Floating point in [bigquery] + /// Floating point in [BigQuery]. /// - /// [bigquery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#floating_point_types - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/float + /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#floating_point_types + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/float Float64, - /// Floating point e.g. REAL + /// Floating point, e.g. REAL. Real, - /// Float8 as alias for Double in [postgresql] + /// Float8 is an alias for Double in [PostgreSQL]. /// - /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Float8, /// Double Double(ExactNumberInfo), - /// Double PRECISION e.g. [standard], [postgresql] + /// Double Precision, see [SQL Standard], [PostgreSQL]. /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#approximate-numeric-type - /// [postgresql]: https://www.postgresql.org/docs/current/datatype-numeric.html + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#approximate-numeric-type + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype-numeric.html DoublePrecision, - /// Bool as alias for Boolean in [postgresql] + /// Bool is an alias for Boolean, see [PostgreSQL]. /// - /// [postgresql]: https://www.postgresql.org/docs/15/datatype.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Bool, - /// Boolean + /// Boolean type. Boolean, - /// Date + /// Date type. Date, - /// Date32 with the same range as Datetime64 + /// Date32 with the same range as Datetime64. /// /// [1]: https://clickhouse.com/docs/en/sql-reference/data-types/date32 Date32, - /// Time with optional time precision and time zone information e.g. [standard][1]. + /// Time with optional time precision and time zone information, see [SQL Standard][1]. /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type Time(Option, TimezoneInfo), - /// Datetime with optional time precision e.g. [MySQL][1]. + /// Datetime with optional time precision, see [MySQL][1]. /// /// [1]: https://dev.mysql.com/doc/refman/8.0/en/datetime.html Datetime(Option), - /// Datetime with time precision and optional timezone e.g. [ClickHouse][1]. + /// Datetime with time precision and optional timezone, see [ClickHouse][1]. /// /// [1]: https://clickhouse.com/docs/en/sql-reference/data-types/datetime64 Datetime64(u64, Option), - /// Timestamp with optional time precision and time zone information e.g. [standard][1]. + /// Timestamp with optional time precision and time zone information, see [SQL Standard][1]. /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type Timestamp(Option, TimezoneInfo), @@ -316,25 +335,27 @@ pub enum DataType { /// /// [1]: https://docs.databricks.com/aws/en/sql/language-manual/data-types/timestamp-ntz-type TimestampNtz, - /// Interval + /// Interval type. Interval, - /// JSON type + /// JSON type. JSON, - /// Binary JSON type + /// Binary JSON type. JSONB, - /// Regclass used in postgresql serial + /// Regclass used in [PostgreSQL] serial. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Regclass, - /// Text + /// Text type. Text, - /// [MySQL] text with up to 2**8 bytes + /// [MySQL] text with up to 2**8 bytes. /// /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/blob.html TinyText, - /// [MySQL] text with up to 2**24 bytes + /// [MySQL] text with up to 2**24 bytes. /// /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/blob.html MediumText, - /// [MySQL] text with up to 2**32 bytes + /// [MySQL] text with up to 2**32 bytes. /// /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/blob.html LongText, @@ -344,75 +365,76 @@ pub enum DataType { /// /// [1]: https://clickhouse.com/docs/en/sql-reference/data-types/fixedstring FixedString(u64), - /// Bytea + /// Bytea type, see [PostgreSQL]. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype-bit.html Bytea, - /// Bit string, e.g. [Postgres], [MySQL], or [MSSQL] + /// Bit string, see [PostgreSQL], [MySQL], or [MSSQL]. /// - /// [Postgres]: https://www.postgresql.org/docs/current/datatype-bit.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype-bit.html /// [MySQL]: https://dev.mysql.com/doc/refman/9.1/en/bit-type.html /// [MSSQL]: https://learn.microsoft.com/en-us/sql/t-sql/data-types/bit-transact-sql?view=sql-server-ver16 Bit(Option), - /// `BIT VARYING(n)`: Variable-length bit string e.g. [Postgres] + /// `BIT VARYING(n)`: Variable-length bit string, see [PostgreSQL]. /// - /// [Postgres]: https://www.postgresql.org/docs/current/datatype-bit.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype-bit.html BitVarying(Option), - /// `VARBIT(n)`: Variable-length bit string. [Postgres] alias for `BIT VARYING` + /// `VARBIT(n)`: Variable-length bit string. [PostgreSQL] alias for `BIT VARYING`. /// - /// [Postgres]: https://www.postgresql.org/docs/current/datatype.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html VarBit(Option), - /// - /// Custom type such as enums + /// Custom types. Custom(ObjectName, Vec), - /// Arrays + /// Arrays. Array(ArrayElemTypeDef), - /// Map + /// Map, see [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/map + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/map Map(Box, Box), - /// Tuple + /// Tuple, see [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/tuple + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/tuple Tuple(Vec), - /// Nested + /// Nested type, see [ClickHouse]. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/nested-data-structures/nested + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/nested-data-structures/nested Nested(Vec), - /// Enums + /// Enum type. Enum(Vec, Option), - /// Set + /// Set type. Set(Vec), - /// Struct + /// Struct type, see [Hive], [BigQuery]. /// - /// [hive]: https://docs.cloudera.com/cdw-runtime/cloud/impala-sql-reference/topics/impala-struct.html - /// [bigquery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#struct_type + /// [Hive]: https://docs.cloudera.com/cdw-runtime/cloud/impala-sql-reference/topics/impala-struct.html + /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#struct_type Struct(Vec, StructBracketKind), - /// Union + /// Union type, see [DuckDB]. /// - /// [duckdb]: https://duckdb.org/docs/sql/data_types/union.html + /// [DuckDB]: https://duckdb.org/docs/sql/data_types/union.html Union(Vec), /// Nullable - special marker NULL represents in ClickHouse as a data type. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/nullable + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/nullable Nullable(Box), /// LowCardinality - changes the internal representation of other data types to be dictionary-encoded. /// - /// [clickhouse]: https://clickhouse.com/docs/en/sql-reference/data-types/lowcardinality + /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/lowcardinality LowCardinality(Box), /// No type specified - only used with /// [`SQLiteDialect`](crate::dialect::SQLiteDialect), from statements such /// as `CREATE TABLE t1 (a)`. Unspecified, - /// Trigger data type, returned by functions associated with triggers + /// Trigger data type, returned by functions associated with triggers, see [PostgreSQL]. /// - /// [postgresql]: https://www.postgresql.org/docs/current/plpgsql-trigger.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/plpgsql-trigger.html Trigger, - /// Any data type, used in BigQuery UDF definitions for templated parameters + /// Any data type, used in BigQuery UDF definitions for templated parameters, see [BigQuery]. /// - /// [bigquery]: https://cloud.google.com/bigquery/docs/user-defined-functions#templated-sql-udf-parameters + /// [BigQuery]: https://cloud.google.com/bigquery/docs/user-defined-functions#templated-sql-udf-parameters AnyType, - /// geometric type + /// Geometric type, see [PostgreSQL]. /// - /// [Postgres]: https://www.postgresql.org/docs/9.5/functions-geometry.html + /// [PostgreSQL]: https://www.postgresql.org/docs/9.5/functions-geometry.html GeometricType(GeometricTypeKind), } @@ -503,6 +525,9 @@ impl fmt::Display for DataType { DataType::Int256 => { write!(f, "Int256") } + DataType::HugeInt => { + write!(f, "HUGEINT") + } DataType::Int4Unsigned(zerofill) => { format_type_with_optional_length(f, "INT4", zerofill, true) } @@ -521,6 +546,18 @@ impl fmt::Display for DataType { DataType::Int8Unsigned(zerofill) => { format_type_with_optional_length(f, "INT8", zerofill, true) } + DataType::UTinyInt => { + write!(f, "UTINYINT") + } + DataType::USmallInt => { + write!(f, "USMALLINT") + } + DataType::UBigInt => { + write!(f, "UBIGINT") + } + DataType::UHugeInt => { + write!(f, "UHUGEINT") + } DataType::UInt8 => { write!(f, "UInt8") } @@ -782,19 +819,19 @@ pub enum StructBracketKind { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum TimezoneInfo { - /// No information about time zone. E.g., TIMESTAMP + /// No information about time zone, e.g. TIMESTAMP None, - /// Temporal type 'WITH TIME ZONE'. E.g., TIMESTAMP WITH TIME ZONE, [standard], [Oracle] + /// Temporal type 'WITH TIME ZONE', e.g. TIMESTAMP WITH TIME ZONE, [SQL Standard], [Oracle] /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type /// [Oracle]: https://docs.oracle.com/en/database/oracle/oracle-database/12.2/nlspg/datetime-data-types-and-time-zone-support.html#GUID-3F1C388E-C651-43D5-ADBC-1A49E5C2CA05 WithTimeZone, - /// Temporal type 'WITHOUT TIME ZONE'. E.g., TIME WITHOUT TIME ZONE, [standard], [Postgresql] + /// Temporal type 'WITHOUT TIME ZONE', e.g. TIME WITHOUT TIME ZONE, [SQL Standard], [Postgresql] /// - /// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type + /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#datetime-type /// [Postgresql]: https://www.postgresql.org/docs/current/datatype-datetime.html WithoutTimeZone, - /// Postgresql specific `WITH TIME ZONE` formatting, for both TIME and TIMESTAMP. E.g., TIMETZ, [Postgresql] + /// Postgresql specific `WITH TIME ZONE` formatting, for both TIME and TIMESTAMP, e.g. TIMETZ, [Postgresql] /// /// [Postgresql]: https://www.postgresql.org/docs/current/datatype-datetime.html Tz, @@ -823,18 +860,18 @@ impl fmt::Display for TimezoneInfo { } /// Additional information for `NUMERIC`, `DECIMAL`, and `DEC` data types -/// following the 2016 [standard]. +/// following the 2016 [SQL Standard]. /// -/// [standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type +/// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum ExactNumberInfo { - /// No additional information e.g. `DECIMAL` + /// No additional information, e.g. `DECIMAL` None, - /// Only precision information e.g. `DECIMAL(10)` + /// Only precision information, e.g. `DECIMAL(10)` Precision(u64), - /// Precision and scale information e.g. `DECIMAL(10,2)` + /// Precision and scale information, e.g. `DECIMAL(10,2)` PrecisionAndScale(u64, u64), } @@ -888,7 +925,7 @@ impl fmt::Display for CharacterLength { } } -/// Possible units for characters, initially based on 2016 ANSI [standard][1]. +/// Possible units for characters, initially based on 2016 ANSI [SQL Standard][1]. /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#char-length-units #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -961,7 +998,7 @@ pub enum ArrayElemTypeDef { /// Represents different types of geometric shapes which are commonly used in /// PostgreSQL/Redshift for spatial operations and geometry-related computations. /// -/// [Postgres]: https://www.postgresql.org/docs/9.5/functions-geometry.html +/// [PostgreSQL]: https://www.postgresql.org/docs/9.5/functions-geometry.html #[derive(Debug, Copy, Clone, PartialEq, Eq, Hash, PartialOrd, Ord)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 6a649b73b..000ab3a4f 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1276,9 +1276,9 @@ impl fmt::Display for IndexOption { } } -/// [Postgres] unique index nulls handling option: `[ NULLS [ NOT ] DISTINCT ]` +/// [PostgreSQL] unique index nulls handling option: `[ NULLS [ NOT ] DISTINCT ]` /// -/// [Postgres]: https://www.postgresql.org/docs/17/sql-altertable.html +/// [PostgreSQL]: https://www.postgresql.org/docs/17/sql-altertable.html #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -2175,15 +2175,15 @@ pub struct CreateFunction { /// /// IMMUTABLE | STABLE | VOLATILE /// - /// [Postgres](https://www.postgresql.org/docs/current/sql-createfunction.html) + /// [PostgreSQL](https://www.postgresql.org/docs/current/sql-createfunction.html) pub behavior: Option, /// CALLED ON NULL INPUT | RETURNS NULL ON NULL INPUT | STRICT /// - /// [Postgres](https://www.postgresql.org/docs/current/sql-createfunction.html) + /// [PostgreSQL](https://www.postgresql.org/docs/current/sql-createfunction.html) pub called_on_null: Option, /// PARALLEL { UNSAFE | RESTRICTED | SAFE } /// - /// [Postgres](https://www.postgresql.org/docs/current/sql-createfunction.html) + /// [PostgreSQL](https://www.postgresql.org/docs/current/sql-createfunction.html) pub parallel: Option, /// USING ... (Hive only) pub using: Option, diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8537bd858..a6dc682d5 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -410,7 +410,7 @@ impl fmt::Display for Interval { /// A field definition within a struct /// -/// [bigquery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#struct_type +/// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#struct_type #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -431,7 +431,7 @@ impl fmt::Display for StructField { /// A field definition within a union /// -/// [duckdb]: https://duckdb.org/docs/sql/data_types/union.html +/// [DuckDB]: https://duckdb.org/docs/sql/data_types/union.html #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -448,7 +448,7 @@ impl fmt::Display for UnionField { /// A dictionary field within a dictionary. /// -/// [duckdb]: https://duckdb.org/docs/sql/data_types/struct#creating-structs +/// [DuckDB]: https://duckdb.org/docs/sql/data_types/struct#creating-structs #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -479,7 +479,7 @@ impl Display for Map { /// A map field within a map. /// -/// [duckdb]: https://duckdb.org/docs/sql/data_types/map.html#creating-maps +/// [DuckDB]: https://duckdb.org/docs/sql/data_types/map.html#creating-maps #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -2385,10 +2385,10 @@ impl fmt::Display for DeclareAssignment { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum DeclareType { - /// Cursor variable type. e.g. [Snowflake] [Postgres] + /// Cursor variable type. e.g. [Snowflake] [PostgreSQL] /// /// [Snowflake]: https://docs.snowflake.com/en/developer-guide/snowflake-scripting/cursors#declaring-a-cursor - /// [Postgres]: https://www.postgresql.org/docs/current/plpgsql-cursors.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/plpgsql-cursors.html Cursor, /// Result set variable type. [Snowflake] @@ -2427,7 +2427,7 @@ impl fmt::Display for DeclareType { } /// A `DECLARE` statement. -/// [Postgres] [Snowflake] [BigQuery] +/// [PostgreSQL] [Snowflake] [BigQuery] /// /// Examples: /// ```sql @@ -2435,7 +2435,7 @@ impl fmt::Display for DeclareType { /// DECLARE liahona CURSOR FOR SELECT * FROM films; /// ``` /// -/// [Postgres]: https://www.postgresql.org/docs/current/sql-declare.html +/// [PostgreSQL]: https://www.postgresql.org/docs/current/sql-declare.html /// [Snowflake]: https://docs.snowflake.com/en/sql-reference/snowflake-scripting/declare /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/procedural-language#declare #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -3020,7 +3020,7 @@ pub enum Statement { /// ```sql /// CREATE ROLE /// ``` - /// See [postgres](https://www.postgresql.org/docs/current/sql-createrole.html) + /// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-createrole.html) CreateRole { names: Vec, if_not_exists: bool, @@ -3046,7 +3046,7 @@ pub enum Statement { /// ```sql /// CREATE SECRET /// ``` - /// See [duckdb](https://duckdb.org/docs/sql/statements/create_secret.html) + /// See [DuckDB](https://duckdb.org/docs/sql/statements/create_secret.html) CreateSecret { or_replace: bool, temporary: Option, @@ -3550,7 +3550,7 @@ pub enum Statement { /// /// Supported variants: /// 1. [Hive](https://cwiki.apache.org/confluence/display/hive/languagemanual+ddl#LanguageManualDDL-Create/Drop/ReloadFunction) - /// 2. [Postgres](https://www.postgresql.org/docs/15/sql-createfunction.html) + /// 2. [PostgreSQL](https://www.postgresql.org/docs/15/sql-createfunction.html) /// 3. [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_function_statement) CreateFunction(CreateFunction), /// CREATE TRIGGER @@ -8281,7 +8281,7 @@ impl fmt::Display for FunctionDeterminismSpecifier { /// where within the statement, the body shows up. /// /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#syntax_11 -/// [Postgres]: https://www.postgresql.org/docs/15/sql-createfunction.html +/// [PostgreSQL]: https://www.postgresql.org/docs/15/sql-createfunction.html #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -8319,7 +8319,7 @@ pub enum CreateFunctionBody { /// RETURN a + b; /// ``` /// - /// [Postgres]: https://www.postgresql.org/docs/current/sql-createfunction.html + /// [PostgreSQL]: https://www.postgresql.org/docs/current/sql-createfunction.html Return(Expr), } @@ -8625,9 +8625,9 @@ impl Display for CreateViewParams { } } -/// Engine of DB. Some warehouse has parameters of engine, e.g. [clickhouse] +/// Engine of DB. Some warehouse has parameters of engine, e.g. [ClickHouse] /// -/// [clickhouse]: https://clickhouse.com/docs/en/engines/table-engines +/// [ClickHouse]: https://clickhouse.com/docs/en/engines/table-engines #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 79184dd70..e41964f45 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -995,7 +995,7 @@ pub trait Dialect: Debug + Any { /// Returns true if the dialect supports `SET NAMES [COLLATE ]`. /// /// - [MySQL](https://dev.mysql.com/doc/refman/8.4/en/set-names.html) - /// - [Postgres](https://www.postgresql.org/docs/17/sql-set.html) + /// - [PostgreSQL](https://www.postgresql.org/docs/17/sql-set.html) /// /// Note: Postgres doesn't support the `COLLATE` clause, but we permissively parse it anyway. fn supports_set_names(&self) -> bool { diff --git a/src/keywords.rs b/src/keywords.rs index a1b4c0c3a..73c244261 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -411,6 +411,7 @@ define_keywords!( HOSTS, HOUR, HOURS, + HUGEINT, ICEBERG, ID, IDENTITY, @@ -908,7 +909,9 @@ define_keywords!( TRY_CONVERT, TUPLE, TYPE, + UBIGINT, UESCAPE, + UHUGEINT, UINT128, UINT16, UINT256, @@ -942,6 +945,8 @@ define_keywords!( USER, USER_RESOURCES, USING, + USMALLINT, + UTINYINT, UUID, VACUUM, VALID, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 50a9aae84..549549973 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3779,7 +3779,7 @@ impl<'a> Parser<'a> { }) } - /// Parse a postgresql casting style which is in the form of `expr::datatype`. + /// Parse a PostgreSQL casting style which is in the form of `expr::datatype`. pub fn parse_pg_cast(&mut self, expr: Expr) -> Result { Ok(Expr::Cast { kind: CastKind::DoubleColon, @@ -4873,9 +4873,9 @@ impl<'a> Parser<'a> { } } - /// Parse `CREATE FUNCTION` for [Postgres] + /// Parse `CREATE FUNCTION` for [PostgreSQL] /// - /// [Postgres]: https://www.postgresql.org/docs/15/sql-createfunction.html + /// [PostgreSQL]: https://www.postgresql.org/docs/15/sql-createfunction.html fn parse_postgres_create_function( &mut self, or_replace: bool, @@ -9171,6 +9171,11 @@ impl<'a> Parser<'a> { Ok(DataType::BigInt(optional_precision?)) } } + Keyword::HUGEINT => Ok(DataType::HugeInt), + Keyword::UBIGINT => Ok(DataType::UBigInt), + Keyword::UHUGEINT => Ok(DataType::UHugeInt), + Keyword::USMALLINT => Ok(DataType::USmallInt), + Keyword::UTINYINT => Ok(DataType::UTinyInt), Keyword::UINT8 => Ok(DataType::UInt8), Keyword::UINT16 => Ok(DataType::UInt16), Keyword::UINT32 => Ok(DataType::UInt32), diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index bed02428d..a421154ad 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -352,6 +352,32 @@ fn test_duckdb_load_extension() { ); } +#[test] +fn test_duckdb_specific_int_types() { + let duckdb_dtypes = vec![ + ("UTINYINT", DataType::UTinyInt), + ("USMALLINT", DataType::USmallInt), + ("UBIGINT", DataType::UBigInt), + ("UHUGEINT", DataType::UHugeInt), + ("HUGEINT", DataType::HugeInt), + ]; + for (dtype_string, data_type) in duckdb_dtypes { + let sql = format!("SELECT 123::{}", dtype_string); + let select = duckdb().verified_only_select(&sql); + assert_eq!( + &Expr::Cast { + kind: CastKind::DoubleColon, + expr: Box::new(Expr::Value( + Value::Number("123".parse().unwrap(), false).with_empty_span() + )), + data_type: data_type.clone(), + format: None, + }, + expr_from_projection(&select.projection[0]) + ); + } +} + #[test] fn test_duckdb_struct_literal() { //struct literal syntax https://duckdb.org/docs/sql/data_types/struct#creating-structs From 0d2976d723f083961d5c90ad998073c6f34d2ee0 Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Sun, 6 Apr 2025 07:09:24 +0200 Subject: [PATCH 095/286] Add support for MSSQL IF/ELSE statements. (#1791) Co-authored-by: Roman Borschel --- src/ast/mod.rs | 166 +++++++++++++++++++++++++------------- src/ast/spans.rs | 81 ++++++++++++------- src/dialect/mssql.rs | 128 +++++++++++++++++++++++++++++ src/parser/mod.rs | 68 +++++++++------- tests/sqlparser_common.rs | 108 +++++++++++++++++++++---- tests/sqlparser_mssql.rs | 103 ++++++++++++++++++++++- 6 files changed, 525 insertions(+), 129 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index a6dc682d5..3ee19043e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -37,7 +37,8 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use crate::tokenizer::Span; +use crate::keywords::Keyword; +use crate::tokenizer::{Span, Token}; pub use self::data_type::{ ArrayElemTypeDef, BinaryLength, CharLengthUnits, CharacterLength, DataType, EnumMember, @@ -2118,20 +2119,23 @@ pub enum Password { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct CaseStatement { + /// The `CASE` token that starts the statement. + pub case_token: AttachedToken, pub match_expr: Option, - pub when_blocks: Vec, - pub else_block: Option>, - /// TRUE if the statement ends with `END CASE` (vs `END`). - pub has_end_case: bool, + pub when_blocks: Vec, + pub else_block: Option, + /// The last token of the statement (`END` or `CASE`). + pub end_case_token: AttachedToken, } impl fmt::Display for CaseStatement { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { let CaseStatement { + case_token: _, match_expr, when_blocks, else_block, - has_end_case, + end_case_token: AttachedToken(end), } = self; write!(f, "CASE")?; @@ -2145,13 +2149,15 @@ impl fmt::Display for CaseStatement { } if let Some(else_block) = else_block { - write!(f, " ELSE ")?; - format_statement_list(f, else_block)?; + write!(f, " {else_block}")?; } write!(f, " END")?; - if *has_end_case { - write!(f, " CASE")?; + + if let Token::Word(w) = &end.token { + if w.keyword == Keyword::CASE { + write!(f, " CASE")?; + } } Ok(()) @@ -2160,7 +2166,7 @@ impl fmt::Display for CaseStatement { /// An `IF` statement. /// -/// Examples: +/// Example (BigQuery or Snowflake): /// ```sql /// IF TRUE THEN /// SELECT 1; @@ -2171,16 +2177,22 @@ impl fmt::Display for CaseStatement { /// SELECT 4; /// END IF /// ``` -/// /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/procedural-language#if) /// [Snowflake](https://docs.snowflake.com/en/sql-reference/snowflake-scripting/if) +/// +/// Example (MSSQL): +/// ```sql +/// IF 1=1 SELECT 1 ELSE SELECT 2 +/// ``` +/// [MSSQL](https://learn.microsoft.com/en-us/sql/t-sql/language-elements/if-else-transact-sql?view=sql-server-ver16) #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct IfStatement { - pub if_block: ConditionalStatements, - pub elseif_blocks: Vec, - pub else_block: Option>, + pub if_block: ConditionalStatementBlock, + pub elseif_blocks: Vec, + pub else_block: Option, + pub end_token: Option, } impl fmt::Display for IfStatement { @@ -2189,82 +2201,128 @@ impl fmt::Display for IfStatement { if_block, elseif_blocks, else_block, + end_token, } = self; write!(f, "{if_block}")?; - if !elseif_blocks.is_empty() { - write!(f, " {}", display_separated(elseif_blocks, " "))?; + for elseif_block in elseif_blocks { + write!(f, " {elseif_block}")?; } if let Some(else_block) = else_block { - write!(f, " ELSE ")?; - format_statement_list(f, else_block)?; + write!(f, " {else_block}")?; } - write!(f, " END IF")?; + if let Some(AttachedToken(end_token)) = end_token { + write!(f, " END {end_token}")?; + } Ok(()) } } -/// Represents a type of [ConditionalStatements] -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub enum ConditionalStatementKind { - /// `WHEN THEN ` - When, - /// `IF THEN ` - If, - /// `ELSEIF THEN ` - ElseIf, -} - /// A block within a [Statement::Case] or [Statement::If]-like statement /// -/// Examples: +/// Example 1: /// ```sql /// WHEN EXISTS(SELECT 1) THEN SELECT 1; +/// ``` /// +/// Example 2: +/// ```sql /// IF TRUE THEN SELECT 1; SELECT 2; /// ``` +/// +/// Example 3: +/// ```sql +/// ELSE SELECT 1; SELECT 2; +/// ``` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct ConditionalStatements { - /// The condition expression. - pub condition: Expr, - /// Statement list of the `THEN` clause. - pub statements: Vec, - pub kind: ConditionalStatementKind, +pub struct ConditionalStatementBlock { + pub start_token: AttachedToken, + pub condition: Option, + pub then_token: Option, + pub conditional_statements: ConditionalStatements, } -impl fmt::Display for ConditionalStatements { +impl ConditionalStatementBlock { + pub fn statements(&self) -> &Vec { + self.conditional_statements.statements() + } +} + +impl fmt::Display for ConditionalStatementBlock { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - let ConditionalStatements { - condition: expr, - statements, - kind, + let ConditionalStatementBlock { + start_token: AttachedToken(start_token), + condition, + then_token, + conditional_statements, } = self; - let kind = match kind { - ConditionalStatementKind::When => "WHEN", - ConditionalStatementKind::If => "IF", - ConditionalStatementKind::ElseIf => "ELSEIF", - }; + write!(f, "{start_token}")?; + + if let Some(condition) = condition { + write!(f, " {condition}")?; + } - write!(f, "{kind} {expr} THEN")?; + if then_token.is_some() { + write!(f, " THEN")?; + } - if !statements.is_empty() { - write!(f, " ")?; - format_statement_list(f, statements)?; + if !conditional_statements.statements().is_empty() { + write!(f, " {conditional_statements}")?; } Ok(()) } } +/// A list of statements in a [ConditionalStatementBlock]. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ConditionalStatements { + /// SELECT 1; SELECT 2; SELECT 3; ... + Sequence { statements: Vec }, + /// BEGIN SELECT 1; SELECT 2; SELECT 3; ... END + BeginEnd { + begin_token: AttachedToken, + statements: Vec, + end_token: AttachedToken, + }, +} + +impl ConditionalStatements { + pub fn statements(&self) -> &Vec { + match self { + ConditionalStatements::Sequence { statements } => statements, + ConditionalStatements::BeginEnd { statements, .. } => statements, + } + } +} + +impl fmt::Display for ConditionalStatements { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ConditionalStatements::Sequence { statements } => { + if !statements.is_empty() { + format_statement_list(f, statements)?; + } + Ok(()) + } + ConditionalStatements::BeginEnd { statements, .. } => { + write!(f, "BEGIN ")?; + format_statement_list(f, statements)?; + write!(f, " END") + } + } + } +} + /// A `RAISE` statement. /// /// Examples: diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 11770d1bc..d253f8914 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -22,21 +22,22 @@ use crate::tokenizer::Span; use super::{ dcl::SecondaryRoles, value::ValueWithSpan, AccessExpr, AlterColumnOperation, - AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, CaseStatement, - CloseCursor, ClusteredIndex, ColumnDef, ColumnOption, ColumnOptionDef, ConditionalStatements, - ConflictTarget, ConnectBy, ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, - CreateTableOptions, Cte, Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, - ExprWithAlias, Fetch, FromTable, Function, FunctionArg, FunctionArgExpr, - FunctionArgumentClause, FunctionArgumentList, FunctionArguments, GroupByExpr, HavingBound, - IfStatement, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, - JoinOperator, JsonPath, JsonPathElem, LateralView, LimitClause, MatchRecognizePattern, Measure, - NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, - OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, - Query, RaiseStatement, RaiseStatementValue, ReferentialAction, RenameSelectItem, - ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, - Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, - TableFactor, TableObject, TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, - Value, Values, ViewColumnDef, WildcardAdditionalOptions, With, WithFill, + AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, AttachedToken, + CaseStatement, CloseCursor, ClusteredIndex, ColumnDef, ColumnOption, ColumnOptionDef, + ConditionalStatementBlock, ConditionalStatements, ConflictTarget, ConnectBy, + ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, + Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, + Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, + FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, Insert, Interpolate, + InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, + LimitClause, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, + Offset, OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, + PivotValueSource, ProjectionSelect, Query, RaiseStatement, RaiseStatementValue, + ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, + SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, + TableAliasColumnDef, TableConstraint, TableFactor, TableObject, TableOptionsClustered, + TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, + WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -739,19 +740,14 @@ impl Spanned for CreateIndex { impl Spanned for CaseStatement { fn span(&self) -> Span { let CaseStatement { - match_expr, - when_blocks, - else_block, - has_end_case: _, + case_token: AttachedToken(start), + match_expr: _, + when_blocks: _, + else_block: _, + end_case_token: AttachedToken(end), } = self; - union_spans( - match_expr - .iter() - .map(|e| e.span()) - .chain(when_blocks.iter().map(|b| b.span())) - .chain(else_block.iter().flat_map(|e| e.iter().map(|s| s.span()))), - ) + union_spans([start.span, end.span].into_iter()) } } @@ -761,25 +757,48 @@ impl Spanned for IfStatement { if_block, elseif_blocks, else_block, + end_token, } = self; union_spans( iter::once(if_block.span()) .chain(elseif_blocks.iter().map(|b| b.span())) - .chain(else_block.iter().flat_map(|e| e.iter().map(|s| s.span()))), + .chain(else_block.as_ref().map(|b| b.span())) + .chain(end_token.as_ref().map(|AttachedToken(t)| t.span)), ) } } impl Spanned for ConditionalStatements { fn span(&self) -> Span { - let ConditionalStatements { + match self { + ConditionalStatements::Sequence { statements } => { + union_spans(statements.iter().map(|s| s.span())) + } + ConditionalStatements::BeginEnd { + begin_token: AttachedToken(start), + statements: _, + end_token: AttachedToken(end), + } => union_spans([start.span, end.span].into_iter()), + } + } +} + +impl Spanned for ConditionalStatementBlock { + fn span(&self) -> Span { + let ConditionalStatementBlock { + start_token: AttachedToken(start_token), condition, - statements, - kind: _, + then_token, + conditional_statements, } = self; - union_spans(iter::once(condition.span()).chain(statements.iter().map(|s| s.span()))) + union_spans( + iter::once(start_token.span) + .chain(condition.as_ref().map(|c| c.span())) + .chain(then_token.as_ref().map(|AttachedToken(t)| t.span)) + .chain(iter::once(conditional_statements.span())), + ) } } diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 18a963a4b..d86d68a20 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -15,7 +15,16 @@ // specific language governing permissions and limitations // under the License. +use crate::ast::helpers::attached_token::AttachedToken; +use crate::ast::{ConditionalStatementBlock, ConditionalStatements, IfStatement, Statement}; use crate::dialect::Dialect; +use crate::keywords::{self, Keyword}; +use crate::parser::{Parser, ParserError}; +use crate::tokenizer::Token; +#[cfg(not(feature = "std"))] +use alloc::{vec, vec::Vec}; + +const RESERVED_FOR_COLUMN_ALIAS: &[Keyword] = &[Keyword::IF, Keyword::ELSE]; /// A [`Dialect`] for [Microsoft SQL Server](https://www.microsoft.com/en-us/sql-server/) #[derive(Debug)] @@ -106,4 +115,123 @@ impl Dialect for MsSqlDialect { fn supports_object_name_double_dot_notation(&self) -> bool { true } + + fn is_column_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { + !keywords::RESERVED_FOR_COLUMN_ALIAS.contains(kw) && !RESERVED_FOR_COLUMN_ALIAS.contains(kw) + } + + fn parse_statement(&self, parser: &mut Parser) -> Option> { + if parser.peek_keyword(Keyword::IF) { + Some(self.parse_if_stmt(parser)) + } else { + None + } + } +} + +impl MsSqlDialect { + /// ```sql + /// IF boolean_expression + /// { sql_statement | statement_block } + /// [ ELSE + /// { sql_statement | statement_block } ] + /// ``` + fn parse_if_stmt(&self, parser: &mut Parser) -> Result { + let if_token = parser.expect_keyword(Keyword::IF)?; + + let condition = parser.parse_expr()?; + + let if_block = if parser.peek_keyword(Keyword::BEGIN) { + let begin_token = parser.expect_keyword(Keyword::BEGIN)?; + let statements = self.parse_statement_list(parser, Some(Keyword::END))?; + let end_token = parser.expect_keyword(Keyword::END)?; + ConditionalStatementBlock { + start_token: AttachedToken(if_token), + condition: Some(condition), + then_token: None, + conditional_statements: ConditionalStatements::BeginEnd { + begin_token: AttachedToken(begin_token), + statements, + end_token: AttachedToken(end_token), + }, + } + } else { + let stmt = parser.parse_statement()?; + ConditionalStatementBlock { + start_token: AttachedToken(if_token), + condition: Some(condition), + then_token: None, + conditional_statements: ConditionalStatements::Sequence { + statements: vec![stmt], + }, + } + }; + + while let Token::SemiColon = parser.peek_token_ref().token { + parser.advance_token(); + } + + let mut else_block = None; + if parser.peek_keyword(Keyword::ELSE) { + let else_token = parser.expect_keyword(Keyword::ELSE)?; + if parser.peek_keyword(Keyword::BEGIN) { + let begin_token = parser.expect_keyword(Keyword::BEGIN)?; + let statements = self.parse_statement_list(parser, Some(Keyword::END))?; + let end_token = parser.expect_keyword(Keyword::END)?; + else_block = Some(ConditionalStatementBlock { + start_token: AttachedToken(else_token), + condition: None, + then_token: None, + conditional_statements: ConditionalStatements::BeginEnd { + begin_token: AttachedToken(begin_token), + statements, + end_token: AttachedToken(end_token), + }, + }); + } else { + let stmt = parser.parse_statement()?; + else_block = Some(ConditionalStatementBlock { + start_token: AttachedToken(else_token), + condition: None, + then_token: None, + conditional_statements: ConditionalStatements::Sequence { + statements: vec![stmt], + }, + }); + } + } + + Ok(Statement::If(IfStatement { + if_block, + else_block, + elseif_blocks: Vec::new(), + end_token: None, + })) + } + + /// Parse a sequence of statements, optionally separated by semicolon. + /// + /// Stops parsing when reaching EOF or the given keyword. + fn parse_statement_list( + &self, + parser: &mut Parser, + terminal_keyword: Option, + ) -> Result, ParserError> { + let mut stmts = Vec::new(); + loop { + if let Token::EOF = parser.peek_token_ref().token { + break; + } + if let Some(term) = terminal_keyword { + if parser.peek_keyword(term) { + break; + } + } + stmts.push(parser.parse_statement()?); + while let Token::SemiColon = parser.peek_token_ref().token { + parser.advance_token(); + } + } + Ok(stmts) + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 549549973..b9076bb77 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -631,7 +631,7 @@ impl<'a> Parser<'a> { /// /// See [Statement::Case] pub fn parse_case_stmt(&mut self) -> Result { - self.expect_keyword_is(Keyword::CASE)?; + let case_token = self.expect_keyword(Keyword::CASE)?; let match_expr = if self.peek_keyword(Keyword::WHEN) { None @@ -641,26 +641,26 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::WHEN)?; let when_blocks = self.parse_keyword_separated(Keyword::WHEN, |parser| { - parser.parse_conditional_statements( - ConditionalStatementKind::When, - &[Keyword::WHEN, Keyword::ELSE, Keyword::END], - ) + parser.parse_conditional_statement_block(&[Keyword::WHEN, Keyword::ELSE, Keyword::END]) })?; let else_block = if self.parse_keyword(Keyword::ELSE) { - Some(self.parse_statement_list(&[Keyword::END])?) + Some(self.parse_conditional_statement_block(&[Keyword::END])?) } else { None }; - self.expect_keyword_is(Keyword::END)?; - let has_end_case = self.parse_keyword(Keyword::CASE); + let mut end_case_token = self.expect_keyword(Keyword::END)?; + if self.peek_keyword(Keyword::CASE) { + end_case_token = self.expect_keyword(Keyword::CASE)?; + } Ok(Statement::Case(CaseStatement { + case_token: AttachedToken(case_token), match_expr, when_blocks, else_block, - has_end_case, + end_case_token: AttachedToken(end_case_token), })) } @@ -669,34 +669,38 @@ impl<'a> Parser<'a> { /// See [Statement::If] pub fn parse_if_stmt(&mut self) -> Result { self.expect_keyword_is(Keyword::IF)?; - let if_block = self.parse_conditional_statements( - ConditionalStatementKind::If, - &[Keyword::ELSE, Keyword::ELSEIF, Keyword::END], - )?; + let if_block = self.parse_conditional_statement_block(&[ + Keyword::ELSE, + Keyword::ELSEIF, + Keyword::END, + ])?; let elseif_blocks = if self.parse_keyword(Keyword::ELSEIF) { self.parse_keyword_separated(Keyword::ELSEIF, |parser| { - parser.parse_conditional_statements( - ConditionalStatementKind::ElseIf, - &[Keyword::ELSEIF, Keyword::ELSE, Keyword::END], - ) + parser.parse_conditional_statement_block(&[ + Keyword::ELSEIF, + Keyword::ELSE, + Keyword::END, + ]) })? } else { vec![] }; let else_block = if self.parse_keyword(Keyword::ELSE) { - Some(self.parse_statement_list(&[Keyword::END])?) + Some(self.parse_conditional_statement_block(&[Keyword::END])?) } else { None }; - self.expect_keywords(&[Keyword::END, Keyword::IF])?; + self.expect_keyword_is(Keyword::END)?; + let end_token = self.expect_keyword(Keyword::IF)?; Ok(Statement::If(IfStatement { if_block, elseif_blocks, else_block, + end_token: Some(AttachedToken(end_token)), })) } @@ -707,19 +711,29 @@ impl<'a> Parser<'a> { /// ```sql /// IF condition THEN statement1; statement2; /// ``` - fn parse_conditional_statements( + fn parse_conditional_statement_block( &mut self, - kind: ConditionalStatementKind, terminal_keywords: &[Keyword], - ) -> Result { - let condition = self.parse_expr()?; - self.expect_keyword_is(Keyword::THEN)?; + ) -> Result { + let start_token = self.get_current_token().clone(); // self.expect_keyword(keyword)?; + let mut then_token = None; + + let condition = match &start_token.token { + Token::Word(w) if w.keyword == Keyword::ELSE => None, + _ => { + let expr = self.parse_expr()?; + then_token = Some(AttachedToken(self.expect_keyword(Keyword::THEN)?)); + Some(expr) + } + }; + let statements = self.parse_statement_list(terminal_keywords)?; - Ok(ConditionalStatements { + Ok(ConditionalStatementBlock { + start_token: AttachedToken(start_token), condition, - statements, - kind, + then_token, + conditional_statements: ConditionalStatements::Sequence { statements }, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 365332170..14716dde9 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14229,9 +14229,12 @@ fn parse_case_statement() { }; assert_eq!(Some(Expr::value(number("1"))), stmt.match_expr); - assert_eq!(Expr::value(number("2")), stmt.when_blocks[0].condition); - assert_eq!(2, stmt.when_blocks[0].statements.len()); - assert_eq!(1, stmt.else_block.unwrap().len()); + assert_eq!( + Some(Expr::value(number("2"))), + stmt.when_blocks[0].condition + ); + assert_eq!(2, stmt.when_blocks[0].statements().len()); + assert_eq!(1, stmt.else_block.unwrap().statements().len()); verified_stmt(concat!( "CASE 1", @@ -14274,17 +14277,35 @@ fn parse_case_statement() { ); } +#[test] +fn test_case_statement_span() { + let sql = "CASE 1 WHEN 2 THEN SELECT 1; SELECT 2; ELSE SELECT 3; END CASE"; + let mut parser = Parser::new(&GenericDialect {}).try_with_sql(sql).unwrap(); + assert_eq!( + parser.parse_statement().unwrap().span(), + Span::new(Location::new(1, 1), Location::new(1, sql.len() as u64 + 1)) + ); +} + #[test] fn parse_if_statement() { + let dialects = all_dialects_except(|d| d.is::()); + let sql = "IF 1 THEN SELECT 1; ELSEIF 2 THEN SELECT 2; ELSE SELECT 3; END IF"; - let Statement::If(stmt) = verified_stmt(sql) else { + let Statement::If(IfStatement { + if_block, + elseif_blocks, + else_block, + .. + }) = dialects.verified_stmt(sql) + else { unreachable!() }; - assert_eq!(Expr::value(number("1")), stmt.if_block.condition); - assert_eq!(Expr::value(number("2")), stmt.elseif_blocks[0].condition); - assert_eq!(1, stmt.else_block.unwrap().len()); + assert_eq!(Some(Expr::value(number("1"))), if_block.condition); + assert_eq!(Some(Expr::value(number("2"))), elseif_blocks[0].condition); + assert_eq!(1, else_block.unwrap().statements().len()); - verified_stmt(concat!( + dialects.verified_stmt(concat!( "IF 1 THEN", " SELECT 1;", " SELECT 2;", @@ -14300,7 +14321,7 @@ fn parse_if_statement() { " SELECT 9;", " END IF" )); - verified_stmt(concat!( + dialects.verified_stmt(concat!( "IF 1 THEN", " SELECT 1;", " SELECT 2;", @@ -14309,7 +14330,7 @@ fn parse_if_statement() { " SELECT 4;", " END IF" )); - verified_stmt(concat!( + dialects.verified_stmt(concat!( "IF 1 THEN", " SELECT 1;", " SELECT 2;", @@ -14319,22 +14340,79 @@ fn parse_if_statement() { " SELECT 4;", " END IF" )); - verified_stmt(concat!("IF 1 THEN", " SELECT 1;", " SELECT 2;", " END IF")); - verified_stmt(concat!( + dialects.verified_stmt(concat!("IF 1 THEN", " SELECT 1;", " SELECT 2;", " END IF")); + dialects.verified_stmt(concat!( "IF (1) THEN", " SELECT 1;", " SELECT 2;", " END IF" )); - verified_stmt("IF 1 THEN END IF"); - verified_stmt("IF 1 THEN SELECT 1; ELSEIF 1 THEN END IF"); + dialects.verified_stmt("IF 1 THEN END IF"); + dialects.verified_stmt("IF 1 THEN SELECT 1; ELSEIF 1 THEN END IF"); assert_eq!( ParserError::ParserError("Expected: IF, found: EOF".to_string()), - parse_sql_statements("IF 1 THEN SELECT 1; ELSEIF 1 THEN SELECT 2; END").unwrap_err() + dialects + .parse_sql_statements("IF 1 THEN SELECT 1; ELSEIF 1 THEN SELECT 2; END") + .unwrap_err() ); } +#[test] +fn test_if_statement_span() { + let sql = "IF 1=1 THEN SELECT 1; ELSEIF 1=2 THEN SELECT 2; ELSE SELECT 3; END IF"; + let mut parser = Parser::new(&GenericDialect {}).try_with_sql(sql).unwrap(); + assert_eq!( + parser.parse_statement().unwrap().span(), + Span::new(Location::new(1, 1), Location::new(1, sql.len() as u64 + 1)) + ); +} + +#[test] +fn test_if_statement_multiline_span() { + let sql_line1 = "IF 1 = 1 THEN SELECT 1;"; + let sql_line2 = "ELSEIF 1 = 2 THEN SELECT 2;"; + let sql_line3 = "ELSE SELECT 3;"; + let sql_line4 = "END IF"; + let sql = [sql_line1, sql_line2, sql_line3, sql_line4].join("\n"); + let mut parser = Parser::new(&GenericDialect {}).try_with_sql(&sql).unwrap(); + assert_eq!( + parser.parse_statement().unwrap().span(), + Span::new( + Location::new(1, 1), + Location::new(4, sql_line4.len() as u64 + 1) + ) + ); +} + +#[test] +fn test_conditional_statement_span() { + let sql = "IF 1=1 THEN SELECT 1; ELSEIF 1=2 THEN SELECT 2; ELSE SELECT 3; END IF"; + let mut parser = Parser::new(&GenericDialect {}).try_with_sql(sql).unwrap(); + match parser.parse_statement().unwrap() { + Statement::If(IfStatement { + if_block, + elseif_blocks, + else_block, + .. + }) => { + assert_eq!( + Span::new(Location::new(1, 1), Location::new(1, 21)), + if_block.span() + ); + assert_eq!( + Span::new(Location::new(1, 23), Location::new(1, 47)), + elseif_blocks[0].span() + ); + assert_eq!( + Span::new(Location::new(1, 49), Location::new(1, 62)), + else_block.unwrap().span() + ); + } + stmt => panic!("Unexpected statement: {:?}", stmt), + } +} + #[test] fn parse_raise_statement() { let sql = "RAISE USING MESSAGE = 42"; diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 5d76fd019..bcaf527c0 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -23,7 +23,7 @@ mod test_utils; use helpers::attached_token::AttachedToken; -use sqlparser::tokenizer::Span; +use sqlparser::tokenizer::{Location, Span}; use test_utils::*; use sqlparser::ast::DataType::{Int, Text, Varbinary}; @@ -31,7 +31,7 @@ use sqlparser::ast::DeclareAssignment::MsSqlAssignment; use sqlparser::ast::Value::SingleQuotedString; use sqlparser::ast::*; use sqlparser::dialect::{GenericDialect, MsSqlDialect}; -use sqlparser::parser::ParserError; +use sqlparser::parser::{Parser, ParserError}; #[test] fn parse_mssql_identifiers() { @@ -1857,6 +1857,104 @@ fn parse_mssql_set_session_value() { ms().verified_stmt("SET ANSI_NULLS, ANSI_PADDING ON"); } +#[test] +fn parse_mssql_if_else() { + // Simple statements and blocks + ms().verified_stmt("IF 1 = 1 SELECT '1'; ELSE SELECT '2';"); + ms().verified_stmt("IF 1 = 1 BEGIN SET @A = 1; END ELSE SET @A = 2;"); + ms().verified_stmt( + "IF DATENAME(weekday, GETDATE()) IN (N'Saturday', N'Sunday') SELECT 'Weekend'; ELSE SELECT 'Weekday';" + ); + ms().verified_stmt( + "IF (SELECT COUNT(*) FROM a.b WHERE c LIKE 'x%') > 1 SELECT 'yes'; ELSE SELECT 'No';", + ); + + // Multiple statements + let stmts = ms() + .parse_sql_statements("DECLARE @A INT; IF 1=1 BEGIN SET @A = 1 END ELSE SET @A = 2") + .unwrap(); + match &stmts[..] { + [Statement::Declare { .. }, Statement::If(stmt)] => { + assert_eq!( + stmt.to_string(), + "IF 1 = 1 BEGIN SET @A = 1; END ELSE SET @A = 2;" + ); + } + _ => panic!("Unexpected statements: {:?}", stmts), + } +} + +#[test] +fn test_mssql_if_else_span() { + let sql = "IF 1 = 1 SELECT '1' ELSE SELECT '2'"; + let mut parser = Parser::new(&MsSqlDialect {}).try_with_sql(sql).unwrap(); + assert_eq!( + parser.parse_statement().unwrap().span(), + Span::new(Location::new(1, 1), Location::new(1, sql.len() as u64 + 1)) + ); +} + +#[test] +fn test_mssql_if_else_multiline_span() { + let sql_line1 = "IF 1 = 1"; + let sql_line2 = "SELECT '1'"; + let sql_line3 = "ELSE SELECT '2'"; + let sql = [sql_line1, sql_line2, sql_line3].join("\n"); + let mut parser = Parser::new(&MsSqlDialect {}).try_with_sql(&sql).unwrap(); + assert_eq!( + parser.parse_statement().unwrap().span(), + Span::new( + Location::new(1, 1), + Location::new(3, sql_line3.len() as u64 + 1) + ) + ); +} + +#[test] +fn test_mssql_if_statements_span() { + // Simple statements + let mut sql = "IF 1 = 1 SELECT '1' ELSE SELECT '2'"; + let mut parser = Parser::new(&MsSqlDialect {}).try_with_sql(sql).unwrap(); + match parser.parse_statement().unwrap() { + Statement::If(IfStatement { + if_block, + else_block: Some(else_block), + .. + }) => { + assert_eq!( + if_block.span(), + Span::new(Location::new(1, 1), Location::new(1, 20)) + ); + assert_eq!( + else_block.span(), + Span::new(Location::new(1, 21), Location::new(1, 36)) + ); + } + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Blocks + sql = "IF 1 = 1 BEGIN SET @A = 1; END ELSE BEGIN SET @A = 2 END"; + parser = Parser::new(&MsSqlDialect {}).try_with_sql(sql).unwrap(); + match parser.parse_statement().unwrap() { + Statement::If(IfStatement { + if_block, + else_block: Some(else_block), + .. + }) => { + assert_eq!( + if_block.span(), + Span::new(Location::new(1, 1), Location::new(1, 31)) + ); + assert_eq!( + else_block.span(), + Span::new(Location::new(1, 32), Location::new(1, 57)) + ); + } + stmt => panic!("Unexpected statement: {:?}", stmt), + } +} + #[test] fn parse_mssql_varbinary_max_length() { let sql = "CREATE TABLE example (var_binary_col VARBINARY(MAX))"; @@ -1918,6 +2016,7 @@ fn parse_mssql_table_identifier_with_default_schema() { fn ms() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {})]) } + fn ms_and_generic() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {}), Box::new(GenericDialect {})]) } From cfd8951452b97de7b59afa328a739858e1da6ce3 Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Thu, 10 Apr 2025 06:59:44 +0200 Subject: [PATCH 096/286] Allow literal backslash escapes for string literals in Redshift dialect. (#1801) Co-authored-by: Roman Borschel --- src/dialect/redshift.rs | 4 ++++ tests/sqlparser_redshift.rs | 5 +++++ 2 files changed, 9 insertions(+) diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index d90eb6e7d..feccca5dd 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -125,4 +125,8 @@ impl Dialect for RedshiftSqlDialect { fn allow_extract_single_quotes(&self) -> bool { true } + + fn supports_string_literal_backslash_escape(&self) -> bool { + true + } } diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index 060e3853d..be2b67223 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -397,3 +397,8 @@ fn parse_extract_single_quotes() { let sql = "SELECT EXTRACT('month' FROM my_timestamp) FROM my_table"; redshift().verified_stmt(sql); } + +#[test] +fn parse_string_literal_backslash_escape() { + redshift().one_statement_parses_to(r#"SELECT 'l\'auto'"#, "SELECT 'l''auto'"); +} From 67c3be075ef072387590c4529d8c704c8e8340aa Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Thu, 10 Apr 2025 12:26:13 +0200 Subject: [PATCH 097/286] Add support for MySQL's STRAIGHT_JOIN join operator. (#1802) Co-authored-by: Roman Borschel --- src/ast/query.rs | 7 +++++++ src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 4 ++++ tests/sqlparser_mysql.rs | 7 +++++++ 5 files changed, 20 insertions(+) diff --git a/src/ast/query.rs b/src/ast/query.rs index 1b30dcf17..abc115a0d 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2157,6 +2157,9 @@ impl fmt::Display for Join { self.relation, suffix(constraint) ), + JoinOperator::StraightJoin(constraint) => { + write!(f, " STRAIGHT_JOIN {}{}", self.relation, suffix(constraint)) + } } } } @@ -2197,6 +2200,10 @@ pub enum JoinOperator { match_condition: Expr, constraint: JoinConstraint, }, + /// STRAIGHT_JOIN (non-standard) + /// + /// See . + StraightJoin(JoinConstraint), } #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index d253f8914..9ff83b760 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2128,6 +2128,7 @@ impl Spanned for JoinOperator { } => match_condition.span().union(&constraint.span()), JoinOperator::Anti(join_constraint) => join_constraint.span(), JoinOperator::Semi(join_constraint) => join_constraint.span(), + JoinOperator::StraightJoin(join_constraint) => join_constraint.span(), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 73c244261..0b947b61c 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -840,6 +840,7 @@ define_keywords!( STORAGE_INTEGRATION, STORAGE_SERIALIZATION_POLICY, STORED, + STRAIGHT_JOIN, STRICT, STRING, STRUCT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b9076bb77..0ccf10d7a 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11826,6 +11826,10 @@ impl<'a> Parser<'a> { Keyword::OUTER => { return self.expected("LEFT, RIGHT, or FULL", self.peek_token()); } + Keyword::STRAIGHT_JOIN => { + let _ = self.next_token(); // consume STRAIGHT_JOIN + JoinOperator::StraightJoin + } _ if natural => { return self.expected("a join type after NATURAL", self.peek_token()); } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 1d4fd6a0d..c60936ca8 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3587,3 +3587,10 @@ fn test_variable_assignment_using_colon_equal() { _ => panic!("Unexpected statement {stmt}"), } } + +#[test] +fn parse_straight_join() { + mysql().verified_stmt( + "SELECT a.*, b.* FROM table_a AS a STRAIGHT_JOIN table_b AS b ON a.b_id = b.id", + ); +} From d090ad4ccfd6c76c87de676fae50bbaec36e752a Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 11 Apr 2025 11:49:43 +0200 Subject: [PATCH 098/286] Snowflake COPY INTO target columns, select items and optional alias (#1805) --- src/ast/helpers/stmt_data_loading.rs | 21 +++- src/ast/mod.rs | 14 ++- src/ast/spans.rs | 1 + src/dialect/snowflake.rs | 164 ++++++++++++++------------- tests/sqlparser_snowflake.rs | 36 ++++-- 5 files changed, 146 insertions(+), 90 deletions(-) diff --git a/src/ast/helpers/stmt_data_loading.rs b/src/ast/helpers/stmt_data_loading.rs index cc4fa12fa..e960bb05b 100644 --- a/src/ast/helpers/stmt_data_loading.rs +++ b/src/ast/helpers/stmt_data_loading.rs @@ -29,7 +29,7 @@ use core::fmt; use serde::{Deserialize, Serialize}; use crate::ast::helpers::key_value_options::KeyValueOptions; -use crate::ast::{Ident, ObjectName}; +use crate::ast::{Ident, ObjectName, SelectItem}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; @@ -44,6 +44,25 @@ pub struct StageParamsObject { pub credentials: KeyValueOptions, } +/// This enum enables support for both standard SQL select item expressions +/// and Snowflake-specific ones for data loading. +#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum StageLoadSelectItemKind { + SelectItem(SelectItem), + StageLoadSelectItem(StageLoadSelectItem), +} + +impl fmt::Display for StageLoadSelectItemKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match &self { + StageLoadSelectItemKind::SelectItem(item) => write!(f, "{item}"), + StageLoadSelectItemKind::StageLoadSelectItem(item) => write!(f, "{item}"), + } + } +} + #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 3ee19043e..4031936e9 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -23,7 +23,10 @@ use alloc::{ string::{String, ToString}, vec::Vec, }; -use helpers::{attached_token::AttachedToken, stmt_data_loading::FileStagingCommand}; +use helpers::{ + attached_token::AttachedToken, + stmt_data_loading::{FileStagingCommand, StageLoadSelectItemKind}, +}; use core::ops::Deref; use core::{ @@ -92,7 +95,7 @@ pub use self::value::{ }; use crate::ast::helpers::key_value_options::KeyValueOptions; -use crate::ast::helpers::stmt_data_loading::{StageLoadSelectItem, StageParamsObject}; +use crate::ast::helpers::stmt_data_loading::StageParamsObject; #[cfg(feature = "visitor")] pub use visitor::*; @@ -2988,10 +2991,11 @@ pub enum Statement { CopyIntoSnowflake { kind: CopyIntoSnowflakeKind, into: ObjectName, + into_columns: Option>, from_obj: Option, from_obj_alias: Option, stage_params: StageParamsObject, - from_transformations: Option>, + from_transformations: Option>, from_query: Option>, files: Option>, pattern: Option, @@ -5583,6 +5587,7 @@ impl fmt::Display for Statement { Statement::CopyIntoSnowflake { kind, into, + into_columns, from_obj, from_obj_alias, stage_params, @@ -5596,6 +5601,9 @@ impl fmt::Display for Statement { partition, } => { write!(f, "COPY INTO {}", into)?; + if let Some(into_columns) = into_columns { + write!(f, " ({})", display_comma_separated(into_columns))?; + } if let Some(from_transformations) = from_transformations { // Data load with transformation if let Some(from_stage) = from_obj { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 9ff83b760..31a036b1a 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -350,6 +350,7 @@ impl Spanned for Statement { } => source.span(), Statement::CopyIntoSnowflake { into: _, + into_columns: _, from_obj: _, from_obj_alias: _, stage_params: _, diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index f303f8218..8b279c7cc 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -20,7 +20,7 @@ use crate::alloc::string::ToString; use crate::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionType, KeyValueOptions}; use crate::ast::helpers::stmt_create_table::CreateTableBuilder; use crate::ast::helpers::stmt_data_loading::{ - FileStagingCommand, StageLoadSelectItem, StageParamsObject, + FileStagingCommand, StageLoadSelectItem, StageLoadSelectItemKind, StageParamsObject, }; use crate::ast::{ ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, @@ -30,7 +30,7 @@ use crate::ast::{ }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; -use crate::parser::{Parser, ParserError}; +use crate::parser::{IsOptional, Parser, ParserError}; use crate::tokenizer::{Token, Word}; #[cfg(not(feature = "std"))] use alloc::boxed::Box; @@ -722,7 +722,7 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { }; let mut files: Vec = vec![]; - let mut from_transformations: Option> = None; + let mut from_transformations: Option> = None; let mut from_stage_alias = None; let mut from_stage = None; let mut stage_params = StageParamsObject { @@ -744,6 +744,11 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { stage_params = parse_stage_params(parser)?; } + let into_columns = match &parser.peek_token().token { + Token::LParen => Some(parser.parse_parenthesized_column_list(IsOptional::Optional, true)?), + _ => None, + }; + parser.expect_keyword_is(Keyword::FROM)?; match parser.next_token().token { Token::LParen if kind == CopyIntoSnowflakeKind::Table => { @@ -755,15 +760,10 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { from_stage = Some(parse_snowflake_stage_name(parser)?); stage_params = parse_stage_params(parser)?; - // as - from_stage_alias = if parser.parse_keyword(Keyword::AS) { - Some(match parser.next_token().token { - Token::Word(w) => Ok(Ident::new(w.value)), - _ => parser.expected("stage alias", parser.peek_token()), - }?) - } else { - None - }; + // Parse an optional alias + from_stage_alias = parser + .maybe_parse_table_alias()? + .map(|table_alias| table_alias.name); parser.expect_token(&Token::RParen)?; } Token::LParen if kind == CopyIntoSnowflakeKind::Location => { @@ -846,6 +846,7 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { Ok(Statement::CopyIntoSnowflake { kind, into, + into_columns, from_obj: from_stage, from_obj_alias: from_stage_alias, stage_params, @@ -866,86 +867,93 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { fn parse_select_items_for_data_load( parser: &mut Parser, -) -> Result>, ParserError> { - // [.]$[.] [ , [.]$[.] ... ] - let mut select_items: Vec = vec![]; +) -> Result>, ParserError> { + let mut select_items: Vec = vec![]; loop { - let mut alias: Option = None; - let mut file_col_num: i32 = 0; - let mut element: Option = None; - let mut item_as: Option = None; + match parser.maybe_parse(parse_select_item_for_data_load)? { + // [.]$[.] [ , [.]$[.] ... ] + Some(item) => select_items.push(StageLoadSelectItemKind::StageLoadSelectItem(item)), + // Fallback, try to parse a standard SQL select item + None => select_items.push(StageLoadSelectItemKind::SelectItem( + parser.parse_select_item()?, + )), + } + if matches!(parser.peek_token_ref().token, Token::Comma) { + parser.advance_token(); + } else { + break; + } + } + Ok(Some(select_items)) +} - let next_token = parser.next_token(); - match next_token.token { +fn parse_select_item_for_data_load( + parser: &mut Parser, +) -> Result { + let mut alias: Option = None; + let mut file_col_num: i32 = 0; + let mut element: Option = None; + let mut item_as: Option = None; + + let next_token = parser.next_token(); + match next_token.token { + Token::Placeholder(w) => { + file_col_num = w.to_string().split_off(1).parse::().map_err(|e| { + ParserError::ParserError(format!("Could not parse '{w}' as i32: {e}")) + })?; + Ok(()) + } + Token::Word(w) => { + alias = Some(Ident::new(w.value)); + Ok(()) + } + _ => parser.expected("alias or file_col_num", next_token), + }?; + + if alias.is_some() { + parser.expect_token(&Token::Period)?; + // now we get col_num token + let col_num_token = parser.next_token(); + match col_num_token.token { Token::Placeholder(w) => { file_col_num = w.to_string().split_off(1).parse::().map_err(|e| { ParserError::ParserError(format!("Could not parse '{w}' as i32: {e}")) })?; Ok(()) } - Token::Word(w) => { - alias = Some(Ident::new(w.value)); - Ok(()) - } - _ => parser.expected("alias or file_col_num", next_token), + _ => parser.expected("file_col_num", col_num_token), }?; + } - if alias.is_some() { - parser.expect_token(&Token::Period)?; - // now we get col_num token - let col_num_token = parser.next_token(); - match col_num_token.token { - Token::Placeholder(w) => { - file_col_num = w.to_string().split_off(1).parse::().map_err(|e| { - ParserError::ParserError(format!("Could not parse '{w}' as i32: {e}")) - })?; - Ok(()) - } - _ => parser.expected("file_col_num", col_num_token), - }?; - } - - // try extracting optional element - match parser.next_token().token { - Token::Colon => { - // parse element - element = Some(Ident::new(match parser.next_token().token { - Token::Word(w) => Ok(w.value), - _ => parser.expected("file_col_num", parser.peek_token()), - }?)); - } - _ => { - // element not present move back - parser.prev_token(); - } + // try extracting optional element + match parser.next_token().token { + Token::Colon => { + // parse element + element = Some(Ident::new(match parser.next_token().token { + Token::Word(w) => Ok(w.value), + _ => parser.expected("file_col_num", parser.peek_token()), + }?)); } - - // as - if parser.parse_keyword(Keyword::AS) { - item_as = Some(match parser.next_token().token { - Token::Word(w) => Ok(Ident::new(w.value)), - _ => parser.expected("column item alias", parser.peek_token()), - }?); + _ => { + // element not present move back + parser.prev_token(); } + } - select_items.push(StageLoadSelectItem { - alias, - file_col_num, - element, - item_as, - }); - - match parser.next_token().token { - Token::Comma => { - // continue - } - _ => { - parser.prev_token(); // need to move back - break; - } - } + // as + if parser.parse_keyword(Keyword::AS) { + item_as = Some(match parser.next_token().token { + Token::Word(w) => Ok(Ident::new(w.value)), + _ => parser.expected("column item alias", parser.peek_token()), + }?); } - Ok(Some(select_items)) + + Ok(StageLoadSelectItem { + alias, + file_col_num, + element, + item_as, + }) } fn parse_stage_params(parser: &mut Parser) -> Result { diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 62e52e2d1..4b6694143 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -20,7 +20,7 @@ //! generic dialect is also tested (on the inputs it can handle). use sqlparser::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionType}; -use sqlparser::ast::helpers::stmt_data_loading::StageLoadSelectItem; +use sqlparser::ast::helpers::stmt_data_loading::{StageLoadSelectItem, StageLoadSelectItemKind}; use sqlparser::ast::*; use sqlparser::dialect::{Dialect, GenericDialect, SnowflakeDialect}; use sqlparser::parser::{ParserError, ParserOptions}; @@ -2256,7 +2256,7 @@ fn test_copy_into_with_files_and_pattern_and_verification() { fn test_copy_into_with_transformations() { let sql = concat!( "COPY INTO my_company.emp_basic FROM ", - "(SELECT t1.$1:st AS st, $1:index, t2.$1 FROM @schema.general_finished AS T) ", + "(SELECT t1.$1:st AS st, $1:index, t2.$1, 4, '5' AS const_str FROM @schema.general_finished AS T) ", "FILES = ('file1.json', 'file2.json') ", "PATTERN = '.*employees0[1-5].csv.gz' ", "VALIDATION_MODE = RETURN_7_ROWS" @@ -2277,35 +2277,55 @@ fn test_copy_into_with_transformations() { ); assert_eq!( from_transformations.as_ref().unwrap()[0], - StageLoadSelectItem { + StageLoadSelectItemKind::StageLoadSelectItem(StageLoadSelectItem { alias: Some(Ident::new("t1")), file_col_num: 1, element: Some(Ident::new("st")), item_as: Some(Ident::new("st")) - } + }) ); assert_eq!( from_transformations.as_ref().unwrap()[1], - StageLoadSelectItem { + StageLoadSelectItemKind::StageLoadSelectItem(StageLoadSelectItem { alias: None, file_col_num: 1, element: Some(Ident::new("index")), item_as: None - } + }) ); assert_eq!( from_transformations.as_ref().unwrap()[2], - StageLoadSelectItem { + StageLoadSelectItemKind::StageLoadSelectItem(StageLoadSelectItem { alias: Some(Ident::new("t2")), file_col_num: 1, element: None, item_as: None - } + }) + ); + assert_eq!( + from_transformations.as_ref().unwrap()[3], + StageLoadSelectItemKind::SelectItem(SelectItem::UnnamedExpr(Expr::Value( + Value::Number("4".parse().unwrap(), false).into() + ))) + ); + assert_eq!( + from_transformations.as_ref().unwrap()[4], + StageLoadSelectItemKind::SelectItem(SelectItem::ExprWithAlias { + expr: Expr::Value(Value::SingleQuotedString("5".parse().unwrap()).into()), + alias: Ident::new("const_str".to_string()) + }) ); } _ => unreachable!(), } assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); + + // Test optional AS keyword to denote an alias for the stage + let sql1 = concat!( + "COPY INTO my_company.emp_basic FROM ", + "(SELECT t1.$1:st AS st, $1:index, t2.$1, 4, '5' AS const_str FROM @schema.general_finished T) " + ); + snowflake().parse_sql_statements(sql1).unwrap(); } #[test] From bbc80d7537d31986821d00e2d6c342c5fe1337da Mon Sep 17 00:00:00 2001 From: Roman Borschel Date: Fri, 11 Apr 2025 20:58:43 +0200 Subject: [PATCH 099/286] Fix tokenization of qualified identifiers with numeric prefix. (#1803) Co-authored-by: Roman Borschel --- src/tokenizer.rs | 76 ++++++++++++++++++++---- tests/sqlparser_mysql.rs | 122 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 186 insertions(+), 12 deletions(-) diff --git a/src/tokenizer.rs b/src/tokenizer.rs index d33a7d8af..13bce0c0d 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -895,7 +895,7 @@ impl<'a> Tokenizer<'a> { }; let mut location = state.location(); - while let Some(token) = self.next_token(&mut state)? { + while let Some(token) = self.next_token(&mut state, buf.last().map(|t| &t.token))? { let span = location.span_to(state.location()); buf.push(TokenWithSpan { token, span }); @@ -932,7 +932,11 @@ impl<'a> Tokenizer<'a> { } /// Get the next token or return None - fn next_token(&self, chars: &mut State) -> Result, TokenizerError> { + fn next_token( + &self, + chars: &mut State, + prev_token: Option<&Token>, + ) -> Result, TokenizerError> { match chars.peek() { Some(&ch) => match ch { ' ' => self.consume_and_return(chars, Token::Whitespace(Whitespace::Space)), @@ -1211,17 +1215,29 @@ impl<'a> Tokenizer<'a> { chars.next(); } + // If the dialect supports identifiers that start with a numeric prefix + // and we have now consumed a dot, check if the previous token was a Word. + // If so, what follows is definitely not part of a decimal number and + // we should yield the dot as a dedicated token so compound identifiers + // starting with digits can be parsed correctly. + if s == "." && self.dialect.supports_numeric_prefix() { + if let Some(Token::Word(_)) = prev_token { + return Ok(Some(Token::Period)); + } + } + + // Consume fractional digits. s += &peeking_next_take_while(chars, |ch, next_ch| { ch.is_ascii_digit() || is_number_separator(ch, next_ch) }); - // No number -> Token::Period + // No fraction -> Token::Period if s == "." { return Ok(Some(Token::Period)); } - let mut exponent_part = String::new(); // Parse exponent as number + let mut exponent_part = String::new(); if chars.peek() == Some(&'e') || chars.peek() == Some(&'E') { let mut char_clone = chars.peekable.clone(); exponent_part.push(char_clone.next().unwrap()); @@ -1250,14 +1266,23 @@ impl<'a> Tokenizer<'a> { } } - // mysql dialect supports identifiers that start with a numeric prefix, - // as long as they aren't an exponent number. - if self.dialect.supports_numeric_prefix() && exponent_part.is_empty() { - let word = - peeking_take_while(chars, |ch| self.dialect.is_identifier_part(ch)); - - if !word.is_empty() { - s += word.as_str(); + // If the dialect supports identifiers that start with a numeric prefix, + // we need to check if the value is in fact an identifier and must thus + // be tokenized as a word. + if self.dialect.supports_numeric_prefix() { + if exponent_part.is_empty() { + // If it is not a number with an exponent, it may be + // an identifier starting with digits. + let word = + peeking_take_while(chars, |ch| self.dialect.is_identifier_part(ch)); + + if !word.is_empty() { + s += word.as_str(); + return Ok(Some(Token::make_word(s.as_str(), None))); + } + } else if prev_token == Some(&Token::Period) { + // If the previous token was a period, thus not belonging to a number, + // the value we have is part of an identifier. return Ok(Some(Token::make_word(s.as_str(), None))); } } @@ -3960,4 +3985,31 @@ mod tests { ], ); } + + #[test] + fn test_tokenize_identifiers_numeric_prefix() { + all_dialects_where(|dialect| dialect.supports_numeric_prefix()) + .tokenizes_to("123abc", vec![Token::make_word("123abc", None)]); + + all_dialects_where(|dialect| dialect.supports_numeric_prefix()) + .tokenizes_to("12e34", vec![Token::Number("12e34".to_string(), false)]); + + all_dialects_where(|dialect| dialect.supports_numeric_prefix()).tokenizes_to( + "t.12e34", + vec![ + Token::make_word("t", None), + Token::Period, + Token::make_word("12e34", None), + ], + ); + + all_dialects_where(|dialect| dialect.supports_numeric_prefix()).tokenizes_to( + "t.1two3", + vec![ + Token::make_word("t", None), + Token::Period, + Token::make_word("1two3", None), + ], + ); + } } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index c60936ca8..3a3d8f002 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1926,6 +1926,128 @@ fn parse_select_with_numeric_prefix_column_name() { } } +#[test] +fn parse_qualified_identifiers_with_numeric_prefix() { + // Case 1: Qualified column name that starts with digits. + match mysql().verified_stmt("SELECT t.15to29 FROM my_table AS t") { + Statement::Query(q) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { + assert_eq!(&[Ident::new("t"), Ident::new("15to29")], &parts[..]); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Case 2: Qualified column name that starts with digits and on its own represents a number. + match mysql().verified_stmt("SELECT t.15e29 FROM my_table AS t") { + Statement::Query(q) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { + assert_eq!(&[Ident::new("t"), Ident::new("15e29")], &parts[..]); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Case 3: Unqualified, the same token is parsed as a number. + match mysql() + .parse_sql_statements("SELECT 15e29 FROM my_table") + .unwrap() + .pop() + { + Some(Statement::Query(q)) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::Value(ValueWithSpan { value, .. }))) => { + assert_eq!(&number("15e29"), value); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Case 4: Quoted simple identifier. + match mysql().verified_stmt("SELECT `15e29` FROM my_table") { + Statement::Query(q) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::Identifier(name))) => { + assert_eq!(&Ident::with_quote('`', "15e29"), name); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Case 5: Quoted compound identifier. + match mysql().verified_stmt("SELECT t.`15e29` FROM my_table AS t") { + Statement::Query(q) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { + assert_eq!( + &[Ident::new("t"), Ident::with_quote('`', "15e29")], + &parts[..] + ); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Case 6: Multi-level compound identifiers. + match mysql().verified_stmt("SELECT 1db.1table.1column") { + Statement::Query(q) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { + assert_eq!( + &[ + Ident::new("1db"), + Ident::new("1table"), + Ident::new("1column") + ], + &parts[..] + ); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } + + // Case 7: Multi-level compound quoted identifiers. + match mysql().verified_stmt("SELECT `1`.`2`.`3`") { + Statement::Query(q) => match *q.body { + SetExpr::Select(s) => match s.projection.last() { + Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { + assert_eq!( + &[ + Ident::with_quote('`', "1"), + Ident::with_quote('`', "2"), + Ident::with_quote('`', "3") + ], + &parts[..] + ); + } + proj => panic!("Unexpected projection: {:?}", proj), + }, + body => panic!("Unexpected statement body: {:?}", body), + }, + stmt => panic!("Unexpected statement: {:?}", stmt), + } +} + // Don't run with bigdecimal as it fails like this on rust beta: // // 'parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column' From 896c088153ac340d18d027ea0c56cd89f794146b Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Sat, 12 Apr 2025 18:03:43 +0200 Subject: [PATCH 100/286] Add support for `INHERITS` option in `CREATE TABLE` statement (#1806) --- src/ast/dml.rs | 8 +++++++ src/ast/helpers/stmt_create_table.rs | 11 +++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 13 ++++++++-- tests/sqlparser_duckdb.rs | 1 + tests/sqlparser_mssql.rs | 2 ++ tests/sqlparser_postgres.rs | 36 ++++++++++++++++++++++++++++ 8 files changed, 71 insertions(+), 2 deletions(-) diff --git a/src/ast/dml.rs b/src/ast/dml.rs index ccea7fbcb..9cdb1ca86 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -182,6 +182,11 @@ pub struct CreateTable { /// BigQuery: Table options list. /// pub options: Option>, + /// Postgres `INHERITs` clause, which contains the list of tables from which + /// the new table inherits. + /// + /// + pub inherits: Option>, /// SQLite "STRICT" clause. /// if the "STRICT" table-option keyword is added to the end, after the closing ")", /// then strict typing rules apply to that table. @@ -405,6 +410,9 @@ impl Display for CreateTable { if let Some(order_by) = &self.order_by { write!(f, " ORDER BY {}", order_by)?; } + if let Some(inherits) = &self.inherits { + write!(f, " INHERITS ({})", display_comma_separated(inherits))?; + } if let Some(partition_by) = self.partition_by.as_ref() { write!(f, " PARTITION BY {partition_by}")?; } diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index 344e9dec9..1c50cb843 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -97,6 +97,7 @@ pub struct CreateTableBuilder { pub cluster_by: Option>>, pub clustered_by: Option, pub options: Option>, + pub inherits: Option>, pub strict: bool, pub copy_grants: bool, pub enable_schema_evolution: Option, @@ -151,6 +152,7 @@ impl CreateTableBuilder { cluster_by: None, clustered_by: None, options: None, + inherits: None, strict: false, copy_grants: false, enable_schema_evolution: None, @@ -331,6 +333,11 @@ impl CreateTableBuilder { self } + pub fn inherits(mut self, inherits: Option>) -> Self { + self.inherits = inherits; + self + } + pub fn strict(mut self, strict: bool) -> Self { self.strict = strict; self @@ -451,6 +458,7 @@ impl CreateTableBuilder { cluster_by: self.cluster_by, clustered_by: self.clustered_by, options: self.options, + inherits: self.inherits, strict: self.strict, copy_grants: self.copy_grants, enable_schema_evolution: self.enable_schema_evolution, @@ -512,6 +520,7 @@ impl TryFrom for CreateTableBuilder { cluster_by, clustered_by, options, + inherits, strict, copy_grants, enable_schema_evolution, @@ -560,6 +569,7 @@ impl TryFrom for CreateTableBuilder { cluster_by, clustered_by, options, + inherits, strict, iceberg, copy_grants, @@ -591,6 +601,7 @@ pub(crate) struct CreateTableConfiguration { pub partition_by: Option>, pub cluster_by: Option>>, pub options: Option>, + pub inherits: Option>, } #[cfg(test)] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 31a036b1a..230341518 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -581,6 +581,7 @@ impl Spanned for CreateTable { cluster_by: _, // todo, BigQuery specific clustered_by: _, // todo, Hive specific options: _, // todo, BigQuery specific + inherits: _, // todo, PostgreSQL specific strict: _, // bool copy_grants: _, // bool enable_schema_evolution: _, // bool diff --git a/src/keywords.rs b/src/keywords.rs index 0b947b61c..fb2734096 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -430,6 +430,7 @@ define_keywords!( INDEX, INDICATOR, INHERIT, + INHERITS, INITIALLY, INNER, INOUT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0ccf10d7a..c4bec1d5c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7050,6 +7050,7 @@ impl<'a> Parser<'a> { .partition_by(create_table_config.partition_by) .cluster_by(create_table_config.cluster_by) .options(create_table_config.options) + .inherits(create_table_config.inherits) .primary_key(primary_key) .strict(strict) .build()) @@ -7070,13 +7071,20 @@ impl<'a> Parser<'a> { } } - /// Parse configuration like partitioning, clustering information during the table creation. + /// Parse configuration like inheritance, partitioning, clustering information during the table creation. /// /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#syntax_2) - /// [PostgreSQL](https://www.postgresql.org/docs/current/ddl-partitioning.html) + /// [PostgreSQL Partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html) + /// [PostgreSQL Inheritance](https://www.postgresql.org/docs/current/ddl-inherit.html) fn parse_optional_create_table_config( &mut self, ) -> Result { + let inherits = if self.parse_keyword(Keyword::INHERITS) { + Some(self.parse_parenthesized_qualified_column_list(IsOptional::Mandatory, false)?) + } else { + None + }; + let partition_by = if dialect_of!(self is BigQueryDialect | PostgreSqlDialect | GenericDialect) && self.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) { @@ -7105,6 +7113,7 @@ impl<'a> Parser<'a> { partition_by, cluster_by, options, + inherits, }) } diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index a421154ad..320583248 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -756,6 +756,7 @@ fn test_duckdb_union_datatype() { cluster_by: Default::default(), clustered_by: Default::default(), options: Default::default(), + inherits: Default::default(), strict: Default::default(), copy_grants: Default::default(), enable_schema_evolution: Default::default(), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index bcaf527c0..44fd01f17 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1594,6 +1594,7 @@ fn parse_create_table_with_valid_options() { cluster_by: None, clustered_by: None, options: None, + inherits: None, strict: false, iceberg: false, copy_grants: false, @@ -1764,6 +1765,7 @@ fn parse_create_table_with_identity_column() { cluster_by: None, clustered_by: None, options: None, + inherits: None, strict: false, copy_grants: false, enable_schema_evolution: None, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index a6d65ec75..098d4b1c4 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2733,6 +2733,41 @@ fn parse_create_brin() { } } +#[test] +fn parse_create_table_with_inherits() { + let single_inheritance_sql = + "CREATE TABLE child_table (child_column INT) INHERITS (public.parent_table)"; + match pg().verified_stmt(single_inheritance_sql) { + Statement::CreateTable(CreateTable { + inherits: Some(inherits), + .. + }) => { + assert_eq_vec(&["public", "parent_table"], &inherits[0].0); + } + _ => unreachable!(), + } + + let double_inheritance_sql = "CREATE TABLE child_table (child_column INT) INHERITS (public.parent_table, pg_catalog.pg_settings)"; + match pg().verified_stmt(double_inheritance_sql) { + Statement::CreateTable(CreateTable { + inherits: Some(inherits), + .. + }) => { + assert_eq_vec(&["public", "parent_table"], &inherits[0].0); + assert_eq_vec(&["pg_catalog", "pg_settings"], &inherits[1].0); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_create_table_with_empty_inherits_fails() { + assert!(matches!( + pg().parse_sql_statements("CREATE TABLE child_table (child_column INT) INHERITS ()"), + Err(ParserError::ParserError(_)) + )); +} + #[test] fn parse_create_index_concurrently() { let sql = "CREATE INDEX CONCURRENTLY IF NOT EXISTS my_index ON my_table(col1,col2)"; @@ -5426,6 +5461,7 @@ fn parse_trigger_related_functions() { cluster_by: None, clustered_by: None, options: None, + inherits: None, strict: false, copy_grants: false, enable_schema_evolution: None, From 6566c47593d3dc713e1f25e99adb59dff148494b Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Tue, 15 Apr 2025 01:50:50 -0400 Subject: [PATCH 101/286] Add `DROP TRIGGER` support for SQL Server (#1813) --- src/parser/mod.rs | 2 +- tests/sqlparser_mssql.rs | 15 +++++++++++++++ 2 files changed, 16 insertions(+), 1 deletion(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c4bec1d5c..adf8d86d6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5177,7 +5177,7 @@ impl<'a> Parser<'a> { /// DROP TRIGGER [ IF EXISTS ] name ON table_name [ CASCADE | RESTRICT ] /// ``` pub fn parse_drop_trigger(&mut self) -> Result { - if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect) { + if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect | MsSqlDialect) { self.prev_token(); return self.expected("an object type after DROP", self.peek_token()); } diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 44fd01f17..644589b58 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2038,3 +2038,18 @@ fn parse_mssql_merge_with_output() { OUTPUT $action, deleted.ProductID INTO dsi.temp_products"; ms_and_generic().verified_stmt(stmt); } + +#[test] +fn parse_drop_trigger() { + let sql_drop_trigger = "DROP TRIGGER emp_stamp;"; + let drop_stmt = ms().one_statement_parses_to(sql_drop_trigger, ""); + assert_eq!( + drop_stmt, + Statement::DropTrigger { + if_exists: false, + trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), + table_name: None, + option: None, + } + ); +} From 514d2ecdaf24af52a5a5e7e5227bb1b1bd94f300 Mon Sep 17 00:00:00 2001 From: bar sela Date: Tue, 15 Apr 2025 08:57:26 +0300 Subject: [PATCH 102/286] Snowflake: support nested join without parentheses (#1799) --- src/parser/mod.rs | 32 ++- tests/sqlparser_snowflake.rs | 410 +++++++++++++++++++++++++++++++++++ 2 files changed, 440 insertions(+), 2 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index adf8d86d6..bb989fef1 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11712,6 +11712,11 @@ impl<'a> Parser<'a> { // Note that for keywords to be properly handled here, they need to be // added to `RESERVED_FOR_TABLE_ALIAS`, otherwise they may be parsed as // a table alias. + let joins = self.parse_joins()?; + Ok(TableWithJoins { relation, joins }) + } + + fn parse_joins(&mut self) -> Result, ParserError> { let mut joins = vec![]; loop { let global = self.parse_keyword(Keyword::GLOBAL); @@ -11844,7 +11849,16 @@ impl<'a> Parser<'a> { } _ => break, }; - let relation = self.parse_table_factor()?; + let mut relation = self.parse_table_factor()?; + + if self.peek_parens_less_nested_join() { + let joins = self.parse_joins()?; + relation = TableFactor::NestedJoin { + table_with_joins: Box::new(TableWithJoins { relation, joins }), + alias: None, + }; + } + let join_constraint = self.parse_join_constraint(natural)?; Join { relation, @@ -11854,7 +11868,21 @@ impl<'a> Parser<'a> { }; joins.push(join); } - Ok(TableWithJoins { relation, joins }) + Ok(joins) + } + + fn peek_parens_less_nested_join(&self) -> bool { + matches!( + self.peek_token_ref().token, + Token::Word(Word { + keyword: Keyword::JOIN + | Keyword::INNER + | Keyword::LEFT + | Keyword::RIGHT + | Keyword::FULL, + .. + }) + ) } /// A table name or a parenthesized subquery, followed by optional `[AS] alias` diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 4b6694143..84c08874e 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3573,3 +3573,413 @@ fn test_alter_session_followed_by_statement() { _ => panic!("Unexpected statements: {:?}", stmts), } } + +#[test] +fn test_nested_join_without_parentheses() { + let query = "SELECT DISTINCT p.product_id FROM orders AS o INNER JOIN customers AS c INNER JOIN products AS p ON p.customer_id = c.customer_id ON c.order_id = o.order_id"; + assert_eq!( + only( + snowflake() + .verified_only_select_with_canonical(query, "SELECT DISTINCT p.product_id FROM orders AS o INNER JOIN (customers AS c INNER JOIN products AS p ON p.customer_id = c.customer_id) ON c.order_id = o.order_id") + .from + ) + .joins, + vec![Join { + relation: TableFactor::NestedJoin { + table_with_joins: Box::new(TableWithJoins { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("customers".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "c".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + joins: vec![Join { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("products".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "p".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + global: false, + join_operator: JoinOperator::Inner(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("p".to_string()), + Ident::new("customer_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("customer_id".to_string()) + ])), + })), + }] + }), + alias: None + }, + global: false, + join_operator: JoinOperator::Inner(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("order_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("o".to_string()), + Ident::new("order_id".to_string()) + ])), + })) + }], + ); + + let query = "SELECT DISTINCT p.product_id FROM orders AS o JOIN customers AS c JOIN products AS p ON p.customer_id = c.customer_id ON c.order_id = o.order_id"; + assert_eq!( + only( + snowflake() + .verified_only_select_with_canonical(query, "SELECT DISTINCT p.product_id FROM orders AS o JOIN (customers AS c JOIN products AS p ON p.customer_id = c.customer_id) ON c.order_id = o.order_id") + .from + ) + .joins, + vec![Join { + relation: TableFactor::NestedJoin { + table_with_joins: Box::new(TableWithJoins { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("customers".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "c".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + joins: vec![Join { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("products".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "p".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + global: false, + join_operator: JoinOperator::Join(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("p".to_string()), + Ident::new("customer_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("customer_id".to_string()) + ])), + })), + }] + }), + alias: None + }, + global: false, + join_operator: JoinOperator::Join(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("order_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("o".to_string()), + Ident::new("order_id".to_string()) + ])), + })) + }], + ); + + let query = "SELECT DISTINCT p.product_id FROM orders AS o LEFT JOIN customers AS c LEFT JOIN products AS p ON p.customer_id = c.customer_id ON c.order_id = o.order_id"; + assert_eq!( + only( + snowflake() + .verified_only_select_with_canonical(query, "SELECT DISTINCT p.product_id FROM orders AS o LEFT JOIN (customers AS c LEFT JOIN products AS p ON p.customer_id = c.customer_id) ON c.order_id = o.order_id") + .from + ) + .joins, + vec![Join { + relation: TableFactor::NestedJoin { + table_with_joins: Box::new(TableWithJoins { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("customers".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "c".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + joins: vec![Join { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("products".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "p".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + global: false, + join_operator: JoinOperator::Left(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("p".to_string()), + Ident::new("customer_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("customer_id".to_string()) + ])), + })), + }] + }), + alias: None + }, + global: false, + join_operator: JoinOperator::Left(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("order_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("o".to_string()), + Ident::new("order_id".to_string()) + ])), + })) + }], + ); + + let query = "SELECT DISTINCT p.product_id FROM orders AS o RIGHT JOIN customers AS c RIGHT JOIN products AS p ON p.customer_id = c.customer_id ON c.order_id = o.order_id"; + assert_eq!( + only( + snowflake() + .verified_only_select_with_canonical(query, "SELECT DISTINCT p.product_id FROM orders AS o RIGHT JOIN (customers AS c RIGHT JOIN products AS p ON p.customer_id = c.customer_id) ON c.order_id = o.order_id") + .from + ) + .joins, + vec![Join { + relation: TableFactor::NestedJoin { + table_with_joins: Box::new(TableWithJoins { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("customers".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "c".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + joins: vec![Join { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("products".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "p".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + global: false, + join_operator: JoinOperator::Right(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("p".to_string()), + Ident::new("customer_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("customer_id".to_string()) + ])), + })), + }] + }), + alias: None + }, + global: false, + join_operator: JoinOperator::Right(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("order_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("o".to_string()), + Ident::new("order_id".to_string()) + ])), + })) + }], + ); + + let query = "SELECT DISTINCT p.product_id FROM orders AS o FULL JOIN customers AS c FULL JOIN products AS p ON p.customer_id = c.customer_id ON c.order_id = o.order_id"; + assert_eq!( + only( + snowflake() + .verified_only_select_with_canonical(query, "SELECT DISTINCT p.product_id FROM orders AS o FULL JOIN (customers AS c FULL JOIN products AS p ON p.customer_id = c.customer_id) ON c.order_id = o.order_id") + .from + ) + .joins, + vec![Join { + relation: TableFactor::NestedJoin { + table_with_joins: Box::new(TableWithJoins { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("customers".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "c".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + joins: vec![Join { + relation: TableFactor::Table { + name: ObjectName::from(vec![Ident::new("products".to_string())]), + alias: Some(TableAlias { + name: Ident { + value: "p".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: vec![], + }), + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }, + global: false, + join_operator: JoinOperator::FullOuter(JoinConstraint::On( + Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("p".to_string()), + Ident::new("customer_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("customer_id".to_string()) + ])), + } + )), + }] + }), + alias: None + }, + global: false, + join_operator: JoinOperator::FullOuter(JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("c".to_string()), + Ident::new("order_id".to_string()) + ])), + op: BinaryOperator::Eq, + right: Box::new(Expr::CompoundIdentifier(vec![ + Ident::new("o".to_string()), + Ident::new("order_id".to_string()) + ])), + })) + }], + ); +} From 3ad13af56318479aad6a6542594e0cd4e9040a6e Mon Sep 17 00:00:00 2001 From: Adam Johnson Date: Tue, 15 Apr 2025 18:01:18 +0100 Subject: [PATCH 103/286] Add support for parenthesized subquery as `IN` predicate (#1793) --- src/ast/mod.rs | 2 +- src/parser/mod.rs | 14 ++++++-------- tests/sqlparser_common.rs | 16 +++++++++++++++- 3 files changed, 22 insertions(+), 10 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 4031936e9..72fb99977 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -728,7 +728,7 @@ pub enum Expr { /// `[ NOT ] IN (SELECT ...)` InSubquery { expr: Box, - subquery: Box, + subquery: Box, negated: bool, }, /// `[ NOT ] IN UNNEST(array_expression)` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index bb989fef1..d43e5d02f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3756,15 +3756,13 @@ impl<'a> Parser<'a> { }); } self.expect_token(&Token::LParen)?; - let in_op = if self.parse_keyword(Keyword::SELECT) || self.parse_keyword(Keyword::WITH) { - self.prev_token(); - Expr::InSubquery { + let in_op = match self.maybe_parse(|p| p.parse_query_body(p.dialect.prec_unknown()))? { + Some(subquery) => Expr::InSubquery { expr: Box::new(expr), - subquery: self.parse_query()?, + subquery, negated, - } - } else { - Expr::InList { + }, + None => Expr::InList { expr: Box::new(expr), list: if self.dialect.supports_in_empty_list() { self.parse_comma_separated0(Parser::parse_expr, Token::RParen)? @@ -3772,7 +3770,7 @@ impl<'a> Parser<'a> { self.parse_comma_separated(Parser::parse_expr)? }, negated, - } + }, }; self.expect_token(&Token::RParen)?; Ok(in_op) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 14716dde9..be848a603 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2224,7 +2224,21 @@ fn parse_in_subquery() { assert_eq!( Expr::InSubquery { expr: Box::new(Expr::Identifier(Ident::new("segment"))), - subquery: Box::new(verified_query("SELECT segm FROM bar")), + subquery: verified_query("SELECT segm FROM bar").body, + negated: false, + }, + select.selection.unwrap() + ); +} + +#[test] +fn parse_in_union() { + let sql = "SELECT * FROM customers WHERE segment IN ((SELECT segm FROM bar) UNION (SELECT segm FROM bar2))"; + let select = verified_only_select(sql); + assert_eq!( + Expr::InSubquery { + expr: Box::new(Expr::Identifier(Ident::new("segment"))), + subquery: verified_query("(SELECT segm FROM bar) UNION (SELECT segm FROM bar2)").body, negated: false, }, select.selection.unwrap() From 81d8909e006f90c9b2f06fdeb4bab22935bec827 Mon Sep 17 00:00:00 2001 From: Bruno Clemente Date: Tue, 15 Apr 2025 14:05:20 -0300 Subject: [PATCH 104/286] Fix `STRAIGHT_JOIN` constraint when table alias is absent (#1812) Co-authored-by: Ifeanyi Ubah --- src/dialect/mysql.rs | 7 ++++++- tests/sqlparser_mysql.rs | 3 +++ 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 2077ea195..f69e42436 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -27,7 +27,12 @@ use crate::{ use super::keywords; -const RESERVED_FOR_TABLE_ALIAS_MYSQL: &[Keyword] = &[Keyword::USE, Keyword::IGNORE, Keyword::FORCE]; +const RESERVED_FOR_TABLE_ALIAS_MYSQL: &[Keyword] = &[ + Keyword::USE, + Keyword::IGNORE, + Keyword::FORCE, + Keyword::STRAIGHT_JOIN, +]; /// A [`Dialect`] for [MySQL](https://www.mysql.com/) #[derive(Debug)] diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 3a3d8f002..9d8d12b57 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3715,4 +3715,7 @@ fn parse_straight_join() { mysql().verified_stmt( "SELECT a.*, b.* FROM table_a AS a STRAIGHT_JOIN table_b AS b ON a.b_id = b.id", ); + // Without table alias + mysql() + .verified_stmt("SELECT a.*, b.* FROM table_a STRAIGHT_JOIN table_b AS b ON a.b_id = b.id"); } From 4a487290ce3c5233c25913bb28bf02ddab3b0fdb Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Fri, 18 Apr 2025 02:59:39 -0400 Subject: [PATCH 105/286] Add support for `PRINT` statement for SQL Server (#1811) --- src/ast/mod.rs | 21 ++++++++++++++++++++- src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 8 ++++++++ tests/sqlparser_mssql.rs | 17 +++++++++++++++++ 5 files changed, 47 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 72fb99977..ab3be35c1 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4054,6 +4054,12 @@ pub enum Statement { arguments: Vec, options: Vec, }, + /// ```sql + /// PRINT msg_str | @local_variable | string_expr + /// ``` + /// + /// See: + Print(PrintStatement), } #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -5745,7 +5751,7 @@ impl fmt::Display for Statement { } Ok(()) } - + Statement::Print(s) => write!(f, "{s}"), Statement::List(command) => write!(f, "LIST {command}"), Statement::Remove(command) => write!(f, "REMOVE {command}"), } @@ -9211,6 +9217,19 @@ pub enum CopyIntoSnowflakeKind { Location, } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct PrintStatement { + pub message: Box, +} + +impl fmt::Display for PrintStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "PRINT {}", self.message) + } +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 230341518..a241fdf4d 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -519,6 +519,7 @@ impl Spanned for Statement { Statement::UNLISTEN { .. } => Span::empty(), Statement::RenameTable { .. } => Span::empty(), Statement::RaisError { .. } => Span::empty(), + Statement::Print { .. } => Span::empty(), Statement::List(..) | Statement::Remove(..) => Span::empty(), } } diff --git a/src/keywords.rs b/src/keywords.rs index fb2734096..a5400a5b0 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -686,6 +686,7 @@ define_keywords!( PRESERVE, PREWHERE, PRIMARY, + PRINT, PRIOR, PRIVILEGES, PROCEDURE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d43e5d02f..a9ddd1837 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -617,6 +617,7 @@ impl<'a> Parser<'a> { } // `COMMENT` is snowflake specific https://docs.snowflake.com/en/sql-reference/sql/comment Keyword::COMMENT if self.dialect.supports_comment_on() => self.parse_comment(), + Keyword::PRINT => self.parse_print(), _ => self.expected("an SQL statement", next_token), }, Token::LParen => { @@ -15056,6 +15057,13 @@ impl<'a> Parser<'a> { } } + /// Parse [Statement::Print] + fn parse_print(&mut self) -> Result { + Ok(Statement::Print(PrintStatement { + message: Box::new(self.parse_expr()?), + })) + } + /// Consume the parser and return its underlying token buffer pub fn into_tokens(self) -> Vec { self.tokens diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 644589b58..2786384b3 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2053,3 +2053,20 @@ fn parse_drop_trigger() { } ); } + +#[test] +fn parse_print() { + let print_string_literal = "PRINT 'Hello, world!'"; + let print_stmt = ms().verified_stmt(print_string_literal); + assert_eq!( + print_stmt, + Statement::Print(PrintStatement { + message: Box::new(Expr::Value( + (Value::SingleQuotedString("Hello, world!".to_string())).with_empty_span() + )), + }) + ); + + let _ = ms().verified_stmt("PRINT N'Hello, ⛄️!'"); + let _ = ms().verified_stmt("PRINT @my_variable"); +} From 3ec80e187d163c4f90c5bfc7c04ef71a2705a631 Mon Sep 17 00:00:00 2001 From: Jax Liu Date: Sat, 19 Apr 2025 19:14:45 +0800 Subject: [PATCH 106/286] enable `supports_filter_during_aggregation` for Generic dialect (#1815) --- src/dialect/generic.rs | 4 ++++ tests/sqlparser_common.rs | 18 +++++++++++++++--- 2 files changed, 19 insertions(+), 3 deletions(-) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 92cfca8fd..8f57e487f 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -163,4 +163,8 @@ impl Dialect for GenericDialect { fn supports_comma_separated_set_assignments(&self) -> bool { true } + + fn supports_filter_during_aggregation(&self) -> bool { + true + } } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index be848a603..1c03a0fa4 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11642,6 +11642,20 @@ fn parse_connect_by() { #[test] fn test_selective_aggregation() { + let testing_dialects = all_dialects_where(|d| d.supports_filter_during_aggregation()); + let expected_dialects: Vec> = vec![ + Box::new(PostgreSqlDialect {}), + Box::new(DatabricksDialect {}), + Box::new(HiveDialect {}), + Box::new(SQLiteDialect {}), + Box::new(DuckDbDialect {}), + Box::new(GenericDialect {}), + ]; + assert_eq!(testing_dialects.dialects.len(), expected_dialects.len()); + expected_dialects + .into_iter() + .for_each(|d| assert!(d.supports_filter_during_aggregation())); + let sql = concat!( "SELECT ", "ARRAY_AGG(name) FILTER (WHERE name IS NOT NULL), ", @@ -11649,9 +11663,7 @@ fn test_selective_aggregation() { "FROM region" ); assert_eq!( - all_dialects_where(|d| d.supports_filter_during_aggregation()) - .verified_only_select(sql) - .projection, + testing_dialects.verified_only_select(sql).projection, vec![ SelectItem::UnnamedExpr(Expr::Function(Function { name: ObjectName::from(vec![Ident::new("ARRAY_AGG")]), From 945f8e0534401556657d35bfcb9a3370ddcc9b7c Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Wed, 23 Apr 2025 18:03:06 +0200 Subject: [PATCH 107/286] Add support for `XMLTABLE` (#1817) --- src/ast/mod.rs | 3 +- src/ast/query.rs | 186 ++++++++++++++++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 3 + src/parser/mod.rs | 97 ++++++++++++++++++++ tests/sqlparser_common.rs | 38 ++++++++ 6 files changed, 327 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index ab3be35c1..74e8cb55c 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -81,7 +81,8 @@ pub use self::query::{ TableSampleBucket, TableSampleKind, TableSampleMethod, TableSampleModifier, TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, UpdateTableFromKind, ValueTableMode, Values, - WildcardAdditionalOptions, With, WithFill, + WildcardAdditionalOptions, With, WithFill, XmlNamespaceDefinition, XmlPassingArgument, + XmlPassingClause, XmlTableColumn, XmlTableColumnOption, }; pub use self::trigger::{ diff --git a/src/ast/query.rs b/src/ast/query.rs index abc115a0d..982985ec3 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1271,6 +1271,37 @@ pub enum TableFactor { symbols: Vec, alias: Option, }, + /// The `XMLTABLE` table-valued function. + /// Part of the SQL standard, supported by PostgreSQL, Oracle, and DB2. + /// + /// + /// + /// ```sql + /// SELECT xmltable.* + /// FROM xmldata, + /// XMLTABLE('//ROWS/ROW' + /// PASSING data + /// COLUMNS id int PATH '@id', + /// ordinality FOR ORDINALITY, + /// "COUNTRY_NAME" text, + /// country_id text PATH 'COUNTRY_ID', + /// size_sq_km float PATH 'SIZE[@unit = "sq_km"]', + /// size_other text PATH 'concat(SIZE[@unit!="sq_km"], " ", SIZE[@unit!="sq_km"]/@unit)', + /// premier_name text PATH 'PREMIER_NAME' DEFAULT 'not specified' + /// ); + /// ```` + XmlTable { + /// Optional XMLNAMESPACES clause (empty if not present) + namespaces: Vec, + /// The row-generating XPath expression. + row_expression: Expr, + /// The PASSING clause specifying the document expression. + passing: XmlPassingClause, + /// The columns to be extracted from each generated row. + columns: Vec, + /// The alias for the table. + alias: Option, + }, } /// The table sample modifier options @@ -1936,6 +1967,31 @@ impl fmt::Display for TableFactor { } Ok(()) } + TableFactor::XmlTable { + row_expression, + passing, + columns, + alias, + namespaces, + } => { + write!(f, "XMLTABLE(")?; + if !namespaces.is_empty() { + write!( + f, + "XMLNAMESPACES({}), ", + display_comma_separated(namespaces) + )?; + } + write!( + f, + "{row_expression}{passing} COLUMNS {columns})", + columns = display_comma_separated(columns) + )?; + if let Some(alias) = alias { + write!(f, " AS {alias}")?; + } + Ok(()) + } } } } @@ -3082,3 +3138,133 @@ pub enum UpdateTableFromKind { /// For Example: `UPDATE SET t1.name='aaa' FROM t1` AfterSet(Vec), } + +/// Defines the options for an XmlTable column: Named or ForOrdinality +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +pub enum XmlTableColumnOption { + /// A named column with a type, optional path, and default value. + NamedInfo { + /// The type of the column to be extracted. + r#type: DataType, + /// The path to the column to be extracted. If None, defaults to the column name. + path: Option, + /// Default value if path does not match + default: Option, + /// Whether the column is nullable (NULL=true, NOT NULL=false) + nullable: bool, + }, + /// The FOR ORDINALITY marker + ForOrdinality, +} + +/// A single column definition in XMLTABLE +/// +/// ```sql +/// COLUMNS +/// id int PATH '@id', +/// ordinality FOR ORDINALITY, +/// "COUNTRY_NAME" text, +/// country_id text PATH 'COUNTRY_ID', +/// size_sq_km float PATH 'SIZE[@unit = "sq_km"]', +/// size_other text PATH 'concat(SIZE[@unit!="sq_km"], " ", SIZE[@unit!="sq_km"]/@unit)', +/// premier_name text PATH 'PREMIER_NAME' DEFAULT 'not specified' +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +pub struct XmlTableColumn { + /// The name of the column. + pub name: Ident, + /// Column options: type/path/default or FOR ORDINALITY + pub option: XmlTableColumnOption, +} + +impl fmt::Display for XmlTableColumn { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!(f, "{}", self.name)?; + match &self.option { + XmlTableColumnOption::NamedInfo { + r#type, + path, + default, + nullable, + } => { + write!(f, " {}", r#type)?; + if let Some(p) = path { + write!(f, " PATH {}", p)?; + } + if let Some(d) = default { + write!(f, " DEFAULT {}", d)?; + } + if !*nullable { + write!(f, " NOT NULL")?; + } + Ok(()) + } + XmlTableColumnOption::ForOrdinality => { + write!(f, " FOR ORDINALITY") + } + } + } +} + +/// Argument passed in the XMLTABLE PASSING clause +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +pub struct XmlPassingArgument { + pub expr: Expr, + pub alias: Option, + pub by_value: bool, // True if BY VALUE is specified +} + +impl fmt::Display for XmlPassingArgument { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + if self.by_value { + write!(f, "BY VALUE ")?; + } + write!(f, "{}", self.expr)?; + if let Some(alias) = &self.alias { + write!(f, " AS {}", alias)?; + } + Ok(()) + } +} + +/// The PASSING clause for XMLTABLE +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +pub struct XmlPassingClause { + pub arguments: Vec, +} + +impl fmt::Display for XmlPassingClause { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + if !self.arguments.is_empty() { + write!(f, " PASSING {}", display_comma_separated(&self.arguments))?; + } + Ok(()) + } +} + +/// Represents a single XML namespace definition in the XMLNAMESPACES clause. +/// +/// `namespace_uri AS namespace_name` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +pub struct XmlNamespaceDefinition { + /// The namespace URI (a text expression). + pub uri: Expr, + /// The alias for the namespace (a simple identifier). + pub name: Ident, +} + +impl fmt::Display for XmlNamespaceDefinition { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!(f, "{} AS {}", self.uri, self.name) + } +} diff --git a/src/ast/spans.rs b/src/ast/spans.rs index a241fdf4d..27d52c26f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1909,6 +1909,7 @@ impl Spanned for TableFactor { .chain(alias.as_ref().map(|alias| alias.span())), ), TableFactor::JsonTable { .. } => Span::empty(), + TableFactor::XmlTable { .. } => Span::empty(), TableFactor::Pivot { table, aggregate_functions, diff --git a/src/keywords.rs b/src/keywords.rs index a5400a5b0..4eaad7ed2 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -654,6 +654,7 @@ define_keywords!( PARTITION, PARTITIONED, PARTITIONS, + PASSING, PASSWORD, PAST, PATH, @@ -989,6 +990,8 @@ define_keywords!( WORK, WRITE, XML, + XMLNAMESPACES, + XMLTABLE, XOR, YEAR, YEARS, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a9ddd1837..77466b97e 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11992,6 +11992,7 @@ impl<'a> Parser<'a> { | TableFactor::Function { alias, .. } | TableFactor::UNNEST { alias, .. } | TableFactor::JsonTable { alias, .. } + | TableFactor::XmlTable { alias, .. } | TableFactor::OpenJsonTable { alias, .. } | TableFactor::TableFunction { alias, .. } | TableFactor::Pivot { alias, .. } @@ -12107,6 +12108,9 @@ impl<'a> Parser<'a> { } else if self.parse_keyword_with_tokens(Keyword::OPENJSON, &[Token::LParen]) { self.prev_token(); self.parse_open_json_table_factor() + } else if self.parse_keyword_with_tokens(Keyword::XMLTABLE, &[Token::LParen]) { + self.prev_token(); + self.parse_xml_table_factor() } else { let name = self.parse_object_name(true)?; @@ -12339,6 +12343,99 @@ impl<'a> Parser<'a> { }) } + fn parse_xml_table_factor(&mut self) -> Result { + self.expect_token(&Token::LParen)?; + let namespaces = if self.parse_keyword(Keyword::XMLNAMESPACES) { + self.expect_token(&Token::LParen)?; + let namespaces = self.parse_comma_separated(Parser::parse_xml_namespace_definition)?; + self.expect_token(&Token::RParen)?; + self.expect_token(&Token::Comma)?; + namespaces + } else { + vec![] + }; + let row_expression = self.parse_expr()?; + let passing = self.parse_xml_passing_clause()?; + self.expect_keyword_is(Keyword::COLUMNS)?; + let columns = self.parse_comma_separated(Parser::parse_xml_table_column)?; + self.expect_token(&Token::RParen)?; + let alias = self.maybe_parse_table_alias()?; + Ok(TableFactor::XmlTable { + namespaces, + row_expression, + passing, + columns, + alias, + }) + } + + fn parse_xml_namespace_definition(&mut self) -> Result { + let uri = self.parse_expr()?; + self.expect_keyword_is(Keyword::AS)?; + let name = self.parse_identifier()?; + Ok(XmlNamespaceDefinition { uri, name }) + } + + fn parse_xml_table_column(&mut self) -> Result { + let name = self.parse_identifier()?; + + let option = if self.parse_keyword(Keyword::FOR) { + self.expect_keyword(Keyword::ORDINALITY)?; + XmlTableColumnOption::ForOrdinality + } else { + let r#type = self.parse_data_type()?; + let mut path = None; + let mut default = None; + + if self.parse_keyword(Keyword::PATH) { + path = Some(self.parse_expr()?); + } + + if self.parse_keyword(Keyword::DEFAULT) { + default = Some(self.parse_expr()?); + } + + let not_null = self.parse_keywords(&[Keyword::NOT, Keyword::NULL]); + if !not_null { + // NULL is the default but can be specified explicitly + let _ = self.parse_keyword(Keyword::NULL); + } + + XmlTableColumnOption::NamedInfo { + r#type, + path, + default, + nullable: !not_null, + } + }; + Ok(XmlTableColumn { name, option }) + } + + fn parse_xml_passing_clause(&mut self) -> Result { + let mut arguments = vec![]; + if self.parse_keyword(Keyword::PASSING) { + loop { + let by_value = + self.parse_keyword(Keyword::BY) && self.expect_keyword(Keyword::VALUE).is_ok(); + let expr = self.parse_expr()?; + let alias = if self.parse_keyword(Keyword::AS) { + Some(self.parse_identifier()?) + } else { + None + }; + arguments.push(XmlPassingArgument { + expr, + alias, + by_value, + }); + if !self.consume_token(&Token::Comma) { + break; + } + } + } + Ok(XmlPassingClause { arguments }) + } + fn parse_match_recognize(&mut self, table: TableFactor) -> Result { self.expect_token(&Token::LParen)?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 1c03a0fa4..466b65ec6 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11741,6 +11741,44 @@ fn test_group_by_grouping_sets() { ); } +#[test] +fn test_xmltable() { + all_dialects() + .verified_only_select("SELECT * FROM XMLTABLE('/root' PASSING data COLUMNS element TEXT)"); + + // Minimal meaningful working example: returns a single row with a single column named y containing the value z + all_dialects().verified_only_select( + "SELECT y FROM XMLTABLE('/X' PASSING 'z' COLUMNS y TEXT)", + ); + + // Test using subqueries + all_dialects().verified_only_select("SELECT y FROM XMLTABLE((SELECT '/X') PASSING (SELECT CAST('z' AS xml)) COLUMNS y TEXT PATH (SELECT 'y'))"); + + // NOT NULL + all_dialects().verified_only_select( + "SELECT y FROM XMLTABLE('/X' PASSING '' COLUMNS y TEXT NOT NULL)", + ); + + all_dialects().verified_only_select("SELECT * FROM XMLTABLE('/root/row' PASSING xmldata COLUMNS id INT PATH '@id', name TEXT PATH 'name/text()', value FLOAT PATH 'value')"); + + all_dialects().verified_only_select("SELECT * FROM XMLTABLE('//ROWS/ROW' PASSING data COLUMNS row_num FOR ORDINALITY, id INT PATH '@id', name TEXT PATH 'NAME' DEFAULT 'unnamed')"); + + // Example from https://www.postgresql.org/docs/15/functions-xml.html#FUNCTIONS-XML-PROCESSING + all_dialects().verified_only_select( + "SELECT xmltable.* FROM xmldata, XMLTABLE('//ROWS/ROW' PASSING data COLUMNS id INT PATH '@id', ordinality FOR ORDINALITY, \"COUNTRY_NAME\" TEXT, country_id TEXT PATH 'COUNTRY_ID', size_sq_km FLOAT PATH 'SIZE[@unit = \"sq_km\"]', size_other TEXT PATH 'concat(SIZE[@unit!=\"sq_km\"], \" \", SIZE[@unit!=\"sq_km\"]/@unit)', premier_name TEXT PATH 'PREMIER_NAME' DEFAULT 'not specified')" + ); + + // Example from DB2 docs without explicit PASSING clause: https://www.ibm.com/docs/en/db2/12.1.0?topic=xquery-simple-column-name-passing-xmlexists-xmlquery-xmltable + all_dialects().verified_only_select( + "SELECT X.* FROM T1, XMLTABLE('$CUSTLIST/customers/customerinfo' COLUMNS \"Cid\" BIGINT PATH '@Cid', \"Info\" XML PATH 'document{.}', \"History\" XML PATH 'NULL') AS X" + ); + + // Example from PostgreSQL with XMLNAMESPACES + all_dialects().verified_only_select( + "SELECT xmltable.* FROM XMLTABLE(XMLNAMESPACES('http://example.com/myns' AS x, 'http://example.com/b' AS \"B\"), '/x:example/x:item' PASSING (SELECT data FROM xmldata) COLUMNS foo INT PATH '@foo', bar INT PATH '@B:bar')" + ); +} + #[test] fn test_match_recognize() { use MatchRecognizePattern::*; From 2eb1e7bdd45cb8a0e9fd8d9b1a2717fa86794332 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Wed, 23 Apr 2025 12:10:57 -0400 Subject: [PATCH 108/286] Add `CREATE FUNCTION` support for SQL Server (#1808) --- src/ast/ddl.rs | 10 +++- src/ast/mod.rs | 100 ++++++++++++++++++++++++++++---- src/ast/spans.rs | 26 ++++++--- src/dialect/mssql.rs | 16 ++++-- src/parser/mod.rs | 110 ++++++++++++++++++++++++++++-------- tests/sqlparser_bigquery.rs | 1 + tests/sqlparser_common.rs | 8 +++ tests/sqlparser_hive.rs | 4 +- tests/sqlparser_mssql.rs | 86 ++++++++++++++++++++++++++++ tests/sqlparser_postgres.rs | 2 + 10 files changed, 313 insertions(+), 50 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 000ab3a4f..c1c113b32 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2157,6 +2157,10 @@ impl fmt::Display for ClusteredBy { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct CreateFunction { + /// True if this is a `CREATE OR ALTER FUNCTION` statement + /// + /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql?view=sql-server-ver16#or-alter) + pub or_alter: bool, pub or_replace: bool, pub temporary: bool, pub if_not_exists: bool, @@ -2219,9 +2223,10 @@ impl fmt::Display for CreateFunction { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!( f, - "CREATE {or_replace}{temp}FUNCTION {if_not_exists}{name}", + "CREATE {or_alter}{or_replace}{temp}FUNCTION {if_not_exists}{name}", name = self.name, temp = if self.temporary { "TEMPORARY " } else { "" }, + or_alter = if self.or_alter { "OR ALTER " } else { "" }, or_replace = if self.or_replace { "OR REPLACE " } else { "" }, if_not_exists = if self.if_not_exists { "IF NOT EXISTS " @@ -2272,6 +2277,9 @@ impl fmt::Display for CreateFunction { if let Some(CreateFunctionBody::AsAfterOptions(function_body)) = &self.function_body { write!(f, " AS {function_body}")?; } + if let Some(CreateFunctionBody::AsBeginEnd(bes)) = &self.function_body { + write!(f, " AS {bes}")?; + } Ok(()) } } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 74e8cb55c..c4bb3fa17 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2293,18 +2293,14 @@ pub enum ConditionalStatements { /// SELECT 1; SELECT 2; SELECT 3; ... Sequence { statements: Vec }, /// BEGIN SELECT 1; SELECT 2; SELECT 3; ... END - BeginEnd { - begin_token: AttachedToken, - statements: Vec, - end_token: AttachedToken, - }, + BeginEnd(BeginEndStatements), } impl ConditionalStatements { pub fn statements(&self) -> &Vec { match self { ConditionalStatements::Sequence { statements } => statements, - ConditionalStatements::BeginEnd { statements, .. } => statements, + ConditionalStatements::BeginEnd(bes) => &bes.statements, } } } @@ -2318,15 +2314,44 @@ impl fmt::Display for ConditionalStatements { } Ok(()) } - ConditionalStatements::BeginEnd { statements, .. } => { - write!(f, "BEGIN ")?; - format_statement_list(f, statements)?; - write!(f, " END") - } + ConditionalStatements::BeginEnd(bes) => write!(f, "{}", bes), } } } +/// Represents a list of statements enclosed within `BEGIN` and `END` keywords. +/// Example: +/// ```sql +/// BEGIN +/// SELECT 1; +/// SELECT 2; +/// END +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct BeginEndStatements { + pub begin_token: AttachedToken, + pub statements: Vec, + pub end_token: AttachedToken, +} + +impl fmt::Display for BeginEndStatements { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let BeginEndStatements { + begin_token: AttachedToken(begin_token), + statements, + end_token: AttachedToken(end_token), + } = self; + + write!(f, "{begin_token} ")?; + if !statements.is_empty() { + format_statement_list(f, statements)?; + } + write!(f, " {end_token}") + } +} + /// A `RAISE` statement. /// /// Examples: @@ -3615,6 +3640,7 @@ pub enum Statement { /// 1. [Hive](https://cwiki.apache.org/confluence/display/hive/languagemanual+ddl#LanguageManualDDL-Create/Drop/ReloadFunction) /// 2. [PostgreSQL](https://www.postgresql.org/docs/15/sql-createfunction.html) /// 3. [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_function_statement) + /// 4. [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql) CreateFunction(CreateFunction), /// CREATE TRIGGER /// @@ -4061,6 +4087,12 @@ pub enum Statement { /// /// See: Print(PrintStatement), + /// ```sql + /// RETURN [ expression ] + /// ``` + /// + /// See [ReturnStatement] + Return(ReturnStatement), } #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -5753,6 +5785,7 @@ impl fmt::Display for Statement { Ok(()) } Statement::Print(s) => write!(f, "{s}"), + Statement::Return(r) => write!(f, "{r}"), Statement::List(command) => write!(f, "LIST {command}"), Statement::Remove(command) => write!(f, "REMOVE {command}"), } @@ -8355,6 +8388,7 @@ impl fmt::Display for FunctionDeterminismSpecifier { /// /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#syntax_11 /// [PostgreSQL]: https://www.postgresql.org/docs/15/sql-createfunction.html +/// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -8383,6 +8417,22 @@ pub enum CreateFunctionBody { /// /// [BigQuery]: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#syntax_11 AsAfterOptions(Expr), + /// Function body with statements before the `RETURN` keyword. + /// + /// Example: + /// ```sql + /// CREATE FUNCTION my_scalar_udf(a INT, b INT) + /// RETURNS INT + /// AS + /// BEGIN + /// DECLARE c INT; + /// SET c = a + b; + /// RETURN c; + /// END + /// ``` + /// + /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql + AsBeginEnd(BeginEndStatements), /// Function body expression using the 'RETURN' keyword. /// /// Example: @@ -9231,6 +9281,34 @@ impl fmt::Display for PrintStatement { } } +/// Represents a `Return` statement. +/// +/// [MsSql triggers](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-trigger-transact-sql) +/// [MsSql functions](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ReturnStatement { + pub value: Option, +} + +impl fmt::Display for ReturnStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match &self.value { + Some(ReturnStatementValue::Expr(expr)) => write!(f, "RETURN {}", expr), + None => write!(f, "RETURN"), + } + } +} + +/// Variants of a `RETURN` statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ReturnStatementValue { + Expr(Expr), +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 27d52c26f..16ff660dc 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -23,8 +23,8 @@ use crate::tokenizer::Span; use super::{ dcl::SecondaryRoles, value::ValueWithSpan, AccessExpr, AlterColumnOperation, AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, AttachedToken, - CaseStatement, CloseCursor, ClusteredIndex, ColumnDef, ColumnOption, ColumnOptionDef, - ConditionalStatementBlock, ConditionalStatements, ConflictTarget, ConnectBy, + BeginEndStatements, CaseStatement, CloseCursor, ClusteredIndex, ColumnDef, ColumnOption, + ColumnOptionDef, ConditionalStatementBlock, ConditionalStatements, ConflictTarget, ConnectBy, ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, @@ -520,6 +520,7 @@ impl Spanned for Statement { Statement::RenameTable { .. } => Span::empty(), Statement::RaisError { .. } => Span::empty(), Statement::Print { .. } => Span::empty(), + Statement::Return { .. } => Span::empty(), Statement::List(..) | Statement::Remove(..) => Span::empty(), } } @@ -778,11 +779,7 @@ impl Spanned for ConditionalStatements { ConditionalStatements::Sequence { statements } => { union_spans(statements.iter().map(|s| s.span())) } - ConditionalStatements::BeginEnd { - begin_token: AttachedToken(start), - statements: _, - end_token: AttachedToken(end), - } => union_spans([start.span, end.span].into_iter()), + ConditionalStatements::BeginEnd(bes) => bes.span(), } } } @@ -2282,6 +2279,21 @@ impl Spanned for TableObject { } } +impl Spanned for BeginEndStatements { + fn span(&self) -> Span { + let BeginEndStatements { + begin_token, + statements, + end_token, + } = self; + union_spans( + core::iter::once(begin_token.0.span) + .chain(statements.iter().map(|i| i.span())) + .chain(core::iter::once(end_token.0.span)), + ) + } +} + #[cfg(test)] pub mod tests { use crate::dialect::{Dialect, GenericDialect, SnowflakeDialect}; diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index d86d68a20..31e324f06 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -16,7 +16,9 @@ // under the License. use crate::ast::helpers::attached_token::AttachedToken; -use crate::ast::{ConditionalStatementBlock, ConditionalStatements, IfStatement, Statement}; +use crate::ast::{ + BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, IfStatement, Statement, +}; use crate::dialect::Dialect; use crate::keywords::{self, Keyword}; use crate::parser::{Parser, ParserError}; @@ -149,11 +151,11 @@ impl MsSqlDialect { start_token: AttachedToken(if_token), condition: Some(condition), then_token: None, - conditional_statements: ConditionalStatements::BeginEnd { + conditional_statements: ConditionalStatements::BeginEnd(BeginEndStatements { begin_token: AttachedToken(begin_token), statements, end_token: AttachedToken(end_token), - }, + }), } } else { let stmt = parser.parse_statement()?; @@ -167,8 +169,10 @@ impl MsSqlDialect { } }; + let mut prior_statement_ended_with_semi_colon = false; while let Token::SemiColon = parser.peek_token_ref().token { parser.advance_token(); + prior_statement_ended_with_semi_colon = true; } let mut else_block = None; @@ -182,11 +186,11 @@ impl MsSqlDialect { start_token: AttachedToken(else_token), condition: None, then_token: None, - conditional_statements: ConditionalStatements::BeginEnd { + conditional_statements: ConditionalStatements::BeginEnd(BeginEndStatements { begin_token: AttachedToken(begin_token), statements, end_token: AttachedToken(end_token), - }, + }), }); } else { let stmt = parser.parse_statement()?; @@ -199,6 +203,8 @@ impl MsSqlDialect { }, }); } + } else if prior_statement_ended_with_semi_colon { + parser.prev_token(); } Ok(Statement::If(IfStatement { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 77466b97e..9b519fe84 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -577,13 +577,7 @@ impl<'a> Parser<'a> { Keyword::GRANT => self.parse_grant(), Keyword::REVOKE => self.parse_revoke(), Keyword::START => self.parse_start_transaction(), - // `BEGIN` is a nonstandard but common alias for the - // standard `START TRANSACTION` statement. It is supported - // by at least PostgreSQL and MySQL. Keyword::BEGIN => self.parse_begin(), - // `END` is a nonstandard but common alias for the - // standard `COMMIT TRANSACTION` statement. It is supported - // by PostgreSQL. Keyword::END => self.parse_end(), Keyword::SAVEPOINT => self.parse_savepoint(), Keyword::RELEASE => self.parse_release(), @@ -618,6 +612,7 @@ impl<'a> Parser<'a> { // `COMMENT` is snowflake specific https://docs.snowflake.com/en/sql-reference/sql/comment Keyword::COMMENT if self.dialect.supports_comment_on() => self.parse_comment(), Keyword::PRINT => self.parse_print(), + Keyword::RETURN => self.parse_return(), _ => self.expected("an SQL statement", next_token), }, Token::LParen => { @@ -4458,7 +4453,6 @@ impl<'a> Parser<'a> { break; } } - values.push(self.parse_statement()?); self.expect_token(&Token::SemiColon)?; } @@ -4560,7 +4554,7 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::EXTERNAL) { self.parse_create_external_table(or_replace) } else if self.parse_keyword(Keyword::FUNCTION) { - self.parse_create_function(or_replace, temporary) + self.parse_create_function(or_alter, or_replace, temporary) } else if self.parse_keyword(Keyword::TRIGGER) { self.parse_create_trigger(or_replace, false) } else if self.parse_keywords(&[Keyword::CONSTRAINT, Keyword::TRIGGER]) { @@ -4869,6 +4863,7 @@ impl<'a> Parser<'a> { pub fn parse_create_function( &mut self, + or_alter: bool, or_replace: bool, temporary: bool, ) -> Result { @@ -4880,6 +4875,8 @@ impl<'a> Parser<'a> { self.parse_create_macro(or_replace, temporary) } else if dialect_of!(self is BigQueryDialect) { self.parse_bigquery_create_function(or_replace, temporary) + } else if dialect_of!(self is MsSqlDialect) { + self.parse_mssql_create_function(or_alter, or_replace, temporary) } else { self.prev_token(); self.expected("an object type after CREATE", self.peek_token()) @@ -4994,6 +4991,7 @@ impl<'a> Parser<'a> { } Ok(Statement::CreateFunction(CreateFunction { + or_alter: false, or_replace, temporary, name, @@ -5027,6 +5025,7 @@ impl<'a> Parser<'a> { let using = self.parse_optional_create_function_using()?; Ok(Statement::CreateFunction(CreateFunction { + or_alter: false, or_replace, temporary, name, @@ -5054,22 +5053,7 @@ impl<'a> Parser<'a> { temporary: bool, ) -> Result { let if_not_exists = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); - let name = self.parse_object_name(false)?; - - let parse_function_param = - |parser: &mut Parser| -> Result { - let name = parser.parse_identifier()?; - let data_type = parser.parse_data_type()?; - Ok(OperateFunctionArg { - mode: None, - name: Some(name), - data_type, - default_expr: None, - }) - }; - self.expect_token(&Token::LParen)?; - let args = self.parse_comma_separated0(parse_function_param, Token::RParen)?; - self.expect_token(&Token::RParen)?; + let (name, args) = self.parse_create_function_name_and_params()?; let return_type = if self.parse_keyword(Keyword::RETURNS) { Some(self.parse_data_type()?) @@ -5116,6 +5100,7 @@ impl<'a> Parser<'a> { }; Ok(Statement::CreateFunction(CreateFunction { + or_alter: false, or_replace, temporary, if_not_exists, @@ -5134,6 +5119,73 @@ impl<'a> Parser<'a> { })) } + /// Parse `CREATE FUNCTION` for [MsSql] + /// + /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql + fn parse_mssql_create_function( + &mut self, + or_alter: bool, + or_replace: bool, + temporary: bool, + ) -> Result { + let (name, args) = self.parse_create_function_name_and_params()?; + + self.expect_keyword(Keyword::RETURNS)?; + let return_type = Some(self.parse_data_type()?); + + self.expect_keyword_is(Keyword::AS)?; + + let begin_token = self.expect_keyword(Keyword::BEGIN)?; + let statements = self.parse_statement_list(&[Keyword::END])?; + let end_token = self.expect_keyword(Keyword::END)?; + + let function_body = Some(CreateFunctionBody::AsBeginEnd(BeginEndStatements { + begin_token: AttachedToken(begin_token), + statements, + end_token: AttachedToken(end_token), + })); + + Ok(Statement::CreateFunction(CreateFunction { + or_alter, + or_replace, + temporary, + if_not_exists: false, + name, + args: Some(args), + return_type, + function_body, + language: None, + determinism_specifier: None, + options: None, + remote_connection: None, + using: None, + behavior: None, + called_on_null: None, + parallel: None, + })) + } + + fn parse_create_function_name_and_params( + &mut self, + ) -> Result<(ObjectName, Vec), ParserError> { + let name = self.parse_object_name(false)?; + let parse_function_param = + |parser: &mut Parser| -> Result { + let name = parser.parse_identifier()?; + let data_type = parser.parse_data_type()?; + Ok(OperateFunctionArg { + mode: None, + name: Some(name), + data_type, + default_expr: None, + }) + }; + self.expect_token(&Token::LParen)?; + let args = self.parse_comma_separated0(parse_function_param, Token::RParen)?; + self.expect_token(&Token::RParen)?; + Ok((name, args)) + } + fn parse_function_arg(&mut self) -> Result { let mode = if self.parse_keyword(Keyword::IN) { Some(ArgMode::In) @@ -15161,6 +15213,16 @@ impl<'a> Parser<'a> { })) } + /// Parse [Statement::Return] + fn parse_return(&mut self) -> Result { + match self.maybe_parse(|p| p.parse_expr())? { + Some(expr) => Ok(Statement::Return(ReturnStatement { + value: Some(ReturnStatementValue::Expr(expr)), + })), + None => Ok(Statement::Return(ReturnStatement { value: None })), + } + } + /// Consume the parser and return its underlying token buffer pub fn into_tokens(self) -> Vec { self.tokens diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 5eb30d15c..416d2e435 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -2134,6 +2134,7 @@ fn test_bigquery_create_function() { assert_eq!( stmt, Statement::CreateFunction(CreateFunction { + or_alter: false, or_replace: true, temporary: true, if_not_exists: false, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 466b65ec6..66800b811 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15079,3 +15079,11 @@ fn parse_set_time_zone_alias() { _ => unreachable!(), } } + +#[test] +fn parse_return() { + let stmt = all_dialects().verified_stmt("RETURN"); + assert_eq!(stmt, Statement::Return(ReturnStatement { value: None })); + + let _ = all_dialects().verified_stmt("RETURN 1"); +} diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 2af93db7d..9b0430947 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -25,7 +25,7 @@ use sqlparser::ast::{ Expr, Function, FunctionArgumentList, FunctionArguments, Ident, ObjectName, OrderByExpr, OrderByOptions, SelectItem, Set, Statement, TableFactor, UnaryOperator, Use, Value, }; -use sqlparser::dialect::{GenericDialect, HiveDialect, MsSqlDialect}; +use sqlparser::dialect::{AnsiDialect, GenericDialect, HiveDialect}; use sqlparser::parser::ParserError; use sqlparser::test_utils::*; @@ -423,7 +423,7 @@ fn parse_create_function() { } // Test error in dialect that doesn't support parsing CREATE FUNCTION - let unsupported_dialects = TestedDialects::new(vec![Box::new(MsSqlDialect {})]); + let unsupported_dialects = TestedDialects::new(vec![Box::new(AnsiDialect {})]); assert_eq!( unsupported_dialects.parse_sql_statements(sql).unwrap_err(), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 2786384b3..b86e1a7d4 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -187,6 +187,92 @@ fn parse_mssql_create_procedure() { let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN UPDATE bar SET col = 'test'; SELECT [foo] FROM BAR WHERE [FOO] > 10 END"); } +#[test] +fn parse_create_function() { + let return_expression_function = "CREATE FUNCTION some_scalar_udf(@foo INT, @bar VARCHAR(256)) RETURNS INT AS BEGIN RETURN 1; END"; + assert_eq!( + ms().verified_stmt(return_expression_function), + sqlparser::ast::Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: false, + temporary: false, + if_not_exists: false, + name: ObjectName::from(vec![Ident::new("some_scalar_udf")]), + args: Some(vec![ + OperateFunctionArg { + mode: None, + name: Some(Ident::new("@foo")), + data_type: DataType::Int(None), + default_expr: None, + }, + OperateFunctionArg { + mode: None, + name: Some(Ident::new("@bar")), + data_type: DataType::Varchar(Some(CharacterLength::IntegerLength { + length: 256, + unit: None + })), + default_expr: None, + }, + ]), + return_type: Some(DataType::Int(None)), + function_body: Some(CreateFunctionBody::AsBeginEnd(BeginEndStatements { + begin_token: AttachedToken::empty(), + statements: vec![Statement::Return(ReturnStatement { + value: Some(ReturnStatementValue::Expr(Expr::Value( + (number("1")).with_empty_span() + ))), + }),], + end_token: AttachedToken::empty(), + })), + behavior: None, + called_on_null: None, + parallel: None, + using: None, + language: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }), + ); + + let multi_statement_function = "\ + CREATE FUNCTION some_scalar_udf(@foo INT, @bar VARCHAR(256)) \ + RETURNS INT \ + AS \ + BEGIN \ + SET @foo = @foo + 1; \ + RETURN @foo; \ + END\ + "; + let _ = ms().verified_stmt(multi_statement_function); + + let create_function_with_conditional = "\ + CREATE FUNCTION some_scalar_udf() \ + RETURNS INT \ + AS \ + BEGIN \ + IF 1 = 2 \ + BEGIN \ + RETURN 1; \ + END; \ + RETURN 0; \ + END\ + "; + let _ = ms().verified_stmt(create_function_with_conditional); + + let create_or_alter_function = "\ + CREATE OR ALTER FUNCTION some_scalar_udf(@foo INT, @bar VARCHAR(256)) \ + RETURNS INT \ + AS \ + BEGIN \ + SET @foo = @foo + 1; \ + RETURN @foo; \ + END\ + "; + let _ = ms().verified_stmt(create_or_alter_function); +} + #[test] fn parse_mssql_apply_join() { let _ = ms_and_generic().verified_only_select( diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 098d4b1c4..27fc7fa17 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -4104,6 +4104,7 @@ fn parse_create_function() { assert_eq!( pg_and_generic().verified_stmt(sql), Statement::CreateFunction(CreateFunction { + or_alter: false, or_replace: false, temporary: false, name: ObjectName::from(vec![Ident::new("add")]), @@ -5485,6 +5486,7 @@ fn parse_trigger_related_functions() { assert_eq!( create_function, Statement::CreateFunction(CreateFunction { + or_alter: false, or_replace: false, temporary: false, if_not_exists: false, From 87d190734c7b978e8252b110c9529d7a93a30cf0 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Wed, 23 Apr 2025 12:55:57 -0400 Subject: [PATCH 109/286] Add `OR ALTER` support for `CREATE VIEW` (#1818) --- src/ast/mod.rs | 8 +++++++- src/ast/spans.rs | 1 + src/parser/mod.rs | 4 +++- tests/sqlparser_common.rs | 16 ++++++++++++++++ 4 files changed, 27 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index c4bb3fa17..b60ade78b 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3063,6 +3063,10 @@ pub enum Statement { /// CREATE VIEW /// ``` CreateView { + /// True if this is a `CREATE OR ALTER VIEW` statement + /// + /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-view-transact-sql) + or_alter: bool, or_replace: bool, materialized: bool, /// View name @@ -4623,6 +4627,7 @@ impl fmt::Display for Statement { Ok(()) } Statement::CreateView { + or_alter, name, or_replace, columns, @@ -4639,7 +4644,8 @@ impl fmt::Display for Statement { } => { write!( f, - "CREATE {or_replace}", + "CREATE {or_alter}{or_replace}", + or_alter = if *or_alter { "OR ALTER " } else { "" }, or_replace = if *or_replace { "OR REPLACE " } else { "" }, )?; if let Some(params) = params { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 16ff660dc..93de5fff2 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -384,6 +384,7 @@ impl Spanned for Statement { ), Statement::Delete(delete) => delete.span(), Statement::CreateView { + or_alter: _, or_replace: _, materialized: _, name, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9b519fe84..fe81b5999 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4548,7 +4548,7 @@ impl<'a> Parser<'a> { self.parse_create_table(or_replace, temporary, global, transient) } else if self.parse_keyword(Keyword::MATERIALIZED) || self.parse_keyword(Keyword::VIEW) { self.prev_token(); - self.parse_create_view(or_replace, temporary, create_view_params) + self.parse_create_view(or_alter, or_replace, temporary, create_view_params) } else if self.parse_keyword(Keyword::POLICY) { self.parse_create_policy() } else if self.parse_keyword(Keyword::EXTERNAL) { @@ -5512,6 +5512,7 @@ impl<'a> Parser<'a> { pub fn parse_create_view( &mut self, + or_alter: bool, or_replace: bool, temporary: bool, create_view_params: Option, @@ -5576,6 +5577,7 @@ impl<'a> Parser<'a> { ]); Ok(Statement::CreateView { + or_alter, name, columns, query, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 66800b811..fa2346c2c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7840,6 +7840,7 @@ fn parse_create_view() { let sql = "CREATE VIEW myschema.myview AS SELECT foo FROM bar"; match verified_stmt(sql) { Statement::CreateView { + or_alter, name, columns, query, @@ -7854,6 +7855,7 @@ fn parse_create_view() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); assert_eq!("SELECT foo FROM bar", query.to_string()); @@ -7870,6 +7872,8 @@ fn parse_create_view() { } _ => unreachable!(), } + + let _ = verified_stmt("CREATE OR ALTER VIEW v AS SELECT 1"); } #[test] @@ -7904,6 +7908,7 @@ fn parse_create_view_with_columns() { // match all_dialects().verified_stmt(sql) { match all_dialects_except(|d| d.is::()).verified_stmt(sql) { Statement::CreateView { + or_alter, name, columns, or_replace, @@ -7918,6 +7923,7 @@ fn parse_create_view_with_columns() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); assert_eq!( columns, @@ -7951,6 +7957,7 @@ fn parse_create_view_temporary() { let sql = "CREATE TEMPORARY VIEW myschema.myview AS SELECT foo FROM bar"; match verified_stmt(sql) { Statement::CreateView { + or_alter, name, columns, query, @@ -7965,6 +7972,7 @@ fn parse_create_view_temporary() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); assert_eq!("SELECT foo FROM bar", query.to_string()); @@ -7988,6 +7996,7 @@ fn parse_create_or_replace_view() { let sql = "CREATE OR REPLACE VIEW v AS SELECT 1"; match verified_stmt(sql) { Statement::CreateView { + or_alter, name, columns, or_replace, @@ -8002,6 +8011,7 @@ fn parse_create_or_replace_view() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); assert_eq!(columns, vec![]); assert_eq!(options, CreateTableOptions::None); @@ -8029,6 +8039,7 @@ fn parse_create_or_replace_materialized_view() { let sql = "CREATE OR REPLACE MATERIALIZED VIEW v AS SELECT 1"; match verified_stmt(sql) { Statement::CreateView { + or_alter, name, columns, or_replace, @@ -8043,6 +8054,7 @@ fn parse_create_or_replace_materialized_view() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); assert_eq!(columns, vec![]); assert_eq!(options, CreateTableOptions::None); @@ -8066,6 +8078,7 @@ fn parse_create_materialized_view() { let sql = "CREATE MATERIALIZED VIEW myschema.myview AS SELECT foo FROM bar"; match verified_stmt(sql) { Statement::CreateView { + or_alter, name, or_replace, columns, @@ -8080,6 +8093,7 @@ fn parse_create_materialized_view() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); assert_eq!("SELECT foo FROM bar", query.to_string()); @@ -8103,6 +8117,7 @@ fn parse_create_materialized_view_with_cluster_by() { let sql = "CREATE MATERIALIZED VIEW myschema.myview CLUSTER BY (foo) AS SELECT foo FROM bar"; match verified_stmt(sql) { Statement::CreateView { + or_alter, name, or_replace, columns, @@ -8117,6 +8132,7 @@ fn parse_create_materialized_view_with_cluster_by() { to, params, } => { + assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); assert_eq!("SELECT foo FROM bar", query.to_string()); From 7703fd0d3180c2e8b347c11394084c3a2458be14 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Thu, 24 Apr 2025 14:16:49 -0400 Subject: [PATCH 110/286] Add `DECLARE ... CURSOR FOR` support for SQL Server (#1821) --- src/ast/mod.rs | 3 ++- src/parser/mod.rs | 22 +++++++++++++++++----- tests/sqlparser_mssql.rs | 4 ++++ 3 files changed, 23 insertions(+), 6 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index b60ade78b..45924579b 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2472,10 +2472,11 @@ impl fmt::Display for DeclareAssignment { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum DeclareType { - /// Cursor variable type. e.g. [Snowflake] [PostgreSQL] + /// Cursor variable type. e.g. [Snowflake] [PostgreSQL] [MsSql] /// /// [Snowflake]: https://docs.snowflake.com/en/developer-guide/snowflake-scripting/cursors#declaring-a-cursor /// [PostgreSQL]: https://www.postgresql.org/docs/current/plpgsql-cursors.html + /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/language-elements/declare-cursor-transact-sql Cursor, /// Result set variable type. [Snowflake] diff --git a/src/parser/mod.rs b/src/parser/mod.rs index fe81b5999..0546548af 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6446,7 +6446,7 @@ impl<'a> Parser<'a> { /// DECLARE // { // { @local_variable [AS] data_type [ = value ] } - // | { @cursor_variable_name CURSOR } + // | { @cursor_variable_name CURSOR [ FOR ] } // } [ ,...n ] /// ``` /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/language-elements/declare-local-variable-transact-sql?view=sql-server-ver16 @@ -6462,14 +6462,19 @@ impl<'a> Parser<'a> { /// ```text // { // { @local_variable [AS] data_type [ = value ] } - // | { @cursor_variable_name CURSOR } + // | { @cursor_variable_name CURSOR [ FOR ]} // } [ ,...n ] /// ``` /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/language-elements/declare-local-variable-transact-sql?view=sql-server-ver16 pub fn parse_mssql_declare_stmt(&mut self) -> Result { let name = { let ident = self.parse_identifier()?; - if !ident.value.starts_with('@') { + if !ident.value.starts_with('@') + && !matches!( + self.peek_token().token, + Token::Word(w) if w.keyword == Keyword::CURSOR + ) + { Err(ParserError::TokenizerError( "Invalid MsSql variable declaration.".to_string(), )) @@ -6493,7 +6498,14 @@ impl<'a> Parser<'a> { _ => (None, Some(self.parse_data_type()?)), }; - let assignment = self.parse_mssql_variable_declaration_expression()?; + let (for_query, assignment) = if self.peek_keyword(Keyword::FOR) { + self.next_token(); + let query = Some(self.parse_query()?); + (query, None) + } else { + let assignment = self.parse_mssql_variable_declaration_expression()?; + (None, assignment) + }; Ok(Declare { names: vec![name], @@ -6504,7 +6516,7 @@ impl<'a> Parser<'a> { sensitive: None, scroll: None, hold: None, - for_query: None, + for_query, }) } diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index b86e1a7d4..ef6103474 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1387,6 +1387,10 @@ fn parse_mssql_declare() { ], ast ); + + let declare_cursor_for_select = + "DECLARE vend_cursor CURSOR FOR SELECT * FROM Purchasing.Vendor"; + let _ = ms().verified_stmt(declare_cursor_for_select); } #[test] From 4e392f5c07ef1855c5de3ddb604be5d718ab1040 Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Mon, 28 Apr 2025 19:03:39 +0200 Subject: [PATCH 111/286] Handle missing login in changelog generate script (#1823) --- dev/release/generate-changelog.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/dev/release/generate-changelog.py b/dev/release/generate-changelog.py index 52fd2e548..6f2b7c41c 100755 --- a/dev/release/generate-changelog.py +++ b/dev/release/generate-changelog.py @@ -28,7 +28,8 @@ def print_pulls(repo_name, title, pulls): print() for (pull, commit) in pulls: url = "https://github.com/%7B%7D/pull/%7B%7D".format(repo_name, pull.number) - print("- {} [#{}]({}) ({})".format(pull.title, pull.number, url, commit.author.login)) + author = f"({commit.author.login})" if commit.author else '' + print("- {} [#{}]({}) {}".format(pull.title, pull.number, url, author)) print() @@ -161,4 +162,4 @@ def cli(args=None): generate_changelog(repo, project, args.tag1, args.tag2, args.version) if __name__ == "__main__": - cli() \ No newline at end of file + cli() From 2b5bdcef0b79274f8987d4c5d71cbc861dcc50cd Mon Sep 17 00:00:00 2001 From: tomershaniii <65544633+tomershaniii@users.noreply.github.com> Date: Tue, 29 Apr 2025 09:44:19 +0300 Subject: [PATCH 112/286] Snowflake: Add support for `CONNECT_BY_ROOT` (#1780) --- src/ast/mod.rs | 10 ++++--- src/ast/spans.rs | 2 +- src/dialect/mod.rs | 6 +++++ src/dialect/snowflake.rs | 6 +++++ src/keywords.rs | 1 + src/parser/mod.rs | 51 +++++++++++++++++++++++++++--------- tests/sqlparser_mysql.rs | 9 ++++--- tests/sqlparser_snowflake.rs | 42 +++++++++++++++++++++++++++++ 8 files changed, 107 insertions(+), 20 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 45924579b..d65889810 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -930,12 +930,14 @@ pub enum Expr { Nested(Box), /// A literal value, such as string, number, date or NULL Value(ValueWithSpan), + /// Prefixed expression, e.g. introducer strings, projection prefix /// - IntroducedString { - introducer: String, + /// + Prefixed { + prefix: Ident, /// The value of the constant. /// Hint: you can unwrap the string value using `value.into_string()`. - value: Value, + value: Box, }, /// A constant of form ` 'value'`. /// This can represent ANSI SQL `DATE`, `TIME`, and `TIMESTAMP` literals (such as `DATE '2020-01-01'`), @@ -1655,7 +1657,7 @@ impl fmt::Display for Expr { Expr::Collate { expr, collation } => write!(f, "{expr} COLLATE {collation}"), Expr::Nested(ast) => write!(f, "({ast})"), Expr::Value(v) => write!(f, "{v}"), - Expr::IntroducedString { introducer, value } => write!(f, "{introducer} {value}"), + Expr::Prefixed { prefix, value } => write!(f, "{prefix} {value}"), Expr::TypedString { data_type, value } => { write!(f, "{data_type}")?; write!(f, " {value}") diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 93de5fff2..28d479f30 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1543,7 +1543,7 @@ impl Spanned for Expr { .map(|items| union_spans(items.iter().map(|i| i.span()))), ), ), - Expr::IntroducedString { value, .. } => value.span(), + Expr::Prefixed { value, .. } => value.span(), Expr::Case { operand, conditions, diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index e41964f45..b2dff065e 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -888,6 +888,12 @@ pub trait Dialect: Debug + Any { keywords::RESERVED_FOR_TABLE_FACTOR } + /// Returns reserved keywords that may prefix a select item expression + /// e.g. `SELECT CONNECT_BY_ROOT name FROM Tbl2` (Snowflake) + fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] { + &[] + } + /// Returns true if this dialect supports the `TABLESAMPLE` option /// before the table alias option. For example: /// diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 8b279c7cc..c4d6a5ad5 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -44,6 +44,7 @@ use alloc::{format, vec}; use super::keywords::RESERVED_FOR_IDENTIFIER; use sqlparser::ast::StorageSerializationPolicy; +const RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR: [Keyword; 1] = [Keyword::CONNECT_BY_ROOT]; /// A [`Dialect`] for [Snowflake](https://www.snowflake.com/) #[derive(Debug, Default)] pub struct SnowflakeDialect; @@ -346,6 +347,11 @@ impl Dialect for SnowflakeDialect { fn supports_group_by_expr(&self) -> bool { true } + + /// See: + fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] { + &RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/keywords.rs b/src/keywords.rs index 4eaad7ed2..15a6f91ad 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -207,6 +207,7 @@ define_keywords!( CONNECT, CONNECTION, CONNECTOR, + CONNECT_BY_ROOT, CONSTRAINT, CONTAINS, CONTINUE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0546548af..03ea91faf 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1388,9 +1388,9 @@ impl<'a> Parser<'a> { | Token::HexStringLiteral(_) if w.value.starts_with('_') => { - Ok(Expr::IntroducedString { - introducer: w.value.clone(), - value: self.parse_introduced_string_value()?, + Ok(Expr::Prefixed { + prefix: w.clone().into_ident(w_span), + value: self.parse_introduced_string_expr()?.into(), }) } // string introducer https://dev.mysql.com/doc/refman/8.0/en/charset-introducer.html @@ -1399,9 +1399,9 @@ impl<'a> Parser<'a> { | Token::HexStringLiteral(_) if w.value.starts_with('_') => { - Ok(Expr::IntroducedString { - introducer: w.value.clone(), - value: self.parse_introduced_string_value()?, + Ok(Expr::Prefixed { + prefix: w.clone().into_ident(w_span), + value: self.parse_introduced_string_expr()?.into(), }) } Token::Arrow if self.dialect.supports_lambda_functions() => { @@ -9035,13 +9035,19 @@ impl<'a> Parser<'a> { } } - fn parse_introduced_string_value(&mut self) -> Result { + fn parse_introduced_string_expr(&mut self) -> Result { let next_token = self.next_token(); let span = next_token.span; match next_token.token { - Token::SingleQuotedString(ref s) => Ok(Value::SingleQuotedString(s.to_string())), - Token::DoubleQuotedString(ref s) => Ok(Value::DoubleQuotedString(s.to_string())), - Token::HexStringLiteral(ref s) => Ok(Value::HexStringLiteral(s.to_string())), + Token::SingleQuotedString(ref s) => Ok(Expr::Value( + Value::SingleQuotedString(s.to_string()).with_span(span), + )), + Token::DoubleQuotedString(ref s) => Ok(Expr::Value( + Value::DoubleQuotedString(s.to_string()).with_span(span), + )), + Token::HexStringLiteral(ref s) => Ok(Expr::Value( + Value::HexStringLiteral(s.to_string()).with_span(span), + )), unexpected => self.expected( "a string value", TokenWithSpan { @@ -13968,6 +13974,13 @@ impl<'a> Parser<'a> { /// Parse a comma-delimited list of projections after SELECT pub fn parse_select_item(&mut self) -> Result { + let prefix = self + .parse_one_of_keywords( + self.dialect + .get_reserved_keywords_for_select_item_operator(), + ) + .map(|keyword| Ident::new(format!("{:?}", keyword))); + match self.parse_wildcard_expr()? { Expr::QualifiedWildcard(prefix, token) => Ok(SelectItem::QualifiedWildcard( SelectItemQualifiedWildcardKind::ObjectName(prefix), @@ -14012,8 +14025,11 @@ impl<'a> Parser<'a> { expr => self .maybe_parse_select_item_alias() .map(|alias| match alias { - Some(alias) => SelectItem::ExprWithAlias { expr, alias }, - None => SelectItem::UnnamedExpr(expr), + Some(alias) => SelectItem::ExprWithAlias { + expr: maybe_prefixed_expr(expr, prefix), + alias, + }, + None => SelectItem::UnnamedExpr(maybe_prefixed_expr(expr, prefix)), }), } } @@ -15375,6 +15391,17 @@ impl<'a> Parser<'a> { } } +fn maybe_prefixed_expr(expr: Expr, prefix: Option) -> Expr { + if let Some(prefix) = prefix { + Expr::Prefixed { + prefix, + value: Box::new(expr), + } + } else { + expr + } +} + impl Word { #[deprecated(since = "0.54.0", note = "please use `into_ident` instead")] pub fn to_ident(&self, span: Span) -> Ident { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 9d8d12b57..f74248b86 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3020,9 +3020,12 @@ fn parse_hex_string_introducer() { distinct: None, top: None, top_before_distinct: false, - projection: vec![SelectItem::UnnamedExpr(Expr::IntroducedString { - introducer: "_latin1".to_string(), - value: Value::HexStringLiteral("4D7953514C".to_string()) + projection: vec![SelectItem::UnnamedExpr(Expr::Prefixed { + prefix: Ident::from("_latin1"), + value: Expr::Value( + Value::HexStringLiteral("4D7953514C".to_string()).with_empty_span() + ) + .into(), })], from: vec![], lateral_views: vec![], diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 84c08874e..aa974115d 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3983,3 +3983,45 @@ fn test_nested_join_without_parentheses() { }], ); } + +#[test] +fn parse_connect_by_root_operator() { + let sql = "SELECT CONNECT_BY_ROOT name AS root_name FROM Tbl1"; + + match snowflake().verified_stmt(sql) { + Statement::Query(query) => { + assert_eq!( + query.body.as_select().unwrap().projection[0], + SelectItem::ExprWithAlias { + expr: Expr::Prefixed { + prefix: Ident::new("CONNECT_BY_ROOT"), + value: Box::new(Expr::Identifier(Ident::new("name"))) + }, + alias: Ident::new("root_name"), + } + ); + } + _ => unreachable!(), + } + + let sql = "SELECT CONNECT_BY_ROOT name FROM Tbl2"; + match snowflake().verified_stmt(sql) { + Statement::Query(query) => { + assert_eq!( + query.body.as_select().unwrap().projection[0], + SelectItem::UnnamedExpr(Expr::Prefixed { + prefix: Ident::new("CONNECT_BY_ROOT"), + value: Box::new(Expr::Identifier(Ident::new("name"))) + }) + ); + } + _ => unreachable!(), + } + + let sql = "SELECT CONNECT_BY_ROOT FROM Tbl2"; + let res = snowflake().parse_sql_statements(sql); + assert_eq!( + res.unwrap_err().to_string(), + "sql parser error: Expected an expression, found: FROM" + ); +} From c0921dceb93218ca97bf7e0d65f1f28d7729289d Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Tue, 29 Apr 2025 21:32:04 +0200 Subject: [PATCH 113/286] Prepare for 0.56.0 release: Version and CHANGELOG (#1822) --- CHANGELOG.md | 1 + Cargo.toml | 2 +- changelog/0.56.0.md | 100 ++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 102 insertions(+), 1 deletion(-) create mode 100644 changelog/0.56.0.md diff --git a/CHANGELOG.md b/CHANGELOG.md index 362a637d9..a5511a053 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -28,6 +28,7 @@ technically be breaking and thus will result in a `0.(N+1)` version. - Unreleased: Check https://github.com/sqlparser-rs/sqlparser-rs/commits/main for undocumented changes. +- `0.56.0`: [changelog/0.56.0.md](changelog/0.56.0.md) - `0.55.0`: [changelog/0.55.0.md](changelog/0.55.0.md) - `0.54.0`: [changelog/0.54.0.md](changelog/0.54.0.md) - `0.53.0`: [changelog/0.53.0.md](changelog/0.53.0.md) diff --git a/Cargo.toml b/Cargo.toml index 99bfdc241..d746775e4 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.55.0" +version = "0.56.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/changelog/0.56.0.md b/changelog/0.56.0.md new file mode 100644 index 000000000..9e6289e88 --- /dev/null +++ b/changelog/0.56.0.md @@ -0,0 +1,100 @@ + + +# sqlparser-rs 0.56.0 Changelog + +This release consists of 45 commits from 19 contributors. See credits at the end of this changelog for more information. + +**Other:** + +- Ignore escaped LIKE wildcards in MySQL [#1735](https://github.com/apache/datafusion-sqlparser-rs/pull/1735) (mvzink) +- Parse SET NAMES syntax in Postgres [#1752](https://github.com/apache/datafusion-sqlparser-rs/pull/1752) (mvzink) +- re-add support for nested comments in mssql [#1754](https://github.com/apache/datafusion-sqlparser-rs/pull/1754) (lovasoa) +- Extend support for INDEX parsing [#1707](https://github.com/apache/datafusion-sqlparser-rs/pull/1707) (LucaCappelletti94) +- Parse MySQL `ALTER TABLE DROP FOREIGN KEY` syntax [#1762](https://github.com/apache/datafusion-sqlparser-rs/pull/1762) (mvzink) +- add support for `with` clauses (CTEs) in `delete` statements [#1764](https://github.com/apache/datafusion-sqlparser-rs/pull/1764) (lovasoa) +- SET with a list of comma separated assignments [#1757](https://github.com/apache/datafusion-sqlparser-rs/pull/1757) (MohamedAbdeen21) +- Preserve MySQL-style `LIMIT , ` syntax [#1765](https://github.com/apache/datafusion-sqlparser-rs/pull/1765) (mvzink) +- Add support for `DROP MATERIALIZED VIEW` [#1743](https://github.com/apache/datafusion-sqlparser-rs/pull/1743) (iffyio) +- Add `CASE` and `IF` statement support [#1741](https://github.com/apache/datafusion-sqlparser-rs/pull/1741) (iffyio) +- BigQuery: Add support for `CREATE SCHEMA` options [#1742](https://github.com/apache/datafusion-sqlparser-rs/pull/1742) (iffyio) +- Snowflake: Support dollar quoted comments [#1755](https://github.com/apache/datafusion-sqlparser-rs/pull/1755) +- Add LOCK operation for ALTER TABLE [#1768](https://github.com/apache/datafusion-sqlparser-rs/pull/1768) (MohamedAbdeen21) +- Add support for `RAISE` statement [#1766](https://github.com/apache/datafusion-sqlparser-rs/pull/1766) (iffyio) +- Add GLOBAL context/modifier to SET statements [#1767](https://github.com/apache/datafusion-sqlparser-rs/pull/1767) (MohamedAbdeen21) +- Parse `SUBSTR` as alias for `SUBSTRING` [#1769](https://github.com/apache/datafusion-sqlparser-rs/pull/1769) (mvzink) +- SET statements: scope modifier for multiple assignments [#1772](https://github.com/apache/datafusion-sqlparser-rs/pull/1772) (MohamedAbdeen21) +- Support qualified column names in `MATCH AGAINST` clause [#1774](https://github.com/apache/datafusion-sqlparser-rs/pull/1774) (tomershaniii) +- Mysql: Add support for := operator [#1779](https://github.com/apache/datafusion-sqlparser-rs/pull/1779) (barsela1) +- Add cipherstash-proxy to list of users in README.md [#1782](https://github.com/apache/datafusion-sqlparser-rs/pull/1782) (coderdan) +- Fix typos [#1785](https://github.com/apache/datafusion-sqlparser-rs/pull/1785) (jayvdb) +- Add support for Databricks TIMESTAMP_NTZ. [#1781](https://github.com/apache/datafusion-sqlparser-rs/pull/1781) (romanb) +- Enable double-dot-notation for mssql. [#1787](https://github.com/apache/datafusion-sqlparser-rs/pull/1787) (romanb) +- Fix: Snowflake ALTER SESSION cannot be followed by other statements. [#1786](https://github.com/apache/datafusion-sqlparser-rs/pull/1786) (romanb) +- Add GreptimeDB to the "Users" in README [#1788](https://github.com/apache/datafusion-sqlparser-rs/pull/1788) (MichaelScofield) +- Extend snowflake grant options support [#1794](https://github.com/apache/datafusion-sqlparser-rs/pull/1794) (yoavcloud) +- Fix clippy lint on rust 1.86 [#1796](https://github.com/apache/datafusion-sqlparser-rs/pull/1796) (iffyio) +- Allow single quotes in EXTRACT() for Redshift. [#1795](https://github.com/apache/datafusion-sqlparser-rs/pull/1795) (romanb) +- MSSQL: Add support for functionality `MERGE` output clause [#1790](https://github.com/apache/datafusion-sqlparser-rs/pull/1790) (dilovancelik) +- Support additional DuckDB integer types such as HUGEINT, UHUGEINT, etc [#1797](https://github.com/apache/datafusion-sqlparser-rs/pull/1797) (alexander-beedie) +- Add support for MSSQL IF/ELSE statements. [#1791](https://github.com/apache/datafusion-sqlparser-rs/pull/1791) (romanb) +- Allow literal backslash escapes for string literals in Redshift dialect. [#1801](https://github.com/apache/datafusion-sqlparser-rs/pull/1801) (romanb) +- Add support for MySQL's STRAIGHT_JOIN join operator. [#1802](https://github.com/apache/datafusion-sqlparser-rs/pull/1802) (romanb) +- Snowflake COPY INTO target columns, select items and optional alias [#1805](https://github.com/apache/datafusion-sqlparser-rs/pull/1805) (yoavcloud) +- Fix tokenization of qualified identifiers with numeric prefix. [#1803](https://github.com/apache/datafusion-sqlparser-rs/pull/1803) (romanb) +- Add support for `INHERITS` option in `CREATE TABLE` statement [#1806](https://github.com/apache/datafusion-sqlparser-rs/pull/1806) (LucaCappelletti94) +- Add `DROP TRIGGER` support for SQL Server [#1813](https://github.com/apache/datafusion-sqlparser-rs/pull/1813) (aharpervc) +- Snowflake: support nested join without parentheses [#1799](https://github.com/apache/datafusion-sqlparser-rs/pull/1799) (barsela1) +- Add support for parenthesized subquery as `IN` predicate [#1793](https://github.com/apache/datafusion-sqlparser-rs/pull/1793) (adamchainz) +- Fix `STRAIGHT_JOIN` constraint when table alias is absent [#1812](https://github.com/apache/datafusion-sqlparser-rs/pull/1812) (killertux) +- Add support for `PRINT` statement for SQL Server [#1811](https://github.com/apache/datafusion-sqlparser-rs/pull/1811) (aharpervc) +- enable `supports_filter_during_aggregation` for Generic dialect [#1815](https://github.com/apache/datafusion-sqlparser-rs/pull/1815) (goldmedal) +- Add support for `XMLTABLE` [#1817](https://github.com/apache/datafusion-sqlparser-rs/pull/1817) (lovasoa) +- Add `CREATE FUNCTION` support for SQL Server [#1808](https://github.com/apache/datafusion-sqlparser-rs/pull/1808) (aharpervc) +- Add `OR ALTER` support for `CREATE VIEW` [#1818](https://github.com/apache/datafusion-sqlparser-rs/pull/1818) (aharpervc) +- Add `DECLARE ... CURSOR FOR` support for SQL Server [#1821](https://github.com/apache/datafusion-sqlparser-rs/pull/1821) (aharpervc) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 8 Roman Borschel + 5 Ifeanyi Ubah + 5 Michael Victor Zink + 4 Andrew Harper + 4 Mohamed Abdeen + 3 Ophir LOJKINE + 2 Luca Cappelletti + 2 Yoav Cohen + 2 bar sela + 1 Adam Johnson + 1 Aleksei Piianin + 1 Alexander Beedie + 1 Bruno Clemente + 1 Dan Draper + 1 DilovanCelik + 1 Jax Liu + 1 John Vandenberg + 1 LFC + 1 tomershaniii +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + From a5b9821d1d2fa9c0b8ee73b698a2b0e5f138beaf Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Tue, 29 Apr 2025 15:55:22 -0400 Subject: [PATCH 114/286] Update `56.0.0` Changelog with latest commits (#1832) --- changelog/0.56.0.md | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/changelog/0.56.0.md b/changelog/0.56.0.md index 9e6289e88..b3c8a67aa 100644 --- a/changelog/0.56.0.md +++ b/changelog/0.56.0.md @@ -19,7 +19,7 @@ under the License. # sqlparser-rs 0.56.0 Changelog -This release consists of 45 commits from 19 contributors. See credits at the end of this changelog for more information. +This release consists of 48 commits from 19 contributors. See credits at the end of this changelog for more information. **Other:** @@ -69,6 +69,8 @@ This release consists of 45 commits from 19 contributors. See credits at the end - Add `CREATE FUNCTION` support for SQL Server [#1808](https://github.com/apache/datafusion-sqlparser-rs/pull/1808) (aharpervc) - Add `OR ALTER` support for `CREATE VIEW` [#1818](https://github.com/apache/datafusion-sqlparser-rs/pull/1818) (aharpervc) - Add `DECLARE ... CURSOR FOR` support for SQL Server [#1821](https://github.com/apache/datafusion-sqlparser-rs/pull/1821) (aharpervc) +- Handle missing login in changelog generate script [#1823](https://github.com/apache/datafusion-sqlparser-rs/pull/1823) (iffyio) +- Snowflake: Add support for `CONNECT_BY_ROOT` [#1780](https://github.com/apache/datafusion-sqlparser-rs/pull/1780) (tomershaniii) ## Credits @@ -76,14 +78,15 @@ Thank you to everyone who contributed to this release. Here is a breakdown of co ``` 8 Roman Borschel - 5 Ifeanyi Ubah + 6 Ifeanyi Ubah + 5 Andrew Harper 5 Michael Victor Zink - 4 Andrew Harper 4 Mohamed Abdeen 3 Ophir LOJKINE 2 Luca Cappelletti 2 Yoav Cohen 2 bar sela + 2 tomershaniii 1 Adam Johnson 1 Aleksei Piianin 1 Alexander Beedie @@ -93,7 +96,6 @@ Thank you to everyone who contributed to this release. Here is a breakdown of co 1 Jax Liu 1 John Vandenberg 1 LFC - 1 tomershaniii ``` Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. From e5d2215267c000fbdb453a0345e9878311086269 Mon Sep 17 00:00:00 2001 From: Simon Vandel Sillesen Date: Fri, 2 May 2025 05:13:47 +0200 Subject: [PATCH 115/286] Support some of pipe operators (#1759) --- src/ast/mod.rs | 34 ++++---- src/ast/query.rs | 155 ++++++++++++++++++++++++++++++++++++ src/ast/spans.rs | 9 ++- src/dialect/bigquery.rs | 4 + src/dialect/mod.rs | 14 ++++ src/keywords.rs | 2 + src/parser/mod.rs | 116 +++++++++++++++++++++++++++ src/tokenizer.rs | 6 ++ tests/sqlparser_common.rs | 84 +++++++++++++++++++ tests/sqlparser_mssql.rs | 4 + tests/sqlparser_mysql.rs | 15 ++++ tests/sqlparser_postgres.rs | 5 ++ 12 files changed, 427 insertions(+), 21 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d65889810..b496403c4 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -66,23 +66,23 @@ pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; pub use self::query::{ AfterMatchSkip, ConnectBy, Cte, CteAsMaterialized, Distinct, EmptyMatchesMode, - ExceptSelectItem, ExcludeSelectItem, ExprWithAlias, Fetch, ForClause, ForJson, ForXml, - FormatClause, GroupByExpr, GroupByWithModifier, IdentWithAlias, IlikeSelectItem, - InputFormatClause, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, - JsonTableColumn, JsonTableColumnErrorHandling, JsonTableNamedColumn, JsonTableNestedColumn, - LateralView, LimitClause, LockClause, LockType, MatchRecognizePattern, MatchRecognizeSymbol, - Measure, NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, OffsetRows, - OpenJsonTableColumn, OrderBy, OrderByExpr, OrderByKind, OrderByOptions, PivotValueSource, - ProjectionSelect, Query, RenameSelectItem, RepetitionQuantifier, ReplaceSelectElement, - ReplaceSelectItem, RowsPerMatch, Select, SelectFlavor, SelectInto, SelectItem, - SelectItemQualifiedWildcardKind, SetExpr, SetOperator, SetQuantifier, Setting, - SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, TableFunctionArgs, - TableIndexHintForClause, TableIndexHintType, TableIndexHints, TableIndexType, TableSample, - TableSampleBucket, TableSampleKind, TableSampleMethod, TableSampleModifier, - TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, TableSampleUnit, TableVersion, - TableWithJoins, Top, TopQuantity, UpdateTableFromKind, ValueTableMode, Values, - WildcardAdditionalOptions, With, WithFill, XmlNamespaceDefinition, XmlPassingArgument, - XmlPassingClause, XmlTableColumn, XmlTableColumnOption, + ExceptSelectItem, ExcludeSelectItem, ExprWithAlias, ExprWithAliasAndOrderBy, Fetch, ForClause, + ForJson, ForXml, FormatClause, GroupByExpr, GroupByWithModifier, IdentWithAlias, + IlikeSelectItem, InputFormatClause, Interpolate, InterpolateExpr, Join, JoinConstraint, + JoinOperator, JsonTableColumn, JsonTableColumnErrorHandling, JsonTableNamedColumn, + JsonTableNestedColumn, LateralView, LimitClause, LockClause, LockType, MatchRecognizePattern, + MatchRecognizeSymbol, Measure, NamedWindowDefinition, NamedWindowExpr, NonBlock, Offset, + OffsetRows, OpenJsonTableColumn, OrderBy, OrderByExpr, OrderByKind, OrderByOptions, + PipeOperator, PivotValueSource, ProjectionSelect, Query, RenameSelectItem, + RepetitionQuantifier, ReplaceSelectElement, ReplaceSelectItem, RowsPerMatch, Select, + SelectFlavor, SelectInto, SelectItem, SelectItemQualifiedWildcardKind, SetExpr, SetOperator, + SetQuantifier, Setting, SymbolDefinition, Table, TableAlias, TableAliasColumnDef, TableFactor, + TableFunctionArgs, TableIndexHintForClause, TableIndexHintType, TableIndexHints, + TableIndexType, TableSample, TableSampleBucket, TableSampleKind, TableSampleMethod, + TableSampleModifier, TableSampleQuantity, TableSampleSeed, TableSampleSeedModifier, + TableSampleUnit, TableVersion, TableWithJoins, Top, TopQuantity, UpdateTableFromKind, + ValueTableMode, Values, WildcardAdditionalOptions, With, WithFill, XmlNamespaceDefinition, + XmlPassingArgument, XmlPassingClause, XmlTableColumn, XmlTableColumnOption, }; pub use self::trigger::{ diff --git a/src/ast/query.rs b/src/ast/query.rs index 982985ec3..a90b61668 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -62,6 +62,9 @@ pub struct Query { /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/select/format) /// (ClickHouse-specific) pub format_clause: Option, + + /// Pipe operator + pub pipe_operators: Vec, } impl fmt::Display for Query { @@ -92,6 +95,9 @@ impl fmt::Display for Query { if let Some(ref format) = self.format_clause { write!(f, " {}", format)?; } + for pipe_operator in &self.pipe_operators { + write!(f, " |> {}", pipe_operator)?; + } Ok(()) } } @@ -1004,6 +1010,26 @@ impl fmt::Display for ExprWithAlias { } } +/// An expression optionally followed by an alias and order by options. +/// +/// Example: +/// ```sql +/// 42 AS myint ASC +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ExprWithAliasAndOrderBy { + pub expr: ExprWithAlias, + pub order_by: OrderByOptions, +} + +impl fmt::Display for ExprWithAliasAndOrderBy { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}{}", self.expr, self.order_by) + } +} + /// Arguments to a table-valued function #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -2513,6 +2539,135 @@ impl fmt::Display for OffsetRows { } } +/// Pipe syntax, first introduced in Google BigQuery. +/// Example: +/// +/// ```sql +/// FROM Produce +/// |> WHERE sales > 0 +/// |> AGGREGATE SUM(sales) AS total_sales, COUNT(*) AS num_sales +/// GROUP BY item; +/// ``` +/// +/// See +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum PipeOperator { + /// Limits the number of rows to return in a query, with an optional OFFSET clause to skip over rows. + /// + /// Syntax: `|> LIMIT [OFFSET ]` + /// + /// See more at + Limit { expr: Expr, offset: Option }, + /// Filters the results of the input table. + /// + /// Syntax: `|> WHERE ` + /// + /// See more at + Where { expr: Expr }, + /// `ORDER BY [ASC|DESC], ...` + OrderBy { exprs: Vec }, + /// Produces a new table with the listed columns, similar to the outermost SELECT clause in a table subquery in standard syntax. + /// + /// Syntax `|> SELECT [[AS] alias], ...` + /// + /// See more at + Select { exprs: Vec }, + /// Propagates the existing table and adds computed columns, similar to SELECT *, new_column in standard syntax. + /// + /// Syntax: `|> EXTEND [[AS] alias], ...` + /// + /// See more at + Extend { exprs: Vec }, + /// Replaces the value of a column in the current table, similar to SELECT * REPLACE (expression AS column) in standard syntax. + /// + /// Syntax: `|> SET = , ...` + /// + /// See more at + Set { assignments: Vec }, + /// Removes listed columns from the current table, similar to SELECT * EXCEPT (column) in standard syntax. + /// + /// Syntax: `|> DROP , ...` + /// + /// See more at + Drop { columns: Vec }, + /// Introduces a table alias for the input table, similar to applying the AS alias clause on a table subquery in standard syntax. + /// + /// Syntax: `|> AS ` + /// + /// See more at + As { alias: Ident }, + /// Performs aggregation on data across grouped rows or an entire table. + /// + /// Syntax: `|> AGGREGATE [[AS] alias], ...` + /// + /// Syntax: + /// ```norust + /// |> AGGREGATE [ [[AS] alias], ...] + /// GROUP BY [AS alias], ... + /// ``` + /// + /// See more at + Aggregate { + full_table_exprs: Vec, + group_by_expr: Vec, + }, +} + +impl fmt::Display for PipeOperator { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + PipeOperator::Select { exprs } => { + write!(f, "SELECT {}", display_comma_separated(exprs.as_slice())) + } + PipeOperator::Extend { exprs } => { + write!(f, "EXTEND {}", display_comma_separated(exprs.as_slice())) + } + PipeOperator::Set { assignments } => { + write!(f, "SET {}", display_comma_separated(assignments.as_slice())) + } + PipeOperator::Drop { columns } => { + write!(f, "DROP {}", display_comma_separated(columns.as_slice())) + } + PipeOperator::As { alias } => { + write!(f, "AS {}", alias) + } + PipeOperator::Limit { expr, offset } => { + write!(f, "LIMIT {}", expr)?; + if let Some(offset) = offset { + write!(f, " OFFSET {}", offset)?; + } + Ok(()) + } + PipeOperator::Aggregate { + full_table_exprs, + group_by_expr, + } => { + write!(f, "AGGREGATE")?; + if !full_table_exprs.is_empty() { + write!( + f, + " {}", + display_comma_separated(full_table_exprs.as_slice()) + )?; + } + if !group_by_expr.is_empty() { + write!(f, " GROUP BY {}", display_comma_separated(group_by_expr))?; + } + Ok(()) + } + + PipeOperator::Where { expr } => { + write!(f, "WHERE {}", expr) + } + PipeOperator::OrderBy { exprs } => { + write!(f, "ORDER BY {}", display_comma_separated(exprs.as_slice())) + } + } + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 28d479f30..661cd03da 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -98,10 +98,11 @@ impl Spanned for Query { order_by, limit_clause, fetch, - locks: _, // todo - for_clause: _, // todo, mssql specific - settings: _, // todo, clickhouse specific - format_clause: _, // todo, clickhouse specific + locks: _, // todo + for_clause: _, // todo, mssql specific + settings: _, // todo, clickhouse specific + format_clause: _, // todo, clickhouse specific + pipe_operators: _, // todo bigquery specific } = self; union_spans( diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index 49fb24f19..68ca1390a 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -136,6 +136,10 @@ impl Dialect for BigQueryDialect { fn is_column_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { !RESERVED_FOR_COLUMN_ALIAS.contains(kw) } + + fn supports_pipe_operator(&self) -> bool { + true + } } impl BigQueryDialect { diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index b2dff065e..b754a04f1 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -518,6 +518,20 @@ pub trait Dialect: Debug + Any { false } + /// Return true if the dialect supports pipe operator. + /// + /// Example: + /// ```sql + /// SELECT * + /// FROM table + /// |> limit 1 + /// ``` + /// + /// See + fn supports_pipe_operator(&self) -> bool { + false + } + /// Does the dialect support MySQL-style `'user'@'host'` grantee syntax? fn supports_user_host_grantee(&self) -> bool { false diff --git a/src/keywords.rs b/src/keywords.rs index 15a6f91ad..d2ccbb2c6 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -83,6 +83,7 @@ define_keywords!( ADMIN, AFTER, AGAINST, + AGGREGATE, AGGREGATION, ALERT, ALGORITHM, @@ -338,6 +339,7 @@ define_keywords!( EXPLAIN, EXPLICIT, EXPORT, + EXTEND, EXTENDED, EXTENSION, EXTERNAL, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 03ea91faf..2e37f1bc9 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1149,6 +1149,25 @@ impl<'a> Parser<'a> { self.parse_subexpr(self.dialect.prec_unknown()) } + pub fn parse_expr_with_alias_and_order_by( + &mut self, + ) -> Result { + let expr = self.parse_expr()?; + + fn validator(explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { + explicit || !&[Keyword::ASC, Keyword::DESC, Keyword::GROUP].contains(kw) + } + let alias = self.parse_optional_alias_inner(None, validator)?; + let order_by = OrderByOptions { + asc: self.parse_asc_desc(), + nulls_first: None, + }; + Ok(ExprWithAliasAndOrderBy { + expr: ExprWithAlias { expr, alias }, + order_by, + }) + } + /// Parse tokens until the precedence changes. pub fn parse_subexpr(&mut self, precedence: u8) -> Result { let _guard = self.recursion_counter.try_decrease()?; @@ -10571,6 +10590,7 @@ impl<'a> Parser<'a> { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], } .into()) } else if self.parse_keyword(Keyword::UPDATE) { @@ -10584,6 +10604,7 @@ impl<'a> Parser<'a> { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], } .into()) } else if self.parse_keyword(Keyword::DELETE) { @@ -10597,6 +10618,7 @@ impl<'a> Parser<'a> { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], } .into()) } else { @@ -10637,6 +10659,12 @@ impl<'a> Parser<'a> { None }; + let pipe_operators = if self.dialect.supports_pipe_operator() { + self.parse_pipe_operators()? + } else { + Vec::new() + }; + Ok(Query { with, body, @@ -10647,11 +10675,98 @@ impl<'a> Parser<'a> { for_clause, settings, format_clause, + pipe_operators, } .into()) } } + fn parse_pipe_operators(&mut self) -> Result, ParserError> { + let mut pipe_operators = Vec::new(); + + while self.consume_token(&Token::VerticalBarRightAngleBracket) { + let kw = self.expect_one_of_keywords(&[ + Keyword::SELECT, + Keyword::EXTEND, + Keyword::SET, + Keyword::DROP, + Keyword::AS, + Keyword::WHERE, + Keyword::LIMIT, + Keyword::AGGREGATE, + Keyword::ORDER, + ])?; + match kw { + Keyword::SELECT => { + let exprs = self.parse_comma_separated(Parser::parse_select_item)?; + pipe_operators.push(PipeOperator::Select { exprs }) + } + Keyword::EXTEND => { + let exprs = self.parse_comma_separated(Parser::parse_select_item)?; + pipe_operators.push(PipeOperator::Extend { exprs }) + } + Keyword::SET => { + let assignments = self.parse_comma_separated(Parser::parse_assignment)?; + pipe_operators.push(PipeOperator::Set { assignments }) + } + Keyword::DROP => { + let columns = self.parse_identifiers()?; + pipe_operators.push(PipeOperator::Drop { columns }) + } + Keyword::AS => { + let alias = self.parse_identifier()?; + pipe_operators.push(PipeOperator::As { alias }) + } + Keyword::WHERE => { + let expr = self.parse_expr()?; + pipe_operators.push(PipeOperator::Where { expr }) + } + Keyword::LIMIT => { + let expr = self.parse_expr()?; + let offset = if self.parse_keyword(Keyword::OFFSET) { + Some(self.parse_expr()?) + } else { + None + }; + pipe_operators.push(PipeOperator::Limit { expr, offset }) + } + Keyword::AGGREGATE => { + let full_table_exprs = if self.peek_keyword(Keyword::GROUP) { + vec![] + } else { + self.parse_comma_separated(|parser| { + parser.parse_expr_with_alias_and_order_by() + })? + }; + + let group_by_expr = if self.parse_keywords(&[Keyword::GROUP, Keyword::BY]) { + self.parse_comma_separated(|parser| { + parser.parse_expr_with_alias_and_order_by() + })? + } else { + vec![] + }; + + pipe_operators.push(PipeOperator::Aggregate { + full_table_exprs, + group_by_expr, + }) + } + Keyword::ORDER => { + self.expect_one_of_keywords(&[Keyword::BY])?; + let exprs = self.parse_comma_separated(Parser::parse_order_by_expr)?; + pipe_operators.push(PipeOperator::OrderBy { exprs }) + } + unhandled => { + return Err(ParserError::ParserError(format!( + "`expect_one_of_keywords` further up allowed unhandled keyword: {unhandled:?}" + ))) + } + } + } + Ok(pipe_operators) + } + fn parse_settings(&mut self) -> Result>, ParserError> { let settings = if dialect_of!(self is ClickHouseDialect|GenericDialect) && self.parse_keyword(Keyword::SETTINGS) @@ -12122,6 +12237,7 @@ impl<'a> Parser<'a> { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }), alias, }) diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 13bce0c0d..4fad54624 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -246,6 +246,8 @@ pub enum Token { ShiftLeftVerticalBar, /// `|>> PostgreSQL/Redshift geometrical binary operator (Is strictly above?) VerticalBarShiftRight, + /// `|> BigQuery pipe operator + VerticalBarRightAngleBracket, /// `#>>`, extracts JSON sub-object at the specified path as text HashLongArrow, /// jsonb @> jsonb -> boolean: Test whether left json contains the right json @@ -359,6 +361,7 @@ impl fmt::Display for Token { Token::AmpersandRightAngleBracket => f.write_str("&>"), Token::AmpersandLeftAngleBracketVerticalBar => f.write_str("&<|"), Token::VerticalBarAmpersandRightAngleBracket => f.write_str("|&>"), + Token::VerticalBarRightAngleBracket => f.write_str("|>"), Token::TwoWayArrow => f.write_str("<->"), Token::LeftAngleBracketCaret => f.write_str("<^"), Token::RightAngleBracketCaret => f.write_str(">^"), @@ -1403,6 +1406,9 @@ impl<'a> Tokenizer<'a> { _ => self.start_binop_opt(chars, "|>", None), } } + Some('>') if self.dialect.supports_pipe_operator() => { + self.consume_for_binop(chars, "|>", Token::VerticalBarRightAngleBracket) + } // Bitshift '|' operator _ => self.start_binop(chars, "|", Token::Pipe), } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index fa2346c2c..6d99929d5 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -489,6 +489,7 @@ fn parse_update_set_from() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }), alias: Some(TableAlias { name: Ident::new("t2"), @@ -4310,6 +4311,7 @@ fn parse_create_table_as_table() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }); match verified_stmt(sql1) { @@ -4335,6 +4337,7 @@ fn parse_create_table_as_table() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }); match verified_stmt(sql2) { @@ -6332,6 +6335,7 @@ fn parse_interval_and_or_xor() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }))]; assert_eq!(actual_ast, expected_ast); @@ -9467,6 +9471,7 @@ fn parse_merge() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }), alias: Some(TableAlias { name: Ident { @@ -11344,6 +11349,7 @@ fn parse_unload() { order_by: None, settings: None, format_clause: None, + pipe_operators: vec![], }), to: Ident { value: "s3://...".to_string(), @@ -12564,6 +12570,7 @@ fn test_extract_seconds_ok() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }))]; assert_eq!(actual_ast, expected_ast); @@ -14641,6 +14648,7 @@ fn test_select_from_first() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }; assert_eq!(expected, ast); assert_eq!(ast.to_string(), q); @@ -15020,6 +15028,82 @@ fn parse_set_names() { dialects.verified_stmt("SET NAMES UTF8 COLLATE bogus"); } +#[test] +fn parse_pipeline_operator() { + let dialects = all_dialects_where(|d| d.supports_pipe_operator()); + + // select pipe operator + dialects.verified_stmt("SELECT * FROM users |> SELECT id"); + dialects.verified_stmt("SELECT * FROM users |> SELECT id, name"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> SELECT id user_id", + "SELECT * FROM users |> SELECT id AS user_id", + ); + dialects.verified_stmt("SELECT * FROM users |> SELECT id AS user_id"); + + // extend pipe operator + dialects.verified_stmt("SELECT * FROM users |> EXTEND id + 1 AS new_id"); + dialects.verified_stmt("SELECT * FROM users |> EXTEND id AS new_id, name AS new_name"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> EXTEND id user_id", + "SELECT * FROM users |> EXTEND id AS user_id", + ); + + // set pipe operator + dialects.verified_stmt("SELECT * FROM users |> SET id = id + 1"); + dialects.verified_stmt("SELECT * FROM users |> SET id = id + 1, name = name + ' Doe'"); + + // drop pipe operator + dialects.verified_stmt("SELECT * FROM users |> DROP id"); + dialects.verified_stmt("SELECT * FROM users |> DROP id, name"); + + // as pipe operator + dialects.verified_stmt("SELECT * FROM users |> AS new_users"); + + // limit pipe operator + dialects.verified_stmt("SELECT * FROM users |> LIMIT 10"); + dialects.verified_stmt("SELECT * FROM users |> LIMIT 10 OFFSET 5"); + dialects.verified_stmt("SELECT * FROM users |> LIMIT 10 |> LIMIT 5"); + dialects.verified_stmt("SELECT * FROM users |> LIMIT 10 |> WHERE true"); + + // where pipe operator + dialects.verified_stmt("SELECT * FROM users |> WHERE id = 1"); + dialects.verified_stmt("SELECT * FROM users |> WHERE id = 1 AND name = 'John'"); + dialects.verified_stmt("SELECT * FROM users |> WHERE id = 1 OR name = 'John'"); + + // aggregate pipe operator full table + dialects.verified_stmt("SELECT * FROM users |> AGGREGATE COUNT(*)"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> AGGREGATE COUNT(*) total_users", + "SELECT * FROM users |> AGGREGATE COUNT(*) AS total_users", + ); + dialects.verified_stmt("SELECT * FROM users |> AGGREGATE COUNT(*) AS total_users"); + dialects.verified_stmt("SELECT * FROM users |> AGGREGATE COUNT(*), MIN(id)"); + + // aggregate pipe opeprator with grouping + dialects.verified_stmt( + "SELECT * FROM users |> AGGREGATE SUM(o_totalprice) AS price, COUNT(*) AS cnt GROUP BY EXTRACT(YEAR FROM o_orderdate) AS year", + ); + dialects.verified_stmt( + "SELECT * FROM users |> AGGREGATE GROUP BY EXTRACT(YEAR FROM o_orderdate) AS year", + ); + dialects + .verified_stmt("SELECT * FROM users |> AGGREGATE GROUP BY EXTRACT(YEAR FROM o_orderdate)"); + dialects.verified_stmt("SELECT * FROM users |> AGGREGATE GROUP BY a, b"); + dialects.verified_stmt("SELECT * FROM users |> AGGREGATE SUM(c) GROUP BY a, b"); + dialects.verified_stmt("SELECT * FROM users |> AGGREGATE SUM(c) ASC"); + + // order by pipe operator + dialects.verified_stmt("SELECT * FROM users |> ORDER BY id ASC"); + dialects.verified_stmt("SELECT * FROM users |> ORDER BY id DESC"); + dialects.verified_stmt("SELECT * FROM users |> ORDER BY id DESC, name ASC"); + + // many pipes + dialects.verified_stmt( + "SELECT * FROM CustomerOrders |> AGGREGATE SUM(cost) AS total_cost GROUP BY customer_id, state, item_type |> EXTEND COUNT(*) OVER (PARTITION BY customer_id) AS num_orders |> WHERE num_orders > 1 |> AGGREGATE AVG(total_cost) AS average GROUP BY state DESC, item_type ASC", + ); +} + #[test] fn parse_multiple_set_statements() -> Result<(), ParserError> { let dialects = all_dialects_where(|d| d.supports_comma_separated_set_assignments()); diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index ef6103474..b2d5210c2 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -114,6 +114,7 @@ fn parse_create_procedure() { order_by: None, settings: None, format_clause: None, + pipe_operators: vec![], body: Box::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), distinct: None, @@ -1252,6 +1253,7 @@ fn parse_substring_in_select() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }), query ); @@ -1354,6 +1356,8 @@ fn parse_mssql_declare() { order_by: None, settings: None, format_clause: None, + pipe_operators: vec![], + body: Box::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), distinct: None, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index f74248b86..990107b25 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1113,6 +1113,7 @@ fn parse_escaped_quote_identifiers_with_escape() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })) ); } @@ -1165,6 +1166,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })) ); } @@ -1211,6 +1213,7 @@ fn parse_escaped_backticks_with_escape() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })) ); } @@ -1261,6 +1264,7 @@ fn parse_escaped_backticks_with_no_escape() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })) ); } @@ -1436,6 +1440,7 @@ fn parse_simple_insert() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1484,6 +1489,7 @@ fn parse_ignore_insert() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1532,6 +1538,7 @@ fn parse_priority_insert() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1577,6 +1584,7 @@ fn parse_priority_insert() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1624,6 +1632,7 @@ fn parse_insert_as() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1686,6 +1695,7 @@ fn parse_insert_as() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1735,6 +1745,7 @@ fn parse_replace_insert() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1775,6 +1786,7 @@ fn parse_empty_row_insert() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -1839,6 +1851,7 @@ fn parse_insert_with_on_duplicate_update() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), source ); @@ -2745,6 +2758,7 @@ fn parse_substring_in_select() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], }), query ); @@ -3051,6 +3065,7 @@ fn parse_hex_string_introducer() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })) ) } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 27fc7fa17..4ad8e00cc 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1326,6 +1326,7 @@ fn parse_copy_to() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), to: true, target: CopyTarget::File { @@ -2994,6 +2995,7 @@ fn parse_array_subquery_expr() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), filter: None, null_treatment: None, @@ -4785,6 +4787,7 @@ fn test_simple_postgres_insert_with_alias() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), assignments: vec![], partitioned: None, @@ -4856,6 +4859,7 @@ fn test_simple_postgres_insert_with_alias() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), assignments: vec![], partitioned: None, @@ -4925,6 +4929,7 @@ fn test_simple_insert_with_quoted_alias() { for_clause: None, settings: None, format_clause: None, + pipe_operators: vec![], })), assignments: vec![], partitioned: None, From 483394cd1a29739ada14cb061da3f5cfc5e33506 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Fri, 2 May 2025 05:16:24 +0200 Subject: [PATCH 116/286] Added support for `DROP DOMAIN` (#1828) --- src/ast/mod.rs | 36 ++++++++++++++++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 16 ++++++++++ tests/sqlparser_postgres.rs | 60 +++++++++++++++++++++++++++++++++++++ 5 files changed, 114 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index b496403c4..b14392665 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3321,6 +3321,14 @@ pub enum Statement { drop_behavior: Option, }, /// ```sql + /// DROP DOMAIN + /// ``` + /// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-dropdomain.html) + /// + /// DROP DOMAIN [ IF EXISTS ] name [, ...] [ CASCADE | RESTRICT ] + /// + DropDomain(DropDomain), + /// ```sql /// DROP PROCEDURE /// ``` DropProcedure { @@ -5094,6 +5102,21 @@ impl fmt::Display for Statement { } Ok(()) } + Statement::DropDomain(DropDomain { + if_exists, + name, + drop_behavior, + }) => { + write!( + f, + "DROP DOMAIN{} {name}", + if *if_exists { " IF EXISTS" } else { "" }, + )?; + if let Some(op) = drop_behavior { + write!(f, " {op}")?; + } + Ok(()) + } Statement::DropProcedure { if_exists, proc_desc, @@ -6829,6 +6852,19 @@ impl fmt::Display for CloseCursor { } } +/// A Drop Domain statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct DropDomain { + /// Whether to drop the domain if it exists + pub if_exists: bool, + /// The name of the domain to drop + pub name: ObjectName, + /// The behavior to apply when dropping the domain + pub drop_behavior: Option, +} + /// A function call #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 661cd03da..33bc07392 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -455,6 +455,7 @@ impl Spanned for Statement { Statement::DetachDuckDBDatabase { .. } => Span::empty(), Statement::Drop { .. } => Span::empty(), Statement::DropFunction { .. } => Span::empty(), + Statement::DropDomain { .. } => Span::empty(), Statement::DropProcedure { .. } => Span::empty(), Statement::DropSecret { .. } => Span::empty(), Statement::Declare { .. } => Span::empty(), diff --git a/src/keywords.rs b/src/keywords.rs index d2ccbb2c6..32612ccd5 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -288,6 +288,7 @@ define_keywords!( DISTRIBUTE, DIV, DO, + DOMAIN, DOUBLE, DOW, DOY, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2e37f1bc9..0d74235b2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6070,6 +6070,8 @@ impl<'a> Parser<'a> { return self.parse_drop_policy(); } else if self.parse_keyword(Keyword::CONNECTOR) { return self.parse_drop_connector(); + } else if self.parse_keyword(Keyword::DOMAIN) { + return self.parse_drop_domain(); } else if self.parse_keyword(Keyword::PROCEDURE) { return self.parse_drop_procedure(); } else if self.parse_keyword(Keyword::SECRET) { @@ -6165,6 +6167,20 @@ impl<'a> Parser<'a> { Ok(Statement::DropConnector { if_exists, name }) } + /// ```sql + /// DROP DOMAIN [ IF EXISTS ] name [ CASCADE | RESTRICT ] + /// ``` + fn parse_drop_domain(&mut self) -> Result { + let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); + let name = self.parse_object_name(false)?; + let drop_behavior = self.parse_optional_drop_behavior(); + Ok(Statement::DropDomain(DropDomain { + if_exists, + name, + drop_behavior, + })) + } + /// ```sql /// DROP PROCEDURE [ IF EXISTS ] name [ ( [ [ argmode ] [ argname ] argtype [, ...] ] ) ] [, ...] /// [ CASCADE | RESTRICT ] diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 4ad8e00cc..6c008c84d 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -4241,6 +4241,66 @@ fn parse_drop_function() { ); } +#[test] +fn parse_drop_domain() { + let sql = "DROP DOMAIN IF EXISTS jpeg_domain"; + assert_eq!( + pg().verified_stmt(sql), + Statement::DropDomain(DropDomain { + if_exists: true, + name: ObjectName::from(vec![Ident { + value: "jpeg_domain".to_string(), + quote_style: None, + span: Span::empty(), + }]), + drop_behavior: None + }) + ); + + let sql = "DROP DOMAIN jpeg_domain"; + assert_eq!( + pg().verified_stmt(sql), + Statement::DropDomain(DropDomain { + if_exists: false, + name: ObjectName::from(vec![Ident { + value: "jpeg_domain".to_string(), + quote_style: None, + span: Span::empty(), + }]), + drop_behavior: None + }) + ); + + let sql = "DROP DOMAIN IF EXISTS jpeg_domain CASCADE"; + assert_eq!( + pg().verified_stmt(sql), + Statement::DropDomain(DropDomain { + if_exists: true, + name: ObjectName::from(vec![Ident { + value: "jpeg_domain".to_string(), + quote_style: None, + span: Span::empty(), + }]), + drop_behavior: Some(DropBehavior::Cascade) + }) + ); + + let sql = "DROP DOMAIN IF EXISTS jpeg_domain RESTRICT"; + + assert_eq!( + pg().verified_stmt(sql), + Statement::DropDomain(DropDomain { + if_exists: true, + name: ObjectName::from(vec![Ident { + value: "jpeg_domain".to_string(), + quote_style: None, + span: Span::empty(), + }]), + drop_behavior: Some(DropBehavior::Restrict) + }) + ); +} + #[test] fn parse_drop_procedure() { let sql = "DROP PROCEDURE IF EXISTS test_proc"; From a464f8e8d7a5057e4e5b8046b0f619acdf7fce74 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Thu, 1 May 2025 23:25:30 -0400 Subject: [PATCH 117/286] Improve support for cursors for SQL Server (#1831) Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 90 +++++++++++++++++++++++++++++++++++++-- src/ast/spans.rs | 31 +++++++++++--- src/keywords.rs | 2 + src/parser/mod.rs | 70 +++++++++++++++++++++++++++--- src/test_utils.rs | 20 +++++++++ tests/sqlparser_common.rs | 12 ++++++ tests/sqlparser_mssql.rs | 84 +++++++++++++++++++++++++++++++++++- 7 files changed, 289 insertions(+), 20 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index b14392665..582922a3e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2228,7 +2228,33 @@ impl fmt::Display for IfStatement { } } -/// A block within a [Statement::Case] or [Statement::If]-like statement +/// A `WHILE` statement. +/// +/// Example: +/// ```sql +/// WHILE @@FETCH_STATUS = 0 +/// BEGIN +/// FETCH NEXT FROM c1 INTO @var1, @var2; +/// END +/// ``` +/// +/// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/language-elements/while-transact-sql) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct WhileStatement { + pub while_block: ConditionalStatementBlock, +} + +impl fmt::Display for WhileStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let WhileStatement { while_block } = self; + write!(f, "{while_block}")?; + Ok(()) + } +} + +/// A block within a [Statement::Case] or [Statement::If] or [Statement::While]-like statement /// /// Example 1: /// ```sql @@ -2244,6 +2270,14 @@ impl fmt::Display for IfStatement { /// ```sql /// ELSE SELECT 1; SELECT 2; /// ``` +/// +/// Example 4: +/// ```sql +/// WHILE @@FETCH_STATUS = 0 +/// BEGIN +/// FETCH NEXT FROM c1 INTO @var1, @var2; +/// END +/// ``` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -2983,6 +3017,8 @@ pub enum Statement { Case(CaseStatement), /// An `IF` statement. If(IfStatement), + /// A `WHILE` statement. + While(WhileStatement), /// A `RAISE` statement. Raise(RaiseStatement), /// ```sql @@ -3034,6 +3070,11 @@ pub enum Statement { partition: Option>, }, /// ```sql + /// OPEN cursor_name + /// ``` + /// Opens a cursor. + Open(OpenStatement), + /// ```sql /// CLOSE /// ``` /// Closes the portal underlying an open cursor. @@ -3413,6 +3454,7 @@ pub enum Statement { /// Cursor name name: Ident, direction: FetchDirection, + position: FetchPosition, /// Optional, It's possible to fetch rows form cursor to the table into: Option, }, @@ -4235,11 +4277,10 @@ impl fmt::Display for Statement { Statement::Fetch { name, direction, + position, into, } => { - write!(f, "FETCH {direction} ")?; - - write!(f, "IN {name}")?; + write!(f, "FETCH {direction} {position} {name}")?; if let Some(into) = into { write!(f, " INTO {into}")?; @@ -4329,6 +4370,9 @@ impl fmt::Display for Statement { Statement::If(stmt) => { write!(f, "{stmt}") } + Statement::While(stmt) => { + write!(f, "{stmt}") + } Statement::Raise(stmt) => { write!(f, "{stmt}") } @@ -4498,6 +4542,7 @@ impl fmt::Display for Statement { Ok(()) } Statement::Delete(delete) => write!(f, "{delete}"), + Statement::Open(open) => write!(f, "{open}"), Statement::Close { cursor } => { write!(f, "CLOSE {cursor}")?; @@ -6187,6 +6232,28 @@ impl fmt::Display for FetchDirection { } } +/// The "position" for a FETCH statement. +/// +/// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/language-elements/fetch-transact-sql) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum FetchPosition { + From, + In, +} + +impl fmt::Display for FetchPosition { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + FetchPosition::From => f.write_str("FROM")?, + FetchPosition::In => f.write_str("IN")?, + }; + + Ok(()) + } +} + /// A privilege on a database object (table, sequence, etc.). #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -9354,6 +9421,21 @@ pub enum ReturnStatementValue { Expr(Expr), } +/// Represents an `OPEN` statement. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct OpenStatement { + /// Cursor name + pub cursor_name: Ident, +} + +impl fmt::Display for OpenStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "OPEN {}", self.cursor_name) + } +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 33bc07392..836f229a2 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -31,13 +31,13 @@ use super::{ FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, LimitClause, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, - Offset, OnConflict, OnConflictAction, OnInsert, OrderBy, OrderByExpr, OrderByKind, Partition, - PivotValueSource, ProjectionSelect, Query, RaiseStatement, RaiseStatementValue, - ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, - SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, - TableAliasColumnDef, TableConstraint, TableFactor, TableObject, TableOptionsClustered, - TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, - WildcardAdditionalOptions, With, WithFill, + Offset, OnConflict, OnConflictAction, OnInsert, OpenStatement, OrderBy, OrderByExpr, + OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, RaiseStatement, + RaiseStatementValue, ReferentialAction, RenameSelectItem, ReplaceSelectElement, + ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, + SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, + TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, + WhileStatement, WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -339,6 +339,7 @@ impl Spanned for Statement { } => source.span(), Statement::Case(stmt) => stmt.span(), Statement::If(stmt) => stmt.span(), + Statement::While(stmt) => stmt.span(), Statement::Raise(stmt) => stmt.span(), Statement::Call(function) => function.span(), Statement::Copy { @@ -365,6 +366,7 @@ impl Spanned for Statement { from_query: _, partition: _, } => Span::empty(), + Statement::Open(open) => open.span(), Statement::Close { cursor } => match cursor { CloseCursor::All => Span::empty(), CloseCursor::Specific { name } => name.span, @@ -776,6 +778,14 @@ impl Spanned for IfStatement { } } +impl Spanned for WhileStatement { + fn span(&self) -> Span { + let WhileStatement { while_block } = self; + + while_block.span() + } +} + impl Spanned for ConditionalStatements { fn span(&self) -> Span { match self { @@ -2297,6 +2307,13 @@ impl Spanned for BeginEndStatements { } } +impl Spanned for OpenStatement { + fn span(&self) -> Span { + let OpenStatement { cursor_name } = self; + cursor_name.span + } +} + #[cfg(test)] pub mod tests { use crate::dialect::{Dialect, GenericDialect, SnowflakeDialect}; diff --git a/src/keywords.rs b/src/keywords.rs index 32612ccd5..bf8a1915d 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -985,6 +985,7 @@ define_keywords!( WHEN, WHENEVER, WHERE, + WHILE, WIDTH_BUCKET, WINDOW, WITH, @@ -1068,6 +1069,7 @@ pub const RESERVED_FOR_TABLE_ALIAS: &[Keyword] = &[ Keyword::SAMPLE, Keyword::TABLESAMPLE, Keyword::FROM, + Keyword::OPEN, ]; /// Can't be used as a column alias, so that `SELECT alias` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0d74235b2..cbd464c34 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -536,6 +536,10 @@ impl<'a> Parser<'a> { self.prev_token(); self.parse_if_stmt() } + Keyword::WHILE => { + self.prev_token(); + self.parse_while() + } Keyword::RAISE => { self.prev_token(); self.parse_raise_stmt() @@ -570,6 +574,10 @@ impl<'a> Parser<'a> { Keyword::ALTER => self.parse_alter(), Keyword::CALL => self.parse_call(), Keyword::COPY => self.parse_copy(), + Keyword::OPEN => { + self.prev_token(); + self.parse_open() + } Keyword::CLOSE => self.parse_close(), Keyword::SET => self.parse_set(), Keyword::SHOW => self.parse_show(), @@ -700,8 +708,18 @@ impl<'a> Parser<'a> { })) } + /// Parse a `WHILE` statement. + /// + /// See [Statement::While] + fn parse_while(&mut self) -> Result { + self.expect_keyword_is(Keyword::WHILE)?; + let while_block = self.parse_conditional_statement_block(&[Keyword::END])?; + + Ok(Statement::While(WhileStatement { while_block })) + } + /// Parses an expression and associated list of statements - /// belonging to a conditional statement like `IF` or `WHEN`. + /// belonging to a conditional statement like `IF` or `WHEN` or `WHILE`. /// /// Example: /// ```sql @@ -716,6 +734,10 @@ impl<'a> Parser<'a> { let condition = match &start_token.token { Token::Word(w) if w.keyword == Keyword::ELSE => None, + Token::Word(w) if w.keyword == Keyword::WHILE => { + let expr = self.parse_expr()?; + Some(expr) + } _ => { let expr = self.parse_expr()?; then_token = Some(AttachedToken(self.expect_keyword(Keyword::THEN)?)); @@ -723,13 +745,25 @@ impl<'a> Parser<'a> { } }; - let statements = self.parse_statement_list(terminal_keywords)?; + let conditional_statements = if self.peek_keyword(Keyword::BEGIN) { + let begin_token = self.expect_keyword(Keyword::BEGIN)?; + let statements = self.parse_statement_list(terminal_keywords)?; + let end_token = self.expect_keyword(Keyword::END)?; + ConditionalStatements::BeginEnd(BeginEndStatements { + begin_token: AttachedToken(begin_token), + statements, + end_token: AttachedToken(end_token), + }) + } else { + let statements = self.parse_statement_list(terminal_keywords)?; + ConditionalStatements::Sequence { statements } + }; Ok(ConditionalStatementBlock { start_token: AttachedToken(start_token), condition, then_token, - conditional_statements: ConditionalStatements::Sequence { statements }, + conditional_statements, }) } @@ -4467,11 +4501,16 @@ impl<'a> Parser<'a> { ) -> Result, ParserError> { let mut values = vec![]; loop { - if let Token::Word(w) = &self.peek_nth_token_ref(0).token { - if w.quote_style.is_none() && terminal_keywords.contains(&w.keyword) { - break; + match &self.peek_nth_token_ref(0).token { + Token::EOF => break, + Token::Word(w) => { + if w.quote_style.is_none() && terminal_keywords.contains(&w.keyword) { + break; + } } + _ => {} } + values.push(self.parse_statement()?); self.expect_token(&Token::SemiColon)?; } @@ -6644,7 +6683,15 @@ impl<'a> Parser<'a> { } }; - self.expect_one_of_keywords(&[Keyword::FROM, Keyword::IN])?; + let position = if self.peek_keyword(Keyword::FROM) { + self.expect_keyword(Keyword::FROM)?; + FetchPosition::From + } else if self.peek_keyword(Keyword::IN) { + self.expect_keyword(Keyword::IN)?; + FetchPosition::In + } else { + return parser_err!("Expected FROM or IN", self.peek_token().span.start); + }; let name = self.parse_identifier()?; @@ -6657,6 +6704,7 @@ impl<'a> Parser<'a> { Ok(Statement::Fetch { name, direction, + position, into, }) } @@ -8770,6 +8818,14 @@ impl<'a> Parser<'a> { }) } + /// Parse [Statement::Open] + fn parse_open(&mut self) -> Result { + self.expect_keyword(Keyword::OPEN)?; + Ok(Statement::Open(OpenStatement { + cursor_name: self.parse_identifier()?, + })) + } + pub fn parse_close(&mut self) -> Result { let cursor = if self.parse_keyword(Keyword::ALL) { CloseCursor::All diff --git a/src/test_utils.rs b/src/test_utils.rs index 6270ac42b..3c22fa911 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -151,6 +151,8 @@ impl TestedDialects { /// /// 2. re-serializing the result of parsing `sql` produces the same /// `canonical` sql string + /// + /// For multiple statements, use [`statements_parse_to`]. pub fn one_statement_parses_to(&self, sql: &str, canonical: &str) -> Statement { let mut statements = self.parse_sql_statements(sql).expect(sql); assert_eq!(statements.len(), 1); @@ -166,6 +168,24 @@ impl TestedDialects { only_statement } + /// The same as [`one_statement_parses_to`] but it works for a multiple statements + pub fn statements_parse_to(&self, sql: &str, canonical: &str) -> Vec { + let statements = self.parse_sql_statements(sql).expect(sql); + if !canonical.is_empty() && sql != canonical { + assert_eq!(self.parse_sql_statements(canonical).unwrap(), statements); + } else { + assert_eq!( + sql, + statements + .iter() + .map(|s| s.to_string()) + .collect::>() + .join("; ") + ); + } + statements + } + /// Ensures that `sql` parses as an [`Expr`], and that /// re-serializing the parse result produces canonical pub fn expr_parses_to(&self, sql: &str, canonical: &str) -> Expr { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 6d99929d5..1ddf3f92e 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15187,3 +15187,15 @@ fn parse_return() { let _ = all_dialects().verified_stmt("RETURN 1"); } + +#[test] +fn test_open() { + let open_cursor = "OPEN Employee_Cursor"; + let stmt = all_dialects().verified_stmt(open_cursor); + assert_eq!( + stmt, + Statement::Open(OpenStatement { + cursor_name: Ident::new("Employee_Cursor"), + }) + ); +} diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index b2d5210c2..88e7a1f17 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -23,7 +23,8 @@ mod test_utils; use helpers::attached_token::AttachedToken; -use sqlparser::tokenizer::{Location, Span}; +use sqlparser::keywords::Keyword; +use sqlparser::tokenizer::{Location, Span, Token, TokenWithSpan, Word}; use test_utils::*; use sqlparser::ast::DataType::{Int, Text, Varbinary}; @@ -223,7 +224,7 @@ fn parse_create_function() { value: Some(ReturnStatementValue::Expr(Expr::Value( (number("1")).with_empty_span() ))), - }),], + })], end_token: AttachedToken::empty(), })), behavior: None, @@ -1397,6 +1398,85 @@ fn parse_mssql_declare() { let _ = ms().verified_stmt(declare_cursor_for_select); } +#[test] +fn test_mssql_cursor() { + let full_cursor_usage = "\ + DECLARE Employee_Cursor CURSOR FOR \ + SELECT LastName, FirstName \ + FROM AdventureWorks2022.HumanResources.vEmployee \ + WHERE LastName LIKE 'B%'; \ + \ + OPEN Employee_Cursor; \ + \ + FETCH NEXT FROM Employee_Cursor; \ + \ + WHILE @@FETCH_STATUS = 0 \ + BEGIN \ + FETCH NEXT FROM Employee_Cursor; \ + END; \ + \ + CLOSE Employee_Cursor; \ + DEALLOCATE Employee_Cursor\ + "; + let _ = ms().statements_parse_to(full_cursor_usage, ""); +} + +#[test] +fn test_mssql_while_statement() { + let while_single_statement = "WHILE 1 = 0 PRINT 'Hello World';"; + let stmt = ms().verified_stmt(while_single_statement); + assert_eq!( + stmt, + Statement::While(sqlparser::ast::WhileStatement { + while_block: ConditionalStatementBlock { + start_token: AttachedToken(TokenWithSpan { + token: Token::Word(Word { + value: "WHILE".to_string(), + quote_style: None, + keyword: Keyword::WHILE + }), + span: Span::empty() + }), + condition: Some(Expr::BinaryOp { + left: Box::new(Expr::Value( + (Value::Number("1".parse().unwrap(), false)).with_empty_span() + )), + op: BinaryOperator::Eq, + right: Box::new(Expr::Value( + (Value::Number("0".parse().unwrap(), false)).with_empty_span() + )), + }), + then_token: None, + conditional_statements: ConditionalStatements::Sequence { + statements: vec![Statement::Print(PrintStatement { + message: Box::new(Expr::Value( + (Value::SingleQuotedString("Hello World".to_string())) + .with_empty_span() + )), + })], + } + } + }) + ); + + let while_begin_end = "\ + WHILE @@FETCH_STATUS = 0 \ + BEGIN \ + FETCH NEXT FROM Employee_Cursor; \ + END\ + "; + let _ = ms().verified_stmt(while_begin_end); + + let while_begin_end_multiple_statements = "\ + WHILE @@FETCH_STATUS = 0 \ + BEGIN \ + FETCH NEXT FROM Employee_Cursor; \ + PRINT 'Hello World'; \ + END\ + "; + let _ = ms().verified_stmt(while_begin_end_multiple_statements); +} + #[test] fn test_parse_raiserror() { let sql = r#"RAISERROR('This is a test', 16, 1)"#; From 728645fb31f8e41640255c955cfe90e7f98e7752 Mon Sep 17 00:00:00 2001 From: benrsatori Date: Fri, 2 May 2025 16:16:59 +0300 Subject: [PATCH 118/286] Add all missing table options to be handled in any order (#1747) Co-authored-by: Tomer Shani --- src/ast/dml.rs | 75 ++---- src/ast/helpers/stmt_create_table.rs | 91 ++------ src/ast/helpers/stmt_data_loading.rs | 2 - src/ast/mod.rs | 111 +++++++-- src/ast/spans.rs | 39 ++-- src/dialect/snowflake.rs | 17 +- src/keywords.rs | 17 ++ src/parser/mod.rs | 334 ++++++++++++++++++++------- tests/sqlparser_bigquery.rs | 6 +- tests/sqlparser_clickhouse.rs | 37 +-- tests/sqlparser_common.rs | 47 ++-- tests/sqlparser_duckdb.rs | 8 +- tests/sqlparser_hive.rs | 4 +- tests/sqlparser_mssql.rs | 18 +- tests/sqlparser_mysql.rs | 304 +++++++++++++++++++++--- tests/sqlparser_postgres.rs | 22 +- tests/sqlparser_snowflake.rs | 17 +- 17 files changed, 767 insertions(+), 382 deletions(-) diff --git a/src/ast/dml.rs b/src/ast/dml.rs index 9cdb1ca86..7ed17be9b 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -33,11 +33,11 @@ pub use super::ddl::{ColumnDef, TableConstraint}; use super::{ display_comma_separated, display_separated, query::InputFormatClause, Assignment, ClusteredBy, - CommentDef, Expr, FileFormat, FromTable, HiveDistributionStyle, HiveFormat, HiveIOFormat, - HiveRowFormat, Ident, IndexType, InsertAliases, MysqlInsertPriority, ObjectName, OnCommit, - OnInsert, OneOrManyWithParens, OrderByExpr, Query, RowAccessPolicy, SelectItem, Setting, - SqlOption, SqliteOnConflict, StorageSerializationPolicy, TableEngine, TableObject, - TableWithJoins, Tag, WrappedCollection, + CommentDef, CreateTableOptions, Expr, FileFormat, FromTable, HiveDistributionStyle, HiveFormat, + HiveIOFormat, HiveRowFormat, Ident, IndexType, InsertAliases, MysqlInsertPriority, ObjectName, + OnCommit, OnInsert, OneOrManyWithParens, OrderByExpr, Query, RowAccessPolicy, SelectItem, + Setting, SqliteOnConflict, StorageSerializationPolicy, TableObject, TableWithJoins, Tag, + WrappedCollection, }; /// Index column type. @@ -146,19 +146,17 @@ pub struct CreateTable { pub constraints: Vec, pub hive_distribution: HiveDistributionStyle, pub hive_formats: Option, - pub table_properties: Vec, - pub with_options: Vec, + pub table_options: CreateTableOptions, pub file_format: Option, pub location: Option, pub query: Option>, pub without_rowid: bool, pub like: Option, pub clone: Option, - pub engine: Option, + // For Hive dialect, the table comment is after the column definitions without `=`, + // so the `comment` field is optional and different than the comment field in the general options list. + // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) pub comment: Option, - pub auto_increment_offset: Option, - pub default_charset: Option, - pub collation: Option, pub on_commit: Option, /// ClickHouse "ON CLUSTER" clause: /// @@ -179,9 +177,6 @@ pub struct CreateTable { /// Hive: Table clustering column list. /// pub clustered_by: Option, - /// BigQuery: Table options list. - /// - pub options: Option>, /// Postgres `INHERITs` clause, which contains the list of tables from which /// the new table inherits. /// @@ -282,7 +277,7 @@ impl Display for CreateTable { // Hive table comment should be after column definitions, please refer to: // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) - if let Some(CommentDef::AfterColumnDefsWithoutEq(comment)) = &self.comment { + if let Some(comment) = &self.comment { write!(f, " COMMENT '{comment}'")?; } @@ -375,35 +370,14 @@ impl Display for CreateTable { } write!(f, " LOCATION '{}'", self.location.as_ref().unwrap())?; } - if !self.table_properties.is_empty() { - write!( - f, - " TBLPROPERTIES ({})", - display_comma_separated(&self.table_properties) - )?; - } - if !self.with_options.is_empty() { - write!(f, " WITH ({})", display_comma_separated(&self.with_options))?; - } - if let Some(engine) = &self.engine { - write!(f, " ENGINE={engine}")?; - } - if let Some(comment_def) = &self.comment { - match comment_def { - CommentDef::WithEq(comment) => { - write!(f, " COMMENT = '{comment}'")?; - } - CommentDef::WithoutEq(comment) => { - write!(f, " COMMENT '{comment}'")?; - } - // For CommentDef::AfterColumnDefsWithoutEq will be displayed after column definition - CommentDef::AfterColumnDefsWithoutEq(_) => (), - } - } - if let Some(auto_increment_offset) = self.auto_increment_offset { - write!(f, " AUTO_INCREMENT {auto_increment_offset}")?; + match &self.table_options { + options @ CreateTableOptions::With(_) + | options @ CreateTableOptions::Plain(_) + | options @ CreateTableOptions::TableProperties(_) => write!(f, " {}", options)?, + _ => (), } + if let Some(primary_key) = &self.primary_key { write!(f, " PRIMARY KEY {}", primary_key)?; } @@ -419,15 +393,9 @@ impl Display for CreateTable { if let Some(cluster_by) = self.cluster_by.as_ref() { write!(f, " CLUSTER BY {cluster_by}")?; } - - if let Some(options) = self.options.as_ref() { - write!( - f, - " OPTIONS({})", - display_comma_separated(options.as_slice()) - )?; + if let options @ CreateTableOptions::Options(_) = &self.table_options { + write!(f, " {}", options)?; } - if let Some(external_volume) = self.external_volume.as_ref() { write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; } @@ -503,13 +471,6 @@ impl Display for CreateTable { write!(f, " WITH TAG ({})", display_comma_separated(tag.as_slice()))?; } - if let Some(default_charset) = &self.default_charset { - write!(f, " DEFAULT CHARSET={default_charset}")?; - } - if let Some(collation) = &self.collation { - write!(f, " COLLATE={collation}")?; - } - if self.on_commit.is_some() { let on_commit = match self.on_commit { Some(OnCommit::DeleteRows) => "ON COMMIT DELETE ROWS", diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index 1c50cb843..542d30ea9 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -26,10 +26,12 @@ use sqlparser_derive::{Visit, VisitMut}; use super::super::dml::CreateTable; use crate::ast::{ - ClusteredBy, ColumnDef, CommentDef, Expr, FileFormat, HiveDistributionStyle, HiveFormat, Ident, - ObjectName, OnCommit, OneOrManyWithParens, Query, RowAccessPolicy, SqlOption, Statement, - StorageSerializationPolicy, TableConstraint, TableEngine, Tag, WrappedCollection, + ClusteredBy, ColumnDef, CommentDef, CreateTableOptions, Expr, FileFormat, + HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, OneOrManyWithParens, Query, + RowAccessPolicy, Statement, StorageSerializationPolicy, TableConstraint, Tag, + WrappedCollection, }; + use crate::parser::ParserError; /// Builder for create table statement variant ([1]). @@ -76,19 +78,13 @@ pub struct CreateTableBuilder { pub constraints: Vec, pub hive_distribution: HiveDistributionStyle, pub hive_formats: Option, - pub table_properties: Vec, - pub with_options: Vec, pub file_format: Option, pub location: Option, pub query: Option>, pub without_rowid: bool, pub like: Option, pub clone: Option, - pub engine: Option, pub comment: Option, - pub auto_increment_offset: Option, - pub default_charset: Option, - pub collation: Option, pub on_commit: Option, pub on_cluster: Option, pub primary_key: Option>, @@ -96,7 +92,6 @@ pub struct CreateTableBuilder { pub partition_by: Option>, pub cluster_by: Option>>, pub clustered_by: Option, - pub options: Option>, pub inherits: Option>, pub strict: bool, pub copy_grants: bool, @@ -113,6 +108,7 @@ pub struct CreateTableBuilder { pub catalog: Option, pub catalog_sync: Option, pub storage_serialization_policy: Option, + pub table_options: CreateTableOptions, } impl CreateTableBuilder { @@ -131,19 +127,13 @@ impl CreateTableBuilder { constraints: vec![], hive_distribution: HiveDistributionStyle::NONE, hive_formats: None, - table_properties: vec![], - with_options: vec![], file_format: None, location: None, query: None, without_rowid: false, like: None, clone: None, - engine: None, comment: None, - auto_increment_offset: None, - default_charset: None, - collation: None, on_commit: None, on_cluster: None, primary_key: None, @@ -151,7 +141,6 @@ impl CreateTableBuilder { partition_by: None, cluster_by: None, clustered_by: None, - options: None, inherits: None, strict: false, copy_grants: false, @@ -168,6 +157,7 @@ impl CreateTableBuilder { catalog: None, catalog_sync: None, storage_serialization_policy: None, + table_options: CreateTableOptions::None, } } pub fn or_replace(mut self, or_replace: bool) -> Self { @@ -230,15 +220,6 @@ impl CreateTableBuilder { self } - pub fn table_properties(mut self, table_properties: Vec) -> Self { - self.table_properties = table_properties; - self - } - - pub fn with_options(mut self, with_options: Vec) -> Self { - self.with_options = with_options; - self - } pub fn file_format(mut self, file_format: Option) -> Self { self.file_format = file_format; self @@ -268,31 +249,11 @@ impl CreateTableBuilder { self } - pub fn engine(mut self, engine: Option) -> Self { - self.engine = engine; - self - } - - pub fn comment(mut self, comment: Option) -> Self { + pub fn comment_after_column_def(mut self, comment: Option) -> Self { self.comment = comment; self } - pub fn auto_increment_offset(mut self, offset: Option) -> Self { - self.auto_increment_offset = offset; - self - } - - pub fn default_charset(mut self, default_charset: Option) -> Self { - self.default_charset = default_charset; - self - } - - pub fn collation(mut self, collation: Option) -> Self { - self.collation = collation; - self - } - pub fn on_commit(mut self, on_commit: Option) -> Self { self.on_commit = on_commit; self @@ -328,11 +289,6 @@ impl CreateTableBuilder { self } - pub fn options(mut self, options: Option>) -> Self { - self.options = options; - self - } - pub fn inherits(mut self, inherits: Option>) -> Self { self.inherits = inherits; self @@ -422,6 +378,11 @@ impl CreateTableBuilder { self } + pub fn table_options(mut self, table_options: CreateTableOptions) -> Self { + self.table_options = table_options; + self + } + pub fn build(self) -> Statement { Statement::CreateTable(CreateTable { or_replace: self.or_replace, @@ -437,19 +398,13 @@ impl CreateTableBuilder { constraints: self.constraints, hive_distribution: self.hive_distribution, hive_formats: self.hive_formats, - table_properties: self.table_properties, - with_options: self.with_options, file_format: self.file_format, location: self.location, query: self.query, without_rowid: self.without_rowid, like: self.like, clone: self.clone, - engine: self.engine, comment: self.comment, - auto_increment_offset: self.auto_increment_offset, - default_charset: self.default_charset, - collation: self.collation, on_commit: self.on_commit, on_cluster: self.on_cluster, primary_key: self.primary_key, @@ -457,7 +412,6 @@ impl CreateTableBuilder { partition_by: self.partition_by, cluster_by: self.cluster_by, clustered_by: self.clustered_by, - options: self.options, inherits: self.inherits, strict: self.strict, copy_grants: self.copy_grants, @@ -474,6 +428,7 @@ impl CreateTableBuilder { catalog: self.catalog, catalog_sync: self.catalog_sync, storage_serialization_policy: self.storage_serialization_policy, + table_options: self.table_options, }) } } @@ -499,19 +454,13 @@ impl TryFrom for CreateTableBuilder { constraints, hive_distribution, hive_formats, - table_properties, - with_options, file_format, location, query, without_rowid, like, clone, - engine, comment, - auto_increment_offset, - default_charset, - collation, on_commit, on_cluster, primary_key, @@ -519,7 +468,6 @@ impl TryFrom for CreateTableBuilder { partition_by, cluster_by, clustered_by, - options, inherits, strict, copy_grants, @@ -536,6 +484,7 @@ impl TryFrom for CreateTableBuilder { catalog, catalog_sync, storage_serialization_policy, + table_options, }) => Ok(Self { or_replace, temporary, @@ -548,19 +497,13 @@ impl TryFrom for CreateTableBuilder { constraints, hive_distribution, hive_formats, - table_properties, - with_options, file_format, location, query, without_rowid, like, clone, - engine, comment, - auto_increment_offset, - default_charset, - collation, on_commit, on_cluster, primary_key, @@ -568,7 +511,6 @@ impl TryFrom for CreateTableBuilder { partition_by, cluster_by, clustered_by, - options, inherits, strict, iceberg, @@ -587,6 +529,7 @@ impl TryFrom for CreateTableBuilder { catalog, catalog_sync, storage_serialization_policy, + table_options, }), _ => Err(ParserError::ParserError(format!( "Expected create table statement, but received: {stmt}" @@ -600,8 +543,8 @@ impl TryFrom for CreateTableBuilder { pub(crate) struct CreateTableConfiguration { pub partition_by: Option>, pub cluster_by: Option>>, - pub options: Option>, pub inherits: Option>, + pub table_options: CreateTableOptions, } #[cfg(test)] diff --git a/src/ast/helpers/stmt_data_loading.rs b/src/ast/helpers/stmt_data_loading.rs index e960bb05b..92a727279 100644 --- a/src/ast/helpers/stmt_data_loading.rs +++ b/src/ast/helpers/stmt_data_loading.rs @@ -21,8 +21,6 @@ #[cfg(not(feature = "std"))] use alloc::string::String; -#[cfg(not(feature = "std"))] -use alloc::vec::Vec; use core::fmt; #[cfg(feature = "serde")] diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 582922a3e..d74d197e3 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2681,6 +2681,18 @@ pub enum CreateTableOptions { /// /// Options(Vec), + + /// Plain options, options which are not part on any declerative statement e.g. WITH/OPTIONS/... + /// + Plain(Vec), + + TableProperties(Vec), +} + +impl Default for CreateTableOptions { + fn default() -> Self { + Self::None + } } impl fmt::Display for CreateTableOptions { @@ -2692,6 +2704,12 @@ impl fmt::Display for CreateTableOptions { CreateTableOptions::Options(options) => { write!(f, "OPTIONS({})", display_comma_separated(options)) } + CreateTableOptions::TableProperties(options) => { + write!(f, "TBLPROPERTIES ({})", display_comma_separated(options)) + } + CreateTableOptions::Plain(options) => { + write!(f, "{}", display_separated(options, " ")) + } CreateTableOptions::None => Ok(()), } } @@ -7560,6 +7578,18 @@ pub enum SqlOption { range_direction: Option, for_values: Vec, }, + /// Comment parameter (supports `=` and no `=` syntax) + Comment(CommentDef), + /// MySQL TableSpace option + /// + TableSpace(TablespaceOption), + /// An option representing a key value pair, where the value is a parenthesized list and with an optional name + /// e.g. + /// + /// UNION = (tbl_name\[,tbl_name\]...) + /// ENGINE = ReplicatedMergeTree('/table_name','{replica}', ver) + /// ENGINE = SummingMergeTree(\[columns\]) + NamedParenthesizedList(NamedParenthesizedList), } impl fmt::Display for SqlOption { @@ -7591,10 +7621,54 @@ impl fmt::Display for SqlOption { display_comma_separated(for_values) ) } + SqlOption::TableSpace(tablespace_option) => { + write!(f, "TABLESPACE {}", tablespace_option.name)?; + match tablespace_option.storage { + Some(StorageType::Disk) => write!(f, " STORAGE DISK"), + Some(StorageType::Memory) => write!(f, " STORAGE MEMORY"), + None => Ok(()), + } + } + SqlOption::Comment(comment) => match comment { + CommentDef::WithEq(comment) => { + write!(f, "COMMENT = '{comment}'") + } + CommentDef::WithoutEq(comment) => { + write!(f, "COMMENT '{comment}'") + } + }, + SqlOption::NamedParenthesizedList(value) => { + write!(f, "{} = ", value.key)?; + if let Some(key) = &value.name { + write!(f, "{}", key)?; + } + if !value.values.is_empty() { + write!(f, "({})", display_comma_separated(&value.values))? + } + Ok(()) + } } } } +#[derive(Debug, Clone, PartialEq, Eq, Hash, PartialOrd, Ord)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum StorageType { + Disk, + Memory, +} + +#[derive(Debug, Clone, PartialEq, Eq, Hash, PartialOrd, Ord)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// MySql TableSpace option +/// +pub struct TablespaceOption { + pub name: String, + pub storage: Option, +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -8860,27 +8934,20 @@ impl Display for CreateViewParams { } } -/// Engine of DB. Some warehouse has parameters of engine, e.g. [ClickHouse] -/// -/// [ClickHouse]: https://clickhouse.com/docs/en/engines/table-engines #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct TableEngine { - pub name: String, - pub parameters: Option>, -} - -impl Display for TableEngine { - fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{}", self.name)?; - - if let Some(parameters) = self.parameters.as_ref() { - write!(f, "({})", display_comma_separated(parameters))?; - } - - Ok(()) - } +/// Key/Value, where the value is a (optionally named) list of identifiers +/// +/// ```sql +/// UNION = (tbl_name[,tbl_name]...) +/// ENGINE = ReplicatedMergeTree('/table_name','{replica}', ver) +/// ENGINE = SummingMergeTree([columns]) +/// ``` +pub struct NamedParenthesizedList { + pub key: Ident, + pub name: Option, + pub values: Vec, } /// Snowflake `WITH ROW ACCESS POLICY policy_name ON (identifier, ...)` @@ -8944,18 +9011,12 @@ pub enum CommentDef { /// Does not include `=` when printing the comment, as `COMMENT 'comment'` WithEq(String), WithoutEq(String), - // For Hive dialect, the table comment is after the column definitions without `=`, - // so we need to add an extra variant to allow to identify this case when displaying. - // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) - AfterColumnDefsWithoutEq(String), } impl Display for CommentDef { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { - CommentDef::WithEq(comment) - | CommentDef::WithoutEq(comment) - | CommentDef::AfterColumnDefsWithoutEq(comment) => write!(f, "{comment}"), + CommentDef::WithEq(comment) | CommentDef::WithoutEq(comment) => write!(f, "{comment}"), } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 836f229a2..3f703ffaf 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -30,10 +30,10 @@ use super::{ Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, Insert, Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, - LimitClause, MatchRecognizePattern, Measure, NamedWindowDefinition, ObjectName, ObjectNamePart, - Offset, OnConflict, OnConflictAction, OnInsert, OpenStatement, OrderBy, OrderByExpr, - OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, RaiseStatement, - RaiseStatementValue, ReferentialAction, RenameSelectItem, ReplaceSelectElement, + LimitClause, MatchRecognizePattern, Measure, NamedParenthesizedList, NamedWindowDefinition, + ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, OnInsert, OpenStatement, + OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, + RaiseStatement, RaiseStatementValue, ReferentialAction, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, @@ -567,27 +567,20 @@ impl Spanned for CreateTable { constraints, hive_distribution: _, // hive specific hive_formats: _, // hive specific - table_properties, - with_options, - file_format: _, // enum - location: _, // string, no span + file_format: _, // enum + location: _, // string, no span query, without_rowid: _, // bool like, clone, - engine: _, // todo - comment: _, // todo, no span - auto_increment_offset: _, // u32, no span - default_charset: _, // string, no span - collation: _, // string, no span - on_commit: _, // enum + comment: _, // todo, no span + on_commit: _, on_cluster: _, // todo, clickhouse specific primary_key: _, // todo, clickhouse specific order_by: _, // todo, clickhouse specific partition_by: _, // todo, BigQuery specific cluster_by: _, // todo, BigQuery specific clustered_by: _, // todo, Hive specific - options: _, // todo, BigQuery specific inherits: _, // todo, PostgreSQL specific strict: _, // bool copy_grants: _, // bool @@ -603,15 +596,15 @@ impl Spanned for CreateTable { base_location: _, // todo, Snowflake specific catalog: _, // todo, Snowflake specific catalog_sync: _, // todo, Snowflake specific - storage_serialization_policy: _, // todo, Snowflake specific + storage_serialization_policy: _, + table_options, } = self; union_spans( core::iter::once(name.span()) + .chain(core::iter::once(table_options.span())) .chain(columns.iter().map(|i| i.span())) .chain(constraints.iter().map(|i| i.span())) - .chain(table_properties.iter().map(|i| i.span())) - .chain(with_options.iter().map(|i| i.span())) .chain(query.iter().map(|i| i.span())) .chain(like.iter().map(|i| i.span())) .chain(clone.iter().map(|i| i.span())), @@ -1004,6 +997,14 @@ impl Spanned for SqlOption { } => union_spans( core::iter::once(column_name.span).chain(for_values.iter().map(|i| i.span())), ), + SqlOption::TableSpace(_) => Span::empty(), + SqlOption::Comment(_) => Span::empty(), + SqlOption::NamedParenthesizedList(NamedParenthesizedList { + key: name, + name: value, + values, + }) => union_spans(core::iter::once(name.span).chain(values.iter().map(|i| i.span))) + .union_opt(&value.as_ref().map(|i| i.span)), } } } @@ -1041,6 +1042,8 @@ impl Spanned for CreateTableOptions { CreateTableOptions::None => Span::empty(), CreateTableOptions::With(vec) => union_spans(vec.iter().map(|i| i.span())), CreateTableOptions::Options(vec) => union_spans(vec.iter().map(|i| i.span())), + CreateTableOptions::Plain(vec) => union_spans(vec.iter().map(|i| i.span())), + CreateTableOptions::TableProperties(vec) => union_spans(vec.iter().map(|i| i.span())), } } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index c4d6a5ad5..ccce16198 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -25,8 +25,8 @@ use crate::ast::helpers::stmt_data_loading::{ use crate::ast::{ ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, - IdentityPropertyOrder, ObjectName, RowAccessPolicy, ShowObjects, Statement, TagsColumnOption, - WrappedCollection, + IdentityPropertyOrder, ObjectName, RowAccessPolicy, ShowObjects, SqlOption, Statement, + TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -417,6 +417,8 @@ pub fn parse_create_table( // "CREATE TABLE x COPY GRANTS (c INT)" and "CREATE TABLE x (c INT) COPY GRANTS" are both // accepted by Snowflake + let mut plain_options = vec![]; + loop { let next_token = parser.next_token(); match &next_token.token { @@ -428,7 +430,9 @@ pub fn parse_create_table( Keyword::COMMENT => { // Rewind the COMMENT keyword parser.prev_token(); - builder = builder.comment(parser.parse_optional_inline_comment()?); + if let Some(comment_def) = parser.parse_optional_inline_comment()? { + plain_options.push(SqlOption::Comment(comment_def)) + } } Keyword::AS => { let query = parser.parse_query()?; @@ -589,6 +593,13 @@ pub fn parse_create_table( } } } + let table_options = if !plain_options.is_empty() { + crate::ast::CreateTableOptions::Plain(plain_options) + } else { + crate::ast::CreateTableOptions::None + }; + + builder = builder.table_options(table_options); if iceberg && builder.base_location.is_none() { return Err(ParserError::ParserError( diff --git a/src/keywords.rs b/src/keywords.rs index bf8a1915d..ddb786650 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -116,9 +116,11 @@ define_keywords!( AUTHENTICATION, AUTHORIZATION, AUTO, + AUTOEXTEND_SIZE, AUTOINCREMENT, AUTO_INCREMENT, AVG, + AVG_ROW_LENGTH, AVRO, BACKWARD, BASE64, @@ -180,6 +182,7 @@ define_keywords!( CHARSET, CHAR_LENGTH, CHECK, + CHECKSUM, CIRCLE, CLEAR, CLOB, @@ -269,6 +272,7 @@ define_keywords!( DEFINED, DEFINER, DELAYED, + DELAY_KEY_WRITE, DELETE, DELIMITED, DELIMITER, @@ -313,6 +317,7 @@ define_keywords!( END_PARTITION, ENFORCED, ENGINE, + ENGINE_ATTRIBUTE, ENUM, ENUM16, ENUM8, @@ -444,6 +449,7 @@ define_keywords!( INPUTFORMAT, INSENSITIVE, INSERT, + INSERT_METHOD, INSTALL, INSTANT, INSTEAD, @@ -480,6 +486,7 @@ define_keywords!( JULIAN, KEY, KEYS, + KEY_BLOCK_SIZE, KILL, LAG, LANGUAGE, @@ -533,6 +540,7 @@ define_keywords!( MAX, MAXVALUE, MAX_DATA_EXTENSION_TIME_IN_DAYS, + MAX_ROWS, MEASURES, MEDIUMBLOB, MEDIUMINT, @@ -554,6 +562,7 @@ define_keywords!( MINUTE, MINUTES, MINVALUE, + MIN_ROWS, MOD, MODE, MODIFIES, @@ -651,6 +660,7 @@ define_keywords!( OWNERSHIP, PACKAGE, PACKAGES, + PACK_KEYS, PARALLEL, PARAMETER, PARQUET, @@ -773,6 +783,7 @@ define_keywords!( ROW, ROWID, ROWS, + ROW_FORMAT, ROW_NUMBER, RULE, RUN, @@ -787,6 +798,7 @@ define_keywords!( SEARCH, SECOND, SECONDARY, + SECONDARY_ENGINE_ATTRIBUTE, SECONDS, SECRET, SECURITY, @@ -838,12 +850,16 @@ define_keywords!( STATEMENT, STATIC, STATISTICS, + STATS_AUTO_RECALC, + STATS_PERSISTENT, + STATS_SAMPLE_PAGES, STATUS, STDDEV_POP, STDDEV_SAMP, STDIN, STDOUT, STEP, + STORAGE, STORAGE_INTEGRATION, STORAGE_SERIALIZATION_POLICY, STORED, @@ -870,6 +886,7 @@ define_keywords!( TABLE, TABLES, TABLESAMPLE, + TABLESPACE, TAG, TARGET, TASK, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index cbd464c34..a347f3d4d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5524,12 +5524,17 @@ impl<'a> Parser<'a> { }; let location = hive_formats.location.clone(); let table_properties = self.parse_options(Keyword::TBLPROPERTIES)?; + let table_options = if !table_properties.is_empty() { + CreateTableOptions::TableProperties(table_properties) + } else { + CreateTableOptions::None + }; Ok(CreateTableBuilder::new(table_name) .columns(columns) .constraints(constraints) .hive_distribution(hive_distribution) .hive_formats(Some(hive_formats)) - .table_properties(table_properties) + .table_options(table_options) .or_replace(or_replace) .if_not_exists(if_not_exists) .external(true) @@ -7041,17 +7046,16 @@ impl<'a> Parser<'a> { // parse optional column list (schema) let (columns, constraints) = self.parse_columns()?; - let mut comment = if dialect_of!(self is HiveDialect) - && self.parse_keyword(Keyword::COMMENT) - { - let next_token = self.next_token(); - match next_token.token { - Token::SingleQuotedString(str) => Some(CommentDef::AfterColumnDefsWithoutEq(str)), - _ => self.expected("comment", next_token)?, - } - } else { - None - }; + let comment_after_column_def = + if dialect_of!(self is HiveDialect) && self.parse_keyword(Keyword::COMMENT) { + let next_token = self.next_token(); + match next_token.token { + Token::SingleQuotedString(str) => Some(CommentDef::WithoutEq(str)), + _ => self.expected("comment", next_token)?, + } + } else { + None + }; // SQLite supports `WITHOUT ROWID` at the end of `CREATE TABLE` let without_rowid = self.parse_keywords(&[Keyword::WITHOUT, Keyword::ROWID]); @@ -7059,39 +7063,8 @@ impl<'a> Parser<'a> { let hive_distribution = self.parse_hive_distribution()?; let clustered_by = self.parse_optional_clustered_by()?; let hive_formats = self.parse_hive_formats()?; - // PostgreSQL supports `WITH ( options )`, before `AS` - let with_options = self.parse_options(Keyword::WITH)?; - let table_properties = self.parse_options(Keyword::TBLPROPERTIES)?; - let engine = if self.parse_keyword(Keyword::ENGINE) { - self.expect_token(&Token::Eq)?; - let next_token = self.next_token(); - match next_token.token { - Token::Word(w) => { - let name = w.value; - let parameters = if self.peek_token() == Token::LParen { - Some(self.parse_parenthesized_identifiers()?) - } else { - None - }; - Some(TableEngine { name, parameters }) - } - _ => self.expected("identifier", next_token)?, - } - } else { - None - }; - - let auto_increment_offset = if self.parse_keyword(Keyword::AUTO_INCREMENT) { - let _ = self.consume_token(&Token::Eq); - let next_token = self.next_token(); - match next_token.token { - Token::Number(s, _) => Some(Self::parse::(s, next_token.span.start)?), - _ => self.expected("literal int", next_token)?, - } - } else { - None - }; + let create_table_config = self.parse_optional_create_table_config()?; // ClickHouse supports `PRIMARY KEY`, before `ORDER BY` // https://clickhouse.com/docs/en/sql-reference/statements/create/table#primary-key @@ -7119,30 +7092,6 @@ impl<'a> Parser<'a> { None }; - let create_table_config = self.parse_optional_create_table_config()?; - - let default_charset = if self.parse_keywords(&[Keyword::DEFAULT, Keyword::CHARSET]) { - self.expect_token(&Token::Eq)?; - let next_token = self.next_token(); - match next_token.token { - Token::Word(w) => Some(w.value), - _ => self.expected("identifier", next_token)?, - } - } else { - None - }; - - let collation = if self.parse_keywords(&[Keyword::COLLATE]) { - self.expect_token(&Token::Eq)?; - let next_token = self.next_token(); - match next_token.token { - Token::Word(w) => Some(w.value), - _ => self.expected("identifier", next_token)?, - } - } else { - None - }; - let on_commit = if self.parse_keywords(&[Keyword::ON, Keyword::COMMIT]) { Some(self.parse_create_table_on_commit()?) } else { @@ -7151,13 +7100,6 @@ impl<'a> Parser<'a> { let strict = self.parse_keyword(Keyword::STRICT); - // Excludes Hive dialect here since it has been handled after table column definitions. - if !dialect_of!(self is HiveDialect) && self.parse_keyword(Keyword::COMMENT) { - // rewind the COMMENT keyword - self.prev_token(); - comment = self.parse_optional_inline_comment()? - }; - // Parse optional `AS ( query )` let query = if self.parse_keyword(Keyword::AS) { Some(self.parse_query()?) @@ -7174,8 +7116,6 @@ impl<'a> Parser<'a> { .temporary(temporary) .columns(columns) .constraints(constraints) - .with_options(with_options) - .table_properties(table_properties) .or_replace(or_replace) .if_not_exists(if_not_exists) .transient(transient) @@ -7186,19 +7126,15 @@ impl<'a> Parser<'a> { .without_rowid(without_rowid) .like(like) .clone_clause(clone) - .engine(engine) - .comment(comment) - .auto_increment_offset(auto_increment_offset) + .comment_after_column_def(comment_after_column_def) .order_by(order_by) - .default_charset(default_charset) - .collation(collation) .on_commit(on_commit) .on_cluster(on_cluster) .clustered_by(clustered_by) .partition_by(create_table_config.partition_by) .cluster_by(create_table_config.cluster_by) - .options(create_table_config.options) .inherits(create_table_config.inherits) + .table_options(create_table_config.table_options) .primary_key(primary_key) .strict(strict) .build()) @@ -7222,17 +7158,29 @@ impl<'a> Parser<'a> { /// Parse configuration like inheritance, partitioning, clustering information during the table creation. /// /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#syntax_2) - /// [PostgreSQL Partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html) - /// [PostgreSQL Inheritance](https://www.postgresql.org/docs/current/ddl-inherit.html) + /// [PostgreSQL](https://www.postgresql.org/docs/current/ddl-partitioning.html) + /// [MySql](https://dev.mysql.com/doc/refman/8.4/en/create-table.html) fn parse_optional_create_table_config( &mut self, ) -> Result { + let mut table_options = CreateTableOptions::None; + let inherits = if self.parse_keyword(Keyword::INHERITS) { Some(self.parse_parenthesized_qualified_column_list(IsOptional::Mandatory, false)?) } else { None }; + // PostgreSQL supports `WITH ( options )`, before `AS` + let with_options = self.parse_options(Keyword::WITH)?; + if !with_options.is_empty() { + table_options = CreateTableOptions::With(with_options) + } + + let table_properties = self.parse_options(Keyword::TBLPROPERTIES)?; + if !table_properties.is_empty() { + table_options = CreateTableOptions::TableProperties(table_properties); + } let partition_by = if dialect_of!(self is BigQueryDialect | PostgreSqlDialect | GenericDialect) && self.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) { @@ -7242,7 +7190,6 @@ impl<'a> Parser<'a> { }; let mut cluster_by = None; - let mut options = None; if dialect_of!(self is BigQueryDialect | GenericDialect) { if self.parse_keywords(&[Keyword::CLUSTER, Keyword::BY]) { cluster_by = Some(WrappedCollection::NoWrapping( @@ -7252,19 +7199,230 @@ impl<'a> Parser<'a> { if let Token::Word(word) = self.peek_token().token { if word.keyword == Keyword::OPTIONS { - options = Some(self.parse_options(Keyword::OPTIONS)?); + table_options = + CreateTableOptions::Options(self.parse_options(Keyword::OPTIONS)?) } }; } + if !dialect_of!(self is HiveDialect) && table_options == CreateTableOptions::None { + let plain_options = self.parse_plain_options()?; + if !plain_options.is_empty() { + table_options = CreateTableOptions::Plain(plain_options) + } + }; + Ok(CreateTableConfiguration { partition_by, cluster_by, - options, inherits, + table_options, }) } + fn parse_plain_option(&mut self) -> Result, ParserError> { + // Single parameter option + // + if self.parse_keywords(&[Keyword::START, Keyword::TRANSACTION]) { + return Ok(Some(SqlOption::Ident(Ident::new("START TRANSACTION")))); + } + + // Custom option + // + if self.parse_keywords(&[Keyword::COMMENT]) { + let has_eq = self.consume_token(&Token::Eq); + let value = self.next_token(); + + let comment = match (has_eq, value.token) { + (true, Token::SingleQuotedString(s)) => { + Ok(Some(SqlOption::Comment(CommentDef::WithEq(s)))) + } + (false, Token::SingleQuotedString(s)) => { + Ok(Some(SqlOption::Comment(CommentDef::WithoutEq(s)))) + } + (_, token) => { + self.expected("Token::SingleQuotedString", TokenWithSpan::wrap(token)) + } + }; + return comment; + } + + // + // + if self.parse_keywords(&[Keyword::ENGINE]) { + let _ = self.consume_token(&Token::Eq); + let value = self.next_token(); + + let engine = match value.token { + Token::Word(w) => { + let parameters = if self.peek_token() == Token::LParen { + self.parse_parenthesized_identifiers()? + } else { + vec![] + }; + + Ok(Some(SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("ENGINE"), + name: Some(Ident::new(w.value)), + values: parameters, + }, + ))) + } + _ => { + return self.expected("Token::Word", value)?; + } + }; + + return engine; + } + + // + if self.parse_keywords(&[Keyword::TABLESPACE]) { + let _ = self.consume_token(&Token::Eq); + let value = self.next_token(); + + let tablespace = match value.token { + Token::Word(Word { value: name, .. }) | Token::SingleQuotedString(name) => { + let storage = match self.parse_keyword(Keyword::STORAGE) { + true => { + let _ = self.consume_token(&Token::Eq); + let storage_token = self.next_token(); + match &storage_token.token { + Token::Word(w) => match w.value.to_uppercase().as_str() { + "DISK" => Some(StorageType::Disk), + "MEMORY" => Some(StorageType::Memory), + _ => self + .expected("Storage type (DISK or MEMORY)", storage_token)?, + }, + _ => self.expected("Token::Word", storage_token)?, + } + } + false => None, + }; + + Ok(Some(SqlOption::TableSpace(TablespaceOption { + name, + storage, + }))) + } + _ => { + return self.expected("Token::Word", value)?; + } + }; + + return tablespace; + } + + // + if self.parse_keyword(Keyword::UNION) { + let _ = self.consume_token(&Token::Eq); + let value = self.next_token(); + + match value.token { + Token::LParen => { + let tables: Vec = + self.parse_comma_separated0(Parser::parse_identifier, Token::RParen)?; + self.expect_token(&Token::RParen)?; + + return Ok(Some(SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("UNION"), + name: None, + values: tables, + }, + ))); + } + _ => { + return self.expected("Token::LParen", value)?; + } + } + } + + // Key/Value parameter option + let key = if self.parse_keywords(&[Keyword::DEFAULT, Keyword::CHARSET]) { + Ident::new("DEFAULT CHARSET") + } else if self.parse_keyword(Keyword::CHARSET) { + Ident::new("CHARSET") + } else if self.parse_keywords(&[Keyword::DEFAULT, Keyword::CHARACTER, Keyword::SET]) { + Ident::new("DEFAULT CHARACTER SET") + } else if self.parse_keywords(&[Keyword::CHARACTER, Keyword::SET]) { + Ident::new("CHARACTER SET") + } else if self.parse_keywords(&[Keyword::DEFAULT, Keyword::COLLATE]) { + Ident::new("DEFAULT COLLATE") + } else if self.parse_keyword(Keyword::COLLATE) { + Ident::new("COLLATE") + } else if self.parse_keywords(&[Keyword::DATA, Keyword::DIRECTORY]) { + Ident::new("DATA DIRECTORY") + } else if self.parse_keywords(&[Keyword::INDEX, Keyword::DIRECTORY]) { + Ident::new("INDEX DIRECTORY") + } else if self.parse_keyword(Keyword::KEY_BLOCK_SIZE) { + Ident::new("KEY_BLOCK_SIZE") + } else if self.parse_keyword(Keyword::ROW_FORMAT) { + Ident::new("ROW_FORMAT") + } else if self.parse_keyword(Keyword::PACK_KEYS) { + Ident::new("PACK_KEYS") + } else if self.parse_keyword(Keyword::STATS_AUTO_RECALC) { + Ident::new("STATS_AUTO_RECALC") + } else if self.parse_keyword(Keyword::STATS_PERSISTENT) { + Ident::new("STATS_PERSISTENT") + } else if self.parse_keyword(Keyword::STATS_SAMPLE_PAGES) { + Ident::new("STATS_SAMPLE_PAGES") + } else if self.parse_keyword(Keyword::DELAY_KEY_WRITE) { + Ident::new("DELAY_KEY_WRITE") + } else if self.parse_keyword(Keyword::COMPRESSION) { + Ident::new("COMPRESSION") + } else if self.parse_keyword(Keyword::ENCRYPTION) { + Ident::new("ENCRYPTION") + } else if self.parse_keyword(Keyword::MAX_ROWS) { + Ident::new("MAX_ROWS") + } else if self.parse_keyword(Keyword::MIN_ROWS) { + Ident::new("MIN_ROWS") + } else if self.parse_keyword(Keyword::AUTOEXTEND_SIZE) { + Ident::new("AUTOEXTEND_SIZE") + } else if self.parse_keyword(Keyword::AVG_ROW_LENGTH) { + Ident::new("AVG_ROW_LENGTH") + } else if self.parse_keyword(Keyword::CHECKSUM) { + Ident::new("CHECKSUM") + } else if self.parse_keyword(Keyword::CONNECTION) { + Ident::new("CONNECTION") + } else if self.parse_keyword(Keyword::ENGINE_ATTRIBUTE) { + Ident::new("ENGINE_ATTRIBUTE") + } else if self.parse_keyword(Keyword::PASSWORD) { + Ident::new("PASSWORD") + } else if self.parse_keyword(Keyword::SECONDARY_ENGINE_ATTRIBUTE) { + Ident::new("SECONDARY_ENGINE_ATTRIBUTE") + } else if self.parse_keyword(Keyword::INSERT_METHOD) { + Ident::new("INSERT_METHOD") + } else if self.parse_keyword(Keyword::AUTO_INCREMENT) { + Ident::new("AUTO_INCREMENT") + } else { + return Ok(None); + }; + + let _ = self.consume_token(&Token::Eq); + + let value = match self + .maybe_parse(|parser| parser.parse_value())? + .map(Expr::Value) + { + Some(expr) => expr, + None => Expr::Identifier(self.parse_identifier()?), + }; + + Ok(Some(SqlOption::KeyValue { key, value })) + } + + pub fn parse_plain_options(&mut self) -> Result, ParserError> { + let mut options = Vec::new(); + + while let Some(option) = self.parse_plain_option()? { + options.push(option); + } + + Ok(options) + } + pub fn parse_optional_inline_comment(&mut self) -> Result, ParserError> { let comment = if self.parse_keyword(Keyword::COMMENT) { let has_eq = self.consume_token(&Token::Eq); diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 416d2e435..8f54f3c97 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -484,7 +484,7 @@ fn parse_create_table_with_options() { columns, partition_by, cluster_by, - options, + table_options, .. }) => { assert_eq!( @@ -539,7 +539,7 @@ fn parse_create_table_with_options() { Ident::new("userid"), Ident::new("age"), ])), - Some(vec![ + CreateTableOptions::Options(vec![ SqlOption::KeyValue { key: Ident::new("partition_expiration_days"), value: Expr::Value( @@ -561,7 +561,7 @@ fn parse_create_table_with_options() { }, ]) ), - (partition_by, cluster_by, options) + (partition_by, cluster_by, table_options) ) } _ => unreachable!(), diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index c56f98860..d0218b6c3 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -219,10 +219,10 @@ fn parse_delimited_identifiers() { #[test] fn parse_create_table() { - clickhouse().verified_stmt(r#"CREATE TABLE "x" ("a" "int") ENGINE=MergeTree ORDER BY ("x")"#); - clickhouse().verified_stmt(r#"CREATE TABLE "x" ("a" "int") ENGINE=MergeTree ORDER BY "x""#); + clickhouse().verified_stmt(r#"CREATE TABLE "x" ("a" "int") ENGINE = MergeTree ORDER BY ("x")"#); + clickhouse().verified_stmt(r#"CREATE TABLE "x" ("a" "int") ENGINE = MergeTree ORDER BY "x""#); clickhouse().verified_stmt( - r#"CREATE TABLE "x" ("a" "int") ENGINE=MergeTree ORDER BY "x" AS SELECT * FROM "t" WHERE true"#, + r#"CREATE TABLE "x" ("a" "int") ENGINE = MergeTree ORDER BY "x" AS SELECT * FROM "t" WHERE true"#, ); } @@ -589,7 +589,7 @@ fn parse_clickhouse_data_types() { #[test] fn parse_create_table_with_nullable() { - let sql = r#"CREATE TABLE table (k UInt8, `a` Nullable(String), `b` Nullable(DateTime64(9, 'UTC')), c Nullable(DateTime64(9)), d Date32 NULL) ENGINE=MergeTree ORDER BY (`k`)"#; + let sql = r#"CREATE TABLE table (k UInt8, `a` Nullable(String), `b` Nullable(DateTime64(9, 'UTC')), c Nullable(DateTime64(9)), d Date32 NULL) ENGINE = MergeTree ORDER BY (`k`)"#; // ClickHouse has a case-sensitive definition of data type, but canonical representation is not let canonical_sql = sql.replace("String", "STRING"); @@ -714,14 +714,14 @@ fn parse_create_table_with_nested_data_types() { fn parse_create_table_with_primary_key() { match clickhouse_and_generic().verified_stmt(concat!( r#"CREATE TABLE db.table (`i` INT, `k` INT)"#, - " ENGINE=SharedMergeTree('/clickhouse/tables/{uuid}/{shard}', '{replica}')", + " ENGINE = SharedMergeTree('/clickhouse/tables/{uuid}/{shard}', '{replica}')", " PRIMARY KEY tuple(i)", " ORDER BY tuple(i)", )) { Statement::CreateTable(CreateTable { name, columns, - engine, + table_options, primary_key, order_by, .. @@ -742,16 +742,23 @@ fn parse_create_table_with_primary_key() { ], columns ); - assert_eq!( - engine, - Some(TableEngine { - name: "SharedMergeTree".to_string(), - parameters: Some(vec![ + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("ENGINE"), + name: Some(Ident::new("SharedMergeTree")), + values: vec![ Ident::with_quote('\'', "/clickhouse/tables/{uuid}/{shard}"), Ident::with_quote('\'', "{replica}"), - ]), - }) - ); + ] + } + ))); + fn assert_function(actual: &Function, name: &str, arg: &str) -> bool { assert_eq!(actual.name, ObjectName::from(vec![Ident::new(name)])); assert_eq!( @@ -798,7 +805,7 @@ fn parse_create_table_with_variant_default_expressions() { " b DATETIME EPHEMERAL now(),", " c DATETIME EPHEMERAL,", " d STRING ALIAS toString(c)", - ") ENGINE=MergeTree" + ") ENGINE = MergeTree" ); match clickhouse_and_generic().verified_stmt(sql) { Statement::CreateTable(CreateTable { columns, .. }) => { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 1ddf3f92e..7a8b8bdaa 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3657,7 +3657,7 @@ fn parse_create_table() { name, columns, constraints, - with_options, + table_options, if_not_exists: false, external: false, file_format: None, @@ -3795,7 +3795,7 @@ fn parse_create_table() { }, ] ); - assert_eq!(with_options, vec![]); + assert_eq!(table_options, CreateTableOptions::None); } _ => unreachable!(), } @@ -3840,7 +3840,7 @@ fn parse_create_table_with_constraint_characteristics() { name, columns, constraints, - with_options, + table_options, if_not_exists: false, external: false, file_format: None, @@ -3934,7 +3934,7 @@ fn parse_create_table_with_constraint_characteristics() { }, ] ); - assert_eq!(with_options, vec![]); + assert_eq!(table_options, CreateTableOptions::None); } _ => unreachable!(), } @@ -4421,7 +4421,11 @@ fn parse_create_table_with_options() { let sql = "CREATE TABLE t (c INT) WITH (foo = 'bar', a = 123)"; match generic.verified_stmt(sql) { - Statement::CreateTable(CreateTable { with_options, .. }) => { + Statement::CreateTable(CreateTable { table_options, .. }) => { + let with_options = match table_options { + CreateTableOptions::With(options) => options, + _ => unreachable!(), + }; assert_eq!( vec![ SqlOption::KeyValue { @@ -4482,7 +4486,7 @@ fn parse_create_external_table() { name, columns, constraints, - with_options, + table_options, if_not_exists, external, file_format, @@ -4525,7 +4529,7 @@ fn parse_create_external_table() { assert_eq!(FileFormat::TEXTFILE, file_format.unwrap()); assert_eq!("/tmp/example.csv", location.unwrap()); - assert_eq!(with_options, vec![]); + assert_eq!(table_options, CreateTableOptions::None); assert!(!if_not_exists); } _ => unreachable!(), @@ -4550,7 +4554,7 @@ fn parse_create_or_replace_external_table() { name, columns, constraints, - with_options, + table_options, if_not_exists, external, file_format, @@ -4579,7 +4583,7 @@ fn parse_create_or_replace_external_table() { assert_eq!(FileFormat::TEXTFILE, file_format.unwrap()); assert_eq!("/tmp/example.csv", location.unwrap()); - assert_eq!(with_options, vec![]); + assert_eq!(table_options, CreateTableOptions::None); assert!(!if_not_exists); assert!(or_replace); } @@ -11420,7 +11424,9 @@ fn test_parse_inline_comment() { // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) match all_dialects_except(|d| d.is::()).verified_stmt(sql) { Statement::CreateTable(CreateTable { - columns, comment, .. + columns, + table_options, + .. }) => { assert_eq!( columns, @@ -11434,8 +11440,10 @@ fn test_parse_inline_comment() { }] ); assert_eq!( - comment.unwrap(), - CommentDef::WithEq("comment with equal".to_string()) + table_options, + CreateTableOptions::Plain(vec![SqlOption::Comment(CommentDef::WithEq( + "comment with equal".to_string() + ))]) ); } _ => unreachable!(), @@ -12460,21 +12468,6 @@ fn parse_select_wildcard_with_except() { ); } -#[test] -fn parse_auto_increment_too_large() { - let dialect = GenericDialect {}; - let u64_max = u64::MAX; - let sql = - format!("CREATE TABLE foo (bar INT NOT NULL AUTO_INCREMENT) AUTO_INCREMENT=1{u64_max}"); - - let res = Parser::new(&dialect) - .try_with_sql(&sql) - .expect("tokenize to work") - .parse_statements(); - - assert!(res.is_err(), "{res:?}"); -} - #[test] fn test_group_by_nothing() { let Select { group_by, .. } = all_dialects_where(|d| d.supports_group_by_expr()) diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 320583248..8e4983655 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -735,19 +735,13 @@ fn test_duckdb_union_datatype() { storage: Default::default(), location: Default::default() }), - table_properties: Default::default(), - with_options: Default::default(), file_format: Default::default(), location: Default::default(), query: Default::default(), without_rowid: Default::default(), like: Default::default(), clone: Default::default(), - engine: Default::default(), comment: Default::default(), - auto_increment_offset: Default::default(), - default_charset: Default::default(), - collation: Default::default(), on_commit: Default::default(), on_cluster: Default::default(), primary_key: Default::default(), @@ -755,7 +749,6 @@ fn test_duckdb_union_datatype() { partition_by: Default::default(), cluster_by: Default::default(), clustered_by: Default::default(), - options: Default::default(), inherits: Default::default(), strict: Default::default(), copy_grants: Default::default(), @@ -772,6 +765,7 @@ fn test_duckdb_union_datatype() { catalog: Default::default(), catalog_sync: Default::default(), storage_serialization_policy: Default::default(), + table_options: CreateTableOptions::None }), stmt ); diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 9b0430947..14dcbffd1 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -133,9 +133,7 @@ fn create_table_with_comment() { Statement::CreateTable(CreateTable { comment, .. }) => { assert_eq!( comment, - Some(CommentDef::AfterColumnDefsWithoutEq( - "table comment".to_string() - )) + Some(CommentDef::WithoutEq("table comment".to_string())) ) } _ => unreachable!(), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 88e7a1f17..8cc5758fd 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1725,7 +1725,6 @@ fn parse_create_table_with_valid_options() { span: Span::empty(), }, data_type: Int(None,), - options: vec![], }, ColumnDef { @@ -1735,7 +1734,6 @@ fn parse_create_table_with_valid_options() { span: Span::empty(), }, data_type: Int(None,), - options: vec![], }, ], @@ -1747,19 +1745,13 @@ fn parse_create_table_with_valid_options() { storage: None, location: None, },), - table_properties: vec![], - with_options, file_format: None, location: None, query: None, without_rowid: false, like: None, clone: None, - engine: None, comment: None, - auto_increment_offset: None, - default_charset: None, - collation: None, on_commit: None, on_cluster: None, primary_key: None, @@ -1767,7 +1759,6 @@ fn parse_create_table_with_valid_options() { partition_by: None, cluster_by: None, clustered_by: None, - options: None, inherits: None, strict: false, iceberg: false, @@ -1785,6 +1776,7 @@ fn parse_create_table_with_valid_options() { catalog: None, catalog_sync: None, storage_serialization_policy: None, + table_options: CreateTableOptions::With(with_options) }) ); } @@ -1918,19 +1910,13 @@ fn parse_create_table_with_identity_column() { storage: None, location: None, },), - table_properties: vec![], - with_options: vec![], file_format: None, location: None, query: None, without_rowid: false, like: None, clone: None, - engine: None, comment: None, - auto_increment_offset: None, - default_charset: None, - collation: None, on_commit: None, on_cluster: None, primary_key: None, @@ -1938,7 +1924,6 @@ fn parse_create_table_with_identity_column() { partition_by: None, cluster_by: None, clustered_by: None, - options: None, inherits: None, strict: false, copy_grants: false, @@ -1955,6 +1940,7 @@ fn parse_create_table_with_identity_column() { catalog: None, catalog_sync: None, storage_serialization_policy: None, + table_options: CreateTableOptions::None }), ); } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 990107b25..4bb1063dd 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -848,9 +848,23 @@ fn parse_create_table_comment() { for sql in [without_equal, with_equal] { match mysql().verified_stmt(sql) { - Statement::CreateTable(CreateTable { name, comment, .. }) => { + Statement::CreateTable(CreateTable { + name, + table_options, + .. + }) => { assert_eq!(name.to_string(), "foo"); - assert_eq!(comment.expect("Should exist").to_string(), "baz"); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + let comment = match plain_options.first().unwrap() { + SqlOption::Comment(CommentDef::WithEq(c)) + | SqlOption::Comment(CommentDef::WithoutEq(c)) => c, + _ => unreachable!(), + }; + assert_eq!(comment, "baz"); } _ => unreachable!(), } @@ -859,29 +873,226 @@ fn parse_create_table_comment() { #[test] fn parse_create_table_auto_increment_offset() { - let canonical = - "CREATE TABLE foo (bar INT NOT NULL AUTO_INCREMENT) ENGINE=InnoDB AUTO_INCREMENT 123"; - let with_equal = - "CREATE TABLE foo (bar INT NOT NULL AUTO_INCREMENT) ENGINE=InnoDB AUTO_INCREMENT=123"; + let sql = + "CREATE TABLE foo (bar INT NOT NULL AUTO_INCREMENT) ENGINE = InnoDB AUTO_INCREMENT = 123"; + + match mysql().verified_stmt(sql) { + Statement::CreateTable(CreateTable { + name, + table_options, + .. + }) => { + assert_eq!(name.to_string(), "foo"); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("AUTO_INCREMENT"), + value: Expr::Value(test_utils::number("123").with_empty_span()) + })); + } + _ => unreachable!(), + } +} - for sql in [canonical, with_equal] { - match mysql().one_statement_parses_to(sql, canonical) { +#[test] +fn parse_create_table_multiple_options_order_independent() { + let sql1 = "CREATE TABLE mytable (id INT) ENGINE=InnoDB ROW_FORMAT=DYNAMIC KEY_BLOCK_SIZE=8 COMMENT='abc'"; + let sql2 = "CREATE TABLE mytable (id INT) KEY_BLOCK_SIZE=8 COMMENT='abc' ENGINE=InnoDB ROW_FORMAT=DYNAMIC"; + let sql3 = "CREATE TABLE mytable (id INT) ROW_FORMAT=DYNAMIC KEY_BLOCK_SIZE=8 COMMENT='abc' ENGINE=InnoDB"; + + for sql in [sql1, sql2, sql3] { + match mysql().parse_sql_statements(sql).unwrap().pop().unwrap() { Statement::CreateTable(CreateTable { name, - auto_increment_offset, + table_options, .. }) => { - assert_eq!(name.to_string(), "foo"); - assert_eq!( - auto_increment_offset.expect("Should exist").to_string(), - "123" - ); + assert_eq!(name.to_string(), "mytable"); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("ENGINE"), + name: Some(Ident::new("InnoDB")), + values: vec![] + } + ))); + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("KEY_BLOCK_SIZE"), + value: Expr::Value(test_utils::number("8").with_empty_span()) + })); + + assert!(plain_options + .contains(&SqlOption::Comment(CommentDef::WithEq("abc".to_owned())))); + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("ROW_FORMAT"), + value: Expr::Identifier(Ident::new("DYNAMIC".to_owned())) + })); } _ => unreachable!(), } } } +#[test] +fn parse_create_table_with_all_table_options() { + let sql = + "CREATE TABLE foo (bar INT NOT NULL AUTO_INCREMENT) ENGINE = InnoDB AUTO_INCREMENT = 123 DEFAULT CHARSET = utf8mb4 COLLATE = utf8mb4_0900_ai_ci INSERT_METHOD = FIRST KEY_BLOCK_SIZE = 8 ROW_FORMAT = DYNAMIC DATA DIRECTORY = '/var/lib/mysql/data' INDEX DIRECTORY = '/var/lib/mysql/index' PACK_KEYS = 1 STATS_AUTO_RECALC = 1 STATS_PERSISTENT = 0 STATS_SAMPLE_PAGES = 128 DELAY_KEY_WRITE = 1 COMPRESSION = 'ZLIB' ENCRYPTION = 'Y' MAX_ROWS = 10000 MIN_ROWS = 10 AUTOEXTEND_SIZE = 64 AVG_ROW_LENGTH = 128 CHECKSUM = 1 CONNECTION = 'mysql://localhost' ENGINE_ATTRIBUTE = 'primary' PASSWORD = 'secure_password' SECONDARY_ENGINE_ATTRIBUTE = 'secondary_attr' START TRANSACTION TABLESPACE my_tablespace STORAGE DISK UNION = (table1, table2, table3)"; + + match mysql().verified_stmt(sql) { + Statement::CreateTable(CreateTable { + name, + table_options, + .. + }) => { + assert_eq!(name, vec![Ident::new("foo".to_owned())].into()); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("ENGINE"), + name: Some(Ident::new("InnoDB")), + values: vec![] + } + ))); + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("COLLATE"), + value: Expr::Identifier(Ident::new("utf8mb4_0900_ai_ci".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("DEFAULT CHARSET"), + value: Expr::Identifier(Ident::new("utf8mb4".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("AUTO_INCREMENT"), + value: Expr::value(test_utils::number("123")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("KEY_BLOCK_SIZE"), + value: Expr::value(test_utils::number("8")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("ROW_FORMAT"), + value: Expr::Identifier(Ident::new("DYNAMIC".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("PACK_KEYS"), + value: Expr::value(test_utils::number("1")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("STATS_AUTO_RECALC"), + value: Expr::value(test_utils::number("1")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("STATS_PERSISTENT"), + value: Expr::value(test_utils::number("0")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("STATS_SAMPLE_PAGES"), + value: Expr::value(test_utils::number("128")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("STATS_SAMPLE_PAGES"), + value: Expr::value(test_utils::number("128")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("INSERT_METHOD"), + value: Expr::Identifier(Ident::new("FIRST".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("COMPRESSION"), + value: Expr::value(Value::SingleQuotedString("ZLIB".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("ENCRYPTION"), + value: Expr::value(Value::SingleQuotedString("Y".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("MAX_ROWS"), + value: Expr::value(test_utils::number("10000")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("MIN_ROWS"), + value: Expr::value(test_utils::number("10")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("AUTOEXTEND_SIZE"), + value: Expr::value(test_utils::number("64")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("AVG_ROW_LENGTH"), + value: Expr::value(test_utils::number("128")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("CHECKSUM"), + value: Expr::value(test_utils::number("1")) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("CONNECTION"), + value: Expr::value(Value::SingleQuotedString("mysql://localhost".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("ENGINE_ATTRIBUTE"), + value: Expr::value(Value::SingleQuotedString("primary".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("PASSWORD"), + value: Expr::value(Value::SingleQuotedString("secure_password".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("SECONDARY_ENGINE_ATTRIBUTE"), + value: Expr::value(Value::SingleQuotedString("secondary_attr".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::Ident(Ident::new( + "START TRANSACTION".to_owned() + )))); + assert!( + plain_options.contains(&SqlOption::TableSpace(TablespaceOption { + name: "my_tablespace".to_string(), + storage: Some(StorageType::Disk), + })) + ); + + assert!(plain_options.contains(&SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("UNION"), + name: None, + values: vec![ + Ident::new("table1".to_string()), + Ident::new("table2".to_string()), + Ident::new("table3".to_string()) + ] + } + ))); + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("DATA DIRECTORY"), + value: Expr::value(Value::SingleQuotedString("/var/lib/mysql/data".to_owned())) + })); + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("INDEX DIRECTORY"), + value: Expr::value(Value::SingleQuotedString("/var/lib/mysql/index".to_owned())) + })); + } + _ => unreachable!(), + } +} + #[test] fn parse_create_table_set_enum() { let sql = "CREATE TABLE foo (bar SET('a', 'b'), baz ENUM('a', 'b'))"; @@ -916,13 +1127,12 @@ fn parse_create_table_set_enum() { #[test] fn parse_create_table_engine_default_charset() { - let sql = "CREATE TABLE foo (id INT(11)) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3"; + let sql = "CREATE TABLE foo (id INT(11)) ENGINE = InnoDB DEFAULT CHARSET = utf8mb3"; match mysql().verified_stmt(sql) { Statement::CreateTable(CreateTable { name, columns, - engine, - default_charset, + table_options, .. }) => { assert_eq!(name.to_string(), "foo"); @@ -934,14 +1144,24 @@ fn parse_create_table_engine_default_charset() { },], columns ); - assert_eq!( - engine, - Some(TableEngine { - name: "InnoDB".to_string(), - parameters: None - }) - ); - assert_eq!(default_charset, Some("utf8mb3".to_string())); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("DEFAULT CHARSET"), + value: Expr::Identifier(Ident::new("utf8mb3".to_owned())) + })); + + assert!(plain_options.contains(&SqlOption::NamedParenthesizedList( + NamedParenthesizedList { + key: Ident::new("ENGINE"), + name: Some(Ident::new("InnoDB")), + values: vec![] + } + ))); } _ => unreachable!(), } @@ -949,12 +1169,12 @@ fn parse_create_table_engine_default_charset() { #[test] fn parse_create_table_collate() { - let sql = "CREATE TABLE foo (id INT(11)) COLLATE=utf8mb4_0900_ai_ci"; + let sql = "CREATE TABLE foo (id INT(11)) COLLATE = utf8mb4_0900_ai_ci"; match mysql().verified_stmt(sql) { Statement::CreateTable(CreateTable { name, columns, - collation, + table_options, .. }) => { assert_eq!(name.to_string(), "foo"); @@ -966,7 +1186,16 @@ fn parse_create_table_collate() { },], columns ); - assert_eq!(collation, Some("utf8mb4_0900_ai_ci".to_string())); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("COLLATE"), + value: Expr::Identifier(Ident::new("utf8mb4_0900_ai_ci".to_owned())) + })); } _ => unreachable!(), } @@ -974,16 +1203,26 @@ fn parse_create_table_collate() { #[test] fn parse_create_table_both_options_and_as_query() { - let sql = "CREATE TABLE foo (id INT(11)) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3 COLLATE=utf8mb4_0900_ai_ci AS SELECT 1"; + let sql = "CREATE TABLE foo (id INT(11)) ENGINE = InnoDB DEFAULT CHARSET = utf8mb3 COLLATE = utf8mb4_0900_ai_ci AS SELECT 1"; match mysql_and_generic().verified_stmt(sql) { Statement::CreateTable(CreateTable { name, - collation, query, + table_options, .. }) => { assert_eq!(name.to_string(), "foo"); - assert_eq!(collation, Some("utf8mb4_0900_ai_ci".to_string())); + + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + + assert!(plain_options.contains(&SqlOption::KeyValue { + key: Ident::new("COLLATE"), + value: Expr::Identifier(Ident::new("utf8mb4_0900_ai_ci".to_owned())) + })); + assert_eq!( query.unwrap().body.as_select().unwrap().projection, vec![SelectItem::UnnamedExpr(Expr::Value( @@ -994,7 +1233,8 @@ fn parse_create_table_both_options_and_as_query() { _ => unreachable!(), } - let sql = r"CREATE TABLE foo (id INT(11)) ENGINE=InnoDB AS SELECT 1 DEFAULT CHARSET=utf8mb3"; + let sql = + r"CREATE TABLE foo (id INT(11)) ENGINE = InnoDB AS SELECT 1 DEFAULT CHARSET = utf8mb3"; assert!(matches!( mysql_and_generic().parse_sql_statements(sql), Err(ParserError::ParserError(_)) diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 6c008c84d..1fb7432a4 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -348,7 +348,7 @@ fn parse_create_table_with_defaults() { name, columns, constraints, - with_options, + table_options, if_not_exists: false, external: false, file_format: None, @@ -485,6 +485,11 @@ fn parse_create_table_with_defaults() { ] ); assert!(constraints.is_empty()); + + let with_options = match table_options { + CreateTableOptions::With(options) => options, + _ => unreachable!(), + }; assert_eq!( with_options, vec![ @@ -4668,7 +4673,6 @@ fn parse_create_table_with_alias() { name, columns, constraints, - with_options: _with_options, if_not_exists: false, external: false, file_format: None, @@ -5078,7 +5082,11 @@ fn parse_at_time_zone() { fn parse_create_table_with_options() { let sql = "CREATE TABLE t (c INT) WITH (foo = 'bar', a = 123)"; match pg().verified_stmt(sql) { - Statement::CreateTable(CreateTable { with_options, .. }) => { + Statement::CreateTable(CreateTable { table_options, .. }) => { + let with_options = match table_options { + CreateTableOptions::With(options) => options, + _ => unreachable!(), + }; assert_eq!( vec![ SqlOption::KeyValue { @@ -5506,19 +5514,13 @@ fn parse_trigger_related_functions() { storage: None, location: None }), - table_properties: vec![], - with_options: vec![], file_format: None, location: None, query: None, without_rowid: false, like: None, clone: None, - engine: None, comment: None, - auto_increment_offset: None, - default_charset: None, - collation: None, on_commit: None, on_cluster: None, primary_key: None, @@ -5526,7 +5528,6 @@ fn parse_trigger_related_functions() { partition_by: None, cluster_by: None, clustered_by: None, - options: None, inherits: None, strict: false, copy_grants: false, @@ -5543,6 +5544,7 @@ fn parse_trigger_related_functions() { catalog: None, catalog_sync: None, storage_serialization_policy: None, + table_options: CreateTableOptions::None } ); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index aa974115d..52be31435 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -470,9 +470,22 @@ fn test_snowflake_create_table_cluster_by() { #[test] fn test_snowflake_create_table_comment() { match snowflake().verified_stmt("CREATE TABLE my_table (a INT) COMMENT = 'some comment'") { - Statement::CreateTable(CreateTable { name, comment, .. }) => { + Statement::CreateTable(CreateTable { + name, + table_options, + .. + }) => { assert_eq!("my_table", name.to_string()); - assert_eq!("some comment", comment.unwrap().to_string()); + let plain_options = match table_options { + CreateTableOptions::Plain(options) => options, + _ => unreachable!(), + }; + let comment = match plain_options.first().unwrap() { + SqlOption::Comment(CommentDef::WithEq(c)) + | SqlOption::Comment(CommentDef::WithoutEq(c)) => c, + _ => unreachable!(), + }; + assert_eq!("some comment", comment); } _ => unreachable!(), } From a497358c3a3ef24cb346f5e8f071c3bd65fd0cdc Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Sat, 3 May 2025 10:59:13 -0400 Subject: [PATCH 119/286] Add `CREATE TRIGGER` support for SQL Server (#1810) --- src/ast/mod.rs | 48 +++++++++++++---- src/ast/trigger.rs | 2 + src/dialect/mssql.rs | 46 ++++++++++++++++ src/parser/mod.rs | 43 ++++++++++----- tests/sqlparser_mssql.rs | 105 ++++++++++++++++++++++++++++++++++++ tests/sqlparser_mysql.rs | 6 ++- tests/sqlparser_postgres.rs | 38 ++++++++----- 7 files changed, 251 insertions(+), 37 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d74d197e3..c3009743d 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2380,11 +2380,16 @@ impl fmt::Display for BeginEndStatements { end_token: AttachedToken(end_token), } = self; - write!(f, "{begin_token} ")?; + if begin_token.token != Token::EOF { + write!(f, "{begin_token} ")?; + } if !statements.is_empty() { format_statement_list(f, statements)?; } - write!(f, " {end_token}") + if end_token.token != Token::EOF { + write!(f, " {end_token}")?; + } + Ok(()) } } @@ -3729,7 +3734,12 @@ pub enum Statement { /// ``` /// /// Postgres: + /// SQL Server: CreateTrigger { + /// True if this is a `CREATE OR ALTER TRIGGER` statement + /// + /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-trigger-transact-sql?view=sql-server-ver16#arguments) + or_alter: bool, /// The `OR REPLACE` clause is used to re-create the trigger if it already exists. /// /// Example: @@ -3790,7 +3800,9 @@ pub enum Statement { /// Triggering conditions condition: Option, /// Execute logic block - exec_body: TriggerExecBody, + exec_body: Option, + /// For SQL dialects with statement(s) for a body + statements: Option, /// The characteristic of the trigger, which include whether the trigger is `DEFERRABLE`, `INITIALLY DEFERRED`, or `INITIALLY IMMEDIATE`, characteristics: Option, }, @@ -4587,6 +4599,7 @@ impl fmt::Display for Statement { } Statement::CreateFunction(create_function) => create_function.fmt(f), Statement::CreateTrigger { + or_alter, or_replace, is_constraint, name, @@ -4599,19 +4612,30 @@ impl fmt::Display for Statement { condition, include_each, exec_body, + statements, characteristics, } => { write!( f, - "CREATE {or_replace}{is_constraint}TRIGGER {name} {period}", + "CREATE {or_alter}{or_replace}{is_constraint}TRIGGER {name} ", + or_alter = if *or_alter { "OR ALTER " } else { "" }, or_replace = if *or_replace { "OR REPLACE " } else { "" }, is_constraint = if *is_constraint { "CONSTRAINT " } else { "" }, )?; - if !events.is_empty() { - write!(f, " {}", display_separated(events, " OR "))?; + if exec_body.is_some() { + write!(f, "{period}")?; + if !events.is_empty() { + write!(f, " {}", display_separated(events, " OR "))?; + } + write!(f, " ON {table_name}")?; + } else { + write!(f, "ON {table_name}")?; + write!(f, " {period}")?; + if !events.is_empty() { + write!(f, " {}", display_separated(events, ", "))?; + } } - write!(f, " ON {table_name}")?; if let Some(referenced_table_name) = referenced_table_name { write!(f, " FROM {referenced_table_name}")?; @@ -4627,13 +4651,19 @@ impl fmt::Display for Statement { if *include_each { write!(f, " FOR EACH {trigger_object}")?; - } else { + } else if exec_body.is_some() { write!(f, " FOR {trigger_object}")?; } if let Some(condition) = condition { write!(f, " WHEN {condition}")?; } - write!(f, " EXECUTE {exec_body}") + if let Some(exec_body) = exec_body { + write!(f, " EXECUTE {exec_body}")?; + } + if let Some(statements) = statements { + write!(f, " AS {statements}")?; + } + Ok(()) } Statement::DropTrigger { if_exists, diff --git a/src/ast/trigger.rs b/src/ast/trigger.rs index cf1c8c466..2c64e4239 100644 --- a/src/ast/trigger.rs +++ b/src/ast/trigger.rs @@ -110,6 +110,7 @@ impl fmt::Display for TriggerEvent { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum TriggerPeriod { + For, After, Before, InsteadOf, @@ -118,6 +119,7 @@ pub enum TriggerPeriod { impl fmt::Display for TriggerPeriod { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { + TriggerPeriod::For => write!(f, "FOR"), TriggerPeriod::After => write!(f, "AFTER"), TriggerPeriod::Before => write!(f, "BEFORE"), TriggerPeriod::InsteadOf => write!(f, "INSTEAD OF"), diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 31e324f06..647e82a2a 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -18,6 +18,7 @@ use crate::ast::helpers::attached_token::AttachedToken; use crate::ast::{ BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, IfStatement, Statement, + TriggerObject, }; use crate::dialect::Dialect; use crate::keywords::{self, Keyword}; @@ -125,6 +126,15 @@ impl Dialect for MsSqlDialect { fn parse_statement(&self, parser: &mut Parser) -> Option> { if parser.peek_keyword(Keyword::IF) { Some(self.parse_if_stmt(parser)) + } else if parser.parse_keywords(&[Keyword::CREATE, Keyword::TRIGGER]) { + Some(self.parse_create_trigger(parser, false)) + } else if parser.parse_keywords(&[ + Keyword::CREATE, + Keyword::OR, + Keyword::ALTER, + Keyword::TRIGGER, + ]) { + Some(self.parse_create_trigger(parser, true)) } else { None } @@ -215,6 +225,42 @@ impl MsSqlDialect { })) } + /// Parse `CREATE TRIGGER` for [MsSql] + /// + /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-trigger-transact-sql + fn parse_create_trigger( + &self, + parser: &mut Parser, + or_alter: bool, + ) -> Result { + let name = parser.parse_object_name(false)?; + parser.expect_keyword_is(Keyword::ON)?; + let table_name = parser.parse_object_name(false)?; + let period = parser.parse_trigger_period()?; + let events = parser.parse_comma_separated(Parser::parse_trigger_event)?; + + parser.expect_keyword_is(Keyword::AS)?; + let statements = Some(parser.parse_conditional_statements(&[Keyword::END])?); + + Ok(Statement::CreateTrigger { + or_alter, + or_replace: false, + is_constraint: false, + name, + period, + events, + table_name, + referenced_table_name: None, + referencing: Vec::new(), + trigger_object: TriggerObject::Statement, + include_each: false, + condition: None, + exec_body: None, + statements, + characteristics: None, + }) + } + /// Parse a sequence of statements, optionally separated by semicolon. /// /// Stops parsing when reaching EOF or the given keyword. diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a347f3d4d..2011d31e3 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -745,26 +745,38 @@ impl<'a> Parser<'a> { } }; + let conditional_statements = self.parse_conditional_statements(terminal_keywords)?; + + Ok(ConditionalStatementBlock { + start_token: AttachedToken(start_token), + condition, + then_token, + conditional_statements, + }) + } + + /// Parse a BEGIN/END block or a sequence of statements + /// This could be inside of a conditional (IF, CASE, WHILE etc.) or an object body defined optionally BEGIN/END and one or more statements. + pub(crate) fn parse_conditional_statements( + &mut self, + terminal_keywords: &[Keyword], + ) -> Result { let conditional_statements = if self.peek_keyword(Keyword::BEGIN) { let begin_token = self.expect_keyword(Keyword::BEGIN)?; let statements = self.parse_statement_list(terminal_keywords)?; let end_token = self.expect_keyword(Keyword::END)?; + ConditionalStatements::BeginEnd(BeginEndStatements { begin_token: AttachedToken(begin_token), statements, end_token: AttachedToken(end_token), }) } else { - let statements = self.parse_statement_list(terminal_keywords)?; - ConditionalStatements::Sequence { statements } + ConditionalStatements::Sequence { + statements: self.parse_statement_list(terminal_keywords)?, + } }; - - Ok(ConditionalStatementBlock { - start_token: AttachedToken(start_token), - condition, - then_token, - conditional_statements, - }) + Ok(conditional_statements) } /// Parse a `RAISE` statement. @@ -4614,9 +4626,9 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::FUNCTION) { self.parse_create_function(or_alter, or_replace, temporary) } else if self.parse_keyword(Keyword::TRIGGER) { - self.parse_create_trigger(or_replace, false) + self.parse_create_trigger(or_alter, or_replace, false) } else if self.parse_keywords(&[Keyword::CONSTRAINT, Keyword::TRIGGER]) { - self.parse_create_trigger(or_replace, true) + self.parse_create_trigger(or_alter, or_replace, true) } else if self.parse_keyword(Keyword::MACRO) { self.parse_create_macro(or_replace, temporary) } else if self.parse_keyword(Keyword::SECRET) { @@ -5314,10 +5326,11 @@ impl<'a> Parser<'a> { pub fn parse_create_trigger( &mut self, + or_alter: bool, or_replace: bool, is_constraint: bool, ) -> Result { - if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect) { + if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect | MsSqlDialect) { self.prev_token(); return self.expected("an object type after CREATE", self.peek_token()); } @@ -5363,6 +5376,7 @@ impl<'a> Parser<'a> { let exec_body = self.parse_trigger_exec_body()?; Ok(Statement::CreateTrigger { + or_alter, or_replace, is_constraint, name, @@ -5374,7 +5388,8 @@ impl<'a> Parser<'a> { trigger_object, include_each, condition, - exec_body, + exec_body: Some(exec_body), + statements: None, characteristics, }) } @@ -5382,10 +5397,12 @@ impl<'a> Parser<'a> { pub fn parse_trigger_period(&mut self) -> Result { Ok( match self.expect_one_of_keywords(&[ + Keyword::FOR, Keyword::BEFORE, Keyword::AFTER, Keyword::INSTEAD, ])? { + Keyword::FOR => TriggerPeriod::For, Keyword::BEFORE => TriggerPeriod::Before, Keyword::AFTER => TriggerPeriod::After, Keyword::INSTEAD => self diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 8cc5758fd..9ff55198f 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -273,6 +273,16 @@ fn parse_create_function() { END\ "; let _ = ms().verified_stmt(create_or_alter_function); + + let create_function_with_return_expression = "\ + CREATE FUNCTION some_scalar_udf(@foo INT, @bar VARCHAR(256)) \ + RETURNS INT \ + AS \ + BEGIN \ + RETURN CONVERT(INT, 1) + 2; \ + END\ + "; + let _ = ms().verified_stmt(create_function_with_return_expression); } #[test] @@ -2199,6 +2209,101 @@ fn parse_mssql_merge_with_output() { ms_and_generic().verified_stmt(stmt); } +#[test] +fn parse_create_trigger() { + let create_trigger = "\ + CREATE OR ALTER TRIGGER reminder1 \ + ON Sales.Customer \ + AFTER INSERT, UPDATE \ + AS RAISERROR('Notify Customer Relations', 16, 10);\ + "; + let create_stmt = ms().verified_stmt(create_trigger); + assert_eq!( + create_stmt, + Statement::CreateTrigger { + or_alter: true, + or_replace: false, + is_constraint: false, + name: ObjectName::from(vec![Ident::new("reminder1")]), + period: TriggerPeriod::After, + events: vec![TriggerEvent::Insert, TriggerEvent::Update(vec![]),], + table_name: ObjectName::from(vec![Ident::new("Sales"), Ident::new("Customer")]), + referenced_table_name: None, + referencing: vec![], + trigger_object: TriggerObject::Statement, + include_each: false, + condition: None, + exec_body: None, + statements: Some(ConditionalStatements::Sequence { + statements: vec![Statement::RaisError { + message: Box::new(Expr::Value( + (Value::SingleQuotedString("Notify Customer Relations".to_string())) + .with_empty_span() + )), + severity: Box::new(Expr::Value( + (Value::Number("16".parse().unwrap(), false)).with_empty_span() + )), + state: Box::new(Expr::Value( + (Value::Number("10".parse().unwrap(), false)).with_empty_span() + )), + arguments: vec![], + options: vec![], + }], + }), + characteristics: None, + } + ); + + let multi_statement_as_trigger = "\ + CREATE TRIGGER some_trigger ON some_table FOR INSERT \ + AS \ + DECLARE @var INT; \ + RAISERROR('Trigger fired', 10, 1);\ + "; + let _ = ms().verified_stmt(multi_statement_as_trigger); + + let multi_statement_trigger = "\ + CREATE TRIGGER some_trigger ON some_table FOR INSERT \ + AS \ + BEGIN \ + DECLARE @var INT; \ + RAISERROR('Trigger fired', 10, 1); \ + END\ + "; + let _ = ms().verified_stmt(multi_statement_trigger); + + let create_trigger_with_return = "\ + CREATE TRIGGER some_trigger ON some_table FOR INSERT \ + AS \ + BEGIN \ + RETURN; \ + END\ + "; + let _ = ms().verified_stmt(create_trigger_with_return); + + let create_trigger_with_return = "\ + CREATE TRIGGER some_trigger ON some_table FOR INSERT \ + AS \ + BEGIN \ + RETURN; \ + END\ + "; + let _ = ms().verified_stmt(create_trigger_with_return); + + let create_trigger_with_conditional = "\ + CREATE TRIGGER some_trigger ON some_table FOR INSERT \ + AS \ + BEGIN \ + IF 1 = 2 \ + BEGIN \ + RAISERROR('Trigger fired', 10, 1); \ + END; \ + RETURN; \ + END\ + "; + let _ = ms().verified_stmt(create_trigger_with_conditional); +} + #[test] fn parse_drop_trigger() { let sql_drop_trigger = "DROP TRIGGER emp_stamp;"; diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 4bb1063dd..27c60b052 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3779,6 +3779,7 @@ fn parse_create_trigger() { assert_eq!( create_stmt, Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), @@ -3790,13 +3791,14 @@ fn parse_create_trigger() { trigger_object: TriggerObject::Row, include_each: true, condition: None, - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("emp_stamp")]), args: None, } - }, + }), + statements: None, characteristics: None, } ); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 1fb7432a4..008d0670a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5157,6 +5157,7 @@ fn test_escaped_string_literal() { fn parse_create_simple_before_insert_trigger() { let sql = "CREATE TRIGGER check_insert BEFORE INSERT ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_insert"; let expected = Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_insert")]), @@ -5168,13 +5169,14 @@ fn parse_create_simple_before_insert_trigger() { trigger_object: TriggerObject::Row, include_each: true, condition: None, - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("check_account_insert")]), args: None, }, - }, + }), + statements: None, characteristics: None, }; @@ -5185,6 +5187,7 @@ fn parse_create_simple_before_insert_trigger() { fn parse_create_after_update_trigger_with_condition() { let sql = "CREATE TRIGGER check_update AFTER UPDATE ON accounts FOR EACH ROW WHEN (NEW.balance > 10000) EXECUTE FUNCTION check_account_update"; let expected = Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_update")]), @@ -5203,13 +5206,14 @@ fn parse_create_after_update_trigger_with_condition() { op: BinaryOperator::Gt, right: Box::new(Expr::value(number("10000"))), }))), - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("check_account_update")]), args: None, }, - }, + }), + statements: None, characteristics: None, }; @@ -5220,6 +5224,7 @@ fn parse_create_after_update_trigger_with_condition() { fn parse_create_instead_of_delete_trigger() { let sql = "CREATE TRIGGER check_delete INSTEAD OF DELETE ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_deletes"; let expected = Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_delete")]), @@ -5231,13 +5236,14 @@ fn parse_create_instead_of_delete_trigger() { trigger_object: TriggerObject::Row, include_each: true, condition: None, - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("check_account_deletes")]), args: None, }, - }, + }), + statements: None, characteristics: None, }; @@ -5248,6 +5254,7 @@ fn parse_create_instead_of_delete_trigger() { fn parse_create_trigger_with_multiple_events_and_deferrable() { let sql = "CREATE CONSTRAINT TRIGGER check_multiple_events BEFORE INSERT OR UPDATE OR DELETE ON accounts DEFERRABLE INITIALLY DEFERRED FOR EACH ROW EXECUTE FUNCTION check_account_changes"; let expected = Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: true, name: ObjectName::from(vec![Ident::new("check_multiple_events")]), @@ -5263,13 +5270,14 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { trigger_object: TriggerObject::Row, include_each: true, condition: None, - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("check_account_changes")]), args: None, }, - }, + }), + statements: None, characteristics: Some(ConstraintCharacteristics { deferrable: Some(true), initially: Some(DeferrableInitial::Deferred), @@ -5284,6 +5292,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { fn parse_create_trigger_with_referencing() { let sql = "CREATE TRIGGER check_referencing BEFORE INSERT ON accounts REFERENCING NEW TABLE AS new_accounts OLD TABLE AS old_accounts FOR EACH ROW EXECUTE FUNCTION check_account_referencing"; let expected = Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_referencing")]), @@ -5306,13 +5315,14 @@ fn parse_create_trigger_with_referencing() { trigger_object: TriggerObject::Row, include_each: true, condition: None, - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("check_account_referencing")]), args: None, }, - }, + }), + statements: None, characteristics: None, }; @@ -5332,7 +5342,7 @@ fn parse_create_trigger_invalid_cases() { ), ( "CREATE TRIGGER check_update TOMORROW UPDATE ON accounts EXECUTE FUNCTION check_account_update", - "Expected: one of BEFORE or AFTER or INSTEAD, found: TOMORROW" + "Expected: one of FOR or BEFORE or AFTER or INSTEAD, found: TOMORROW" ), ( "CREATE TRIGGER check_update BEFORE SAVE ON accounts EXECUTE FUNCTION check_account_update", @@ -5590,6 +5600,7 @@ fn parse_trigger_related_functions() { assert_eq!( create_trigger, Statement::CreateTrigger { + or_alter: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), @@ -5601,13 +5612,14 @@ fn parse_trigger_related_functions() { trigger_object: TriggerObject::Row, include_each: true, condition: None, - exec_body: TriggerExecBody { + exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("emp_stamp")]), args: None, } - }, + }), + statements: None, characteristics: None } ); From ac1c339666c68779ed7f20dc0fb3b7473b298f83 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Sun, 4 May 2025 23:21:44 +0200 Subject: [PATCH 120/286] Added support for `CREATE DOMAIN` (#1830) --- src/ast/ddl.rs | 49 +++++++++++++++++++ src/ast/mod.rs | 15 +++--- src/ast/spans.rs | 1 + src/parser/mod.rs | 31 +++++++++++++ tests/sqlparser_postgres.rs | 93 +++++++++++++++++++++++++++++++++++++ 5 files changed, 183 insertions(+), 6 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index c1c113b32..a457a0655 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2153,6 +2153,55 @@ impl fmt::Display for ClusteredBy { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// ```sql +/// CREATE DOMAIN name [ AS ] data_type +/// [ COLLATE collation ] +/// [ DEFAULT expression ] +/// [ domain_constraint [ ... ] ] +/// +/// where domain_constraint is: +/// +/// [ CONSTRAINT constraint_name ] +/// { NOT NULL | NULL | CHECK (expression) } +/// ``` +/// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-createdomain.html) +pub struct CreateDomain { + /// The name of the domain to be created. + pub name: ObjectName, + /// The data type of the domain. + pub data_type: DataType, + /// The collation of the domain. + pub collation: Option, + /// The default value of the domain. + pub default: Option, + /// The constraints of the domain. + pub constraints: Vec, +} + +impl fmt::Display for CreateDomain { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "CREATE DOMAIN {name} AS {data_type}", + name = self.name, + data_type = self.data_type + )?; + if let Some(collation) = &self.collation { + write!(f, " COLLATE {collation}")?; + } + if let Some(default) = &self.default { + write!(f, " DEFAULT {default}")?; + } + if !self.constraints.is_empty() { + write!(f, " {}", display_separated(&self.constraints, " "))?; + } + Ok(()) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/mod.rs b/src/ast/mod.rs index c3009743d..a72cb8add 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -55,12 +55,12 @@ pub use self::ddl::{ AlterTableAlgorithm, AlterTableLock, AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, - ConstraintCharacteristics, CreateConnector, CreateFunction, Deduplicate, DeferrableInitial, - DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, - IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexOption, - IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, - ReferentialAction, TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, - UserDefinedTypeRepresentation, ViewColumnDef, + ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, Deduplicate, + DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, + IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, + IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, + ProcedureParam, ReferentialAction, TableConstraint, TagsColumnOption, + UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; @@ -4049,6 +4049,8 @@ pub enum Statement { sequence_options: Vec, owned_by: Option, }, + /// A `CREATE DOMAIN` statement. + CreateDomain(CreateDomain), /// ```sql /// CREATE TYPE /// ``` @@ -4598,6 +4600,7 @@ impl fmt::Display for Statement { Ok(()) } Statement::CreateFunction(create_function) => create_function.fmt(f), + Statement::CreateDomain(create_domain) => create_domain.fmt(f), Statement::CreateTrigger { or_alter, or_replace, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 3f703ffaf..ff2a61cf3 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -483,6 +483,7 @@ impl Spanned for Statement { Statement::CreateSchema { .. } => Span::empty(), Statement::CreateDatabase { .. } => Span::empty(), Statement::CreateFunction { .. } => Span::empty(), + Statement::CreateDomain { .. } => Span::empty(), Statement::CreateTrigger { .. } => Span::empty(), Statement::DropTrigger { .. } => Span::empty(), Statement::CreateProcedure { .. } => Span::empty(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2011d31e3..fc6f44376 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4625,6 +4625,8 @@ impl<'a> Parser<'a> { self.parse_create_external_table(or_replace) } else if self.parse_keyword(Keyword::FUNCTION) { self.parse_create_function(or_alter, or_replace, temporary) + } else if self.parse_keyword(Keyword::DOMAIN) { + self.parse_create_domain() } else if self.parse_keyword(Keyword::TRIGGER) { self.parse_create_trigger(or_alter, or_replace, false) } else if self.parse_keywords(&[Keyword::CONSTRAINT, Keyword::TRIGGER]) { @@ -5974,6 +5976,35 @@ impl<'a> Parser<'a> { Ok(owner) } + /// Parses a [Statement::CreateDomain] statement. + fn parse_create_domain(&mut self) -> Result { + let name = self.parse_object_name(false)?; + self.expect_keyword_is(Keyword::AS)?; + let data_type = self.parse_data_type()?; + let collation = if self.parse_keyword(Keyword::COLLATE) { + Some(self.parse_identifier()?) + } else { + None + }; + let default = if self.parse_keyword(Keyword::DEFAULT) { + Some(self.parse_expr()?) + } else { + None + }; + let mut constraints = Vec::new(); + while let Some(constraint) = self.parse_optional_table_constraint()? { + constraints.push(constraint); + } + + Ok(Statement::CreateDomain(CreateDomain { + name, + data_type, + collation, + default, + constraints, + })) + } + /// ```sql /// CREATE POLICY name ON table_name [ AS { PERMISSIVE | RESTRICTIVE } ] /// [ FOR { ALL | SELECT | INSERT | UPDATE | DELETE } ] diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 008d0670a..859eca453 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5153,6 +5153,99 @@ fn test_escaped_string_literal() { } } +#[test] +fn parse_create_domain() { + let sql1 = "CREATE DOMAIN my_domain AS INTEGER CHECK (VALUE > 0)"; + let expected = Statement::CreateDomain(CreateDomain { + name: ObjectName::from(vec![Ident::new("my_domain")]), + data_type: DataType::Integer(None), + collation: None, + default: None, + constraints: vec![TableConstraint::Check { + name: None, + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("VALUE"))), + op: BinaryOperator::Gt, + right: Box::new(Expr::Value(test_utils::number("0").into())), + }), + }], + }); + + assert_eq!(pg().verified_stmt(sql1), expected); + + let sql2 = "CREATE DOMAIN my_domain AS INTEGER COLLATE \"en_US\" CHECK (VALUE > 0)"; + let expected = Statement::CreateDomain(CreateDomain { + name: ObjectName::from(vec![Ident::new("my_domain")]), + data_type: DataType::Integer(None), + collation: Some(Ident::with_quote('"', "en_US")), + default: None, + constraints: vec![TableConstraint::Check { + name: None, + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("VALUE"))), + op: BinaryOperator::Gt, + right: Box::new(Expr::Value(test_utils::number("0").into())), + }), + }], + }); + + assert_eq!(pg().verified_stmt(sql2), expected); + + let sql3 = "CREATE DOMAIN my_domain AS INTEGER DEFAULT 1 CHECK (VALUE > 0)"; + let expected = Statement::CreateDomain(CreateDomain { + name: ObjectName::from(vec![Ident::new("my_domain")]), + data_type: DataType::Integer(None), + collation: None, + default: Some(Expr::Value(test_utils::number("1").into())), + constraints: vec![TableConstraint::Check { + name: None, + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("VALUE"))), + op: BinaryOperator::Gt, + right: Box::new(Expr::Value(test_utils::number("0").into())), + }), + }], + }); + + assert_eq!(pg().verified_stmt(sql3), expected); + + let sql4 = "CREATE DOMAIN my_domain AS INTEGER COLLATE \"en_US\" DEFAULT 1 CHECK (VALUE > 0)"; + let expected = Statement::CreateDomain(CreateDomain { + name: ObjectName::from(vec![Ident::new("my_domain")]), + data_type: DataType::Integer(None), + collation: Some(Ident::with_quote('"', "en_US")), + default: Some(Expr::Value(test_utils::number("1").into())), + constraints: vec![TableConstraint::Check { + name: None, + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("VALUE"))), + op: BinaryOperator::Gt, + right: Box::new(Expr::Value(test_utils::number("0").into())), + }), + }], + }); + + assert_eq!(pg().verified_stmt(sql4), expected); + + let sql5 = "CREATE DOMAIN my_domain AS INTEGER CONSTRAINT my_constraint CHECK (VALUE > 0)"; + let expected = Statement::CreateDomain(CreateDomain { + name: ObjectName::from(vec![Ident::new("my_domain")]), + data_type: DataType::Integer(None), + collation: None, + default: None, + constraints: vec![TableConstraint::Check { + name: Some(Ident::new("my_constraint")), + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("VALUE"))), + op: BinaryOperator::Gt, + right: Box::new(Expr::Value(test_utils::number("0").into())), + }), + }], + }); + + assert_eq!(pg().verified_stmt(sql5), expected); +} + #[test] fn parse_create_simple_before_insert_trigger() { let sql = "CREATE TRIGGER check_insert BEFORE INSERT ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_insert"; From 6cd237ea43e5363469418475e61fa503eba2db7b Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Thu, 8 May 2025 19:40:03 -0400 Subject: [PATCH 121/286] Allow stored procedures to be defined without `BEGIN`/`END` (#1834) --- src/ast/mod.rs | 9 ++-- src/parser/mod.rs | 8 ++-- tests/sqlparser_mssql.rs | 93 +++++++++++++++++++++------------------- 3 files changed, 56 insertions(+), 54 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index a72cb8add..6b7ba12d9 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3826,7 +3826,7 @@ pub enum Statement { or_alter: bool, name: ObjectName, params: Option>, - body: Vec, + body: ConditionalStatements, }, /// ```sql /// CREATE MACRO @@ -4705,11 +4705,8 @@ impl fmt::Display for Statement { write!(f, " ({})", display_comma_separated(p))?; } } - write!( - f, - " AS BEGIN {body} END", - body = display_separated(body, "; ") - ) + + write!(f, " AS {body}") } Statement::CreateMacro { or_replace, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index fc6f44376..d18c7f694 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15505,14 +15505,14 @@ impl<'a> Parser<'a> { let name = self.parse_object_name(false)?; let params = self.parse_optional_procedure_parameters()?; self.expect_keyword_is(Keyword::AS)?; - self.expect_keyword_is(Keyword::BEGIN)?; - let statements = self.parse_statements()?; - self.expect_keyword_is(Keyword::END)?; + + let body = self.parse_conditional_statements(&[Keyword::END])?; + Ok(Statement::CreateProcedure { name, or_alter, params, - body: statements, + body, }) } diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 9ff55198f..1c0a00b16 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -100,48 +100,52 @@ fn parse_mssql_delimited_identifiers() { #[test] fn parse_create_procedure() { - let sql = "CREATE OR ALTER PROCEDURE test (@foo INT, @bar VARCHAR(256)) AS BEGIN SELECT 1 END"; + let sql = "CREATE OR ALTER PROCEDURE test (@foo INT, @bar VARCHAR(256)) AS BEGIN SELECT 1; END"; assert_eq!( ms().verified_stmt(sql), Statement::CreateProcedure { or_alter: true, - body: vec![Statement::Query(Box::new(Query { - with: None, - limit_clause: None, - fetch: None, - locks: vec![], - for_clause: None, - order_by: None, - settings: None, - format_clause: None, - pipe_operators: vec![], - body: Box::new(SetExpr::Select(Box::new(Select { - select_token: AttachedToken::empty(), - distinct: None, - top: None, - top_before_distinct: false, - projection: vec![SelectItem::UnnamedExpr(Expr::Value( - (number("1")).with_empty_span() - ))], - into: None, - from: vec![], - lateral_views: vec![], - prewhere: None, - selection: None, - group_by: GroupByExpr::Expressions(vec![], vec![]), - cluster_by: vec![], - distribute_by: vec![], - sort_by: vec![], - having: None, - named_window: vec![], - window_before_qualify: false, - qualify: None, - value_table_mode: None, - connect_by: None, - flavor: SelectFlavor::Standard, - }))) - }))], + body: ConditionalStatements::BeginEnd(BeginEndStatements { + begin_token: AttachedToken::empty(), + statements: vec![Statement::Query(Box::new(Query { + with: None, + limit_clause: None, + fetch: None, + locks: vec![], + for_clause: None, + order_by: None, + settings: None, + format_clause: None, + pipe_operators: vec![], + body: Box::new(SetExpr::Select(Box::new(Select { + select_token: AttachedToken::empty(), + distinct: None, + top: None, + top_before_distinct: false, + projection: vec![SelectItem::UnnamedExpr(Expr::Value( + (number("1")).with_empty_span() + ))], + into: None, + from: vec![], + lateral_views: vec![], + prewhere: None, + selection: None, + group_by: GroupByExpr::Expressions(vec![], vec![]), + cluster_by: vec![], + distribute_by: vec![], + sort_by: vec![], + having: None, + named_window: vec![], + window_before_qualify: false, + qualify: None, + value_table_mode: None, + connect_by: None, + flavor: SelectFlavor::Standard, + }))) + }))], + end_token: AttachedToken::empty(), + }), params: Some(vec![ ProcedureParam { name: Ident { @@ -174,19 +178,20 @@ fn parse_create_procedure() { #[test] fn parse_mssql_create_procedure() { - let _ = ms_and_generic().verified_stmt("CREATE OR ALTER PROCEDURE foo AS BEGIN SELECT 1 END"); - let _ = ms_and_generic().verified_stmt("CREATE PROCEDURE foo AS BEGIN SELECT 1 END"); + let _ = ms_and_generic().verified_stmt("CREATE OR ALTER PROCEDURE foo AS SELECT 1;"); + let _ = ms_and_generic().verified_stmt("CREATE OR ALTER PROCEDURE foo AS BEGIN SELECT 1; END"); + let _ = ms_and_generic().verified_stmt("CREATE PROCEDURE foo AS BEGIN SELECT 1; END"); let _ = ms().verified_stmt( - "CREATE PROCEDURE foo AS BEGIN SELECT [myColumn] FROM [myschema].[mytable] END", + "CREATE PROCEDURE foo AS BEGIN SELECT [myColumn] FROM [myschema].[mytable]; END", ); let _ = ms_and_generic().verified_stmt( - "CREATE PROCEDURE foo (@CustomerName NVARCHAR(50)) AS BEGIN SELECT * FROM DEV END", + "CREATE PROCEDURE foo (@CustomerName NVARCHAR(50)) AS BEGIN SELECT * FROM DEV; END", ); - let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN UPDATE bar SET col = 'test' END"); + let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN UPDATE bar SET col = 'test'; END"); // Test a statement with END in it - let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN SELECT [foo], CASE WHEN [foo] IS NULL THEN 'empty' ELSE 'notempty' END AS [foo] END"); + let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN SELECT [foo], CASE WHEN [foo] IS NULL THEN 'empty' ELSE 'notempty' END AS [foo]; END"); // Multiple statements - let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN UPDATE bar SET col = 'test'; SELECT [foo] FROM BAR WHERE [FOO] > 10 END"); + let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN UPDATE bar SET col = 'test'; SELECT [foo] FROM BAR WHERE [FOO] > 10; END"); } #[test] From 2182f7ea71242a7e9674932d559baef67dae522d Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Fri, 9 May 2025 01:48:23 +0200 Subject: [PATCH 122/286] Add support for the MATCH and REGEXP binary operators (#1840) --- src/ast/operator.rs | 7 +++++++ src/dialect/mod.rs | 2 ++ src/dialect/sqlite.rs | 27 ++++++++++++++++++++++++++- tests/sqlparser_sqlite.rs | 30 ++++++++++++++++++++++++++++++ 4 files changed, 65 insertions(+), 1 deletion(-) diff --git a/src/ast/operator.rs b/src/ast/operator.rs index 73fe9cf42..d0bb05e3c 100644 --- a/src/ast/operator.rs +++ b/src/ast/operator.rs @@ -139,6 +139,11 @@ pub enum BinaryOperator { DuckIntegerDivide, /// MySQL [`DIV`](https://dev.mysql.com/doc/refman/8.0/en/arithmetic-functions.html) integer division MyIntegerDivide, + /// MATCH operator, e.g. `a MATCH b` (SQLite-specific) + /// See + Match, + /// REGEXP operator, e.g. `a REGEXP b` (SQLite-specific) + Regexp, /// Support for custom operators (such as Postgres custom operators) Custom(String), /// Bitwise XOR, e.g. `a # b` (PostgreSQL-specific) @@ -350,6 +355,8 @@ impl fmt::Display for BinaryOperator { BinaryOperator::BitwiseXor => f.write_str("^"), BinaryOperator::DuckIntegerDivide => f.write_str("//"), BinaryOperator::MyIntegerDivide => f.write_str("DIV"), + BinaryOperator::Match => f.write_str("MATCH"), + BinaryOperator::Regexp => f.write_str("REGEXP"), BinaryOperator::Custom(s) => f.write_str(s), BinaryOperator::PGBitwiseXor => f.write_str("#"), BinaryOperator::PGBitwiseShiftLeft => f.write_str("<<"), diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index b754a04f1..6fbbc7a23 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -619,6 +619,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::ILIKE => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::RLIKE => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::REGEXP => Ok(p!(Like)), + Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), _ => Ok(self.prec_unknown()), }, @@ -630,6 +631,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::ILIKE => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::RLIKE => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::REGEXP => Ok(p!(Like)), + Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::OPERATOR => Ok(p!(Between)), Token::Word(w) if w.keyword == Keyword::DIV => Ok(p!(MulDivModOp)), diff --git a/src/dialect/sqlite.rs b/src/dialect/sqlite.rs index 138c4692c..847e0d135 100644 --- a/src/dialect/sqlite.rs +++ b/src/dialect/sqlite.rs @@ -15,7 +15,11 @@ // specific language governing permissions and limitations // under the License. -use crate::ast::Statement; +#[cfg(not(feature = "std"))] +use alloc::boxed::Box; + +use crate::ast::BinaryOperator; +use crate::ast::{Expr, Statement}; use crate::dialect::Dialect; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; @@ -70,6 +74,27 @@ impl Dialect for SQLiteDialect { } } + fn parse_infix( + &self, + parser: &mut crate::parser::Parser, + expr: &crate::ast::Expr, + _precedence: u8, + ) -> Option> { + // Parse MATCH and REGEXP as operators + // See + for (keyword, op) in [ + (Keyword::REGEXP, BinaryOperator::Regexp), + (Keyword::MATCH, BinaryOperator::Match), + ] { + if parser.parse_keyword(keyword) { + let left = Box::new(expr.clone()); + let right = Box::new(parser.parse_expr().unwrap()); + return Some(Ok(Expr::BinaryOp { left, op, right })); + } + } + None + } + fn supports_in_empty_list(&self) -> bool { true } diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 361c9b051..b759065f3 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -562,6 +562,36 @@ fn test_dollar_identifier_as_placeholder() { } } +#[test] +fn test_match_operator() { + assert_eq!( + sqlite().verified_expr("col MATCH 'pattern'"), + Expr::BinaryOp { + op: BinaryOperator::Match, + left: Box::new(Expr::Identifier(Ident::new("col"))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("pattern".to_string())).with_empty_span() + )) + } + ); + sqlite().verified_only_select("SELECT * FROM email WHERE email MATCH 'fts5'"); +} + +#[test] +fn test_regexp_operator() { + assert_eq!( + sqlite().verified_expr("col REGEXP 'pattern'"), + Expr::BinaryOp { + op: BinaryOperator::Regexp, + left: Box::new(Expr::Identifier(Ident::new("col"))), + right: Box::new(Expr::Value( + (Value::SingleQuotedString("pattern".to_string())).with_empty_span() + )) + } + ); + sqlite().verified_only_select(r#"SELECT count(*) FROM messages WHERE msg_text REGEXP '\d+'"#); +} + fn sqlite() -> TestedDialects { TestedDialects::new(vec![Box::new(SQLiteDialect {})]) } From 052ad4a75981f1d411f8440e618217c7f010bbb9 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Sat, 10 May 2025 01:14:25 +0100 Subject: [PATCH 123/286] Fix: parsing ident starting with underscore in certain dialects (#1835) --- src/tokenizer.rs | 52 ++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 52 insertions(+) diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 4fad54624..afe1e35c7 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -1191,6 +1191,22 @@ impl<'a> Tokenizer<'a> { } // numbers and period '0'..='9' | '.' => { + // special case where if ._ is encountered after a word then that word + // is a table and the _ is the start of the col name. + // if the prev token is not a word, then this is not a valid sql + // word or number. + if ch == '.' && chars.peekable.clone().nth(1) == Some('_') { + if let Some(Token::Word(_)) = prev_token { + chars.next(); + return Ok(Some(Token::Period)); + } + + return self.tokenizer_error( + chars.location(), + "Unexpected character '_'".to_string(), + ); + } + // Some dialects support underscore as number separator // There can only be one at a time and it must be followed by another digit let is_number_separator = |ch: char, next_char: Option| { @@ -4018,4 +4034,40 @@ mod tests { ], ); } + + #[test] + fn tokenize_period_underscore() { + let sql = String::from("SELECT table._col"); + // a dialect that supports underscores in numeric literals + let dialect = PostgreSqlDialect {}; + let tokens = Tokenizer::new(&dialect, &sql).tokenize().unwrap(); + + let expected = vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Word(Word { + value: "table".to_string(), + quote_style: None, + keyword: Keyword::TABLE, + }), + Token::Period, + Token::Word(Word { + value: "_col".to_string(), + quote_style: None, + keyword: Keyword::NoKeyword, + }), + ]; + + compare(expected, tokens); + + let sql = String::from("SELECT ._123"); + if let Ok(tokens) = Tokenizer::new(&dialect, &sql).tokenize() { + panic!("Tokenizer should have failed on {sql}, but it succeeded with {tokens:?}"); + } + + let sql = String::from("SELECT ._abc"); + if let Ok(tokens) = Tokenizer::new(&dialect, &sql).tokenize() { + panic!("Tokenizer should have failed on {sql}, but it succeeded with {tokens:?}"); + } + } } From 6120bb59cc9cab0403366d644e914945155031a2 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Tue, 13 May 2025 15:25:07 +0200 Subject: [PATCH 124/286] implement pretty-printing with `{:#}` (#1847) --- src/ast/mod.rs | 102 ++++++++++--- src/ast/query.rs | 337 +++++++++++++++++++++++++----------------- src/display_utils.rs | 133 +++++++++++++++++ src/lib.rs | 22 +++ tests/pretty_print.rs | 157 ++++++++++++++++++++ 5 files changed, 591 insertions(+), 160 deletions(-) create mode 100644 src/display_utils.rs create mode 100644 tests/pretty_print.rs diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 6b7ba12d9..79b0e0d52 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -40,8 +40,14 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use crate::keywords::Keyword; -use crate::tokenizer::{Span, Token}; +use crate::{ + display_utils::SpaceOrNewline, + tokenizer::{Span, Token}, +}; +use crate::{ + display_utils::{Indent, NewLine}, + keywords::Keyword, +}; pub use self::data_type::{ ArrayElemTypeDef, BinaryLength, CharLengthUnits, CharacterLength, DataType, EnumMember, @@ -134,9 +140,9 @@ where fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { let mut delim = ""; for t in self.slice { - write!(f, "{delim}")?; + f.write_str(delim)?; delim = self.sep; - write!(f, "{t}")?; + t.fmt(f)?; } Ok(()) } @@ -628,7 +634,12 @@ pub struct CaseWhen { impl fmt::Display for CaseWhen { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "WHEN {} THEN {}", self.condition, self.result) + f.write_str("WHEN ")?; + self.condition.fmt(f)?; + f.write_str(" THEN")?; + SpaceOrNewline.fmt(f)?; + Indent(&self.result).fmt(f)?; + Ok(()) } } @@ -1662,23 +1673,29 @@ impl fmt::Display for Expr { write!(f, "{data_type}")?; write!(f, " {value}") } - Expr::Function(fun) => write!(f, "{fun}"), + Expr::Function(fun) => fun.fmt(f), Expr::Case { operand, conditions, else_result, } => { - write!(f, "CASE")?; + f.write_str("CASE")?; if let Some(operand) = operand { - write!(f, " {operand}")?; + f.write_str(" ")?; + operand.fmt(f)?; } for when in conditions { - write!(f, " {when}")?; + SpaceOrNewline.fmt(f)?; + Indent(when).fmt(f)?; } if let Some(else_result) = else_result { - write!(f, " ELSE {else_result}")?; + SpaceOrNewline.fmt(f)?; + Indent("ELSE").fmt(f)?; + SpaceOrNewline.fmt(f)?; + Indent(Indent(else_result)).fmt(f)?; } - write!(f, " END") + SpaceOrNewline.fmt(f)?; + f.write_str("END") } Expr::Exists { subquery, negated } => write!( f, @@ -1867,8 +1884,14 @@ pub enum WindowType { impl Display for WindowType { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { - WindowType::WindowSpec(spec) => write!(f, "({})", spec), - WindowType::NamedWindow(name) => write!(f, "{}", name), + WindowType::WindowSpec(spec) => { + f.write_str("(")?; + NewLine.fmt(f)?; + Indent(spec).fmt(f)?; + NewLine.fmt(f)?; + f.write_str(")") + } + WindowType::NamedWindow(name) => name.fmt(f), } } } @@ -1896,14 +1919,19 @@ pub struct WindowSpec { impl fmt::Display for WindowSpec { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - let mut delim = ""; + let mut is_first = true; if let Some(window_name) = &self.window_name { - delim = " "; + if !is_first { + SpaceOrNewline.fmt(f)?; + } + is_first = false; write!(f, "{window_name}")?; } if !self.partition_by.is_empty() { - f.write_str(delim)?; - delim = " "; + if !is_first { + SpaceOrNewline.fmt(f)?; + } + is_first = false; write!( f, "PARTITION BY {}", @@ -1911,12 +1939,16 @@ impl fmt::Display for WindowSpec { )?; } if !self.order_by.is_empty() { - f.write_str(delim)?; - delim = " "; + if !is_first { + SpaceOrNewline.fmt(f)?; + } + is_first = false; write!(f, "ORDER BY {}", display_comma_separated(&self.order_by))?; } if let Some(window_frame) = &self.window_frame { - f.write_str(delim)?; + if !is_first { + SpaceOrNewline.fmt(f)?; + } if let Some(end_bound) = &window_frame.end_bound { write!( f, @@ -4204,6 +4236,28 @@ impl fmt::Display for RaisErrorOption { } impl fmt::Display for Statement { + /// Formats a SQL statement with support for pretty printing. + /// + /// When using the alternate flag (`{:#}`), the statement will be formatted with proper + /// indentation and line breaks. For example: + /// + /// ``` + /// # use sqlparser::dialect::GenericDialect; + /// # use sqlparser::parser::Parser; + /// let sql = "SELECT a, b FROM table_1"; + /// let ast = Parser::parse_sql(&GenericDialect, sql).unwrap(); + /// + /// // Regular formatting + /// assert_eq!(format!("{}", ast[0]), "SELECT a, b FROM table_1"); + /// + /// // Pretty printing + /// assert_eq!(format!("{:#}", ast[0]), + /// r#"SELECT + /// a, + /// b + /// FROM + /// table_1"#); + /// ``` // Clippy thinks this function is too complicated, but it is painful to // split up without extracting structs for each `Statement` variant. #[allow(clippy::cognitive_complexity)] @@ -4219,7 +4273,8 @@ impl fmt::Display for Statement { } => { write!(f, "FLUSH")?; if let Some(location) = location { - write!(f, " {location}")?; + f.write_str(" ")?; + location.fmt(f)?; } write!(f, " {object_type}")?; @@ -4301,7 +4356,7 @@ impl fmt::Display for Statement { write!(f, "{statement}") } - Statement::Query(s) => write!(f, "{s}"), + Statement::Query(s) => s.fmt(f), Statement::Declare { stmts } => { write!(f, "DECLARE ")?; write!(f, "{}", display_separated(stmts, "; ")) @@ -7056,7 +7111,8 @@ impl fmt::Display for Function { } if let Some(o) = &self.over { - write!(f, " OVER {o}")?; + f.write_str(" OVER ")?; + o.fmt(f)?; } if self.uses_odbc_syntax { diff --git a/src/ast/query.rs b/src/ast/query.rs index a90b61668..33168695f 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -27,6 +27,7 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::{ ast::*, + display_utils::{indented_list, SpaceOrNewline}, tokenizer::{Token, TokenWithSpan}, }; @@ -70,33 +71,41 @@ pub struct Query { impl fmt::Display for Query { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { if let Some(ref with) = self.with { - write!(f, "{with} ")?; + with.fmt(f)?; + SpaceOrNewline.fmt(f)?; } - write!(f, "{}", self.body)?; + self.body.fmt(f)?; if let Some(ref order_by) = self.order_by { - write!(f, " {order_by}")?; + f.write_str(" ")?; + order_by.fmt(f)?; } if let Some(ref limit_clause) = self.limit_clause { limit_clause.fmt(f)?; } if let Some(ref settings) = self.settings { - write!(f, " SETTINGS {}", display_comma_separated(settings))?; + f.write_str(" SETTINGS ")?; + display_comma_separated(settings).fmt(f)?; } if let Some(ref fetch) = self.fetch { - write!(f, " {fetch}")?; + f.write_str(" ")?; + fetch.fmt(f)?; } if !self.locks.is_empty() { - write!(f, " {}", display_separated(&self.locks, " "))?; + f.write_str(" ")?; + display_separated(&self.locks, " ").fmt(f)?; } if let Some(ref for_clause) = self.for_clause { - write!(f, " {}", for_clause)?; + f.write_str(" ")?; + for_clause.fmt(f)?; } if let Some(ref format) = self.format_clause { - write!(f, " {}", format)?; + f.write_str(" ")?; + format.fmt(f)?; } for pipe_operator in &self.pipe_operators { - write!(f, " |> {}", pipe_operator)?; + f.write_str(" |> ")?; + pipe_operator.fmt(f)?; } Ok(()) } @@ -169,29 +178,39 @@ impl SetExpr { impl fmt::Display for SetExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - SetExpr::Select(s) => write!(f, "{s}"), - SetExpr::Query(q) => write!(f, "({q})"), - SetExpr::Values(v) => write!(f, "{v}"), - SetExpr::Insert(v) => write!(f, "{v}"), - SetExpr::Update(v) => write!(f, "{v}"), - SetExpr::Delete(v) => write!(f, "{v}"), - SetExpr::Table(t) => write!(f, "{t}"), + SetExpr::Select(s) => s.fmt(f), + SetExpr::Query(q) => { + f.write_str("(")?; + q.fmt(f)?; + f.write_str(")") + } + SetExpr::Values(v) => v.fmt(f), + SetExpr::Insert(v) => v.fmt(f), + SetExpr::Update(v) => v.fmt(f), + SetExpr::Delete(v) => v.fmt(f), + SetExpr::Table(t) => t.fmt(f), SetExpr::SetOperation { left, right, op, set_quantifier, } => { - write!(f, "{left} {op}")?; + left.fmt(f)?; + SpaceOrNewline.fmt(f)?; + op.fmt(f)?; match set_quantifier { SetQuantifier::All | SetQuantifier::Distinct | SetQuantifier::ByName | SetQuantifier::AllByName - | SetQuantifier::DistinctByName => write!(f, " {set_quantifier}")?, - SetQuantifier::None => write!(f, "{set_quantifier}")?, + | SetQuantifier::DistinctByName => { + f.write_str(" ")?; + set_quantifier.fmt(f)?; + } + SetQuantifier::None => {} } - write!(f, " {right}")?; + SpaceOrNewline.fmt(f)?; + right.fmt(f)?; Ok(()) } } @@ -242,7 +261,7 @@ impl fmt::Display for SetQuantifier { SetQuantifier::ByName => write!(f, "BY NAME"), SetQuantifier::AllByName => write!(f, "ALL BY NAME"), SetQuantifier::DistinctByName => write!(f, "DISTINCT BY NAME"), - SetQuantifier::None => write!(f, ""), + SetQuantifier::None => Ok(()), } } } @@ -357,90 +376,122 @@ impl fmt::Display for Select { } if let Some(value_table_mode) = self.value_table_mode { - write!(f, " {value_table_mode}")?; + f.write_str(" ")?; + value_table_mode.fmt(f)?; } if let Some(ref top) = self.top { if self.top_before_distinct { - write!(f, " {top}")?; + f.write_str(" ")?; + top.fmt(f)?; } } if let Some(ref distinct) = self.distinct { - write!(f, " {distinct}")?; + f.write_str(" ")?; + distinct.fmt(f)?; } if let Some(ref top) = self.top { if !self.top_before_distinct { - write!(f, " {top}")?; + f.write_str(" ")?; + top.fmt(f)?; } } if !self.projection.is_empty() { - write!(f, " {}", display_comma_separated(&self.projection))?; + indented_list(f, &self.projection)?; } if let Some(ref into) = self.into { - write!(f, " {into}")?; + f.write_str(" ")?; + into.fmt(f)?; } if self.flavor == SelectFlavor::Standard && !self.from.is_empty() { - write!(f, " FROM {}", display_comma_separated(&self.from))?; + SpaceOrNewline.fmt(f)?; + f.write_str("FROM")?; + indented_list(f, &self.from)?; } if !self.lateral_views.is_empty() { for lv in &self.lateral_views { - write!(f, "{lv}")?; + lv.fmt(f)?; } } if let Some(ref prewhere) = self.prewhere { - write!(f, " PREWHERE {prewhere}")?; + f.write_str(" PREWHERE ")?; + prewhere.fmt(f)?; } if let Some(ref selection) = self.selection { - write!(f, " WHERE {selection}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("WHERE")?; + SpaceOrNewline.fmt(f)?; + Indent(selection).fmt(f)?; } match &self.group_by { - GroupByExpr::All(_) => write!(f, " {}", self.group_by)?, + GroupByExpr::All(_) => { + SpaceOrNewline.fmt(f)?; + self.group_by.fmt(f)?; + } GroupByExpr::Expressions(exprs, _) => { if !exprs.is_empty() { - write!(f, " {}", self.group_by)? + SpaceOrNewline.fmt(f)?; + self.group_by.fmt(f)?; } } } if !self.cluster_by.is_empty() { - write!( - f, - " CLUSTER BY {}", - display_comma_separated(&self.cluster_by) - )?; + SpaceOrNewline.fmt(f)?; + f.write_str("CLUSTER BY")?; + SpaceOrNewline.fmt(f)?; + Indent(display_comma_separated(&self.cluster_by)).fmt(f)?; } if !self.distribute_by.is_empty() { - write!( - f, - " DISTRIBUTE BY {}", - display_comma_separated(&self.distribute_by) - )?; + SpaceOrNewline.fmt(f)?; + f.write_str("DISTRIBUTE BY")?; + SpaceOrNewline.fmt(f)?; + display_comma_separated(&self.distribute_by).fmt(f)?; } if !self.sort_by.is_empty() { - write!(f, " SORT BY {}", display_comma_separated(&self.sort_by))?; + SpaceOrNewline.fmt(f)?; + f.write_str("SORT BY")?; + SpaceOrNewline.fmt(f)?; + Indent(display_comma_separated(&self.sort_by)).fmt(f)?; } if let Some(ref having) = self.having { - write!(f, " HAVING {having}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("HAVING")?; + SpaceOrNewline.fmt(f)?; + Indent(having).fmt(f)?; } if self.window_before_qualify { if !self.named_window.is_empty() { - write!(f, " WINDOW {}", display_comma_separated(&self.named_window))?; + SpaceOrNewline.fmt(f)?; + f.write_str("WINDOW")?; + SpaceOrNewline.fmt(f)?; + display_comma_separated(&self.named_window).fmt(f)?; } if let Some(ref qualify) = self.qualify { - write!(f, " QUALIFY {qualify}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("QUALIFY")?; + SpaceOrNewline.fmt(f)?; + qualify.fmt(f)?; } } else { if let Some(ref qualify) = self.qualify { - write!(f, " QUALIFY {qualify}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("QUALIFY")?; + SpaceOrNewline.fmt(f)?; + qualify.fmt(f)?; } if !self.named_window.is_empty() { - write!(f, " WINDOW {}", display_comma_separated(&self.named_window))?; + SpaceOrNewline.fmt(f)?; + f.write_str("WINDOW")?; + SpaceOrNewline.fmt(f)?; + display_comma_separated(&self.named_window).fmt(f)?; } } if let Some(ref connect_by) = self.connect_by { - write!(f, " {connect_by}")?; + SpaceOrNewline.fmt(f)?; + connect_by.fmt(f)?; } Ok(()) } @@ -546,12 +597,12 @@ pub struct With { impl fmt::Display for With { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!( - f, - "WITH {}{}", - if self.recursive { "RECURSIVE " } else { "" }, - display_comma_separated(&self.cte_tables) - ) + f.write_str("WITH ")?; + if self.recursive { + f.write_str("RECURSIVE ")?; + } + display_comma_separated(&self.cte_tables).fmt(f)?; + Ok(()) } } @@ -598,8 +649,24 @@ pub struct Cte { impl fmt::Display for Cte { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self.materialized.as_ref() { - None => write!(f, "{} AS ({})", self.alias, self.query)?, - Some(materialized) => write!(f, "{} AS {materialized} ({})", self.alias, self.query)?, + None => { + self.alias.fmt(f)?; + f.write_str(" AS (")?; + NewLine.fmt(f)?; + Indent(&self.query).fmt(f)?; + NewLine.fmt(f)?; + f.write_str(")")?; + } + Some(materialized) => { + self.alias.fmt(f)?; + f.write_str(" AS ")?; + materialized.fmt(f)?; + f.write_str(" (")?; + NewLine.fmt(f)?; + Indent(&self.query).fmt(f)?; + NewLine.fmt(f)?; + f.write_str(")")?; + } }; if let Some(ref fr) = self.from { write!(f, " FROM {fr}")?; @@ -912,18 +979,21 @@ impl fmt::Display for ReplaceSelectElement { impl fmt::Display for SelectItem { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + use core::fmt::Write; match &self { - SelectItem::UnnamedExpr(expr) => write!(f, "{expr}"), - SelectItem::ExprWithAlias { expr, alias } => write!(f, "{expr} AS {alias}"), + SelectItem::UnnamedExpr(expr) => expr.fmt(f), + SelectItem::ExprWithAlias { expr, alias } => { + expr.fmt(f)?; + f.write_str(" AS ")?; + alias.fmt(f) + } SelectItem::QualifiedWildcard(kind, additional_options) => { - write!(f, "{kind}")?; - write!(f, "{additional_options}")?; - Ok(()) + kind.fmt(f)?; + additional_options.fmt(f) } SelectItem::Wildcard(additional_options) => { - write!(f, "*")?; - write!(f, "{additional_options}")?; - Ok(()) + f.write_char('*')?; + additional_options.fmt(f) } } } @@ -939,9 +1009,10 @@ pub struct TableWithJoins { impl fmt::Display for TableWithJoins { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{}", self.relation)?; + self.relation.fmt(f)?; for join in &self.joins { - write!(f, "{join}")?; + SpaceOrNewline.fmt(f)?; + join.fmt(f)?; } Ok(()) } @@ -1334,7 +1405,6 @@ pub enum TableFactor { #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] - pub enum TableSampleKind { /// Table sample located before the table alias option BeforeTableAlias(Box), @@ -1769,9 +1839,9 @@ impl fmt::Display for TableFactor { sample, index_hints, } => { - write!(f, "{name}")?; + name.fmt(f)?; if let Some(json_path) = json_path { - write!(f, "{json_path}")?; + json_path.fmt(f)?; } if !partitions.is_empty() { write!(f, "PARTITION ({})", display_comma_separated(partitions))?; @@ -1818,7 +1888,11 @@ impl fmt::Display for TableFactor { if *lateral { write!(f, "LATERAL ")?; } - write!(f, "({subquery})")?; + f.write_str("(")?; + NewLine.fmt(f)?; + Indent(subquery).fmt(f)?; + NewLine.fmt(f)?; + f.write_str(")")?; if let Some(alias) = alias { write!(f, " AS {alias}")?; } @@ -2132,116 +2206,104 @@ impl fmt::Display for Join { Suffix(constraint) } if self.global { - write!(f, " GLOBAL")?; + write!(f, "GLOBAL ")?; } match &self.join_operator { - JoinOperator::Join(constraint) => write!( - f, - " {}JOIN {}{}", + JoinOperator::Join(constraint) => f.write_fmt(format_args!( + "{}JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::Inner(constraint) => write!( - f, - " {}INNER JOIN {}{}", + )), + JoinOperator::Inner(constraint) => f.write_fmt(format_args!( + "{}INNER JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::Left(constraint) => write!( - f, - " {}LEFT JOIN {}{}", + )), + JoinOperator::Left(constraint) => f.write_fmt(format_args!( + "{}LEFT JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::LeftOuter(constraint) => write!( - f, - " {}LEFT OUTER JOIN {}{}", + )), + JoinOperator::LeftOuter(constraint) => f.write_fmt(format_args!( + "{}LEFT OUTER JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::Right(constraint) => write!( - f, - " {}RIGHT JOIN {}{}", + )), + JoinOperator::Right(constraint) => f.write_fmt(format_args!( + "{}RIGHT JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::RightOuter(constraint) => write!( - f, - " {}RIGHT OUTER JOIN {}{}", + )), + JoinOperator::RightOuter(constraint) => f.write_fmt(format_args!( + "{}RIGHT OUTER JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::FullOuter(constraint) => write!( - f, - " {}FULL JOIN {}{}", + )), + JoinOperator::FullOuter(constraint) => f.write_fmt(format_args!( + "{}FULL JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::CrossJoin => write!(f, " CROSS JOIN {}", self.relation), - JoinOperator::Semi(constraint) => write!( - f, - " {}SEMI JOIN {}{}", + )), + JoinOperator::CrossJoin => f.write_fmt(format_args!("CROSS JOIN {}", self.relation)), + JoinOperator::Semi(constraint) => f.write_fmt(format_args!( + "{}SEMI JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::LeftSemi(constraint) => write!( - f, - " {}LEFT SEMI JOIN {}{}", + )), + JoinOperator::LeftSemi(constraint) => f.write_fmt(format_args!( + "{}LEFT SEMI JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::RightSemi(constraint) => write!( - f, - " {}RIGHT SEMI JOIN {}{}", + )), + JoinOperator::RightSemi(constraint) => f.write_fmt(format_args!( + "{}RIGHT SEMI JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::Anti(constraint) => write!( - f, - " {}ANTI JOIN {}{}", + )), + JoinOperator::Anti(constraint) => f.write_fmt(format_args!( + "{}ANTI JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::LeftAnti(constraint) => write!( - f, - " {}LEFT ANTI JOIN {}{}", + )), + JoinOperator::LeftAnti(constraint) => f.write_fmt(format_args!( + "{}LEFT ANTI JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::RightAnti(constraint) => write!( - f, - " {}RIGHT ANTI JOIN {}{}", + )), + JoinOperator::RightAnti(constraint) => f.write_fmt(format_args!( + "{}RIGHT ANTI JOIN {}{}", prefix(constraint), self.relation, suffix(constraint) - ), - JoinOperator::CrossApply => write!(f, " CROSS APPLY {}", self.relation), - JoinOperator::OuterApply => write!(f, " OUTER APPLY {}", self.relation), + )), + JoinOperator::CrossApply => f.write_fmt(format_args!("CROSS APPLY {}", self.relation)), + JoinOperator::OuterApply => f.write_fmt(format_args!("OUTER APPLY {}", self.relation)), JoinOperator::AsOf { match_condition, constraint, - } => write!( - f, - " ASOF JOIN {} MATCH_CONDITION ({match_condition}){}", + } => f.write_fmt(format_args!( + "ASOF JOIN {} MATCH_CONDITION ({match_condition}){}", self.relation, suffix(constraint) - ), - JoinOperator::StraightJoin(constraint) => { - write!(f, " STRAIGHT_JOIN {}{}", self.relation, suffix(constraint)) - } + )), + JoinOperator::StraightJoin(constraint) => f.write_fmt(format_args!( + "STRAIGHT_JOIN {}{}", + self.relation, + suffix(constraint) + )), } } } @@ -2914,8 +2976,9 @@ impl fmt::Display for GroupByExpr { Ok(()) } GroupByExpr::Expressions(col_names, modifiers) => { - let col_names = display_comma_separated(col_names); - write!(f, "GROUP BY {col_names}")?; + f.write_str("GROUP BY")?; + SpaceOrNewline.fmt(f)?; + Indent(display_comma_separated(col_names)).fmt(f)?; if !modifiers.is_empty() { write!(f, " {}", display_separated(modifiers, " "))?; } diff --git a/src/display_utils.rs b/src/display_utils.rs new file mode 100644 index 000000000..e7e1272fb --- /dev/null +++ b/src/display_utils.rs @@ -0,0 +1,133 @@ +//! Utilities for formatting SQL AST nodes with pretty printing support. +//! +//! The module provides formatters that implement the `Display` trait with support +//! for both regular (`{}`) and pretty (`{:#}`) formatting modes. Pretty printing +//! adds proper indentation and line breaks to make SQL statements more readable. + +use core::fmt::{self, Display, Write}; + +/// A wrapper around a value that adds an indent to the value when displayed with {:#}. +pub(crate) struct Indent(pub T); + +const INDENT: &str = " "; + +impl Display for Indent +where + T: Display, +{ + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + if f.alternate() { + f.write_str(INDENT)?; + write!(Indent(f), "{:#}", self.0) + } else { + self.0.fmt(f) + } + } +} + +/// Adds an indent to the inner writer +impl Write for Indent +where + T: Write, +{ + fn write_str(&mut self, s: &str) -> fmt::Result { + let mut first = true; + for line in s.split('\n') { + if !first { + write!(self.0, "\n{INDENT}")?; + } + self.0.write_str(line)?; + first = false; + } + Ok(()) + } +} + +/// A value that inserts a newline when displayed with {:#}, but not when displayed with {}. +pub(crate) struct NewLine; + +impl Display for NewLine { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + if f.alternate() { + f.write_char('\n') + } else { + Ok(()) + } + } +} + +/// A value that inserts a space when displayed with {}, but a newline when displayed with {:#}. +pub(crate) struct SpaceOrNewline; + +impl Display for SpaceOrNewline { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + if f.alternate() { + f.write_char('\n') + } else { + f.write_char(' ') + } + } +} + +/// A value that displays a comma-separated list of values. +/// When pretty-printed (using {:#}), it displays each value on a new line. +pub struct DisplayCommaSeparated<'a, T: fmt::Display>(&'a [T]); + +impl fmt::Display for DisplayCommaSeparated<'_, T> { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let mut first = true; + for t in self.0 { + if !first { + f.write_char(',')?; + SpaceOrNewline.fmt(f)?; + } + first = false; + t.fmt(f)?; + } + Ok(()) + } +} + +/// Displays a whitespace, followed by a comma-separated list that is indented when pretty-printed. +pub(crate) fn indented_list(f: &mut fmt::Formatter, slice: &[T]) -> fmt::Result { + SpaceOrNewline.fmt(f)?; + Indent(DisplayCommaSeparated(slice)).fmt(f) +} + +#[cfg(test)] +mod tests { + use super::*; + + struct DisplayCharByChar(T); + + impl Display for DisplayCharByChar { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + for c in self.0.to_string().chars() { + write!(f, "{}", c)?; + } + Ok(()) + } + } + + #[test] + fn test_indent() { + let original = "line 1\nline 2"; + let indent = Indent(original); + assert_eq!( + indent.to_string(), + original, + "Only the alternate form should be indented" + ); + let expected = " line 1\n line 2"; + assert_eq!(format!("{:#}", indent), expected); + let display_char_by_char = DisplayCharByChar(original); + assert_eq!(format!("{:#}", Indent(display_char_by_char)), expected); + } + + #[test] + fn test_space_or_newline() { + let space_or_newline = SpaceOrNewline; + assert_eq!(format!("{}", space_or_newline), " "); + assert_eq!(format!("{:#}", space_or_newline), "\n"); + } +} diff --git a/src/lib.rs b/src/lib.rs index 5d72f9f0e..c81ab5002 100644 --- a/src/lib.rs +++ b/src/lib.rs @@ -64,6 +64,27 @@ //! // The original SQL text can be generated from the AST //! assert_eq!(ast[0].to_string(), sql); //! ``` +//! +//! # Pretty Printing +//! +//! SQL statements can be pretty-printed with proper indentation and line breaks using the alternate flag (`{:#}`): +//! +//! ``` +//! # use sqlparser::dialect::GenericDialect; +//! # use sqlparser::parser::Parser; +//! let sql = "SELECT a, b FROM table_1"; +//! let ast = Parser::parse_sql(&GenericDialect, sql).unwrap(); +//! +//! // Pretty print with indentation and line breaks +//! let pretty_sql = format!("{:#}", ast[0]); +//! assert_eq!(pretty_sql, r#" +//! SELECT +//! a, +//! b +//! FROM +//! table_1 +//! "#.trim()); +//! ``` //! [sqlparser crates.io page]: https://crates.io/crates/sqlparser //! [`Parser::parse_sql`]: crate::parser::Parser::parse_sql //! [`Parser::new`]: crate::parser::Parser::new @@ -142,6 +163,7 @@ extern crate pretty_assertions; pub mod ast; #[macro_use] pub mod dialect; +mod display_utils; pub mod keywords; pub mod parser; pub mod tokenizer; diff --git a/tests/pretty_print.rs b/tests/pretty_print.rs new file mode 100644 index 000000000..1eb8ca41c --- /dev/null +++ b/tests/pretty_print.rs @@ -0,0 +1,157 @@ +use sqlparser::dialect::GenericDialect; +use sqlparser::parser::Parser; + +fn prettify(sql: &str) -> String { + let ast = Parser::parse_sql(&GenericDialect {}, sql).unwrap(); + format!("{:#}", ast[0]) +} + +#[test] +fn test_pretty_print_select() { + assert_eq!( + prettify("SELECT a, b, c FROM my_table WHERE x = 1 AND y = 2"), + r#" +SELECT + a, + b, + c +FROM + my_table +WHERE + x = 1 AND y = 2 +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_join() { + assert_eq!( + prettify("SELECT a FROM table1 JOIN table2 ON table1.id = table2.id"), + r#" +SELECT + a +FROM + table1 + JOIN table2 ON table1.id = table2.id +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_subquery() { + assert_eq!( + prettify("SELECT * FROM (SELECT a, b FROM my_table) AS subquery"), + r#" +SELECT + * +FROM + ( + SELECT + a, + b + FROM + my_table + ) AS subquery +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_union() { + assert_eq!( + prettify("SELECT a FROM table1 UNION SELECT b FROM table2"), + r#" +SELECT + a +FROM + table1 +UNION +SELECT + b +FROM + table2 +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_group_by() { + assert_eq!( + prettify("SELECT a, COUNT(*) FROM my_table GROUP BY a HAVING COUNT(*) > 1"), + r#" +SELECT + a, + COUNT(*) +FROM + my_table +GROUP BY + a +HAVING + COUNT(*) > 1 +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_cte() { + assert_eq!( + prettify("WITH cte AS (SELECT a, b FROM my_table) SELECT * FROM cte"), + r#" +WITH cte AS ( + SELECT + a, + b + FROM + my_table +) +SELECT + * +FROM + cte +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_case_when() { + assert_eq!( + prettify("SELECT CASE WHEN x > 0 THEN 'positive' WHEN x < 0 THEN 'negative' ELSE 'zero' END FROM my_table"), + r#" +SELECT + CASE + WHEN x > 0 THEN + 'positive' + WHEN x < 0 THEN + 'negative' + ELSE + 'zero' + END +FROM + my_table +"#.trim() + ); +} + +#[test] +fn test_pretty_print_window_function() { + assert_eq!( + prettify("SELECT id, value, ROW_NUMBER() OVER (PARTITION BY category ORDER BY value DESC) as rank FROM my_table"), + r#" +SELECT + id, + value, + ROW_NUMBER() OVER ( + PARTITION BY category + ORDER BY value DESC + ) AS rank +FROM + my_table +"#.trim() + ); +} From 178a351812e34f8ca9cf0e6662f9e717addebebb Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Tue, 13 May 2025 17:36:27 +0200 Subject: [PATCH 125/286] Fix big performance issue in string serialization (#1848) --- src/ast/value.rs | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/src/ast/value.rs b/src/ast/value.rs index 77e2e0e81..98616407c 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -456,30 +456,38 @@ impl fmt::Display for EscapeQuotedString<'_> { // | `"A\"B\"A"` | default | `DoubleQuotedString(String::from("A\"B\"A"))` | `"A""B""A"` | let quote = self.quote; let mut previous_char = char::default(); - let mut peekable_chars = self.string.chars().peekable(); - while let Some(&ch) = peekable_chars.peek() { + let mut start_idx = 0; + let mut peekable_chars = self.string.char_indices().peekable(); + while let Some(&(idx, ch)) = peekable_chars.peek() { match ch { char if char == quote => { if previous_char == '\\' { - write!(f, "{char}")?; + // the quote is already escaped with a backslash, skip peekable_chars.next(); continue; } peekable_chars.next(); - if peekable_chars.peek().map(|c| *c == quote).unwrap_or(false) { - write!(f, "{char}{char}")?; - peekable_chars.next(); - } else { - write!(f, "{char}{char}")?; + match peekable_chars.peek() { + Some((_, c)) if *c == quote => { + // the quote is already escaped with another quote, skip + peekable_chars.next(); + } + _ => { + // The quote is not escaped. + // Including idx in the range, so the quote at idx will be printed twice: + // in this call to write_str() and in the next one. + f.write_str(&self.string[start_idx..=idx])?; + start_idx = idx; + } } } _ => { - write!(f, "{ch}")?; peekable_chars.next(); } } previous_char = ch; } + f.write_str(&self.string[start_idx..])?; Ok(()) } } From 74a95fdbd1467b6f7b0f6708d27d5d9e8598eea9 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Wed, 14 May 2025 03:21:23 -0400 Subject: [PATCH 126/286] Add support for `DENY` statements (#1836) --- src/ast/mod.rs | 50 +++++++++++++++++++++++ src/ast/spans.rs | 1 + src/dialect/mod.rs | 7 +++- src/dialect/mssql.rs | 13 +++++- src/keywords.rs | 1 + src/parser/mod.rs | 84 +++++++++++++++++++++++++++++++++------ tests/sqlparser_common.rs | 33 +++++++++++++++ tests/sqlparser_mssql.rs | 10 +++++ tests/sqlparser_mysql.rs | 1 + 9 files changed, 185 insertions(+), 15 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 79b0e0d52..3791154d1 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3903,9 +3903,14 @@ pub enum Statement { objects: Option, grantees: Vec, with_grant_option: bool, + as_grantor: Option, granted_by: Option, }, /// ```sql + /// DENY privileges ON object TO grantees + /// ``` + Deny(DenyStatement), + /// ```sql /// REVOKE privileges ON objects FROM grantees /// ``` Revoke { @@ -5580,6 +5585,7 @@ impl fmt::Display for Statement { objects, grantees, with_grant_option, + as_grantor, granted_by, } => { write!(f, "GRANT {privileges} ")?; @@ -5590,11 +5596,15 @@ impl fmt::Display for Statement { if *with_grant_option { write!(f, " WITH GRANT OPTION")?; } + if let Some(grantor) = as_grantor { + write!(f, " AS {grantor}")?; + } if let Some(grantor) = granted_by { write!(f, " GRANTED BY {grantor}")?; } Ok(()) } + Statement::Deny(s) => write!(f, "{s}"), Statement::Revoke { privileges, objects, @@ -6380,6 +6390,9 @@ pub enum Action { }, Delete, EvolveSchema, + Exec { + obj_type: Option, + }, Execute { obj_type: Option, }, @@ -6446,6 +6459,12 @@ impl fmt::Display for Action { Action::DatabaseRole { role } => write!(f, "DATABASE ROLE {role}")?, Action::Delete => f.write_str("DELETE")?, Action::EvolveSchema => f.write_str("EVOLVE SCHEMA")?, + Action::Exec { obj_type } => { + f.write_str("EXEC")?; + if let Some(obj_type) = obj_type { + write!(f, " {obj_type}")? + } + } Action::Execute { obj_type } => { f.write_str("EXECUTE")?; if let Some(obj_type) = obj_type { @@ -6867,6 +6886,37 @@ impl fmt::Display for GrantObjects { } } +/// A `DENY` statement +/// +/// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/deny-transact-sql) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct DenyStatement { + pub privileges: Privileges, + pub objects: GrantObjects, + pub grantees: Vec, + pub granted_by: Option, + pub cascade: Option, +} + +impl fmt::Display for DenyStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "DENY {}", self.privileges)?; + write!(f, " ON {}", self.objects)?; + if !self.grantees.is_empty() { + write!(f, " TO {}", display_comma_separated(&self.grantees))?; + } + if let Some(cascade) = &self.cascade { + write!(f, " {cascade}")?; + } + if let Some(granted_by) = &self.granted_by { + write!(f, " AS {granted_by}")?; + } + Ok(()) + } +} + /// SQL assignment `foo = expr` as used in SQLUpdate #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index ff2a61cf3..c09359480 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -491,6 +491,7 @@ impl Spanned for Statement { Statement::CreateStage { .. } => Span::empty(), Statement::Assert { .. } => Span::empty(), Statement::Grant { .. } => Span::empty(), + Statement::Deny { .. } => Span::empty(), Statement::Revoke { .. } => Span::empty(), Statement::Deallocate { .. } => Span::empty(), Statement::Execute { .. } => Span::empty(), diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 6fbbc7a23..a4c899e6b 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -49,7 +49,7 @@ pub use self::postgresql::PostgreSqlDialect; pub use self::redshift::RedshiftSqlDialect; pub use self::snowflake::SnowflakeDialect; pub use self::sqlite::SQLiteDialect; -use crate::ast::{ColumnOption, Expr, Statement}; +use crate::ast::{ColumnOption, Expr, GranteesType, Statement}; pub use crate::keywords; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; @@ -910,6 +910,11 @@ pub trait Dialect: Debug + Any { &[] } + /// Returns grantee types that should be treated as identifiers + fn get_reserved_grantees_types(&self) -> &[GranteesType] { + &[] + } + /// Returns true if this dialect supports the `TABLESAMPLE` option /// before the table alias option. For example: /// diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 647e82a2a..36bd222b8 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -17,8 +17,8 @@ use crate::ast::helpers::attached_token::AttachedToken; use crate::ast::{ - BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, IfStatement, Statement, - TriggerObject, + BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, GranteesType, + IfStatement, Statement, TriggerObject, }; use crate::dialect::Dialect; use crate::keywords::{self, Keyword}; @@ -52,6 +52,10 @@ impl Dialect for MsSqlDialect { || ch == '_' } + fn identifier_quote_style(&self, _identifier: &str) -> Option { + Some('[') + } + /// SQL Server has `CONVERT(type, value)` instead of `CONVERT(value, type)` /// fn convert_type_before_value(&self) -> bool { @@ -119,6 +123,11 @@ impl Dialect for MsSqlDialect { true } + /// See + fn get_reserved_grantees_types(&self) -> &[GranteesType] { + &[GranteesType::Public] + } + fn is_column_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { !keywords::RESERVED_FOR_COLUMN_ALIAS.contains(kw) && !RESERVED_FOR_COLUMN_ALIAS.contains(kw) } diff --git a/src/keywords.rs b/src/keywords.rs index ddb786650..aaa2e167f 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -278,6 +278,7 @@ define_keywords!( DELIMITER, DELTA, DENSE_RANK, + DENY, DEREF, DESC, DESCRIBE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d18c7f694..1e0227271 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -583,6 +583,10 @@ impl<'a> Parser<'a> { Keyword::SHOW => self.parse_show(), Keyword::USE => self.parse_use(), Keyword::GRANT => self.parse_grant(), + Keyword::DENY => { + self.prev_token(); + self.parse_deny() + } Keyword::REVOKE => self.parse_revoke(), Keyword::START => self.parse_start_transaction(), Keyword::BEGIN => self.parse_begin(), @@ -13381,7 +13385,7 @@ impl<'a> Parser<'a> { /// Parse a GRANT statement. pub fn parse_grant(&mut self) -> Result { - let (privileges, objects) = self.parse_grant_revoke_privileges_objects()?; + let (privileges, objects) = self.parse_grant_deny_revoke_privileges_objects()?; self.expect_keyword_is(Keyword::TO)?; let grantees = self.parse_grantees()?; @@ -13389,15 +13393,24 @@ impl<'a> Parser<'a> { let with_grant_option = self.parse_keywords(&[Keyword::WITH, Keyword::GRANT, Keyword::OPTION]); - let granted_by = self - .parse_keywords(&[Keyword::GRANTED, Keyword::BY]) - .then(|| self.parse_identifier().unwrap()); + let as_grantor = if self.parse_keywords(&[Keyword::AS]) { + Some(self.parse_identifier()?) + } else { + None + }; + + let granted_by = if self.parse_keywords(&[Keyword::GRANTED, Keyword::BY]) { + Some(self.parse_identifier()?) + } else { + None + }; Ok(Statement::Grant { privileges, objects, grantees, with_grant_option, + as_grantor, granted_by, }) } @@ -13406,7 +13419,7 @@ impl<'a> Parser<'a> { let mut values = vec![]; let mut grantee_type = GranteesType::None; loop { - grantee_type = if self.parse_keyword(Keyword::ROLE) { + let new_grantee_type = if self.parse_keyword(Keyword::ROLE) { GranteesType::Role } else if self.parse_keyword(Keyword::USER) { GranteesType::User @@ -13423,9 +13436,19 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::APPLICATION) { GranteesType::Application } else { - grantee_type // keep from previous iteraton, if not specified + grantee_type.clone() // keep from previous iteraton, if not specified }; + if self + .dialect + .get_reserved_grantees_types() + .contains(&new_grantee_type) + { + self.prev_token(); + } else { + grantee_type = new_grantee_type; + } + let grantee = if grantee_type == GranteesType::Public { Grantee { grantee_type: grantee_type.clone(), @@ -13460,7 +13483,7 @@ impl<'a> Parser<'a> { Ok(values) } - pub fn parse_grant_revoke_privileges_objects( + pub fn parse_grant_deny_revoke_privileges_objects( &mut self, ) -> Result<(Privileges, Option), ParserError> { let privileges = if self.parse_keyword(Keyword::ALL) { @@ -13510,7 +13533,6 @@ impl<'a> Parser<'a> { let object_type = self.parse_one_of_keywords(&[ Keyword::SEQUENCE, Keyword::DATABASE, - Keyword::DATABASE, Keyword::SCHEMA, Keyword::TABLE, Keyword::VIEW, @@ -13605,6 +13627,9 @@ impl<'a> Parser<'a> { Ok(Action::Create { obj_type }) } else if self.parse_keyword(Keyword::DELETE) { Ok(Action::Delete) + } else if self.parse_keyword(Keyword::EXEC) { + let obj_type = self.maybe_parse_action_execute_obj_type(); + Ok(Action::Exec { obj_type }) } else if self.parse_keyword(Keyword::EXECUTE) { let obj_type = self.maybe_parse_action_execute_obj_type(); Ok(Action::Execute { obj_type }) @@ -13803,16 +13828,51 @@ impl<'a> Parser<'a> { } } + /// Parse [`Statement::Deny`] + pub fn parse_deny(&mut self) -> Result { + self.expect_keyword(Keyword::DENY)?; + + let (privileges, objects) = self.parse_grant_deny_revoke_privileges_objects()?; + let objects = match objects { + Some(o) => o, + None => { + return parser_err!( + "DENY statements must specify an object", + self.peek_token().span.start + ) + } + }; + + self.expect_keyword_is(Keyword::TO)?; + let grantees = self.parse_grantees()?; + let cascade = self.parse_cascade_option(); + let granted_by = if self.parse_keywords(&[Keyword::AS]) { + Some(self.parse_identifier()?) + } else { + None + }; + + Ok(Statement::Deny(DenyStatement { + privileges, + objects, + grantees, + cascade, + granted_by, + })) + } + /// Parse a REVOKE statement pub fn parse_revoke(&mut self) -> Result { - let (privileges, objects) = self.parse_grant_revoke_privileges_objects()?; + let (privileges, objects) = self.parse_grant_deny_revoke_privileges_objects()?; self.expect_keyword_is(Keyword::FROM)?; let grantees = self.parse_grantees()?; - let granted_by = self - .parse_keywords(&[Keyword::GRANTED, Keyword::BY]) - .then(|| self.parse_identifier().unwrap()); + let granted_by = if self.parse_keywords(&[Keyword::GRANTED, Keyword::BY]) { + Some(self.parse_identifier()?) + } else { + None + }; let cascade = self.parse_cascade_option(); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 7a8b8bdaa..a07bcc68e 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9331,6 +9331,39 @@ fn parse_grant() { verified_stmt("GRANT USAGE ON WAREHOUSE wh1 TO ROLE role1"); verified_stmt("GRANT OWNERSHIP ON INTEGRATION int1 TO ROLE role1"); verified_stmt("GRANT SELECT ON VIEW view1 TO ROLE role1"); + verified_stmt("GRANT EXEC ON my_sp TO runner"); + verified_stmt("GRANT UPDATE ON my_table TO updater_role AS dbo"); + + all_dialects_where(|d| d.identifier_quote_style("none") == Some('[')) + .verified_stmt("GRANT SELECT ON [my_table] TO [public]"); +} + +#[test] +fn parse_deny() { + let sql = "DENY INSERT, DELETE ON users TO analyst CASCADE AS admin"; + match verified_stmt(sql) { + Statement::Deny(deny) => { + assert_eq!( + Privileges::Actions(vec![Action::Insert { columns: None }, Action::Delete]), + deny.privileges + ); + assert_eq!( + &GrantObjects::Tables(vec![ObjectName::from(vec![Ident::new("users")])]), + &deny.objects + ); + assert_eq_vec(&["analyst"], &deny.grantees); + assert_eq!(Some(CascadeOption::Cascade), deny.cascade); + assert_eq!(Some(Ident::from("admin")), deny.granted_by); + } + _ => unreachable!(), + } + + verified_stmt("DENY SELECT, INSERT, UPDATE, DELETE ON db1.sc1 TO role1, role2"); + verified_stmt("DENY ALL ON db1.sc1 TO role1"); + verified_stmt("DENY EXEC ON my_sp TO runner"); + + all_dialects_where(|d| d.identifier_quote_style("none") == Some('[')) + .verified_stmt("DENY SELECT ON [my_table] TO [public]"); } #[test] diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 1c0a00b16..7b3769ec8 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2340,3 +2340,13 @@ fn parse_print() { let _ = ms().verified_stmt("PRINT N'Hello, ⛄️!'"); let _ = ms().verified_stmt("PRINT @my_variable"); } + +#[test] +fn parse_mssql_grant() { + ms().verified_stmt("GRANT SELECT ON my_table TO public, db_admin"); +} + +#[test] +fn parse_mssql_deny() { + ms().verified_stmt("DENY SELECT ON my_table TO public, db_admin"); +} diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 27c60b052..dd279894a 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3538,6 +3538,7 @@ fn parse_grant() { objects, grantees, with_grant_option, + as_grantor: _, granted_by, } = stmt { From c6e897dc12428891fec1a1ec5421c5e43e7080b2 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Wed, 14 May 2025 08:40:44 +0100 Subject: [PATCH 127/286] Postgresql: Add `REPLICA IDENTITY` operation for `ALTER TABLE` (#1844) --- src/ast/ddl.rs | 33 +++++++++++++++++++++++++++++++++ src/ast/mod.rs | 2 +- src/ast/spans.rs | 1 + src/parser/mod.rs | 19 ++++++++++++++++++- tests/sqlparser_postgres.rs | 27 +++++++++++++++++++++++++++ 5 files changed, 80 insertions(+), 2 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index a457a0655..270897130 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -39,6 +39,29 @@ use crate::ast::{ use crate::keywords::Keyword; use crate::tokenizer::Token; +/// ALTER TABLE operation REPLICA IDENTITY values +/// See [Postgres ALTER TABLE docs](https://www.postgresql.org/docs/current/sql-altertable.html) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ReplicaIdentity { + None, + Full, + Default, + Index(Ident), +} + +impl fmt::Display for ReplicaIdentity { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + ReplicaIdentity::None => f.write_str("NONE"), + ReplicaIdentity::Full => f.write_str("FULL"), + ReplicaIdentity::Default => f.write_str("DEFAULT"), + ReplicaIdentity::Index(idx) => write!(f, "USING INDEX {}", idx), + } + } +} + /// An `ALTER TABLE` (`Statement::AlterTable`) operation #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -208,6 +231,13 @@ pub enum AlterTableOperation { old_partitions: Vec, new_partitions: Vec, }, + /// REPLICA IDENTITY { DEFAULT | USING INDEX index_name | FULL | NOTHING } + /// + /// Note: this is a PostgreSQL-specific operation. + /// Please refer to [PostgreSQL documentation](https://www.postgresql.org/docs/current/sql-altertable.html) + ReplicaIdentity { + identity: ReplicaIdentity, + }, /// Add Partitions AddPartitions { if_not_exists: bool, @@ -729,6 +759,9 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::Lock { equals, lock } => { write!(f, "LOCK {}{}", if *equals { "= " } else { "" }, lock) } + AlterTableOperation::ReplicaIdentity { identity } => { + write!(f, "REPLICA IDENTITY {identity}") + } } } } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 3791154d1..d1d706e46 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -65,7 +65,7 @@ pub use self::ddl::{ DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, - ProcedureParam, ReferentialAction, TableConstraint, TagsColumnOption, + ProcedureParam, ReferentialAction, ReplicaIdentity, TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index c09359480..cb1c48cab 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1175,6 +1175,7 @@ impl Spanned for AlterTableOperation { AlterTableOperation::Algorithm { .. } => Span::empty(), AlterTableOperation::AutoIncrement { value, .. } => value.span(), AlterTableOperation::Lock { .. } => Span::empty(), + AlterTableOperation::ReplicaIdentity { .. } => Span::empty(), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 1e0227271..4665bb769 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8774,6 +8774,23 @@ impl<'a> Parser<'a> { let equals = self.consume_token(&Token::Eq); let value = self.parse_number_value()?; AlterTableOperation::AutoIncrement { equals, value } + } else if self.parse_keywords(&[Keyword::REPLICA, Keyword::IDENTITY]) { + let identity = if self.parse_keyword(Keyword::NONE) { + ReplicaIdentity::None + } else if self.parse_keyword(Keyword::FULL) { + ReplicaIdentity::Full + } else if self.parse_keyword(Keyword::DEFAULT) { + ReplicaIdentity::Default + } else if self.parse_keywords(&[Keyword::USING, Keyword::INDEX]) { + ReplicaIdentity::Index(self.parse_identifier()?) + } else { + return self.expected( + "NONE, FULL, DEFAULT, or USING INDEX index_name after REPLICA IDENTITY", + self.peek_token(), + ); + }; + + AlterTableOperation::ReplicaIdentity { identity } } else { let options: Vec = self.parse_options_with_keywords(&[Keyword::SET, Keyword::TBLPROPERTIES])?; @@ -8783,7 +8800,7 @@ impl<'a> Parser<'a> { } } else { return self.expected( - "ADD, RENAME, PARTITION, SWAP, DROP, or SET TBLPROPERTIES after ALTER TABLE", + "ADD, RENAME, PARTITION, SWAP, DROP, REPLICA IDENTITY, or SET TBLPROPERTIES after ALTER TABLE", self.peek_token(), ); } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 859eca453..9e71883cc 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5959,3 +5959,30 @@ fn parse_varbit_datatype() { _ => unreachable!(), } } + +#[test] +fn parse_alter_table_replica_identity() { + match pg_and_generic().verified_stmt("ALTER TABLE foo REPLICA IDENTITY FULL") { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::ReplicaIdentity { + identity: ReplicaIdentity::Full + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("ALTER TABLE foo REPLICA IDENTITY USING INDEX foo_idx") { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::ReplicaIdentity { + identity: ReplicaIdentity::Index("foo_idx".into()) + }] + ); + } + _ => unreachable!(), + } +} From 3c5995006073be1f9d9d445f578dd206ee889d9e Mon Sep 17 00:00:00 2001 From: Denys Tsomenko Date: Thu, 15 May 2025 17:40:21 +0300 Subject: [PATCH 128/286] Add support for INCLUDE/EXCLUDE NULLS for UNPIVOT (#1849) --- src/ast/mod.rs | 20 +++++++ src/ast/query.rs | 11 ++-- src/ast/spans.rs | 1 + src/parser/mod.rs | 10 ++++ tests/sqlparser_common.rs | 109 ++++++++++++++++++++++++-------------- 5 files changed, 109 insertions(+), 42 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d1d706e46..1ad1b6da7 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -9633,6 +9633,26 @@ impl fmt::Display for OpenStatement { } } +/// Specifies Include / Exclude NULL within UNPIVOT command. +/// For example +/// `UNPIVOT (column1 FOR new_column IN (col3, col4, col5, col6))` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum NullInclusion { + IncludeNulls, + ExcludeNulls, +} + +impl fmt::Display for NullInclusion { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + NullInclusion::IncludeNulls => write!(f, "INCLUDE NULLS"), + NullInclusion::ExcludeNulls => write!(f, "EXCLUDE NULLS"), + } + } +} + #[cfg(test)] mod tests { use super::*; diff --git a/src/ast/query.rs b/src/ast/query.rs index 33168695f..adb8516f9 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1336,7 +1336,7 @@ pub enum TableFactor { /// /// Syntax: /// ```sql - /// table UNPIVOT(value FOR name IN (column1, [ column2, ... ])) [ alias ] + /// table UNPIVOT [ { INCLUDE | EXCLUDE } NULLS ] (value FOR name IN (column1, [ column2, ... ])) [ alias ] /// ``` /// /// See . @@ -1345,6 +1345,7 @@ pub enum TableFactor { value: Ident, name: Ident, columns: Vec, + null_inclusion: Option, alias: Option, }, /// A `MATCH_RECOGNIZE` operation on a table. @@ -2015,15 +2016,19 @@ impl fmt::Display for TableFactor { } TableFactor::Unpivot { table, + null_inclusion, value, name, columns, alias, } => { + write!(f, "{table} UNPIVOT")?; + if let Some(null_inclusion) = null_inclusion { + write!(f, " {null_inclusion} ")?; + } write!( f, - "{} UNPIVOT({} FOR {} IN ({}))", - table, + "({} FOR {} IN ({}))", value, name, display_comma_separated(columns) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index cb1c48cab..bffd11722 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1944,6 +1944,7 @@ impl Spanned for TableFactor { TableFactor::Unpivot { table, value, + null_inclusion: _, name, columns, alias, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 4665bb769..838996130 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13368,6 +13368,15 @@ impl<'a> Parser<'a> { &mut self, table: TableFactor, ) -> Result { + let null_inclusion = if self.parse_keyword(Keyword::INCLUDE) { + self.expect_keyword_is(Keyword::NULLS)?; + Some(NullInclusion::IncludeNulls) + } else if self.parse_keyword(Keyword::EXCLUDE) { + self.expect_keyword_is(Keyword::NULLS)?; + Some(NullInclusion::ExcludeNulls) + } else { + None + }; self.expect_token(&Token::LParen)?; let value = self.parse_identifier()?; self.expect_keyword_is(Keyword::FOR)?; @@ -13379,6 +13388,7 @@ impl<'a> Parser<'a> { Ok(TableFactor::Unpivot { table: Box::new(table), value, + null_inclusion, name, columns, alias, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a07bcc68e..8e3bc0021 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -10746,49 +10746,47 @@ fn parse_unpivot_table() { "SELECT * FROM sales AS s ", "UNPIVOT(quantity FOR quarter IN (Q1, Q2, Q3, Q4)) AS u (product, quarter, quantity)" ); - - pretty_assertions::assert_eq!( - verified_only_select(sql).from[0].relation, - Unpivot { - table: Box::new(TableFactor::Table { - name: ObjectName::from(vec![Ident::new("sales")]), - alias: Some(TableAlias { - name: Ident::new("s"), - columns: vec![] - }), - args: None, - with_hints: vec![], - version: None, - partitions: vec![], - with_ordinality: false, - json_path: None, - sample: None, - index_hints: vec![], + let base_unpivot = Unpivot { + table: Box::new(TableFactor::Table { + name: ObjectName::from(vec![Ident::new("sales")]), + alias: Some(TableAlias { + name: Ident::new("s"), + columns: vec![], }), - value: Ident { - value: "quantity".to_string(), - quote_style: None, - span: Span::empty() - }, + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }), + null_inclusion: None, + value: Ident { + value: "quantity".to_string(), + quote_style: None, + span: Span::empty(), + }, - name: Ident { - value: "quarter".to_string(), - quote_style: None, - span: Span::empty() - }, - columns: ["Q1", "Q2", "Q3", "Q4"] + name: Ident { + value: "quarter".to_string(), + quote_style: None, + span: Span::empty(), + }, + columns: ["Q1", "Q2", "Q3", "Q4"] + .into_iter() + .map(Ident::new) + .collect(), + alias: Some(TableAlias { + name: Ident::new("u"), + columns: ["product", "quarter", "quantity"] .into_iter() - .map(Ident::new) + .map(TableAliasColumnDef::from_name) .collect(), - alias: Some(TableAlias { - name: Ident::new("u"), - columns: ["product", "quarter", "quantity"] - .into_iter() - .map(TableAliasColumnDef::from_name) - .collect(), - }), - } - ); + }), + }; + pretty_assertions::assert_eq!(verified_only_select(sql).from[0].relation, base_unpivot); assert_eq!(verified_stmt(sql).to_string(), sql); let sql_without_aliases = concat!( @@ -10808,6 +10806,38 @@ fn parse_unpivot_table() { verified_stmt(sql_without_aliases).to_string(), sql_without_aliases ); + + let sql_unpivot_exclude_nulls = concat!( + "SELECT * FROM sales AS s ", + "UNPIVOT EXCLUDE NULLS (quantity FOR quarter IN (Q1, Q2, Q3, Q4)) AS u (product, quarter, quantity)" + ); + + if let Unpivot { null_inclusion, .. } = + &verified_only_select(sql_unpivot_exclude_nulls).from[0].relation + { + assert_eq!(*null_inclusion, Some(NullInclusion::ExcludeNulls)); + } + + assert_eq!( + verified_stmt(sql_unpivot_exclude_nulls).to_string(), + sql_unpivot_exclude_nulls + ); + + let sql_unpivot_include_nulls = concat!( + "SELECT * FROM sales AS s ", + "UNPIVOT INCLUDE NULLS (quantity FOR quarter IN (Q1, Q2, Q3, Q4)) AS u (product, quarter, quantity)" + ); + + if let Unpivot { null_inclusion, .. } = + &verified_only_select(sql_unpivot_include_nulls).from[0].relation + { + assert_eq!(*null_inclusion, Some(NullInclusion::IncludeNulls)); + } + + assert_eq!( + verified_stmt(sql_unpivot_include_nulls).to_string(), + sql_unpivot_include_nulls + ); } #[test] @@ -10904,6 +10934,7 @@ fn parse_pivot_unpivot_table() { sample: None, index_hints: vec![], }), + null_inclusion: None, value: Ident { value: "population".to_string(), quote_style: None, From ae587dcbec9f6d2d3cde415ead7e8e376167355c Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Thu, 15 May 2025 16:43:16 +0200 Subject: [PATCH 129/286] pretty print improvements (#1851) --- README.md | 4 + src/ast/dml.rs | 57 ++++++---- src/ast/mod.rs | 36 ++++--- src/ast/query.rs | 9 +- src/display_utils.rs | 51 ++++----- tests/pretty_print.rs | 242 ++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 332 insertions(+), 67 deletions(-) diff --git a/README.md b/README.md index 6acfbcef7..666be17c0 100644 --- a/README.md +++ b/README.md @@ -89,10 +89,14 @@ keywords, the following should hold true for all SQL: ```rust // Parse SQL +let sql = "SELECT 'hello'"; let ast = Parser::parse_sql(&GenericDialect, sql).unwrap(); // The original SQL text can be generated from the AST assert_eq!(ast[0].to_string(), sql); + +// The SQL can also be pretty-printed with newlines and indentation +assert_eq!(format!("{:#}", ast[0]), "SELECT\n 'hello'"); ``` There are still some cases in this crate where different SQL with seemingly diff --git a/src/ast/dml.rs b/src/ast/dml.rs index 7ed17be9b..e4081a637 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -29,6 +29,8 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; +use crate::display_utils::{indented_list, Indent, SpaceOrNewline}; + pub use super::ddl::{ColumnDef, TableConstraint}; use super::{ @@ -579,28 +581,32 @@ impl Display for Insert { )?; } if !self.columns.is_empty() { - write!(f, "({}) ", display_comma_separated(&self.columns))?; + write!(f, "({})", display_comma_separated(&self.columns))?; + SpaceOrNewline.fmt(f)?; } if let Some(ref parts) = self.partitioned { if !parts.is_empty() { - write!(f, "PARTITION ({}) ", display_comma_separated(parts))?; + write!(f, "PARTITION ({})", display_comma_separated(parts))?; + SpaceOrNewline.fmt(f)?; } } if !self.after_columns.is_empty() { - write!(f, "({}) ", display_comma_separated(&self.after_columns))?; + write!(f, "({})", display_comma_separated(&self.after_columns))?; + SpaceOrNewline.fmt(f)?; } if let Some(settings) = &self.settings { - write!(f, "SETTINGS {} ", display_comma_separated(settings))?; + write!(f, "SETTINGS {}", display_comma_separated(settings))?; + SpaceOrNewline.fmt(f)?; } if let Some(source) = &self.source { - write!(f, "{source}")?; + source.fmt(f)?; } else if !self.assignments.is_empty() { - write!(f, "SET ")?; - write!(f, "{}", display_comma_separated(&self.assignments))?; + write!(f, "SET")?; + indented_list(f, &self.assignments)?; } else if let Some(format_clause) = &self.format_clause { - write!(f, "{format_clause}")?; + format_clause.fmt(f)?; } else if self.columns.is_empty() { write!(f, "DEFAULT VALUES")?; } @@ -620,7 +626,9 @@ impl Display for Insert { } if let Some(returning) = &self.returning { - write!(f, " RETURNING {}", display_comma_separated(returning))?; + SpaceOrNewline.fmt(f)?; + f.write_str("RETURNING")?; + indented_list(f, returning)?; } Ok(()) } @@ -649,32 +657,45 @@ pub struct Delete { impl Display for Delete { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - write!(f, "DELETE ")?; + f.write_str("DELETE")?; if !self.tables.is_empty() { - write!(f, "{} ", display_comma_separated(&self.tables))?; + indented_list(f, &self.tables)?; } match &self.from { FromTable::WithFromKeyword(from) => { - write!(f, "FROM {}", display_comma_separated(from))?; + f.write_str(" FROM")?; + indented_list(f, from)?; } FromTable::WithoutKeyword(from) => { - write!(f, "{}", display_comma_separated(from))?; + indented_list(f, from)?; } } if let Some(using) = &self.using { - write!(f, " USING {}", display_comma_separated(using))?; + SpaceOrNewline.fmt(f)?; + f.write_str("USING")?; + indented_list(f, using)?; } if let Some(selection) = &self.selection { - write!(f, " WHERE {selection}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("WHERE")?; + SpaceOrNewline.fmt(f)?; + Indent(selection).fmt(f)?; } if let Some(returning) = &self.returning { - write!(f, " RETURNING {}", display_comma_separated(returning))?; + SpaceOrNewline.fmt(f)?; + f.write_str("RETURNING")?; + indented_list(f, returning)?; } if !self.order_by.is_empty() { - write!(f, " ORDER BY {}", display_comma_separated(&self.order_by))?; + SpaceOrNewline.fmt(f)?; + f.write_str("ORDER BY")?; + indented_list(f, &self.order_by)?; } if let Some(limit) = &self.limit { - write!(f, " LIMIT {limit}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("LIMIT")?; + SpaceOrNewline.fmt(f)?; + Indent(limit).fmt(f)?; } Ok(()) } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 1ad1b6da7..d711a1062 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -41,7 +41,7 @@ use serde::{Deserialize, Serialize}; use sqlparser_derive::{Visit, VisitMut}; use crate::{ - display_utils::SpaceOrNewline, + display_utils::{indented_list, SpaceOrNewline}, tokenizer::{Span, Token}, }; use crate::{ @@ -4548,7 +4548,7 @@ impl fmt::Display for Statement { } Ok(()) } - Statement::Insert(insert) => write!(f, "{insert}"), + Statement::Insert(insert) => insert.fmt(f), Statement::Install { extension_name: name, } => write!(f, "INSTALL {name}"), @@ -4611,30 +4611,42 @@ impl fmt::Display for Statement { returning, or, } => { - write!(f, "UPDATE ")?; + f.write_str("UPDATE ")?; if let Some(or) = or { - write!(f, "{or} ")?; + or.fmt(f)?; + f.write_str(" ")?; } - write!(f, "{table}")?; + table.fmt(f)?; if let Some(UpdateTableFromKind::BeforeSet(from)) = from { - write!(f, " FROM {}", display_comma_separated(from))?; + SpaceOrNewline.fmt(f)?; + f.write_str("FROM")?; + indented_list(f, from)?; } if !assignments.is_empty() { - write!(f, " SET {}", display_comma_separated(assignments))?; + SpaceOrNewline.fmt(f)?; + f.write_str("SET")?; + indented_list(f, assignments)?; } if let Some(UpdateTableFromKind::AfterSet(from)) = from { - write!(f, " FROM {}", display_comma_separated(from))?; + SpaceOrNewline.fmt(f)?; + f.write_str("FROM")?; + indented_list(f, from)?; } if let Some(selection) = selection { - write!(f, " WHERE {selection}")?; + SpaceOrNewline.fmt(f)?; + f.write_str("WHERE")?; + SpaceOrNewline.fmt(f)?; + Indent(selection).fmt(f)?; } if let Some(returning) = returning { - write!(f, " RETURNING {}", display_comma_separated(returning))?; + SpaceOrNewline.fmt(f)?; + f.write_str("RETURNING")?; + indented_list(f, returning)?; } Ok(()) } - Statement::Delete(delete) => write!(f, "{delete}"), - Statement::Open(open) => write!(f, "{open}"), + Statement::Delete(delete) => delete.fmt(f), + Statement::Open(open) => open.fmt(f), Statement::Close { cursor } => { write!(f, "CLOSE {cursor}")?; diff --git a/src/ast/query.rs b/src/ast/query.rs index adb8516f9..5b784b199 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2888,13 +2888,14 @@ pub struct Values { impl fmt::Display for Values { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "VALUES ")?; + f.write_str("VALUES")?; let prefix = if self.explicit_row { "ROW" } else { "" }; let mut delim = ""; for row in &self.rows { - write!(f, "{delim}")?; - delim = ", "; - write!(f, "{prefix}({})", display_comma_separated(row))?; + f.write_str(delim)?; + delim = ","; + SpaceOrNewline.fmt(f)?; + Indent(format_args!("{prefix}({})", display_comma_separated(row))).fmt(f)?; } Ok(()) } diff --git a/src/display_utils.rs b/src/display_utils.rs index e7e1272fb..849aea941 100644 --- a/src/display_utils.rs +++ b/src/display_utils.rs @@ -31,13 +31,10 @@ where T: Write, { fn write_str(&mut self, s: &str) -> fmt::Result { - let mut first = true; - for line in s.split('\n') { - if !first { - write!(self.0, "\n{INDENT}")?; - } - self.0.write_str(line)?; - first = false; + self.0.write_str(s)?; + // Our NewLine and SpaceOrNewline utils always print individual newlines as a single-character string. + if s == "\n" { + self.0.write_str(INDENT)?; } Ok(()) } @@ -71,7 +68,7 @@ impl Display for SpaceOrNewline { /// A value that displays a comma-separated list of values. /// When pretty-printed (using {:#}), it displays each value on a new line. -pub struct DisplayCommaSeparated<'a, T: fmt::Display>(&'a [T]); +pub(crate) struct DisplayCommaSeparated<'a, T: fmt::Display>(&'a [T]); impl fmt::Display for DisplayCommaSeparated<'_, T> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { @@ -89,45 +86,33 @@ impl fmt::Display for DisplayCommaSeparated<'_, T> { } /// Displays a whitespace, followed by a comma-separated list that is indented when pretty-printed. -pub(crate) fn indented_list(f: &mut fmt::Formatter, slice: &[T]) -> fmt::Result { +pub(crate) fn indented_list(f: &mut fmt::Formatter, items: &[T]) -> fmt::Result { SpaceOrNewline.fmt(f)?; - Indent(DisplayCommaSeparated(slice)).fmt(f) + Indent(DisplayCommaSeparated(items)).fmt(f) } #[cfg(test)] mod tests { use super::*; - struct DisplayCharByChar(T); + #[test] + fn test_indent() { + struct TwoLines; - impl Display for DisplayCharByChar { - fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - for c in self.0.to_string().chars() { - write!(f, "{}", c)?; + impl Display for TwoLines { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + f.write_str("line 1")?; + SpaceOrNewline.fmt(f)?; + f.write_str("line 2") } - Ok(()) } - } - #[test] - fn test_indent() { - let original = "line 1\nline 2"; - let indent = Indent(original); + let indent = Indent(TwoLines); assert_eq!( indent.to_string(), - original, + TwoLines.to_string(), "Only the alternate form should be indented" ); - let expected = " line 1\n line 2"; - assert_eq!(format!("{:#}", indent), expected); - let display_char_by_char = DisplayCharByChar(original); - assert_eq!(format!("{:#}", Indent(display_char_by_char)), expected); - } - - #[test] - fn test_space_or_newline() { - let space_or_newline = SpaceOrNewline; - assert_eq!(format!("{}", space_or_newline), " "); - assert_eq!(format!("{:#}", space_or_newline), "\n"); + assert_eq!(format!("{:#}", indent), " line 1\n line 2"); } } diff --git a/tests/pretty_print.rs b/tests/pretty_print.rs index 1eb8ca41c..b4adbe352 100644 --- a/tests/pretty_print.rs +++ b/tests/pretty_print.rs @@ -155,3 +155,245 @@ FROM "#.trim() ); } + +#[test] +fn test_pretty_print_multiline_string() { + assert_eq!( + prettify("SELECT 'multiline\nstring' AS str"), + r#" +SELECT + 'multiline +string' AS str +"# + .trim(), + "A literal string with a newline should be kept as is. The contents of the string should not be indented." + ); +} + +#[test] +fn test_pretty_print_insert_values() { + assert_eq!( + prettify("INSERT INTO my_table (a, b, c) VALUES (1, 2, 3), (4, 5, 6)"), + r#" +INSERT INTO my_table (a, b, c) +VALUES + (1, 2, 3), + (4, 5, 6) +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_insert_select() { + assert_eq!( + prettify("INSERT INTO my_table (a, b) SELECT x, y FROM source_table RETURNING a AS id"), + r#" +INSERT INTO my_table (a, b) +SELECT + x, + y +FROM + source_table +RETURNING + a AS id +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_update() { + assert_eq!( + prettify("UPDATE my_table SET a = 1, b = 2 WHERE x > 0 RETURNING id, name"), + r#" +UPDATE my_table +SET + a = 1, + b = 2 +WHERE + x > 0 +RETURNING + id, + name +"# + .trim() + ); +} + +#[test] +fn test_pretty_print_delete() { + assert_eq!( + prettify("DELETE FROM my_table WHERE x > 0 RETURNING id, name"), + r#" +DELETE FROM + my_table +WHERE + x > 0 +RETURNING + id, + name +"# + .trim() + ); + + assert_eq!( + prettify("DELETE table1, table2"), + r#" +DELETE + table1, + table2 +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_create_table() { + assert_eq!( + prettify("CREATE TABLE my_table (id INT PRIMARY KEY, name VARCHAR(255) NOT NULL, CONSTRAINT fk_other FOREIGN KEY (id) REFERENCES other_table(id))"), + r#" +CREATE TABLE my_table ( + id INT PRIMARY KEY, + name VARCHAR(255) NOT NULL, + CONSTRAINT fk_other FOREIGN KEY (id) REFERENCES other_table(id) +) +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_create_view() { + assert_eq!( + prettify("CREATE VIEW my_view AS SELECT a, b FROM my_table WHERE x > 0"), + r#" +CREATE VIEW my_view AS +SELECT + a, + b +FROM + my_table +WHERE + x > 0 +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_create_function() { + assert_eq!( + prettify("CREATE FUNCTION my_func() RETURNS INT BEGIN SELECT COUNT(*) INTO @count FROM my_table; RETURN @count; END"), + r#" +CREATE FUNCTION my_func() RETURNS INT +BEGIN + SELECT COUNT(*) INTO @count FROM my_table; + RETURN @count; +END +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_json_table() { + assert_eq!( + prettify("SELECT * FROM JSON_TABLE(@json, '$[*]' COLUMNS (id INT PATH '$.id', name VARCHAR(255) PATH '$.name')) AS jt"), + r#" +SELECT + * +FROM + JSON_TABLE( + @json, + '$[*]' COLUMNS ( + id INT PATH '$.id', + name VARCHAR(255) PATH '$.name' + ) + ) AS jt +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_transaction_blocks() { + assert_eq!( + prettify("BEGIN; UPDATE my_table SET x = 1; COMMIT;"), + r#" +BEGIN; +UPDATE my_table SET x = 1; +COMMIT; +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_control_flow() { + assert_eq!( + prettify("IF x > 0 THEN SELECT 'positive'; ELSE SELECT 'negative'; END IF;"), + r#" +IF x > 0 THEN + SELECT 'positive'; +ELSE + SELECT 'negative'; +END IF; +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_merge() { + assert_eq!( + prettify("MERGE INTO target_table t USING source_table s ON t.id = s.id WHEN MATCHED THEN UPDATE SET t.value = s.value WHEN NOT MATCHED THEN INSERT (id, value) VALUES (s.id, s.value)"), + r#" +MERGE INTO target_table t +USING source_table s ON t.id = s.id +WHEN MATCHED THEN + UPDATE SET t.value = s.value +WHEN NOT MATCHED THEN + INSERT (id, value) VALUES (s.id, s.value) +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_create_index() { + assert_eq!( + prettify("CREATE INDEX idx_name ON my_table (column1, column2)"), + r#" +CREATE INDEX idx_name +ON my_table (column1, column2) +"# + .trim() + ); +} + +#[test] +#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] +fn test_pretty_print_explain() { + assert_eq!( + prettify("EXPLAIN ANALYZE SELECT * FROM my_table WHERE x > 0"), + r#" +EXPLAIN ANALYZE +SELECT + * +FROM + my_table +WHERE + x > 0 +"# + .trim() + ); +} From e7bf186e44a35628af2c2581ad7e82ea8e8df2d3 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Mon, 19 May 2025 13:09:00 +0200 Subject: [PATCH 130/286] fix new rust 1.87 cargo clippy warnings (#1856) --- src/ast/visitor.rs | 13 ++++++++----- src/lib.rs | 4 ++++ src/parser/mod.rs | 5 +---- tests/sqlparser_common.rs | 2 +- 4 files changed, 14 insertions(+), 10 deletions(-) diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index 50985a3e7..ab4f73aa4 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -741,7 +741,7 @@ mod tests { } } - fn do_visit(sql: &str, visitor: &mut V) -> Statement { + fn do_visit>(sql: &str, visitor: &mut V) -> Statement { let dialect = GenericDialect {}; let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); let s = Parser::new(&dialect) @@ -749,7 +749,8 @@ mod tests { .parse_statement() .unwrap(); - s.visit(visitor); + let flow = s.visit(visitor); + assert_eq!(flow, ControlFlow::Continue(())); s } @@ -938,7 +939,8 @@ mod tests { .unwrap(); let mut visitor = QuickVisitor {}; - s.visit(&mut visitor); + let flow = s.visit(&mut visitor); + assert_eq!(flow, ControlFlow::Continue(())); } } @@ -969,7 +971,7 @@ mod visit_mut_tests { } } - fn do_visit_mut(sql: &str, visitor: &mut V) -> Statement { + fn do_visit_mut>(sql: &str, visitor: &mut V) -> Statement { let dialect = GenericDialect {}; let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); let mut s = Parser::new(&dialect) @@ -977,7 +979,8 @@ mod visit_mut_tests { .parse_statement() .unwrap(); - s.visit(visitor); + let flow = s.visit(visitor); + assert_eq!(flow, ControlFlow::Continue(())); s } diff --git a/src/lib.rs b/src/lib.rs index c81ab5002..dbfd1791a 100644 --- a/src/lib.rs +++ b/src/lib.rs @@ -149,6 +149,10 @@ #![cfg_attr(not(feature = "std"), no_std)] #![allow(clippy::upper_case_acronyms)] +// Permit large enum variants to keep a unified, expressive AST. +// Splitting complex nodes (expressions, statements, types) into separate types +// would bloat the API and hide intent. Extra memory is a worthwhile tradeoff. +#![allow(clippy::large_enum_variant)] // Allow proc-macros to find this crate extern crate self as sqlparser; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 838996130..992d19c49 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2558,10 +2558,7 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; let mut trim_where = None; if let Token::Word(word) = self.peek_token().token { - if [Keyword::BOTH, Keyword::LEADING, Keyword::TRAILING] - .iter() - .any(|d| word.keyword == *d) - { + if [Keyword::BOTH, Keyword::LEADING, Keyword::TRAILING].contains(&word.keyword) { trim_where = Some(self.parse_trim_where()?); } } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 8e3bc0021..86c473d7d 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -14346,7 +14346,7 @@ fn test_visit_order() { let sql = "SELECT CASE a WHEN 1 THEN 2 WHEN 3 THEN 4 ELSE 5 END"; let stmt = verified_stmt(sql); let mut visited = vec![]; - sqlparser::ast::visit_expressions(&stmt, |expr| { + let _ = sqlparser::ast::visit_expressions(&stmt, |expr| { visited.push(expr.to_string()); core::ops::ControlFlow::<()>::Continue(()) }); From 525ed81fdeffb3d765be99fc3cc4fd31060e0df1 Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Tue, 20 May 2025 06:21:41 +0200 Subject: [PATCH 131/286] Update criterion requirement from 0.5 to 0.6 in /sqlparser_bench (#1857) Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: ifeanyi --- sqlparser_bench/Cargo.toml | 2 +- sqlparser_bench/benches/sqlparser_bench.rs | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/sqlparser_bench/Cargo.toml b/sqlparser_bench/Cargo.toml index 2c1f0ae4d..01c59be75 100644 --- a/sqlparser_bench/Cargo.toml +++ b/sqlparser_bench/Cargo.toml @@ -26,7 +26,7 @@ edition = "2018" sqlparser = { path = "../" } [dev-dependencies] -criterion = "0.5" +criterion = "0.6" [[bench]] name = "sqlparser_bench" diff --git a/sqlparser_bench/benches/sqlparser_bench.rs b/sqlparser_bench/benches/sqlparser_bench.rs index a7768cbc9..24c59c076 100644 --- a/sqlparser_bench/benches/sqlparser_bench.rs +++ b/sqlparser_bench/benches/sqlparser_bench.rs @@ -68,7 +68,7 @@ fn basic_queries(c: &mut Criterion) { }; group.bench_function("parse_large_statement", |b| { - b.iter(|| Parser::parse_sql(&dialect, criterion::black_box(large_statement.as_str()))); + b.iter(|| Parser::parse_sql(&dialect, std::hint::black_box(large_statement.as_str()))); }); let large_statement = Parser::parse_sql(&dialect, large_statement.as_str()) From a496f78803d1ceda4a09f5b5d952636f19175f16 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Tue, 20 May 2025 18:34:48 +0200 Subject: [PATCH 132/286] pretty-print CREATE TABLE statements (#1854) --- src/ast/dml.rs | 15 ++++++++++----- src/display_utils.rs | 2 +- tests/pretty_print.rs | 1 - 3 files changed, 11 insertions(+), 7 deletions(-) diff --git a/src/ast/dml.rs b/src/ast/dml.rs index e4081a637..da82a4ede 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -29,7 +29,7 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use crate::display_utils::{indented_list, Indent, SpaceOrNewline}; +use crate::display_utils::{indented_list, DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; pub use super::ddl::{ColumnDef, TableConstraint}; @@ -267,14 +267,19 @@ impl Display for CreateTable { write!(f, " ON CLUSTER {}", on_cluster)?; } if !self.columns.is_empty() || !self.constraints.is_empty() { - write!(f, " ({}", display_comma_separated(&self.columns))?; + f.write_str(" (")?; + NewLine.fmt(f)?; + Indent(DisplayCommaSeparated(&self.columns)).fmt(f)?; if !self.columns.is_empty() && !self.constraints.is_empty() { - write!(f, ", ")?; + f.write_str(",")?; + SpaceOrNewline.fmt(f)?; } - write!(f, "{})", display_comma_separated(&self.constraints))?; + Indent(DisplayCommaSeparated(&self.constraints)).fmt(f)?; + NewLine.fmt(f)?; + f.write_str(")")?; } else if self.query.is_none() && self.like.is_none() && self.clone.is_none() { // PostgreSQL allows `CREATE TABLE t ();`, but requires empty parens - write!(f, " ()")?; + f.write_str(" ()")?; } // Hive table comment should be after column definitions, please refer to: diff --git a/src/display_utils.rs b/src/display_utils.rs index 849aea941..e594a34ef 100644 --- a/src/display_utils.rs +++ b/src/display_utils.rs @@ -68,7 +68,7 @@ impl Display for SpaceOrNewline { /// A value that displays a comma-separated list of values. /// When pretty-printed (using {:#}), it displays each value on a new line. -pub(crate) struct DisplayCommaSeparated<'a, T: fmt::Display>(&'a [T]); +pub(crate) struct DisplayCommaSeparated<'a, T: fmt::Display>(pub(crate) &'a [T]); impl fmt::Display for DisplayCommaSeparated<'_, T> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { diff --git a/tests/pretty_print.rs b/tests/pretty_print.rs index b4adbe352..d6794218b 100644 --- a/tests/pretty_print.rs +++ b/tests/pretty_print.rs @@ -249,7 +249,6 @@ DELETE } #[test] -#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] fn test_pretty_print_create_table() { assert_eq!( prettify("CREATE TABLE my_table (id INT PRIMARY KEY, name VARCHAR(255) NOT NULL, CONSTRAINT fk_other FOREIGN KEY (id) REFERENCES other_table(id))"), From 3f4d5f96ee2622a0a2ad3fedf56dfd93baeb9a03 Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Wed, 21 May 2025 05:44:33 +0200 Subject: [PATCH 133/286] pretty-print CREATE VIEW statements (#1855) --- src/ast/mod.rs | 4 +++- tests/pretty_print.rs | 1 - 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d711a1062..e18251ea2 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4858,7 +4858,9 @@ impl fmt::Display for Statement { if matches!(options, CreateTableOptions::Options(_)) { write!(f, " {options}")?; } - write!(f, " AS {query}")?; + f.write_str(" AS")?; + SpaceOrNewline.fmt(f)?; + query.fmt(f)?; if *with_no_schema_binding { write!(f, " WITH NO SCHEMA BINDING")?; } diff --git a/tests/pretty_print.rs b/tests/pretty_print.rs index d6794218b..e1d35eb04 100644 --- a/tests/pretty_print.rs +++ b/tests/pretty_print.rs @@ -264,7 +264,6 @@ CREATE TABLE my_table ( } #[test] -#[ignore = "https://github.com/apache/datafusion-sqlparser-rs/issues/1850"] fn test_pretty_print_create_view() { assert_eq!( prettify("CREATE VIEW my_view AS SELECT a, b FROM my_table WHERE x > 0"), From 05d7ffb1d5ef6e4c4852a200e0aa08fec224aa3c Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Wed, 21 May 2025 05:49:28 +0200 Subject: [PATCH 134/286] Handle optional datatypes properly in `CREATE FUNCTION` statements (#1826) Co-authored-by: Ifeanyi Ubah --- src/parser/mod.rs | 19 +++- tests/sqlparser_postgres.rs | 211 ++++++++++++++++++++++++++++++++++++ 2 files changed, 225 insertions(+), 5 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 992d19c49..f6a45adad 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5273,12 +5273,21 @@ impl<'a> Parser<'a> { // parse: [ argname ] argtype let mut name = None; let mut data_type = self.parse_data_type()?; - if let DataType::Custom(n, _) = &data_type { - // the first token is actually a name - match n.0[0].clone() { - ObjectNamePart::Identifier(ident) => name = Some(ident), + + // To check whether the first token is a name or a type, we need to + // peek the next token, which if it is another type keyword, then the + // first token is a name and not a type in itself. + let data_type_idx = self.get_current_index(); + if let Some(next_data_type) = self.maybe_parse(|parser| parser.parse_data_type())? { + let token = self.token_at(data_type_idx); + + // We ensure that the token is a `Word` token, and not other special tokens. + if !matches!(token.token, Token::Word(_)) { + return self.expected("a name or type", token.clone()); } - data_type = self.parse_data_type()?; + + name = Some(Ident::new(token.to_string())); + data_type = next_data_type; } let default_expr = if self.parse_keyword(Keyword::DEFAULT) || self.consume_token(&Token::Eq) diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 9e71883cc..682c0d6cc 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -21,6 +21,7 @@ #[macro_use] mod test_utils; + use helpers::attached_token::AttachedToken; use sqlparser::tokenizer::Span; use test_utils::*; @@ -4105,6 +4106,216 @@ fn parse_update_in_with_subquery() { pg_and_generic().verified_stmt(r#"WITH "result" AS (UPDATE "Hero" SET "name" = 'Captain America', "number_of_movies" = "number_of_movies" + 1 WHERE "secret_identity" = 'Sam Wilson' RETURNING "id", "name", "secret_identity", "number_of_movies") SELECT * FROM "result""#); } +#[test] +fn parser_create_function_with_args() { + let sql1 = r#"CREATE OR REPLACE FUNCTION check_strings_different(str1 VARCHAR, str2 VARCHAR) RETURNS BOOLEAN LANGUAGE plpgsql AS $$ +BEGIN + IF str1 <> str2 THEN + RETURN TRUE; + ELSE + RETURN FALSE; + END IF; +END; +$$"#; + + assert_eq!( + pg_and_generic().verified_stmt(sql1), + Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: true, + temporary: false, + name: ObjectName::from(vec![Ident::new("check_strings_different")]), + args: Some(vec![ + OperateFunctionArg::with_name( + "str1", + DataType::Varchar(None), + ), + OperateFunctionArg::with_name( + "str2", + DataType::Varchar(None), + ), + ]), + return_type: Some(DataType::Boolean), + language: Some("plpgsql".into()), + behavior: None, + called_on_null: None, + parallel: None, + function_body: Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( + (Value::DollarQuotedString(DollarQuotedString {value: "\nBEGIN\n IF str1 <> str2 THEN\n RETURN TRUE;\n ELSE\n RETURN FALSE;\n END IF;\nEND;\n".to_owned(), tag: None})).with_empty_span() + ))), + if_not_exists: false, + using: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }) + ); + + let sql2 = r#"CREATE OR REPLACE FUNCTION check_not_zero(int1 INT) RETURNS BOOLEAN LANGUAGE plpgsql AS $$ +BEGIN + IF int1 <> 0 THEN + RETURN TRUE; + ELSE + RETURN FALSE; + END IF; +END; +$$"#; + assert_eq!( + pg_and_generic().verified_stmt(sql2), + Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: true, + temporary: false, + name: ObjectName::from(vec![Ident::new("check_not_zero")]), + args: Some(vec![ + OperateFunctionArg::with_name( + "int1", + DataType::Int(None) + ) + ]), + return_type: Some(DataType::Boolean), + language: Some("plpgsql".into()), + behavior: None, + called_on_null: None, + parallel: None, + function_body: Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( + (Value::DollarQuotedString(DollarQuotedString {value: "\nBEGIN\n IF int1 <> 0 THEN\n RETURN TRUE;\n ELSE\n RETURN FALSE;\n END IF;\nEND;\n".to_owned(), tag: None})).with_empty_span() + ))), + if_not_exists: false, + using: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }) + ); + + let sql3 = r#"CREATE OR REPLACE FUNCTION check_values_different(a INT, b INT) RETURNS BOOLEAN LANGUAGE plpgsql AS $$ +BEGIN + IF a <> b THEN + RETURN TRUE; + ELSE + RETURN FALSE; + END IF; +END; +$$"#; + assert_eq!( + pg_and_generic().verified_stmt(sql3), + Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: true, + temporary: false, + name: ObjectName::from(vec![Ident::new("check_values_different")]), + args: Some(vec![ + OperateFunctionArg::with_name( + "a", + DataType::Int(None) + ), + OperateFunctionArg::with_name( + "b", + DataType::Int(None) + ), + ]), + return_type: Some(DataType::Boolean), + language: Some("plpgsql".into()), + behavior: None, + called_on_null: None, + parallel: None, + function_body: Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( + (Value::DollarQuotedString(DollarQuotedString {value: "\nBEGIN\n IF a <> b THEN\n RETURN TRUE;\n ELSE\n RETURN FALSE;\n END IF;\nEND;\n".to_owned(), tag: None})).with_empty_span() + ))), + if_not_exists: false, + using: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }) + ); + + let sql4 = r#"CREATE OR REPLACE FUNCTION check_values_different(int1 INT, int2 INT) RETURNS BOOLEAN LANGUAGE plpgsql AS $$ +BEGIN + IF int1 <> int2 THEN + RETURN TRUE; + ELSE + RETURN FALSE; + END IF; +END; +$$"#; + assert_eq!( + pg_and_generic().verified_stmt(sql4), + Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: true, + temporary: false, + name: ObjectName::from(vec![Ident::new("check_values_different")]), + args: Some(vec![ + OperateFunctionArg::with_name( + "int1", + DataType::Int(None) + ), + OperateFunctionArg::with_name( + "int2", + DataType::Int(None) + ), + ]), + return_type: Some(DataType::Boolean), + language: Some("plpgsql".into()), + behavior: None, + called_on_null: None, + parallel: None, + function_body: Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( + (Value::DollarQuotedString(DollarQuotedString {value: "\nBEGIN\n IF int1 <> int2 THEN\n RETURN TRUE;\n ELSE\n RETURN FALSE;\n END IF;\nEND;\n".to_owned(), tag: None})).with_empty_span() + ))), + if_not_exists: false, + using: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }) + ); + + let sql5 = r#"CREATE OR REPLACE FUNCTION foo(a TIMESTAMP WITH TIME ZONE, b VARCHAR) RETURNS BOOLEAN LANGUAGE plpgsql AS $$ + BEGIN + RETURN TRUE; + END; + $$"#; + assert_eq!( + pg_and_generic().verified_stmt(sql5), + Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: true, + temporary: false, + name: ObjectName::from(vec![Ident::new("foo")]), + args: Some(vec![ + OperateFunctionArg::with_name( + "a", + DataType::Timestamp(None, TimezoneInfo::WithTimeZone) + ), + OperateFunctionArg::with_name("b", DataType::Varchar(None)), + ]), + return_type: Some(DataType::Boolean), + language: Some("plpgsql".into()), + behavior: None, + called_on_null: None, + parallel: None, + function_body: Some(CreateFunctionBody::AsBeforeOptions(Expr::Value( + (Value::DollarQuotedString(DollarQuotedString { + value: "\n BEGIN\n RETURN TRUE;\n END;\n ".to_owned(), + tag: None + })) + .with_empty_span() + ))), + if_not_exists: false, + using: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }) + ); + + let incorrect_sql = "CREATE FUNCTION add(function(struct int64), b INTEGER) RETURNS INTEGER LANGUAGE SQL IMMUTABLE STRICT PARALLEL SAFE AS 'select $1 + $2;'"; + assert!(pg().parse_sql_statements(incorrect_sql).is_err(),); +} + #[test] fn parse_create_function() { let sql = "CREATE FUNCTION add(INTEGER, INTEGER) RETURNS INTEGER LANGUAGE SQL IMMUTABLE STRICT PARALLEL SAFE AS 'select $1 + $2;'"; From bf2b72fbe0ae08ec5faeffc6e81e3b830c302c5d Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Fri, 23 May 2025 06:09:05 +0100 Subject: [PATCH 135/286] Mysql: Add `SRID` column option (#1852) --- src/ast/ddl.rs | 10 ++++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 4 ++++ tests/sqlparser_mysql.rs | 5 +++++ 5 files changed, 21 insertions(+) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 270897130..a8a1fdbae 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1758,6 +1758,13 @@ pub enum ColumnOption { /// ``` /// [Snowflake]: https://docs.snowflake.com/en/sql-reference/sql/create-table Tags(TagsColumnOption), + /// MySQL specific: Spatial reference identifier + /// Syntax: + /// ```sql + /// CREATE TABLE geom (g GEOMETRY NOT NULL SRID 4326); + /// ``` + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/creating-spatial-indexes.html + Srid(Box), } impl fmt::Display for ColumnOption { @@ -1873,6 +1880,9 @@ impl fmt::Display for ColumnOption { Tags(tags) => { write!(f, "{tags}") } + Srid(srid) => { + write!(f, "SRID {srid}") + } } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index bffd11722..1c28b62cc 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -871,6 +871,7 @@ impl Spanned for ColumnOption { ColumnOption::OnConflict(..) => Span::empty(), ColumnOption::Policy(..) => Span::empty(), ColumnOption::Tags(..) => Span::empty(), + ColumnOption::Srid(..) => Span::empty(), } } } diff --git a/src/keywords.rs b/src/keywords.rs index aaa2e167f..f5c5e567e 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -844,6 +844,7 @@ define_keywords!( SQLSTATE, SQLWARNING, SQRT, + SRID, STABLE, STAGE, START, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f6a45adad..47b321d9c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7754,6 +7754,10 @@ impl<'a> Parser<'a> { && dialect_of!(self is MySqlDialect | SQLiteDialect | DuckDbDialect | GenericDialect) { self.parse_optional_column_option_as() + } else if self.parse_keyword(Keyword::SRID) + && dialect_of!(self is MySqlDialect | GenericDialect) + { + Ok(Some(ColumnOption::Srid(Box::new(self.parse_expr()?)))) } else if self.parse_keyword(Keyword::IDENTITY) && dialect_of!(self is MsSqlDialect | GenericDialect) { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index dd279894a..bcde14ee5 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3745,6 +3745,11 @@ fn parse_begin_without_transaction() { mysql().verified_stmt("BEGIN"); } +#[test] +fn parse_geometric_types_srid_option() { + mysql_and_generic().verified_stmt("CREATE TABLE t (a geometry SRID 4326)"); +} + #[test] fn parse_double_precision() { mysql().verified_stmt("CREATE TABLE foo (bar DOUBLE)"); From 301726541a9c34d979bbfe82014752fd656afc15 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Fri, 23 May 2025 01:19:16 -0400 Subject: [PATCH 136/286] Add support for table valued functions for SQL Server (#1839) --- src/ast/data_type.rs | 24 ++++++++++- src/ast/ddl.rs | 6 +++ src/ast/mod.rs | 24 +++++++++++ src/parser/mod.rs | 81 ++++++++++++++++++++++++++++--------- tests/sqlparser_mssql.rs | 87 ++++++++++++++++++++++++++++++++++++++++ 5 files changed, 201 insertions(+), 21 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 52919de8a..3a4958c9f 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -48,7 +48,17 @@ pub enum DataType { /// Table type in [PostgreSQL], e.g. CREATE FUNCTION RETURNS TABLE(...). /// /// [PostgreSQL]: https://www.postgresql.org/docs/15/sql-createfunction.html - Table(Vec), + /// [MsSQL]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql?view=sql-server-ver16#c-create-a-multi-statement-table-valued-function + Table(Option>), + /// Table type with a name, e.g. CREATE FUNCTION RETURNS @result TABLE(...). + /// + /// [MsSQl]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql?view=sql-server-ver16#table + NamedTable { + /// Table name. + name: ObjectName, + /// Table columns. + columns: Vec, + }, /// Fixed-length character type, e.g. CHARACTER(10). Character(Option), /// Fixed-length char type, e.g. CHAR(10). @@ -716,7 +726,17 @@ impl fmt::Display for DataType { DataType::Unspecified => Ok(()), DataType::Trigger => write!(f, "TRIGGER"), DataType::AnyType => write!(f, "ANY TYPE"), - DataType::Table(fields) => write!(f, "TABLE({})", display_comma_separated(fields)), + DataType::Table(fields) => match fields { + Some(fields) => { + write!(f, "TABLE({})", display_comma_separated(fields)) + } + None => { + write!(f, "TABLE") + } + }, + DataType::NamedTable { name, columns } => { + write!(f, "{} TABLE ({})", name, display_comma_separated(columns)) + } DataType::GeometricType(kind) => write!(f, "{}", kind), } } diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index a8a1fdbae..bbfa7d3c9 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2356,6 +2356,12 @@ impl fmt::Display for CreateFunction { if let Some(CreateFunctionBody::Return(function_body)) = &self.function_body { write!(f, " RETURN {function_body}")?; } + if let Some(CreateFunctionBody::AsReturnExpr(function_body)) = &self.function_body { + write!(f, " AS RETURN {function_body}")?; + } + if let Some(CreateFunctionBody::AsReturnSelect(function_body)) = &self.function_body { + write!(f, " AS RETURN {function_body}")?; + } if let Some(using) = &self.using { write!(f, " {using}")?; } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index e18251ea2..8e7bec3f4 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -8780,6 +8780,30 @@ pub enum CreateFunctionBody { /// /// [PostgreSQL]: https://www.postgresql.org/docs/current/sql-createfunction.html Return(Expr), + + /// Function body expression using the 'AS RETURN' keywords + /// + /// Example: + /// ```sql + /// CREATE FUNCTION myfunc(a INT, b INT) + /// RETURNS TABLE + /// AS RETURN (SELECT a + b AS sum); + /// ``` + /// + /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql + AsReturnExpr(Expr), + + /// Function body expression using the 'AS RETURN' keywords, with an un-parenthesized SELECT query + /// + /// Example: + /// ```sql + /// CREATE FUNCTION myfunc(a INT, b INT) + /// RETURNS TABLE + /// AS RETURN SELECT a + b AS sum; + /// ``` + /// + /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql?view=sql-server-ver16#select_stmt + AsReturnSelect(Select), } #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 47b321d9c..4299d1566 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5204,19 +5204,62 @@ impl<'a> Parser<'a> { let (name, args) = self.parse_create_function_name_and_params()?; self.expect_keyword(Keyword::RETURNS)?; - let return_type = Some(self.parse_data_type()?); - self.expect_keyword_is(Keyword::AS)?; + let return_table = self.maybe_parse(|p| { + let return_table_name = p.parse_identifier()?; + + p.expect_keyword_is(Keyword::TABLE)?; + p.prev_token(); + + let table_column_defs = match p.parse_data_type()? { + DataType::Table(Some(table_column_defs)) if !table_column_defs.is_empty() => { + table_column_defs + } + _ => parser_err!( + "Expected table column definitions after TABLE keyword", + p.peek_token().span.start + )?, + }; + + Ok(DataType::NamedTable { + name: ObjectName(vec![ObjectNamePart::Identifier(return_table_name)]), + columns: table_column_defs, + }) + })?; - let begin_token = self.expect_keyword(Keyword::BEGIN)?; - let statements = self.parse_statement_list(&[Keyword::END])?; - let end_token = self.expect_keyword(Keyword::END)?; + let return_type = if return_table.is_some() { + return_table + } else { + Some(self.parse_data_type()?) + }; - let function_body = Some(CreateFunctionBody::AsBeginEnd(BeginEndStatements { - begin_token: AttachedToken(begin_token), - statements, - end_token: AttachedToken(end_token), - })); + let _ = self.parse_keyword(Keyword::AS); + + let function_body = if self.peek_keyword(Keyword::BEGIN) { + let begin_token = self.expect_keyword(Keyword::BEGIN)?; + let statements = self.parse_statement_list(&[Keyword::END])?; + let end_token = self.expect_keyword(Keyword::END)?; + + Some(CreateFunctionBody::AsBeginEnd(BeginEndStatements { + begin_token: AttachedToken(begin_token), + statements, + end_token: AttachedToken(end_token), + })) + } else if self.parse_keyword(Keyword::RETURN) { + if self.peek_token() == Token::LParen { + Some(CreateFunctionBody::AsReturnExpr(self.parse_expr()?)) + } else if self.peek_keyword(Keyword::SELECT) { + let select = self.parse_select()?; + Some(CreateFunctionBody::AsReturnSelect(select)) + } else { + parser_err!( + "Expected a subquery (or bare SELECT statement) after RETURN", + self.peek_token().span.start + )? + } + } else { + parser_err!("Unparsable function body", self.peek_token().span.start)? + }; Ok(Statement::CreateFunction(CreateFunction { or_alter, @@ -9797,8 +9840,14 @@ impl<'a> Parser<'a> { Ok(DataType::AnyType) } Keyword::TABLE => { - let columns = self.parse_returns_table_columns()?; - Ok(DataType::Table(columns)) + // an LParen after the TABLE keyword indicates that table columns are being defined + // whereas no LParen indicates an anonymous table expression will be returned + if self.peek_token() == Token::LParen { + let columns = self.parse_returns_table_columns()?; + Ok(DataType::Table(Some(columns))) + } else { + Ok(DataType::Table(None)) + } } Keyword::SIGNED => { if self.parse_keyword(Keyword::INTEGER) { @@ -9839,13 +9888,7 @@ impl<'a> Parser<'a> { } fn parse_returns_table_column(&mut self) -> Result { - let name = self.parse_identifier()?; - let data_type = self.parse_data_type()?; - Ok(ColumnDef { - name, - data_type, - options: Vec::new(), // No constraints expected here - }) + self.parse_column_def() } fn parse_returns_table_columns(&mut self) -> Result, ParserError> { diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 7b3769ec8..32388c44a 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -254,6 +254,12 @@ fn parse_create_function() { "; let _ = ms().verified_stmt(multi_statement_function); + let multi_statement_function_without_as = multi_statement_function.replace(" AS", ""); + let _ = ms().one_statement_parses_to( + &multi_statement_function_without_as, + multi_statement_function, + ); + let create_function_with_conditional = "\ CREATE FUNCTION some_scalar_udf() \ RETURNS INT \ @@ -288,6 +294,87 @@ fn parse_create_function() { END\ "; let _ = ms().verified_stmt(create_function_with_return_expression); + + let create_inline_table_value_function = "\ + CREATE FUNCTION some_inline_tvf(@foo INT, @bar VARCHAR(256)) \ + RETURNS TABLE \ + AS \ + RETURN (SELECT 1 AS col_1)\ + "; + let _ = ms().verified_stmt(create_inline_table_value_function); + + let create_inline_table_value_function_without_parentheses = "\ + CREATE FUNCTION some_inline_tvf(@foo INT, @bar VARCHAR(256)) \ + RETURNS TABLE \ + AS \ + RETURN SELECT 1 AS col_1\ + "; + let _ = ms().verified_stmt(create_inline_table_value_function_without_parentheses); + + let create_inline_table_value_function_without_as = + create_inline_table_value_function.replace(" AS", ""); + let _ = ms().one_statement_parses_to( + &create_inline_table_value_function_without_as, + create_inline_table_value_function, + ); + + let create_multi_statement_table_value_function = "\ + CREATE FUNCTION some_multi_statement_tvf(@foo INT, @bar VARCHAR(256)) \ + RETURNS @t TABLE (col_1 INT) \ + AS \ + BEGIN \ + INSERT INTO @t SELECT 1; \ + RETURN; \ + END\ + "; + let _ = ms().verified_stmt(create_multi_statement_table_value_function); + + let create_multi_statement_table_value_function_without_as = + create_multi_statement_table_value_function.replace(" AS", ""); + let _ = ms().one_statement_parses_to( + &create_multi_statement_table_value_function_without_as, + create_multi_statement_table_value_function, + ); + + let create_multi_statement_table_value_function_with_constraints = "\ + CREATE FUNCTION some_multi_statement_tvf(@foo INT, @bar VARCHAR(256)) \ + RETURNS @t TABLE (col_1 INT NOT NULL) \ + AS \ + BEGIN \ + INSERT INTO @t SELECT 1; \ + RETURN @t; \ + END\ + "; + let _ = ms().verified_stmt(create_multi_statement_table_value_function_with_constraints); + + let create_multi_statement_tvf_without_table_definition = "\ + CREATE FUNCTION incorrect_tvf(@foo INT, @bar VARCHAR(256)) \ + RETURNS @t TABLE () + AS \ + BEGIN \ + INSERT INTO @t SELECT 1; \ + RETURN @t; \ + END\ + "; + assert_eq!( + ParserError::ParserError("Unparsable function body".to_owned()), + ms().parse_sql_statements(create_multi_statement_tvf_without_table_definition) + .unwrap_err() + ); + + let create_inline_tvf_without_subquery_or_bare_select = "\ + CREATE FUNCTION incorrect_tvf(@foo INT, @bar VARCHAR(256)) \ + RETURNS TABLE + AS \ + RETURN 'hi'\ + "; + assert_eq!( + ParserError::ParserError( + "Expected a subquery (or bare SELECT statement) after RETURN".to_owned() + ), + ms().parse_sql_statements(create_inline_tvf_without_subquery_or_bare_select) + .unwrap_err() + ); } #[test] From 9159d08c5ed4f08457e05878476051e27f6daa34 Mon Sep 17 00:00:00 2001 From: hulk Date: Wed, 28 May 2025 13:09:40 +0800 Subject: [PATCH 137/286] Keep the COLUMN keyword only if it exists when dropping the column (#1862) --- src/ast/ddl.rs | 5 ++++- src/ast/spans.rs | 1 + src/parser/mod.rs | 3 ++- tests/sqlparser_common.rs | 5 +++-- tests/sqlparser_mysql.rs | 2 ++ 5 files changed, 12 insertions(+), 4 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index bbfa7d3c9..06b85b0f1 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -139,6 +139,7 @@ pub enum AlterTableOperation { }, /// `DROP [ COLUMN ] [ IF EXISTS ] [ CASCADE ]` DropColumn { + has_column_keyword: bool, column_name: Ident, if_exists: bool, drop_behavior: Option, @@ -606,12 +607,14 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::DropPrimaryKey => write!(f, "DROP PRIMARY KEY"), AlterTableOperation::DropForeignKey { name } => write!(f, "DROP FOREIGN KEY {name}"), AlterTableOperation::DropColumn { + has_column_keyword, column_name, if_exists, drop_behavior, } => write!( f, - "DROP COLUMN {}{}{}", + "DROP {}{}{}{}", + if *has_column_keyword { "COLUMN " } else { "" }, if *if_exists { "IF EXISTS " } else { "" }, column_name, match drop_behavior { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 1c28b62cc..d612738c5 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1090,6 +1090,7 @@ impl Spanned for AlterTableOperation { drop_behavior: _, } => name.span, AlterTableOperation::DropColumn { + has_column_keyword: _, column_name, if_exists: _, drop_behavior: _, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 4299d1566..fcd07aa46 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8608,11 +8608,12 @@ impl<'a> Parser<'a> { } else if self.parse_keywords(&[Keyword::CLUSTERING, Keyword::KEY]) { AlterTableOperation::DropClusteringKey } else { - let _ = self.parse_keyword(Keyword::COLUMN); // [ COLUMN ] + let has_column_keyword = self.parse_keyword(Keyword::COLUMN); // [ COLUMN ] let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); let column_name = self.parse_identifier()?; let drop_behavior = self.parse_optional_drop_behavior(); AlterTableOperation::DropColumn { + has_column_keyword, column_name, if_exists, drop_behavior, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 86c473d7d..d02d7d837 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4926,17 +4926,18 @@ fn parse_alter_table_drop_column() { check_one("DROP COLUMN IF EXISTS is_active CASCADE"); check_one("DROP COLUMN IF EXISTS is_active RESTRICT"); one_statement_parses_to( - "ALTER TABLE tab DROP IF EXISTS is_active CASCADE", + "ALTER TABLE tab DROP COLUMN IF EXISTS is_active CASCADE", "ALTER TABLE tab DROP COLUMN IF EXISTS is_active CASCADE", ); one_statement_parses_to( "ALTER TABLE tab DROP is_active CASCADE", - "ALTER TABLE tab DROP COLUMN is_active CASCADE", + "ALTER TABLE tab DROP is_active CASCADE", ); fn check_one(constraint_text: &str) { match alter_table_op(verified_stmt(&format!("ALTER TABLE tab {constraint_text}"))) { AlterTableOperation::DropColumn { + has_column_keyword: true, column_name, if_exists, drop_behavior, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index bcde14ee5..71a5d9051 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2801,6 +2801,7 @@ fn parse_alter_table_with_algorithm() { operations, vec![ AlterTableOperation::DropColumn { + has_column_keyword: true, column_name: Ident::new("password_digest"), if_exists: false, drop_behavior: None, @@ -2848,6 +2849,7 @@ fn parse_alter_table_with_lock() { operations, vec![ AlterTableOperation::DropColumn { + has_column_keyword: true, column_name: Ident::new("password_digest"), if_exists: false, drop_behavior: None, From eacf00d269bf3d1323100c33a18f139d6921adb5 Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Thu, 29 May 2025 05:49:28 -0400 Subject: [PATCH 138/286] Add support for parameter default values in SQL Server (#1866) --- src/parser/mod.rs | 8 +++++++- tests/sqlparser_mssql.rs | 40 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 47 insertions(+), 1 deletion(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index fcd07aa46..c1be87642 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5289,11 +5289,17 @@ impl<'a> Parser<'a> { |parser: &mut Parser| -> Result { let name = parser.parse_identifier()?; let data_type = parser.parse_data_type()?; + let default_expr = if parser.consume_token(&Token::Eq) { + Some(parser.parse_expr()?) + } else { + None + }; + Ok(OperateFunctionArg { mode: None, name: Some(name), data_type, - default_expr: None, + default_expr, }) }; self.expect_token(&Token::LParen)?; diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 32388c44a..2a3145028 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -377,6 +377,46 @@ fn parse_create_function() { ); } +#[test] +fn parse_create_function_parameter_default_values() { + let single_default_sql = + "CREATE FUNCTION test_func(@param1 INT = 42) RETURNS INT AS BEGIN RETURN @param1; END"; + assert_eq!( + ms().verified_stmt(single_default_sql), + Statement::CreateFunction(CreateFunction { + or_alter: false, + or_replace: false, + temporary: false, + if_not_exists: false, + name: ObjectName::from(vec![Ident::new("test_func")]), + args: Some(vec![OperateFunctionArg { + mode: None, + name: Some(Ident::new("@param1")), + data_type: DataType::Int(None), + default_expr: Some(Expr::Value((number("42")).with_empty_span())), + },]), + return_type: Some(DataType::Int(None)), + function_body: Some(CreateFunctionBody::AsBeginEnd(BeginEndStatements { + begin_token: AttachedToken::empty(), + statements: vec![Statement::Return(ReturnStatement { + value: Some(ReturnStatementValue::Expr(Expr::Identifier(Ident::new( + "@param1" + )))), + })], + end_token: AttachedToken::empty(), + })), + behavior: None, + called_on_null: None, + parallel: None, + using: None, + language: None, + determinism_specifier: None, + options: None, + remote_connection: None, + }), + ); +} + #[test] fn parse_mssql_apply_join() { let _ = ms_and_generic().verified_only_select( From a8bde39efb4c3568fb3dc685b440962d50403fc3 Mon Sep 17 00:00:00 2001 From: Hendrik Makait Date: Fri, 30 May 2025 09:14:36 +0200 Subject: [PATCH 139/286] Add support for `TABLESAMPLE` pipe operator (#1860) --- src/ast/query.rs | 14 +++++++++++--- src/parser/mod.rs | 15 +++++++++++++-- tests/sqlparser_common.rs | 5 +++++ 3 files changed, 29 insertions(+), 5 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 5b784b199..ffe1e4023 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1559,7 +1559,7 @@ impl fmt::Display for TableSampleBucket { } impl fmt::Display for TableSample { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - write!(f, " {}", self.modifier)?; + write!(f, "{}", self.modifier)?; if let Some(name) = &self.name { write!(f, " {}", name)?; } @@ -1862,7 +1862,7 @@ impl fmt::Display for TableFactor { write!(f, " WITH ORDINALITY")?; } if let Some(TableSampleKind::BeforeTableAlias(sample)) = sample { - write!(f, "{sample}")?; + write!(f, " {sample}")?; } if let Some(alias) = alias { write!(f, " AS {alias}")?; @@ -1877,7 +1877,7 @@ impl fmt::Display for TableFactor { write!(f, "{version}")?; } if let Some(TableSampleKind::AfterTableAlias(sample)) = sample { - write!(f, "{sample}")?; + write!(f, " {sample}")?; } Ok(()) } @@ -2680,6 +2680,10 @@ pub enum PipeOperator { full_table_exprs: Vec, group_by_expr: Vec, }, + /// Selects a random sample of rows from the input table. + /// Syntax: `|> TABLESAMPLE SYSTEM (10 PERCENT) + /// See more at + TableSample { sample: Box }, } impl fmt::Display for PipeOperator { @@ -2731,6 +2735,10 @@ impl fmt::Display for PipeOperator { PipeOperator::OrderBy { exprs } => { write!(f, "ORDER BY {}", display_comma_separated(exprs.as_slice())) } + + PipeOperator::TableSample { sample } => { + write!(f, "{}", sample) + } } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c1be87642..6d6423842 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11054,6 +11054,7 @@ impl<'a> Parser<'a> { Keyword::LIMIT, Keyword::AGGREGATE, Keyword::ORDER, + Keyword::TABLESAMPLE, ])?; match kw { Keyword::SELECT => { @@ -11116,6 +11117,10 @@ impl<'a> Parser<'a> { let exprs = self.parse_comma_separated(Parser::parse_order_by_expr)?; pipe_operators.push(PipeOperator::OrderBy { exprs }) } + Keyword::TABLESAMPLE => { + let sample = self.parse_table_sample(TableSampleModifier::TableSample)?; + pipe_operators.push(PipeOperator::TableSample { sample }); + } unhandled => { return Err(ParserError::ParserError(format!( "`expect_one_of_keywords` further up allowed unhandled keyword: {unhandled:?}" @@ -12760,7 +12765,13 @@ impl<'a> Parser<'a> { } else { return Ok(None); }; + self.parse_table_sample(modifier).map(Some) + } + fn parse_table_sample( + &mut self, + modifier: TableSampleModifier, + ) -> Result, ParserError> { let name = match self.parse_one_of_keywords(&[ Keyword::BERNOULLI, Keyword::ROW, @@ -12842,14 +12853,14 @@ impl<'a> Parser<'a> { None }; - Ok(Some(Box::new(TableSample { + Ok(Box::new(TableSample { modifier, name, quantity, seed, bucket, offset, - }))) + })) } fn parse_table_sample_seed( diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index d02d7d837..1cc793175 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15156,6 +15156,11 @@ fn parse_pipeline_operator() { dialects.verified_stmt("SELECT * FROM users |> ORDER BY id DESC"); dialects.verified_stmt("SELECT * FROM users |> ORDER BY id DESC, name ASC"); + // tablesample pipe operator + dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE BERNOULLI (50)"); + dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE SYSTEM (50 PERCENT)"); + dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE SYSTEM (50) REPEATABLE (10)"); + // many pipes dialects.verified_stmt( "SELECT * FROM CustomerOrders |> AGGREGATE SUM(cost) AS total_cost GROUP BY customer_id, state, item_type |> EXTEND COUNT(*) OVER (PARTITION BY customer_id) AS num_orders |> WHERE num_orders > 1 |> AGGREGATE AVG(total_cost) AS average GROUP BY state DESC, item_type ASC", From 80d47eee84f0954ad24918b046e47f08e3762142 Mon Sep 17 00:00:00 2001 From: Dmitriy Mazurin Date: Fri, 30 May 2025 08:16:36 +0100 Subject: [PATCH 140/286] Adds support for mysql's drop index (#1864) Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 32 +++++++++++++++++++++----------- src/parser/mod.rs | 6 ++++++ tests/sqlparser_mysql.rs | 31 +++++++++++++++++++++++++++++++ 3 files changed, 58 insertions(+), 11 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8e7bec3f4..653f58e4d 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3405,6 +3405,9 @@ pub enum Statement { purge: bool, /// MySQL-specific "TEMPORARY" keyword temporary: bool, + /// MySQL-specific drop index syntax, which requires table specification + /// See + table: Option, }, /// ```sql /// DROP FUNCTION @@ -5242,17 +5245,24 @@ impl fmt::Display for Statement { restrict, purge, temporary, - } => write!( - f, - "DROP {}{}{} {}{}{}{}", - if *temporary { "TEMPORARY " } else { "" }, - object_type, - if *if_exists { " IF EXISTS" } else { "" }, - display_comma_separated(names), - if *cascade { " CASCADE" } else { "" }, - if *restrict { " RESTRICT" } else { "" }, - if *purge { " PURGE" } else { "" } - ), + table, + } => { + write!( + f, + "DROP {}{}{} {}{}{}{}", + if *temporary { "TEMPORARY " } else { "" }, + object_type, + if *if_exists { " IF EXISTS" } else { "" }, + display_comma_separated(names), + if *cascade { " CASCADE" } else { "" }, + if *restrict { " RESTRICT" } else { "" }, + if *purge { " PURGE" } else { "" }, + )?; + if let Some(table_name) = table.as_ref() { + write!(f, " ON {}", table_name)?; + }; + Ok(()) + } Statement::DropFunction { if_exists, func_desc, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6d6423842..f2da659f6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6255,6 +6255,11 @@ impl<'a> Parser<'a> { loc ); } + let table = if self.parse_keyword(Keyword::ON) { + Some(self.parse_object_name(false)?) + } else { + None + }; Ok(Statement::Drop { object_type, if_exists, @@ -6263,6 +6268,7 @@ impl<'a> Parser<'a> { restrict, purge, temporary, + table, }) } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 71a5d9051..337707c99 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3987,3 +3987,34 @@ fn parse_straight_join() { mysql() .verified_stmt("SELECT a.*, b.* FROM table_a STRAIGHT_JOIN table_b AS b ON a.b_id = b.id"); } + +#[test] +fn parse_drop_index() { + let sql = "DROP INDEX idx_name ON table_name"; + match mysql().verified_stmt(sql) { + Statement::Drop { + object_type, + if_exists, + names, + cascade, + restrict, + purge, + temporary, + table, + } => { + assert!(!if_exists); + assert_eq!(ObjectType::Index, object_type); + assert_eq!( + vec!["idx_name"], + names.iter().map(ToString::to_string).collect::>() + ); + assert!(!cascade); + assert!(!restrict); + assert!(!purge); + assert!(!temporary); + assert!(table.is_some()); + assert_eq!("table_name", table.unwrap().to_string()); + } + _ => unreachable!(), + } +} From 394a53448678ef16b715482e4b643badd035fb5e Mon Sep 17 00:00:00 2001 From: Denys Tsomenko Date: Mon, 2 Jun 2025 19:04:35 +0300 Subject: [PATCH 141/286] Fix: GROUPING SETS accept values without parenthesis (#1867) --- src/parser/mod.rs | 8 +++++++- tests/sqlparser_common.rs | 32 ++++++++++++++++++++++++++++++++ 2 files changed, 39 insertions(+), 1 deletion(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f2da659f6..a28540d14 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10111,7 +10111,13 @@ impl<'a> Parser<'a> { } if self.parse_keywords(&[Keyword::GROUPING, Keyword::SETS]) { self.expect_token(&Token::LParen)?; - let result = self.parse_comma_separated(|p| p.parse_tuple(true, true))?; + let result = self.parse_comma_separated(|p| { + if p.peek_token_ref().token == Token::LParen { + p.parse_tuple(true, true) + } else { + Ok(vec![p.parse_expr()?]) + } + })?; self.expect_token(&Token::RParen)?; modifiers.push(GroupByWithModifier::GroupingSets(Expr::GroupingSets( result, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 1cc793175..a1a8fc3b3 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2822,6 +2822,38 @@ fn parse_group_by_special_grouping_sets() { } } +#[test] +fn parse_group_by_grouping_sets_single_values() { + let sql = "SELECT a, b, SUM(c) FROM tab1 GROUP BY a, b GROUPING SETS ((a, b), a, (b), c, ())"; + let canonical = + "SELECT a, b, SUM(c) FROM tab1 GROUP BY a, b GROUPING SETS ((a, b), (a), (b), (c), ())"; + match all_dialects().one_statement_parses_to(sql, canonical) { + Statement::Query(query) => { + let group_by = &query.body.as_select().unwrap().group_by; + assert_eq!( + group_by, + &GroupByExpr::Expressions( + vec![ + Expr::Identifier(Ident::new("a")), + Expr::Identifier(Ident::new("b")) + ], + vec![GroupByWithModifier::GroupingSets(Expr::GroupingSets(vec![ + vec![ + Expr::Identifier(Ident::new("a")), + Expr::Identifier(Ident::new("b")) + ], + vec![Expr::Identifier(Ident::new("a"))], + vec![Expr::Identifier(Ident::new("b"))], + vec![Expr::Identifier(Ident::new("c"))], + vec![] + ]))] + ) + ); + } + _ => unreachable!(), + } +} + #[test] fn parse_select_having() { let sql = "SELECT foo FROM bar GROUP BY foo HAVING COUNT(*) > 1"; From 5327f0ce132e12de71db7d03711397c5ac6c0031 Mon Sep 17 00:00:00 2001 From: Artem Osipov <59066880+osipovartem@users.noreply.github.com> Date: Wed, 4 Jun 2025 20:49:07 +0300 Subject: [PATCH 142/286] Add ICEBERG keyword support to ALTER TABLE statement (#1869) --- src/ast/mod.rs | 13 ++++++-- src/ast/spans.rs | 1 + src/parser/mod.rs | 65 ++++++++++++++++++++---------------- src/test_utils.rs | 2 ++ tests/sqlparser_mysql.rs | 8 ++--- tests/sqlparser_postgres.rs | 6 ++-- tests/sqlparser_snowflake.rs | 7 ++++ 7 files changed, 64 insertions(+), 38 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 653f58e4d..711e580df 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3281,6 +3281,9 @@ pub enum Statement { /// For example: `ALTER TABLE table_name ON CLUSTER cluster_name ADD COLUMN c UInt32` /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/alter/update) on_cluster: Option, + /// Snowflake "ICEBERG" clause for Iceberg tables + /// + iceberg: bool, }, /// ```sql /// ALTER INDEX @@ -3405,7 +3408,7 @@ pub enum Statement { purge: bool, /// MySQL-specific "TEMPORARY" keyword temporary: bool, - /// MySQL-specific drop index syntax, which requires table specification + /// MySQL-specific drop index syntax, which requires table specification /// See table: Option, }, @@ -5139,8 +5142,14 @@ impl fmt::Display for Statement { operations, location, on_cluster, + iceberg, } => { - write!(f, "ALTER TABLE ")?; + if *iceberg { + write!(f, "ALTER ICEBERG TABLE ")?; + } else { + write!(f, "ALTER TABLE ")?; + } + if *if_exists { write!(f, "IF EXISTS ")?; } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index d612738c5..dd918c34b 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -431,6 +431,7 @@ impl Spanned for Statement { operations, location: _, on_cluster, + iceberg: _, } => union_spans( core::iter::once(name.span()) .chain(operations.iter().map(|i| i.span())) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a28540d14..3e721072b 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8893,38 +8893,15 @@ impl<'a> Parser<'a> { Keyword::ROLE, Keyword::POLICY, Keyword::CONNECTOR, + Keyword::ICEBERG, ])?; match object_type { Keyword::VIEW => self.parse_alter_view(), Keyword::TYPE => self.parse_alter_type(), - Keyword::TABLE => { - let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); - let only = self.parse_keyword(Keyword::ONLY); // [ ONLY ] - let table_name = self.parse_object_name(false)?; - let on_cluster = self.parse_optional_on_cluster()?; - let operations = self.parse_comma_separated(Parser::parse_alter_table_operation)?; - - let mut location = None; - if self.parse_keyword(Keyword::LOCATION) { - location = Some(HiveSetLocation { - has_set: false, - location: self.parse_identifier()?, - }); - } else if self.parse_keywords(&[Keyword::SET, Keyword::LOCATION]) { - location = Some(HiveSetLocation { - has_set: true, - location: self.parse_identifier()?, - }); - } - - Ok(Statement::AlterTable { - name: table_name, - if_exists, - only, - operations, - location, - on_cluster, - }) + Keyword::TABLE => self.parse_alter_table(false), + Keyword::ICEBERG => { + self.expect_keyword(Keyword::TABLE)?; + self.parse_alter_table(true) } Keyword::INDEX => { let index_name = self.parse_object_name(false)?; @@ -8952,6 +8929,38 @@ impl<'a> Parser<'a> { } } + /// Parse a [Statement::AlterTable] + pub fn parse_alter_table(&mut self, iceberg: bool) -> Result { + let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); + let only = self.parse_keyword(Keyword::ONLY); // [ ONLY ] + let table_name = self.parse_object_name(false)?; + let on_cluster = self.parse_optional_on_cluster()?; + let operations = self.parse_comma_separated(Parser::parse_alter_table_operation)?; + + let mut location = None; + if self.parse_keyword(Keyword::LOCATION) { + location = Some(HiveSetLocation { + has_set: false, + location: self.parse_identifier()?, + }); + } else if self.parse_keywords(&[Keyword::SET, Keyword::LOCATION]) { + location = Some(HiveSetLocation { + has_set: true, + location: self.parse_identifier()?, + }); + } + + Ok(Statement::AlterTable { + name: table_name, + if_exists, + only, + operations, + location, + on_cluster, + iceberg, + }) + } + pub fn parse_alter_view(&mut self) -> Result { let name = self.parse_object_name(false)?; let columns = self.parse_parenthesized_column_list(Optional, false)?; diff --git a/src/test_utils.rs b/src/test_utils.rs index 3c22fa911..24c0ca575 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -345,10 +345,12 @@ pub fn alter_table_op_with_name(stmt: Statement, expected_name: &str) -> AlterTa operations, on_cluster: _, location: _, + iceberg, } => { assert_eq!(name.to_string(), expected_name); assert!(!if_exists); assert!(!is_only); + assert!(!iceberg); only(operations) } _ => panic!("Expected ALTER TABLE statement"), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 337707c99..0800c3297 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2507,11 +2507,13 @@ fn parse_alter_table_add_column() { if_exists, only, operations, + iceberg, location: _, on_cluster: _, } => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); + assert!(!iceberg); assert!(!only); assert_eq!( operations, @@ -2536,8 +2538,7 @@ fn parse_alter_table_add_column() { if_exists, only, operations, - location: _, - on_cluster: _, + .. } => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); @@ -2574,8 +2575,7 @@ fn parse_alter_table_add_columns() { if_exists, only, operations, - location: _, - on_cluster: _, + .. } => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 682c0d6cc..d54d4e2af 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -834,8 +834,7 @@ fn parse_alter_table_add_columns() { if_exists, only, operations, - location: _, - on_cluster: _, + .. } => { assert_eq!(name.to_string(), "tab"); assert!(if_exists); @@ -915,8 +914,7 @@ fn parse_alter_table_owner_to() { if_exists: _, only: _, operations, - location: _, - on_cluster: _, + .. } => { assert_eq!(name.to_string(), "tab"); assert_eq!( diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 52be31435..b4d62506d 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -1591,6 +1591,13 @@ fn test_alter_table_clustering() { snowflake_and_generic().verified_stmt("ALTER TABLE tbl RESUME RECLUSTER"); } +#[test] +fn test_alter_iceberg_table() { + snowflake_and_generic().verified_stmt("ALTER ICEBERG TABLE tbl DROP CLUSTERING KEY"); + snowflake_and_generic().verified_stmt("ALTER ICEBERG TABLE tbl SUSPEND RECLUSTER"); + snowflake_and_generic().verified_stmt("ALTER ICEBERG TABLE tbl RESUME RECLUSTER"); +} + #[test] fn test_drop_stage() { match snowflake_and_generic().verified_stmt("DROP STAGE s1") { From de2cc7b50233a5e5c0a5c9fe7ab08d23b18b9d5c Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Fri, 6 Jun 2025 08:03:59 +0100 Subject: [PATCH 143/286] MySQL: Support `index_name` in FK constraints (#1871) --- src/ast/ddl.rs | 7 ++++++- src/ast/spans.rs | 2 ++ src/parser/mod.rs | 14 ++++++++------ tests/sqlparser_common.rs | 8 ++++++++ tests/sqlparser_mysql.rs | 7 +++++++ 5 files changed, 31 insertions(+), 7 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 06b85b0f1..bbc15704e 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1019,6 +1019,9 @@ pub enum TableConstraint { /// }`). ForeignKey { name: Option, + /// MySQL-specific field + /// + index_name: Option, columns: Vec, foreign_table: ObjectName, referred_columns: Vec, @@ -1129,6 +1132,7 @@ impl fmt::Display for TableConstraint { } TableConstraint::ForeignKey { name, + index_name, columns, foreign_table, referred_columns, @@ -1138,8 +1142,9 @@ impl fmt::Display for TableConstraint { } => { write!( f, - "{}FOREIGN KEY ({}) REFERENCES {}", + "{}FOREIGN KEY{} ({}) REFERENCES {}", display_constraint_name(name), + display_option_spaced(index_name), display_comma_separated(columns), foreign_table, )?; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index dd918c34b..a1bad2c54 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -671,6 +671,7 @@ impl Spanned for TableConstraint { TableConstraint::ForeignKey { name, columns, + index_name, foreign_table, referred_columns, on_delete, @@ -679,6 +680,7 @@ impl Spanned for TableConstraint { } => union_spans( name.iter() .map(|i| i.span) + .chain(index_name.iter().map(|i| i.span)) .chain(columns.iter().map(|i| i.span)) .chain(core::iter::once(foreign_table.span())) .chain(referred_columns.iter().map(|i| i.span)) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 3e721072b..677566c56 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8061,7 +8061,7 @@ impl<'a> Parser<'a> { let nulls_distinct = self.parse_optional_nulls_distinct()?; // optional index name - let index_name = self.parse_optional_indent()?; + let index_name = self.parse_optional_ident()?; let index_type = self.parse_optional_using_then_index_type()?; let columns = self.parse_parenthesized_column_list(Mandatory, false)?; @@ -8083,7 +8083,7 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::KEY)?; // optional index name - let index_name = self.parse_optional_indent()?; + let index_name = self.parse_optional_ident()?; let index_type = self.parse_optional_using_then_index_type()?; let columns = self.parse_parenthesized_column_list(Mandatory, false)?; @@ -8100,6 +8100,7 @@ impl<'a> Parser<'a> { } Token::Word(w) if w.keyword == Keyword::FOREIGN => { self.expect_keyword_is(Keyword::KEY)?; + let index_name = self.parse_optional_ident()?; let columns = self.parse_parenthesized_column_list(Mandatory, false)?; self.expect_keyword_is(Keyword::REFERENCES)?; let foreign_table = self.parse_object_name(false)?; @@ -8122,6 +8123,7 @@ impl<'a> Parser<'a> { Ok(Some(TableConstraint::ForeignKey { name, + index_name, columns, foreign_table, referred_columns, @@ -8145,7 +8147,7 @@ impl<'a> Parser<'a> { let name = match self.peek_token().token { Token::Word(word) if word.keyword == Keyword::USING => None, - _ => self.parse_optional_indent()?, + _ => self.parse_optional_ident()?, }; let index_type = self.parse_optional_using_then_index_type()?; @@ -8176,7 +8178,7 @@ impl<'a> Parser<'a> { let index_type_display = self.parse_index_type_display(); - let opt_index_name = self.parse_optional_indent()?; + let opt_index_name = self.parse_optional_ident()?; let columns = self.parse_parenthesized_column_list(Mandatory, false)?; @@ -8286,7 +8288,7 @@ impl<'a> Parser<'a> { /// Parse `[ident]`, mostly `ident` is name, like: /// `window_name`, `index_name`, ... - pub fn parse_optional_indent(&mut self) -> Result, ParserError> { + pub fn parse_optional_ident(&mut self) -> Result, ParserError> { self.maybe_parse(|parser| parser.parse_identifier()) } @@ -15698,7 +15700,7 @@ impl<'a> Parser<'a> { pub fn parse_window_spec(&mut self) -> Result { let window_name = match self.peek_token().token { Token::Word(word) if word.keyword == Keyword::NoKeyword => { - self.parse_optional_indent()? + self.parse_optional_ident()? } _ => None, }; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a1a8fc3b3..0431c720d 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3791,6 +3791,7 @@ fn parse_create_table() { vec![ TableConstraint::ForeignKey { name: Some("fkey".into()), + index_name: None, columns: vec!["lat".into()], foreign_table: ObjectName::from(vec!["othertable3".into()]), referred_columns: vec!["lat".into()], @@ -3800,6 +3801,7 @@ fn parse_create_table() { }, TableConstraint::ForeignKey { name: Some("fkey2".into()), + index_name: None, columns: vec!["lat".into()], foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], @@ -3809,6 +3811,7 @@ fn parse_create_table() { }, TableConstraint::ForeignKey { name: None, + index_name: None, columns: vec!["lat".into()], foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], @@ -3818,6 +3821,7 @@ fn parse_create_table() { }, TableConstraint::ForeignKey { name: None, + index_name: None, columns: vec!["lng".into()], foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["longitude".into()], @@ -3914,6 +3918,7 @@ fn parse_create_table_with_constraint_characteristics() { vec![ TableConstraint::ForeignKey { name: Some("fkey".into()), + index_name: None, columns: vec!["lat".into()], foreign_table: ObjectName::from(vec!["othertable3".into()]), referred_columns: vec!["lat".into()], @@ -3927,6 +3932,7 @@ fn parse_create_table_with_constraint_characteristics() { }, TableConstraint::ForeignKey { name: Some("fkey2".into()), + index_name: None, columns: vec!["lat".into()], foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], @@ -3940,6 +3946,7 @@ fn parse_create_table_with_constraint_characteristics() { }, TableConstraint::ForeignKey { name: None, + index_name: None, columns: vec!["lat".into()], foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["lat".into()], @@ -3953,6 +3960,7 @@ fn parse_create_table_with_constraint_characteristics() { }, TableConstraint::ForeignKey { name: None, + index_name: None, columns: vec!["lng".into()], foreign_table: ObjectName::from(vec!["othertable4".into()]), referred_columns: vec!["longitude".into()], diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 0800c3297..b1b7d539e 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3988,6 +3988,13 @@ fn parse_straight_join() { .verified_stmt("SELECT a.*, b.* FROM table_a STRAIGHT_JOIN table_b AS b ON a.b_id = b.id"); } +#[test] +fn mysql_foreign_key_with_index_name() { + mysql().verified_stmt( + "CREATE TABLE orders (customer_id INT, INDEX idx_customer (customer_id), CONSTRAINT fk_customer FOREIGN KEY idx_customer (customer_id) REFERENCES customers(id))", + ); +} + #[test] fn parse_drop_index() { let sql = "DROP INDEX idx_name ON table_name"; From 4cf5e571d3baad070ca76b5e17c9e145af0e3b1a Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Fri, 6 Jun 2025 08:10:03 +0100 Subject: [PATCH 144/286] Postgres: Apply `ONLY` keyword per table in TRUNCATE stmt (#1872) --- src/ast/mod.rs | 14 ++++++++------ src/ast/spans.rs | 1 - src/parser/mod.rs | 8 ++++---- tests/sqlparser_common.rs | 28 ++++++++++++++++++++++++++++ tests/sqlparser_postgres.rs | 7 ++++--- 5 files changed, 44 insertions(+), 14 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 711e580df..3ccce0619 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3014,9 +3014,6 @@ pub enum Statement { /// TABLE - optional keyword; table: bool, /// Postgres-specific option - /// [ TRUNCATE TABLE ONLY ] - only: bool, - /// Postgres-specific option /// [ RESTART IDENTITY | CONTINUE IDENTITY ] identity: Option, /// Postgres-specific option @@ -4425,17 +4422,15 @@ impl fmt::Display for Statement { table_names, partitions, table, - only, identity, cascade, on_cluster, } => { let table = if *table { "TABLE " } else { "" }; - let only = if *only { "ONLY " } else { "" }; write!( f, - "TRUNCATE {table}{only}{table_names}", + "TRUNCATE {table}{table_names}", table_names = display_comma_separated(table_names) )?; @@ -6106,10 +6101,17 @@ pub struct TruncateTableTarget { /// name of the table being truncated #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] pub name: ObjectName, + /// Postgres-specific option + /// [ TRUNCATE TABLE ONLY ] + /// + pub only: bool, } impl fmt::Display for TruncateTableTarget { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + if self.only { + write!(f, "ONLY ")?; + }; write!(f, "{}", self.name) } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index a1bad2c54..11986f8c4 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -311,7 +311,6 @@ impl Spanned for Statement { table_names, partitions, table: _, - only: _, identity: _, cascade: _, on_cluster: _, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 677566c56..7c9be1985 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -960,12 +960,13 @@ impl<'a> Parser<'a> { pub fn parse_truncate(&mut self) -> Result { let table = self.parse_keyword(Keyword::TABLE); - let only = self.parse_keyword(Keyword::ONLY); let table_names = self - .parse_comma_separated(|p| p.parse_object_name(false))? + .parse_comma_separated(|p| { + Ok((p.parse_keyword(Keyword::ONLY), p.parse_object_name(false)?)) + })? .into_iter() - .map(|n| TruncateTableTarget { name: n }) + .map(|(only, name)| TruncateTableTarget { name, only }) .collect(); let mut partitions = None; @@ -996,7 +997,6 @@ impl<'a> Parser<'a> { table_names, partitions, table, - only, identity, cascade, on_cluster, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 0431c720d..2cb51de38 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15302,3 +15302,31 @@ fn test_open() { }) ); } + +#[test] +fn parse_truncate_only() { + let truncate = all_dialects().verified_stmt("TRUNCATE TABLE employee, ONLY dept"); + + let table_names = vec![ + TruncateTableTarget { + name: ObjectName::from(vec![Ident::new("employee")]), + only: false, + }, + TruncateTableTarget { + name: ObjectName::from(vec![Ident::new("dept")]), + only: true, + }, + ]; + + assert_eq!( + Statement::Truncate { + table_names, + partitions: None, + table: true, + identity: None, + cascade: None, + on_cluster: None, + }, + truncate + ); +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index d54d4e2af..c50f066a6 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -4788,13 +4788,13 @@ fn parse_truncate() { let table_name = ObjectName::from(vec![Ident::new("db"), Ident::new("table_name")]); let table_names = vec![TruncateTableTarget { name: table_name.clone(), + only: false, }]; assert_eq!( Statement::Truncate { table_names, partitions: None, table: false, - only: false, identity: None, cascade: None, on_cluster: None, @@ -4811,6 +4811,7 @@ fn parse_truncate_with_options() { let table_name = ObjectName::from(vec![Ident::new("db"), Ident::new("table_name")]); let table_names = vec![TruncateTableTarget { name: table_name.clone(), + only: true, }]; assert_eq!( @@ -4818,7 +4819,6 @@ fn parse_truncate_with_options() { table_names, partitions: None, table: true, - only: true, identity: Some(TruncateIdentityOption::Restart), cascade: Some(CascadeOption::Cascade), on_cluster: None, @@ -4839,9 +4839,11 @@ fn parse_truncate_with_table_list() { let table_names = vec![ TruncateTableTarget { name: table_name_a.clone(), + only: false, }, TruncateTableTarget { name: table_name_b.clone(), + only: false, }, ]; @@ -4850,7 +4852,6 @@ fn parse_truncate_with_table_list() { table_names, partitions: None, table: true, - only: false, identity: Some(TruncateIdentityOption::Restart), cascade: Some(CascadeOption::Cascade), on_cluster: None, From e2b1ae36e94a8cb6d99234a2d9a25f56a3501f30 Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Fri, 6 Jun 2025 15:11:44 +0800 Subject: [PATCH 145/286] feat: Hive: support `SORT BY` direction (#1873) --- src/ast/query.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_hive.rs | 3 +++ 3 files changed, 5 insertions(+), 2 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index ffe1e4023..4398531cb 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -341,7 +341,7 @@ pub struct Select { /// DISTRIBUTE BY (Hive) pub distribute_by: Vec, /// SORT BY (Hive) - pub sort_by: Vec, + pub sort_by: Vec, /// HAVING pub having: Option, /// WINDOW AS diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 7c9be1985..821ce8439 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11591,7 +11591,7 @@ impl<'a> Parser<'a> { }; let sort_by = if self.parse_keywords(&[Keyword::SORT, Keyword::BY]) { - self.parse_comma_separated(Parser::parse_expr)? + self.parse_comma_separated(Parser::parse_order_by_expr)? } else { vec![] }; diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index 14dcbffd1..fd52b7730 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -341,6 +341,9 @@ fn lateral_view() { fn sort_by() { let sort_by = "SELECT * FROM db.table SORT BY a"; hive().verified_stmt(sort_by); + + let sort_by_with_direction = "SELECT * FROM db.table SORT BY a, b DESC"; + hive().verified_stmt(sort_by_with_direction); } #[test] From ff29dd25b2b8f4b10dd958f0c601f98d1c51ab36 Mon Sep 17 00:00:00 2001 From: Elia Perantoni Date: Fri, 6 Jun 2025 16:06:33 +0200 Subject: [PATCH 146/286] Fix `CASE` expression spans (#1874) --- src/ast/mod.rs | 4 ++++ src/ast/spans.rs | 34 ++++++++++++++++++++++++++-------- src/parser/mod.rs | 5 ++++- tests/sqlparser_common.rs | 6 ++++++ tests/sqlparser_databricks.rs | 3 +++ 5 files changed, 43 insertions(+), 9 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 3ccce0619..6f47ae7f7 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -967,6 +967,8 @@ pub enum Expr { /// not `< 0` nor `1, 2, 3` as allowed in a `` per /// Case { + case_token: AttachedToken, + end_token: AttachedToken, operand: Option>, conditions: Vec, else_result: Option>, @@ -1675,6 +1677,8 @@ impl fmt::Display for Expr { } Expr::Function(fun) => fun.fmt(f), Expr::Case { + case_token: _, + end_token: _, operand, conditions, else_result, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 11986f8c4..f957194ae 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1567,18 +1567,24 @@ impl Spanned for Expr { ), Expr::Prefixed { value, .. } => value.span(), Expr::Case { + case_token, + end_token, operand, conditions, else_result, } => union_spans( - operand - .as_ref() - .map(|i| i.span()) - .into_iter() - .chain(conditions.iter().flat_map(|case_when| { - [case_when.condition.span(), case_when.result.span()] - })) - .chain(else_result.as_ref().map(|i| i.span())), + iter::once(case_token.0.span) + .chain( + operand + .as_ref() + .map(|i| i.span()) + .into_iter() + .chain(conditions.iter().flat_map(|case_when| { + [case_when.condition.span(), case_when.result.span()] + })) + .chain(else_result.as_ref().map(|i| i.span())), + ) + .chain(iter::once(end_token.0.span)), ), Expr::Exists { subquery, .. } => subquery.span(), Expr::Subquery(query) => query.span(), @@ -2464,4 +2470,16 @@ pub mod tests { assert_eq!(test.get_source(body_span), "SELECT cte.* FROM cte"); } + + #[test] + fn test_case_expr_span() { + let dialect = &GenericDialect; + let mut test = SpanTest::new(dialect, "CASE 1 WHEN 2 THEN 3 ELSE 4 END"); + let expr = test.0.parse_expr().unwrap(); + let expr_span = expr.span(); + assert_eq!( + test.get_source(expr_span), + "CASE 1 WHEN 2 THEN 3 ELSE 4 END" + ); + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 821ce8439..f8c307dc9 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2274,6 +2274,7 @@ impl<'a> Parser<'a> { } pub fn parse_case_expr(&mut self) -> Result { + let case_token = AttachedToken(self.get_current_token().clone()); let mut operand = None; if !self.parse_keyword(Keyword::WHEN) { operand = Some(Box::new(self.parse_expr()?)); @@ -2294,8 +2295,10 @@ impl<'a> Parser<'a> { } else { None }; - self.expect_keyword_is(Keyword::END)?; + let end_token = AttachedToken(self.expect_keyword(Keyword::END)?); Ok(Expr::Case { + case_token, + end_token, operand, conditions, else_result, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 2cb51de38..5b96dcd7f 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -6869,6 +6869,8 @@ fn parse_searched_case_expr() { let select = verified_only_select(sql); assert_eq!( &Case { + case_token: AttachedToken::empty(), + end_token: AttachedToken::empty(), operand: None, conditions: vec![ CaseWhen { @@ -6908,6 +6910,8 @@ fn parse_simple_case_expr() { use self::Expr::{Case, Identifier}; assert_eq!( &Case { + case_token: AttachedToken::empty(), + end_token: AttachedToken::empty(), operand: Some(Box::new(Identifier(Ident::new("foo")))), conditions: vec![CaseWhen { condition: Expr::value(number("1")), @@ -14650,6 +14654,8 @@ fn test_lambdas() { Expr::Lambda(LambdaFunction { params: OneOrManyWithParens::Many(vec![Ident::new("p1"), Ident::new("p2")]), body: Box::new(Expr::Case { + case_token: AttachedToken::empty(), + end_token: AttachedToken::empty(), operand: None, conditions: vec![ CaseWhen { diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 88aae499a..99b7eecde 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -15,6 +15,7 @@ // specific language governing permissions and limitations // under the License. +use sqlparser::ast::helpers::attached_token::AttachedToken; use sqlparser::ast::*; use sqlparser::dialect::{DatabricksDialect, GenericDialect}; use sqlparser::parser::ParserError; @@ -108,6 +109,8 @@ fn test_databricks_lambdas() { Expr::Lambda(LambdaFunction { params: OneOrManyWithParens::Many(vec![Ident::new("p1"), Ident::new("p2")]), body: Box::new(Expr::Case { + case_token: AttachedToken::empty(), + end_token: AttachedToken::empty(), operand: None, conditions: vec![ CaseWhen { From 84c3a1b325c39c879b68ab712e3b9b3e3e40ed56 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Sat, 7 Jun 2025 05:48:40 +0100 Subject: [PATCH 147/286] MySQL: `[[NOT] ENFORCED]` in CHECK constraint (#1870) --- src/ast/ddl.rs | 18 +++++++++++++++--- src/ast/spans.rs | 8 +++++--- src/parser/mod.rs | 15 ++++++++++++++- tests/sqlparser_common.rs | 7 +++++++ tests/sqlparser_postgres.rs | 5 +++++ 5 files changed, 46 insertions(+), 7 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index bbc15704e..b0a3708c1 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1029,10 +1029,13 @@ pub enum TableConstraint { on_update: Option, characteristics: Option, }, - /// `[ CONSTRAINT ] CHECK ()` + /// `[ CONSTRAINT ] CHECK () [[NOT] ENFORCED]` Check { name: Option, expr: Box, + /// MySQL-specific syntax + /// + enforced: Option, }, /// MySQLs [index definition][1] for index creation. Not present on ANSI so, for now, the usage /// is restricted to MySQL, as no other dialects that support this syntax were found. @@ -1162,8 +1165,17 @@ impl fmt::Display for TableConstraint { } Ok(()) } - TableConstraint::Check { name, expr } => { - write!(f, "{}CHECK ({})", display_constraint_name(name), expr) + TableConstraint::Check { + name, + expr, + enforced, + } => { + write!(f, "{}CHECK ({})", display_constraint_name(name), expr)?; + if let Some(b) = enforced { + write!(f, " {}", if *b { "ENFORCED" } else { "NOT ENFORCED" }) + } else { + Ok(()) + } } TableConstraint::Index { display_as_key, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index f957194ae..39d30df95 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -687,9 +687,11 @@ impl Spanned for TableConstraint { .chain(on_update.iter().map(|i| i.span())) .chain(characteristics.iter().map(|i| i.span())), ), - TableConstraint::Check { name, expr } => { - expr.span().union_opt(&name.as_ref().map(|i| i.span)) - } + TableConstraint::Check { + name, + expr, + enforced: _, + } => expr.span().union_opt(&name.as_ref().map(|i| i.span)), TableConstraint::Index { display_as_key: _, name, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f8c307dc9..9cef22edb 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8139,7 +8139,20 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; let expr = Box::new(self.parse_expr()?); self.expect_token(&Token::RParen)?; - Ok(Some(TableConstraint::Check { name, expr })) + + let enforced = if self.parse_keyword(Keyword::ENFORCED) { + Some(true) + } else if self.parse_keywords(&[Keyword::NOT, Keyword::ENFORCED]) { + Some(false) + } else { + None + }; + + Ok(Some(TableConstraint::Check { + name, + expr, + enforced, + })) } Token::Word(w) if (w.keyword == Keyword::INDEX || w.keyword == Keyword::KEY) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5b96dcd7f..399fdb3dc 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15336,3 +15336,10 @@ fn parse_truncate_only() { truncate ); } + +#[test] +fn check_enforced() { + all_dialects().verified_stmt( + "CREATE TABLE t (a INT, b INT, c INT, CHECK (a > 0) NOT ENFORCED, CHECK (b > 0) ENFORCED, CHECK (c > 0))", + ); +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index c50f066a6..6f0ba9c69 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5378,6 +5378,7 @@ fn parse_create_domain() { op: BinaryOperator::Gt, right: Box::new(Expr::Value(test_utils::number("0").into())), }), + enforced: None, }], }); @@ -5396,6 +5397,7 @@ fn parse_create_domain() { op: BinaryOperator::Gt, right: Box::new(Expr::Value(test_utils::number("0").into())), }), + enforced: None, }], }); @@ -5414,6 +5416,7 @@ fn parse_create_domain() { op: BinaryOperator::Gt, right: Box::new(Expr::Value(test_utils::number("0").into())), }), + enforced: None, }], }); @@ -5432,6 +5435,7 @@ fn parse_create_domain() { op: BinaryOperator::Gt, right: Box::new(Expr::Value(test_utils::number("0").into())), }), + enforced: None, }], }); @@ -5450,6 +5454,7 @@ fn parse_create_domain() { op: BinaryOperator::Gt, right: Box::new(Expr::Value(test_utils::number("0").into())), }), + enforced: None, }], }); From 40d12b98bd195eecf8f1b89dcfd9d163c96fd0af Mon Sep 17 00:00:00 2001 From: Yannick Utard Date: Tue, 10 Jun 2025 06:50:10 +0200 Subject: [PATCH 148/286] Add support for `CREATE SCHEMA WITH ( )` (#1877) --- src/ast/mod.rs | 13 +++++++++++++ src/parser/mod.rs | 7 +++++++ tests/sqlparser_common.rs | 3 +++ 3 files changed, 23 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 6f47ae7f7..a1d8ff6fc 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3725,6 +3725,14 @@ pub enum Statement { /// ` | AUTHORIZATION | AUTHORIZATION ` schema_name: SchemaName, if_not_exists: bool, + /// Schema properties. + /// + /// ```sql + /// CREATE SCHEMA myschema WITH (key1='value1'); + /// ``` + /// + /// [Trino](https://trino.io/docs/current/sql/create-schema.html) + with: Option>, /// Schema options. /// /// ```sql @@ -5585,6 +5593,7 @@ impl fmt::Display for Statement { Statement::CreateSchema { schema_name, if_not_exists, + with, options, default_collate_spec, } => { @@ -5599,6 +5608,10 @@ impl fmt::Display for Statement { write!(f, " DEFAULT COLLATE {collate}")?; } + if let Some(with) = with { + write!(f, " WITH ({})", display_comma_separated(with))?; + } + if let Some(options) = options { write!(f, " OPTIONS({})", display_comma_separated(options))?; } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9cef22edb..b582e7934 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4862,6 +4862,12 @@ impl<'a> Parser<'a> { None }; + let with = if self.peek_keyword(Keyword::WITH) { + Some(self.parse_options(Keyword::WITH)?) + } else { + None + }; + let options = if self.peek_keyword(Keyword::OPTIONS) { Some(self.parse_options(Keyword::OPTIONS)?) } else { @@ -4871,6 +4877,7 @@ impl<'a> Parser<'a> { Ok(Statement::CreateSchema { schema_name, if_not_exists, + with, options, default_collate_spec, }) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 399fdb3dc..abcadb458 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4268,6 +4268,9 @@ fn parse_create_schema() { verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a OPTIONS(key1 = 'value1')"#); verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a OPTIONS()"#); verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a DEFAULT COLLATE 'und:ci' OPTIONS()"#); + verified_stmt(r#"CREATE SCHEMA a.b.c WITH (key1 = 'value1', key2 = 'value2')"#); + verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a WITH (key1 = 'value1')"#); + verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a WITH ()"#); } #[test] From 559b7e36d9bafbc6da9e3972136ccfae5c22aeb8 Mon Sep 17 00:00:00 2001 From: vimko <1399950+vimko@users.noreply.github.com> Date: Wed, 11 Jun 2025 00:26:07 +0800 Subject: [PATCH 149/286] Add support for `ALTER TABLE DROP INDEX` (#1865) --- src/ast/ddl.rs | 7 +++++++ src/ast/spans.rs | 1 + src/parser/mod.rs | 3 +++ tests/sqlparser_mysql.rs | 10 ++++++++++ 4 files changed, 21 insertions(+) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index b0a3708c1..059c61967 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -187,6 +187,12 @@ pub enum AlterTableOperation { DropForeignKey { name: Ident, }, + /// `DROP INDEX ` + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html + DropIndex { + name: Ident, + }, /// `ENABLE ALWAYS RULE rewrite_rule_name` /// /// Note: this is a PostgreSQL-specific operation. @@ -606,6 +612,7 @@ impl fmt::Display for AlterTableOperation { } AlterTableOperation::DropPrimaryKey => write!(f, "DROP PRIMARY KEY"), AlterTableOperation::DropForeignKey { name } => write!(f, "DROP FOREIGN KEY {name}"), + AlterTableOperation::DropIndex { name } => write!(f, "DROP INDEX {name}"), AlterTableOperation::DropColumn { has_column_keyword, column_name, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 39d30df95..14664b4cd 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1115,6 +1115,7 @@ impl Spanned for AlterTableOperation { .union_opt(&with_name.as_ref().map(|n| n.span)), AlterTableOperation::DropPrimaryKey => Span::empty(), AlterTableOperation::DropForeignKey { name } => name.span, + AlterTableOperation::DropIndex { name } => name.span, AlterTableOperation::EnableAlwaysRule { name } => name.span, AlterTableOperation::EnableAlwaysTrigger { name } => name.span, AlterTableOperation::EnableReplicaRule { name } => name.span, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b582e7934..6831d52e0 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8636,6 +8636,9 @@ impl<'a> Parser<'a> { } else if self.parse_keywords(&[Keyword::FOREIGN, Keyword::KEY]) { let name = self.parse_identifier()?; AlterTableOperation::DropForeignKey { name } + } else if self.parse_keyword(Keyword::INDEX) { + let name = self.parse_identifier()?; + AlterTableOperation::DropIndex { name } } else if self.parse_keyword(Keyword::PROJECTION) && dialect_of!(self is ClickHouseDialect|GenericDialect) { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index b1b7d539e..540348ff3 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -4025,3 +4025,13 @@ fn parse_drop_index() { _ => unreachable!(), } } + +#[test] +fn parse_alter_table_drop_index() { + assert_matches!( + alter_table_op( + mysql_and_generic().verified_stmt("ALTER TABLE tab DROP INDEX idx_index") + ), + AlterTableOperation::DropIndex { name } if name.value == "idx_index" + ); +} From 9fc9009b94c67f0b3314b7ca2b6b0a5f808c6027 Mon Sep 17 00:00:00 2001 From: Jacob Wujciak-Jens Date: Wed, 11 Jun 2025 01:54:21 +0200 Subject: [PATCH 150/286] chore: Replace archived actions-rs/install action (#1876) --- .github/workflows/rust.yml | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml index b5744e863..3abf9d387 100644 --- a/.github/workflows/rust.yml +++ b/.github/workflows/rust.yml @@ -19,6 +19,9 @@ name: Rust on: [push, pull_request] +permissions: + contents: read + jobs: codestyle: @@ -85,11 +88,8 @@ jobs: uses: ./.github/actions/setup-builder with: rust-version: ${{ matrix.rust }} + - uses: Swatinem/rust-cache@9d47c6ad4b02e050fd481d890b2ea34778fd09d6 # v2.7.8 - name: Install Tarpaulin - uses: actions-rs/install@v0.1 - with: - crate: cargo-tarpaulin - version: 0.14.2 - use-tool-cache: true + run: cargo install cargo-tarpaulin - name: Test run: cargo test --all-features From c2e83d49f6442bd93faa1abd011d4fac170bec7f Mon Sep 17 00:00:00 2001 From: Simon Sawert Date: Wed, 11 Jun 2025 18:11:07 +0200 Subject: [PATCH 151/286] Allow `IF NOT EXISTS` after table name for Snowflake (#1881) --- src/dialect/snowflake.rs | 3 +++ tests/sqlparser_snowflake.rs | 21 +++++++++++++++++++++ 2 files changed, 24 insertions(+) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index ccce16198..990e2ea29 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -560,6 +560,9 @@ pub fn parse_create_table( builder.storage_serialization_policy = Some(parse_storage_serialization_policy(parser)?); } + Keyword::IF if parser.parse_keywords(&[Keyword::NOT, Keyword::EXISTS]) => { + builder = builder.if_not_exists(true); + } _ => { return parser.expected("end of statement", next_token); } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index b4d62506d..7a164c248 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -446,6 +446,27 @@ fn test_snowflake_create_table_if_not_exists() { } _ => unreachable!(), } + + for (sql, parse_to) in [ + ( + r#"CREATE TABLE IF NOT EXISTS "A"."B"."C" (v VARIANT)"#, + r#"CREATE TABLE IF NOT EXISTS "A"."B"."C" (v VARIANT)"#, + ), + ( + r#"CREATE TABLE "A"."B"."C" IF NOT EXISTS (v VARIANT)"#, + r#"CREATE TABLE IF NOT EXISTS "A"."B"."C" (v VARIANT)"#, + ), + ( + r#"CREATE TRANSIENT TABLE IF NOT EXISTS "A"."B"."C" (v VARIANT)"#, + r#"CREATE TRANSIENT TABLE IF NOT EXISTS "A"."B"."C" (v VARIANT)"#, + ), + ( + r#"CREATE TRANSIENT TABLE "A"."B"."C" IF NOT EXISTS (v VARIANT)"#, + r#"CREATE TRANSIENT TABLE IF NOT EXISTS "A"."B"."C" (v VARIANT)"#, + ), + ] { + snowflake().one_statement_parses_to(sql, parse_to); + } } #[test] From 703ba2cf4867c3c7c768836fdacb0193eb5d0bca Mon Sep 17 00:00:00 2001 From: Simon Sawert Date: Wed, 11 Jun 2025 18:12:30 +0200 Subject: [PATCH 152/286] Support `DISTINCT AS { STRUCT | VALUE }` for BigQuery (#1880) --- src/ast/query.rs | 10 ++++++++-- src/parser/mod.rs | 39 +++++++++++++++++++++++++------------ tests/sqlparser_bigquery.rs | 38 ++++++++++++++++++++++++++++++++---- 3 files changed, 69 insertions(+), 18 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 4398531cb..1fb93b6c6 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -3338,15 +3338,19 @@ impl fmt::Display for OpenJsonTableColumn { } /// BigQuery supports ValueTables which have 2 modes: -/// `SELECT AS STRUCT` -/// `SELECT AS VALUE` +/// `SELECT [ALL | DISTINCT] AS STRUCT` +/// `SELECT [ALL | DISTINCT] AS VALUE` +/// /// +/// #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum ValueTableMode { AsStruct, AsValue, + DistinctAsStruct, + DistinctAsValue, } impl fmt::Display for ValueTableMode { @@ -3354,6 +3358,8 @@ impl fmt::Display for ValueTableMode { match self { ValueTableMode::AsStruct => write!(f, "AS STRUCT"), ValueTableMode::AsValue => write!(f, "AS VALUE"), + ValueTableMode::DistinctAsStruct => write!(f, "DISTINCT AS STRUCT"), + ValueTableMode::DistinctAsValue => write!(f, "DISTINCT AS VALUE"), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6831d52e0..2c208e2e5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11505,18 +11505,7 @@ impl<'a> Parser<'a> { } let select_token = self.expect_keyword(Keyword::SELECT)?; - let value_table_mode = - if dialect_of!(self is BigQueryDialect) && self.parse_keyword(Keyword::AS) { - if self.parse_keyword(Keyword::VALUE) { - Some(ValueTableMode::AsValue) - } else if self.parse_keyword(Keyword::STRUCT) { - Some(ValueTableMode::AsStruct) - } else { - self.expected("VALUE or STRUCT", self.peek_token())? - } - } else { - None - }; + let value_table_mode = self.parse_value_table_mode()?; let mut top_before_distinct = false; let mut top = None; @@ -11692,6 +11681,32 @@ impl<'a> Parser<'a> { }) } + fn parse_value_table_mode(&mut self) -> Result, ParserError> { + if !dialect_of!(self is BigQueryDialect) { + return Ok(None); + } + + let mode = if self.parse_keywords(&[Keyword::DISTINCT, Keyword::AS, Keyword::VALUE]) { + Some(ValueTableMode::DistinctAsValue) + } else if self.parse_keywords(&[Keyword::DISTINCT, Keyword::AS, Keyword::STRUCT]) { + Some(ValueTableMode::DistinctAsStruct) + } else if self.parse_keywords(&[Keyword::AS, Keyword::VALUE]) + || self.parse_keywords(&[Keyword::ALL, Keyword::AS, Keyword::VALUE]) + { + Some(ValueTableMode::AsValue) + } else if self.parse_keywords(&[Keyword::AS, Keyword::STRUCT]) + || self.parse_keywords(&[Keyword::ALL, Keyword::AS, Keyword::STRUCT]) + { + Some(ValueTableMode::AsStruct) + } else if self.parse_keyword(Keyword::AS) { + self.expected("VALUE or STRUCT", self.peek_token())? + } else { + None + }; + + Ok(mode) + } + /// Invoke `f` after first setting the parser's `ParserState` to `state`. /// /// Upon return, restores the parser's state to what it started at. diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 8f54f3c97..b64f190f6 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -2313,16 +2313,46 @@ fn bigquery_select_expr_star() { #[test] fn test_select_as_struct() { - bigquery().verified_only_select("SELECT * FROM (SELECT AS VALUE STRUCT(123 AS a, false AS b))"); + for (sql, parse_to) in [ + ( + "SELECT * FROM (SELECT AS STRUCT STRUCT(123 AS a, false AS b))", + "SELECT * FROM (SELECT AS STRUCT STRUCT(123 AS a, false AS b))", + ), + ( + "SELECT * FROM (SELECT DISTINCT AS STRUCT STRUCT(123 AS a, false AS b))", + "SELECT * FROM (SELECT DISTINCT AS STRUCT STRUCT(123 AS a, false AS b))", + ), + ( + "SELECT * FROM (SELECT ALL AS STRUCT STRUCT(123 AS a, false AS b))", + "SELECT * FROM (SELECT AS STRUCT STRUCT(123 AS a, false AS b))", + ), + ] { + bigquery().one_statement_parses_to(sql, parse_to); + } + let select = bigquery().verified_only_select("SELECT AS STRUCT 1 AS a, 2 AS b"); assert_eq!(Some(ValueTableMode::AsStruct), select.value_table_mode); } #[test] fn test_select_as_value() { - bigquery().verified_only_select( - "SELECT * FROM (SELECT AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", - ); + for (sql, parse_to) in [ + ( + "SELECT * FROM (SELECT AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", + "SELECT * FROM (SELECT AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", + ), + ( + "SELECT * FROM (SELECT DISTINCT AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", + "SELECT * FROM (SELECT DISTINCT AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", + ), + ( + "SELECT * FROM (SELECT ALL AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", + "SELECT * FROM (SELECT AS VALUE STRUCT(5 AS star_rating, false AS up_down_rating))", + ), + ] { + bigquery().one_statement_parses_to(sql, parse_to); + } + let select = bigquery().verified_only_select("SELECT AS VALUE STRUCT(1 AS a, 2 AS b) AS xyz"); assert_eq!(Some(ValueTableMode::AsValue), select.value_table_mode); } From 0f2208d293d4eb8c9ca7502881377609aaf0a5dc Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Sat, 14 Jun 2025 00:45:34 -0400 Subject: [PATCH 153/286] Prepare for `0.57.0` release (#1885) --- Cargo.toml | 2 +- changelog/0.57.0.md | 95 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 96 insertions(+), 1 deletion(-) create mode 100644 changelog/0.57.0.md diff --git a/Cargo.toml b/Cargo.toml index d746775e4..07e44f66b 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.56.0" +version = "0.57.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/changelog/0.57.0.md b/changelog/0.57.0.md new file mode 100644 index 000000000..200bb73af --- /dev/null +++ b/changelog/0.57.0.md @@ -0,0 +1,95 @@ + + +# sqlparser-rs 0.57.0 Changelog + +This release consists of 39 commits from 19 contributors. See credits at the end of this changelog for more information. + +**Implemented enhancements:** + +- feat: Hive: support `SORT BY` direction [#1873](https://github.com/apache/datafusion-sqlparser-rs/pull/1873) (chenkovsky) + +**Other:** + +- Support some of pipe operators [#1759](https://github.com/apache/datafusion-sqlparser-rs/pull/1759) (simonvandel) +- Added support for `DROP DOMAIN` [#1828](https://github.com/apache/datafusion-sqlparser-rs/pull/1828) (LucaCappelletti94) +- Improve support for cursors for SQL Server [#1831](https://github.com/apache/datafusion-sqlparser-rs/pull/1831) (aharpervc) +- Add all missing table options to be handled in any order [#1747](https://github.com/apache/datafusion-sqlparser-rs/pull/1747) (benrsatori) +- Add `CREATE TRIGGER` support for SQL Server [#1810](https://github.com/apache/datafusion-sqlparser-rs/pull/1810) (aharpervc) +- Added support for `CREATE DOMAIN` [#1830](https://github.com/apache/datafusion-sqlparser-rs/pull/1830) (LucaCappelletti94) +- Allow stored procedures to be defined without `BEGIN`/`END` [#1834](https://github.com/apache/datafusion-sqlparser-rs/pull/1834) (aharpervc) +- Add support for the MATCH and REGEXP binary operators [#1840](https://github.com/apache/datafusion-sqlparser-rs/pull/1840) (lovasoa) +- Fix: parsing ident starting with underscore in certain dialects [#1835](https://github.com/apache/datafusion-sqlparser-rs/pull/1835) (MohamedAbdeen21) +- implement pretty-printing with `{:#}` [#1847](https://github.com/apache/datafusion-sqlparser-rs/pull/1847) (lovasoa) +- Fix big performance issue in string serialization [#1848](https://github.com/apache/datafusion-sqlparser-rs/pull/1848) (lovasoa) +- Add support for `DENY` statements [#1836](https://github.com/apache/datafusion-sqlparser-rs/pull/1836) (aharpervc) +- Postgresql: Add `REPLICA IDENTITY` operation for `ALTER TABLE` [#1844](https://github.com/apache/datafusion-sqlparser-rs/pull/1844) (MohamedAbdeen21) +- Add support for INCLUDE/EXCLUDE NULLS for UNPIVOT [#1849](https://github.com/apache/datafusion-sqlparser-rs/pull/1849) (Vedin) +- pretty print improvements [#1851](https://github.com/apache/datafusion-sqlparser-rs/pull/1851) (lovasoa) +- fix new rust 1.87 cargo clippy warnings [#1856](https://github.com/apache/datafusion-sqlparser-rs/pull/1856) (lovasoa) +- Update criterion requirement from 0.5 to 0.6 in /sqlparser_bench [#1857](https://github.com/apache/datafusion-sqlparser-rs/pull/1857) (dependabot[bot]) +- pretty-print CREATE TABLE statements [#1854](https://github.com/apache/datafusion-sqlparser-rs/pull/1854) (lovasoa) +- pretty-print CREATE VIEW statements [#1855](https://github.com/apache/datafusion-sqlparser-rs/pull/1855) (lovasoa) +- Handle optional datatypes properly in `CREATE FUNCTION` statements [#1826](https://github.com/apache/datafusion-sqlparser-rs/pull/1826) (LucaCappelletti94) +- Mysql: Add `SRID` column option [#1852](https://github.com/apache/datafusion-sqlparser-rs/pull/1852) (MohamedAbdeen21) +- Add support for table valued functions for SQL Server [#1839](https://github.com/apache/datafusion-sqlparser-rs/pull/1839) (aharpervc) +- Keep the COLUMN keyword only if it exists when dropping the column [#1862](https://github.com/apache/datafusion-sqlparser-rs/pull/1862) (git-hulk) +- Add support for parameter default values in SQL Server [#1866](https://github.com/apache/datafusion-sqlparser-rs/pull/1866) (aharpervc) +- Add support for `TABLESAMPLE` pipe operator [#1860](https://github.com/apache/datafusion-sqlparser-rs/pull/1860) (hendrikmakait) +- Adds support for mysql's drop index [#1864](https://github.com/apache/datafusion-sqlparser-rs/pull/1864) (dmzmk) +- Fix: GROUPING SETS accept values without parenthesis [#1867](https://github.com/apache/datafusion-sqlparser-rs/pull/1867) (Vedin) +- Add ICEBERG keyword support to ALTER TABLE statement [#1869](https://github.com/apache/datafusion-sqlparser-rs/pull/1869) (osipovartem) +- MySQL: Support `index_name` in FK constraints [#1871](https://github.com/apache/datafusion-sqlparser-rs/pull/1871) (MohamedAbdeen21) +- Postgres: Apply `ONLY` keyword per table in TRUNCATE stmt [#1872](https://github.com/apache/datafusion-sqlparser-rs/pull/1872) (MohamedAbdeen21) +- Fix `CASE` expression spans [#1874](https://github.com/apache/datafusion-sqlparser-rs/pull/1874) (eliaperantoni) +- MySQL: `[[NOT] ENFORCED]` in CHECK constraint [#1870](https://github.com/apache/datafusion-sqlparser-rs/pull/1870) (MohamedAbdeen21) +- Add support for `CREATE SCHEMA WITH ( )` [#1877](https://github.com/apache/datafusion-sqlparser-rs/pull/1877) (utay) +- Add support for `ALTER TABLE DROP INDEX` [#1865](https://github.com/apache/datafusion-sqlparser-rs/pull/1865) (vimko) +- chore: Replace archived actions-rs/install action [#1876](https://github.com/apache/datafusion-sqlparser-rs/pull/1876) (assignUser) +- Allow `IF NOT EXISTS` after table name for Snowflake [#1881](https://github.com/apache/datafusion-sqlparser-rs/pull/1881) (bombsimon) +- Support `DISTINCT AS { STRUCT | VALUE }` for BigQuery [#1880](https://github.com/apache/datafusion-sqlparser-rs/pull/1880) (bombsimon) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 7 Ophir LOJKINE + 6 Andrew Harper + 6 Mohamed Abdeen + 3 Luca Cappelletti + 2 Denys Tsomenko + 2 Simon Sawert + 1 Andrew Lamb + 1 Artem Osipov + 1 Chen Chongchen + 1 Dmitriy Mazurin + 1 Elia Perantoni + 1 Hendrik Makait + 1 Jacob Wujciak-Jens + 1 Simon Vandel Sillesen + 1 Yannick Utard + 1 benrsatori + 1 dependabot[bot] + 1 hulk + 1 vimko +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + From e406422bacb6a8d1a09f053faaa1f9a063dc0f40 Mon Sep 17 00:00:00 2001 From: Artem Osipov <59066880+osipovartem@users.noreply.github.com> Date: Mon, 16 Jun 2025 13:33:16 +0300 Subject: [PATCH 154/286] Add support for cluster by expressions (#1883) Co-authored-by: osipovartem --- src/ast/dml.rs | 4 +++- src/ast/helpers/stmt_create_table.rs | 6 +++--- src/dialect/snowflake.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_bigquery.rs | 4 ++-- tests/sqlparser_snowflake.rs | 26 +++++++++++++++++++++----- 6 files changed, 31 insertions(+), 13 deletions(-) diff --git a/src/ast/dml.rs b/src/ast/dml.rs index da82a4ede..292650c88 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -175,7 +175,9 @@ pub struct CreateTable { pub partition_by: Option>, /// BigQuery: Table clustering column list. /// - pub cluster_by: Option>>, + /// Snowflake: Table clustering list which contains base column, expressions on base columns. + /// + pub cluster_by: Option>>, /// Hive: Table clustering column list. /// pub clustered_by: Option, diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index 542d30ea9..d66a869bf 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -90,7 +90,7 @@ pub struct CreateTableBuilder { pub primary_key: Option>, pub order_by: Option>, pub partition_by: Option>, - pub cluster_by: Option>>, + pub cluster_by: Option>>, pub clustered_by: Option, pub inherits: Option>, pub strict: bool, @@ -279,7 +279,7 @@ impl CreateTableBuilder { self } - pub fn cluster_by(mut self, cluster_by: Option>>) -> Self { + pub fn cluster_by(mut self, cluster_by: Option>>) -> Self { self.cluster_by = cluster_by; self } @@ -542,7 +542,7 @@ impl TryFrom for CreateTableBuilder { #[derive(Default)] pub(crate) struct CreateTableConfiguration { pub partition_by: Option>, - pub cluster_by: Option>>, + pub cluster_by: Option>>, pub inherits: Option>, pub table_options: CreateTableOptions, } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 990e2ea29..ea0b94a63 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -453,7 +453,7 @@ pub fn parse_create_table( parser.expect_keyword_is(Keyword::BY)?; parser.expect_token(&Token::LParen)?; let cluster_by = Some(WrappedCollection::Parentheses( - parser.parse_comma_separated(|p| p.parse_identifier())?, + parser.parse_comma_separated(|p| p.parse_expr())?, )); parser.expect_token(&Token::RParen)?; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2c208e2e5..73cc3e0ed 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7316,7 +7316,7 @@ impl<'a> Parser<'a> { if dialect_of!(self is BigQueryDialect | GenericDialect) { if self.parse_keywords(&[Keyword::CLUSTER, Keyword::BY]) { cluster_by = Some(WrappedCollection::NoWrapping( - self.parse_comma_separated(|p| p.parse_identifier())?, + self.parse_comma_separated(|p| p.parse_expr())?, )); }; diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index b64f190f6..6a303577f 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -536,8 +536,8 @@ fn parse_create_table_with_options() { ( Some(Box::new(Expr::Identifier(Ident::new("_PARTITIONDATE")))), Some(WrappedCollection::NoWrapping(vec![ - Ident::new("userid"), - Ident::new("age"), + Expr::Identifier(Ident::new("userid")), + Expr::Identifier(Ident::new("age")), ])), CreateTableOptions::Options(vec![ SqlOption::KeyValue { diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 7a164c248..7d734ca26 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -471,15 +471,31 @@ fn test_snowflake_create_table_if_not_exists() { #[test] fn test_snowflake_create_table_cluster_by() { - match snowflake().verified_stmt("CREATE TABLE my_table (a INT) CLUSTER BY (a, b)") { + match snowflake().verified_stmt("CREATE TABLE my_table (a INT) CLUSTER BY (a, b, my_func(c))") { Statement::CreateTable(CreateTable { name, cluster_by, .. }) => { assert_eq!("my_table", name.to_string()); assert_eq!( Some(WrappedCollection::Parentheses(vec![ - Ident::new("a"), - Ident::new("b"), + Expr::Identifier(Ident::new("a")), + Expr::Identifier(Ident::new("b")), + Expr::Function(Function { + name: ObjectName::from(vec![Ident::new("my_func")]), + uses_odbc_syntax: false, + parameters: FunctionArguments::None, + args: FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr( + Expr::Identifier(Ident::new("c")) + ))], + duplicate_treatment: None, + clauses: vec![], + }), + filter: None, + null_treatment: None, + over: None, + within_group: vec![], + }), ])), cluster_by ) @@ -903,8 +919,8 @@ fn test_snowflake_create_iceberg_table_all_options() { assert_eq!("my_table", name.to_string()); assert_eq!( Some(WrappedCollection::Parentheses(vec![ - Ident::new("a"), - Ident::new("b"), + Expr::Identifier(Ident::new("a")), + Expr::Identifier(Ident::new("b")), ])), cluster_by ); From 6f423969b0e723eeded2e5b7c0587aefae91ec54 Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Mon, 16 Jun 2025 12:37:18 -0400 Subject: [PATCH 155/286] Add license header to display_utils.rs and pretty_print.rs (#1887) --- src/display_utils.rs | 17 +++++++++++++++++ tests/pretty_print.rs | 17 +++++++++++++++++ 2 files changed, 34 insertions(+) diff --git a/src/display_utils.rs b/src/display_utils.rs index e594a34ef..ba36fccdd 100644 --- a/src/display_utils.rs +++ b/src/display_utils.rs @@ -1,3 +1,20 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + //! Utilities for formatting SQL AST nodes with pretty printing support. //! //! The module provides formatters that implement the `Display` trait with support diff --git a/tests/pretty_print.rs b/tests/pretty_print.rs index e1d35eb04..f5a9d8613 100644 --- a/tests/pretty_print.rs +++ b/tests/pretty_print.rs @@ -1,3 +1,20 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + use sqlparser::dialect::GenericDialect; use sqlparser::parser::Parser; From be30697efb94cecdca749cc540c5a662c63fad3f Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Tue, 17 Jun 2025 10:46:58 -0400 Subject: [PATCH 156/286] Add license header check to CI (#1888) --- .github/workflows/license.yml | 39 +++++++++++++++++++++++++++++++ dev/release/rat_exclude_files.txt | 9 +++---- 2 files changed, 44 insertions(+), 4 deletions(-) create mode 100644 .github/workflows/license.yml diff --git a/.github/workflows/license.yml b/.github/workflows/license.yml new file mode 100644 index 000000000..c851bff31 --- /dev/null +++ b/.github/workflows/license.yml @@ -0,0 +1,39 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, +# software distributed under the License is distributed on an +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +# KIND, either express or implied. See the License for the +# specific language governing permissions and limitations +# under the License. + +name: license + +# trigger for all PRs and changes to main +on: + push: + branches: + - main + pull_request: + +jobs: + + rat: + name: Release Audit Tool (RAT) + runs-on: ubuntu-latest + steps: + - uses: actions/checkout@v4 + - name: Setup Python + uses: actions/setup-python@v5 + with: + python-version: 3.8 + - name: Audit licenses + run: ./dev/release/run-rat.sh . diff --git a/dev/release/rat_exclude_files.txt b/dev/release/rat_exclude_files.txt index 562eec2f1..280b1bce6 100644 --- a/dev/release/rat_exclude_files.txt +++ b/dev/release/rat_exclude_files.txt @@ -1,7 +1,8 @@ -# Files to exclude from the Apache Rat (license) check -.gitignore .tool-versions +target/* +**.gitignore +rat.txt dev/release/rat_exclude_files.txt -fuzz/.gitignore sqlparser_bench/img/flamegraph.svg - +**Cargo.lock +filtered_rat.txt From b1b379e57086ad7fcfa67136c0c64b14d4efab9d Mon Sep 17 00:00:00 2001 From: hulk Date: Wed, 18 Jun 2025 13:00:53 +0800 Subject: [PATCH 157/286] Add support of parsing struct field's options in BigQuery (#1890) Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 12 ++- src/parser/mod.rs | 3 + tests/sqlparser_bigquery.rs | 159 ++++++++++++++++++++++++---------- tests/sqlparser_clickhouse.rs | 10 ++- tests/sqlparser_duckdb.rs | 6 ++ 5 files changed, 138 insertions(+), 52 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index a1d8ff6fc..04401a48b 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -428,14 +428,22 @@ impl fmt::Display for Interval { pub struct StructField { pub field_name: Option, pub field_type: DataType, + /// Struct field options. + /// See [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#column_name_and_column_schema) + pub options: Option>, } impl fmt::Display for StructField { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { if let Some(name) = &self.field_name { - write!(f, "{name} {}", self.field_type) + write!(f, "{name} {}", self.field_type)?; } else { - write!(f, "{}", self.field_type) + write!(f, "{}", self.field_type)?; + } + if let Some(options) = &self.options { + write!(f, " OPTIONS({})", display_separated(options, ", ")) + } else { + Ok(()) } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 73cc3e0ed..19c11d296 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3076,6 +3076,7 @@ impl<'a> Parser<'a> { Ok(StructField { field_name: Some(field_name), field_type, + options: None, }) }); self.expect_token(&Token::RParen)?; @@ -3109,10 +3110,12 @@ impl<'a> Parser<'a> { let (field_type, trailing_bracket) = self.parse_data_type_helper()?; + let options = self.maybe_parse_options(Keyword::OPTIONS)?; Ok(( StructField { field_name, field_type, + options, }, trailing_bracket, )) diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 6a303577f..e37e4a449 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -601,11 +601,13 @@ fn parse_nested_data_types() { field_name: Some("a".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket( Box::new(DataType::Int64,) - )) + )), + options: None, }, StructField { field_name: Some("b".into()), - field_type: DataType::Bytes(Some(42)) + field_type: DataType::Bytes(Some(42)), + options: None, }, ], StructBracketKind::AngleBrackets @@ -619,6 +621,7 @@ fn parse_nested_data_types() { vec![StructField { field_name: None, field_type: DataType::Int64, + options: None, }], StructBracketKind::AngleBrackets ), @@ -771,6 +774,7 @@ fn parse_typed_struct_syntax_bigquery() { fields: vec![StructField { field_name: None, field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -799,7 +803,8 @@ fn parse_typed_struct_syntax_bigquery() { quote_style: None, span: Span::empty(), }), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident { @@ -807,7 +812,8 @@ fn parse_typed_struct_syntax_bigquery() { quote_style: None, span: Span::empty(), }), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }, ] }, @@ -825,17 +831,20 @@ fn parse_typed_struct_syntax_bigquery() { field_name: Some("arr".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( DataType::Float64 - ))) + ))), + options: None, }, StructField { field_name: Some("str".into()), field_type: DataType::Struct( vec![StructField { field_name: None, - field_type: DataType::Bool + field_type: DataType::Bool, + options: None, }], StructBracketKind::AngleBrackets - ) + ), + options: None, }, ] }, @@ -858,13 +867,15 @@ fn parse_typed_struct_syntax_bigquery() { field_type: DataType::Struct( Default::default(), StructBracketKind::AngleBrackets - ) + ), + options: None, }, StructField { field_name: Some("y".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( DataType::Struct(Default::default(), StructBracketKind::AngleBrackets) - ))) + ))), + options: None, }, ] }, @@ -879,7 +890,8 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::Value(Value::Boolean(true).with_empty_span())], fields: vec![StructField { field_name: None, - field_type: DataType::Bool + field_type: DataType::Bool, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -891,7 +903,8 @@ fn parse_typed_struct_syntax_bigquery() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Bytes(Some(42)) + field_type: DataType::Bytes(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -907,7 +920,8 @@ fn parse_typed_struct_syntax_bigquery() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Date + field_type: DataType::Date, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -920,7 +934,8 @@ fn parse_typed_struct_syntax_bigquery() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Datetime(None) + field_type: DataType::Datetime(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -930,7 +945,8 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::value(number("5.0"))], fields: vec![StructField { field_name: None, - field_type: DataType::Float64 + field_type: DataType::Float64, + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -940,7 +956,8 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::value(number("1"))], fields: vec![StructField { field_name: None, - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[3]) @@ -962,7 +979,8 @@ fn parse_typed_struct_syntax_bigquery() { })], fields: vec![StructField { field_name: None, - field_type: DataType::Interval + field_type: DataType::Interval, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -977,7 +995,8 @@ fn parse_typed_struct_syntax_bigquery() { }], fields: vec![StructField { field_name: None, - field_type: DataType::JSON + field_type: DataType::JSON, + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -993,7 +1012,8 @@ fn parse_typed_struct_syntax_bigquery() { )], fields: vec![StructField { field_name: None, - field_type: DataType::String(Some(42)) + field_type: DataType::String(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1006,7 +1026,8 @@ fn parse_typed_struct_syntax_bigquery() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Timestamp(None, TimezoneInfo::None) + field_type: DataType::Timestamp(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1020,7 +1041,8 @@ fn parse_typed_struct_syntax_bigquery() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Time(None, TimezoneInfo::None) + field_type: DataType::Time(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -1037,7 +1059,8 @@ fn parse_typed_struct_syntax_bigquery() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Numeric(ExactNumberInfo::None) + field_type: DataType::Numeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1050,7 +1073,8 @@ fn parse_typed_struct_syntax_bigquery() { }], fields: vec![StructField { field_name: None, - field_type: DataType::BigNumeric(ExactNumberInfo::None) + field_type: DataType::BigNumeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1067,10 +1091,12 @@ fn parse_typed_struct_syntax_bigquery() { StructField { field_name: Some("key".into()), field_type: DataType::Int64, + options: None, }, StructField { field_name: Some("value".into()), field_type: DataType::Int64, + options: None, }, ] }, @@ -1092,6 +1118,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { fields: vec![StructField { field_name: None, field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1120,7 +1147,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { quote_style: None, span: Span::empty(), }), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident { @@ -1128,7 +1156,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { quote_style: None, span: Span::empty(), }), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }, ] }, @@ -1151,13 +1180,15 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { field_type: DataType::Struct( Default::default(), StructBracketKind::AngleBrackets - ) + ), + options: None, }, StructField { field_name: Some("y".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( DataType::Struct(Default::default(), StructBracketKind::AngleBrackets) - ))) + ))), + options: None, }, ] }, @@ -1172,7 +1203,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::Value(Value::Boolean(true).with_empty_span())], fields: vec![StructField { field_name: None, - field_type: DataType::Bool + field_type: DataType::Bool, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1184,7 +1216,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Bytes(Some(42)) + field_type: DataType::Bytes(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1200,7 +1233,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Date + field_type: DataType::Date, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1213,7 +1247,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Datetime(None) + field_type: DataType::Datetime(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1223,7 +1258,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::value(number("5.0"))], fields: vec![StructField { field_name: None, - field_type: DataType::Float64 + field_type: DataType::Float64, + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -1233,7 +1269,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::value(number("1"))], fields: vec![StructField { field_name: None, - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[3]) @@ -1255,7 +1292,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { })], fields: vec![StructField { field_name: None, - field_type: DataType::Interval + field_type: DataType::Interval, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1270,7 +1308,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }], fields: vec![StructField { field_name: None, - field_type: DataType::JSON + field_type: DataType::JSON, + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1286,7 +1325,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { )], fields: vec![StructField { field_name: None, - field_type: DataType::String(Some(42)) + field_type: DataType::String(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1299,7 +1339,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Timestamp(None, TimezoneInfo::None) + field_type: DataType::Timestamp(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1313,7 +1354,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Time(None, TimezoneInfo::None) + field_type: DataType::Time(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -1330,7 +1372,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }], fields: vec![StructField { field_name: None, - field_type: DataType::Numeric(ExactNumberInfo::None) + field_type: DataType::Numeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1343,7 +1386,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { }], fields: vec![StructField { field_name: None, - field_type: DataType::BigNumeric(ExactNumberInfo::None) + field_type: DataType::BigNumeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1360,7 +1404,8 @@ fn parse_typed_struct_with_field_name_bigquery() { values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1372,7 +1417,8 @@ fn parse_typed_struct_with_field_name_bigquery() { )], fields: vec![StructField { field_name: Some(Ident::from("y")), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1387,11 +1433,13 @@ fn parse_typed_struct_with_field_name_bigquery() { fields: vec![ StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident::from("y")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, } ] }, @@ -1409,7 +1457,8 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1421,7 +1470,8 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { )], fields: vec![StructField { field_name: Some(Ident::from("y")), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1436,11 +1486,13 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { fields: vec![ StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident::from("y")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, } ] }, @@ -2407,3 +2459,16 @@ fn test_any_type() { fn test_any_type_dont_break_custom_type() { bigquery_and_generic().verified_stmt("CREATE TABLE foo (x ANY)"); } + +#[test] +fn test_struct_field_options() { + bigquery().verified_stmt(concat!( + "CREATE TABLE my_table (", + "f0 STRUCT, ", + "f1 STRUCT<", + "a STRING OPTIONS(description = 'This is a string', type = 'string'), ", + "b INT64", + "> OPTIONS(description = 'This is a struct field')", + ")", + )); +} diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index d0218b6c3..93b4c4f59 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -669,11 +669,13 @@ fn parse_create_table_with_nested_data_types() { DataType::Tuple(vec![ StructField { field_name: None, - field_type: DataType::FixedString(128) + field_type: DataType::FixedString(128), + options: None, }, StructField { field_name: None, - field_type: DataType::Int128 + field_type: DataType::Int128, + options: None, } ]) ))), @@ -685,12 +687,14 @@ fn parse_create_table_with_nested_data_types() { StructField { field_name: Some("a".into()), field_type: DataType::Datetime64(9, None), + options: None, }, StructField { field_name: Some("b".into()), field_type: DataType::Array(ArrayElemTypeDef::Parenthesis( Box::new(DataType::Uuid) - )) + )), + options: None, }, ]), options: vec![], diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 8e4983655..f503ed551 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -44,10 +44,12 @@ fn test_struct() { StructField { field_name: Some(Ident::new("v")), field_type: DataType::Varchar(None), + options: None, }, StructField { field_name: Some(Ident::new("i")), field_type: DataType::Integer(None), + options: None, }, ], StructBracketKind::Parentheses, @@ -84,6 +86,7 @@ fn test_struct() { StructField { field_name: Some(Ident::new("v")), field_type: DataType::Varchar(None), + options: None, }, StructField { field_name: Some(Ident::new("s")), @@ -92,14 +95,17 @@ fn test_struct() { StructField { field_name: Some(Ident::new("a1")), field_type: DataType::Integer(None), + options: None, }, StructField { field_name: Some(Ident::new("a2")), field_type: DataType::Varchar(None), + options: None, }, ], StructBracketKind::Parentheses, ), + options: None, }, ], StructBracketKind::Parentheses, From 185a4902189565031fa066d165594325172f2893 Mon Sep 17 00:00:00 2001 From: hulk Date: Fri, 20 Jun 2025 22:56:26 +0800 Subject: [PATCH 158/286] Fix parsing error when having fields after nested struct in BigQuery (#1897) --- src/parser/mod.rs | 11 +--- tests/sqlparser_bigquery.rs | 104 ++++++++++++++++++++++++++---------- 2 files changed, 77 insertions(+), 38 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 19c11d296..ece9ba65c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3034,7 +3034,6 @@ impl<'a> Parser<'a> { where F: FnMut(&mut Parser<'a>) -> Result<(StructField, MatchedTrailingBracket), ParserError>, { - let start_token = self.peek_token(); self.expect_keyword_is(Keyword::STRUCT)?; // Nothing to do if we have no type information. @@ -3047,16 +3046,10 @@ impl<'a> Parser<'a> { let trailing_bracket = loop { let (def, trailing_bracket) = elem_parser(self)?; field_defs.push(def); - if !self.consume_token(&Token::Comma) { + // The struct field definition is finished if it occurs `>>` or comma. + if trailing_bracket.0 || !self.consume_token(&Token::Comma) { break trailing_bracket; } - - // Angle brackets are balanced so we only expect the trailing `>>` after - // we've matched all field types for the current struct. - // e.g. this is invalid syntax `STRUCT>>, INT>(NULL)` - if trailing_bracket.0 { - return parser_err!("unmatched > in STRUCT definition", start_token.span.start); - } }; Ok(( diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index e37e4a449..810d8480a 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -635,35 +635,6 @@ fn parse_nested_data_types() { } } -#[test] -fn parse_invalid_brackets() { - let sql = "SELECT STRUCT>(NULL)"; - assert_eq!( - bigquery_and_generic() - .parse_sql_statements(sql) - .unwrap_err(), - ParserError::ParserError("unmatched > in STRUCT literal".to_string()) - ); - - let sql = "SELECT STRUCT>>(NULL)"; - assert_eq!( - bigquery_and_generic() - .parse_sql_statements(sql) - .unwrap_err(), - ParserError::ParserError("Expected: (, found: >".to_string()) - ); - - let sql = "CREATE TABLE table (x STRUCT>>)"; - assert_eq!( - bigquery_and_generic() - .parse_sql_statements(sql) - .unwrap_err(), - ParserError::ParserError( - "Expected: ',' or ')' after column definition, found: >".to_string() - ) - ); -} - #[test] fn parse_tuple_struct_literal() { // tuple syntax: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#tuple_syntax @@ -2472,3 +2443,78 @@ fn test_struct_field_options() { ")", )); } + +#[test] +fn test_struct_trailing_and_nested_bracket() { + bigquery().verified_stmt(concat!( + "CREATE TABLE my_table (", + "f0 STRING, ", + "f1 STRUCT>, ", + "f2 STRING", + ")", + )); + + // More complex nested structs + bigquery().verified_stmt(concat!( + "CREATE TABLE my_table (", + "f0 STRING, ", + "f1 STRUCT>>, ", + "f2 STRUCT>>>, ", + "f3 STRUCT>", + ")", + )); + + // Bad case with missing closing bracket + assert_eq!( + ParserError::ParserError("Expected: >, found: )".to_owned()), + bigquery() + .parse_sql_statements("CREATE TABLE my_table(f1 STRUCT after parsing data type STRUCT)".to_owned() + ), + bigquery() + .parse_sql_statements("CREATE TABLE my_table(f1 STRUCT>)") + .unwrap_err() + ); + + // Base case with redundant closing bracket in nested struct + assert_eq!( + ParserError::ParserError( + "Expected: ',' or ')' after column definition, found: >".to_owned() + ), + bigquery() + .parse_sql_statements("CREATE TABLE my_table(f1 STRUCT>>, c INT64)") + .unwrap_err() + ); + + let sql = "SELECT STRUCT>(NULL)"; + assert_eq!( + bigquery_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError::ParserError("unmatched > in STRUCT literal".to_string()) + ); + + let sql = "SELECT STRUCT>>(NULL)"; + assert_eq!( + bigquery_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError::ParserError("Expected: (, found: >".to_string()) + ); + + let sql = "CREATE TABLE table (x STRUCT>>)"; + assert_eq!( + bigquery_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError::ParserError( + "Expected: ',' or ')' after column definition, found: >".to_string() + ) + ); +} From 204d3b484db4a659396ae13e7bd2c20ac55fb07c Mon Sep 17 00:00:00 2001 From: Simon Sawert Date: Sat, 21 Jun 2025 08:12:07 +0200 Subject: [PATCH 159/286] Extend exception handling (#1884) --- src/ast/mod.rs | 60 +++++++++++++++++++++++++-------- src/dialect/bigquery.rs | 51 +++------------------------- src/dialect/snowflake.rs | 4 +++ src/keywords.rs | 1 + src/parser/mod.rs | 49 +++++++++++++++++++++++++-- tests/sqlparser_bigquery.rs | 9 +++-- tests/sqlparser_common.rs | 7 ++-- tests/sqlparser_snowflake.rs | 64 ++++++++++++++++++++++++++++++++++++ 8 files changed, 178 insertions(+), 67 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 04401a48b..369489f56 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2990,6 +2990,36 @@ impl From for Statement { } } +/// A representation of a `WHEN` arm with all the identifiers catched and the statements to execute +/// for the arm. +/// +/// Snowflake: +/// BigQuery: +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ExceptionWhen { + pub idents: Vec, + pub statements: Vec, +} + +impl Display for ExceptionWhen { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "WHEN {idents} THEN", + idents = display_separated(&self.idents, " OR ") + )?; + + if !self.statements.is_empty() { + write!(f, " ")?; + format_statement_list(f, &self.statements)?; + } + + Ok(()) + } +} + /// A top-level statement (SELECT, INSERT, CREATE, etc.) #[allow(clippy::large_enum_variant)] #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -3678,17 +3708,20 @@ pub enum Statement { /// END; /// ``` statements: Vec, - /// Statements of an exception clause. + /// Exception handling with exception clauses. /// Example: /// ```sql - /// BEGIN - /// SELECT 1; - /// EXCEPTION WHEN ERROR THEN - /// SELECT 2; - /// SELECT 3; - /// END; + /// EXCEPTION + /// WHEN EXCEPTION_1 THEN + /// SELECT 2; + /// WHEN EXCEPTION_2 OR EXCEPTION_3 THEN + /// SELECT 3; + /// WHEN OTHER THEN + /// SELECT 4; + /// ``` /// - exception_statements: Option>, + /// + exception: Option>, /// TRUE if the statement has an `END` keyword. has_end_keyword: bool, }, @@ -5533,7 +5566,7 @@ impl fmt::Display for Statement { transaction, modifier, statements, - exception_statements, + exception, has_end_keyword, } => { if *syntax_begin { @@ -5555,11 +5588,10 @@ impl fmt::Display for Statement { write!(f, " ")?; format_statement_list(f, statements)?; } - if let Some(exception_statements) = exception_statements { - write!(f, " EXCEPTION WHEN ERROR THEN")?; - if !exception_statements.is_empty() { - write!(f, " ")?; - format_statement_list(f, exception_statements)?; + if let Some(exception_when) = exception { + write!(f, " EXCEPTION")?; + for when in exception_when { + write!(f, " {when}")?; } } if *has_end_keyword { diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index 68ca1390a..c2cd507ce 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -46,7 +46,11 @@ pub struct BigQueryDialect; impl Dialect for BigQueryDialect { fn parse_statement(&self, parser: &mut Parser) -> Option> { - self.maybe_parse_statement(parser) + if parser.parse_keyword(Keyword::BEGIN) { + return Some(parser.parse_begin_exception_end()); + } + + None } /// See @@ -141,48 +145,3 @@ impl Dialect for BigQueryDialect { true } } - -impl BigQueryDialect { - fn maybe_parse_statement(&self, parser: &mut Parser) -> Option> { - if parser.peek_keyword(Keyword::BEGIN) { - return Some(self.parse_begin(parser)); - } - None - } - - /// Parse a `BEGIN` statement. - /// - fn parse_begin(&self, parser: &mut Parser) -> Result { - parser.expect_keyword(Keyword::BEGIN)?; - - let statements = parser.parse_statement_list(&[Keyword::EXCEPTION, Keyword::END])?; - - let has_exception_when_clause = parser.parse_keywords(&[ - Keyword::EXCEPTION, - Keyword::WHEN, - Keyword::ERROR, - Keyword::THEN, - ]); - let exception_statements = if has_exception_when_clause { - if !parser.peek_keyword(Keyword::END) { - Some(parser.parse_statement_list(&[Keyword::END])?) - } else { - Some(Default::default()) - } - } else { - None - }; - - parser.expect_keyword(Keyword::END)?; - - Ok(Statement::StartTransaction { - begin: true, - statements, - exception_statements, - has_end_keyword: true, - transaction: None, - modifier: None, - modes: Default::default(), - }) - } -} diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index ea0b94a63..66e04ac24 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -131,6 +131,10 @@ impl Dialect for SnowflakeDialect { } fn parse_statement(&self, parser: &mut Parser) -> Option> { + if parser.parse_keyword(Keyword::BEGIN) { + return Some(parser.parse_begin_exception_end()); + } + if parser.parse_keywords(&[Keyword::ALTER, Keyword::SESSION]) { // ALTER SESSION let set = match parser.parse_one_of_keywords(&[Keyword::SET, Keyword::UNSET]) { diff --git a/src/keywords.rs b/src/keywords.rs index f5c5e567e..cb6c7a6e2 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -646,6 +646,7 @@ define_keywords!( ORDER, ORDINALITY, ORGANIZATION, + OTHER, OUT, OUTER, OUTPUT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ece9ba65c..be32093f3 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15092,7 +15092,7 @@ impl<'a> Parser<'a> { transaction: Some(BeginTransactionKind::Transaction), modifier: None, statements: vec![], - exception_statements: None, + exception: None, has_end_keyword: false, }) } @@ -15124,11 +15124,56 @@ impl<'a> Parser<'a> { transaction, modifier, statements: vec![], - exception_statements: None, + exception: None, has_end_keyword: false, }) } + pub fn parse_begin_exception_end(&mut self) -> Result { + let statements = self.parse_statement_list(&[Keyword::EXCEPTION, Keyword::END])?; + + let exception = if self.parse_keyword(Keyword::EXCEPTION) { + let mut when = Vec::new(); + + // We can have multiple `WHEN` arms so we consume all cases until `END` + while !self.peek_keyword(Keyword::END) { + self.expect_keyword(Keyword::WHEN)?; + + // Each `WHEN` case can have one or more conditions, e.g. + // WHEN EXCEPTION_1 [OR EXCEPTION_2] THEN + // So we parse identifiers until the `THEN` keyword. + let mut idents = Vec::new(); + + while !self.parse_keyword(Keyword::THEN) { + let ident = self.parse_identifier()?; + idents.push(ident); + + self.maybe_parse(|p| p.expect_keyword(Keyword::OR))?; + } + + let statements = self.parse_statement_list(&[Keyword::WHEN, Keyword::END])?; + + when.push(ExceptionWhen { idents, statements }); + } + + Some(when) + } else { + None + }; + + self.expect_keyword(Keyword::END)?; + + Ok(Statement::StartTransaction { + begin: true, + statements, + exception, + has_end_keyword: true, + transaction: None, + modifier: None, + modes: Default::default(), + }) + } + pub fn parse_end(&mut self) -> Result { let modifier = if !self.dialect.supports_end_transaction_modifier() { None diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 810d8480a..0de0b12ba 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -261,10 +261,10 @@ fn parse_at_at_identifier() { #[test] fn parse_begin() { - let sql = r#"BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT 2; END"#; + let sql = r#"BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT 2; RAISE USING MESSAGE = FORMAT('ERR: %s', 'Bad'); END"#; let Statement::StartTransaction { statements, - exception_statements, + exception, has_end_keyword, .. } = bigquery().verified_stmt(sql) @@ -272,7 +272,10 @@ fn parse_begin() { unreachable!(); }; assert_eq!(1, statements.len()); - assert_eq!(1, exception_statements.unwrap().len()); + assert!(exception.is_some()); + + let exception = exception.unwrap(); + assert_eq!(1, exception.len()); assert!(has_end_keyword); bigquery().verified_stmt( diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index abcadb458..e4363ff6e 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8592,8 +8592,11 @@ fn lateral_function() { #[test] fn parse_start_transaction() { let dialects = all_dialects_except(|d| - // BigQuery does not support this syntax - d.is::()); + // BigQuery and Snowflake does not support this syntax + // + // BigQuery: + // Snowflake: + d.is::() || d.is::()); match dialects .verified_stmt("START TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE") { diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 7d734ca26..b11a2cb05 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4082,3 +4082,67 @@ fn parse_connect_by_root_operator() { "sql parser error: Expected an expression, found: FROM" ); } + +#[test] +fn test_begin_exception_end() { + for sql in [ + "BEGIN SELECT 1; EXCEPTION WHEN OTHER THEN SELECT 2; RAISE; END", + "BEGIN SELECT 1; EXCEPTION WHEN OTHER THEN SELECT 2; RAISE EX_1; END", + "BEGIN SELECT 1; EXCEPTION WHEN FOO THEN SELECT 2; WHEN OTHER THEN SELECT 3; RAISE; END", + "BEGIN BEGIN SELECT 1; EXCEPTION WHEN OTHER THEN SELECT 2; RAISE; END; END", + ] { + snowflake().verified_stmt(sql); + } + + let sql = r#" +DECLARE + EXCEPTION_1 EXCEPTION (-20001, 'I caught the expected exception.'); + EXCEPTION_2 EXCEPTION (-20002, 'Not the expected exception!'); + EXCEPTION_3 EXCEPTION (-20003, 'The worst exception...'); +BEGIN + BEGIN + SELECT 1; + EXCEPTION + WHEN EXCEPTION_1 THEN + SELECT 1; + WHEN EXCEPTION_2 OR EXCEPTION_3 THEN + SELECT 2; + SELECT 3; + WHEN OTHER THEN + SELECT 4; + RAISE; + END; +END +"#; + + // Outer `BEGIN` of the two nested `BEGIN` statements. + let Statement::StartTransaction { mut statements, .. } = snowflake() + .parse_sql_statements(sql) + .unwrap() + .pop() + .unwrap() + else { + unreachable!(); + }; + + // Inner `BEGIN` of the two nested `BEGIN` statements. + let Statement::StartTransaction { + statements, + exception, + has_end_keyword, + .. + } = statements.pop().unwrap() + else { + unreachable!(); + }; + + assert_eq!(1, statements.len()); + assert!(has_end_keyword); + + let exception = exception.unwrap(); + assert_eq!(3, exception.len()); + assert_eq!(1, exception[0].idents.len()); + assert_eq!(1, exception[0].statements.len()); + assert_eq!(2, exception[1].idents.len()); + assert_eq!(2, exception[1].statements.len()); +} From 1d0dc7cdd8209d5e68f41675622dd17e2b336e26 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Sun, 22 Jun 2025 08:02:51 +0100 Subject: [PATCH 160/286] Postgres: Add support for text search types (#1889) --- src/ast/data_type.rs | 10 ++++++++++ src/keywords.rs | 2 ++ src/parser/mod.rs | 6 ++++++ tests/sqlparser_postgres.rs | 31 +++++++++++++++++++++++++++++++ 4 files changed, 49 insertions(+) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 3a4958c9f..ef04b7e4b 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -446,6 +446,14 @@ pub enum DataType { /// /// [PostgreSQL]: https://www.postgresql.org/docs/9.5/functions-geometry.html GeometricType(GeometricTypeKind), + /// PostgreSQL text search vectors, see [PostgreSQL]. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/17/datatype-textsearch.html + TsVector, + /// PostgreSQL text search query, see [PostgreSQL]. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/17/datatype-textsearch.html + TsQuery, } impl fmt::Display for DataType { @@ -738,6 +746,8 @@ impl fmt::Display for DataType { write!(f, "{} TABLE ({})", name, display_comma_separated(columns)) } DataType::GeometricType(kind) => write!(f, "{}", kind), + DataType::TsVector => write!(f, "TSVECTOR"), + DataType::TsQuery => write!(f, "TSQUERY"), } } } diff --git a/src/keywords.rs b/src/keywords.rs index cb6c7a6e2..f56178c17 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -935,6 +935,8 @@ define_keywords!( TRY, TRY_CAST, TRY_CONVERT, + TSQUERY, + TSVECTOR, TUPLE, TYPE, UBIGINT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index be32093f3..f60323b22 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9909,6 +9909,12 @@ impl<'a> Parser<'a> { Ok(DataType::Unsigned) } } + Keyword::TSVECTOR if dialect_is!(dialect is PostgreSqlDialect | GenericDialect) => { + Ok(DataType::TsVector) + } + Keyword::TSQUERY if dialect_is!(dialect is PostgreSqlDialect | GenericDialect) => { + Ok(DataType::TsQuery) + } _ => { self.prev_token(); let type_name = self.parse_object_name(false)?; diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 6f0ba9c69..b6605cf1a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -6201,3 +6201,34 @@ fn parse_alter_table_replica_identity() { _ => unreachable!(), } } + +#[test] +fn parse_ts_datatypes() { + match pg_and_generic().verified_stmt("CREATE TABLE foo (x TSVECTOR)") { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ColumnDef { + name: "x".into(), + data_type: DataType::TsVector, + options: vec![], + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("CREATE TABLE foo (x TSQUERY)") { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ColumnDef { + name: "x".into(), + data_type: DataType::TsQuery, + options: vec![], + }] + ); + } + _ => unreachable!(), + } +} From 7865de015f8a45d9545dc7ca0659f37a0c3f7496 Mon Sep 17 00:00:00 2001 From: Dima <111751109+Dimchikkk@users.noreply.github.com> Date: Sun, 22 Jun 2025 08:22:45 +0100 Subject: [PATCH 161/286] Fix `limit` in subqueries (#1899) --- src/ast/mod.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_common.rs | 11 +++++++++-- 3 files changed, 11 insertions(+), 4 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 369489f56..17464f4ac 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -748,7 +748,7 @@ pub enum Expr { /// `[ NOT ] IN (SELECT ...)` InSubquery { expr: Box, - subquery: Box, + subquery: Box, negated: bool, }, /// `[ NOT ] IN UNNEST(array_expression)` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f60323b22..44bf58d0f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3817,7 +3817,7 @@ impl<'a> Parser<'a> { }); } self.expect_token(&Token::LParen)?; - let in_op = match self.maybe_parse(|p| p.parse_query_body(p.dialect.prec_unknown()))? { + let in_op = match self.maybe_parse(|p| p.parse_query())? { Some(subquery) => Expr::InSubquery { expr: Box::new(expr), subquery, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index e4363ff6e..52054604d 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2225,7 +2225,7 @@ fn parse_in_subquery() { assert_eq!( Expr::InSubquery { expr: Box::new(Expr::Identifier(Ident::new("segment"))), - subquery: verified_query("SELECT segm FROM bar").body, + subquery: Box::new(verified_query("SELECT segm FROM bar")), negated: false, }, select.selection.unwrap() @@ -2239,7 +2239,9 @@ fn parse_in_union() { assert_eq!( Expr::InSubquery { expr: Box::new(Expr::Identifier(Ident::new("segment"))), - subquery: verified_query("(SELECT segm FROM bar) UNION (SELECT segm FROM bar2)").body, + subquery: Box::new(verified_query( + "(SELECT segm FROM bar) UNION (SELECT segm FROM bar2)" + )), negated: false, }, select.selection.unwrap() @@ -15303,6 +15305,11 @@ fn parse_return() { let _ = all_dialects().verified_stmt("RETURN 1"); } +#[test] +fn parse_subquery_limit() { + let _ = all_dialects().verified_stmt("SELECT t1_id, t1_name FROM t1 WHERE t1_id IN (SELECT t2_id FROM t2 WHERE t1_name = t2_name LIMIT 10)"); +} + #[test] fn test_open() { let open_cursor = "OPEN Employee_Cursor"; From 5d63663bc6cd78d63d0cb2581ffc71a1dee3559d Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Mon, 23 Jun 2025 23:18:03 -0700 Subject: [PATCH 162/286] Use `IndexColumn` in all index definitions (#1900) --- src/ast/ddl.rs | 14 ++++---- src/ast/spans.rs | 35 +++++++++++-------- src/parser/mod.rs | 48 +++++++++++++++++-------- src/test_utils.rs | 44 +++++++++++++++++++++++ tests/sqlparser_mysql.rs | 75 ++++++++++++++++++++++++++++++++++++++++ 5 files changed, 181 insertions(+), 35 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 059c61967..e35332da9 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -32,9 +32,9 @@ use crate::ast::value::escape_single_quote_string; use crate::ast::{ display_comma_separated, display_separated, CommentDef, CreateFunctionBody, CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, - FunctionDeterminismSpecifier, FunctionParallel, Ident, MySQLColumnPosition, ObjectName, - OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, Value, - ValueWithSpan, + FunctionDeterminismSpecifier, FunctionParallel, Ident, IndexColumn, MySQLColumnPosition, + ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, + Value, ValueWithSpan, }; use crate::keywords::Keyword; use crate::tokenizer::Token; @@ -979,7 +979,7 @@ pub enum TableConstraint { /// [1]: IndexType index_type: Option, /// Identifiers of the columns that are unique. - columns: Vec, + columns: Vec, index_options: Vec, characteristics: Option, /// Optional Postgres nulls handling: `[ NULLS [ NOT ] DISTINCT ]` @@ -1015,7 +1015,7 @@ pub enum TableConstraint { /// [1]: IndexType index_type: Option, /// Identifiers of the columns that form the primary key. - columns: Vec, + columns: Vec, index_options: Vec, characteristics: Option, }, @@ -1060,7 +1060,7 @@ pub enum TableConstraint { /// [1]: IndexType index_type: Option, /// Referred column identifier list. - columns: Vec, + columns: Vec, }, /// MySQLs [fulltext][1] definition. Since the [`SPATIAL`][2] definition is exactly the same, /// and MySQL displays both the same way, it is part of this definition as well. @@ -1083,7 +1083,7 @@ pub enum TableConstraint { /// Optional index name. opt_index_name: Option, /// Referred column identifier list. - columns: Vec, + columns: Vec, }, } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 14664b4cd..ca321cc2e 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -28,16 +28,17 @@ use super::{ ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, - FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, Insert, Interpolate, - InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, - LimitClause, MatchRecognizePattern, Measure, NamedParenthesizedList, NamedWindowDefinition, - ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, OnInsert, OpenStatement, - OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, - RaiseStatement, RaiseStatementValue, ReferentialAction, RenameSelectItem, ReplaceSelectElement, - ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, - SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, - TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, - WhileStatement, WildcardAdditionalOptions, With, WithFill, + FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, IndexColumn, Insert, + Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, + LateralView, LimitClause, MatchRecognizePattern, Measure, NamedParenthesizedList, + NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, + OnInsert, OpenStatement, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, + ProjectionSelect, Query, RaiseStatement, RaiseStatementValue, ReferentialAction, + RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, + SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, + TableConstraint, TableFactor, TableObject, TableOptionsClustered, TableWithJoins, + UpdateTableFromKind, Use, Value, Values, ViewColumnDef, WhileStatement, + WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -650,7 +651,7 @@ impl Spanned for TableConstraint { name.iter() .map(|i| i.span) .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span)) + .chain(columns.iter().map(|i| i.span())) .chain(characteristics.iter().map(|i| i.span())), ), TableConstraint::PrimaryKey { @@ -664,7 +665,7 @@ impl Spanned for TableConstraint { name.iter() .map(|i| i.span) .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span)) + .chain(columns.iter().map(|i| i.span())) .chain(characteristics.iter().map(|i| i.span())), ), TableConstraint::ForeignKey { @@ -700,7 +701,7 @@ impl Spanned for TableConstraint { } => union_spans( name.iter() .map(|i| i.span) - .chain(columns.iter().map(|i| i.span)), + .chain(columns.iter().map(|i| i.span())), ), TableConstraint::FulltextOrSpatial { fulltext: _, @@ -711,7 +712,7 @@ impl Spanned for TableConstraint { opt_index_name .iter() .map(|i| i.span) - .chain(columns.iter().map(|i| i.span)), + .chain(columns.iter().map(|i| i.span())), ), } } @@ -745,6 +746,12 @@ impl Spanned for CreateIndex { } } +impl Spanned for IndexColumn { + fn span(&self) -> Span { + self.column.span() + } +} + impl Spanned for CaseStatement { fn span(&self) -> Span { let CaseStatement { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 44bf58d0f..35c05dfbe 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6868,9 +6868,7 @@ impl<'a> Parser<'a> { None }; - self.expect_token(&Token::LParen)?; - let columns = self.parse_comma_separated(Parser::parse_create_index_expr)?; - self.expect_token(&Token::RParen)?; + let columns = self.parse_parenthesized_index_column_list()?; let include = if self.parse_keyword(Keyword::INCLUDE) { self.expect_token(&Token::LParen)?; @@ -8070,7 +8068,7 @@ impl<'a> Parser<'a> { let index_name = self.parse_optional_ident()?; let index_type = self.parse_optional_using_then_index_type()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; let characteristics = self.parse_constraint_characteristics()?; Ok(Some(TableConstraint::Unique { @@ -8092,7 +8090,7 @@ impl<'a> Parser<'a> { let index_name = self.parse_optional_ident()?; let index_type = self.parse_optional_using_then_index_type()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; let characteristics = self.parse_constraint_characteristics()?; Ok(Some(TableConstraint::PrimaryKey { @@ -8170,7 +8168,7 @@ impl<'a> Parser<'a> { }; let index_type = self.parse_optional_using_then_index_type()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; Ok(Some(TableConstraint::Index { display_as_key, @@ -8199,7 +8197,7 @@ impl<'a> Parser<'a> { let opt_index_name = self.parse_optional_ident()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; Ok(Some(TableConstraint::FulltextOrSpatial { fulltext, @@ -10601,6 +10599,14 @@ impl<'a> Parser<'a> { self.parse_parenthesized_column_list_inner(optional, allow_empty, |p| p.parse_identifier()) } + /// Parses a parenthesized comma-separated list of index columns, which can be arbitrary + /// expressions with ordering information (and an opclass in some dialects). + fn parse_parenthesized_index_column_list(&mut self) -> Result, ParserError> { + self.parse_parenthesized_column_list_inner(Mandatory, false, |p| { + p.parse_create_index_expr() + }) + } + /// Parses a parenthesized comma-separated list of qualified, possibly quoted identifiers. /// For example: `(db1.sc1.tbl1.col1, db1.sc1.tbl1."col 2", ...)` pub fn parse_parenthesized_qualified_column_list( @@ -16527,6 +16533,20 @@ mod tests { }}; } + fn mk_expected_col(name: &str) -> IndexColumn { + IndexColumn { + column: OrderByExpr { + expr: Expr::Identifier(name.into()), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }, + operator_class: None, + } + } + let dialect = TestedDialects::new(vec![Box::new(GenericDialect {}), Box::new(MySqlDialect {})]); @@ -16537,7 +16557,7 @@ mod tests { display_as_key: false, name: None, index_type: None, - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16548,7 +16568,7 @@ mod tests { display_as_key: true, name: None, index_type: None, - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16559,7 +16579,7 @@ mod tests { display_as_key: false, name: Some(Ident::with_quote('\'', "index")), index_type: None, - columns: vec![Ident::new("c1"), Ident::new("c2")], + columns: vec![mk_expected_col("c1"), mk_expected_col("c2")], } ); @@ -16570,7 +16590,7 @@ mod tests { display_as_key: false, name: None, index_type: Some(IndexType::BTree), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16581,7 +16601,7 @@ mod tests { display_as_key: false, name: None, index_type: Some(IndexType::Hash), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16592,7 +16612,7 @@ mod tests { display_as_key: false, name: Some(Ident::new("idx_name")), index_type: Some(IndexType::BTree), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16603,7 +16623,7 @@ mod tests { display_as_key: false, name: Some(Ident::new("idx_name")), index_type: Some(IndexType::Hash), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); } diff --git a/src/test_utils.rs b/src/test_utils.rs index 24c0ca575..1af9f454b 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -448,3 +448,47 @@ pub fn call(function: &str, args: impl IntoIterator) -> Expr { within_group: vec![], }) } + +/// Gets the first index column (mysql calls it a key part) of the first index found in a +/// [`Statement::CreateIndex`], [`Statement::CreateTable`], or [`Statement::AlterTable`]. +pub fn index_column(stmt: Statement) -> Expr { + match stmt { + Statement::CreateIndex(CreateIndex { columns, .. }) => { + columns.first().unwrap().column.expr.clone() + } + Statement::CreateTable(CreateTable { constraints, .. }) => { + match constraints.first().unwrap() { + TableConstraint::Index { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::Unique { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::PrimaryKey { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::FulltextOrSpatial { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), + } + } + Statement::AlterTable { operations, .. } => match operations.first().unwrap() { + AlterTableOperation::AddConstraint(TableConstraint::Index { columns, .. }) => { + columns.first().unwrap().column.expr.clone() + } + AlterTableOperation::AddConstraint(TableConstraint::Unique { columns, .. }) => { + columns.first().unwrap().column.expr.clone() + } + AlterTableOperation::AddConstraint(TableConstraint::PrimaryKey { columns, .. }) => { + columns.first().unwrap().column.expr.clone() + } + AlterTableOperation::AddConstraint(TableConstraint::FulltextOrSpatial { + columns, + .. + }) => columns.first().unwrap().column.expr.clone(), + _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), + }, + _ => panic!("Expected CREATE INDEX, ALTER TABLE, or CREATE TABLE, got: {stmt:?}"), + } +} diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 540348ff3..b11d76dde 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -670,6 +670,20 @@ fn table_constraint_unique_primary_ctor( characteristics: Option, unique_index_type_display: Option, ) -> TableConstraint { + let columns = columns + .into_iter() + .map(|ident| IndexColumn { + column: OrderByExpr { + expr: Expr::Identifier(ident), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }, + operator_class: None, + }) + .collect(); match unique_index_type_display { Some(index_type_display) => TableConstraint::Unique { name, @@ -795,6 +809,67 @@ fn parse_create_table_primary_and_unique_key_with_index_options() { } } +#[test] +fn parse_prefix_key_part() { + let expected = vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::value( + number("10"), + )))]; + for sql in [ + "CREATE INDEX idx_index ON t(textcol(10))", + "ALTER TABLE tab ADD INDEX idx_index (textcol(10))", + "ALTER TABLE tab ADD PRIMARY KEY (textcol(10))", + "ALTER TABLE tab ADD UNIQUE KEY (textcol(10))", + "ALTER TABLE tab ADD UNIQUE KEY (textcol(10))", + "ALTER TABLE tab ADD FULLTEXT INDEX (textcol(10))", + "CREATE TABLE t (textcol TEXT, INDEX idx_index (textcol(10)))", + ] { + match index_column(mysql_and_generic().verified_stmt(sql)) { + Expr::Function(Function { + name, + args: FunctionArguments::List(FunctionArgumentList { args, .. }), + .. + }) => { + assert_eq!(name.to_string(), "textcol"); + assert_eq!(args, expected); + } + expr => panic!("unexpected expression {expr} for {sql}"), + } + } +} + +#[test] +fn test_functional_key_part() { + assert_eq!( + index_column( + mysql_and_generic() + .verified_stmt("CREATE INDEX idx_index ON t((col COLLATE utf8mb4_bin) DESC)") + ), + Expr::Nested(Box::new(Expr::Collate { + expr: Box::new(Expr::Identifier("col".into())), + collation: ObjectName(vec![sqlparser::ast::ObjectNamePart::Identifier( + Ident::new("utf8mb4_bin") + )]), + })) + ); + assert_eq!( + index_column(mysql_and_generic().verified_stmt( + r#"CREATE TABLE t (jsoncol JSON, PRIMARY KEY ((CAST(col ->> '$.id' AS UNSIGNED)) ASC))"# + )), + Expr::Nested(Box::new(Expr::Cast { + kind: CastKind::Cast, + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("col"))), + op: BinaryOperator::LongArrow, + right: Box::new(Expr::Value( + Value::SingleQuotedString("$.id".to_string()).with_empty_span() + )), + }), + data_type: DataType::Unsigned, + format: None, + })), + ); +} + #[test] fn parse_create_table_primary_and_unique_key_with_index_type() { let sqls = ["UNIQUE", "PRIMARY KEY"].map(|key_ty| { From 44f3be38e5c9199479a4b986fdfaf417e925f0d4 Mon Sep 17 00:00:00 2001 From: Denys Tsomenko Date: Tue, 24 Jun 2025 09:29:44 +0300 Subject: [PATCH 163/286] fix: parse snowflake fetch clause (#1894) --- src/parser/mod.rs | 11 ++++++----- tests/sqlparser_snowflake.rs | 19 +++++++++++++++++++ 2 files changed, 25 insertions(+), 5 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 35c05dfbe..a5e890692 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15023,7 +15023,8 @@ impl<'a> Parser<'a> { /// Parse a FETCH clause pub fn parse_fetch(&mut self) -> Result { - self.expect_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT])?; + let _ = self.parse_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]); + let (quantity, percent) = if self .parse_one_of_keywords(&[Keyword::ROW, Keyword::ROWS]) .is_some() @@ -15032,16 +15033,16 @@ impl<'a> Parser<'a> { } else { let quantity = Expr::Value(self.parse_value()?); let percent = self.parse_keyword(Keyword::PERCENT); - self.expect_one_of_keywords(&[Keyword::ROW, Keyword::ROWS])?; + let _ = self.parse_one_of_keywords(&[Keyword::ROW, Keyword::ROWS]); (Some(quantity), percent) }; + let with_ties = if self.parse_keyword(Keyword::ONLY) { false - } else if self.parse_keywords(&[Keyword::WITH, Keyword::TIES]) { - true } else { - return self.expected("one of ONLY or WITH TIES", self.peek_token()); + self.parse_keywords(&[Keyword::WITH, Keyword::TIES]) }; + Ok(Fetch { with_ties, percent, diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index b11a2cb05..7dc00f9a8 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4146,3 +4146,22 @@ END assert_eq!(2, exception[1].idents.len()); assert_eq!(2, exception[1].statements.len()); } + +#[test] +fn test_snowflake_fetch_clause_syntax() { + let canonical = "SELECT c1 FROM fetch_test FETCH FIRST 2 ROWS ONLY"; + snowflake().verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH 2", canonical); + + snowflake() + .verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH FIRST 2", canonical); + snowflake() + .verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH NEXT 2", canonical); + + snowflake() + .verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH 2 ROW", canonical); + + snowflake().verified_only_select_with_canonical( + "SELECT c1 FROM fetch_test FETCH FIRST 2 ROWS", + canonical, + ); +} From b9365b3853cf8878284746af07f7781915ae7052 Mon Sep 17 00:00:00 2001 From: ZacJW Date: Tue, 24 Jun 2025 07:39:02 +0100 Subject: [PATCH 164/286] Support procedure argmode (#1901) --- src/ast/ddl.rs | 9 ++++-- src/parser/mod.rs | 15 +++++++++- tests/sqlparser_common.rs | 62 +++++++++++++++++++++++++++++++++++++++ tests/sqlparser_mssql.rs | 6 ++-- 4 files changed, 87 insertions(+), 5 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index e35332da9..f81c6fc20 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -30,7 +30,7 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, CommentDef, CreateFunctionBody, + display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, Ident, IndexColumn, MySQLColumnPosition, ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, @@ -1367,11 +1367,16 @@ impl fmt::Display for NullsDistinctOption { pub struct ProcedureParam { pub name: Ident, pub data_type: DataType, + pub mode: Option, } impl fmt::Display for ProcedureParam { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{} {}", self.name, self.data_type) + if let Some(mode) = &self.mode { + write!(f, "{mode} {} {}", self.name, self.data_type) + } else { + write!(f, "{} {}", self.name, self.data_type) + } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a5e890692..ca658d7da 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7624,9 +7624,22 @@ impl<'a> Parser<'a> { } pub fn parse_procedure_param(&mut self) -> Result { + let mode = if self.parse_keyword(Keyword::IN) { + Some(ArgMode::In) + } else if self.parse_keyword(Keyword::OUT) { + Some(ArgMode::Out) + } else if self.parse_keyword(Keyword::INOUT) { + Some(ArgMode::InOut) + } else { + None + }; let name = self.parse_identifier()?; let data_type = self.parse_data_type()?; - Ok(ProcedureParam { name, data_type }) + Ok(ProcedureParam { + name, + data_type, + mode, + }) } pub fn parse_column_def(&mut self) -> Result { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 52054604d..d45da69ae 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15356,3 +15356,65 @@ fn check_enforced() { "CREATE TABLE t (a INT, b INT, c INT, CHECK (a > 0) NOT ENFORCED, CHECK (b > 0) ENFORCED, CHECK (c > 0))", ); } + +#[test] +fn parse_create_procedure_with_parameter_modes() { + let sql = r#"CREATE PROCEDURE test_proc (IN a INTEGER, OUT b TEXT, INOUT c TIMESTAMP, d BOOL) AS BEGIN SELECT 1; END"#; + match verified_stmt(sql) { + Statement::CreateProcedure { + or_alter, + name, + params, + .. + } => { + assert_eq!(or_alter, false); + assert_eq!(name.to_string(), "test_proc"); + let fake_span = Span { + start: Location { line: 0, column: 0 }, + end: Location { line: 0, column: 0 }, + }; + assert_eq!( + params, + Some(vec![ + ProcedureParam { + name: Ident { + value: "a".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Integer(None), + mode: Some(ArgMode::In) + }, + ProcedureParam { + name: Ident { + value: "b".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Text, + mode: Some(ArgMode::Out) + }, + ProcedureParam { + name: Ident { + value: "c".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Timestamp(None, TimezoneInfo::None), + mode: Some(ArgMode::InOut) + }, + ProcedureParam { + name: Ident { + value: "d".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Bool, + mode: None + }, + ]) + ); + } + _ => unreachable!(), + } +} diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 2a3145028..8edb100aa 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -153,7 +153,8 @@ fn parse_create_procedure() { quote_style: None, span: Span::empty(), }, - data_type: DataType::Int(None) + data_type: DataType::Int(None), + mode: None, }, ProcedureParam { name: Ident { @@ -164,7 +165,8 @@ fn parse_create_procedure() { data_type: DataType::Varchar(Some(CharacterLength::IntegerLength { length: 256, unit: None - })) + })), + mode: None, } ]), name: ObjectName::from(vec![Ident { From b2ab0061c172aa55e06ece592309a96083d2220d Mon Sep 17 00:00:00 2001 From: Elia Perantoni Date: Wed, 25 Jun 2025 12:21:59 +0200 Subject: [PATCH 165/286] Fix `impl Ord for Ident` (#1893) --- src/ast/mod.rs | 48 ++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 46 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 17464f4ac..ef0c4dc92 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -28,6 +28,7 @@ use helpers::{ stmt_data_loading::{FileStagingCommand, StageLoadSelectItemKind}, }; +use core::cmp::Ordering; use core::ops::Deref; use core::{ fmt::{self, Display}, @@ -172,7 +173,7 @@ fn format_statement_list(f: &mut fmt::Formatter, statements: &[Statement]) -> fm } /// An identifier, decomposed into its value or character data and the quote style. -#[derive(Debug, Clone, PartialOrd, Ord)] +#[derive(Debug, Clone)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct Ident { @@ -214,6 +215,35 @@ impl core::hash::Hash for Ident { impl Eq for Ident {} +impl PartialOrd for Ident { + fn partial_cmp(&self, other: &Self) -> Option { + Some(self.cmp(other)) + } +} + +impl Ord for Ident { + fn cmp(&self, other: &Self) -> Ordering { + let Ident { + value, + quote_style, + // exhaustiveness check; we ignore spans in ordering + span: _, + } = self; + + let Ident { + value: other_value, + quote_style: other_quote_style, + // exhaustiveness check; we ignore spans in ordering + span: _, + } = other; + + // First compare by value, then by quote_style + value + .cmp(other_value) + .then_with(|| quote_style.cmp(other_quote_style)) + } +} + impl Ident { /// Create a new identifier with the given value and no quotes and an empty span. pub fn new(value: S) -> Self @@ -4214,7 +4244,7 @@ pub enum Statement { /// ```sql /// NOTIFY channel [ , payload ] /// ``` - /// send a notification event together with an optional “payload” string to channel + /// send a notification event together with an optional "payload" string to channel /// /// See Postgres NOTIFY { @@ -9771,6 +9801,8 @@ impl fmt::Display for NullInclusion { #[cfg(test)] mod tests { + use crate::tokenizer::Location; + use super::*; #[test] @@ -10066,4 +10098,16 @@ mod tests { test_steps(OneOrManyWithParens::Many(vec![2]), vec![2], 3); test_steps(OneOrManyWithParens::Many(vec![3, 4]), vec![3, 4], 4); } + + // Tests that the position in the code of an `Ident` does not affect its + // ordering. + #[test] + fn test_ident_ord() { + let mut a = Ident::with_span(Span::new(Location::new(1, 1), Location::new(1, 1)), "a"); + let mut b = Ident::with_span(Span::new(Location::new(2, 2), Location::new(2, 2)), "b"); + + assert!(a < b); + std::mem::swap(&mut a.span, &mut b.span); + assert!(a < b); + } } From 1bbc05cdff7952182e2cdafe55091de0f1f7b3ce Mon Sep 17 00:00:00 2001 From: Elia Perantoni Date: Wed, 25 Jun 2025 16:10:01 +0200 Subject: [PATCH 166/286] Snowflake: support multiple column options in `CREATE VIEW` (#1891) --- src/ast/ddl.rs | 28 ++++++++++++++++++++++++++-- src/ast/mod.rs | 15 ++++++++------- src/ast/spans.rs | 17 +++++++++++------ src/dialect/mod.rs | 4 ++++ src/dialect/snowflake.rs | 4 ++++ src/parser/mod.rs | 31 ++++++++++++++++++++----------- tests/sqlparser_bigquery.rs | 18 ++++++++++-------- tests/sqlparser_clickhouse.rs | 4 ++-- tests/sqlparser_common.rs | 2 +- tests/sqlparser_snowflake.rs | 15 ++++++++++++--- 10 files changed, 98 insertions(+), 40 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index f81c6fc20..d2863c3a1 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1426,7 +1426,24 @@ impl fmt::Display for ColumnDef { pub struct ViewColumnDef { pub name: Ident, pub data_type: Option, - pub options: Option>, + pub options: Option, +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ColumnOptions { + CommaSeparated(Vec), + SpaceSeparated(Vec), +} + +impl ColumnOptions { + pub fn as_slice(&self) -> &[ColumnOption] { + match self { + ColumnOptions::CommaSeparated(options) => options.as_slice(), + ColumnOptions::SpaceSeparated(options) => options.as_slice(), + } + } } impl fmt::Display for ViewColumnDef { @@ -1436,7 +1453,14 @@ impl fmt::Display for ViewColumnDef { write!(f, " {}", data_type)?; } if let Some(options) = self.options.as_ref() { - write!(f, " {}", display_comma_separated(options.as_slice()))?; + match options { + ColumnOptions::CommaSeparated(column_options) => { + write!(f, " {}", display_comma_separated(column_options.as_slice()))?; + } + ColumnOptions::SpaceSeparated(column_options) => { + write!(f, " {}", display_separated(column_options.as_slice(), " "))? + } + } } Ok(()) } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index ef0c4dc92..bfed91fb6 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -61,13 +61,14 @@ pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, AlterTableAlgorithm, AlterTableLock, AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, - ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, - ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, Deduplicate, - DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, - IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, - IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, - ProcedureParam, ReferentialAction, ReplicaIdentity, TableConstraint, TagsColumnOption, - UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, + ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, + ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, + Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, + IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, + IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, + Partition, ProcedureParam, ReferentialAction, ReplicaIdentity, TableConstraint, + TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, + ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index ca321cc2e..78ed772bd 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -15,7 +15,7 @@ // specific language governing permissions and limitations // under the License. -use crate::ast::query::SelectItemQualifiedWildcardKind; +use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions}; use core::iter; use crate::tokenizer::Span; @@ -991,10 +991,13 @@ impl Spanned for ViewColumnDef { options, } = self; - union_spans( - core::iter::once(name.span) - .chain(options.iter().flat_map(|i| i.iter().map(|k| k.span()))), - ) + name.span.union_opt(&options.as_ref().map(|o| o.span())) + } +} + +impl Spanned for ColumnOptions { + fn span(&self) -> Span { + union_spans(self.as_slice().iter().map(|i| i.span())) } } @@ -1055,7 +1058,9 @@ impl Spanned for CreateTableOptions { match self { CreateTableOptions::None => Span::empty(), CreateTableOptions::With(vec) => union_spans(vec.iter().map(|i| i.span())), - CreateTableOptions::Options(vec) => union_spans(vec.iter().map(|i| i.span())), + CreateTableOptions::Options(vec) => { + union_spans(vec.as_slice().iter().map(|i| i.span())) + } CreateTableOptions::Plain(vec) => union_spans(vec.iter().map(|i| i.span())), CreateTableOptions::TableProperties(vec) => union_spans(vec.iter().map(|i| i.span())), } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index a4c899e6b..bc92948d1 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1028,6 +1028,10 @@ pub trait Dialect: Debug + Any { fn supports_set_names(&self) -> bool { false } + + fn supports_space_separated_column_options(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 66e04ac24..5ebb7e37c 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -356,6 +356,10 @@ impl Dialect for SnowflakeDialect { fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] { &RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR } + + fn supports_space_separated_column_options(&self) -> bool { + true + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ca658d7da..44ba0bb7c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10579,17 +10579,7 @@ impl<'a> Parser<'a> { /// Parses a column definition within a view. fn parse_view_column(&mut self) -> Result { let name = self.parse_identifier()?; - let options = if (dialect_of!(self is BigQueryDialect | GenericDialect) - && self.parse_keyword(Keyword::OPTIONS)) - || (dialect_of!(self is SnowflakeDialect | GenericDialect) - && self.parse_keyword(Keyword::COMMENT)) - { - self.prev_token(); - self.parse_optional_column_option()? - .map(|option| vec![option]) - } else { - None - }; + let options = self.parse_view_column_options()?; let data_type = if dialect_of!(self is ClickHouseDialect) { Some(self.parse_data_type()?) } else { @@ -10602,6 +10592,25 @@ impl<'a> Parser<'a> { }) } + fn parse_view_column_options(&mut self) -> Result, ParserError> { + let mut options = Vec::new(); + loop { + let option = self.parse_optional_column_option()?; + if let Some(option) = option { + options.push(option); + } else { + break; + } + } + if options.is_empty() { + Ok(None) + } else if self.dialect.supports_space_separated_column_options() { + Ok(Some(ColumnOptions::SpaceSeparated(options))) + } else { + Ok(Some(ColumnOptions::CommaSeparated(options))) + } + } + /// Parses a parenthesized comma-separated list of unqualified, possibly quoted identifiers. /// For example: `(col1, "col 2", ...)` pub fn parse_parenthesized_column_list( diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 0de0b12ba..2bcdb4e59 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -355,14 +355,16 @@ fn parse_create_view_with_options() { ViewColumnDef { name: Ident::new("age"), data_type: None, - options: Some(vec![ColumnOption::Options(vec![SqlOption::KeyValue { - key: Ident::new("description"), - value: Expr::Value( - Value::DoubleQuotedString("field age".to_string()).with_span( - Span::new(Location::new(1, 42), Location::new(1, 52)) - ) - ), - }])]), + options: Some(ColumnOptions::CommaSeparated(vec![ColumnOption::Options( + vec![SqlOption::KeyValue { + key: Ident::new("description"), + value: Expr::Value( + Value::DoubleQuotedString("field age".to_string()).with_span( + Span::new(Location::new(1, 42), Location::new(1, 52)) + ) + ), + }] + )])), }, ], columns diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 93b4c4f59..ed5d7ad2f 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -914,7 +914,7 @@ fn parse_create_view_with_fields_data_types() { }]), vec![] )), - options: None + options: None, }, ViewColumnDef { name: "f".into(), @@ -926,7 +926,7 @@ fn parse_create_view_with_fields_data_types() { }]), vec![] )), - options: None + options: None, }, ] ); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index d45da69ae..3c20f82a6 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7990,7 +7990,7 @@ fn parse_create_view_with_columns() { .map(|name| ViewColumnDef { name, data_type: None, - options: None + options: None, }) .collect::>() ); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 7dc00f9a8..8b3988d92 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3124,7 +3124,7 @@ fn view_comment_option_should_be_after_column_list() { "CREATE OR REPLACE VIEW v (a COMMENT 'a comment', b, c COMMENT 'c comment') COMMENT = 'Comment' AS SELECT a FROM t", "CREATE OR REPLACE VIEW v (a COMMENT 'a comment', b, c COMMENT 'c comment') WITH (foo = bar) COMMENT = 'Comment' AS SELECT a FROM t", ] { - snowflake_and_generic() + snowflake() .verified_stmt(sql); } } @@ -3133,7 +3133,7 @@ fn view_comment_option_should_be_after_column_list() { fn parse_view_column_descriptions() { let sql = "CREATE OR REPLACE VIEW v (a COMMENT 'Comment', b) AS SELECT a, b FROM table1"; - match snowflake_and_generic().verified_stmt(sql) { + match snowflake().verified_stmt(sql) { Statement::CreateView { name, columns, .. } => { assert_eq!(name.to_string(), "v"); assert_eq!( @@ -3142,7 +3142,9 @@ fn parse_view_column_descriptions() { ViewColumnDef { name: Ident::new("a"), data_type: None, - options: Some(vec![ColumnOption::Comment("Comment".to_string())]), + options: Some(ColumnOptions::SpaceSeparated(vec![ColumnOption::Comment( + "Comment".to_string() + )])), }, ViewColumnDef { name: Ident::new("b"), @@ -4165,3 +4167,10 @@ fn test_snowflake_fetch_clause_syntax() { canonical, ); } + +#[test] +fn test_snowflake_create_view_with_multiple_column_options() { + let create_view_with_tag = + r#"CREATE VIEW X (COL WITH TAG (pii='email') COMMENT 'foobar') AS SELECT * FROM Y"#; + snowflake().verified_stmt(create_view_with_tag); +} From 95d16e3b2d08ba74ada44dbcfe4ece3400b2d89b Mon Sep 17 00:00:00 2001 From: ZacJW Date: Fri, 27 Jun 2025 17:22:21 +0100 Subject: [PATCH 167/286] Add support for `LANGUAGE` clause in `CREATE PROCEDURE` (#1903) --- src/ast/mod.rs | 6 ++++++ src/parser/mod.rs | 8 ++++++++ tests/sqlparser_common.rs | 30 ++++++++++++++++++++++++++++++ tests/sqlparser_mssql.rs | 3 ++- 4 files changed, 46 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index bfed91fb6..1ea0b5c26 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3945,6 +3945,7 @@ pub enum Statement { or_alter: bool, name: ObjectName, params: Option>, + language: Option, body: ConditionalStatements, }, /// ```sql @@ -4848,6 +4849,7 @@ impl fmt::Display for Statement { name, or_alter, params, + language, body, } => { write!( @@ -4863,6 +4865,10 @@ impl fmt::Display for Statement { } } + if let Some(language) = language { + write!(f, " LANGUAGE {language}")?; + } + write!(f, " AS {body}") } Statement::CreateMacro { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 44ba0bb7c..e09b7e338 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15802,6 +15802,13 @@ impl<'a> Parser<'a> { pub fn parse_create_procedure(&mut self, or_alter: bool) -> Result { let name = self.parse_object_name(false)?; let params = self.parse_optional_procedure_parameters()?; + + let language = if self.parse_keyword(Keyword::LANGUAGE) { + Some(self.parse_identifier()?) + } else { + None + }; + self.expect_keyword_is(Keyword::AS)?; let body = self.parse_conditional_statements(&[Keyword::END])?; @@ -15810,6 +15817,7 @@ impl<'a> Parser<'a> { name, or_alter, params, + language, body, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 3c20f82a6..93cc89e55 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15357,6 +15357,36 @@ fn check_enforced() { ); } +#[test] +fn parse_create_procedure_with_language() { + let sql = r#"CREATE PROCEDURE test_proc LANGUAGE sql AS BEGIN SELECT 1; END"#; + match verified_stmt(sql) { + Statement::CreateProcedure { + or_alter, + name, + params, + language, + .. + } => { + assert_eq!(or_alter, false); + assert_eq!(name.to_string(), "test_proc"); + assert_eq!(params, Some(vec![])); + assert_eq!( + language, + Some(Ident { + value: "sql".into(), + quote_style: None, + span: Span { + start: Location::empty(), + end: Location::empty() + } + }) + ); + } + _ => unreachable!(), + } +} + #[test] fn parse_create_procedure_with_parameter_modes() { let sql = r#"CREATE PROCEDURE test_proc (IN a INTEGER, OUT b TEXT, INOUT c TIMESTAMP, d BOOL) AS BEGIN SELECT 1; END"#; diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 8edb100aa..9ec28f421 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -173,7 +173,8 @@ fn parse_create_procedure() { value: "test".into(), quote_style: None, span: Span::empty(), - }]) + }]), + language: None, } ) } From 5f2b5fe7beb0698663f9ba9572c711c6bee9ef6e Mon Sep 17 00:00:00 2001 From: Ifeanyi Ubah Date: Fri, 27 Jun 2025 20:21:17 +0200 Subject: [PATCH 168/286] Fix clippy lints on 1.88.0 (#1910) --- examples/cli.rs | 2 +- sqlparser_bench/benches/sqlparser_bench.rs | 11 ++- src/ast/data_type.rs | 16 ++-- src/ast/dcl.rs | 18 ++-- src/ast/ddl.rs | 28 +++---- src/ast/dml.rs | 12 +-- src/ast/helpers/key_value_options.rs | 2 +- src/ast/mod.rs | 96 +++++++++++----------- src/ast/query.rs | 76 ++++++++--------- src/ast/value.rs | 6 +- src/ast/visitor.rs | 4 +- src/dialect/mod.rs | 2 +- src/dialect/postgresql.rs | 2 +- src/parser/mod.rs | 13 ++- src/test_utils.rs | 2 +- src/tokenizer.rs | 10 +-- tests/sqlparser_clickhouse.rs | 6 +- tests/sqlparser_common.rs | 20 ++--- tests/sqlparser_databricks.rs | 12 +-- tests/sqlparser_duckdb.rs | 10 ++- tests/sqlparser_hive.rs | 4 +- tests/sqlparser_mssql.rs | 10 +-- tests/sqlparser_mysql.rs | 47 ++++++----- tests/sqlparser_postgres.rs | 11 ++- tests/sqlparser_snowflake.rs | 32 ++++---- tests/sqlparser_sqlite.rs | 6 +- 26 files changed, 226 insertions(+), 232 deletions(-) diff --git a/examples/cli.rs b/examples/cli.rs index 0252fca74..08a40a6dd 100644 --- a/examples/cli.rs +++ b/examples/cli.rs @@ -63,7 +63,7 @@ $ cargo run --example cli - [--dialectname] }; let contents = if filename == "-" { - println!("Parsing from stdin using {:?}", dialect); + println!("Parsing from stdin using {dialect:?}"); let mut buf = Vec::new(); stdin() .read_to_end(&mut buf) diff --git a/sqlparser_bench/benches/sqlparser_bench.rs b/sqlparser_bench/benches/sqlparser_bench.rs index 24c59c076..6132ee432 100644 --- a/sqlparser_bench/benches/sqlparser_bench.rs +++ b/sqlparser_bench/benches/sqlparser_bench.rs @@ -45,25 +45,24 @@ fn basic_queries(c: &mut Criterion) { let large_statement = { let expressions = (0..1000) - .map(|n| format!("FN_{}(COL_{})", n, n)) + .map(|n| format!("FN_{n}(COL_{n})")) .collect::>() .join(", "); let tables = (0..1000) - .map(|n| format!("TABLE_{}", n)) + .map(|n| format!("TABLE_{n}")) .collect::>() .join(" JOIN "); let where_condition = (0..1000) - .map(|n| format!("COL_{} = {}", n, n)) + .map(|n| format!("COL_{n} = {n}")) .collect::>() .join(" OR "); let order_condition = (0..1000) - .map(|n| format!("COL_{} DESC", n)) + .map(|n| format!("COL_{n} DESC")) .collect::>() .join(", "); format!( - "SELECT {} FROM {} WHERE {} ORDER BY {}", - expressions, tables, where_condition, order_condition + "SELECT {expressions} FROM {tables} WHERE {where_condition} ORDER BY {order_condition}" ) }; diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index ef04b7e4b..0897f2db7 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -666,7 +666,7 @@ impl fmt::Display for DataType { } DataType::Enum(vals, bits) => { match bits { - Some(bits) => write!(f, "ENUM{}", bits), + Some(bits) => write!(f, "ENUM{bits}"), None => write!(f, "ENUM"), }?; write!(f, "(")?; @@ -714,16 +714,16 @@ impl fmt::Display for DataType { } // ClickHouse DataType::Nullable(data_type) => { - write!(f, "Nullable({})", data_type) + write!(f, "Nullable({data_type})") } DataType::FixedString(character_length) => { - write!(f, "FixedString({})", character_length) + write!(f, "FixedString({character_length})") } DataType::LowCardinality(data_type) => { - write!(f, "LowCardinality({})", data_type) + write!(f, "LowCardinality({data_type})") } DataType::Map(key_data_type, value_data_type) => { - write!(f, "Map({}, {})", key_data_type, value_data_type) + write!(f, "Map({key_data_type}, {value_data_type})") } DataType::Tuple(fields) => { write!(f, "Tuple({})", display_comma_separated(fields)) @@ -745,7 +745,7 @@ impl fmt::Display for DataType { DataType::NamedTable { name, columns } => { write!(f, "{} TABLE ({})", name, display_comma_separated(columns)) } - DataType::GeometricType(kind) => write!(f, "{}", kind), + DataType::GeometricType(kind) => write!(f, "{kind}"), DataType::TsVector => write!(f, "TSVECTOR"), DataType::TsQuery => write!(f, "TSQUERY"), } @@ -942,7 +942,7 @@ impl fmt::Display for CharacterLength { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { CharacterLength::IntegerLength { length, unit } => { - write!(f, "{}", length)?; + write!(f, "{length}")?; if let Some(unit) = unit { write!(f, " {unit}")?; } @@ -997,7 +997,7 @@ impl fmt::Display for BinaryLength { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { BinaryLength::IntegerLength { length } => { - write!(f, "{}", length)?; + write!(f, "{length}")?; } BinaryLength::Max => { write!(f, "MAX")?; diff --git a/src/ast/dcl.rs b/src/ast/dcl.rs index 735ab0cce..079894075 100644 --- a/src/ast/dcl.rs +++ b/src/ast/dcl.rs @@ -173,7 +173,7 @@ impl fmt::Display for AlterRoleOperation { in_database, } => { if let Some(database_name) = in_database { - write!(f, "IN DATABASE {} ", database_name)?; + write!(f, "IN DATABASE {database_name} ")?; } match config_value { @@ -187,7 +187,7 @@ impl fmt::Display for AlterRoleOperation { in_database, } => { if let Some(database_name) = in_database { - write!(f, "IN DATABASE {} ", database_name)?; + write!(f, "IN DATABASE {database_name} ")?; } match config_name { @@ -218,15 +218,15 @@ impl fmt::Display for Use { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { f.write_str("USE ")?; match self { - Use::Catalog(name) => write!(f, "CATALOG {}", name), - Use::Schema(name) => write!(f, "SCHEMA {}", name), - Use::Database(name) => write!(f, "DATABASE {}", name), - Use::Warehouse(name) => write!(f, "WAREHOUSE {}", name), - Use::Role(name) => write!(f, "ROLE {}", name), + Use::Catalog(name) => write!(f, "CATALOG {name}"), + Use::Schema(name) => write!(f, "SCHEMA {name}"), + Use::Database(name) => write!(f, "DATABASE {name}"), + Use::Warehouse(name) => write!(f, "WAREHOUSE {name}"), + Use::Role(name) => write!(f, "ROLE {name}"), Use::SecondaryRoles(secondary_roles) => { - write!(f, "SECONDARY ROLES {}", secondary_roles) + write!(f, "SECONDARY ROLES {secondary_roles}") } - Use::Object(name) => write!(f, "{}", name), + Use::Object(name) => write!(f, "{name}"), Use::Default => write!(f, "DEFAULT"), } } diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index d2863c3a1..afe9f228c 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -57,7 +57,7 @@ impl fmt::Display for ReplicaIdentity { ReplicaIdentity::None => f.write_str("NONE"), ReplicaIdentity::Full => f.write_str("FULL"), ReplicaIdentity::Default => f.write_str("DEFAULT"), - ReplicaIdentity::Index(idx) => write!(f, "USING INDEX {}", idx), + ReplicaIdentity::Index(idx) => write!(f, "USING INDEX {idx}"), } } } @@ -450,7 +450,7 @@ pub enum Owner { impl fmt::Display for Owner { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - Owner::Ident(ident) => write!(f, "{}", ident), + Owner::Ident(ident) => write!(f, "{ident}"), Owner::CurrentRole => write!(f, "CURRENT_ROLE"), Owner::CurrentUser => write!(f, "CURRENT_USER"), Owner::SessionUser => write!(f, "SESSION_USER"), @@ -525,7 +525,7 @@ impl fmt::Display for AlterTableOperation { if *if_not_exists { write!(f, " IF NOT EXISTS")?; } - write!(f, " {} ({})", name, query) + write!(f, " {name} ({query})") } AlterTableOperation::Algorithm { equals, algorithm } => { write!( @@ -540,7 +540,7 @@ impl fmt::Display for AlterTableOperation { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {}", name) + write!(f, " {name}") } AlterTableOperation::MaterializeProjection { if_exists, @@ -551,9 +551,9 @@ impl fmt::Display for AlterTableOperation { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {}", name)?; + write!(f, " {name}")?; if let Some(partition) = partition { - write!(f, " IN PARTITION {}", partition)?; + write!(f, " IN PARTITION {partition}")?; } Ok(()) } @@ -566,9 +566,9 @@ impl fmt::Display for AlterTableOperation { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {}", name)?; + write!(f, " {name}")?; if let Some(partition) = partition { - write!(f, " IN PARTITION {}", partition)?; + write!(f, " IN PARTITION {partition}")?; } Ok(()) } @@ -1168,7 +1168,7 @@ impl fmt::Display for TableConstraint { write!(f, " ON UPDATE {action}")?; } if let Some(characteristics) = characteristics { - write!(f, " {}", characteristics)?; + write!(f, " {characteristics}")?; } Ok(()) } @@ -1308,7 +1308,7 @@ impl fmt::Display for IndexType { Self::SPGiST => write!(f, "SPGIST"), Self::BRIN => write!(f, "BRIN"), Self::Bloom => write!(f, "BLOOM"), - Self::Custom(name) => write!(f, "{}", name), + Self::Custom(name) => write!(f, "{name}"), } } } @@ -1450,7 +1450,7 @@ impl fmt::Display for ViewColumnDef { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.name)?; if let Some(data_type) = self.data_type.as_ref() { - write!(f, " {}", data_type)?; + write!(f, " {data_type}")?; } if let Some(options) = self.options.as_ref() { match options { @@ -1845,7 +1845,7 @@ impl fmt::Display for ColumnOption { } => { write!(f, "{}", if *is_primary { "PRIMARY KEY" } else { "UNIQUE" })?; if let Some(characteristics) = characteristics { - write!(f, " {}", characteristics)?; + write!(f, " {characteristics}")?; } Ok(()) } @@ -1867,7 +1867,7 @@ impl fmt::Display for ColumnOption { write!(f, " ON UPDATE {action}")?; } if let Some(characteristics) = characteristics { - write!(f, " {}", characteristics)?; + write!(f, " {characteristics}")?; } Ok(()) } @@ -1927,7 +1927,7 @@ impl fmt::Display for ColumnOption { write!(f, "{parameters}") } OnConflict(keyword) => { - write!(f, "ON CONFLICT {:?}", keyword)?; + write!(f, "ON CONFLICT {keyword:?}")?; Ok(()) } Policy(parameters) => { diff --git a/src/ast/dml.rs b/src/ast/dml.rs index 292650c88..e179f5d70 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -55,7 +55,7 @@ impl Display for IndexColumn { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.column)?; if let Some(operator_class) = &self.operator_class { - write!(f, " {}", operator_class)?; + write!(f, " {operator_class}")?; } Ok(()) } @@ -266,7 +266,7 @@ impl Display for CreateTable { name = self.name, )?; if let Some(on_cluster) = &self.on_cluster { - write!(f, " ON CLUSTER {}", on_cluster)?; + write!(f, " ON CLUSTER {on_cluster}")?; } if !self.columns.is_empty() || !self.constraints.is_empty() { f.write_str(" (")?; @@ -383,15 +383,15 @@ impl Display for CreateTable { match &self.table_options { options @ CreateTableOptions::With(_) | options @ CreateTableOptions::Plain(_) - | options @ CreateTableOptions::TableProperties(_) => write!(f, " {}", options)?, + | options @ CreateTableOptions::TableProperties(_) => write!(f, " {options}")?, _ => (), } if let Some(primary_key) = &self.primary_key { - write!(f, " PRIMARY KEY {}", primary_key)?; + write!(f, " PRIMARY KEY {primary_key}")?; } if let Some(order_by) = &self.order_by { - write!(f, " ORDER BY {}", order_by)?; + write!(f, " ORDER BY {order_by}")?; } if let Some(inherits) = &self.inherits { write!(f, " INHERITS ({})", display_comma_separated(inherits))?; @@ -403,7 +403,7 @@ impl Display for CreateTable { write!(f, " CLUSTER BY {cluster_by}")?; } if let options @ CreateTableOptions::Options(_) = &self.table_options { - write!(f, " {}", options)?; + write!(f, " {options}")?; } if let Some(external_volume) = self.external_volume.as_ref() { write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; diff --git a/src/ast/helpers/key_value_options.rs b/src/ast/helpers/key_value_options.rs index 06f028dd2..796bfd5e3 100644 --- a/src/ast/helpers/key_value_options.rs +++ b/src/ast/helpers/key_value_options.rs @@ -67,7 +67,7 @@ impl fmt::Display for KeyValueOptions { } else { f.write_str(" ")?; } - write!(f, "{}", option)?; + write!(f, "{option}")?; } } Ok(()) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 1ea0b5c26..0f6820626 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -357,7 +357,7 @@ impl ObjectNamePart { impl fmt::Display for ObjectNamePart { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - ObjectNamePart::Identifier(ident) => write!(f, "{}", ident), + ObjectNamePart::Identifier(ident) => write!(f, "{ident}"), } } } @@ -1210,8 +1210,8 @@ pub enum AccessExpr { impl fmt::Display for AccessExpr { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { - AccessExpr::Dot(expr) => write!(f, ".{}", expr), - AccessExpr::Subscript(subscript) => write!(f, "[{}]", subscript), + AccessExpr::Dot(expr) => write!(f, ".{expr}"), + AccessExpr::Subscript(subscript) => write!(f, "[{subscript}]"), } } } @@ -1413,12 +1413,12 @@ impl fmt::Display for Expr { match self { Expr::Identifier(s) => write!(f, "{s}"), Expr::Wildcard(_) => f.write_str("*"), - Expr::QualifiedWildcard(prefix, _) => write!(f, "{}.*", prefix), + Expr::QualifiedWildcard(prefix, _) => write!(f, "{prefix}.*"), Expr::CompoundIdentifier(s) => write!(f, "{}", display_separated(s, ".")), Expr::CompoundFieldAccess { root, access_chain } => { - write!(f, "{}", root)?; + write!(f, "{root}")?; for field in access_chain { - write!(f, "{}", field)?; + write!(f, "{field}")?; } Ok(()) } @@ -1547,7 +1547,7 @@ impl fmt::Display for Expr { } => { let not_ = if *negated { "NOT " } else { "" }; if form.is_none() { - write!(f, "{} IS {}NORMALIZED", expr, not_) + write!(f, "{expr} IS {not_}NORMALIZED") } else { write!( f, @@ -1869,7 +1869,7 @@ impl fmt::Display for Expr { } } Expr::Named { expr, name } => { - write!(f, "{} AS {}", expr, name) + write!(f, "{expr} AS {name}") } Expr::Dictionary(fields) => { write!(f, "{{{}}}", display_comma_separated(fields)) @@ -2425,7 +2425,7 @@ impl fmt::Display for ConditionalStatements { } Ok(()) } - ConditionalStatements::BeginEnd(bes) => write!(f, "{}", bes), + ConditionalStatements::BeginEnd(bes) => write!(f, "{bes}"), } } } @@ -2945,9 +2945,7 @@ impl Display for Set { write!( f, "SET {modifier}ROLE {role_name}", - modifier = context_modifier - .map(|m| format!("{}", m)) - .unwrap_or_default() + modifier = context_modifier.map(|m| format!("{m}")).unwrap_or_default() ) } Self::SetSessionParam(kind) => write!(f, "SET {kind}"), @@ -2980,7 +2978,7 @@ impl Display for Set { charset_name, collation_name, } => { - write!(f, "SET NAMES {}", charset_name)?; + write!(f, "SET NAMES {charset_name}")?; if let Some(collation) = collation_name { f.write_str(" COLLATE ")?; @@ -3003,7 +3001,7 @@ impl Display for Set { write!( f, "SET {}{}{} = {}", - scope.map(|s| format!("{}", s)).unwrap_or_default(), + scope.map(|s| format!("{s}")).unwrap_or_default(), if *hivevar { "HIVEVAR:" } else { "" }, variable, display_comma_separated(values) @@ -4405,7 +4403,7 @@ impl fmt::Display for Statement { write!(f, "{describe_alias} ")?; if let Some(format) = hive_format { - write!(f, "{} ", format)?; + write!(f, "{format} ")?; } if *has_table_keyword { write!(f, "TABLE ")?; @@ -5241,7 +5239,7 @@ impl fmt::Display for Statement { if *only { write!(f, "ONLY ")?; } - write!(f, "{name} ", name = name)?; + write!(f, "{name} ")?; if let Some(cluster) = on_cluster { write!(f, "ON CLUSTER {cluster} ")?; } @@ -5319,7 +5317,7 @@ impl fmt::Display for Statement { )?; if !session_params.options.is_empty() { if *set { - write!(f, " {}", session_params)?; + write!(f, " {session_params}")?; } else { let options = session_params .options @@ -5353,7 +5351,7 @@ impl fmt::Display for Statement { if *purge { " PURGE" } else { "" }, )?; if let Some(table_name) = table.as_ref() { - write!(f, " ON {}", table_name)?; + write!(f, " ON {table_name}")?; }; Ok(()) } @@ -5608,7 +5606,7 @@ impl fmt::Display for Statement { } => { if *syntax_begin { if let Some(modifier) = *modifier { - write!(f, "BEGIN {}", modifier)?; + write!(f, "BEGIN {modifier}")?; } else { write!(f, "BEGIN")?; } @@ -5644,7 +5642,7 @@ impl fmt::Display for Statement { if *end_syntax { write!(f, "END")?; if let Some(modifier) = *modifier { - write!(f, " {}", modifier)?; + write!(f, " {modifier}")?; } if *chain { write!(f, " AND CHAIN")?; @@ -5743,7 +5741,7 @@ impl fmt::Display for Statement { write!(f, " GRANTED BY {grantor}")?; } if let Some(cascade) = cascade { - write!(f, " {}", cascade)?; + write!(f, " {cascade}")?; } Ok(()) } @@ -5922,13 +5920,13 @@ impl fmt::Display for Statement { if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, )?; if !directory_table_params.options.is_empty() { - write!(f, " DIRECTORY=({})", directory_table_params)?; + write!(f, " DIRECTORY=({directory_table_params})")?; } if !file_format.options.is_empty() { - write!(f, " FILE_FORMAT=({})", file_format)?; + write!(f, " FILE_FORMAT=({file_format})")?; } if !copy_options.options.is_empty() { - write!(f, " COPY_OPTIONS=({})", copy_options)?; + write!(f, " COPY_OPTIONS=({copy_options})")?; } if comment.is_some() { write!(f, " COMMENT='{}'", comment.as_ref().unwrap())?; @@ -5951,7 +5949,7 @@ impl fmt::Display for Statement { validation_mode, partition, } => { - write!(f, "COPY INTO {}", into)?; + write!(f, "COPY INTO {into}")?; if let Some(into_columns) = into_columns { write!(f, " ({})", display_comma_separated(into_columns))?; } @@ -5967,12 +5965,12 @@ impl fmt::Display for Statement { )?; } if let Some(from_obj_alias) = from_obj_alias { - write!(f, " AS {}", from_obj_alias)?; + write!(f, " AS {from_obj_alias}")?; } write!(f, ")")?; } else if let Some(from_obj) = from_obj { // Standard data load - write!(f, " FROM {}{}", from_obj, stage_params)?; + write!(f, " FROM {from_obj}{stage_params}")?; if let Some(from_obj_alias) = from_obj_alias { write!(f, " AS {from_obj_alias}")?; } @@ -5985,24 +5983,24 @@ impl fmt::Display for Statement { write!(f, " FILES = ('{}')", display_separated(files, "', '"))?; } if let Some(pattern) = pattern { - write!(f, " PATTERN = '{}'", pattern)?; + write!(f, " PATTERN = '{pattern}'")?; } if let Some(partition) = partition { write!(f, " PARTITION BY {partition}")?; } if !file_format.options.is_empty() { - write!(f, " FILE_FORMAT=({})", file_format)?; + write!(f, " FILE_FORMAT=({file_format})")?; } if !copy_options.options.is_empty() { match kind { CopyIntoSnowflakeKind::Table => { - write!(f, " COPY_OPTIONS=({})", copy_options)? + write!(f, " COPY_OPTIONS=({copy_options})")? } CopyIntoSnowflakeKind::Location => write!(f, " {copy_options}")?, } } if let Some(validation_mode) = validation_mode { - write!(f, " VALIDATION_MODE = {}", validation_mode)?; + write!(f, " VALIDATION_MODE = {validation_mode}")?; } Ok(()) } @@ -6048,10 +6046,10 @@ impl fmt::Display for Statement { } => { write!(f, "OPTIMIZE TABLE {name}")?; if let Some(on_cluster) = on_cluster { - write!(f, " ON CLUSTER {on_cluster}", on_cluster = on_cluster)?; + write!(f, " ON CLUSTER {on_cluster}")?; } if let Some(partition) = partition { - write!(f, " {partition}", partition = partition)?; + write!(f, " {partition}")?; } if *include_final { write!(f, " FINAL")?; @@ -6178,7 +6176,7 @@ impl fmt::Display for SetAssignment { write!( f, "{}{} = {}", - self.scope.map(|s| format!("{}", s)).unwrap_or_default(), + self.scope.map(|s| format!("{s}")).unwrap_or_default(), self.name, self.value ) @@ -6907,7 +6905,7 @@ impl fmt::Display for GranteeName { match self { GranteeName::ObjectName(name) => name.fmt(f), GranteeName::UserHost { user, host } => { - write!(f, "{}@{}", user, host) + write!(f, "{user}@{host}") } } } @@ -7077,7 +7075,7 @@ pub enum AssignmentTarget { impl fmt::Display for AssignmentTarget { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - AssignmentTarget::ColumnName(column) => write!(f, "{}", column), + AssignmentTarget::ColumnName(column) => write!(f, "{column}"), AssignmentTarget::Tuple(columns) => write!(f, "({})", display_comma_separated(columns)), } } @@ -7322,8 +7320,8 @@ impl fmt::Display for FunctionArguments { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { FunctionArguments::None => Ok(()), - FunctionArguments::Subquery(query) => write!(f, "({})", query), - FunctionArguments::List(args) => write!(f, "({})", args), + FunctionArguments::Subquery(query) => write!(f, "({query})"), + FunctionArguments::List(args) => write!(f, "({args})"), } } } @@ -7344,7 +7342,7 @@ pub struct FunctionArgumentList { impl fmt::Display for FunctionArgumentList { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { if let Some(duplicate_treatment) = self.duplicate_treatment { - write!(f, "{} ", duplicate_treatment)?; + write!(f, "{duplicate_treatment} ")?; } write!(f, "{}", display_comma_separated(&self.args))?; if !self.clauses.is_empty() { @@ -7404,7 +7402,7 @@ impl fmt::Display for FunctionArgumentClause { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { FunctionArgumentClause::IgnoreOrRespectNulls(null_treatment) => { - write!(f, "{}", null_treatment) + write!(f, "{null_treatment}") } FunctionArgumentClause::OrderBy(order_by) => { write!(f, "ORDER BY {}", display_comma_separated(order_by)) @@ -7860,12 +7858,12 @@ pub enum SqlOption { impl fmt::Display for SqlOption { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - SqlOption::Clustered(c) => write!(f, "{}", c), + SqlOption::Clustered(c) => write!(f, "{c}"), SqlOption::Ident(ident) => { - write!(f, "{}", ident) + write!(f, "{ident}") } SqlOption::KeyValue { key: name, value } => { - write!(f, "{} = {}", name, value) + write!(f, "{name} = {value}") } SqlOption::Partition { column_name, @@ -7905,7 +7903,7 @@ impl fmt::Display for SqlOption { SqlOption::NamedParenthesizedList(value) => { write!(f, "{} = ", value.key)?; if let Some(key) = &value.name { - write!(f, "{}", key)?; + write!(f, "{key}")?; } if !value.values.is_empty() { write!(f, "({})", display_comma_separated(&value.values))? @@ -7962,7 +7960,7 @@ impl fmt::Display for AttachDuckDBDatabaseOption { AttachDuckDBDatabaseOption::ReadOnly(Some(true)) => write!(f, "READ_ONLY true"), AttachDuckDBDatabaseOption::ReadOnly(Some(false)) => write!(f, "READ_ONLY false"), AttachDuckDBDatabaseOption::ReadOnly(None) => write!(f, "READ_ONLY"), - AttachDuckDBDatabaseOption::Type(t) => write!(f, "TYPE {}", t), + AttachDuckDBDatabaseOption::Type(t) => write!(f, "TYPE {t}"), } } } @@ -9485,10 +9483,10 @@ impl fmt::Display for ShowStatementIn { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.clause)?; if let Some(parent_type) = &self.parent_type { - write!(f, " {}", parent_type)?; + write!(f, " {parent_type}")?; } if let Some(parent_name) = &self.parent_name { - write!(f, " {}", parent_name)?; + write!(f, " {parent_name}")?; } Ok(()) } @@ -9569,7 +9567,7 @@ impl fmt::Display for TableObject { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { Self::TableName(table_name) => write!(f, "{table_name}"), - Self::TableFunction(func) => write!(f, "FUNCTION {}", func), + Self::TableFunction(func) => write!(f, "FUNCTION {func}"), } } } @@ -9757,7 +9755,7 @@ pub struct ReturnStatement { impl fmt::Display for ReturnStatement { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match &self.value { - Some(ReturnStatementValue::Expr(expr)) => write!(f, "RETURN {}", expr), + Some(ReturnStatementValue::Expr(expr)) => write!(f, "RETURN {expr}"), None => write!(f, "RETURN"), } } diff --git a/src/ast/query.rs b/src/ast/query.rs index 1fb93b6c6..c79ec110c 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1183,7 +1183,7 @@ impl fmt::Display for TableIndexHints { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{} {} ", self.hint_type, self.index_type)?; if let Some(for_clause) = &self.for_clause { - write!(f, "FOR {} ", for_clause)?; + write!(f, "FOR {for_clause} ")?; } write!(f, "({})", display_comma_separated(&self.index_names)) } @@ -1459,7 +1459,7 @@ impl fmt::Display for TableSampleQuantity { } write!(f, "{}", self.value)?; if let Some(unit) = &self.unit { - write!(f, " {}", unit)?; + write!(f, " {unit}")?; } if self.parenthesized { write!(f, ")")?; @@ -1552,7 +1552,7 @@ impl fmt::Display for TableSampleBucket { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { write!(f, "BUCKET {} OUT OF {}", self.bucket, self.total)?; if let Some(on) = &self.on { - write!(f, " ON {}", on)?; + write!(f, " ON {on}")?; } Ok(()) } @@ -1561,19 +1561,19 @@ impl fmt::Display for TableSample { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { write!(f, "{}", self.modifier)?; if let Some(name) = &self.name { - write!(f, " {}", name)?; + write!(f, " {name}")?; } if let Some(quantity) = &self.quantity { - write!(f, " {}", quantity)?; + write!(f, " {quantity}")?; } if let Some(seed) = &self.seed { - write!(f, " {}", seed)?; + write!(f, " {seed}")?; } if let Some(bucket) = &self.bucket { - write!(f, " ({})", bucket)?; + write!(f, " ({bucket})")?; } if let Some(offset) = &self.offset { - write!(f, " OFFSET {}", offset)?; + write!(f, " OFFSET {offset}")?; } Ok(()) } @@ -1651,7 +1651,7 @@ impl fmt::Display for RowsPerMatch { RowsPerMatch::AllRows(mode) => { write!(f, "ALL ROWS PER MATCH")?; if let Some(mode) = mode { - write!(f, " {}", mode)?; + write!(f, " {mode}")?; } Ok(()) } @@ -1777,7 +1777,7 @@ impl fmt::Display for MatchRecognizePattern { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { use MatchRecognizePattern::*; match self { - Symbol(symbol) => write!(f, "{}", symbol), + Symbol(symbol) => write!(f, "{symbol}"), Exclude(symbol) => write!(f, "{{- {symbol} -}}"), Permute(symbols) => write!(f, "PERMUTE({})", display_comma_separated(symbols)), Concat(patterns) => write!(f, "{}", display_separated(patterns, " ")), @@ -2148,7 +2148,7 @@ impl fmt::Display for TableAliasColumnDef { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.name)?; if let Some(ref data_type) = self.data_type { - write!(f, " {}", data_type)?; + write!(f, " {data_type}")?; } Ok(()) } @@ -2398,7 +2398,7 @@ impl fmt::Display for OrderBy { write!(f, " {}", display_comma_separated(exprs))?; } OrderByKind::All(all) => { - write!(f, " ALL{}", all)?; + write!(f, " ALL{all}")?; } } @@ -2429,7 +2429,7 @@ impl fmt::Display for OrderByExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}{}", self.expr, self.options)?; if let Some(ref with_fill) = self.with_fill { - write!(f, " {}", with_fill)? + write!(f, " {with_fill}")? } Ok(()) } @@ -2452,13 +2452,13 @@ impl fmt::Display for WithFill { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "WITH FILL")?; if let Some(ref from) = self.from { - write!(f, " FROM {}", from)?; + write!(f, " FROM {from}")?; } if let Some(ref to) = self.to { - write!(f, " TO {}", to)?; + write!(f, " TO {to}")?; } if let Some(ref step) = self.step { - write!(f, " STEP {}", step)?; + write!(f, " STEP {step}")?; } Ok(()) } @@ -2487,7 +2487,7 @@ impl fmt::Display for InterpolateExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.column)?; if let Some(ref expr) = self.expr { - write!(f, " AS {}", expr)?; + write!(f, " AS {expr}")?; } Ok(()) } @@ -2565,7 +2565,7 @@ impl fmt::Display for LimitClause { Ok(()) } LimitClause::OffsetCommaLimit { offset, limit } => { - write!(f, " LIMIT {}, {}", offset, limit) + write!(f, " LIMIT {offset}, {limit}") } } } @@ -2702,12 +2702,12 @@ impl fmt::Display for PipeOperator { write!(f, "DROP {}", display_comma_separated(columns.as_slice())) } PipeOperator::As { alias } => { - write!(f, "AS {}", alias) + write!(f, "AS {alias}") } PipeOperator::Limit { expr, offset } => { - write!(f, "LIMIT {}", expr)?; + write!(f, "LIMIT {expr}")?; if let Some(offset) = offset { - write!(f, " OFFSET {}", offset)?; + write!(f, " OFFSET {offset}")?; } Ok(()) } @@ -2730,14 +2730,14 @@ impl fmt::Display for PipeOperator { } PipeOperator::Where { expr } => { - write!(f, "WHERE {}", expr) + write!(f, "WHERE {expr}") } PipeOperator::OrderBy { exprs } => { write!(f, "ORDER BY {}", display_comma_separated(exprs.as_slice())) } PipeOperator::TableSample { sample } => { - write!(f, "{}", sample) + write!(f, "{sample}") } } } @@ -3016,7 +3016,7 @@ pub enum FormatClause { impl fmt::Display for FormatClause { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - FormatClause::Identifier(ident) => write!(f, "FORMAT {}", ident), + FormatClause::Identifier(ident) => write!(f, "FORMAT {ident}"), FormatClause::Null => write!(f, "FORMAT NULL"), } } @@ -3078,9 +3078,9 @@ impl fmt::Display for ForClause { without_array_wrapper, } => { write!(f, "FOR JSON ")?; - write!(f, "{}", for_json)?; + write!(f, "{for_json}")?; if let Some(root) = root { - write!(f, ", ROOT('{}')", root)?; + write!(f, ", ROOT('{root}')")?; } if *include_null_values { write!(f, ", INCLUDE_NULL_VALUES")?; @@ -3098,7 +3098,7 @@ impl fmt::Display for ForClause { r#type, } => { write!(f, "FOR XML ")?; - write!(f, "{}", for_xml)?; + write!(f, "{for_xml}")?; if *binary_base64 { write!(f, ", BINARY BASE64")?; } @@ -3106,7 +3106,7 @@ impl fmt::Display for ForClause { write!(f, ", TYPE")?; } if let Some(root) = root { - write!(f, ", ROOT('{}')", root)?; + write!(f, ", ROOT('{root}')")?; } if *elements { write!(f, ", ELEMENTS")?; @@ -3133,7 +3133,7 @@ impl fmt::Display for ForXml { ForXml::Raw(root) => { write!(f, "RAW")?; if let Some(root) = root { - write!(f, "('{}')", root)?; + write!(f, "('{root}')")?; } Ok(()) } @@ -3142,7 +3142,7 @@ impl fmt::Display for ForXml { ForXml::Path(root) => { write!(f, "PATH")?; if let Some(root) = root { - write!(f, "('{}')", root)?; + write!(f, "('{root}')")?; } Ok(()) } @@ -3205,7 +3205,7 @@ impl fmt::Display for JsonTableColumn { JsonTableColumn::Named(json_table_named_column) => { write!(f, "{json_table_named_column}") } - JsonTableColumn::ForOrdinality(ident) => write!(f, "{} FOR ORDINALITY", ident), + JsonTableColumn::ForOrdinality(ident) => write!(f, "{ident} FOR ORDINALITY"), JsonTableColumn::Nested(json_table_nested_column) => { write!(f, "{json_table_nested_column}") } @@ -3271,10 +3271,10 @@ impl fmt::Display for JsonTableNamedColumn { self.path )?; if let Some(on_empty) = &self.on_empty { - write!(f, " {} ON EMPTY", on_empty)?; + write!(f, " {on_empty} ON EMPTY")?; } if let Some(on_error) = &self.on_error { - write!(f, " {} ON ERROR", on_error)?; + write!(f, " {on_error} ON ERROR")?; } Ok(()) } @@ -3296,7 +3296,7 @@ impl fmt::Display for JsonTableColumnErrorHandling { match self { JsonTableColumnErrorHandling::Null => write!(f, "NULL"), JsonTableColumnErrorHandling::Default(json_string) => { - write!(f, "DEFAULT {}", json_string) + write!(f, "DEFAULT {json_string}") } JsonTableColumnErrorHandling::Error => write!(f, "ERROR"), } @@ -3429,12 +3429,12 @@ impl fmt::Display for XmlTableColumn { default, nullable, } => { - write!(f, " {}", r#type)?; + write!(f, " {type}")?; if let Some(p) = path { - write!(f, " PATH {}", p)?; + write!(f, " PATH {p}")?; } if let Some(d) = default { - write!(f, " DEFAULT {}", d)?; + write!(f, " DEFAULT {d}")?; } if !*nullable { write!(f, " NOT NULL")?; @@ -3465,7 +3465,7 @@ impl fmt::Display for XmlPassingArgument { } write!(f, "{}", self.expr)?; if let Some(alias) = &self.alias { - write!(f, " AS {}", alias)?; + write!(f, " AS {alias}")?; } Ok(()) } diff --git a/src/ast/value.rs b/src/ast/value.rs index 98616407c..90dbccbf9 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -551,16 +551,16 @@ impl fmt::Display for EscapeUnicodeStringLiteral<'_> { write!(f, r#"\\"#)?; } x if x.is_ascii() => { - write!(f, "{}", c)?; + write!(f, "{c}")?; } _ => { let codepoint = c as u32; // if the character fits in 32 bits, we can use the \XXXX format // otherwise, we need to use the \+XXXXXX format if codepoint <= 0xFFFF { - write!(f, "\\{:04X}", codepoint)?; + write!(f, "\\{codepoint:04X}")?; } else { - write!(f, "\\+{:06X}", codepoint)?; + write!(f, "\\+{codepoint:06X}")?; } } } diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index ab4f73aa4..8e0a3139a 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -926,10 +926,10 @@ mod tests { #[test] fn overflow() { let cond = (0..1000) - .map(|n| format!("X = {}", n)) + .map(|n| format!("X = {n}")) .collect::>() .join(" OR "); - let sql = format!("SELECT x where {0}", cond); + let sql = format!("SELECT x where {cond}"); let dialect = GenericDialect {}; let tokens = Tokenizer::new(&dialect, sql.as_str()).tokenize().unwrap(); diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index bc92948d1..c79b45178 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -587,7 +587,7 @@ pub trait Dialect: Debug + Any { } let token = parser.peek_token(); - debug!("get_next_precedence_full() {:?}", token); + debug!("get_next_precedence_full() {token:?}"); match token.token { Token::Word(w) if w.keyword == Keyword::OR => Ok(p!(Or)), Token::Word(w) if w.keyword == Keyword::AND => Ok(p!(And)), diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 9b08b8f32..a91ab598b 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -104,7 +104,7 @@ impl Dialect for PostgreSqlDialect { fn get_next_precedence(&self, parser: &Parser) -> Option> { let token = parser.peek_token(); - debug!("get_next_precedence() {:?}", token); + debug!("get_next_precedence() {token:?}"); // we only return some custom value here when the behaviour (not merely the numeric value) differs // from the default implementation diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e09b7e338..f1b09afbf 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -436,7 +436,7 @@ impl<'a> Parser<'a> { /// /// See example on [`Parser::new()`] for an example pub fn try_with_sql(self, sql: &str) -> Result { - debug!("Parsing sql '{}'...", sql); + debug!("Parsing sql '{sql}'..."); let tokens = Tokenizer::new(self.dialect, sql) .with_unescape(self.options.unescape) .tokenize_with_location()?; @@ -1226,10 +1226,10 @@ impl<'a> Parser<'a> { expr = self.parse_compound_expr(expr, vec![])?; - debug!("prefix: {:?}", expr); + debug!("prefix: {expr:?}"); loop { let next_precedence = self.get_next_precedence()?; - debug!("next precedence: {:?}", next_precedence); + debug!("next precedence: {next_precedence:?}"); if precedence >= next_precedence { break; @@ -1631,8 +1631,7 @@ impl<'a> Parser<'a> { Token::QuestionPipe => UnaryOperator::QuestionPipe, _ => { return Err(ParserError::ParserError(format!( - "Unexpected token in unary operator parsing: {:?}", - tok + "Unexpected token in unary operator parsing: {tok:?}" ))) } }; @@ -13655,7 +13654,7 @@ impl<'a> Parser<'a> { let ident = self.parse_identifier()?; if let GranteeName::ObjectName(namespace) = name { name = GranteeName::ObjectName(ObjectName::from(vec![Ident::new( - format!("{}:{}", namespace, ident), + format!("{namespace}:{ident}"), )])); }; } @@ -14625,7 +14624,7 @@ impl<'a> Parser<'a> { self.dialect .get_reserved_keywords_for_select_item_operator(), ) - .map(|keyword| Ident::new(format!("{:?}", keyword))); + .map(|keyword| Ident::new(format!("{keyword:?}"))); match self.parse_wildcard_expr()? { Expr::QualifiedWildcard(prefix, token) => Ok(SelectItem::QualifiedWildcard( diff --git a/src/test_utils.rs b/src/test_utils.rs index 1af9f454b..c7965c3fd 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -270,7 +270,7 @@ impl TestedDialects { tokenizer = tokenizer.with_unescape(options.unescape); } let tokens = tokenizer.tokenize().unwrap(); - assert_eq!(expected, tokens, "Tokenized differently for {:?}", dialect); + assert_eq!(expected, tokens, "Tokenized differently for {dialect:?}"); }); } } diff --git a/src/tokenizer.rs b/src/tokenizer.rs index afe1e35c7..8382a5344 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -1751,7 +1751,7 @@ impl<'a> Tokenizer<'a> { (None, Some(tok)) => Ok(Some(tok)), (None, None) => self.tokenizer_error( chars.location(), - format!("Expected a valid binary operator after '{}'", prefix), + format!("Expected a valid binary operator after '{prefix}'"), ), } } @@ -1809,7 +1809,7 @@ impl<'a> Tokenizer<'a> { chars.next(); let mut temp = String::new(); - let end_delimiter = format!("${}$", value); + let end_delimiter = format!("${value}$"); loop { match chars.next() { @@ -2402,13 +2402,13 @@ fn take_char_from_hex_digits( location: chars.location(), })?; let digit = next_char.to_digit(16).ok_or_else(|| TokenizerError { - message: format!("Invalid hex digit in escaped unicode string: {}", next_char), + message: format!("Invalid hex digit in escaped unicode string: {next_char}"), location: chars.location(), })?; result = result * 16 + digit; } char::from_u32(result).ok_or_else(|| TokenizerError { - message: format!("Invalid unicode character: {:x}", result), + message: format!("Invalid unicode character: {result:x}"), location: chars.location(), }) } @@ -3504,7 +3504,7 @@ mod tests { } fn check_unescape(s: &str, expected: Option<&str>) { - let s = format!("'{}'", s); + let s = format!("'{s}'"); let mut state = State { peekable: s.chars().peekable(), line: 0, diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index ed5d7ad2f..0a60c9c4f 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -1345,7 +1345,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - clickhouse().verified_stmt(&format!("USE {}", object_name)), + clickhouse().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -1353,7 +1353,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - clickhouse().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), + clickhouse().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -1367,7 +1367,7 @@ fn parse_use() { fn test_query_with_format_clause() { let format_options = vec!["TabSeparated", "JSONCompact", "NULL"]; for format in &format_options { - let sql = format!("SELECT * FROM t FORMAT {}", format); + let sql = format!("SELECT * FROM t FORMAT {format}"); match clickhouse_and_generic().verified_stmt(&sql) { Statement::Query(query) => { if *format == "NULL" { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 93cc89e55..6c4ed08c0 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3563,7 +3563,7 @@ fn test_double_value() { for (input, expected) in test_cases { for (i, expr) in input.iter().enumerate() { if let Statement::Query(query) = - dialects.one_statement_parses_to(&format!("SELECT {}", expr), "") + dialects.one_statement_parses_to(&format!("SELECT {expr}"), "") { if let SetExpr::Select(select) = *query.body { assert_eq!(expected[i], select.projection[0]); @@ -4023,13 +4023,13 @@ fn parse_create_table_column_constraint_characteristics() { syntax }; - let sql = format!("CREATE TABLE t (a int UNIQUE {})", syntax); + let sql = format!("CREATE TABLE t (a int UNIQUE {syntax})"); let expected_clause = if syntax.is_empty() { String::new() } else { format!(" {syntax}") }; - let expected = format!("CREATE TABLE t (a INT UNIQUE{})", expected_clause); + let expected = format!("CREATE TABLE t (a INT UNIQUE{expected_clause})"); let ast = one_statement_parses_to(&sql, &expected); let expected_value = if deferrable.is_some() || initially.is_some() || enforced.is_some() { @@ -7499,7 +7499,7 @@ fn parse_cte_in_data_modification_statements() { assert_eq!(query.with.unwrap().to_string(), "WITH x AS (SELECT 1)"); assert!(matches!(*query.body, SetExpr::Update(_))); } - other => panic!("Expected: UPDATE, got: {:?}", other), + other => panic!("Expected: UPDATE, got: {other:?}"), } match verified_stmt("WITH t (x) AS (SELECT 9) DELETE FROM q WHERE id IN (SELECT x FROM t)") { @@ -7507,7 +7507,7 @@ fn parse_cte_in_data_modification_statements() { assert_eq!(query.with.unwrap().to_string(), "WITH t (x) AS (SELECT 9)"); assert!(matches!(*query.body, SetExpr::Delete(_))); } - other => panic!("Expected: DELETE, got: {:?}", other), + other => panic!("Expected: DELETE, got: {other:?}"), } match verified_stmt("WITH x AS (SELECT 42) INSERT INTO t SELECT foo FROM x") { @@ -7515,7 +7515,7 @@ fn parse_cte_in_data_modification_statements() { assert_eq!(query.with.unwrap().to_string(), "WITH x AS (SELECT 42)"); assert!(matches!(*query.body, SetExpr::Insert(_))); } - other => panic!("Expected: INSERT, got: {:?}", other), + other => panic!("Expected: INSERT, got: {other:?}"), } } @@ -10043,7 +10043,7 @@ fn parse_offset_and_limit() { #[test] fn parse_time_functions() { fn test_time_function(func_name: &'static str) { - let sql = format!("SELECT {}()", func_name); + let sql = format!("SELECT {func_name}()"); let select = verified_only_select(&sql); let select_localtime_func_call_ast = Function { name: ObjectName::from(vec![Ident::new(func_name)]), @@ -10065,7 +10065,7 @@ fn parse_time_functions() { ); // Validating Parenthesis - let sql_without_parens = format!("SELECT {}", func_name); + let sql_without_parens = format!("SELECT {func_name}"); let mut ast_without_parens = select_localtime_func_call_ast; ast_without_parens.args = FunctionArguments::None; assert_eq!( @@ -14306,7 +14306,7 @@ fn overflow() { let expr = std::iter::repeat_n("1", 1000) .collect::>() .join(" + "); - let sql = format!("SELECT {}", expr); + let sql = format!("SELECT {expr}"); let mut statements = Parser::parse_sql(&GenericDialect {}, sql.as_str()).unwrap(); let statement = statements.pop().unwrap(); @@ -14606,7 +14606,7 @@ fn test_conditional_statement_span() { else_block.unwrap().span() ); } - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } } diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 99b7eecde..baf279fae 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -213,7 +213,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - databricks().verified_stmt(&format!("USE {}", object_name)), + databricks().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -221,7 +221,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - databricks().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), + databricks().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -233,21 +233,21 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with keyword and different type of quotes assert_eq!( - databricks().verified_stmt(&format!("USE CATALOG {0}my_catalog{0}", quote)), + databricks().verified_stmt(&format!("USE CATALOG {quote}my_catalog{quote}")), Statement::Use(Use::Catalog(ObjectName::from(vec![Ident::with_quote( quote, "my_catalog".to_string(), )]))) ); assert_eq!( - databricks().verified_stmt(&format!("USE DATABASE {0}my_database{0}", quote)), + databricks().verified_stmt(&format!("USE DATABASE {quote}my_database{quote}")), Statement::Use(Use::Database(ObjectName::from(vec![Ident::with_quote( quote, "my_database".to_string(), )]))) ); assert_eq!( - databricks().verified_stmt(&format!("USE SCHEMA {0}my_schema{0}", quote)), + databricks().verified_stmt(&format!("USE SCHEMA {quote}my_schema{quote}")), Statement::Use(Use::Schema(ObjectName::from(vec![Ident::with_quote( quote, "my_schema".to_string(), @@ -357,6 +357,6 @@ fn data_type_timestamp_ntz() { }] ); } - s => panic!("Unexpected statement: {:?}", s), + s => panic!("Unexpected statement: {s:?}"), } } diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index f503ed551..44cb22cea 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -368,7 +368,7 @@ fn test_duckdb_specific_int_types() { ("HUGEINT", DataType::HugeInt), ]; for (dtype_string, data_type) in duckdb_dtypes { - let sql = format!("SELECT 123::{}", dtype_string); + let sql = format!("SELECT 123::{dtype_string}"); let select = duckdb().verified_only_select(&sql); assert_eq!( &Expr::Cast { @@ -792,7 +792,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - duckdb().verified_stmt(&format!("USE {}", object_name)), + duckdb().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -800,7 +800,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - duckdb().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), + duckdb().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -812,7 +812,9 @@ fn parse_use() { for "e in "e_styles { // Test double identifier with different type of quotes assert_eq!( - duckdb().verified_stmt(&format!("USE {0}CATALOG{0}.{0}my_schema{0}", quote)), + duckdb().verified_stmt(&format!( + "USE {quote}CATALOG{quote}.{quote}my_schema{quote}" + )), Statement::Use(Use::Object(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index fd52b7730..56a72ec84 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -524,7 +524,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - hive().verified_stmt(&format!("USE {}", object_name)), + hive().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -532,7 +532,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - hive().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + hive().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 9ec28f421..ebbec25fb 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1673,7 +1673,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - ms().verified_stmt(&format!("USE {}", object_name)), + ms().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -1681,7 +1681,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - ms().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + ms().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -2187,7 +2187,7 @@ fn parse_mssql_if_else() { "IF 1 = 1 BEGIN SET @A = 1; END ELSE SET @A = 2;" ); } - _ => panic!("Unexpected statements: {:?}", stmts), + _ => panic!("Unexpected statements: {stmts:?}"), } } @@ -2237,7 +2237,7 @@ fn test_mssql_if_statements_span() { Span::new(Location::new(1, 21), Location::new(1, 36)) ); } - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Blocks @@ -2258,7 +2258,7 @@ fn test_mssql_if_statements_span() { Span::new(Location::new(1, 32), Location::new(1, 57)) ); } - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index b11d76dde..d2feee035 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -593,7 +593,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - mysql_and_generic().verified_stmt(&format!("USE {}", object_name)), + mysql_and_generic().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -601,8 +601,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - mysql_and_generic() - .verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + mysql_and_generic().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -2263,11 +2262,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { assert_eq!(&[Ident::new("t"), Ident::new("15to29")], &parts[..]); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 2: Qualified column name that starts with digits and on its own represents a number. @@ -2277,11 +2276,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { assert_eq!(&[Ident::new("t"), Ident::new("15e29")], &parts[..]); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 3: Unqualified, the same token is parsed as a number. @@ -2295,11 +2294,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::Value(ValueWithSpan { value, .. }))) => { assert_eq!(&number("15e29"), value); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 4: Quoted simple identifier. @@ -2309,11 +2308,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::Identifier(name))) => { assert_eq!(&Ident::with_quote('`', "15e29"), name); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 5: Quoted compound identifier. @@ -2326,11 +2325,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { &parts[..] ); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 6: Multi-level compound identifiers. @@ -2347,11 +2346,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { &parts[..] ); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 7: Multi-level compound quoted identifiers. @@ -2368,11 +2367,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { &parts[..] ); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index b6605cf1a..7b0a8c5da 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2535,12 +2535,12 @@ fn parse_create_indices_with_operator_classes() { for expected_operator_class in &operator_classes { let single_column_sql_statement = format!( "CREATE INDEX the_index_name ON users USING {expected_index_type} (concat_users_name(first_name, last_name){})", - expected_operator_class.as_ref().map(|oc| format!(" {}", oc)) + expected_operator_class.as_ref().map(|oc| format!(" {oc}")) .unwrap_or_default() ); let multi_column_sql_statement = format!( "CREATE INDEX the_index_name ON users USING {expected_index_type} (column_name,concat_users_name(first_name, last_name){})", - expected_operator_class.as_ref().map(|oc| format!(" {}", oc)) + expected_operator_class.as_ref().map(|oc| format!(" {oc}")) .unwrap_or_default() ); @@ -3273,7 +3273,7 @@ fn test_fn_arg_with_value_operator() { assert!(matches!( &args[..], &[FunctionArg::ExprNamed { operator: FunctionArgOperator::Value, .. }] - ), "Invalid function argument: {:?}", args); + ), "Invalid function argument: {args:?}"); } other => panic!("Expected: JSON_OBJECT('name' VALUE 'value') to be parsed as a function, but got {other:?}"), } @@ -5679,7 +5679,7 @@ fn parse_drop_trigger() { "DROP TRIGGER{} check_update ON table_name{}", if if_exists { " IF EXISTS" } else { "" }, option - .map(|o| format!(" {}", o)) + .map(|o| format!(" {o}")) .unwrap_or_else(|| "".to_string()) ); assert_eq!( @@ -5773,8 +5773,7 @@ fn parse_trigger_related_functions() { // Now we parse the statements and check if they are parsed correctly. let mut statements = pg() .parse_sql_statements(&format!( - "{}{}{}{}", - sql_table_creation, sql_create_function, sql_create_trigger, sql_drop_trigger + "{sql_table_creation}{sql_create_function}{sql_create_trigger}{sql_drop_trigger}" )) .unwrap(); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 8b3988d92..3ffd33392 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -2510,10 +2510,7 @@ fn test_snowflake_stage_object_names_into_location() { .zip(allowed_object_names.iter_mut()) { let (formatted_name, object_name) = it; - let sql = format!( - "COPY INTO {} FROM 'gcs://mybucket/./../a.csv'", - formatted_name - ); + let sql = format!("COPY INTO {formatted_name} FROM 'gcs://mybucket/./../a.csv'"); match snowflake().verified_stmt(&sql) { Statement::CopyIntoSnowflake { into, .. } => { assert_eq!(into.0, object_name.0) @@ -2536,10 +2533,7 @@ fn test_snowflake_stage_object_names_into_table() { .zip(allowed_object_names.iter_mut()) { let (formatted_name, object_name) = it; - let sql = format!( - "COPY INTO {} FROM 'gcs://mybucket/./../a.csv'", - formatted_name - ); + let sql = format!("COPY INTO {formatted_name} FROM 'gcs://mybucket/./../a.csv'"); match snowflake().verified_stmt(&sql) { Statement::CopyIntoSnowflake { into, .. } => { assert_eq!(into.0, object_name.0) @@ -3020,7 +3014,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - snowflake().verified_stmt(&format!("USE {}", object_name)), + snowflake().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -3028,7 +3022,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - snowflake().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + snowflake().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -3040,7 +3034,9 @@ fn parse_use() { for "e in "e_styles { // Test double identifier with different type of quotes assert_eq!( - snowflake().verified_stmt(&format!("USE {0}CATALOG{0}.{0}my_schema{0}", quote)), + snowflake().verified_stmt(&format!( + "USE {quote}CATALOG{quote}.{quote}my_schema{quote}" + )), Statement::Use(Use::Object(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") @@ -3059,35 +3055,37 @@ fn parse_use() { for "e in "e_styles { // Test single and double identifier with keyword and different type of quotes assert_eq!( - snowflake().verified_stmt(&format!("USE DATABASE {0}my_database{0}", quote)), + snowflake().verified_stmt(&format!("USE DATABASE {quote}my_database{quote}")), Statement::Use(Use::Database(ObjectName::from(vec![Ident::with_quote( quote, "my_database".to_string(), )]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE SCHEMA {0}my_schema{0}", quote)), + snowflake().verified_stmt(&format!("USE SCHEMA {quote}my_schema{quote}")), Statement::Use(Use::Schema(ObjectName::from(vec![Ident::with_quote( quote, "my_schema".to_string(), )]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE SCHEMA {0}CATALOG{0}.{0}my_schema{0}", quote)), + snowflake().verified_stmt(&format!( + "USE SCHEMA {quote}CATALOG{quote}.{quote}my_schema{quote}" + )), Statement::Use(Use::Schema(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") ]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE ROLE {0}my_role{0}", quote)), + snowflake().verified_stmt(&format!("USE ROLE {quote}my_role{quote}")), Statement::Use(Use::Role(ObjectName::from(vec![Ident::with_quote( quote, "my_role".to_string(), )]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE WAREHOUSE {0}my_wh{0}", quote)), + snowflake().verified_stmt(&format!("USE WAREHOUSE {quote}my_wh{quote}")), Statement::Use(Use::Warehouse(ObjectName::from(vec![Ident::with_quote( quote, "my_wh".to_string(), @@ -3629,7 +3627,7 @@ fn test_alter_session_followed_by_statement() { .unwrap(); match stmts[..] { [Statement::AlterSession { .. }, Statement::Query { .. }] => {} - _ => panic!("Unexpected statements: {:?}", stmts), + _ => panic!("Unexpected statements: {stmts:?}"), } } diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index b759065f3..06496f0ce 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -324,7 +324,7 @@ fn parse_create_table_on_conflict_col() { Keyword::IGNORE, Keyword::REPLACE, ] { - let sql = format!("CREATE TABLE t1 (a INT, b INT ON CONFLICT {:?})", keyword); + let sql = format!("CREATE TABLE t1 (a INT, b INT ON CONFLICT {keyword:?})"); match sqlite_and_generic().verified_stmt(&sql) { Statement::CreateTable(CreateTable { columns, .. }) => { assert_eq!( @@ -410,7 +410,7 @@ fn parse_window_function_with_filter() { "count", "user_defined_function", ] { - let sql = format!("SELECT {}(x) FILTER (WHERE y) OVER () FROM t", func_name); + let sql = format!("SELECT {func_name}(x) FILTER (WHERE y) OVER () FROM t"); let select = sqlite().verified_only_select(&sql); assert_eq!(select.to_string(), sql); assert_eq!( @@ -444,7 +444,7 @@ fn parse_window_function_with_filter() { fn parse_attach_database() { let sql = "ATTACH DATABASE 'test.db' AS test"; let verified_stmt = sqlite().verified_stmt(sql); - assert_eq!(sql, format!("{}", verified_stmt)); + assert_eq!(sql, format!("{verified_stmt}")); match verified_stmt { Statement::AttachDatabase { schema_name, From 6c38cdcadb45ff5b343ea49bad23fdfeebc92cf5 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 27 Jun 2025 21:21:38 +0200 Subject: [PATCH 169/286] Snowflake: Add support for future grants (#1906) --- src/ast/mod.rs | 27 +++++++++++++++++++++++++++ src/keywords.rs | 1 + src/parser/mod.rs | 27 +++++++++++++++++++++++++++ tests/sqlparser_common.rs | 4 +++- 4 files changed, 58 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 0f6820626..19966d21c 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6920,6 +6920,12 @@ pub enum GrantObjects { AllSequencesInSchema { schemas: Vec }, /// Grant privileges on `ALL TABLES IN SCHEMA [, ...]` AllTablesInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE SCHEMAS IN DATABASE [, ...]` + FutureSchemasInDatabase { databases: Vec }, + /// Grant privileges on `FUTURE TABLES IN SCHEMA [, ...]` + FutureTablesInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE VIEWS IN SCHEMA [, ...]` + FutureViewsInSchema { schemas: Vec }, /// Grant privileges on specific databases Databases(Vec), /// Grant privileges on specific schemas @@ -6988,6 +6994,27 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::FutureSchemasInDatabase { databases } => { + write!( + f, + "FUTURE SCHEMAS IN DATABASE {}", + display_comma_separated(databases) + ) + } + GrantObjects::FutureTablesInSchema { schemas } => { + write!( + f, + "FUTURE TABLES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureViewsInSchema { schemas } => { + write!( + f, + "FUTURE VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } GrantObjects::ResourceMonitors(objects) => { write!(f, "RESOURCE MONITOR {}", display_comma_separated(objects)) } diff --git a/src/keywords.rs b/src/keywords.rs index f56178c17..a8bbca3d3 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -395,6 +395,7 @@ define_keywords!( FUNCTION, FUNCTIONS, FUSION, + FUTURE, GENERAL, GENERATE, GENERATED, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f1b09afbf..68d89a1e1 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13691,6 +13691,33 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllTablesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::SCHEMAS, + Keyword::IN, + Keyword::DATABASE, + ]) { + Some(GrantObjects::FutureSchemasInDatabase { + databases: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::TABLES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[ Keyword::ALL, Keyword::SEQUENCES, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 6c4ed08c0..17b46e6f7 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9386,9 +9386,11 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON VIEW view1 TO ROLE role1"); verified_stmt("GRANT EXEC ON my_sp TO runner"); verified_stmt("GRANT UPDATE ON my_table TO updater_role AS dbo"); - all_dialects_where(|d| d.identifier_quote_style("none") == Some('[')) .verified_stmt("GRANT SELECT ON [my_table] TO [public]"); + verified_stmt("GRANT SELECT ON FUTURE SCHEMAS IN DATABASE db1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE TABLES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); } #[test] From 50c605a47138b1602575731c196d04b0d280ad3d Mon Sep 17 00:00:00 2001 From: Sergey Olontsev Date: Sat, 28 Jun 2025 07:13:11 +0100 Subject: [PATCH 170/286] Support for Map values in ClickHouse settings (#1896) Co-authored-by: Ifeanyi Ubah --- src/ast/query.rs | 2 +- src/ast/value.rs | 1 - src/parser/mod.rs | 17 ++--- src/test_utils.rs | 5 ++ tests/sqlparser_clickhouse.rs | 130 +++++++++++++++++++++++++--------- 5 files changed, 112 insertions(+), 43 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index c79ec110c..99cd2ef28 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1047,7 +1047,7 @@ impl fmt::Display for ConnectBy { #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct Setting { pub key: Ident, - pub value: Value, + pub value: Expr, } impl fmt::Display for Setting { diff --git a/src/ast/value.rs b/src/ast/value.rs index 90dbccbf9..fdfa6a674 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -116,7 +116,6 @@ impl From for Value { derive(Visit, VisitMut), visit(with = "visit_value") )] - pub enum Value { /// Numeric literal #[cfg(not(feature = "bigdecimal"))] diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 68d89a1e1..adf50a8f0 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2770,7 +2770,7 @@ impl<'a> Parser<'a> { if self.dialect.supports_dictionary_syntax() { self.prev_token(); // Put back the '{' - return self.parse_duckdb_struct_literal(); + return self.parse_dictionary(); } self.expected("an expression", token) @@ -3139,7 +3139,7 @@ impl<'a> Parser<'a> { Ok(fields) } - /// DuckDB specific: Parse a duckdb [dictionary] + /// DuckDB and ClickHouse specific: Parse a duckdb [dictionary] or a clickhouse [map] setting /// /// Syntax: /// @@ -3148,18 +3148,18 @@ impl<'a> Parser<'a> { /// ``` /// /// [dictionary]: https://duckdb.org/docs/sql/data_types/struct#creating-structs - fn parse_duckdb_struct_literal(&mut self) -> Result { + /// [map]: https://clickhouse.com/docs/operations/settings/settings#additional_table_filters + fn parse_dictionary(&mut self) -> Result { self.expect_token(&Token::LBrace)?; - let fields = - self.parse_comma_separated0(Self::parse_duckdb_dictionary_field, Token::RBrace)?; + let fields = self.parse_comma_separated0(Self::parse_dictionary_field, Token::RBrace)?; self.expect_token(&Token::RBrace)?; Ok(Expr::Dictionary(fields)) } - /// Parse a field for a duckdb [dictionary] + /// Parse a field for a duckdb [dictionary] or a clickhouse [map] setting /// /// Syntax /// @@ -3168,7 +3168,8 @@ impl<'a> Parser<'a> { /// ``` /// /// [dictionary]: https://duckdb.org/docs/sql/data_types/struct#creating-structs - fn parse_duckdb_dictionary_field(&mut self) -> Result { + /// [map]: https://clickhouse.com/docs/operations/settings/settings#additional_table_filters + fn parse_dictionary_field(&mut self) -> Result { let key = self.parse_identifier()?; self.expect_token(&Token::Colon)?; @@ -11216,7 +11217,7 @@ impl<'a> Parser<'a> { let key_values = self.parse_comma_separated(|p| { let key = p.parse_identifier()?; p.expect_token(&Token::Eq)?; - let value = p.parse_value()?.value; + let value = p.parse_expr()?; Ok(Setting { key, value }) })?; Some(key_values) diff --git a/src/test_utils.rs b/src/test_utils.rs index c7965c3fd..db7b3dd60 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -366,6 +366,11 @@ pub fn number(n: &str) -> Value { Value::Number(n.parse().unwrap(), false) } +/// Creates a [Value::SingleQuotedString] +pub fn single_quoted_string(s: impl Into) -> Value { + Value::SingleQuotedString(s.into()) +} + pub fn table_alias(name: impl Into) -> Option { Some(TableAlias { name: Ident::new(name), diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 0a60c9c4f..0288c6d2c 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -28,7 +28,7 @@ use test_utils::*; use sqlparser::ast::Expr::{BinaryOp, Identifier}; use sqlparser::ast::SelectItem::UnnamedExpr; use sqlparser::ast::TableFactor::Table; -use sqlparser::ast::Value::Number; +use sqlparser::ast::Value::Boolean; use sqlparser::ast::*; use sqlparser::dialect::ClickHouseDialect; use sqlparser::dialect::GenericDialect; @@ -965,38 +965,103 @@ fn parse_limit_by() { #[test] fn parse_settings_in_query() { - match clickhouse_and_generic() - .verified_stmt(r#"SELECT * FROM t SETTINGS max_threads = 1, max_block_size = 10000"#) - { - Statement::Query(query) => { - assert_eq!( - query.settings, - Some(vec![ - Setting { - key: Ident::new("max_threads"), - value: Number("1".parse().unwrap(), false) - }, - Setting { - key: Ident::new("max_block_size"), - value: Number("10000".parse().unwrap(), false) - }, - ]) - ); + fn check_settings(sql: &str, expected: Vec) { + match clickhouse_and_generic().verified_stmt(sql) { + Statement::Query(q) => { + assert_eq!(q.settings, Some(expected)); + } + _ => unreachable!(), } - _ => unreachable!(), + } + + for (sql, expected_settings) in [ + ( + r#"SELECT * FROM t SETTINGS max_threads = 1, max_block_size = 10000"#, + vec![ + Setting { + key: Ident::new("max_threads"), + value: Expr::value(number("1")), + }, + Setting { + key: Ident::new("max_block_size"), + value: Expr::value(number("10000")), + }, + ], + ), + ( + r#"SELECT * FROM t SETTINGS additional_table_filters = {'table_1': 'x != 2'}"#, + vec![Setting { + key: Ident::new("additional_table_filters"), + value: Expr::Dictionary(vec![DictionaryField { + key: Ident::with_quote('\'', "table_1"), + value: Expr::value(single_quoted_string("x != 2")).into(), + }]), + }], + ), + ( + r#"SELECT * FROM t SETTINGS additional_result_filter = 'x != 2', query_plan_optimize_lazy_materialization = false"#, + vec![ + Setting { + key: Ident::new("additional_result_filter"), + value: Expr::value(single_quoted_string("x != 2")), + }, + Setting { + key: Ident::new("query_plan_optimize_lazy_materialization"), + value: Expr::value(Boolean(false)), + }, + ], + ), + ] { + check_settings(sql, expected_settings); } let invalid_cases = vec![ - "SELECT * FROM t SETTINGS a", - "SELECT * FROM t SETTINGS a=", - "SELECT * FROM t SETTINGS a=1, b", - "SELECT * FROM t SETTINGS a=1, b=", - "SELECT * FROM t SETTINGS a=1, b=c", + ("SELECT * FROM t SETTINGS a", "Expected: =, found: EOF"), + ( + "SELECT * FROM t SETTINGS a=", + "Expected: an expression, found: EOF", + ), + ("SELECT * FROM t SETTINGS a=1, b", "Expected: =, found: EOF"), + ( + "SELECT * FROM t SETTINGS a=1, b=", + "Expected: an expression, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {", + "Expected: identifier, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {'b'", + "Expected: :, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': ", + "Expected: an expression, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': 'c',}", + "Expected: identifier, found: }", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': 'c', 'd'}", + "Expected: :, found: }", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': 'c', 'd': }", + "Expected: an expression, found: }", + ), + ( + "SELECT * FROM t SETTINGS a = {ANY(b)}", + "Expected: :, found: (", + ), ]; - for sql in invalid_cases { - clickhouse_and_generic() - .parse_sql_statements(sql) - .expect_err("Expected: SETTINGS key = value, found: "); + for (sql, error_msg) in invalid_cases { + assert_eq!( + clickhouse_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError(error_msg.to_string()) + ); } } #[test] @@ -1550,11 +1615,11 @@ fn parse_select_table_function_settings() { settings: Some(vec![ Setting { key: "s0".into(), - value: Value::Number("3".parse().unwrap(), false), + value: Expr::value(number("3")), }, Setting { key: "s1".into(), - value: Value::SingleQuotedString("s".into()), + value: Expr::value(single_quoted_string("s")), }, ]), }, @@ -1575,11 +1640,11 @@ fn parse_select_table_function_settings() { settings: Some(vec![ Setting { key: "s0".into(), - value: Value::Number("3".parse().unwrap(), false), + value: Expr::value(number("3")), }, Setting { key: "s1".into(), - value: Value::SingleQuotedString("s".into()), + value: Expr::value(single_quoted_string("s")), }, ]), }, @@ -1589,7 +1654,6 @@ fn parse_select_table_function_settings() { "SELECT * FROM t(SETTINGS a=)", "SELECT * FROM t(SETTINGS a=1, b)", "SELECT * FROM t(SETTINGS a=1, b=)", - "SELECT * FROM t(SETTINGS a=1, b=c)", ]; for sql in invalid_cases { clickhouse_and_generic() From 3bc94234dfcf73d091c592edd1dab3fdd075ab82 Mon Sep 17 00:00:00 2001 From: Dima <111751109+Dimchikkk@users.noreply.github.com> Date: Sat, 28 Jun 2025 19:24:25 +0100 Subject: [PATCH 171/286] Fix join precedence for non-snowflake queries (#1905) --- src/dialect/generic.rs | 4 ++++ src/dialect/mod.rs | 28 ++++++++++++++++++++++++++++ src/dialect/snowflake.rs | 4 ++++ src/parser/mod.rs | 6 +++++- tests/sqlparser_common.rs | 23 +++++++++++++++++++++++ 5 files changed, 64 insertions(+), 1 deletion(-) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 8f57e487f..0b8822474 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -52,6 +52,10 @@ impl Dialect for GenericDialect { true } + fn supports_left_associative_joins_without_parens(&self) -> bool { + true + } + fn supports_connect_by(&self) -> bool { true } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index c79b45178..028aa58a7 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -278,6 +278,34 @@ pub trait Dialect: Debug + Any { false } + /// Indicates whether the dialect supports left-associative join parsing + /// by default when parentheses are omitted in nested joins. + /// + /// Most dialects (like MySQL or Postgres) assume **left-associative** precedence, + /// so a query like: + /// + /// ```sql + /// SELECT * FROM t1 NATURAL JOIN t5 INNER JOIN t0 ON ... + /// ``` + /// is interpreted as: + /// ```sql + /// ((t1 NATURAL JOIN t5) INNER JOIN t0 ON ...) + /// ``` + /// and internally represented as a **flat list** of joins. + /// + /// In contrast, some dialects (e.g. **Snowflake**) assume **right-associative** + /// precedence and interpret the same query as: + /// ```sql + /// (t1 NATURAL JOIN (t5 INNER JOIN t0 ON ...)) + /// ``` + /// which results in a **nested join** structure in the AST. + /// + /// If this method returns `false`, the parser must build nested join trees + /// even in the absence of parentheses to reflect the correct associativity + fn supports_left_associative_joins_without_parens(&self) -> bool { + true + } + /// Returns true if the dialect supports the `(+)` syntax for OUTER JOIN. fn supports_outer_join_operator(&self) -> bool { false diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 5ebb7e37c..ba28a8ec5 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -283,6 +283,10 @@ impl Dialect for SnowflakeDialect { true } + fn supports_left_associative_joins_without_parens(&self) -> bool { + false + } + fn is_reserved_for_identifier(&self, kw: Keyword) -> bool { // Unreserve some keywords that Snowflake accepts as identifiers // See: https://docs.snowflake.com/en/sql-reference/reserved-keywords diff --git a/src/parser/mod.rs b/src/parser/mod.rs index adf50a8f0..a079488f8 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12495,7 +12495,11 @@ impl<'a> Parser<'a> { }; let mut relation = self.parse_table_factor()?; - if self.peek_parens_less_nested_join() { + if !self + .dialect + .supports_left_associative_joins_without_parens() + && self.peek_parens_less_nested_join() + { let joins = self.parse_joins()?; relation = TableFactor::NestedJoin { table_with_joins: Box::new(TableWithJoins { relation, joins }), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 17b46e6f7..f6f51459c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15359,6 +15359,29 @@ fn check_enforced() { ); } +#[test] +fn join_precedence() { + all_dialects_except(|d| !d.supports_left_associative_joins_without_parens()) + .verified_query_with_canonical( + "SELECT * + FROM t1 + NATURAL JOIN t5 + INNER JOIN t0 ON (t0.v1 + t5.v0) > 0 + WHERE t0.v1 = t1.v0", + // canonical string without parentheses + "SELECT * FROM t1 NATURAL JOIN t5 INNER JOIN t0 ON (t0.v1 + t5.v0) > 0 WHERE t0.v1 = t1.v0", + ); + all_dialects_except(|d| d.supports_left_associative_joins_without_parens()).verified_query_with_canonical( + "SELECT * + FROM t1 + NATURAL JOIN t5 + INNER JOIN t0 ON (t0.v1 + t5.v0) > 0 + WHERE t0.v1 = t1.v0", + // canonical string with parentheses + "SELECT * FROM t1 NATURAL JOIN (t5 INNER JOIN t0 ON (t0.v1 + t5.v0) > 0) WHERE t0.v1 = t1.v0", + ); +} + #[test] fn parse_create_procedure_with_language() { let sql = r#"CREATE PROCEDURE test_proc LANGUAGE sql AS BEGIN SELECT 1; END"#; From abd80f9ecbf11c9d4b533a59010f079cd9df5b6c Mon Sep 17 00:00:00 2001 From: Simon Vandel Sillesen Date: Mon, 30 Jun 2025 17:51:55 +0200 Subject: [PATCH 172/286] Support remaining pipe operators (#1879) --- src/ast/query.rs | 153 ++++++++++++++ src/parser/mod.rs | 170 ++++++++++++++++ tests/sqlparser_common.rs | 416 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 739 insertions(+) diff --git a/src/ast/query.rs b/src/ast/query.rs index 99cd2ef28..febf1fc6d 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2684,6 +2684,79 @@ pub enum PipeOperator { /// Syntax: `|> TABLESAMPLE SYSTEM (10 PERCENT) /// See more at TableSample { sample: Box }, + /// Renames columns in the input table. + /// + /// Syntax: `|> RENAME old_name AS new_name, ...` + /// + /// See more at + Rename { mappings: Vec }, + /// Combines the input table with one or more tables using UNION. + /// + /// Syntax: `|> UNION [ALL|DISTINCT] (), (), ...` + /// + /// See more at + Union { + set_quantifier: SetQuantifier, + queries: Vec, + }, + /// Returns only the rows that are present in both the input table and the specified tables. + /// + /// Syntax: `|> INTERSECT [DISTINCT] (), (), ...` + /// + /// See more at + Intersect { + set_quantifier: SetQuantifier, + queries: Vec, + }, + /// Returns only the rows that are present in the input table but not in the specified tables. + /// + /// Syntax: `|> EXCEPT DISTINCT (), (), ...` + /// + /// See more at + Except { + set_quantifier: SetQuantifier, + queries: Vec, + }, + /// Calls a table function or procedure that returns a table. + /// + /// Syntax: `|> CALL function_name(args) [AS alias]` + /// + /// See more at + Call { + function: Function, + alias: Option, + }, + /// Pivots data from rows to columns. + /// + /// Syntax: `|> PIVOT(aggregate_function(column) FOR pivot_column IN (value1, value2, ...)) [AS alias]` + /// + /// See more at + Pivot { + aggregate_functions: Vec, + value_column: Vec, + value_source: PivotValueSource, + alias: Option, + }, + /// The `UNPIVOT` pipe operator transforms columns into rows. + /// + /// Syntax: + /// ```sql + /// |> UNPIVOT(value_column FOR name_column IN (column1, column2, ...)) [alias] + /// ``` + /// + /// See more at + Unpivot { + value_column: Ident, + name_column: Ident, + unpivot_columns: Vec, + alias: Option, + }, + /// Joins the input table with another table. + /// + /// Syntax: `|> [JOIN_TYPE] JOIN
[alias] ON ` or `|> [JOIN_TYPE] JOIN
[alias] USING ()` + /// + /// See more at + Join(Join), } impl fmt::Display for PipeOperator { @@ -2739,7 +2812,87 @@ impl fmt::Display for PipeOperator { PipeOperator::TableSample { sample } => { write!(f, "{sample}") } + PipeOperator::Rename { mappings } => { + write!(f, "RENAME {}", display_comma_separated(mappings)) + } + PipeOperator::Union { + set_quantifier, + queries, + } => Self::fmt_set_operation(f, "UNION", set_quantifier, queries), + PipeOperator::Intersect { + set_quantifier, + queries, + } => Self::fmt_set_operation(f, "INTERSECT", set_quantifier, queries), + PipeOperator::Except { + set_quantifier, + queries, + } => Self::fmt_set_operation(f, "EXCEPT", set_quantifier, queries), + PipeOperator::Call { function, alias } => { + write!(f, "CALL {function}")?; + Self::fmt_optional_alias(f, alias) + } + PipeOperator::Pivot { + aggregate_functions, + value_column, + value_source, + alias, + } => { + write!( + f, + "PIVOT({} FOR {} IN ({}))", + display_comma_separated(aggregate_functions), + Expr::CompoundIdentifier(value_column.to_vec()), + value_source + )?; + Self::fmt_optional_alias(f, alias) + } + PipeOperator::Unpivot { + value_column, + name_column, + unpivot_columns, + alias, + } => { + write!( + f, + "UNPIVOT({} FOR {} IN ({}))", + value_column, + name_column, + display_comma_separated(unpivot_columns) + )?; + Self::fmt_optional_alias(f, alias) + } + PipeOperator::Join(join) => write!(f, "{join}"), + } + } +} + +impl PipeOperator { + /// Helper function to format optional alias for pipe operators + fn fmt_optional_alias(f: &mut fmt::Formatter<'_>, alias: &Option) -> fmt::Result { + if let Some(alias) = alias { + write!(f, " AS {alias}")?; } + Ok(()) + } + + /// Helper function to format set operations (UNION, INTERSECT, EXCEPT) with queries + fn fmt_set_operation( + f: &mut fmt::Formatter<'_>, + operation: &str, + set_quantifier: &SetQuantifier, + queries: &[Query], + ) -> fmt::Result { + write!(f, "{operation}")?; + match set_quantifier { + SetQuantifier::None => {} + _ => { + write!(f, " {set_quantifier}")?; + } + } + write!(f, " ")?; + let parenthesized_queries: Vec = + queries.iter().map(|query| format!("({query})")).collect(); + write!(f, "{}", display_comma_separated(&parenthesized_queries)) } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a079488f8..4f8f1b85a 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9988,6 +9988,48 @@ impl<'a> Parser<'a> { Ok(IdentWithAlias { ident, alias }) } + /// Parse `identifier [AS] identifier` where the AS keyword is optional + fn parse_identifier_with_optional_alias(&mut self) -> Result { + let ident = self.parse_identifier()?; + let _after_as = self.parse_keyword(Keyword::AS); + let alias = self.parse_identifier()?; + Ok(IdentWithAlias { ident, alias }) + } + + /// Parse comma-separated list of parenthesized queries for pipe operators + fn parse_pipe_operator_queries(&mut self) -> Result, ParserError> { + self.parse_comma_separated(|parser| { + parser.expect_token(&Token::LParen)?; + let query = parser.parse_query()?; + parser.expect_token(&Token::RParen)?; + Ok(*query) + }) + } + + /// Parse set quantifier for pipe operators that require DISTINCT. E.g. INTERSECT and EXCEPT + fn parse_distinct_required_set_quantifier( + &mut self, + operator_name: &str, + ) -> Result { + let quantifier = self.parse_set_quantifier(&Some(SetOperator::Intersect)); + match quantifier { + SetQuantifier::Distinct | SetQuantifier::DistinctByName => Ok(quantifier), + _ => Err(ParserError::ParserError(format!( + "{operator_name} pipe operator requires DISTINCT modifier", + ))), + } + } + + /// Parse optional identifier alias (with or without AS keyword) + fn parse_identifier_optional_alias(&mut self) -> Result, ParserError> { + if self.parse_keyword(Keyword::AS) { + Ok(Some(self.parse_identifier()?)) + } else { + // Check if the next token is an identifier (implicit alias) + self.maybe_parse(|parser| parser.parse_identifier()) + } + } + /// Optionally parses an alias for a select list item fn maybe_parse_select_item_alias(&mut self) -> Result, ParserError> { fn validator(explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { @@ -11134,6 +11176,19 @@ impl<'a> Parser<'a> { Keyword::AGGREGATE, Keyword::ORDER, Keyword::TABLESAMPLE, + Keyword::RENAME, + Keyword::UNION, + Keyword::INTERSECT, + Keyword::EXCEPT, + Keyword::CALL, + Keyword::PIVOT, + Keyword::UNPIVOT, + Keyword::JOIN, + Keyword::INNER, + Keyword::LEFT, + Keyword::RIGHT, + Keyword::FULL, + Keyword::CROSS, ])?; match kw { Keyword::SELECT => { @@ -11200,6 +11255,121 @@ impl<'a> Parser<'a> { let sample = self.parse_table_sample(TableSampleModifier::TableSample)?; pipe_operators.push(PipeOperator::TableSample { sample }); } + Keyword::RENAME => { + let mappings = + self.parse_comma_separated(Parser::parse_identifier_with_optional_alias)?; + pipe_operators.push(PipeOperator::Rename { mappings }); + } + Keyword::UNION => { + let set_quantifier = self.parse_set_quantifier(&Some(SetOperator::Union)); + let queries = self.parse_pipe_operator_queries()?; + pipe_operators.push(PipeOperator::Union { + set_quantifier, + queries, + }); + } + Keyword::INTERSECT => { + let set_quantifier = + self.parse_distinct_required_set_quantifier("INTERSECT")?; + let queries = self.parse_pipe_operator_queries()?; + pipe_operators.push(PipeOperator::Intersect { + set_quantifier, + queries, + }); + } + Keyword::EXCEPT => { + let set_quantifier = self.parse_distinct_required_set_quantifier("EXCEPT")?; + let queries = self.parse_pipe_operator_queries()?; + pipe_operators.push(PipeOperator::Except { + set_quantifier, + queries, + }); + } + Keyword::CALL => { + let function_name = self.parse_object_name(false)?; + let function_expr = self.parse_function(function_name)?; + if let Expr::Function(function) = function_expr { + let alias = self.parse_identifier_optional_alias()?; + pipe_operators.push(PipeOperator::Call { function, alias }); + } else { + return Err(ParserError::ParserError( + "Expected function call after CALL".to_string(), + )); + } + } + Keyword::PIVOT => { + self.expect_token(&Token::LParen)?; + let aggregate_functions = + self.parse_comma_separated(Self::parse_aliased_function_call)?; + self.expect_keyword_is(Keyword::FOR)?; + let value_column = self.parse_period_separated(|p| p.parse_identifier())?; + self.expect_keyword_is(Keyword::IN)?; + + self.expect_token(&Token::LParen)?; + let value_source = if self.parse_keyword(Keyword::ANY) { + let order_by = if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) { + self.parse_comma_separated(Parser::parse_order_by_expr)? + } else { + vec![] + }; + PivotValueSource::Any(order_by) + } else if self.peek_sub_query() { + PivotValueSource::Subquery(self.parse_query()?) + } else { + PivotValueSource::List( + self.parse_comma_separated(Self::parse_expr_with_alias)?, + ) + }; + self.expect_token(&Token::RParen)?; + self.expect_token(&Token::RParen)?; + + let alias = self.parse_identifier_optional_alias()?; + + pipe_operators.push(PipeOperator::Pivot { + aggregate_functions, + value_column, + value_source, + alias, + }); + } + Keyword::UNPIVOT => { + self.expect_token(&Token::LParen)?; + let value_column = self.parse_identifier()?; + self.expect_keyword(Keyword::FOR)?; + let name_column = self.parse_identifier()?; + self.expect_keyword(Keyword::IN)?; + + self.expect_token(&Token::LParen)?; + let unpivot_columns = self.parse_comma_separated(Parser::parse_identifier)?; + self.expect_token(&Token::RParen)?; + + self.expect_token(&Token::RParen)?; + + let alias = self.parse_identifier_optional_alias()?; + + pipe_operators.push(PipeOperator::Unpivot { + value_column, + name_column, + unpivot_columns, + alias, + }); + } + Keyword::JOIN + | Keyword::INNER + | Keyword::LEFT + | Keyword::RIGHT + | Keyword::FULL + | Keyword::CROSS => { + self.prev_token(); + let mut joins = self.parse_joins()?; + if joins.len() != 1 { + return Err(ParserError::ParserError( + "Join pipe operator must have a single join".to_string(), + )); + } + let join = joins.swap_remove(0); + pipe_operators.push(PipeOperator::Join(join)) + } unhandled => { return Err(ParserError::ParserError(format!( "`expect_one_of_keywords` further up allowed unhandled keyword: {unhandled:?}" diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index f6f51459c..42bdeae17 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -15217,10 +15217,426 @@ fn parse_pipeline_operator() { dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE SYSTEM (50 PERCENT)"); dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE SYSTEM (50) REPEATABLE (10)"); + // rename pipe operator + dialects.verified_stmt("SELECT * FROM users |> RENAME old_name AS new_name"); + dialects.verified_stmt("SELECT * FROM users |> RENAME id AS user_id, name AS user_name"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> RENAME id user_id", + "SELECT * FROM users |> RENAME id AS user_id", + ); + + // union pipe operator + dialects.verified_stmt("SELECT * FROM users |> UNION ALL (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION DISTINCT (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION (SELECT * FROM admins)"); + + // union pipe operator with multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> UNION ALL (SELECT * FROM admins), (SELECT * FROM guests)", + ); + dialects.verified_stmt("SELECT * FROM users |> UNION DISTINCT (SELECT * FROM admins), (SELECT * FROM guests), (SELECT * FROM employees)"); + dialects.verified_stmt( + "SELECT * FROM users |> UNION (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // union pipe operator with BY NAME modifier + dialects.verified_stmt("SELECT * FROM users |> UNION BY NAME (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION ALL BY NAME (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION DISTINCT BY NAME (SELECT * FROM admins)"); + + // union pipe operator with BY NAME and multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> UNION BY NAME (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // intersect pipe operator (BigQuery requires DISTINCT modifier for INTERSECT) + dialects.verified_stmt("SELECT * FROM users |> INTERSECT DISTINCT (SELECT * FROM admins)"); + + // intersect pipe operator with BY NAME modifier + dialects + .verified_stmt("SELECT * FROM users |> INTERSECT DISTINCT BY NAME (SELECT * FROM admins)"); + + // intersect pipe operator with multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> INTERSECT DISTINCT (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // intersect pipe operator with BY NAME and multiple queries + dialects.verified_stmt("SELECT * FROM users |> INTERSECT DISTINCT BY NAME (SELECT * FROM admins), (SELECT * FROM guests)"); + + // except pipe operator (BigQuery requires DISTINCT modifier for EXCEPT) + dialects.verified_stmt("SELECT * FROM users |> EXCEPT DISTINCT (SELECT * FROM admins)"); + + // except pipe operator with BY NAME modifier + dialects.verified_stmt("SELECT * FROM users |> EXCEPT DISTINCT BY NAME (SELECT * FROM admins)"); + + // except pipe operator with multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> EXCEPT DISTINCT (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // except pipe operator with BY NAME and multiple queries + dialects.verified_stmt("SELECT * FROM users |> EXCEPT DISTINCT BY NAME (SELECT * FROM admins), (SELECT * FROM guests)"); + + // call pipe operator + dialects.verified_stmt("SELECT * FROM users |> CALL my_function()"); + dialects.verified_stmt("SELECT * FROM users |> CALL process_data(5, 'test')"); + dialects.verified_stmt( + "SELECT * FROM users |> CALL namespace.function_name(col1, col2, 'literal')", + ); + + // call pipe operator with complex arguments + dialects.verified_stmt("SELECT * FROM users |> CALL transform_data(col1 + col2)"); + dialects.verified_stmt("SELECT * FROM users |> CALL analyze_data('param1', 100, true)"); + + // call pipe operator with aliases + dialects.verified_stmt("SELECT * FROM input_table |> CALL tvf1(arg1) AS al"); + dialects.verified_stmt("SELECT * FROM users |> CALL process_data(5) AS result_table"); + dialects.verified_stmt("SELECT * FROM users |> CALL namespace.func() AS my_alias"); + + // multiple call pipe operators in sequence + dialects.verified_stmt("SELECT * FROM input_table |> CALL tvf1(arg1) |> CALL tvf2(arg2, arg3)"); + dialects.verified_stmt( + "SELECT * FROM data |> CALL transform(col1) |> CALL validate() |> CALL process(param)", + ); + + // multiple call pipe operators with aliases + dialects.verified_stmt( + "SELECT * FROM input_table |> CALL tvf1(arg1) AS step1 |> CALL tvf2(arg2) AS step2", + ); + dialects.verified_stmt( + "SELECT * FROM data |> CALL preprocess() AS clean_data |> CALL analyze(mode) AS results", + ); + + // call pipe operators mixed with other pipe operators + dialects.verified_stmt( + "SELECT * FROM users |> CALL transform() |> WHERE status = 'active' |> CALL process(param)", + ); + dialects.verified_stmt( + "SELECT * FROM data |> CALL preprocess() AS clean |> SELECT col1, col2 |> CALL validate()", + ); + + // pivot pipe operator + dialects.verified_stmt( + "SELECT * FROM monthly_sales |> PIVOT(SUM(amount) FOR quarter IN ('Q1', 'Q2', 'Q3', 'Q4'))", + ); + dialects.verified_stmt("SELECT * FROM sales_data |> PIVOT(AVG(revenue) FOR region IN ('North', 'South', 'East', 'West'))"); + + // pivot pipe operator with multiple aggregate functions + dialects.verified_stmt("SELECT * FROM data |> PIVOT(SUM(sales) AS total_sales, COUNT(*) AS num_transactions FOR month IN ('Jan', 'Feb', 'Mar'))"); + + // pivot pipe operator with compound column names + dialects.verified_stmt("SELECT * FROM sales |> PIVOT(SUM(amount) FOR product.category IN ('Electronics', 'Clothing'))"); + + // pivot pipe operator mixed with other pipe operators + dialects.verified_stmt("SELECT * FROM sales_data |> WHERE year = 2023 |> PIVOT(SUM(revenue) FOR quarter IN ('Q1', 'Q2', 'Q3', 'Q4'))"); + + // pivot pipe operator with aliases + dialects.verified_stmt("SELECT * FROM monthly_sales |> PIVOT(SUM(sales) FOR quarter IN ('Q1', 'Q2')) AS quarterly_sales"); + dialects.verified_stmt("SELECT * FROM data |> PIVOT(AVG(price) FOR category IN ('A', 'B', 'C')) AS avg_by_category"); + dialects.verified_stmt("SELECT * FROM sales |> PIVOT(COUNT(*) AS transactions, SUM(amount) AS total FOR region IN ('North', 'South')) AS regional_summary"); + + // pivot pipe operator with implicit aliases (without AS keyword) + dialects.verified_query_with_canonical( + "SELECT * FROM monthly_sales |> PIVOT(SUM(sales) FOR quarter IN ('Q1', 'Q2')) quarterly_sales", + "SELECT * FROM monthly_sales |> PIVOT(SUM(sales) FOR quarter IN ('Q1', 'Q2')) AS quarterly_sales", + ); + dialects.verified_query_with_canonical( + "SELECT * FROM data |> PIVOT(AVG(price) FOR category IN ('A', 'B', 'C')) avg_by_category", + "SELECT * FROM data |> PIVOT(AVG(price) FOR category IN ('A', 'B', 'C')) AS avg_by_category", + ); + + // unpivot pipe operator basic usage + dialects + .verified_stmt("SELECT * FROM sales |> UNPIVOT(revenue FOR quarter IN (Q1, Q2, Q3, Q4))"); + dialects.verified_stmt("SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C))"); + dialects.verified_stmt( + "SELECT * FROM metrics |> UNPIVOT(measurement FOR metric_type IN (cpu, memory, disk))", + ); + + // unpivot pipe operator with multiple columns + dialects.verified_stmt("SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (jan, feb, mar, apr, may, jun))"); + dialects.verified_stmt( + "SELECT * FROM report |> UNPIVOT(score FOR subject IN (math, science, english, history))", + ); + + // unpivot pipe operator mixed with other pipe operators + dialects.verified_stmt("SELECT * FROM sales_data |> WHERE year = 2023 |> UNPIVOT(revenue FOR quarter IN (Q1, Q2, Q3, Q4))"); + + // unpivot pipe operator with aliases + dialects.verified_stmt("SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (Q1, Q2)) AS unpivoted_sales"); + dialects.verified_stmt( + "SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C)) AS transformed_data", + ); + dialects.verified_stmt("SELECT * FROM metrics |> UNPIVOT(measurement FOR metric_type IN (cpu, memory)) AS metric_measurements"); + + // unpivot pipe operator with implicit aliases (without AS keyword) + dialects.verified_query_with_canonical( + "SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (Q1, Q2)) unpivoted_sales", + "SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (Q1, Q2)) AS unpivoted_sales", + ); + dialects.verified_query_with_canonical( + "SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C)) transformed_data", + "SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C)) AS transformed_data", + ); + // many pipes dialects.verified_stmt( "SELECT * FROM CustomerOrders |> AGGREGATE SUM(cost) AS total_cost GROUP BY customer_id, state, item_type |> EXTEND COUNT(*) OVER (PARTITION BY customer_id) AS num_orders |> WHERE num_orders > 1 |> AGGREGATE AVG(total_cost) AS average GROUP BY state DESC, item_type ASC", ); + + // join pipe operator - INNER JOIN + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt("SELECT * FROM users |> INNER JOIN orders ON users.id = orders.user_id"); + + // join pipe operator - LEFT JOIN + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt( + "SELECT * FROM users |> LEFT OUTER JOIN orders ON users.id = orders.user_id", + ); + + // join pipe operator - RIGHT JOIN + dialects.verified_stmt("SELECT * FROM users |> RIGHT JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt( + "SELECT * FROM users |> RIGHT OUTER JOIN orders ON users.id = orders.user_id", + ); + + // join pipe operator - FULL JOIN + dialects.verified_stmt("SELECT * FROM users |> FULL JOIN orders ON users.id = orders.user_id"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> FULL OUTER JOIN orders ON users.id = orders.user_id", + "SELECT * FROM users |> FULL JOIN orders ON users.id = orders.user_id", + ); + + // join pipe operator - CROSS JOIN + dialects.verified_stmt("SELECT * FROM users |> CROSS JOIN orders"); + + // join pipe operator with USING + dialects.verified_query_with_canonical( + "SELECT * FROM users |> JOIN orders USING (user_id)", + "SELECT * FROM users |> JOIN orders USING(user_id)", + ); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> LEFT JOIN orders USING (user_id, order_date)", + "SELECT * FROM users |> LEFT JOIN orders USING(user_id, order_date)", + ); + + // join pipe operator with alias + dialects.verified_query_with_canonical( + "SELECT * FROM users |> JOIN orders o ON users.id = o.user_id", + "SELECT * FROM users |> JOIN orders AS o ON users.id = o.user_id", + ); + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders AS o ON users.id = o.user_id"); + + // join pipe operator with complex ON condition + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id AND orders.status = 'active'"); + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders ON users.id = orders.user_id AND orders.amount > 100"); + + // multiple join pipe operators + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id |> JOIN products ON orders.product_id = products.id"); + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders ON users.id = orders.user_id |> RIGHT JOIN products ON orders.product_id = products.id"); + + // join pipe operator with other pipe operators + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id |> WHERE orders.amount > 100"); + dialects.verified_stmt("SELECT * FROM users |> WHERE users.active = true |> LEFT JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id |> SELECT users.name, orders.amount"); +} + +#[test] +fn parse_pipeline_operator_negative_tests() { + let dialects = all_dialects_where(|d| d.supports_pipe_operator()); + + // Test that plain EXCEPT without DISTINCT fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> EXCEPT (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that EXCEPT ALL fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> EXCEPT ALL (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that EXCEPT BY NAME without DISTINCT fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> EXCEPT BY NAME (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that EXCEPT ALL BY NAME fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements( + "SELECT * FROM users |> EXCEPT ALL BY NAME (SELECT * FROM admins)" + ) + .unwrap_err() + ); + + // Test that plain INTERSECT without DISTINCT fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> INTERSECT (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that INTERSECT ALL fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> INTERSECT ALL (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that INTERSECT BY NAME without DISTINCT fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> INTERSECT BY NAME (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that INTERSECT ALL BY NAME fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements( + "SELECT * FROM users |> INTERSECT ALL BY NAME (SELECT * FROM admins)" + ) + .unwrap_err() + ); + + // Test that CALL without function name fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL") + .is_err()); + + // Test that CALL without parentheses fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL my_function") + .is_err()); + + // Test that CALL with invalid function syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL 123invalid") + .is_err()); + + // Test that CALL with malformed arguments fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL my_function(,)") + .is_err()); + + // Test that CALL with invalid alias syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL my_function() AS") + .is_err()); + + // Test that PIVOT without parentheses fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT SUM(amount) FOR month IN ('Jan')") + .is_err()); + + // Test that PIVOT without FOR keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) month IN ('Jan'))") + .is_err()); + + // Test that PIVOT without IN keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) FOR month ('Jan'))") + .is_err()); + + // Test that PIVOT with empty IN list fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) FOR month IN ())") + .is_err()); + + // Test that PIVOT with invalid alias syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) FOR month IN ('Jan')) AS") + .is_err()); + + // Test UNPIVOT negative cases + + // Test that UNPIVOT without parentheses fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT value FOR name IN col1, col2") + .is_err()); + + // Test that UNPIVOT without FOR keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value name IN (col1, col2))") + .is_err()); + + // Test that UNPIVOT without IN keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name (col1, col2))") + .is_err()); + + // Test that UNPIVOT with missing value column fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(FOR name IN (col1, col2))") + .is_err()); + + // Test that UNPIVOT with missing name column fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR IN (col1, col2))") + .is_err()); + + // Test that UNPIVOT with empty IN list fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name IN ())") + .is_err()); + + // Test that UNPIVOT with invalid alias syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name IN (col1, col2)) AS") + .is_err()); + + // Test that UNPIVOT with missing closing parenthesis fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name IN (col1, col2)") + .is_err()); + + // Test that JOIN without table name fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN ON users.id = orders.user_id") + .is_err()); + + // Test that CROSS JOIN with ON condition fails + assert!(dialects + .parse_sql_statements( + "SELECT * FROM users |> CROSS JOIN orders ON users.id = orders.user_id" + ) + .is_err()); + + // Test that CROSS JOIN with USING condition fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CROSS JOIN orders USING (user_id)") + .is_err()); + + // Test that JOIN with empty USING list fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN orders USING ()") + .is_err()); + + // Test that JOIN with malformed ON condition fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN orders ON") + .is_err()); + + // Test that JOIN with invalid USING syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN orders USING user_id") + .is_err()); } #[test] From 9ffc546870a46eebb96a64e127b2190a05fa2baf Mon Sep 17 00:00:00 2001 From: Simon Vandel Sillesen Date: Tue, 1 Jul 2025 13:19:40 +0200 Subject: [PATCH 173/286] Make `GenericDialect` support from-first syntax (#1911) --- src/dialect/generic.rs | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 0b8822474..59671e211 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -112,6 +112,10 @@ impl Dialect for GenericDialect { true } + fn supports_from_first_select(&self) -> bool { + true + } + fn supports_asc_desc_in_column_definition(&self) -> bool { true } From f32a41a00451888fea24ee3487e303c85b8204b3 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Tue, 1 Jul 2025 13:52:29 +0200 Subject: [PATCH 174/286] Redshift utf8 idents (#1915) --- src/dialect/redshift.rs | 10 ++++++---- tests/sqlparser_common.rs | 9 ++++++++- tests/sqlparser_redshift.rs | 5 +++++ 3 files changed, 19 insertions(+), 5 deletions(-) diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index feccca5dd..9ad9c5fdb 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -80,13 +80,15 @@ impl Dialect for RedshiftSqlDialect { } fn is_identifier_start(&self, ch: char) -> bool { - // Extends Postgres dialect with sharp - PostgreSqlDialect {}.is_identifier_start(ch) || ch == '#' + // Extends Postgres dialect with sharp and UTF-8 multibyte chars + // https://docs.aws.amazon.com/redshift/latest/dg/r_names.html + PostgreSqlDialect {}.is_identifier_start(ch) || ch == '#' || !ch.is_ascii() } fn is_identifier_part(&self, ch: char) -> bool { - // Extends Postgres dialect with sharp - PostgreSqlDialect {}.is_identifier_part(ch) || ch == '#' + // Extends Postgres dialect with sharp and UTF-8 multibyte chars + // https://docs.aws.amazon.com/redshift/latest/dg/r_names.html + PostgreSqlDialect {}.is_identifier_part(ch) || ch == '#' || !ch.is_ascii() } /// redshift has `CONVERT(type, value)` instead of `CONVERT(value, type)` diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 42bdeae17..10a792528 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11068,10 +11068,17 @@ fn parse_non_latin_identifiers() { Box::new(RedshiftSqlDialect {}), Box::new(MySqlDialect {}), ]); - supported_dialects.verified_stmt("SELECT a.説明 FROM test.public.inter01 AS a"); supported_dialects.verified_stmt("SELECT a.説明 FROM inter01 AS a, inter01_transactions AS b WHERE a.説明 = b.取引 GROUP BY a.説明"); supported_dialects.verified_stmt("SELECT 説明, hühnervögel, garçon, Москва, 東京 FROM inter01"); + + let supported_dialects = TestedDialects::new(vec![ + Box::new(GenericDialect {}), + Box::new(DuckDbDialect {}), + Box::new(PostgreSqlDialect {}), + Box::new(MsSqlDialect {}), + Box::new(MySqlDialect {}), + ]); assert!(supported_dialects .parse_sql_statements("SELECT 💝 FROM table1") .is_err()); diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index be2b67223..d539adf66 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -402,3 +402,8 @@ fn parse_extract_single_quotes() { fn parse_string_literal_backslash_escape() { redshift().one_statement_parses_to(r#"SELECT 'l\'auto'"#, "SELECT 'l''auto'"); } + +#[test] +fn parse_utf8_multibyte_idents() { + redshift().verified_stmt("SELECT 🚀.city AS 🎸 FROM customers AS 🚀"); +} From a3398223d7d6fcc0e5329aec76ceef040aeabaa9 Mon Sep 17 00:00:00 2001 From: Ryan Schneider Date: Wed, 2 Jul 2025 05:57:08 -0700 Subject: [PATCH 175/286] DuckDB: Add support for multiple `TRIM` arguments (#1916) Co-authored-by: Ifeanyi Ubah --- src/parser/mod.rs | 2 +- tests/sqlparser_common.rs | 1 - tests/sqlparser_duckdb.rs | 30 ++++++++++++++++++++++++++++++ 3 files changed, 31 insertions(+), 2 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 4f8f1b85a..a3856d708 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -2576,7 +2576,7 @@ impl<'a> Parser<'a> { trim_characters: None, }) } else if self.consume_token(&Token::Comma) - && dialect_of!(self is SnowflakeDialect | BigQueryDialect | GenericDialect) + && dialect_of!(self is DuckDbDialect | SnowflakeDialect | BigQueryDialect | GenericDialect) { let characters = self.parse_comma_separated(Parser::parse_expr)?; self.expect_token(&Token::RParen)?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 10a792528..61a4de402 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7762,7 +7762,6 @@ fn parse_trim() { Box::new(MySqlDialect {}), //Box::new(BigQueryDialect {}), Box::new(SQLiteDialect {}), - Box::new(DuckDbDialect {}), ]); assert_eq!( diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 44cb22cea..371d3aace 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -24,6 +24,7 @@ use test_utils::*; use sqlparser::ast::*; use sqlparser::dialect::{DuckDbDialect, GenericDialect}; +use sqlparser::parser::ParserError; fn duckdb() -> TestedDialects { TestedDialects::new(vec![Box::new(DuckDbDialect {})]) @@ -830,3 +831,32 @@ fn parse_use() { ]))) ); } + +#[test] +fn test_duckdb_trim() { + let real_sql = r#"SELECT customer_id, TRIM(item_price_id, '"', "a") AS item_price_id FROM models_staging.subscriptions"#; + assert_eq!(duckdb().verified_stmt(real_sql).to_string(), real_sql); + + let sql_only_select = "SELECT TRIM('xyz', 'a')"; + let select = duckdb().verified_only_select(sql_only_select); + assert_eq!( + &Expr::Trim { + expr: Box::new(Expr::Value( + Value::SingleQuotedString("xyz".to_owned()).with_empty_span() + )), + trim_where: None, + trim_what: None, + trim_characters: Some(vec![Expr::Value( + Value::SingleQuotedString("a".to_owned()).with_empty_span() + )]), + }, + expr_from_projection(only(&select.projection)) + ); + + // missing comma separation + let error_sql = "SELECT TRIM('xyz' 'a')"; + assert_eq!( + ParserError::ParserError("Expected: ), found: 'a'".to_owned()), + duckdb().parse_sql_statements(error_sql).unwrap_err() + ); +} From 015caca611f67c22d7e1da5df11a5313afd97509 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 3 Jul 2025 18:16:21 +0200 Subject: [PATCH 176/286] Redshift alter column type no set (#1912) --- src/ast/ddl.rs | 18 ++++++++++++++---- src/ast/spans.rs | 1 + src/dialect/mod.rs | 9 +++++++++ src/dialect/postgresql.rs | 4 ++++ src/parser/mod.rs | 30 ++++++++++++++++++++---------- tests/sqlparser_common.rs | 17 ++++++++--------- tests/sqlparser_postgres.rs | 6 ++---- 7 files changed, 58 insertions(+), 27 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index afe9f228c..9134412c4 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -893,7 +893,10 @@ pub enum AlterColumnOperation { data_type: DataType, /// PostgreSQL specific using: Option, + /// Set to true if the statement includes the `SET DATA TYPE` keywords + had_set: bool, }, + /// `ADD GENERATED { ALWAYS | BY DEFAULT } AS IDENTITY [ ( sequence_options ) ]` /// /// Note: this is a PostgreSQL-specific operation. @@ -914,12 +917,19 @@ impl fmt::Display for AlterColumnOperation { AlterColumnOperation::DropDefault => { write!(f, "DROP DEFAULT") } - AlterColumnOperation::SetDataType { data_type, using } => { + AlterColumnOperation::SetDataType { + data_type, + using, + had_set, + } => { + if *had_set { + write!(f, "SET DATA ")?; + } + write!(f, "TYPE {data_type}")?; if let Some(expr) = using { - write!(f, "SET DATA TYPE {data_type} USING {expr}") - } else { - write!(f, "SET DATA TYPE {data_type}") + write!(f, " USING {expr}")?; } + Ok(()) } AlterColumnOperation::AddGenerated { generated_as, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 78ed772bd..00882602c 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -924,6 +924,7 @@ impl Spanned for AlterColumnOperation { AlterColumnOperation::SetDataType { data_type: _, using, + had_set: _, } => using.as_ref().map_or(Span::empty(), |u| u.span()), AlterColumnOperation::AddGenerated { .. } => Span::empty(), } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 028aa58a7..b214699cf 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1060,6 +1060,15 @@ pub trait Dialect: Debug + Any { fn supports_space_separated_column_options(&self) -> bool { false } + + /// Returns true if the dialect supports the `USING` clause in an `ALTER COLUMN` statement. + /// Example: + /// ```sql + /// ALTER TABLE tbl ALTER COLUMN col SET DATA TYPE USING ` + /// ``` + fn supports_alter_column_type_using(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index a91ab598b..b2d4014cb 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -258,4 +258,8 @@ impl Dialect for PostgreSqlDialect { fn supports_set_names(&self) -> bool { true } + + fn supports_alter_column_type_using(&self) -> bool { + true + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a3856d708..6360817f6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8734,16 +8734,10 @@ impl<'a> Parser<'a> { } } else if self.parse_keywords(&[Keyword::DROP, Keyword::DEFAULT]) { AlterColumnOperation::DropDefault {} - } else if self.parse_keywords(&[Keyword::SET, Keyword::DATA, Keyword::TYPE]) - || (is_postgresql && self.parse_keyword(Keyword::TYPE)) - { - let data_type = self.parse_data_type()?; - let using = if is_postgresql && self.parse_keyword(Keyword::USING) { - Some(self.parse_expr()?) - } else { - None - }; - AlterColumnOperation::SetDataType { data_type, using } + } else if self.parse_keywords(&[Keyword::SET, Keyword::DATA, Keyword::TYPE]) { + self.parse_set_data_type(true)? + } else if self.parse_keyword(Keyword::TYPE) { + self.parse_set_data_type(false)? } else if self.parse_keywords(&[Keyword::ADD, Keyword::GENERATED]) { let generated_as = if self.parse_keyword(Keyword::ALWAYS) { Some(GeneratedAs::Always) @@ -8909,6 +8903,22 @@ impl<'a> Parser<'a> { Ok(operation) } + fn parse_set_data_type(&mut self, had_set: bool) -> Result { + let data_type = self.parse_data_type()?; + let using = if self.dialect.supports_alter_column_type_using() + && self.parse_keyword(Keyword::USING) + { + Some(self.parse_expr()?) + } else { + None + }; + Ok(AlterColumnOperation::SetDataType { + data_type, + using, + had_set, + }) + } + fn parse_part_or_partition(&mut self) -> Result { let keyword = self.expect_one_of_keywords(&[Keyword::PART, Keyword::PARTITION])?; match keyword { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 61a4de402..380bd4757 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -5057,22 +5057,21 @@ fn parse_alter_table_alter_column_type() { AlterColumnOperation::SetDataType { data_type: DataType::Text, using: None, + had_set: true, } ); } _ => unreachable!(), } + verified_stmt(&format!("{alter_stmt} ALTER COLUMN is_active TYPE TEXT")); - let dialect = TestedDialects::new(vec![Box::new(GenericDialect {})]); - - let res = - dialect.parse_sql_statements(&format!("{alter_stmt} ALTER COLUMN is_active TYPE TEXT")); - assert_eq!( - ParserError::ParserError("Expected: SET/DROP NOT NULL, SET DEFAULT, or SET DATA TYPE after ALTER COLUMN, found: TYPE".to_string()), - res.unwrap_err() - ); + let dialects = all_dialects_where(|d| d.supports_alter_column_type_using()); + dialects.verified_stmt(&format!( + "{alter_stmt} ALTER COLUMN is_active SET DATA TYPE TEXT USING 'text'" + )); - let res = dialect.parse_sql_statements(&format!( + let dialects = all_dialects_except(|d| d.supports_alter_column_type_using()); + let res = dialects.parse_sql_statements(&format!( "{alter_stmt} ALTER COLUMN is_active SET DATA TYPE TEXT USING 'text'" )); assert_eq!( diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 7b0a8c5da..e1a49c69c 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -764,10 +764,7 @@ fn parse_drop_extension() { #[test] fn parse_alter_table_alter_column() { - pg().one_statement_parses_to( - "ALTER TABLE tab ALTER COLUMN is_active TYPE TEXT USING 'text'", - "ALTER TABLE tab ALTER COLUMN is_active SET DATA TYPE TEXT USING 'text'", - ); + pg().verified_stmt("ALTER TABLE tab ALTER COLUMN is_active TYPE TEXT USING 'text'"); match alter_table_op( pg().verified_stmt( @@ -783,6 +780,7 @@ fn parse_alter_table_alter_column() { AlterColumnOperation::SetDataType { data_type: DataType::Text, using: Some(using_expr), + had_set: true, } ); } From 418b94227a5cf6ae721629f95a279c40b6a196ad Mon Sep 17 00:00:00 2001 From: carl <44021312+achristmascarl@users.noreply.github.com> Date: Thu, 3 Jul 2025 12:19:26 -0400 Subject: [PATCH 177/286] Postgres: support `ADD CONSTRAINT NOT VALID` and `VALIDATE CONSTRAINT` (#1908) --- src/ast/ddl.rs | 25 +++++++++++++++++--- src/ast/spans.rs | 6 ++++- src/keywords.rs | 1 + src/parser/mod.rs | 9 ++++++- src/test_utils.rs | 31 ++++++++++++++---------- tests/sqlparser_common.rs | 2 +- tests/sqlparser_postgres.rs | 47 ++++++++++++++++++++++++++++++++++--- 7 files changed, 99 insertions(+), 22 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 9134412c4..9d5002039 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -67,8 +67,11 @@ impl fmt::Display for ReplicaIdentity { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum AlterTableOperation { - /// `ADD ` - AddConstraint(TableConstraint), + /// `ADD [NOT VALID]` + AddConstraint { + constraint: TableConstraint, + not_valid: bool, + }, /// `ADD [COLUMN] [IF NOT EXISTS] ` AddColumn { /// `[COLUMN]`. @@ -344,6 +347,10 @@ pub enum AlterTableOperation { equals: bool, value: ValueWithSpan, }, + /// `VALIDATE CONSTRAINT ` + ValidateConstraint { + name: Ident, + }, } /// An `ALTER Policy` (`Statement::AlterPolicy`) operation @@ -494,7 +501,16 @@ impl fmt::Display for AlterTableOperation { display_separated(new_partitions, " "), ine = if *if_not_exists { " IF NOT EXISTS" } else { "" } ), - AlterTableOperation::AddConstraint(c) => write!(f, "ADD {c}"), + AlterTableOperation::AddConstraint { + not_valid, + constraint, + } => { + write!(f, "ADD {constraint}")?; + if *not_valid { + write!(f, " NOT VALID")?; + } + Ok(()) + } AlterTableOperation::AddColumn { column_keyword, if_not_exists, @@ -772,6 +788,9 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::ReplicaIdentity { identity } => { write!(f, "REPLICA IDENTITY {identity}") } + AlterTableOperation::ValidateConstraint { name } => { + write!(f, "VALIDATE CONSTRAINT {name}") + } } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 00882602c..26205496a 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1075,7 +1075,10 @@ impl Spanned for CreateTableOptions { impl Spanned for AlterTableOperation { fn span(&self) -> Span { match self { - AlterTableOperation::AddConstraint(table_constraint) => table_constraint.span(), + AlterTableOperation::AddConstraint { + constraint, + not_valid: _, + } => constraint.span(), AlterTableOperation::AddColumn { column_keyword: _, if_not_exists: _, @@ -1196,6 +1199,7 @@ impl Spanned for AlterTableOperation { AlterTableOperation::AutoIncrement { value, .. } => value.span(), AlterTableOperation::Lock { .. } => Span::empty(), AlterTableOperation::ReplicaIdentity { .. } => Span::empty(), + AlterTableOperation::ValidateConstraint { name } => name.span, } } } diff --git a/src/keywords.rs b/src/keywords.rs index a8bbca3d3..49a54e8a9 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -981,6 +981,7 @@ define_keywords!( UUID, VACUUM, VALID, + VALIDATE, VALIDATION_MODE, VALUE, VALUES, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6360817f6..1c40a9649 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8477,7 +8477,11 @@ impl<'a> Parser<'a> { pub fn parse_alter_table_operation(&mut self) -> Result { let operation = if self.parse_keyword(Keyword::ADD) { if let Some(constraint) = self.parse_optional_table_constraint()? { - AlterTableOperation::AddConstraint(constraint) + let not_valid = self.parse_keywords(&[Keyword::NOT, Keyword::VALID]); + AlterTableOperation::AddConstraint { + constraint, + not_valid, + } } else if dialect_of!(self is ClickHouseDialect|GenericDialect) && self.parse_keyword(Keyword::PROJECTION) { @@ -8886,6 +8890,9 @@ impl<'a> Parser<'a> { }; AlterTableOperation::ReplicaIdentity { identity } + } else if self.parse_keywords(&[Keyword::VALIDATE, Keyword::CONSTRAINT]) { + let name = self.parse_identifier()?; + AlterTableOperation::ValidateConstraint { name } } else { let options: Vec = self.parse_options_with_keywords(&[Keyword::SET, Keyword::TBLPROPERTIES])?; diff --git a/src/test_utils.rs b/src/test_utils.rs index db7b3dd60..544ceaef8 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -479,20 +479,25 @@ pub fn index_column(stmt: Statement) -> Expr { } } Statement::AlterTable { operations, .. } => match operations.first().unwrap() { - AlterTableOperation::AddConstraint(TableConstraint::Index { columns, .. }) => { - columns.first().unwrap().column.expr.clone() - } - AlterTableOperation::AddConstraint(TableConstraint::Unique { columns, .. }) => { - columns.first().unwrap().column.expr.clone() - } - AlterTableOperation::AddConstraint(TableConstraint::PrimaryKey { columns, .. }) => { - columns.first().unwrap().column.expr.clone() + AlterTableOperation::AddConstraint { constraint, .. } => { + match constraint { + TableConstraint::Index { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::Unique { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::PrimaryKey { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::FulltextOrSpatial { + columns, + .. + } => columns.first().unwrap().column.expr.clone(), + _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), + } } - AlterTableOperation::AddConstraint(TableConstraint::FulltextOrSpatial { - columns, - .. - }) => columns.first().unwrap().column.expr.clone(), - _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), + _ => panic!("Expected a constraint"), }, _ => panic!("Expected CREATE INDEX, ALTER TABLE, or CREATE TABLE, got: {stmt:?}"), } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 380bd4757..9ca985b35 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4956,7 +4956,7 @@ fn parse_alter_table_constraints() { match alter_table_op(verified_stmt(&format!( "ALTER TABLE tab ADD {constraint_text}" ))) { - AlterTableOperation::AddConstraint(constraint) => { + AlterTableOperation::AddConstraint { constraint, .. } => { assert_eq!(constraint_text, constraint.to_string()); } _ => unreachable!(), diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index e1a49c69c..16ba5f237 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -606,9 +606,10 @@ fn parse_alter_table_constraints_unique_nulls_distinct() { .verified_stmt("ALTER TABLE t ADD CONSTRAINT b UNIQUE NULLS NOT DISTINCT (c)") { Statement::AlterTable { operations, .. } => match &operations[0] { - AlterTableOperation::AddConstraint(TableConstraint::Unique { - nulls_distinct, .. - }) => { + AlterTableOperation::AddConstraint { + constraint: TableConstraint::Unique { nulls_distinct, .. }, + .. + } => { assert_eq!(nulls_distinct, &NullsDistinctOption::NotDistinct) } _ => unreachable!(), @@ -6229,3 +6230,43 @@ fn parse_ts_datatypes() { _ => unreachable!(), } } + +#[test] +fn parse_alter_table_constraint_not_valid() { + match pg_and_generic().verified_stmt( + "ALTER TABLE foo ADD CONSTRAINT bar FOREIGN KEY (baz) REFERENCES other(ref) NOT VALID", + ) { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::AddConstraint { + constraint: TableConstraint::ForeignKey { + name: Some("bar".into()), + index_name: None, + columns: vec!["baz".into()], + foreign_table: ObjectName::from(vec!["other".into()]), + referred_columns: vec!["ref".into()], + on_delete: None, + on_update: None, + characteristics: None, + }, + not_valid: true, + }] + ); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_alter_table_validate_constraint() { + match pg_and_generic().verified_stmt("ALTER TABLE foo VALIDATE CONSTRAINT bar") { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::ValidateConstraint { name: "bar".into() }] + ); + } + _ => unreachable!(), + } +} From be2d2f14e740ac753d51609b026dae012b697353 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 3 Jul 2025 18:22:17 +0200 Subject: [PATCH 178/286] Add support for MySQL MEMBER OF (#1917) --- src/ast/mod.rs | 24 ++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/dialect/mod.rs | 2 ++ src/parser/mod.rs | 13 +++++++++++++ tests/sqlparser_mysql.rs | 25 +++++++++++++++++++++++++ 5 files changed, 65 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 19966d21c..9e5022609 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -1124,6 +1124,8 @@ pub enum Expr { /// [Databricks](https://docs.databricks.com/en/sql/language-manual/sql-ref-lambda-functions.html) /// [DuckDb](https://duckdb.org/docs/sql/functions/lambda.html) Lambda(LambdaFunction), + /// Checks membership of a value in a JSON array + MemberOf(MemberOf), } impl Expr { @@ -1912,6 +1914,7 @@ impl fmt::Display for Expr { } Expr::Prior(expr) => write!(f, "PRIOR {expr}"), Expr::Lambda(lambda) => write!(f, "{lambda}"), + Expr::MemberOf(member_of) => write!(f, "{member_of}"), } } } @@ -9831,6 +9834,27 @@ impl fmt::Display for NullInclusion { } } +/// Checks membership of a value in a JSON array +/// +/// Syntax: +/// ```sql +/// MEMBER OF() +/// ``` +/// [MySQL](https://dev.mysql.com/doc/refman/8.4/en/json-search-functions.html#operator_member-of) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct MemberOf { + pub value: Box, + pub array: Box, +} + +impl fmt::Display for MemberOf { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{} MEMBER OF({})", self.value, self.array) + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 26205496a..f8ffeb544 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1624,6 +1624,7 @@ impl Spanned for Expr { Expr::OuterJoin(expr) => expr.span(), Expr::Prior(expr) => expr.span(), Expr::Lambda(_) => Span::empty(), + Expr::MemberOf(member_of) => member_of.value.span().union(&member_of.array.span()), } } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index b214699cf..3345380cb 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -649,6 +649,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::REGEXP => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), + Token::Word(w) if w.keyword == Keyword::MEMBER => Ok(p!(Like)), _ => Ok(self.prec_unknown()), }, Token::Word(w) if w.keyword == Keyword::IS => Ok(p!(Is)), @@ -661,6 +662,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::REGEXP => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), + Token::Word(w) if w.keyword == Keyword::MEMBER => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::OPERATOR => Ok(p!(Between)), Token::Word(w) if w.keyword == Keyword::DIV => Ok(p!(MulDivModOp)), Token::Period => Ok(p!(Period)), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 1c40a9649..bfd753857 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3609,6 +3609,19 @@ impl<'a> Parser<'a> { self.expected("IN or BETWEEN after NOT", self.peek_token()) } } + Keyword::MEMBER => { + if self.parse_keyword(Keyword::OF) { + self.expect_token(&Token::LParen)?; + let array = self.parse_expr()?; + self.expect_token(&Token::RParen)?; + Ok(Expr::MemberOf(MemberOf { + value: Box::new(expr), + array: Box::new(array), + })) + } else { + self.expected("OF after MEMBER", self.peek_token()) + } + } // Can only happen if `get_next_precedence` got out of sync with this function _ => parser_err!( format!("No infix parser for token {:?}", tok.token), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index d2feee035..9224a003e 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -4109,3 +4109,28 @@ fn parse_alter_table_drop_index() { AlterTableOperation::DropIndex { name } if name.value == "idx_index" ); } + +#[test] +fn parse_json_member_of() { + mysql().verified_stmt(r#"SELECT 17 MEMBER OF('[23, "abc", 17, "ab", 10]')"#); + let sql = r#"SELECT 'ab' MEMBER OF('[23, "abc", 17, "ab", 10]')"#; + let stmt = mysql().verified_stmt(sql); + match stmt { + Statement::Query(query) => { + let select = query.body.as_select().unwrap(); + assert_eq!( + select.projection, + vec![SelectItem::UnnamedExpr(Expr::MemberOf(MemberOf { + value: Box::new(Expr::Value( + Value::SingleQuotedString("ab".to_string()).into() + )), + array: Box::new(Expr::Value( + Value::SingleQuotedString(r#"[23, "abc", 17, "ab", 10]"#.to_string()) + .into() + )), + }))] + ); + } + _ => panic!("Unexpected statement {stmt}"), + } +} From 9020385c027239979adbef5506c7be4a07cc594c Mon Sep 17 00:00:00 2001 From: feral-dot-io <70718208+feral-dot-io@users.noreply.github.com> Date: Thu, 3 Jul 2025 16:24:51 +0000 Subject: [PATCH 179/286] Add span for `Expr::TypedString` (#1919) --- src/ast/mod.rs | 2 +- src/ast/spans.rs | 1 - src/parser/mod.rs | 5 +- tests/sqlparser_bigquery.rs | 77 ++++++++++++++++++------ tests/sqlparser_common.rs | 106 ++++++++++++++++++++++++++-------- tests/sqlparser_databricks.rs | 6 +- tests/sqlparser_postgres.rs | 5 +- 7 files changed, 155 insertions(+), 47 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 9e5022609..d7e342bd8 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -996,7 +996,7 @@ pub enum Expr { data_type: DataType, /// The value of the constant. /// Hint: you can unwrap the string value using `value.into_string()`. - value: Value, + value: ValueWithSpan, }, /// Scalar function call e.g. `LEFT(foo, 5)` Function(Function), diff --git a/src/ast/spans.rs b/src/ast/spans.rs index f8ffeb544..0895b844f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1415,7 +1415,6 @@ impl Spanned for AssignmentTarget { /// f.e. `IS NULL ` reports as `::span`. /// /// Missing spans: -/// - [Expr::TypedString] # missing span for data_type /// - [Expr::MatchAgainst] # MySQL specific /// - [Expr::RLike] # MySQL specific /// - [Expr::Struct] # BigQuery specific diff --git a/src/parser/mod.rs b/src/parser/mod.rs index bfd753857..f38a360e5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1521,7 +1521,7 @@ impl<'a> Parser<'a> { DataType::Custom(..) => parser_err!("dummy", loc), data_type => Ok(Expr::TypedString { data_type, - value: parser.parse_value()?.value, + value: parser.parse_value()?, }), } })?; @@ -1708,10 +1708,9 @@ impl<'a> Parser<'a> { } fn parse_geometric_type(&mut self, kind: GeometricTypeKind) -> Result { - let value: Value = self.parse_value()?.value; Ok(Expr::TypedString { data_type: DataType::GeometricType(kind), - value, + value: self.parse_value()?, }) } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 2bcdb4e59..2ba54d3e1 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -906,7 +906,10 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), - value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + span: Span::empty(), + }, }], fields: vec![StructField { field_name: None, @@ -965,9 +968,12 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString( - r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() - ) + value: ValueWithSpan { + value: Value::SingleQuotedString( + r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() + ), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -998,7 +1004,12 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString( + "2008-12-25 15:30:00 America/Los_Angeles".into() + ), + span: Span::empty(), + }, }], fields: vec![StructField { field_name: None, @@ -1013,7 +1024,10 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: Value::SingleQuotedString("15:30:00".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("15:30:00".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1031,7 +1045,10 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1045,7 +1062,10 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1219,7 +1239,10 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), - value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1278,9 +1301,12 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString( - r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() - ) + value: ValueWithSpan { + value: Value::SingleQuotedString( + r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() + ), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1311,7 +1337,12 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString( + "2008-12-25 15:30:00 America/Los_Angeles".into() + ), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1326,7 +1357,10 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: Value::SingleQuotedString("15:30:00".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("15:30:00".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1344,7 +1378,10 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -1358,7 +1395,10 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, @@ -2393,7 +2433,10 @@ fn test_triple_quote_typed_strings() { assert_eq!( Expr::TypedString { data_type: DataType::JSON, - value: Value::TripleDoubleQuotedString(r#"{"foo":"bar's"}"#.into()) + value: ValueWithSpan { + value: Value::TripleDoubleQuotedString(r#"{"foo":"bar's"}"#.into()), + span: Span::empty(), + } }, expr ); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 9ca985b35..ac461bb28 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -5851,7 +5851,10 @@ fn parse_literal_date() { assert_eq!( &Expr::TypedString { data_type: DataType::Date, - value: Value::SingleQuotedString("1999-01-01".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01".into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -5864,7 +5867,10 @@ fn parse_literal_time() { assert_eq!( &Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: Value::SingleQuotedString("01:23:34".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("01:23:34".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -5877,7 +5883,10 @@ fn parse_literal_datetime() { assert_eq!( &Expr::TypedString { data_type: DataType::Datetime(None), - value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -5890,7 +5899,10 @@ fn parse_literal_timestamp_without_time_zone() { assert_eq!( &Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("1999-01-01 01:23:34".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -5905,7 +5917,10 @@ fn parse_literal_timestamp_with_time_zone() { assert_eq!( &Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::Tz), - value: Value::SingleQuotedString("1999-01-01 01:23:34Z".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34Z".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -6477,8 +6492,9 @@ fn parse_json_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString( - r#"{ + value: ValueWithSpan { + value: Value::SingleQuotedString( + r#"{ "id": 10, "type": "fruit", "name": "apple", @@ -6498,8 +6514,10 @@ fn parse_json_keyword() { ] } }"# - .to_string() - ) + .to_string() + ), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6511,7 +6529,10 @@ fn parse_typed_strings() { assert_eq!( Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString(r#"{"foo":"bar"}"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"{"foo":"bar"}"#.into()), + span: Span::empty(), + } }, expr ); @@ -6529,7 +6550,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"0"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"0"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6540,7 +6564,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"123456"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"123456"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6551,7 +6578,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"-3.14"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"-3.14"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6562,7 +6592,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"-0.54321"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"-0.54321"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6573,7 +6606,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"1.23456e05"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"1.23456e05"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6584,7 +6620,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"-9.876e-3"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"-9.876e-3"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -14833,7 +14872,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Point), - value: Value::SingleQuotedString("1,2".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2".to_string()), + span: Span::empty(), + }, } ); @@ -14842,7 +14884,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Line), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); @@ -14851,7 +14896,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::GeometricPath), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); let sql = "box '1,2,3,4'"; @@ -14859,7 +14907,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::GeometricBox), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); @@ -14868,7 +14919,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Circle), - value: Value::SingleQuotedString("1,2,3".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3".to_string()), + span: Span::empty(), + }, } ); @@ -14877,7 +14931,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Polygon), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); let sql = "lseg '1,2,3,4'"; @@ -14885,7 +14942,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::LineSegment), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); } diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index baf279fae..a27e0699b 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -19,6 +19,7 @@ use sqlparser::ast::helpers::attached_token::AttachedToken; use sqlparser::ast::*; use sqlparser::dialect::{DatabricksDialect, GenericDialect}; use sqlparser::parser::ParserError; +use sqlparser::tokenizer::Span; use test_utils::*; #[macro_use] @@ -328,7 +329,10 @@ fn data_type_timestamp_ntz() { databricks().verified_expr("TIMESTAMP_NTZ '2025-03-29T18:52:00'"), Expr::TypedString { data_type: DataType::TimestampNtz, - value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()) + value: ValueWithSpan { + value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()), + span: Span::empty(), + } } ); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 16ba5f237..db5c1611e 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5257,7 +5257,10 @@ fn parse_at_time_zone() { left: Box::new(Expr::AtTimeZone { timestamp: Box::new(Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("2001-09-28 01:00".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("2001-09-28 01:00".to_string()), + span: Span::empty(), + }, }), time_zone: Box::new(Expr::Cast { kind: CastKind::DoubleColon, From 239e30a97c5f088d39f1747fb108c8d8783a2e5c Mon Sep 17 00:00:00 2001 From: Sergey Olontsev Date: Thu, 3 Jul 2025 18:04:32 +0100 Subject: [PATCH 180/286] Support for Postgres `CREATE SERVER` (#1914) Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 69 ++++++++++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 2 + src/parser/mod.rs | 45 +++++++++++++++++++++ tests/sqlparser_postgres.rs | 79 +++++++++++++++++++++++++++++++++++++ 5 files changed, 196 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d7e342bd8..cffd32925 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3318,6 +3318,8 @@ pub enum Statement { secret_type: Ident, options: Vec, }, + /// A `CREATE SERVER` statement. + CreateServer(CreateServerStatement), /// ```sql /// CREATE POLICY /// ``` @@ -5178,6 +5180,9 @@ impl fmt::Display for Statement { write!(f, " )")?; Ok(()) } + Statement::CreateServer(stmt) => { + write!(f, "{stmt}") + } Statement::CreatePolicy { name, table_name, @@ -7976,6 +7981,70 @@ impl fmt::Display for SecretOption { } } +/// A `CREATE SERVER` statement. +/// +/// [PostgreSQL Documentation](https://www.postgresql.org/docs/current/sql-createserver.html) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateServerStatement { + pub name: ObjectName, + pub if_not_exists: bool, + pub server_type: Option, + pub version: Option, + pub foreign_data_wrapper: ObjectName, + pub options: Option>, +} + +impl fmt::Display for CreateServerStatement { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + let CreateServerStatement { + name, + if_not_exists, + server_type, + version, + foreign_data_wrapper, + options, + } = self; + + write!( + f, + "CREATE SERVER {if_not_exists}{name} ", + if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, + )?; + + if let Some(st) = server_type { + write!(f, "TYPE {st} ")?; + } + + if let Some(v) = version { + write!(f, "VERSION {v} ")?; + } + + write!(f, "FOREIGN DATA WRAPPER {foreign_data_wrapper}")?; + + if let Some(o) = options { + write!(f, " OPTIONS ({o})", o = display_comma_separated(o))?; + } + + Ok(()) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateServerOption { + pub key: Ident, + pub value: Ident, +} + +impl fmt::Display for CreateServerOption { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{} {}", self.key, self.value) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 0895b844f..1d790ab64 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -423,6 +423,7 @@ impl Spanned for Statement { Statement::CreateIndex(create_index) => create_index.span(), Statement::CreateRole { .. } => Span::empty(), Statement::CreateSecret { .. } => Span::empty(), + Statement::CreateServer { .. } => Span::empty(), Statement::CreateConnector { .. } => Span::empty(), Statement::AlterTable { name, diff --git a/src/keywords.rs b/src/keywords.rs index 49a54e8a9..738651504 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -816,6 +816,7 @@ define_keywords!( SERDE, SERDEPROPERTIES, SERIALIZABLE, + SERVER, SERVICE, SESSION, SESSION_USER, @@ -1017,6 +1018,7 @@ define_keywords!( WITHOUT, WITHOUT_ARRAY_WRAPPER, WORK, + WRAPPER, WRITE, XML, XMLNAMESPACES, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f38a360e5..289dd6b33 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4674,6 +4674,8 @@ impl<'a> Parser<'a> { self.parse_create_procedure(or_alter) } else if self.parse_keyword(Keyword::CONNECTOR) { self.parse_create_connector() + } else if self.parse_keyword(Keyword::SERVER) { + self.parse_pg_create_server() } else { self.expected("an object type after CREATE", self.peek_token()) } @@ -16009,6 +16011,49 @@ impl<'a> Parser<'a> { Ok(sequence_options) } + /// Parse a `CREATE SERVER` statement. + /// + /// See [Statement::CreateServer] + pub fn parse_pg_create_server(&mut self) -> Result { + let ine = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let name = self.parse_object_name(false)?; + + let server_type = if self.parse_keyword(Keyword::TYPE) { + Some(self.parse_identifier()?) + } else { + None + }; + + let version = if self.parse_keyword(Keyword::VERSION) { + Some(self.parse_identifier()?) + } else { + None + }; + + self.expect_keywords(&[Keyword::FOREIGN, Keyword::DATA, Keyword::WRAPPER])?; + let foreign_data_wrapper = self.parse_object_name(false)?; + + let mut options = None; + if self.parse_keyword(Keyword::OPTIONS) { + self.expect_token(&Token::LParen)?; + options = Some(self.parse_comma_separated(|p| { + let key = p.parse_identifier()?; + let value = p.parse_identifier()?; + Ok(CreateServerOption { key, value }) + })?); + self.expect_token(&Token::RParen)?; + } + + Ok(Statement::CreateServer(CreateServerStatement { + name, + if_not_exists: ine, + server_type, + version, + foreign_data_wrapper, + options, + })) + } + /// The index of the first unprocessed token. pub fn index(&self) -> usize { self.index diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index db5c1611e..48792025a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -6273,3 +6273,82 @@ fn parse_alter_table_validate_constraint() { _ => unreachable!(), } } + +#[test] +fn parse_create_server() { + let test_cases = vec![ + ( + "CREATE SERVER myserver FOREIGN DATA WRAPPER postgres_fdw", + CreateServerStatement { + name: ObjectName::from(vec!["myserver".into()]), + if_not_exists: false, + server_type: None, + version: None, + foreign_data_wrapper: ObjectName::from(vec!["postgres_fdw".into()]), + options: None, + }, + ), + ( + "CREATE SERVER IF NOT EXISTS myserver TYPE 'server_type' VERSION 'server_version' FOREIGN DATA WRAPPER postgres_fdw", + CreateServerStatement { + name: ObjectName::from(vec!["myserver".into()]), + if_not_exists: true, + server_type: Some(Ident { + value: "server_type".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }), + version: Some(Ident { + value: "server_version".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }), + foreign_data_wrapper: ObjectName::from(vec!["postgres_fdw".into()]), + options: None, + } + ), + ( + "CREATE SERVER myserver2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host 'foo', dbname 'foodb', port '5432')", + CreateServerStatement { + name: ObjectName::from(vec!["myserver2".into()]), + if_not_exists: false, + server_type: None, + version: None, + foreign_data_wrapper: ObjectName::from(vec!["postgres_fdw".into()]), + options: Some(vec![ + CreateServerOption { + key: "host".into(), + value: Ident { + value: "foo".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }, + }, + CreateServerOption { + key: "dbname".into(), + value: Ident { + value: "foodb".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }, + }, + CreateServerOption { + key: "port".into(), + value: Ident { + value: "5432".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }, + }, + ]), + } + ) + ]; + + for (sql, expected) in test_cases { + let Statement::CreateServer(stmt) = pg_and_generic().verified_stmt(sql) else { + unreachable!() + }; + assert_eq!(stmt, expected); + } +} From 942d747d893ad1c25fc32b002066b3bda72071b6 Mon Sep 17 00:00:00 2001 From: Elia Perantoni Date: Fri, 4 Jul 2025 18:21:31 +0200 Subject: [PATCH 181/286] Change tag and policy names to `ObjectName` (#1892) --- src/ast/ddl.rs | 2 +- src/ast/mod.rs | 4 +-- src/dialect/snowflake.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_snowflake.rs | 64 +++++++++++++++++++++++++++--------- 5 files changed, 53 insertions(+), 21 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 9d5002039..7e46a59fc 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1709,7 +1709,7 @@ pub struct ColumnPolicyProperty { /// ``` /// [Snowflake]: https://docs.snowflake.com/en/sql-reference/sql/create-table pub with: bool, - pub policy_name: Ident, + pub policy_name: ObjectName, pub using_columns: Option>, } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index cffd32925..7db4af111 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -9372,12 +9372,12 @@ impl Display for RowAccessPolicy { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct Tag { - pub key: Ident, + pub key: ObjectName, pub value: String, } impl Tag { - pub fn new(key: Ident, value: String) -> Self { + pub fn new(key: ObjectName, value: String) -> Self { Self { key, value } } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index ba28a8ec5..ee770b0ff 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -1189,7 +1189,7 @@ fn parse_column_policy_property( parser: &mut Parser, with: bool, ) -> Result { - let policy_name = parser.parse_identifier()?; + let policy_name = parser.parse_object_name(false)?; let using_columns = if parser.parse_keyword(Keyword::USING) { parser.expect_token(&Token::LParen)?; let columns = parser.parse_comma_separated(|p| p.parse_identifier())?; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 289dd6b33..612144a61 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7884,7 +7884,7 @@ impl<'a> Parser<'a> { } pub(crate) fn parse_tag(&mut self) -> Result { - let name = self.parse_identifier()?; + let name = self.parse_object_name(false)?; self.expect_token(&Token::Eq)?; let value = self.parse_literal_string()?; diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 3ffd33392..8a1558b2d 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -270,8 +270,8 @@ fn test_snowflake_create_table_with_tag() { assert_eq!("my_table", name.to_string()); assert_eq!( Some(vec![ - Tag::new("A".into(), "TAG A".to_string()), - Tag::new("B".into(), "TAG B".to_string()) + Tag::new(ObjectName::from(vec![Ident::new("A")]), "TAG A".to_string()), + Tag::new(ObjectName::from(vec![Ident::new("B")]), "TAG B".to_string()) ]), with_tags ); @@ -291,8 +291,8 @@ fn test_snowflake_create_table_with_tag() { assert_eq!("my_table", name.to_string()); assert_eq!( Some(vec![ - Tag::new("A".into(), "TAG A".to_string()), - Tag::new("B".into(), "TAG B".to_string()) + Tag::new(ObjectName::from(vec![Ident::new("A")]), "TAG A".to_string()), + Tag::new(ObjectName::from(vec![Ident::new("B")]), "TAG B".to_string()) ]), with_tags ); @@ -731,7 +731,7 @@ fn test_snowflake_create_table_with_columns_masking_policy() { option: ColumnOption::Policy(ColumnPolicy::MaskingPolicy( ColumnPolicyProperty { with, - policy_name: "p".into(), + policy_name: ObjectName::from(vec![Ident::new("p")]), using_columns, } )) @@ -765,7 +765,7 @@ fn test_snowflake_create_table_with_columns_projection_policy() { option: ColumnOption::Policy(ColumnPolicy::ProjectionPolicy( ColumnPolicyProperty { with, - policy_name: "p".into(), + policy_name: ObjectName::from(vec![Ident::new("p")]), using_columns: None, } )) @@ -802,8 +802,14 @@ fn test_snowflake_create_table_with_columns_tags() { option: ColumnOption::Tags(TagsColumnOption { with, tags: vec![ - Tag::new("A".into(), "TAG A".into()), - Tag::new("B".into(), "TAG B".into()), + Tag::new( + ObjectName::from(vec![Ident::new("A")]), + "TAG A".into() + ), + Tag::new( + ObjectName::from(vec![Ident::new("B")]), + "TAG B".into() + ), ] }), }], @@ -846,7 +852,7 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Policy(ColumnPolicy::MaskingPolicy( ColumnPolicyProperty { with: true, - policy_name: "p1".into(), + policy_name: ObjectName::from(vec![Ident::new("p1")]), using_columns: Some(vec!["a".into(), "b".into()]), } )), @@ -856,8 +862,14 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Tags(TagsColumnOption { with: true, tags: vec![ - Tag::new("A".into(), "TAG A".into()), - Tag::new("B".into(), "TAG B".into()), + Tag::new( + ObjectName::from(vec![Ident::new("A")]), + "TAG A".into() + ), + Tag::new( + ObjectName::from(vec![Ident::new("B")]), + "TAG B".into() + ), ] }), } @@ -878,7 +890,7 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Policy(ColumnPolicy::ProjectionPolicy( ColumnPolicyProperty { with: false, - policy_name: "p2".into(), + policy_name: ObjectName::from(vec![Ident::new("p2")]), using_columns: None, } )), @@ -888,8 +900,14 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Tags(TagsColumnOption { with: false, tags: vec![ - Tag::new("C".into(), "TAG C".into()), - Tag::new("D".into(), "TAG D".into()), + Tag::new( + ObjectName::from(vec![Ident::new("C")]), + "TAG C".into() + ), + Tag::new( + ObjectName::from(vec![Ident::new("D")]), + "TAG D".into() + ), ] }), } @@ -942,8 +960,8 @@ fn test_snowflake_create_iceberg_table_all_options() { with_aggregation_policy.map(|name| name.to_string()) ); assert_eq!(Some(vec![ - Tag::new("A".into(), "TAG A".into()), - Tag::new("B".into(), "TAG B".into()), + Tag::new(ObjectName::from(vec![Ident::new("A")]), "TAG A".into()), + Tag::new(ObjectName::from(vec![Ident::new("B")]), "TAG B".into()), ]), with_tags); } @@ -4172,3 +4190,17 @@ fn test_snowflake_create_view_with_multiple_column_options() { r#"CREATE VIEW X (COL WITH TAG (pii='email') COMMENT 'foobar') AS SELECT * FROM Y"#; snowflake().verified_stmt(create_view_with_tag); } + +#[test] +fn test_snowflake_create_view_with_composite_tag() { + let create_view_with_tag = + r#"CREATE VIEW X (COL WITH TAG (foo.bar.baz.pii='email')) AS SELECT * FROM Y"#; + snowflake().verified_stmt(create_view_with_tag); +} + +#[test] +fn test_snowflake_create_view_with_composite_policy_name() { + let create_view_with_tag = + r#"CREATE VIEW X (COL WITH MASKING POLICY foo.bar.baz) AS SELECT * FROM Y"#; + snowflake().verified_stmt(create_view_with_tag); +} From b0bcc46e227162b4a9147c9c1e0a88bf3e0b5e1f Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 4 Jul 2025 21:04:51 +0200 Subject: [PATCH 182/286] Add support for NULL escape char in pattern match searches (#1913) --- src/ast/mod.rs | 12 ++++++------ src/parser/mod.rs | 4 ++-- tests/sqlparser_common.rs | 25 +++++++++++++++++++++---- 3 files changed, 29 insertions(+), 12 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 7db4af111..425e1fb60 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -809,7 +809,7 @@ pub enum Expr { any: bool, expr: Box, pattern: Box, - escape_char: Option, + escape_char: Option, }, /// `ILIKE` (case-insensitive `LIKE`) ILike { @@ -819,14 +819,14 @@ pub enum Expr { any: bool, expr: Box, pattern: Box, - escape_char: Option, + escape_char: Option, }, /// SIMILAR TO regex SimilarTo { negated: bool, expr: Box, pattern: Box, - escape_char: Option, + escape_char: Option, }, /// MySQL: RLIKE regex or REGEXP regex RLike { @@ -1488,7 +1488,7 @@ impl fmt::Display for Expr { } => match escape_char { Some(ch) => write!( f, - "{} {}LIKE {}{} ESCAPE '{}'", + "{} {}LIKE {}{} ESCAPE {}", expr, if *negated { "NOT " } else { "" }, if *any { "ANY " } else { "" }, @@ -1513,7 +1513,7 @@ impl fmt::Display for Expr { } => match escape_char { Some(ch) => write!( f, - "{} {}ILIKE {}{} ESCAPE '{}'", + "{} {}ILIKE {}{} ESCAPE {}", expr, if *negated { "NOT " } else { "" }, if *any { "ANY" } else { "" }, @@ -1568,7 +1568,7 @@ impl fmt::Display for Expr { } => match escape_char { Some(ch) => write!( f, - "{} {}SIMILAR TO {} ESCAPE '{}'", + "{} {}SIMILAR TO {} ESCAPE {}", expr, if *negated { "NOT " } else { "" }, pattern, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 612144a61..32f8a97e4 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3654,9 +3654,9 @@ impl<'a> Parser<'a> { } /// Parse the `ESCAPE CHAR` portion of `LIKE`, `ILIKE`, and `SIMILAR TO` - pub fn parse_escape_char(&mut self) -> Result, ParserError> { + pub fn parse_escape_char(&mut self) -> Result, ParserError> { if self.parse_keyword(Keyword::ESCAPE) { - Ok(Some(self.parse_literal_string()?)) + Ok(Some(self.parse_value()?.into())) } else { Ok(None) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index ac461bb28..1bdd302e8 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -2040,7 +2040,7 @@ fn parse_ilike() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), any: false, }, select.selection.unwrap() @@ -2104,7 +2104,7 @@ fn parse_like() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), any: false, }, select.selection.unwrap() @@ -2167,7 +2167,24 @@ fn parse_similar_to() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), + }, + select.selection.unwrap() + ); + + let sql = &format!( + "SELECT * FROM customers WHERE name {}SIMILAR TO '%a' ESCAPE NULL", + if negated { "NOT " } else { "" } + ); + let select = verified_only_select(sql); + assert_eq!( + Expr::SimilarTo { + expr: Box::new(Expr::Identifier(Ident::new("name"))), + negated, + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), + escape_char: Some(Value::Null), }, select.selection.unwrap() ); @@ -2185,7 +2202,7 @@ fn parse_similar_to() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), })), select.selection.unwrap() ); From d2466af20a5336c1554bdb82b0c84d04a20db26e Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sat, 5 Jul 2025 08:18:58 +0200 Subject: [PATCH 183/286] Add support for dropping multiple columns in Snowflake (#1918) --- src/ast/ddl.rs | 8 ++++---- src/ast/spans.rs | 4 ++-- src/dialect/mod.rs | 5 +++++ src/dialect/snowflake.rs | 4 ++++ src/parser/mod.rs | 8 ++++++-- tests/sqlparser_common.rs | 7 +++++-- tests/sqlparser_mysql.rs | 4 ++-- 7 files changed, 28 insertions(+), 12 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 7e46a59fc..51e057840 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -140,10 +140,10 @@ pub enum AlterTableOperation { name: Ident, drop_behavior: Option, }, - /// `DROP [ COLUMN ] [ IF EXISTS ] [ CASCADE ]` + /// `DROP [ COLUMN ] [ IF EXISTS ] [ , , ... ] [ CASCADE ]` DropColumn { has_column_keyword: bool, - column_name: Ident, + column_names: Vec, if_exists: bool, drop_behavior: Option, }, @@ -631,7 +631,7 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::DropIndex { name } => write!(f, "DROP INDEX {name}"), AlterTableOperation::DropColumn { has_column_keyword, - column_name, + column_names: column_name, if_exists, drop_behavior, } => write!( @@ -639,7 +639,7 @@ impl fmt::Display for AlterTableOperation { "DROP {}{}{}{}", if *has_column_keyword { "COLUMN " } else { "" }, if *if_exists { "IF EXISTS " } else { "" }, - column_name, + display_comma_separated(column_name), match drop_behavior { None => "", Some(DropBehavior::Restrict) => " RESTRICT", diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 1d790ab64..144de5923 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1112,10 +1112,10 @@ impl Spanned for AlterTableOperation { } => name.span, AlterTableOperation::DropColumn { has_column_keyword: _, - column_name, + column_names, if_exists: _, drop_behavior: _, - } => column_name.span, + } => union_spans(column_names.iter().map(|i| i.span)), AlterTableOperation::AttachPartition { partition } => partition.span(), AlterTableOperation::DetachPartition { partition } => partition.span(), AlterTableOperation::FreezePartition { diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 3345380cb..bc3c55554 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1071,6 +1071,11 @@ pub trait Dialect: Debug + Any { fn supports_alter_column_type_using(&self) -> bool { false } + + /// Returns true if the dialect supports `ALTER TABLE tbl DROP COLUMN c1, ..., cn` + fn supports_comma_separated_drop_column_list(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index ee770b0ff..f56db9b83 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -364,6 +364,10 @@ impl Dialect for SnowflakeDialect { fn supports_space_separated_column_options(&self) -> bool { true } + + fn supports_comma_separated_drop_column_list(&self) -> bool { + true + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 32f8a97e4..839d36459 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8675,11 +8675,15 @@ impl<'a> Parser<'a> { } else { let has_column_keyword = self.parse_keyword(Keyword::COLUMN); // [ COLUMN ] let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); - let column_name = self.parse_identifier()?; + let column_names = if self.dialect.supports_comma_separated_drop_column_list() { + self.parse_comma_separated(Parser::parse_identifier)? + } else { + vec![self.parse_identifier()?] + }; let drop_behavior = self.parse_optional_drop_behavior(); AlterTableOperation::DropColumn { has_column_keyword, - column_name, + column_names, if_exists, drop_behavior, } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 1bdd302e8..ed9bb704d 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4996,15 +4996,18 @@ fn parse_alter_table_drop_column() { "ALTER TABLE tab DROP is_active CASCADE", ); + let dialects = all_dialects_where(|d| d.supports_comma_separated_drop_column_list()); + dialects.verified_stmt("ALTER TABLE tbl DROP COLUMN c1, c2, c3"); + fn check_one(constraint_text: &str) { match alter_table_op(verified_stmt(&format!("ALTER TABLE tab {constraint_text}"))) { AlterTableOperation::DropColumn { has_column_keyword: true, - column_name, + column_names, if_exists, drop_behavior, } => { - assert_eq!("is_active", column_name.to_string()); + assert_eq!("is_active", column_names.first().unwrap().to_string()); assert!(if_exists); match drop_behavior { None => assert!(constraint_text.ends_with(" is_active")), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 9224a003e..79e2259b2 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2876,7 +2876,7 @@ fn parse_alter_table_with_algorithm() { vec![ AlterTableOperation::DropColumn { has_column_keyword: true, - column_name: Ident::new("password_digest"), + column_names: vec![Ident::new("password_digest")], if_exists: false, drop_behavior: None, }, @@ -2924,7 +2924,7 @@ fn parse_alter_table_with_lock() { vec![ AlterTableOperation::DropColumn { has_column_keyword: true, - column_name: Ident::new("password_digest"), + column_names: vec![Ident::new("password_digest")], if_exists: false, drop_behavior: None, }, From ed8757f2f0596c3e27b7f878e012bfc4d6da32a2 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sat, 5 Jul 2025 08:40:35 +0200 Subject: [PATCH 184/286] Align Snowflake dialect to new test of reserved keywords (#1924) --- src/dialect/snowflake.rs | 5 ++--- tests/sqlparser_snowflake.rs | 30 +++++++++++++++++++++++++++++- 2 files changed, 31 insertions(+), 4 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index f56db9b83..212cf217d 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -301,9 +301,8 @@ impl Dialect for SnowflakeDialect { true } - fn is_select_item_alias(&self, explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { - explicit - || match kw { + fn is_column_alias(&self, kw: &Keyword, parser: &mut Parser) -> bool { + match kw { // The following keywords can be considered an alias as long as // they are not followed by other tokens that may change their meaning // e.g. `SELECT * EXCEPT (col1) FROM tbl` diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 8a1558b2d..e7393d3f8 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3415,10 +3415,38 @@ fn parse_ls_and_rm() { .unwrap(); } +#[test] +fn test_sql_keywords_as_select_item_ident() { + // Some keywords that should be parsed as an alias + let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT", "SORT"]; + for kw in unreserved_kws { + snowflake().verified_stmt(&format!("SELECT 1, {kw}")); + } + + // Some keywords that should not be parsed as an alias + let reserved_kws = vec![ + "FROM", + "GROUP", + "HAVING", + "INTERSECT", + "INTO", + "ORDER", + "SELECT", + "UNION", + "WHERE", + "WITH", + ]; + for kw in reserved_kws { + assert!(snowflake() + .parse_sql_statements(&format!("SELECT 1, {kw}")) + .is_err()); + } +} + #[test] fn test_sql_keywords_as_select_item_aliases() { // Some keywords that should be parsed as an alias - let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT"]; + let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT", "SORT"]; for kw in unreserved_kws { snowflake() .one_statement_parses_to(&format!("SELECT 1 {kw}"), &format!("SELECT 1 AS {kw}")); From cf9e50474ec6c59027bcc3e2f418c5ff6947920a Mon Sep 17 00:00:00 2001 From: Simon Vandel Sillesen Date: Sun, 6 Jul 2025 08:57:20 +0200 Subject: [PATCH 185/286] Make `GenericDialect` support trailing commas in projections (#1921) --- src/dialect/generic.rs | 4 ++++ tests/sqlparser_common.rs | 2 +- 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 59671e211..5e9f2e4e2 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -116,6 +116,10 @@ impl Dialect for GenericDialect { true } + fn supports_projection_trailing_commas(&self) -> bool { + true + } + fn supports_asc_desc_in_column_definition(&self) -> bool { true } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index ed9bb704d..56b015fc5 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11184,7 +11184,7 @@ fn parse_trailing_comma() { trailing_commas.verified_stmt(r#"SELECT "from" FROM "from""#); // doesn't allow any trailing commas - let trailing_commas = TestedDialects::new(vec![Box::new(GenericDialect {})]); + let trailing_commas = TestedDialects::new(vec![Box::new(PostgreSqlDialect {})]); assert_eq!( trailing_commas From f2fba48a7a7c2cba43baafa3c87b9f7abe85c869 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sun, 6 Jul 2025 08:58:19 +0200 Subject: [PATCH 186/286] Add support for several Snowflake grant statements (#1922) --- src/ast/mod.rs | 54 ++++++++++++++++++++++++++++++++++++ src/parser/mod.rs | 58 +++++++++++++++++++++++++++++++++++++++ tests/sqlparser_common.rs | 6 ++++ 3 files changed, 118 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 425e1fb60..93ef76639 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6928,12 +6928,24 @@ pub enum GrantObjects { AllSequencesInSchema { schemas: Vec }, /// Grant privileges on `ALL TABLES IN SCHEMA [, ...]` AllTablesInSchema { schemas: Vec }, + /// Grant privileges on `ALL VIEWS IN SCHEMA [, ...]` + AllViewsInSchema { schemas: Vec }, + /// Grant privileges on `ALL MATERIALIZED VIEWS IN SCHEMA [, ...]` + AllMaterializedViewsInSchema { schemas: Vec }, + /// Grant privileges on `ALL EXTERNAL TABLES IN SCHEMA [, ...]` + AllExternalTablesInSchema { schemas: Vec }, /// Grant privileges on `FUTURE SCHEMAS IN DATABASE [, ...]` FutureSchemasInDatabase { databases: Vec }, /// Grant privileges on `FUTURE TABLES IN SCHEMA [, ...]` FutureTablesInSchema { schemas: Vec }, /// Grant privileges on `FUTURE VIEWS IN SCHEMA [, ...]` FutureViewsInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE EXTERNAL TABLES IN SCHEMA [, ...]` + FutureExternalTablesInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE MATERIALIZED VIEWS IN SCHEMA [, ...]` + FutureMaterializedViewsInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE SEQUENCES IN SCHEMA [, ...]` + FutureSequencesInSchema { schemas: Vec }, /// Grant privileges on specific databases Databases(Vec), /// Grant privileges on specific schemas @@ -7002,6 +7014,27 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::AllExternalTablesInSchema { schemas } => { + write!( + f, + "ALL EXTERNAL TABLES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::AllViewsInSchema { schemas } => { + write!( + f, + "ALL VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::AllMaterializedViewsInSchema { schemas } => { + write!( + f, + "ALL MATERIALIZED VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } GrantObjects::FutureSchemasInDatabase { databases } => { write!( f, @@ -7016,6 +7049,13 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::FutureExternalTablesInSchema { schemas } => { + write!( + f, + "FUTURE EXTERNAL TABLES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } GrantObjects::FutureViewsInSchema { schemas } => { write!( f, @@ -7023,6 +7063,20 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::FutureMaterializedViewsInSchema { schemas } => { + write!( + f, + "FUTURE MATERIALIZED VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureSequencesInSchema { schemas } => { + write!( + f, + "FUTURE SEQUENCES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } GrantObjects::ResourceMonitors(objects) => { write!(f, "RESOURCE MONITOR {}", display_comma_separated(objects)) } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 839d36459..a94c7b433 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13901,6 +13901,35 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllTablesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::EXTERNAL, + Keyword::TABLES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllExternalTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::MATERIALIZED, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllMaterializedViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[ Keyword::FUTURE, Keyword::SCHEMAS, @@ -13919,6 +13948,16 @@ impl<'a> Parser<'a> { Some(GrantObjects::FutureTablesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::EXTERNAL, + Keyword::TABLES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureExternalTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[ Keyword::FUTURE, Keyword::VIEWS, @@ -13928,6 +13967,16 @@ impl<'a> Parser<'a> { Some(GrantObjects::FutureViewsInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::MATERIALIZED, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureMaterializedViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[ Keyword::ALL, Keyword::SEQUENCES, @@ -13937,6 +13986,15 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllSequencesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::SEQUENCES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureSequencesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[Keyword::RESOURCE, Keyword::MONITOR]) { Some(GrantObjects::ResourceMonitors(self.parse_comma_separated( |p| p.parse_object_name_with_wildcards(false, true), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 56b015fc5..0d3c4d824 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9434,6 +9434,9 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO APPLICATION role1"); verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO APPLICATION ROLE role1"); verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO SHARE share1"); + verified_stmt("GRANT SELECT ON ALL VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON ALL MATERIALIZED VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON ALL EXTERNAL TABLES IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO a:b"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO GROUP group1"); verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST"); @@ -9447,7 +9450,10 @@ fn parse_grant() { .verified_stmt("GRANT SELECT ON [my_table] TO [public]"); verified_stmt("GRANT SELECT ON FUTURE SCHEMAS IN DATABASE db1 TO ROLE role1"); verified_stmt("GRANT SELECT ON FUTURE TABLES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE EXTERNAL TABLES IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT SELECT ON FUTURE VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE MATERIALIZED VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE SEQUENCES IN SCHEMA db1.sc1 TO ROLE role1"); } #[test] From 1a33abda63658f19902026df99826cce74812c02 Mon Sep 17 00:00:00 2001 From: Sergey Olontsev Date: Sun, 6 Jul 2025 08:06:20 +0100 Subject: [PATCH 187/286] Clickhouse: support empty parenthesized options (#1925) --- src/parser/mod.rs | 4 ++-- tests/sqlparser_clickhouse.rs | 4 ++++ 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a94c7b433..61944b8f7 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -16253,9 +16253,9 @@ impl<'a> Parser<'a> { fn parse_parenthesized_identifiers(&mut self) -> Result, ParserError> { self.expect_token(&Token::LParen)?; - let partitions = self.parse_comma_separated(|p| p.parse_identifier())?; + let idents = self.parse_comma_separated0(|p| p.parse_identifier(), Token::RParen)?; self.expect_token(&Token::RParen)?; - Ok(partitions) + Ok(idents) } fn parse_column_position(&mut self) -> Result, ParserError> { diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 0288c6d2c..1d8669a22 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -224,6 +224,10 @@ fn parse_create_table() { clickhouse().verified_stmt( r#"CREATE TABLE "x" ("a" "int") ENGINE = MergeTree ORDER BY "x" AS SELECT * FROM "t" WHERE true"#, ); + clickhouse().one_statement_parses_to( + "CREATE TABLE x (a int) ENGINE = MergeTree() ORDER BY a", + "CREATE TABLE x (a INT) ENGINE = MergeTree ORDER BY a", + ); } #[test] From 93450cc2505388f546623c9807f1320acf4fe880 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 7 Jul 2025 17:13:57 +0200 Subject: [PATCH 188/286] Add Snowflake `COPY/REVOKE CURRENT GRANTS` option (#1926) --- src/ast/mod.rs | 27 +++++++++++++++++++++++++++ src/parser/mod.rs | 10 ++++++++++ tests/sqlparser_common.rs | 2 ++ tests/sqlparser_mysql.rs | 1 + 4 files changed, 40 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 93ef76639..8da6bbe76 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3996,6 +3996,7 @@ pub enum Statement { with_grant_option: bool, as_grantor: Option, granted_by: Option, + current_grants: Option, }, /// ```sql /// DENY privileges ON object TO grantees @@ -4312,6 +4313,28 @@ pub enum Statement { Return(ReturnStatement), } +/// ```sql +/// {COPY | REVOKE} CURRENT GRANTS +/// ``` +/// +/// - [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/grant-ownership#optional-parameters) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum CurrentGrantsKind { + CopyCurrentGrants, + RevokeCurrentGrants, +} + +impl fmt::Display for CurrentGrantsKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + CurrentGrantsKind::CopyCurrentGrants => write!(f, "COPY CURRENT GRANTS"), + CurrentGrantsKind::RevokeCurrentGrants => write!(f, "REVOKE CURRENT GRANTS"), + } + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -5715,6 +5738,7 @@ impl fmt::Display for Statement { with_grant_option, as_grantor, granted_by, + current_grants, } => { write!(f, "GRANT {privileges} ")?; if let Some(objects) = objects { @@ -5724,6 +5748,9 @@ impl fmt::Display for Statement { if *with_grant_option { write!(f, " WITH GRANT OPTION")?; } + if let Some(current_grants) = current_grants { + write!(f, " {current_grants}")?; + } if let Some(grantor) = as_grantor { write!(f, " AS {grantor}")?; } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 61944b8f7..c4d0508df 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13794,6 +13794,15 @@ impl<'a> Parser<'a> { let with_grant_option = self.parse_keywords(&[Keyword::WITH, Keyword::GRANT, Keyword::OPTION]); + let current_grants = + if self.parse_keywords(&[Keyword::COPY, Keyword::CURRENT, Keyword::GRANTS]) { + Some(CurrentGrantsKind::CopyCurrentGrants) + } else if self.parse_keywords(&[Keyword::REVOKE, Keyword::CURRENT, Keyword::GRANTS]) { + Some(CurrentGrantsKind::RevokeCurrentGrants) + } else { + None + }; + let as_grantor = if self.parse_keywords(&[Keyword::AS]) { Some(self.parse_identifier()?) } else { @@ -13813,6 +13822,7 @@ impl<'a> Parser<'a> { with_grant_option, as_grantor, granted_by, + current_grants, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 0d3c4d824..27fe09c73 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9440,6 +9440,8 @@ fn parse_grant() { verified_stmt("GRANT USAGE ON SCHEMA sc1 TO a:b"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO GROUP group1"); verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST"); + verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST COPY CURRENT GRANTS"); + verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST REVOKE CURRENT GRANTS"); verified_stmt("GRANT USAGE ON DATABASE db1 TO ROLE role1"); verified_stmt("GRANT USAGE ON WAREHOUSE wh1 TO ROLE role1"); verified_stmt("GRANT OWNERSHIP ON INTEGRATION int1 TO ROLE role1"); diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 79e2259b2..44f365844 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3616,6 +3616,7 @@ fn parse_grant() { with_grant_option, as_grantor: _, granted_by, + current_grants: _, } = stmt { assert_eq!( From b1a6d11e123ff5432d24e4baf3028176649692f3 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 10 Jul 2025 17:15:50 +0200 Subject: [PATCH 189/286] Add support for Snowflake identifier function (#1929) --- src/ast/mod.rs | 22 ++++++ src/ast/spans.rs | 4 ++ src/dialect/mod.rs | 15 +++- src/dialect/snowflake.rs | 21 +++++- src/parser/mod.rs | 116 ++++++++++++++++++------------ tests/sqlparser_common.rs | 6 +- tests/sqlparser_snowflake.rs | 135 +++++++++++++++++++++++++++++++++++ 7 files changed, 268 insertions(+), 51 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8da6bbe76..619730e21 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -344,12 +344,14 @@ impl fmt::Display for ObjectName { #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum ObjectNamePart { Identifier(Ident), + Function(ObjectNamePartFunction), } impl ObjectNamePart { pub fn as_ident(&self) -> Option<&Ident> { match self { ObjectNamePart::Identifier(ident) => Some(ident), + ObjectNamePart::Function(_) => None, } } } @@ -358,10 +360,30 @@ impl fmt::Display for ObjectNamePart { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { ObjectNamePart::Identifier(ident) => write!(f, "{ident}"), + ObjectNamePart::Function(func) => write!(f, "{func}"), } } } +/// An object name part that consists of a function that dynamically +/// constructs identifiers. +/// +/// - [Snowflake](https://docs.snowflake.com/en/sql-reference/identifier-literal) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ObjectNamePartFunction { + pub name: Ident, + pub args: Vec, +} + +impl fmt::Display for ObjectNamePartFunction { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}(", self.name)?; + write!(f, "{})", display_comma_separated(&self.args)) + } +} + /// Represents an Array Expression, either /// `ARRAY[..]`, or `[..]` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 144de5923..a1b2e4e0d 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1671,6 +1671,10 @@ impl Spanned for ObjectNamePart { fn span(&self) -> Span { match self { ObjectNamePart::Identifier(ident) => ident.span, + ObjectNamePart::Function(func) => func + .name + .span + .union(&union_spans(func.args.iter().map(|i| i.span()))), } } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index bc3c55554..8f9dd617f 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -49,7 +49,7 @@ pub use self::postgresql::PostgreSqlDialect; pub use self::redshift::RedshiftSqlDialect; pub use self::snowflake::SnowflakeDialect; pub use self::sqlite::SQLiteDialect; -use crate::ast::{ColumnOption, Expr, GranteesType, Statement}; +use crate::ast::{ColumnOption, Expr, GranteesType, Ident, ObjectNamePart, Statement}; pub use crate::keywords; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; @@ -1076,6 +1076,19 @@ pub trait Dialect: Debug + Any { fn supports_comma_separated_drop_column_list(&self) -> bool { false } + + /// Returns true if the dialect considers the specified ident as a function + /// that returns an identifier. Typically used to generate identifiers + /// programmatically. + /// + /// - [Snowflake](https://docs.snowflake.com/en/sql-reference/identifier-literal) + fn is_identifier_generating_function_name( + &self, + _ident: &Ident, + _name_parts: &[ObjectNamePart], + ) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 212cf217d..e005300ef 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -25,8 +25,8 @@ use crate::ast::helpers::stmt_data_loading::{ use crate::ast::{ ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, - IdentityPropertyOrder, ObjectName, RowAccessPolicy, ShowObjects, SqlOption, Statement, - TagsColumnOption, WrappedCollection, + IdentityPropertyOrder, ObjectName, ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, + Statement, TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -367,6 +367,23 @@ impl Dialect for SnowflakeDialect { fn supports_comma_separated_drop_column_list(&self) -> bool { true } + + fn is_identifier_generating_function_name( + &self, + ident: &Ident, + name_parts: &[ObjectNamePart], + ) -> bool { + ident.quote_style.is_none() + && ident.value.to_lowercase() == "identifier" + && !name_parts + .iter() + .any(|p| matches!(p, ObjectNamePart::Function(_))) + } + + // For example: `SELECT IDENTIFIER('alias1').* FROM tbl AS alias1` + fn supports_select_expr_star(&self) -> bool { + true + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c4d0508df..5a42bb9e1 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10353,70 +10353,92 @@ impl<'a> Parser<'a> { } } - /// Parse a possibly qualified, possibly quoted identifier, optionally allowing for wildcards, + /// Parse a possibly qualified, possibly quoted identifier, e.g. + /// `foo` or `myschema."table" + /// + /// The `in_table_clause` parameter indicates whether the object name is a table in a FROM, JOIN, + /// or similar table clause. Currently, this is used only to support unquoted hyphenated identifiers + /// in this context on BigQuery. + pub fn parse_object_name(&mut self, in_table_clause: bool) -> Result { + self.parse_object_name_inner(in_table_clause, false) + } + + /// Parse a possibly qualified, possibly quoted identifier, e.g. + /// `foo` or `myschema."table" + /// + /// The `in_table_clause` parameter indicates whether the object name is a table in a FROM, JOIN, + /// or similar table clause. Currently, this is used only to support unquoted hyphenated identifiers + /// in this context on BigQuery. + /// + /// The `allow_wildcards` parameter indicates whether to allow for wildcards in the object name /// e.g. *, *.*, `foo`.*, or "foo"."bar" - fn parse_object_name_with_wildcards( + fn parse_object_name_inner( &mut self, in_table_clause: bool, allow_wildcards: bool, ) -> Result { - let mut idents = vec![]; - + let mut parts = vec![]; if dialect_of!(self is BigQueryDialect) && in_table_clause { loop { let (ident, end_with_period) = self.parse_unquoted_hyphenated_identifier()?; - idents.push(ident); + parts.push(ObjectNamePart::Identifier(ident)); if !self.consume_token(&Token::Period) && !end_with_period { break; } } } else { loop { - let ident = if allow_wildcards && self.peek_token().token == Token::Mul { + if allow_wildcards && self.peek_token().token == Token::Mul { let span = self.next_token().span; - Ident { + parts.push(ObjectNamePart::Identifier(Ident { value: Token::Mul.to_string(), quote_style: None, span, + })); + } else if dialect_of!(self is BigQueryDialect) && in_table_clause { + let (ident, end_with_period) = self.parse_unquoted_hyphenated_identifier()?; + parts.push(ObjectNamePart::Identifier(ident)); + if !self.consume_token(&Token::Period) && !end_with_period { + break; } + } else if self.dialect.supports_object_name_double_dot_notation() + && parts.len() == 1 + && matches!(self.peek_token().token, Token::Period) + { + // Empty string here means default schema + parts.push(ObjectNamePart::Identifier(Ident::new(""))); } else { - if self.dialect.supports_object_name_double_dot_notation() - && idents.len() == 1 - && self.consume_token(&Token::Period) + let ident = self.parse_identifier()?; + let part = if self + .dialect + .is_identifier_generating_function_name(&ident, &parts) { - // Empty string here means default schema - idents.push(Ident::new("")); - } - self.parse_identifier()? - }; - idents.push(ident); + self.expect_token(&Token::LParen)?; + let args: Vec = + self.parse_comma_separated0(Self::parse_function_args, Token::RParen)?; + self.expect_token(&Token::RParen)?; + ObjectNamePart::Function(ObjectNamePartFunction { name: ident, args }) + } else { + ObjectNamePart::Identifier(ident) + }; + parts.push(part); + } + if !self.consume_token(&Token::Period) { break; } } } - Ok(ObjectName::from(idents)) - } - - /// Parse a possibly qualified, possibly quoted identifier, e.g. - /// `foo` or `myschema."table" - /// - /// The `in_table_clause` parameter indicates whether the object name is a table in a FROM, JOIN, - /// or similar table clause. Currently, this is used only to support unquoted hyphenated identifiers - /// in this context on BigQuery. - pub fn parse_object_name(&mut self, in_table_clause: bool) -> Result { - let ObjectName(mut idents) = - self.parse_object_name_with_wildcards(in_table_clause, false)?; // BigQuery accepts any number of quoted identifiers of a table name. // https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#quoted_identifiers if dialect_of!(self is BigQueryDialect) - && idents.iter().any(|part| { + && parts.iter().any(|part| { part.as_ident() .is_some_and(|ident| ident.value.contains('.')) }) { - idents = idents + parts = parts .into_iter() .flat_map(|part| match part.as_ident() { Some(ident) => ident @@ -10435,7 +10457,7 @@ impl<'a> Parser<'a> { .collect() } - Ok(ObjectName(idents)) + Ok(ObjectName(parts)) } /// Parse identifiers @@ -14006,25 +14028,25 @@ impl<'a> Parser<'a> { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) } else if self.parse_keywords(&[Keyword::RESOURCE, Keyword::MONITOR]) { - Some(GrantObjects::ResourceMonitors(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ResourceMonitors( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::COMPUTE, Keyword::POOL]) { - Some(GrantObjects::ComputePools(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ComputePools( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::FAILOVER, Keyword::GROUP]) { - Some(GrantObjects::FailoverGroup(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::FailoverGroup( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::REPLICATION, Keyword::GROUP]) { - Some(GrantObjects::ReplicationGroup(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ReplicationGroup( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::EXTERNAL, Keyword::VOLUME]) { - Some(GrantObjects::ExternalVolumes(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ExternalVolumes( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else { let object_type = self.parse_one_of_keywords(&[ Keyword::SEQUENCE, @@ -14041,7 +14063,7 @@ impl<'a> Parser<'a> { Keyword::CONNECTION, ]); let objects = - self.parse_comma_separated(|p| p.parse_object_name_with_wildcards(false, true)); + self.parse_comma_separated(|p| p.parse_object_name_inner(false, true)); match object_type { Some(Keyword::DATABASE) => Some(GrantObjects::Databases(objects?)), Some(Keyword::SCHEMA) => Some(GrantObjects::Schemas(objects?)), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 27fe09c73..1de1d93fc 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -1232,7 +1232,6 @@ fn parse_select_expr_star() { "SELECT 2. * 3 FROM T", ); dialects.verified_only_select("SELECT myfunc().* FROM T"); - dialects.verified_only_select("SELECT myfunc().* EXCEPT (foo) FROM T"); // Invalid let res = dialects.parse_sql_statements("SELECT foo.*.* FROM T"); @@ -1240,6 +1239,11 @@ fn parse_select_expr_star() { ParserError::ParserError("Expected: end of statement, found: .".to_string()), res.unwrap_err() ); + + let dialects = all_dialects_where(|d| { + d.supports_select_expr_star() && d.supports_select_wildcard_except() + }); + dialects.verified_only_select("SELECT myfunc().* EXCEPT (foo) FROM T"); } #[test] diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index e7393d3f8..6081e3183 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4232,3 +4232,138 @@ fn test_snowflake_create_view_with_composite_policy_name() { r#"CREATE VIEW X (COL WITH MASKING POLICY foo.bar.baz) AS SELECT * FROM Y"#; snowflake().verified_stmt(create_view_with_tag); } + +#[test] +fn test_snowflake_identifier_function() { + // Using IDENTIFIER to reference a column + match &snowflake() + .verified_only_select("SELECT identifier('email') FROM customers") + .projection[0] + { + SelectItem::UnnamedExpr(Expr::Function(Function { name, args, .. })) => { + assert_eq!(*name, ObjectName::from(vec![Ident::new("identifier")])); + assert_eq!( + *args, + FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("email".to_string()).into() + )))], + clauses: vec![], + duplicate_treatment: None + }) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a case-sensitive column + match &snowflake() + .verified_only_select(r#"SELECT identifier('"Email"') FROM customers"#) + .projection[0] + { + SelectItem::UnnamedExpr(Expr::Function(Function { name, args, .. })) => { + assert_eq!(*name, ObjectName::from(vec![Ident::new("identifier")])); + assert_eq!( + *args, + FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("\"Email\"".to_string()).into() + )))], + clauses: vec![], + duplicate_treatment: None + }) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference an alias of a table + match &snowflake() + .verified_only_select("SELECT identifier('alias1').* FROM tbl AS alias1") + .projection[0] + { + SelectItem::QualifiedWildcard( + SelectItemQualifiedWildcardKind::Expr(Expr::Function(Function { name, args, .. })), + _, + ) => { + assert_eq!(*name, ObjectName::from(vec![Ident::new("identifier")])); + assert_eq!( + *args, + FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("alias1".to_string()).into() + )))], + clauses: vec![], + duplicate_treatment: None + }) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a database + match snowflake().verified_stmt("CREATE DATABASE IDENTIFIER('tbl')") { + Statement::CreateDatabase { db_name, .. } => { + assert_eq!( + db_name, + ObjectName(vec![ObjectNamePart::Function(ObjectNamePartFunction { + name: Ident::new("IDENTIFIER"), + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("tbl".to_string()).into() + )))] + })]) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a schema + match snowflake().verified_stmt("CREATE SCHEMA IDENTIFIER('db1.sc1')") { + Statement::CreateSchema { schema_name, .. } => { + assert_eq!( + schema_name, + SchemaName::Simple(ObjectName(vec![ObjectNamePart::Function( + ObjectNamePartFunction { + name: Ident::new("IDENTIFIER"), + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("db1.sc1".to_string()).into() + )))] + } + )])) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a table + match snowflake().verified_stmt("CREATE TABLE IDENTIFIER('tbl') (id INT)") { + Statement::CreateTable(CreateTable { name, .. }) => { + assert_eq!( + name, + ObjectName(vec![ObjectNamePart::Function(ObjectNamePartFunction { + name: Ident::new("IDENTIFIER"), + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("tbl".to_string()).into() + )))] + })]) + ); + } + _ => unreachable!(), + } + + // Cannot have more than one IDENTIFIER part in an object name + assert_eq!( + snowflake() + .parse_sql_statements( + "CREATE TABLE IDENTIFIER('db1').IDENTIFIER('sc1').IDENTIFIER('tbl') (id INT)" + ) + .is_err(), + true + ); + assert_eq!( + snowflake() + .parse_sql_statements("CREATE TABLE IDENTIFIER('db1')..IDENTIFIER('tbl') (id INT)") + .is_err(), + true + ); +} From 8f1414efffa530af17c0853a8de4f0f0c73edacc Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 10 Jul 2025 17:28:51 +0200 Subject: [PATCH 190/286] Add support for granting privileges to procedures and functions in Snowflake (#1930) --- src/ast/mod.rs | 33 +++++++++++++++++++++++++++++++++ src/parser/mod.rs | 34 ++++++++++++++++++++++++++++++++++ tests/sqlparser_common.rs | 2 ++ 3 files changed, 69 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 619730e21..75e88f8a8 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -7023,6 +7023,25 @@ pub enum GrantObjects { ReplicationGroup(Vec), /// Grant privileges on external volumes ExternalVolumes(Vec), + /// Grant privileges on a procedure. In dialects that + /// support overloading, the argument types must be specified. + /// + /// For example: + /// `GRANT USAGE ON PROCEDURE foo(varchar) TO ROLE role1` + Procedure { + name: ObjectName, + arg_types: Vec, + }, + + /// Grant privileges on a function. In dialects that + /// support overloading, the argument types must be specified. + /// + /// For example: + /// `GRANT USAGE ON FUNCTION foo(varchar) TO ROLE role1` + Function { + name: ObjectName, + arg_types: Vec, + }, } impl fmt::Display for GrantObjects { @@ -7147,6 +7166,20 @@ impl fmt::Display for GrantObjects { GrantObjects::ExternalVolumes(objects) => { write!(f, "EXTERNAL VOLUME {}", display_comma_separated(objects)) } + GrantObjects::Procedure { name, arg_types } => { + write!(f, "PROCEDURE {name}")?; + if !arg_types.is_empty() { + write!(f, "({})", display_comma_separated(arg_types))?; + } + Ok(()) + } + GrantObjects::Function { name, arg_types } => { + write!(f, "FUNCTION {name}")?; + if !arg_types.is_empty() { + write!(f, "({})", display_comma_separated(arg_types))?; + } + Ok(()) + } } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 5a42bb9e1..b00cd16d7 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -14061,6 +14061,8 @@ impl<'a> Parser<'a> { Keyword::INTEGRATION, Keyword::USER, Keyword::CONNECTION, + Keyword::PROCEDURE, + Keyword::FUNCTION, ]); let objects = self.parse_comma_separated(|p| p.parse_object_name_inner(false, true)); @@ -14073,6 +14075,13 @@ impl<'a> Parser<'a> { Some(Keyword::VIEW) => Some(GrantObjects::Views(objects?)), Some(Keyword::USER) => Some(GrantObjects::Users(objects?)), Some(Keyword::CONNECTION) => Some(GrantObjects::Connections(objects?)), + kw @ (Some(Keyword::PROCEDURE) | Some(Keyword::FUNCTION)) => { + if let Some(name) = objects?.first() { + self.parse_grant_procedure_or_function(name, &kw)? + } else { + self.expected("procedure or function name", self.peek_token())? + } + } Some(Keyword::TABLE) | None => Some(GrantObjects::Tables(objects?)), _ => unreachable!(), } @@ -14084,6 +14093,31 @@ impl<'a> Parser<'a> { Ok((privileges, objects)) } + fn parse_grant_procedure_or_function( + &mut self, + name: &ObjectName, + kw: &Option, + ) -> Result, ParserError> { + let arg_types = if self.consume_token(&Token::LParen) { + let list = self.parse_comma_separated0(Self::parse_data_type, Token::RParen)?; + self.expect_token(&Token::RParen)?; + list + } else { + vec![] + }; + match kw { + Some(Keyword::PROCEDURE) => Ok(Some(GrantObjects::Procedure { + name: name.clone(), + arg_types, + })), + Some(Keyword::FUNCTION) => Ok(Some(GrantObjects::Function { + name: name.clone(), + arg_types, + })), + _ => self.expected("procedure or function keywords", self.peek_token())?, + } + } + pub fn parse_grant_permission(&mut self) -> Result { fn parse_columns(parser: &mut Parser) -> Result>, ParserError> { let columns = parser.parse_parenthesized_column_list(Optional, false)?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 1de1d93fc..a00405aaf 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9460,6 +9460,8 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON FUTURE VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT SELECT ON FUTURE MATERIALIZED VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT SELECT ON FUTURE SEQUENCES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT USAGE ON PROCEDURE db1.sc1.foo(INT) TO ROLE role1"); + verified_stmt("GRANT USAGE ON FUNCTION db1.sc1.foo(INT) TO ROLE role1"); } #[test] From fd4934ec74ab23cab6ef89750dc161bdeceff25e Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 10 Jul 2025 18:01:08 +0200 Subject: [PATCH 191/286] Add support for `+` char in Snowflake stage names (#1935) --- src/dialect/snowflake.rs | 1 + tests/sqlparser_snowflake.rs | 20 ++++++++++++++++++++ 2 files changed, 21 insertions(+) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index e005300ef..21bc95930 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -735,6 +735,7 @@ pub fn parse_stage_name_identifier(parser: &mut Parser) -> Result ident.push('~'), Token::Mod => ident.push('%'), Token::Div => ident.push('/'), + Token::Plus => ident.push('+'), Token::Word(w) => ident.push_str(&w.to_string()), _ => return parser.expected("stage name identifier", parser.peek_token()), } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 6081e3183..389f47034 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -2581,6 +2581,26 @@ fn test_snowflake_copy_into() { } _ => unreachable!(), } + + // Test for non-ident characters in stage names + let sql = "COPY INTO a.b FROM @namespace.stage_name/x@x~x%x+"; + assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); + match snowflake().verified_stmt(sql) { + Statement::CopyIntoSnowflake { into, from_obj, .. } => { + assert_eq!( + into, + ObjectName::from(vec![Ident::new("a"), Ident::new("b")]) + ); + assert_eq!( + from_obj, + Some(ObjectName::from(vec![ + Ident::new("@namespace"), + Ident::new("stage_name/x@x~x%x+") + ])) + ) + } + _ => unreachable!(), + } } #[test] From 15f35e14765e58724e37704818a8c16426f2f145 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 11 Jul 2025 11:13:20 +0200 Subject: [PATCH 192/286] Snowflake Reserved SQL Keywords as Implicit Table Alias (#1934) --- src/dialect/mod.rs | 10 +++- src/dialect/snowflake.rs | 88 ++++++++++++++++++++++++++++++++++-- tests/sqlparser_common.rs | 26 ++++++++--- tests/sqlparser_snowflake.rs | 51 +++++++++++++++++++++ 4 files changed, 163 insertions(+), 12 deletions(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 8f9dd617f..861dfe265 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -992,11 +992,17 @@ pub trait Dialect: Debug + Any { explicit || self.is_column_alias(kw, parser) } + /// Returns true if the specified keyword should be parsed as a table identifier. + /// See [keywords::RESERVED_FOR_TABLE_ALIAS] + fn is_table_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { + !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) + } + /// Returns true if the specified keyword should be parsed as a table factor alias. /// When explicit is true, the keyword is preceded by an `AS` word. Parser is provided /// to enable looking ahead if needed. - fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { - explicit || !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) + fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { + explicit || self.is_table_alias(kw, parser) } /// Returns true if this dialect supports querying historical table data diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 21bc95930..85c09acb3 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -318,9 +318,11 @@ impl Dialect for SnowflakeDialect { } // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` - // which would give it a different meanings, for example: `SELECT 1 FETCH FIRST 10 ROWS` - not an alias - Keyword::FETCH - if parser.peek_keyword(Keyword::FIRST) || parser.peek_keyword(Keyword::NEXT) => + // which would give it a different meanings, for example: + // `SELECT 1 FETCH FIRST 10 ROWS` - not an alias + // `SELECT 1 FETCH 10` - not an alias + Keyword::FETCH if parser.peek_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]).is_some() + || matches!(parser.peek_token().token, Token::Number(_, _)) => { false } @@ -345,6 +347,86 @@ impl Dialect for SnowflakeDialect { } } + fn is_table_alias(&self, kw: &Keyword, parser: &mut Parser) -> bool { + match kw { + // The following keywords can be considered an alias as long as + // they are not followed by other tokens that may change their meaning + Keyword::LIMIT + | Keyword::RETURNING + | Keyword::INNER + | Keyword::USING + | Keyword::PIVOT + | Keyword::UNPIVOT + | Keyword::EXCEPT + | Keyword::MATCH_RECOGNIZE + | Keyword::OFFSET + if !matches!(parser.peek_token_ref().token, Token::SemiColon | Token::EOF) => + { + false + } + + // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` + // which would give it a different meanings, for example: + // `SELECT * FROM tbl FETCH FIRST 10 ROWS` - not an alias + // `SELECT * FROM tbl FETCH 10` - not an alias + Keyword::FETCH + if parser + .peek_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]) + .is_some() + || matches!(parser.peek_token().token, Token::Number(_, _)) => + { + false + } + + // All sorts of join-related keywords can be considered aliases unless additional + // keywords change their meaning. + Keyword::RIGHT | Keyword::LEFT | Keyword::SEMI | Keyword::ANTI + if parser + .peek_one_of_keywords(&[Keyword::JOIN, Keyword::OUTER]) + .is_some() => + { + false + } + Keyword::GLOBAL if parser.peek_keyword(Keyword::FULL) => false, + + // Reserved keywords by the Snowflake dialect, which seem to be less strictive + // than what is listed in `keywords::RESERVED_FOR_TABLE_ALIAS`. The following + // keywords were tested with the this statement: `SELECT .* FROM tbl `. + Keyword::WITH + | Keyword::ORDER + | Keyword::SELECT + | Keyword::WHERE + | Keyword::GROUP + | Keyword::HAVING + | Keyword::LATERAL + | Keyword::UNION + | Keyword::INTERSECT + | Keyword::MINUS + | Keyword::ON + | Keyword::JOIN + | Keyword::INNER + | Keyword::CROSS + | Keyword::FULL + | Keyword::LEFT + | Keyword::RIGHT + | Keyword::NATURAL + | Keyword::USING + | Keyword::ASOF + | Keyword::MATCH_CONDITION + | Keyword::SET + | Keyword::QUALIFY + | Keyword::FOR + | Keyword::START + | Keyword::CONNECT + | Keyword::SAMPLE + | Keyword::TABLESAMPLE + | Keyword::FROM => false, + + // Any other word is considered an alias + _ => true, + } + } + /// See: fn supports_timestamp_versioning(&self) -> bool { true diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a00405aaf..906be9bef 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -5548,7 +5548,8 @@ fn parse_named_window_functions() { WINDOW w AS (PARTITION BY x), win AS (ORDER BY y)"; supported_dialects.verified_stmt(sql); - let select = verified_only_select(sql); + let select = all_dialects_except(|d| d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d))) + .verified_only_select(sql); const EXPECTED_PROJ_QTY: usize = 2; assert_eq!(EXPECTED_PROJ_QTY, select.projection.len()); @@ -5578,6 +5579,7 @@ fn parse_named_window_functions() { #[test] fn parse_window_clause() { + let dialects = all_dialects_except(|d| d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d))); let sql = "SELECT * \ FROM mytable \ WINDOW \ @@ -5590,10 +5592,14 @@ fn parse_window_clause() { window7 AS (window1 ROWS UNBOUNDED PRECEDING), \ window8 AS (window1 PARTITION BY a ORDER BY b ROWS UNBOUNDED PRECEDING) \ ORDER BY C3"; - verified_only_select(sql); + dialects.verified_only_select(sql); let sql = "SELECT * from mytable WINDOW window1 AS window2"; - let dialects = all_dialects_except(|d| d.is::() || d.is::()); + let dialects = all_dialects_except(|d| { + d.is::() + || d.is::() + || d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d)) + }); let res = dialects.parse_sql_statements(sql); assert_eq!( ParserError::ParserError("Expected: (, found: window2".to_string()), @@ -5603,6 +5609,7 @@ fn parse_window_clause() { #[test] fn test_parse_named_window() { + let dialects = all_dialects_except(|d| d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d))); let sql = "SELECT \ MIN(c12) OVER window1 AS min1, \ MAX(c12) OVER window2 AS max1 \ @@ -5610,7 +5617,7 @@ fn test_parse_named_window() { WINDOW window1 AS (ORDER BY C12), \ window2 AS (PARTITION BY C11) \ ORDER BY C3"; - let actual_select_only = verified_only_select(sql); + let actual_select_only = dialects.verified_only_select(sql); let expected = Select { select_token: AttachedToken::empty(), distinct: None, @@ -5759,6 +5766,10 @@ fn test_parse_named_window() { #[test] fn parse_window_and_qualify_clause() { + let dialects = all_dialects_except(|d| { + d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d)) + || d.is_table_alias(&Keyword::QUALIFY, &mut Parser::new(d)) + }); let sql = "SELECT \ MIN(c12) OVER window1 AS min1 \ FROM aggregate_test_100 \ @@ -5766,7 +5777,7 @@ fn parse_window_and_qualify_clause() { WINDOW window1 AS (ORDER BY C12), \ window2 AS (PARTITION BY C11) \ ORDER BY C3"; - verified_only_select(sql); + dialects.verified_only_select(sql); let sql = "SELECT \ MIN(c12) OVER window1 AS min1 \ @@ -5775,7 +5786,7 @@ fn parse_window_and_qualify_clause() { window2 AS (PARTITION BY C11) \ QUALIFY ROW_NUMBER() OVER my_window \ ORDER BY C3"; - verified_only_select(sql); + dialects.verified_only_select(sql); } #[test] @@ -7443,7 +7454,8 @@ fn parse_join_syntax_variants() { "SELECT c1 FROM t1 FULL JOIN t2 USING(c1)", ); - let res = parse_sql_statements("SELECT * FROM a OUTER JOIN b ON 1"); + let dialects = all_dialects_except(|d| d.is_table_alias(&Keyword::OUTER, &mut Parser::new(d))); + let res = dialects.parse_sql_statements("SELECT * FROM a OUTER JOIN b ON 1"); assert_eq!( ParserError::ParserError("Expected: APPLY, found: JOIN".to_string()), res.unwrap_err() diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 389f47034..65546bee0 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3492,6 +3492,57 @@ fn test_sql_keywords_as_select_item_aliases() { } } +#[test] +fn test_sql_keywords_as_table_aliases() { + // Some keywords that should be parsed as an alias implicitly + let unreserved_kws = vec![ + "VIEW", + "EXPLAIN", + "ANALYZE", + "SORT", + "PIVOT", + "UNPIVOT", + "TOP", + "LIMIT", + "OFFSET", + "FETCH", + "EXCEPT", + "CLUSTER", + "DISTRIBUTE", + "GLOBAL", + "ANTI", + "SEMI", + "RETURNING", + "OUTER", + "WINDOW", + "END", + "PARTITION", + "PREWHERE", + "SETTINGS", + "FORMAT", + "MATCH_RECOGNIZE", + "OPEN", + ]; + + for kw in unreserved_kws { + snowflake().verified_stmt(&format!("SELECT * FROM tbl AS {kw}")); + snowflake().one_statement_parses_to( + &format!("SELECT * FROM tbl {kw}"), + &format!("SELECT * FROM tbl AS {kw}"), + ); + } + + // Some keywords that should not be parsed as an alias implicitly + let reserved_kws = vec![ + "FROM", "GROUP", "HAVING", "ORDER", "SELECT", "UNION", "WHERE", "WITH", + ]; + for kw in reserved_kws { + assert!(snowflake() + .parse_sql_statements(&format!("SELECT * FROM tbl {kw}")) + .is_err()); + } +} + #[test] fn test_timetravel_at_before() { snowflake().verified_only_select("SELECT * FROM tbl AT(TIMESTAMP => '2024-12-15 00:00:00')"); From ee31b64f9e9c0959aa6fcada159d8092517966e3 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 11 Jul 2025 11:39:29 +0200 Subject: [PATCH 193/286] Add support for Redshift `SELECT * EXCLUDE` (#1936) --- src/ast/query.rs | 9 +++ src/ast/spans.rs | 1 + src/dialect/duckdb.rs | 4 ++ src/dialect/generic.rs | 4 ++ src/dialect/mod.rs | 20 ++++++ src/dialect/redshift.rs | 8 +++ src/dialect/snowflake.rs | 4 ++ src/keywords.rs | 1 + src/parser/mod.rs | 11 +++- tests/sqlparser_clickhouse.rs | 1 + tests/sqlparser_common.rs | 115 ++++++++++++++++++++++++++++++++++ tests/sqlparser_duckdb.rs | 2 + tests/sqlparser_mssql.rs | 3 + tests/sqlparser_mysql.rs | 9 ++- tests/sqlparser_postgres.rs | 3 + 15 files changed, 192 insertions(+), 3 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index febf1fc6d..7ffb64d9b 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -321,6 +321,11 @@ pub struct Select { pub top_before_distinct: bool, /// projection expressions pub projection: Vec, + /// Excluded columns from the projection expression which are not specified + /// directly after a wildcard. + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_EXCLUDE_list.html) + pub exclude: Option, /// INTO pub into: Option, /// FROM @@ -401,6 +406,10 @@ impl fmt::Display for Select { indented_list(f, &self.projection)?; } + if let Some(exclude) = &self.exclude { + write!(f, " {exclude}")?; + } + if let Some(ref into) = self.into { f.write_str(" ")?; into.fmt(f)?; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index a1b2e4e0d..3e82905e1 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2220,6 +2220,7 @@ impl Spanned for Select { distinct: _, // todo top: _, // todo, mysql specific projection, + exclude: _, into, from, lateral_views, diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index 3366c6705..fa18463ae 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -94,4 +94,8 @@ impl Dialect for DuckDbDialect { fn supports_order_by_all(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } } diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 5e9f2e4e2..be2cc0076 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -179,4 +179,8 @@ impl Dialect for GenericDialect { fn supports_filter_during_aggregation(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 861dfe265..deb5719d5 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -570,6 +570,26 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports an exclude option + /// following a wildcard in the projection section. For example: + /// `SELECT * EXCLUDE col1 FROM tbl`. + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_EXCLUDE_list.html) + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/select) + fn supports_select_wildcard_exclude(&self) -> bool { + false + } + + /// Returns true if the dialect supports an exclude option + /// as the last item in the projection section, not necessarily + /// after a wildcard. For example: + /// `SELECT *, c1, c2 EXCLUDE c3 FROM tbl` + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_EXCLUDE_list.html) + fn supports_select_exclude(&self) -> bool { + false + } + /// Dialect-specific infix parser override /// /// This method is called to parse the next infix expression. diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index 9ad9c5fdb..8ffed98af 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -131,4 +131,12 @@ impl Dialect for RedshiftSqlDialect { fn supports_string_literal_backslash_escape(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } + + fn supports_select_exclude(&self) -> bool { + true + } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 85c09acb3..3b1eff39a 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -466,6 +466,10 @@ impl Dialect for SnowflakeDialect { fn supports_select_expr_star(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { diff --git a/src/keywords.rs b/src/keywords.rs index 738651504..9e689a6df 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -1119,6 +1119,7 @@ pub const RESERVED_FOR_COLUMN_ALIAS: &[Keyword] = &[ Keyword::FETCH, Keyword::UNION, Keyword::EXCEPT, + Keyword::EXCLUDE, Keyword::INTERSECT, Keyword::MINUS, Keyword::CLUSTER, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b00cd16d7..a2d01f136 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11740,6 +11740,7 @@ impl<'a> Parser<'a> { top: None, top_before_distinct: false, projection: vec![], + exclude: None, into: None, from, lateral_views: vec![], @@ -11782,6 +11783,12 @@ impl<'a> Parser<'a> { self.parse_projection()? }; + let exclude = if self.dialect.supports_select_exclude() { + self.parse_optional_select_item_exclude()? + } else { + None + }; + let into = if self.parse_keyword(Keyword::INTO) { Some(self.parse_select_into()?) } else { @@ -11915,6 +11922,7 @@ impl<'a> Parser<'a> { top, top_before_distinct, projection, + exclude, into, from, lateral_views, @@ -15052,8 +15060,7 @@ impl<'a> Parser<'a> { } else { None }; - let opt_exclude = if opt_ilike.is_none() - && dialect_of!(self is GenericDialect | DuckDbDialect | SnowflakeDialect) + let opt_exclude = if opt_ilike.is_none() && self.dialect.supports_select_wildcard_exclude() { self.parse_optional_select_item_exclude()? } else { diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 1d8669a22..9e5b6ce8a 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -60,6 +60,7 @@ fn parse_map_access_expr() { ), })], })], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("foos")])), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 906be9bef..b7b5b630b 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -459,6 +459,7 @@ fn parse_update_set_from() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("name"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("id"))), ], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("t1")])), @@ -5695,6 +5696,7 @@ fn test_parse_named_window() { }, }, ], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -6351,6 +6353,7 @@ fn parse_interval_and_or_xor() { quote_style: None, span: Span::empty(), }))], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -8620,6 +8623,7 @@ fn lateral_function() { distinct: None, top: None, projection: vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -9616,6 +9620,7 @@ fn parse_merge() { projection: vec![SelectItem::Wildcard( WildcardAdditionalOptions::default() )], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![ @@ -11534,6 +11539,7 @@ fn parse_unload() { top: None, top_before_distinct: false, projection: vec![UnnamedExpr(Expr::Identifier(Ident::new("cola"))),], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("tab")])), @@ -11734,6 +11740,7 @@ fn parse_connect_by() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("manager_id"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("title"))), ], + exclude: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("employees")])), joins: vec![], @@ -11815,6 +11822,7 @@ fn parse_connect_by() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("manager_id"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("title"))), ], + exclude: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("employees")])), joins: vec![], @@ -12748,6 +12756,7 @@ fn test_extract_seconds_ok() { format: None, }), })], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -14820,6 +14829,7 @@ fn test_select_from_first() { distinct: None, top: None, projection, + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -16000,3 +16010,108 @@ fn parse_create_procedure_with_parameter_modes() { _ => unreachable!(), } } + +#[test] +fn test_select_exclude() { + let dialects = all_dialects_where(|d| d.supports_select_wildcard_exclude()); + match &dialects + .verified_only_select("SELECT * EXCLUDE c1 FROM test") + .projection[0] + { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + } + _ => unreachable!(), + } + match &dialects + .verified_only_select("SELECT * EXCLUDE (c1, c2) FROM test") + .projection[0] + { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Multiple(vec![ + Ident::new("c1"), + Ident::new("c2") + ])) + ); + } + _ => unreachable!(), + } + let select = dialects.verified_only_select("SELECT * EXCLUDE c1, c2 FROM test"); + match &select.projection[0] { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + } + _ => unreachable!(), + } + match &select.projection[1] { + SelectItem::UnnamedExpr(Expr::Identifier(ident)) => { + assert_eq!(*ident, Ident::new("c2")); + } + _ => unreachable!(), + } + + let dialects = all_dialects_where(|d| d.supports_select_exclude()); + let select = dialects.verified_only_select("SELECT *, c1 EXCLUDE c1 FROM test"); + match &select.projection[0] { + SelectItem::Wildcard(additional_options) => { + assert_eq!(*additional_options, WildcardAdditionalOptions::default()); + } + _ => unreachable!(), + } + assert_eq!( + select.exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + + let dialects = all_dialects_where(|d| { + d.supports_select_wildcard_exclude() && !d.supports_select_exclude() + }); + let select = dialects.verified_only_select("SELECT * EXCLUDE c1 FROM test"); + match &select.projection[0] { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + } + _ => unreachable!(), + } + + // Dialects that only support the wildcard form and do not accept EXCLUDE as an implicity alias + // will fail when encountered with the `c2` ident + let dialects = all_dialects_where(|d| { + d.supports_select_wildcard_exclude() + && !d.supports_select_exclude() + && d.is_column_alias(&Keyword::EXCLUDE, &mut Parser::new(d)) + }); + assert_eq!( + dialects + .parse_sql_statements("SELECT *, c1 EXCLUDE c2 FROM test") + .err() + .unwrap(), + ParserError::ParserError("Expected: end of statement, found: c2".to_string()) + ); + + // Dialects that only support the wildcard form and accept EXCLUDE as an implicity alias + // will fail when encountered with the `EXCLUDE` keyword + let dialects = all_dialects_where(|d| { + d.supports_select_wildcard_exclude() + && !d.supports_select_exclude() + && !d.is_column_alias(&Keyword::EXCLUDE, &mut Parser::new(d)) + }); + assert_eq!( + dialects + .parse_sql_statements("SELECT *, c1 EXCLUDE c2 FROM test") + .err() + .unwrap(), + ParserError::ParserError("Expected: end of statement, found: EXCLUDE".to_string()) + ); +} diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 371d3aace..fe14b7ba5 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -269,6 +269,7 @@ fn test_select_union_by_name() { distinct: None, top: None, projection: vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -299,6 +300,7 @@ fn test_select_union_by_name() { distinct: None, top: None, projection: vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index ebbec25fb..b4a650cee 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -126,6 +126,7 @@ fn parse_create_procedure() { projection: vec![SelectItem::UnnamedExpr(Expr::Value( (number("1")).with_empty_span() ))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1368,6 +1369,7 @@ fn parse_substring_in_select() { special: true, shorthand: false, })], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -1516,6 +1518,7 @@ fn parse_mssql_declare() { (Value::Number("4".parse().unwrap(), false)).with_empty_span() )), })], + exclude: None, into: None, from: vec![], lateral_views: vec![], diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 44f365844..e8c6719e0 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1403,6 +1403,7 @@ fn parse_escaped_quote_identifiers_with_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1456,6 +1457,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1503,6 +1505,7 @@ fn parse_escaped_backticks_with_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1554,6 +1557,7 @@ fn parse_escaped_backticks_with_no_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -2225,6 +2229,7 @@ fn parse_select_with_numeric_prefix_column_name() { projection: vec![SelectItem::UnnamedExpr(Expr::Identifier(Ident::new( "123col_$@123abc" )))], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::with_quote( @@ -2392,7 +2397,6 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { q.body, Box::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), - distinct: None, top: None, top_before_distinct: false, @@ -2400,6 +2404,7 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { SelectItem::UnnamedExpr(Expr::value(number("123e4"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("123col_$@123abc"))) ], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::with_quote( @@ -3043,6 +3048,7 @@ fn parse_substring_in_select() { special: true, shorthand: false, })], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -3357,6 +3363,7 @@ fn parse_hex_string_introducer() { ) .into(), })], + exclude: None, from: vec![], lateral_views: vec![], prewhere: None, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 48792025a..0d1d138cd 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1305,6 +1305,7 @@ fn parse_copy_to() { }, } ], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -2948,6 +2949,7 @@ fn parse_array_subquery_expr() { projection: vec![SelectItem::UnnamedExpr(Expr::Value( (number("1")).with_empty_span() ))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -2973,6 +2975,7 @@ fn parse_array_subquery_expr() { projection: vec![SelectItem::UnnamedExpr(Expr::Value( (number("2")).with_empty_span() ))], + exclude: None, into: None, from: vec![], lateral_views: vec![], From bc2c4e263d85a2440c87ea56f0c595a458e40fc8 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 11 Jul 2025 14:46:48 +0200 Subject: [PATCH 194/286] Support optional semicolon between statements (#1937) --- src/parser/mod.rs | 8 ++++++++ src/test_utils.rs | 5 +++++ tests/sqlparser_common.rs | 22 ++++++++++++++++++++-- tests/sqlparser_mssql.rs | 26 +++++++++++++++++++++++++- tests/sqlparser_mysql.rs | 1 + 5 files changed, 59 insertions(+), 3 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a2d01f136..47b63da87 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -222,6 +222,9 @@ pub struct ParserOptions { /// Controls how literal values are unescaped. See /// [`Tokenizer::with_unescape`] for more details. pub unescape: bool, + /// Controls if the parser expects a semi-colon token + /// between statements. Default is `true`. + pub require_semicolon_stmt_delimiter: bool, } impl Default for ParserOptions { @@ -229,6 +232,7 @@ impl Default for ParserOptions { Self { trailing_commas: false, unescape: true, + require_semicolon_stmt_delimiter: true, } } } @@ -467,6 +471,10 @@ impl<'a> Parser<'a> { expecting_statement_delimiter = false; } + if !self.options.require_semicolon_stmt_delimiter { + expecting_statement_delimiter = false; + } + match self.peek_token().token { Token::EOF => break, diff --git a/src/test_utils.rs b/src/test_utils.rs index 544ceaef8..654f2723e 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -294,6 +294,11 @@ pub fn all_dialects() -> TestedDialects { ]) } +// Returns all available dialects with the specified parser options +pub fn all_dialects_with_options(options: ParserOptions) -> TestedDialects { + TestedDialects::new_with_options(all_dialects().dialects, options) +} + /// Returns all dialects matching the given predicate. pub fn all_dialects_where(predicate: F) -> TestedDialects where diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index b7b5b630b..15144479c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -40,8 +40,9 @@ use sqlparser::parser::{Parser, ParserError, ParserOptions}; use sqlparser::tokenizer::Tokenizer; use sqlparser::tokenizer::{Location, Span}; use test_utils::{ - all_dialects, all_dialects_where, alter_table_op, assert_eq_vec, call, expr_from_projection, - join, number, only, table, table_alias, table_from_name, TestedDialects, + all_dialects, all_dialects_where, all_dialects_with_options, alter_table_op, assert_eq_vec, + call, expr_from_projection, join, number, only, table, table_alias, table_from_name, + TestedDialects, }; #[macro_use] @@ -16115,3 +16116,20 @@ fn test_select_exclude() { ParserError::ParserError("Expected: end of statement, found: EXCLUDE".to_string()) ); } + +#[test] +fn test_no_semicolon_required_between_statements() { + let sql = r#" +SELECT * FROM tbl1 +SELECT * FROM tbl2 + "#; + + let dialects = all_dialects_with_options(ParserOptions { + trailing_commas: false, + unescape: true, + require_semicolon_stmt_delimiter: false, + }); + let stmts = dialects.parse_sql_statements(sql).unwrap(); + assert_eq!(stmts.len(), 2); + assert!(stmts.iter().all(|s| matches!(s, Statement::Query { .. }))); +} diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index b4a650cee..50c6448d9 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -32,7 +32,7 @@ use sqlparser::ast::DeclareAssignment::MsSqlAssignment; use sqlparser::ast::Value::SingleQuotedString; use sqlparser::ast::*; use sqlparser::dialect::{GenericDialect, MsSqlDialect}; -use sqlparser::parser::{Parser, ParserError}; +use sqlparser::parser::{Parser, ParserError, ParserOptions}; #[test] fn parse_mssql_identifiers() { @@ -2327,6 +2327,18 @@ fn ms() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {})]) } +// MS SQL dialect with support for optional semi-colon statement delimiters +fn tsql() -> TestedDialects { + TestedDialects::new_with_options( + vec![Box::new(MsSqlDialect {})], + ParserOptions { + trailing_commas: false, + unescape: true, + require_semicolon_stmt_delimiter: false, + }, + ) +} + fn ms_and_generic() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {}), Box::new(GenericDialect {})]) } @@ -2483,3 +2495,15 @@ fn parse_mssql_grant() { fn parse_mssql_deny() { ms().verified_stmt("DENY SELECT ON my_table TO public, db_admin"); } + +#[test] +fn test_tsql_no_semicolon_delimiter() { + let sql = r#" +DECLARE @X AS NVARCHAR(MAX)='x' +DECLARE @Y AS NVARCHAR(MAX)='y' + "#; + + let stmts = tsql().parse_sql_statements(sql).unwrap(); + assert_eq!(stmts.len(), 2); + assert!(stmts.iter().all(|s| matches!(s, Statement::Declare { .. }))); +} diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index e8c6719e0..9068ed9c5 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1442,6 +1442,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { ParserOptions { trailing_commas: false, unescape: false, + require_semicolon_stmt_delimiter: true, } ) .verified_stmt(sql), From 750a7aa0549a51f4b7fb03feae80ca07d3808e75 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 14 Jul 2025 10:16:20 +0200 Subject: [PATCH 195/286] Snowflake: support trailing options in `CREATE TABLE` (#1931) --- src/ast/helpers/stmt_create_table.rs | 20 +++++++++++++ src/dialect/bigquery.rs | 4 +++ src/dialect/mod.rs | 7 +++++ src/dialect/snowflake.rs | 7 ++--- tests/sqlparser_common.rs | 17 ++--------- tests/sqlparser_snowflake.rs | 45 ++++++++++++++++++++++++++++ 6 files changed, 80 insertions(+), 20 deletions(-) diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index d66a869bf..60b8fb2a0 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -383,6 +383,26 @@ impl CreateTableBuilder { self } + /// Returns true if the statement has exactly one source of info on the schema of the new table. + /// This is Snowflake-specific, some dialects allow more than one source. + pub(crate) fn validate_schema_info(&self) -> bool { + let mut sources = 0; + if !self.columns.is_empty() { + sources += 1; + } + if self.query.is_some() { + sources += 1; + } + if self.like.is_some() { + sources += 1; + } + if self.clone.is_some() { + sources += 1; + } + + sources == 1 + } + pub fn build(self) -> Statement { Statement::CreateTable(CreateTable { or_replace: self.or_replace, diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index c2cd507ce..d53c9db05 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -144,4 +144,8 @@ impl Dialect for BigQueryDialect { fn supports_pipe_operator(&self) -> bool { true } + + fn supports_create_table_multi_schema_info_sources(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index deb5719d5..c79b279df 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -590,6 +590,13 @@ pub trait Dialect: Debug + Any { false } + /// Returne true if the dialect supports specifying multiple options + /// in a `CREATE TABLE` statement for the structure of the new table. For example: + /// `CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a` + fn supports_create_table_multi_schema_info_sources(&self) -> bool { + false + } + /// Dialect-specific infix parser override /// /// This method is called to parse the next infix expression. diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 3b1eff39a..fcf94ee75 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -555,17 +555,14 @@ pub fn parse_create_table( Keyword::AS => { let query = parser.parse_query()?; builder = builder.query(Some(query)); - break; } Keyword::CLONE => { let clone = parser.parse_object_name(false).ok(); builder = builder.clone_clause(clone); - break; } Keyword::LIKE => { let like = parser.parse_object_name(false).ok(); builder = builder.like(like); - break; } Keyword::CLUSTER => { parser.expect_keyword_is(Keyword::BY)?; @@ -691,7 +688,7 @@ pub fn parse_create_table( builder = builder.columns(columns).constraints(constraints); } Token::EOF => { - if builder.columns.is_empty() { + if !builder.validate_schema_info() { return Err(ParserError::ParserError( "unexpected end of input".to_string(), )); @@ -700,7 +697,7 @@ pub fn parse_create_table( break; } Token::SemiColon => { - if builder.columns.is_empty() { + if !builder.validate_schema_info() { return Err(ParserError::ParserError( "unexpected end of input".to_string(), )); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 15144479c..fd573c600 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4347,8 +4347,9 @@ fn parse_create_table_as() { // BigQuery allows specifying table schema in CTAS // ANSI SQL and PostgreSQL let you only specify the list of columns // (without data types) in a CTAS, but we have yet to support that. + let dialects = all_dialects_where(|d| d.supports_create_table_multi_schema_info_sources()); let sql = "CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a"; - match verified_stmt(sql) { + match dialects.verified_stmt(sql) { Statement::CreateTable(CreateTable { columns, query, .. }) => { assert_eq!(columns.len(), 2); assert_eq!(columns[0].to_string(), "a INT".to_string()); @@ -4453,20 +4454,6 @@ fn parse_create_or_replace_table() { } _ => unreachable!(), } - - let sql = "CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a"; - match verified_stmt(sql) { - Statement::CreateTable(CreateTable { columns, query, .. }) => { - assert_eq!(columns.len(), 2); - assert_eq!(columns[0].to_string(), "a INT".to_string()); - assert_eq!(columns[1].to_string(), "b INT".to_string()); - assert_eq!( - query, - Some(Box::new(verified_query("SELECT 1 AS b, 2 AS a"))) - ); - } - _ => unreachable!(), - } } #[test] diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 65546bee0..562ddfea7 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -995,6 +995,51 @@ fn test_snowflake_create_iceberg_table_without_location() { ); } +#[test] +fn test_snowflake_create_table_trailing_options() { + // Serialization to SQL assume that in `CREATE TABLE AS` the options come before the `AS ()` + // but Snowflake supports also the other way around + snowflake() + .verified_stmt("CREATE TEMPORARY TABLE dst ON COMMIT PRESERVE ROWS AS (SELECT * FROM src)"); + snowflake() + .parse_sql_statements( + "CREATE TEMPORARY TABLE dst AS (SELECT * FROM src) ON COMMIT PRESERVE ROWS", + ) + .unwrap(); + + // Same for `CREATE TABLE LIKE|CLONE`: + snowflake().verified_stmt("CREATE TEMPORARY TABLE dst LIKE src ON COMMIT PRESERVE ROWS"); + snowflake() + .parse_sql_statements("CREATE TEMPORARY TABLE dst ON COMMIT PRESERVE ROWS LIKE src") + .unwrap(); + + snowflake().verified_stmt("CREATE TEMPORARY TABLE dst CLONE src ON COMMIT PRESERVE ROWS"); + snowflake() + .parse_sql_statements("CREATE TEMPORARY TABLE dst ON COMMIT PRESERVE ROWS CLONE src") + .unwrap(); +} + +#[test] +fn test_snowflake_create_table_valid_schema_info() { + // Validate there's exactly one source of information on the schema of the new table + assert_eq!( + snowflake() + .parse_sql_statements("CREATE TABLE dst") + .is_err(), + true + ); + assert_eq!( + snowflake().parse_sql_statements("CREATE OR REPLACE TEMP TABLE dst LIKE src AS (SELECT * FROM CUSTOMERS) ON COMMIT PRESERVE ROWS").is_err(), + true + ); + assert_eq!( + snowflake() + .parse_sql_statements("CREATE OR REPLACE TEMP TABLE dst CLONE customers LIKE customer2") + .is_err(), + true + ); +} + #[test] fn parse_sf_create_or_replace_view_with_comment_missing_equal() { assert!(snowflake_and_generic() From 9b9ffe450c85cbc8c0fcb0c75eafbcdad603030f Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 14 Jul 2025 10:19:28 +0200 Subject: [PATCH 196/286] MSSQL: Add support for EXEC output and default keywords (#1940) --- src/ast/mod.rs | 14 ++++++++++++++ src/parser/mod.rs | 9 ++++++++- tests/sqlparser_common.rs | 16 ++++++++++++++++ tests/sqlparser_postgres.rs | 12 +++++++++--- 4 files changed, 47 insertions(+), 4 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 75e88f8a8..8bf750765 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4059,6 +4059,12 @@ pub enum Statement { immediate: bool, into: Vec, using: Vec, + /// Whether the last parameter is the return value of the procedure + /// MSSQL: + output: bool, + /// Whether to invoke the procedure with the default parameter values + /// MSSQL: + default: bool, }, /// ```sql /// PREPARE name [ ( data_type [, ...] ) ] AS statement @@ -5815,6 +5821,8 @@ impl fmt::Display for Statement { immediate, into, using, + output, + default, } => { let (open, close) = if *has_parentheses { ("(", ")") @@ -5835,6 +5843,12 @@ impl fmt::Display for Statement { if !using.is_empty() { write!(f, " USING {}", display_comma_separated(using))?; }; + if *output { + write!(f, " OUTPUT")?; + } + if *default { + write!(f, " DEFAULT")?; + } Ok(()) } Statement::Prepare { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 47b63da87..b3ceec7e2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15734,10 +15734,11 @@ impl<'a> Parser<'a> { let has_parentheses = self.consume_token(&Token::LParen); + let end_kws = &[Keyword::USING, Keyword::OUTPUT, Keyword::DEFAULT]; let end_token = match (has_parentheses, self.peek_token().token) { (true, _) => Token::RParen, (false, Token::EOF) => Token::EOF, - (false, Token::Word(w)) if w.keyword == Keyword::USING => Token::Word(w), + (false, Token::Word(w)) if end_kws.contains(&w.keyword) => Token::Word(w), (false, _) => Token::SemiColon, }; @@ -15759,6 +15760,10 @@ impl<'a> Parser<'a> { vec![] }; + let output = self.parse_keyword(Keyword::OUTPUT); + + let default = self.parse_keyword(Keyword::DEFAULT); + Ok(Statement::Execute { immediate: name.is_none(), name, @@ -15766,6 +15771,8 @@ impl<'a> Parser<'a> { has_parentheses, into, using, + output, + default, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index fd573c600..2d2008c84 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11393,6 +11393,8 @@ fn parse_execute_stored_procedure() { immediate: false, using: vec![], into: vec![], + output: false, + default: false, }; assert_eq!( // Microsoft SQL Server does not use parentheses around arguments for EXECUTE @@ -11407,6 +11409,18 @@ fn parse_execute_stored_procedure() { ), expected ); + match ms_and_generic().verified_stmt("EXECUTE dbo.proc1 @ReturnVal = @X OUTPUT") { + Statement::Execute { output, .. } => { + assert!(output); + } + _ => unreachable!(), + } + match ms_and_generic().verified_stmt("EXECUTE dbo.proc1 DEFAULT") { + Statement::Execute { default, .. } => { + assert!(default); + } + _ => unreachable!(), + } } #[test] @@ -11425,6 +11439,8 @@ fn parse_execute_immediate() { into: vec![Ident::new("a")], name: None, has_parentheses: false, + output: false, + default: false, }; let stmt = dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1' INTO a USING 1 AS b"); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 0d1d138cd..461277d81 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -1666,7 +1666,9 @@ fn parse_execute() { has_parentheses: false, using: vec![], immediate: false, - into: vec![] + into: vec![], + output: false, + default: false, } ); @@ -1682,7 +1684,9 @@ fn parse_execute() { has_parentheses: true, using: vec![], immediate: false, - into: vec![] + into: vec![], + output: false, + default: false, } ); @@ -1719,7 +1723,9 @@ fn parse_execute() { }, ], immediate: false, - into: vec![] + into: vec![], + output: false, + default: false, } ); } From c5e6ba5e7d746161a59d6fd5e09846262a2fc01e Mon Sep 17 00:00:00 2001 From: etgarperets Date: Mon, 14 Jul 2025 11:24:13 +0300 Subject: [PATCH 197/286] Add identifier unicode support in Mysql, Postgres and Redshift (#1933) --- src/dialect/mysql.rs | 4 +++- src/dialect/postgresql.rs | 4 +++- src/dialect/redshift.rs | 4 ++-- tests/sqlparser_common.rs | 11 +++++++++++ 4 files changed, 19 insertions(+), 4 deletions(-) diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index f69e42436..b50c8df50 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -51,7 +51,9 @@ impl Dialect for MySqlDialect { } fn is_identifier_part(&self, ch: char) -> bool { - self.is_identifier_start(ch) || ch.is_ascii_digit() + self.is_identifier_start(ch) || ch.is_ascii_digit() || + // MySQL implements Unicode characters in identifiers. + !ch.is_ascii() } fn is_delimited_identifier_start(&self, ch: char) -> bool { diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index b2d4014cb..c1f025574 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -72,7 +72,9 @@ impl Dialect for PostgreSqlDialect { } fn is_identifier_part(&self, ch: char) -> bool { - ch.is_alphabetic() || ch.is_ascii_digit() || ch == '$' || ch == '_' + ch.is_alphabetic() || ch.is_ascii_digit() || ch == '$' || ch == '_' || + // PostgreSQL implements Unicode characters in identifiers. + !ch.is_ascii() } fn supports_unicode_string_literal(&self) -> bool { diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index 8ffed98af..c910e4c77 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -86,9 +86,9 @@ impl Dialect for RedshiftSqlDialect { } fn is_identifier_part(&self, ch: char) -> bool { - // Extends Postgres dialect with sharp and UTF-8 multibyte chars + // UTF-8 multibyte characters are supported in identifiers via the PostgreSqlDialect. // https://docs.aws.amazon.com/redshift/latest/dg/r_names.html - PostgreSqlDialect {}.is_identifier_part(ch) || ch == '#' || !ch.is_ascii() + PostgreSqlDialect {}.is_identifier_part(ch) || ch == '#' } /// redshift has `CONVERT(type, value)` instead of `CONVERT(value, type)` diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 2d2008c84..ba72399f9 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16136,3 +16136,14 @@ SELECT * FROM tbl2 assert_eq!(stmts.len(), 2); assert!(stmts.iter().all(|s| matches!(s, Statement::Query { .. }))); } + +#[test] +fn test_identifier_unicode_support() { + let sql = r#"SELECT phoneǤЖשचᎯ⻩☯♜🦄⚛🀄ᚠ⌛🌀 AS tbl FROM customers"#; + let dialects = TestedDialects::new(vec![ + Box::new(MySqlDialect {}), + Box::new(RedshiftSqlDialect {}), + Box::new(PostgreSqlDialect {}), + ]); + let _ = dialects.verified_stmt(sql); +} From ecd5d88638efc20e5be213d71e441d9c6d48798a Mon Sep 17 00:00:00 2001 From: etgarperets Date: Tue, 15 Jul 2025 10:26:11 +0300 Subject: [PATCH 198/286] Add identifier start unicode support for Postegres, MySql and Redshift (#1944) --- src/dialect/mysql.rs | 2 ++ src/dialect/postgresql.rs | 7 +++---- src/dialect/redshift.rs | 4 ++-- tests/sqlparser_common.rs | 13 +++++++++++-- 4 files changed, 18 insertions(+), 8 deletions(-) diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index b50c8df50..f7b5f574e 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -43,11 +43,13 @@ impl Dialect for MySqlDialect { // See https://dev.mysql.com/doc/refman/8.0/en/identifiers.html. // Identifiers which begin with a digit are recognized while tokenizing numbers, // so they can be distinguished from exponent numeric literals. + // MySQL also implements non ascii utf-8 charecters ch.is_alphabetic() || ch == '_' || ch == '$' || ch == '@' || ('\u{0080}'..='\u{ffff}').contains(&ch) + || !ch.is_ascii() } fn is_identifier_part(&self, ch: char) -> bool { diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index c1f025574..9cea252c8 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -65,10 +65,9 @@ impl Dialect for PostgreSqlDialect { } fn is_identifier_start(&self, ch: char) -> bool { - // See https://www.postgresql.org/docs/11/sql-syntax-lexical.html#SQL-SYNTAX-IDENTIFIERS - // We don't yet support identifiers beginning with "letters with - // diacritical marks" - ch.is_alphabetic() || ch == '_' + ch.is_alphabetic() || ch == '_' || + // PostgreSQL implements Unicode characters in identifiers. + !ch.is_ascii() } fn is_identifier_part(&self, ch: char) -> bool { diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index c910e4c77..68e025d18 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -80,9 +80,9 @@ impl Dialect for RedshiftSqlDialect { } fn is_identifier_start(&self, ch: char) -> bool { - // Extends Postgres dialect with sharp and UTF-8 multibyte chars + // UTF-8 multibyte characters are supported in identifiers via the PostgreSqlDialect. // https://docs.aws.amazon.com/redshift/latest/dg/r_names.html - PostgreSqlDialect {}.is_identifier_start(ch) || ch == '#' || !ch.is_ascii() + PostgreSqlDialect {}.is_identifier_start(ch) || ch == '#' } fn is_identifier_part(&self, ch: char) -> bool { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index ba72399f9..e95c7e7b6 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11151,9 +11151,7 @@ fn parse_non_latin_identifiers() { let supported_dialects = TestedDialects::new(vec![ Box::new(GenericDialect {}), Box::new(DuckDbDialect {}), - Box::new(PostgreSqlDialect {}), Box::new(MsSqlDialect {}), - Box::new(MySqlDialect {}), ]); assert!(supported_dialects .parse_sql_statements("SELECT 💝 FROM table1") @@ -16147,3 +16145,14 @@ fn test_identifier_unicode_support() { ]); let _ = dialects.verified_stmt(sql); } + +#[test] +fn test_identifier_unicode_start() { + let sql = r#"SELECT 💝phone AS 💝 FROM customers"#; + let dialects = TestedDialects::new(vec![ + Box::new(MySqlDialect {}), + Box::new(RedshiftSqlDialect {}), + Box::new(PostgreSqlDialect {}), + ]); + let _ = dialects.verified_stmt(sql); +} From 650681422a6f6a075cfa66cfde8a22e7539e96af Mon Sep 17 00:00:00 2001 From: Olexandr88 Date: Tue, 15 Jul 2025 10:57:44 +0300 Subject: [PATCH 199/286] docs: Update rust badge (#1943) --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 666be17c0..9dfe50810 100644 --- a/README.md +++ b/README.md @@ -21,7 +21,7 @@ [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0) [![Version](https://img.shields.io/crates/v/sqlparser.svg)](https://crates.io/crates/sqlparser) -[![Build Status](https://github.com/sqlparser-rs/sqlparser-rs/workflows/Rust/badge.svg?branch=main)](https://github.com/sqlparser-rs/sqlparser-rs/actions?query=workflow%3ARust+branch%3Amain) +[![Build Status](https://github.com/apache/datafusion-sqlparser-rs/actions/workflows/rust.yml/badge.svg)](https://github.com/sqlparser-rs/sqlparser-rs/actions?query=workflow%3ARust+branch%3Amain) [![Coverage Status](https://coveralls.io/repos/github/sqlparser-rs/sqlparser-rs/badge.svg?branch=main)](https://coveralls.io/github/sqlparser-rs/sqlparser-rs?branch=main) [![Gitter Chat](https://badges.gitter.im/sqlparser-rs/community.svg)](https://gitter.im/sqlparser-rs/community?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge) From 4d9338638f0470ff3fe5d67ee89eed0bf4b90078 Mon Sep 17 00:00:00 2001 From: Sergey Olontsev Date: Thu, 17 Jul 2025 09:07:28 +0100 Subject: [PATCH 200/286] Fix for Postgres regex and like binary operators (#1928) --- src/parser/mod.rs | 10 ++++++- tests/sqlparser_postgres.rs | 60 ++++++++++++++++++++++++++++--------- 2 files changed, 55 insertions(+), 15 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b3ceec7e2..e219b20ae 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -3486,10 +3486,18 @@ impl<'a> Parser<'a> { | BinaryOperator::LtEq | BinaryOperator::Eq | BinaryOperator::NotEq + | BinaryOperator::PGRegexMatch + | BinaryOperator::PGRegexIMatch + | BinaryOperator::PGRegexNotMatch + | BinaryOperator::PGRegexNotIMatch + | BinaryOperator::PGLikeMatch + | BinaryOperator::PGILikeMatch + | BinaryOperator::PGNotLikeMatch + | BinaryOperator::PGNotILikeMatch ) { return parser_err!( format!( - "Expected one of [=, >, <, =>, =<, !=] as comparison operator, found: {op}" + "Expected one of [=, >, <, =>, =<, !=, ~, ~*, !~, !~*, ~~, ~~*, !~~, !~~*] as comparison operator, found: {op}" ), span.start ); diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 461277d81..d163096c4 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2187,21 +2187,39 @@ fn parse_pg_regex_match_ops() { ("!~*", BinaryOperator::PGRegexNotIMatch), ]; + // Match against a single value for (str_op, op) in pg_regex_match_ops { - let select = pg().verified_only_select(&format!("SELECT 'abc' {} '^a'", &str_op)); + let select = pg().verified_only_select(&format!("SELECT 'abc' {str_op} '^a'")); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value( - (Value::SingleQuotedString("abc".into())).with_empty_span() - )), + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), op: op.clone(), - right: Box::new(Expr::Value( - (Value::SingleQuotedString("^a".into())).with_empty_span() - )), + right: Box::new(Expr::Value(single_quoted_string("^a").with_empty_span(),)), }), select.projection[0] ); } + + // Match against any value from an array + for (str_op, op) in pg_regex_match_ops { + let select = + pg().verified_only_select(&format!("SELECT 'abc' {str_op} ANY(ARRAY['^a', 'x'])")); + assert_eq!( + SelectItem::UnnamedExpr(Expr::AnyOp { + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), + compare_op: op.clone(), + right: Box::new(Expr::Array(Array { + elem: vec![ + Expr::Value(single_quoted_string("^a").with_empty_span()), + Expr::Value(single_quoted_string("x").with_empty_span()), + ], + named: true, + })), + is_some: false, + }), + select.projection[0] + ) + } } #[test] @@ -2213,21 +2231,35 @@ fn parse_pg_like_match_ops() { ("!~~*", BinaryOperator::PGNotILikeMatch), ]; + // Match against a single value for (str_op, op) in pg_like_match_ops { - let select = pg().verified_only_select(&format!("SELECT 'abc' {} 'a_c%'", &str_op)); + let select = pg().verified_only_select(&format!("SELECT 'abc' {str_op} 'a_c%'")); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value( - (Value::SingleQuotedString("abc".into())).with_empty_span() - )), + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), op: op.clone(), - right: Box::new(Expr::Value( - (Value::SingleQuotedString("a_c%".into())).with_empty_span() - )), + right: Box::new(Expr::Value(single_quoted_string("a_c%").with_empty_span(),)), }), select.projection[0] ); } + + // Match against all values from an array + for (str_op, op) in pg_like_match_ops { + let select = + pg().verified_only_select(&format!("SELECT 'abc' {str_op} ALL(ARRAY['a_c%'])")); + assert_eq!( + SelectItem::UnnamedExpr(Expr::AllOp { + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), + compare_op: op.clone(), + right: Box::new(Expr::Array(Array { + elem: vec![Expr::Value(single_quoted_string("a_c%").with_empty_span())], + named: true, + })), + }), + select.projection[0] + ) + } } #[test] From 92db20673b29b4be471ed0a75c3d770a46631be8 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 17 Jul 2025 11:08:47 +0300 Subject: [PATCH 201/286] Snowflake: Improve accuracy of lookahead in implicit LIMIT alias (#1941) --- src/dialect/snowflake.rs | 38 +++++++++++++++++++++++++----------- tests/sqlparser_snowflake.rs | 18 +++++++++++++++++ 2 files changed, 45 insertions(+), 11 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index fcf94ee75..baf99b84b 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -23,8 +23,8 @@ use crate::ast::helpers::stmt_data_loading::{ FileStagingCommand, StageLoadSelectItem, StageLoadSelectItemKind, StageParamsObject, }; use crate::ast::{ - ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, - IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, + ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, DollarQuotedString, + Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, ObjectName, ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, Statement, TagsColumnOption, WrappedCollection, }; @@ -307,22 +307,22 @@ impl Dialect for SnowflakeDialect { // they are not followed by other tokens that may change their meaning // e.g. `SELECT * EXCEPT (col1) FROM tbl` Keyword::EXCEPT - // e.g. `SELECT 1 LIMIT 5` - | Keyword::LIMIT - // e.g. `SELECT 1 OFFSET 5 ROWS` - | Keyword::OFFSET // e.g. `INSERT INTO t SELECT 1 RETURNING *` | Keyword::RETURNING if !matches!(parser.peek_token_ref().token, Token::Comma | Token::EOF) => { false } + // e.g. `SELECT 1 LIMIT 5` - not an alias + // e.g. `SELECT 1 OFFSET 5 ROWS` - not an alias + Keyword::LIMIT | Keyword::OFFSET if peek_for_limit_options(parser) => false, + // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` // which would give it a different meanings, for example: // `SELECT 1 FETCH FIRST 10 ROWS` - not an alias // `SELECT 1 FETCH 10` - not an alias Keyword::FETCH if parser.peek_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]).is_some() - || matches!(parser.peek_token().token, Token::Number(_, _)) => + || peek_for_limit_options(parser) => { false } @@ -351,20 +351,23 @@ impl Dialect for SnowflakeDialect { match kw { // The following keywords can be considered an alias as long as // they are not followed by other tokens that may change their meaning - Keyword::LIMIT - | Keyword::RETURNING + Keyword::RETURNING | Keyword::INNER | Keyword::USING | Keyword::PIVOT | Keyword::UNPIVOT | Keyword::EXCEPT | Keyword::MATCH_RECOGNIZE - | Keyword::OFFSET if !matches!(parser.peek_token_ref().token, Token::SemiColon | Token::EOF) => { false } + // `LIMIT` can be considered an alias as long as it's not followed by a value. For example: + // `SELECT * FROM tbl LIMIT WHERE 1=1` - alias + // `SELECT * FROM tbl LIMIT 3` - not an alias + Keyword::LIMIT | Keyword::OFFSET if peek_for_limit_options(parser) => false, + // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` // which would give it a different meanings, for example: // `SELECT * FROM tbl FETCH FIRST 10 ROWS` - not an alias @@ -373,7 +376,7 @@ impl Dialect for SnowflakeDialect { if parser .peek_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]) .is_some() - || matches!(parser.peek_token().token, Token::Number(_, _)) => + || peek_for_limit_options(parser) => { false } @@ -387,6 +390,7 @@ impl Dialect for SnowflakeDialect { { false } + Keyword::GLOBAL if parser.peek_keyword(Keyword::FULL) => false, // Reserved keywords by the Snowflake dialect, which seem to be less strictive @@ -472,6 +476,18 @@ impl Dialect for SnowflakeDialect { } } +// Peeks ahead to identify tokens that are expected after +// a LIMIT/FETCH keyword. +fn peek_for_limit_options(parser: &Parser) -> bool { + match &parser.peek_token_ref().token { + Token::Number(_, _) | Token::Placeholder(_) => true, + Token::SingleQuotedString(val) if val.is_empty() => true, + Token::DollarQuotedString(DollarQuotedString { value, .. }) if value.is_empty() => true, + Token::Word(w) if w.keyword == Keyword::NULL => true, + _ => false, + } +} + fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { let stage = parse_snowflake_stage_name(parser)?; let pattern = if parser.parse_keyword(Keyword::PATTERN) { diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 562ddfea7..a7a633152 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3535,6 +3535,15 @@ fn test_sql_keywords_as_select_item_aliases() { .parse_sql_statements(&format!("SELECT 1 {kw}")) .is_err()); } + + // LIMIT is alias + snowflake().one_statement_parses_to("SELECT 1 LIMIT", "SELECT 1 AS LIMIT"); + // LIMIT is not an alias + snowflake().verified_stmt("SELECT 1 LIMIT 1"); + snowflake().verified_stmt("SELECT 1 LIMIT $1"); + snowflake().verified_stmt("SELECT 1 LIMIT ''"); + snowflake().verified_stmt("SELECT 1 LIMIT NULL"); + snowflake().verified_stmt("SELECT 1 LIMIT $$$$"); } #[test] @@ -3586,6 +3595,15 @@ fn test_sql_keywords_as_table_aliases() { .parse_sql_statements(&format!("SELECT * FROM tbl {kw}")) .is_err()); } + + // LIMIT is alias + snowflake().one_statement_parses_to("SELECT * FROM tbl LIMIT", "SELECT * FROM tbl AS LIMIT"); + // LIMIT is not an alias + snowflake().verified_stmt("SELECT * FROM tbl LIMIT 1"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT $1"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT ''"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT NULL"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT $$$$"); } #[test] From 5f69df2693ef3b6c153b1878a075437743998830 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 17 Jul 2025 11:24:14 +0300 Subject: [PATCH 202/286] Add support for `DROP USER` statement (#1951) --- src/ast/mod.rs | 2 ++ src/parser/mod.rs | 4 +++- tests/sqlparser_common.rs | 18 ++++++++++++++++++ 3 files changed, 23 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8bf750765..c2ec8c686 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -7768,6 +7768,7 @@ pub enum ObjectType { Sequence, Stage, Type, + User, } impl fmt::Display for ObjectType { @@ -7783,6 +7784,7 @@ impl fmt::Display for ObjectType { ObjectType::Sequence => "SEQUENCE", ObjectType::Stage => "STAGE", ObjectType::Type => "TYPE", + ObjectType::User => "USER", }) } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e219b20ae..d96a3002e 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6251,6 +6251,8 @@ impl<'a> Parser<'a> { ObjectType::Stage } else if self.parse_keyword(Keyword::TYPE) { ObjectType::Type + } else if self.parse_keyword(Keyword::USER) { + ObjectType::User } else if self.parse_keyword(Keyword::FUNCTION) { return self.parse_drop_function(); } else if self.parse_keyword(Keyword::POLICY) { @@ -6269,7 +6271,7 @@ impl<'a> Parser<'a> { return self.parse_drop_extension(); } else { return self.expected( - "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, VIEW, or MATERIALIZED VIEW after DROP", + "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, VIEW, MATERIALIZED VIEW or USER after DROP", self.peek_token(), ); }; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index e95c7e7b6..32bc70e65 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8354,6 +8354,24 @@ fn parse_drop_view() { verified_stmt("DROP MATERIALIZED VIEW IF EXISTS a.b.c"); } +#[test] +fn parse_drop_user() { + let sql = "DROP USER u1"; + match verified_stmt(sql) { + Statement::Drop { + names, object_type, .. + } => { + assert_eq!( + vec!["u1"], + names.iter().map(ToString::to_string).collect::>() + ); + assert_eq!(ObjectType::User, object_type); + } + _ => unreachable!(), + } + verified_stmt("DROP USER IF EXISTS u1"); +} + #[test] fn parse_invalid_subquery_without_parens() { let res = parse_sql_statements("SELECT SELECT 1 FROM bar WHERE 1=1 FROM baz"); From 40bbcc98340051efe6dc5fc8ea3c853c426d2072 Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Fri, 18 Jul 2025 09:05:25 -0400 Subject: [PATCH 203/286] Prepare 0.58.0 release: update version + Changelog (#1955) --- Cargo.toml | 2 +- changelog/0.58.0.md | 106 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 107 insertions(+), 1 deletion(-) create mode 100644 changelog/0.58.0.md diff --git a/Cargo.toml b/Cargo.toml index 07e44f66b..2e48f9d0d 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.57.0" +version = "0.58.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/changelog/0.58.0.md b/changelog/0.58.0.md new file mode 100644 index 000000000..27a985d7c --- /dev/null +++ b/changelog/0.58.0.md @@ -0,0 +1,106 @@ + + +# sqlparser-rs 0.58.0 Changelog + +This release consists of 47 commits from 18 contributors. See credits at the end of this changelog for more information. + +**Fixed bugs:** + +- fix: parse snowflake fetch clause [#1894](https://github.com/apache/datafusion-sqlparser-rs/pull/1894) (Vedin) + +**Documentation updates:** + +- docs: Update rust badge [#1943](https://github.com/apache/datafusion-sqlparser-rs/pull/1943) (Olexandr88) + +**Other:** + +- Add license header check to CI [#1888](https://github.com/apache/datafusion-sqlparser-rs/pull/1888) (alamb) +- Add support of parsing struct field's options in BigQuery [#1890](https://github.com/apache/datafusion-sqlparser-rs/pull/1890) (git-hulk) +- Fix parsing error when having fields after nested struct in BigQuery [#1897](https://github.com/apache/datafusion-sqlparser-rs/pull/1897) (git-hulk) +- Extend exception handling [#1884](https://github.com/apache/datafusion-sqlparser-rs/pull/1884) (bombsimon) +- Postgres: Add support for text search types [#1889](https://github.com/apache/datafusion-sqlparser-rs/pull/1889) (MohamedAbdeen21) +- Fix `limit` in subqueries [#1899](https://github.com/apache/datafusion-sqlparser-rs/pull/1899) (Dimchikkk) +- Use `IndexColumn` in all index definitions [#1900](https://github.com/apache/datafusion-sqlparser-rs/pull/1900) (mvzink) +- Support procedure argmode [#1901](https://github.com/apache/datafusion-sqlparser-rs/pull/1901) (ZacJW) +- Fix `impl Ord for Ident` [#1893](https://github.com/apache/datafusion-sqlparser-rs/pull/1893) (eliaperantoni) +- Snowflake: support multiple column options in `CREATE VIEW` [#1891](https://github.com/apache/datafusion-sqlparser-rs/pull/1891) (eliaperantoni) +- Add support for `LANGUAGE` clause in `CREATE PROCEDURE` [#1903](https://github.com/apache/datafusion-sqlparser-rs/pull/1903) (ZacJW) +- Fix clippy lints on 1.88.0 [#1910](https://github.com/apache/datafusion-sqlparser-rs/pull/1910) (iffyio) +- Snowflake: Add support for future grants [#1906](https://github.com/apache/datafusion-sqlparser-rs/pull/1906) (yoavcloud) +- Support for Map values in ClickHouse settings [#1896](https://github.com/apache/datafusion-sqlparser-rs/pull/1896) (solontsev) +- Fix join precedence for non-snowflake queries [#1905](https://github.com/apache/datafusion-sqlparser-rs/pull/1905) (Dimchikkk) +- Support remaining pipe operators [#1879](https://github.com/apache/datafusion-sqlparser-rs/pull/1879) (simonvandel) +- Make `GenericDialect` support from-first syntax [#1911](https://github.com/apache/datafusion-sqlparser-rs/pull/1911) (simonvandel) +- Redshift utf8 idents [#1915](https://github.com/apache/datafusion-sqlparser-rs/pull/1915) (yoavcloud) +- DuckDB: Add support for multiple `TRIM` arguments [#1916](https://github.com/apache/datafusion-sqlparser-rs/pull/1916) (ryanschneider) +- Redshift alter column type no set [#1912](https://github.com/apache/datafusion-sqlparser-rs/pull/1912) (yoavcloud) +- Postgres: support `ADD CONSTRAINT NOT VALID` and `VALIDATE CONSTRAINT` [#1908](https://github.com/apache/datafusion-sqlparser-rs/pull/1908) (achristmascarl) +- Add support for MySQL MEMBER OF [#1917](https://github.com/apache/datafusion-sqlparser-rs/pull/1917) (yoavcloud) +- Add span for `Expr::TypedString` [#1919](https://github.com/apache/datafusion-sqlparser-rs/pull/1919) (feral-dot-io) +- Support for Postgres `CREATE SERVER` [#1914](https://github.com/apache/datafusion-sqlparser-rs/pull/1914) (solontsev) +- Change tag and policy names to `ObjectName` [#1892](https://github.com/apache/datafusion-sqlparser-rs/pull/1892) (eliaperantoni) +- Add support for NULL escape char in pattern match searches [#1913](https://github.com/apache/datafusion-sqlparser-rs/pull/1913) (yoavcloud) +- Add support for dropping multiple columns in Snowflake [#1918](https://github.com/apache/datafusion-sqlparser-rs/pull/1918) (yoavcloud) +- Align Snowflake dialect to new test of reserved keywords [#1924](https://github.com/apache/datafusion-sqlparser-rs/pull/1924) (yoavcloud) +- Make `GenericDialect` support trailing commas in projections [#1921](https://github.com/apache/datafusion-sqlparser-rs/pull/1921) (simonvandel) +- Add support for several Snowflake grant statements [#1922](https://github.com/apache/datafusion-sqlparser-rs/pull/1922) (yoavcloud) +- Clickhouse: support empty parenthesized options [#1925](https://github.com/apache/datafusion-sqlparser-rs/pull/1925) (solontsev) +- Add Snowflake `COPY/REVOKE CURRENT GRANTS` option [#1926](https://github.com/apache/datafusion-sqlparser-rs/pull/1926) (yoavcloud) +- Add support for Snowflake identifier function [#1929](https://github.com/apache/datafusion-sqlparser-rs/pull/1929) (yoavcloud) +- Add support for granting privileges to procedures and functions in Snowflake [#1930](https://github.com/apache/datafusion-sqlparser-rs/pull/1930) (yoavcloud) +- Add support for `+` char in Snowflake stage names [#1935](https://github.com/apache/datafusion-sqlparser-rs/pull/1935) (yoavcloud) +- Snowflake Reserved SQL Keywords as Implicit Table Alias [#1934](https://github.com/apache/datafusion-sqlparser-rs/pull/1934) (yoavcloud) +- Add support for Redshift `SELECT * EXCLUDE` [#1936](https://github.com/apache/datafusion-sqlparser-rs/pull/1936) (yoavcloud) +- Support optional semicolon between statements [#1937](https://github.com/apache/datafusion-sqlparser-rs/pull/1937) (yoavcloud) +- Snowflake: support trailing options in `CREATE TABLE` [#1931](https://github.com/apache/datafusion-sqlparser-rs/pull/1931) (yoavcloud) +- MSSQL: Add support for EXEC output and default keywords [#1940](https://github.com/apache/datafusion-sqlparser-rs/pull/1940) (yoavcloud) +- Add identifier unicode support in Mysql, Postgres and Redshift [#1933](https://github.com/apache/datafusion-sqlparser-rs/pull/1933) (etgarperets) +- Add identifier start unicode support for Postegres, MySql and Redshift [#1944](https://github.com/apache/datafusion-sqlparser-rs/pull/1944) (etgarperets) +- Fix for Postgres regex and like binary operators [#1928](https://github.com/apache/datafusion-sqlparser-rs/pull/1928) (solontsev) +- Snowflake: Improve accuracy of lookahead in implicit LIMIT alias [#1941](https://github.com/apache/datafusion-sqlparser-rs/pull/1941) (yoavcloud) +- Add support for `DROP USER` statement [#1951](https://github.com/apache/datafusion-sqlparser-rs/pull/1951) (yoavcloud) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 19 Yoav Cohen + 4 Sergey Olontsev + 3 Elia Perantoni + 3 Simon Vandel Sillesen + 2 Dima + 2 ZacJW + 2 etgarperets + 2 hulk + 1 Andrew Lamb + 1 Denys Tsomenko + 1 Ifeanyi Ubah + 1 Michael Victor Zink + 1 Mohamed Abdeen + 1 Olexandr88 + 1 Ryan Schneider + 1 Simon Sawert + 1 carl + 1 feral-dot-io +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + From 23f40cdc409708c1d71cd72d9a2c57d8c0d9ad1b Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Fri, 18 Jul 2025 16:41:50 +0300 Subject: [PATCH 204/286] MySQL: Support `EXPLAIN ANALYZE` format variants (#1945) --- src/ast/mod.rs | 27 +++++++++++++++++++++++++-- src/parser/mod.rs | 10 +++++++++- tests/sqlparser_common.rs | 17 +++++++++++++---- 3 files changed, 47 insertions(+), 7 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index c2ec8c686..1f5311aff 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4126,7 +4126,7 @@ pub enum Statement { /// A SQL query that specifies what to explain statement: Box, /// Optional output format of explain - format: Option, + format: Option, /// Postgres style utility options, `(analyze, verbose true)` options: Option>, }, @@ -4494,7 +4494,7 @@ impl fmt::Display for Statement { } if let Some(format) = format { - write!(f, "FORMAT {format} ")?; + write!(f, "{format} ")?; } if let Some(options) = options { @@ -7641,6 +7641,25 @@ impl fmt::Display for DuplicateTreatment { } } +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AnalyzeFormatKind { + /// e.g. `EXPLAIN ANALYZE FORMAT JSON SELECT * FROM tbl` + Keyword(AnalyzeFormat), + /// e.g. `EXPLAIN ANALYZE FORMAT=JSON SELECT * FROM tbl` + Assignment(AnalyzeFormat), +} + +impl fmt::Display for AnalyzeFormatKind { + fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result { + match self { + AnalyzeFormatKind::Keyword(format) => write!(f, "FORMAT {format}"), + AnalyzeFormatKind::Assignment(format) => write!(f, "FORMAT={format}"), + } + } +} + #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -7648,6 +7667,8 @@ pub enum AnalyzeFormat { TEXT, GRAPHVIZ, JSON, + TRADITIONAL, + TREE, } impl fmt::Display for AnalyzeFormat { @@ -7656,6 +7677,8 @@ impl fmt::Display for AnalyzeFormat { AnalyzeFormat::TEXT => "TEXT", AnalyzeFormat::GRAPHVIZ => "GRAPHVIZ", AnalyzeFormat::JSON => "JSON", + AnalyzeFormat::TRADITIONAL => "TRADITIONAL", + AnalyzeFormat::TREE => "TREE", }) } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d96a3002e..3bb913118 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5674,6 +5674,14 @@ impl<'a> Parser<'a> { } } + fn parse_analyze_format_kind(&mut self) -> Result { + if self.consume_token(&Token::Eq) { + Ok(AnalyzeFormatKind::Assignment(self.parse_analyze_format()?)) + } else { + Ok(AnalyzeFormatKind::Keyword(self.parse_analyze_format()?)) + } + } + pub fn parse_analyze_format(&mut self) -> Result { let next_token = self.next_token(); match &next_token.token { @@ -11074,7 +11082,7 @@ impl<'a> Parser<'a> { analyze = self.parse_keyword(Keyword::ANALYZE); verbose = self.parse_keyword(Keyword::VERBOSE); if self.parse_keyword(Keyword::FORMAT) { - format = Some(self.parse_analyze_format()?); + format = Some(self.parse_analyze_format_kind()?); } } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 32bc70e65..4183c5539 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -5161,7 +5161,7 @@ fn run_explain_analyze( query: &str, expected_verbose: bool, expected_analyze: bool, - expected_format: Option, + expected_format: Option, expected_options: Option>, ) { match dialect.verified_stmt(query) { @@ -5272,7 +5272,7 @@ fn parse_explain_analyze_with_simple_select() { "EXPLAIN ANALYZE FORMAT GRAPHVIZ SELECT sqrt(id) FROM foo", false, true, - Some(AnalyzeFormat::GRAPHVIZ), + Some(AnalyzeFormatKind::Keyword(AnalyzeFormat::GRAPHVIZ)), None, ); @@ -5281,7 +5281,16 @@ fn parse_explain_analyze_with_simple_select() { "EXPLAIN ANALYZE VERBOSE FORMAT JSON SELECT sqrt(id) FROM foo", true, true, - Some(AnalyzeFormat::JSON), + Some(AnalyzeFormatKind::Keyword(AnalyzeFormat::JSON)), + None, + ); + + run_explain_analyze( + all_dialects(), + "EXPLAIN ANALYZE VERBOSE FORMAT=JSON SELECT sqrt(id) FROM foo", + true, + true, + Some(AnalyzeFormatKind::Assignment(AnalyzeFormat::JSON)), None, ); @@ -5290,7 +5299,7 @@ fn parse_explain_analyze_with_simple_select() { "EXPLAIN VERBOSE FORMAT TEXT SELECT sqrt(id) FROM foo", true, false, - Some(AnalyzeFormat::TEXT), + Some(AnalyzeFormatKind::Keyword(AnalyzeFormat::TEXT)), None, ); } From a73577c29fc86b365f14545a0734e619b439dbf4 Mon Sep 17 00:00:00 2001 From: Ryan Schneider Date: Mon, 21 Jul 2025 03:58:20 -0700 Subject: [PATCH 205/286] Add support for `NOT NULL` and `NOTNULL` expressions (#1927) Co-authored-by: Ifeanyi Ubah --- src/dialect/duckdb.rs | 6 ++ src/dialect/mod.rs | 14 +++++ src/dialect/postgresql.rs | 6 ++ src/dialect/sqlite.rs | 6 ++ src/keywords.rs | 1 + src/parser/mod.rs | 104 +++++++++++++++++++++++++++++++--- tests/sqlparser_clickhouse.rs | 24 ++++++++ tests/sqlparser_common.rs | 47 +++++++++++++++ 8 files changed, 200 insertions(+), 8 deletions(-) diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index fa18463ae..aee7ee935 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -98,4 +98,10 @@ impl Dialect for DuckDbDialect { fn supports_select_wildcard_exclude(&self) -> bool { true } + + /// DuckDB supports `NOTNULL` as an alias for `IS NOT NULL`, + /// see DuckDB Comparisons + fn supports_notnull_operator(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index c79b279df..c78b00033 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -677,8 +677,16 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::MEMBER => Ok(p!(Like)), + Token::Word(w) + if w.keyword == Keyword::NULL && !parser.in_column_definition_state() => + { + Ok(p!(Is)) + } _ => Ok(self.prec_unknown()), }, + Token::Word(w) if w.keyword == Keyword::NOTNULL && self.supports_notnull_operator() => { + Ok(p!(Is)) + } Token::Word(w) if w.keyword == Keyword::IS => Ok(p!(Is)), Token::Word(w) if w.keyword == Keyword::IN => Ok(p!(Between)), Token::Word(w) if w.keyword == Keyword::BETWEEN => Ok(p!(Between)), @@ -1122,6 +1130,12 @@ pub trait Dialect: Debug + Any { ) -> bool { false } + + /// Returns true if the dialect supports the `x NOTNULL` + /// operator expression. + fn supports_notnull_operator(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 9cea252c8..b12abaaf7 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -263,4 +263,10 @@ impl Dialect for PostgreSqlDialect { fn supports_alter_column_type_using(&self) -> bool { true } + + /// Postgres supports `NOTNULL` as an alias for `IS NOT NULL` + /// See: + fn supports_notnull_operator(&self) -> bool { + true + } } diff --git a/src/dialect/sqlite.rs b/src/dialect/sqlite.rs index 847e0d135..64a8d532f 100644 --- a/src/dialect/sqlite.rs +++ b/src/dialect/sqlite.rs @@ -110,4 +110,10 @@ impl Dialect for SQLiteDialect { fn supports_dollar_placeholder(&self) -> bool { true } + + /// SQLite supports `NOTNULL` as aliases for `IS NOT NULL` + /// See: + fn supports_notnull_operator(&self) -> bool { + true + } } diff --git a/src/keywords.rs b/src/keywords.rs index 9e689a6df..7781939bc 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -608,6 +608,7 @@ define_keywords!( NOT, NOTHING, NOTIFY, + NOTNULL, NOWAIT, NO_WRITE_TO_BINLOG, NTH_VALUE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 3bb913118..19acd4b55 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -38,6 +38,7 @@ use crate::ast::*; use crate::dialect::*; use crate::keywords::{Keyword, ALL_KEYWORDS}; use crate::tokenizer::*; +use sqlparser::parser::ParserState::ColumnDefinition; mod alter; @@ -275,6 +276,12 @@ enum ParserState { /// PRIOR expressions while still allowing prior as an identifier name /// in other contexts. ConnectBy, + /// The state when parsing column definitions. This state prohibits + /// NOT NULL as an alias for IS NOT NULL. For example: + /// ```sql + /// CREATE TABLE foo (abc BIGINT NOT NULL); + /// ``` + ColumnDefinition, } /// A SQL Parser @@ -3578,6 +3585,11 @@ impl<'a> Parser<'a> { let negated = self.parse_keyword(Keyword::NOT); let regexp = self.parse_keyword(Keyword::REGEXP); let rlike = self.parse_keyword(Keyword::RLIKE); + let null = if !self.in_column_definition_state() { + self.parse_keyword(Keyword::NULL) + } else { + false + }; if regexp || rlike { Ok(Expr::RLike { negated, @@ -3587,6 +3599,8 @@ impl<'a> Parser<'a> { ), regexp, }) + } else if negated && null { + Ok(Expr::IsNotNull(Box::new(expr))) } else if self.parse_keyword(Keyword::IN) { self.parse_in(expr, negated) } else if self.parse_keyword(Keyword::BETWEEN) { @@ -3624,6 +3638,9 @@ impl<'a> Parser<'a> { self.expected("IN or BETWEEN after NOT", self.peek_token()) } } + Keyword::NOTNULL if dialect.supports_notnull_operator() => { + Ok(Expr::IsNotNull(Box::new(expr))) + } Keyword::MEMBER => { if self.parse_keyword(Keyword::OF) { self.expect_token(&Token::LParen)?; @@ -7742,6 +7759,15 @@ impl<'a> Parser<'a> { return option; } + self.with_state( + ColumnDefinition, + |parser| -> Result, ParserError> { + parser.parse_optional_column_option_inner() + }, + ) + } + + fn parse_optional_column_option_inner(&mut self) -> Result, ParserError> { if self.parse_keywords(&[Keyword::CHARACTER, Keyword::SET]) { Ok(Some(ColumnOption::CharacterSet( self.parse_object_name(false)?, @@ -7757,15 +7783,19 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::NULL) { Ok(Some(ColumnOption::Null)) } else if self.parse_keyword(Keyword::DEFAULT) { - Ok(Some(ColumnOption::Default(self.parse_expr()?))) + Ok(Some(ColumnOption::Default( + self.parse_column_option_expr()?, + ))) } else if dialect_of!(self is ClickHouseDialect| GenericDialect) && self.parse_keyword(Keyword::MATERIALIZED) { - Ok(Some(ColumnOption::Materialized(self.parse_expr()?))) + Ok(Some(ColumnOption::Materialized( + self.parse_column_option_expr()?, + ))) } else if dialect_of!(self is ClickHouseDialect| GenericDialect) && self.parse_keyword(Keyword::ALIAS) { - Ok(Some(ColumnOption::Alias(self.parse_expr()?))) + Ok(Some(ColumnOption::Alias(self.parse_column_option_expr()?))) } else if dialect_of!(self is ClickHouseDialect| GenericDialect) && self.parse_keyword(Keyword::EPHEMERAL) { @@ -7774,7 +7804,9 @@ impl<'a> Parser<'a> { if matches!(self.peek_token().token, Token::Comma | Token::RParen) { Ok(Some(ColumnOption::Ephemeral(None))) } else { - Ok(Some(ColumnOption::Ephemeral(Some(self.parse_expr()?)))) + Ok(Some(ColumnOption::Ephemeral(Some( + self.parse_column_option_expr()?, + )))) } } else if self.parse_keywords(&[Keyword::PRIMARY, Keyword::KEY]) { let characteristics = self.parse_constraint_characteristics()?; @@ -7817,7 +7849,8 @@ impl<'a> Parser<'a> { })) } else if self.parse_keyword(Keyword::CHECK) { self.expect_token(&Token::LParen)?; - let expr = self.parse_expr()?; + // since `CHECK` requires parentheses, we can parse the inner expression in ParserState::Normal + let expr: Expr = self.with_state(ParserState::Normal, |p| p.parse_expr())?; self.expect_token(&Token::RParen)?; Ok(Some(ColumnOption::Check(expr))) } else if self.parse_keyword(Keyword::AUTO_INCREMENT) @@ -7851,7 +7884,7 @@ impl<'a> Parser<'a> { } else if self.parse_keywords(&[Keyword::ON, Keyword::UPDATE]) && dialect_of!(self is MySqlDialect | GenericDialect) { - let expr = self.parse_expr()?; + let expr = self.parse_column_option_expr()?; Ok(Some(ColumnOption::OnUpdate(expr))) } else if self.parse_keyword(Keyword::GENERATED) { self.parse_optional_column_option_generated() @@ -7869,7 +7902,9 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::SRID) && dialect_of!(self is MySqlDialect | GenericDialect) { - Ok(Some(ColumnOption::Srid(Box::new(self.parse_expr()?)))) + Ok(Some(ColumnOption::Srid(Box::new( + self.parse_column_option_expr()?, + )))) } else if self.parse_keyword(Keyword::IDENTITY) && dialect_of!(self is MsSqlDialect | GenericDialect) { @@ -7909,6 +7944,31 @@ impl<'a> Parser<'a> { } } + /// When parsing some column option expressions we need to revert to [ParserState::Normal] since + /// `NOT NULL` is allowed as an alias for `IS NOT NULL`. + /// In those cases we use this helper instead of calling [Parser::parse_expr] directly. + /// + /// For example, consider these `CREATE TABLE` statements: + /// ```sql + /// CREATE TABLE foo (abc BOOL DEFAULT (42 NOT NULL) NOT NULL); + /// ``` + /// vs + /// ```sql + /// CREATE TABLE foo (abc BOOL NOT NULL); + /// ``` + /// + /// In the first we should parse the inner portion of `(42 NOT NULL)` as [Expr::IsNotNull], + /// whereas is both statements that trailing `NOT NULL` should only be parsed as a + /// [ColumnOption::NotNull]. + fn parse_column_option_expr(&mut self) -> Result { + if self.peek_token_ref().token == Token::LParen { + let expr: Expr = self.with_state(ParserState::Normal, |p| p.parse_prefix())?; + Ok(expr) + } else { + Ok(self.parse_expr()?) + } + } + pub(crate) fn parse_tag(&mut self) -> Result { let name = self.parse_object_name(false)?; self.expect_token(&Token::Eq)?; @@ -7953,7 +8013,7 @@ impl<'a> Parser<'a> { })) } else if self.parse_keywords(&[Keyword::ALWAYS, Keyword::AS]) { if self.expect_token(&Token::LParen).is_ok() { - let expr = self.parse_expr()?; + let expr: Expr = self.with_state(ParserState::Normal, |p| p.parse_expr())?; self.expect_token(&Token::RParen)?; let (gen_as, expr_mode) = if self.parse_keywords(&[Keyword::STORED]) { Ok(( @@ -16532,6 +16592,10 @@ impl<'a> Parser<'a> { Ok(None) } } + + pub(crate) fn in_column_definition_state(&self) -> bool { + matches!(self.state, ColumnDefinition) + } } fn maybe_prefixed_expr(expr: Expr, prefix: Option) -> Expr { @@ -17267,4 +17331,28 @@ mod tests { assert!(Parser::parse_sql(&MySqlDialect {}, sql).is_err()); } + + #[test] + fn test_parse_not_null_in_column_options() { + let canonical = concat!( + "CREATE TABLE foo (", + "abc INT DEFAULT (42 IS NOT NULL) NOT NULL,", + " def INT,", + " def_null BOOL GENERATED ALWAYS AS (def IS NOT NULL) STORED,", + " CHECK (abc IS NOT NULL)", + ")" + ); + all_dialects().verified_stmt(canonical); + all_dialects().one_statement_parses_to( + concat!( + "CREATE TABLE foo (", + "abc INT DEFAULT (42 NOT NULL) NOT NULL,", + " def INT,", + " def_null BOOL GENERATED ALWAYS AS (def NOT NULL) STORED,", + " CHECK (abc NOT NULL)", + ")" + ), + canonical, + ); + } } diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index 9e5b6ce8a..bc1431f9c 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -1705,6 +1705,30 @@ fn parse_table_sample() { clickhouse().verified_stmt("SELECT * FROM tbl SAMPLE 1 / 10 OFFSET 1 / 2"); } +#[test] +fn test_parse_not_null_in_column_options() { + // In addition to DEFAULT and CHECK ClickHouse also supports MATERIALIZED, all of which + // can contain `IS NOT NULL` and thus `NOT NULL` as an alias. + let canonical = concat!( + "CREATE TABLE foo (", + "abc INT DEFAULT (42 IS NOT NULL) NOT NULL,", + " not_null BOOL MATERIALIZED (abc IS NOT NULL),", + " CHECK (abc IS NOT NULL)", + ")", + ); + clickhouse().verified_stmt(canonical); + clickhouse().one_statement_parses_to( + concat!( + "CREATE TABLE foo (", + "abc INT DEFAULT (42 NOT NULL) NOT NULL,", + " not_null BOOL MATERIALIZED (abc NOT NULL),", + " CHECK (abc NOT NULL)", + ")", + ), + canonical, + ); +} + fn clickhouse() -> TestedDialects { TestedDialects::new(vec![Box::new(ClickHouseDialect {})]) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 4183c5539..4995b9415 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16040,6 +16040,27 @@ fn parse_create_procedure_with_parameter_modes() { } } +#[test] +fn parse_not_null() { + let _ = all_dialects().expr_parses_to("x NOT NULL", "x IS NOT NULL"); + let _ = all_dialects().expr_parses_to("NULL NOT NULL", "NULL IS NOT NULL"); + + assert_matches!( + all_dialects().expr_parses_to("NOT NULL NOT NULL", "NOT NULL IS NOT NULL"), + Expr::UnaryOp { + op: UnaryOperator::Not, + .. + } + ); + assert_matches!( + all_dialects().expr_parses_to("NOT x NOT NULL", "NOT x IS NOT NULL"), + Expr::UnaryOp { + op: UnaryOperator::Not, + .. + } + ); +} + #[test] fn test_select_exclude() { let dialects = all_dialects_where(|d| d.supports_select_wildcard_exclude()); @@ -16183,3 +16204,29 @@ fn test_identifier_unicode_start() { ]); let _ = dialects.verified_stmt(sql); } + +#[test] +fn parse_notnull() { + // Some dialects support `x NOTNULL` as an expression while others consider + // `x NOTNULL` like `x AS NOTNULL` and thus consider `NOTNULL` an alias for x. + let notnull_unsupported_dialects = all_dialects_except(|d| d.supports_notnull_operator()); + let _ = notnull_unsupported_dialects + .verified_only_select_with_canonical("SELECT NULL NOTNULL", "SELECT NULL AS NOTNULL"); + + // Supported dialects consider `x NOTNULL` as an alias for `x IS NOT NULL` + let notnull_supported_dialects = all_dialects_where(|d| d.supports_notnull_operator()); + let _ = notnull_supported_dialects.expr_parses_to("x NOTNULL", "x IS NOT NULL"); + + // For dialects which support it, `NOT NULL NOTNULL` should + // parse as `(NOT (NULL IS NOT NULL))` + assert_matches!( + notnull_supported_dialects.expr_parses_to("NOT NULL NOTNULL", "NOT NULL IS NOT NULL"), + Expr::UnaryOp { + op: UnaryOperator::Not, + .. + } + ); + + // for unsupported dialects, parsing should stop at `NOT NULL` + notnull_unsupported_dialects.expr_parses_to("NOT NULL NOTNULL", "NOT NULL"); +} From 799c1f748d32b72010cffe6e4a350d17861fed04 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 21 Jul 2025 14:41:20 +0300 Subject: [PATCH 206/286] Snowflake: Support `CLONE` option in `CREATE DATABASE/SCHEMA` statements (#1958) --- src/ast/mod.rs | 24 ++++++++++++++++++++++++ src/parser/mod.rs | 14 ++++++++++++++ tests/sqlparser_common.rs | 25 +++++++++++++++++++++++++ 3 files changed, 63 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 1f5311aff..b321bb535 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3846,6 +3846,14 @@ pub enum Statement { /// /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_schema_statement) default_collate_spec: Option, + /// Clones a schema + /// + /// ```sql + /// CREATE SCHEMA myschema CLONE otherschema + /// ``` + /// + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-clone#databases-schemas) + clone: Option, }, /// ```sql /// CREATE DATABASE @@ -3855,6 +3863,14 @@ pub enum Statement { if_not_exists: bool, location: Option, managed_location: Option, + /// Clones a database + /// + /// ```sql + /// CREATE DATABASE mydb CLONE otherdb + /// ``` + /// + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-clone#databases-schemas) + clone: Option, }, /// ```sql /// CREATE FUNCTION @@ -4797,6 +4813,7 @@ impl fmt::Display for Statement { if_not_exists, location, managed_location, + clone, } => { write!(f, "CREATE DATABASE")?; if *if_not_exists { @@ -4809,6 +4826,9 @@ impl fmt::Display for Statement { if let Some(ml) = managed_location { write!(f, " MANAGEDLOCATION '{ml}'")?; } + if let Some(clone) = clone { + write!(f, " CLONE {clone}")?; + } Ok(()) } Statement::CreateFunction(create_function) => create_function.fmt(f), @@ -5730,6 +5750,7 @@ impl fmt::Display for Statement { with, options, default_collate_spec, + clone, } => { write!( f, @@ -5750,6 +5771,9 @@ impl fmt::Display for Statement { write!(f, " OPTIONS({})", display_comma_separated(options))?; } + if let Some(clone) = clone { + write!(f, " CLONE {clone}")?; + } Ok(()) } Statement::Assert { condition, message } => { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 19acd4b55..7cccde370 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4917,12 +4917,19 @@ impl<'a> Parser<'a> { None }; + let clone = if self.parse_keyword(Keyword::CLONE) { + Some(self.parse_object_name(false)?) + } else { + None + }; + Ok(Statement::CreateSchema { schema_name, if_not_exists, with, options, default_collate_spec, + clone, }) } @@ -4957,11 +4964,18 @@ impl<'a> Parser<'a> { _ => break, } } + let clone = if self.parse_keyword(Keyword::CLONE) { + Some(self.parse_object_name(false)?) + } else { + None + }; + Ok(Statement::CreateDatabase { db_name, if_not_exists: ine, location, managed_location, + clone, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 4995b9415..bbe202264 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4296,6 +4296,7 @@ fn parse_create_schema() { verified_stmt(r#"CREATE SCHEMA a.b.c WITH (key1 = 'value1', key2 = 'value2')"#); verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a WITH (key1 = 'value1')"#); verified_stmt(r#"CREATE SCHEMA IF NOT EXISTS a WITH ()"#); + verified_stmt(r#"CREATE SCHEMA a CLONE b"#); } #[test] @@ -7900,11 +7901,33 @@ fn parse_create_database() { if_not_exists, location, managed_location, + clone, } => { assert_eq!("mydb", db_name.to_string()); assert!(!if_not_exists); assert_eq!(None, location); assert_eq!(None, managed_location); + assert_eq!(None, clone); + } + _ => unreachable!(), + } + let sql = "CREATE DATABASE mydb CLONE otherdb"; + match verified_stmt(sql) { + Statement::CreateDatabase { + db_name, + if_not_exists, + location, + managed_location, + clone, + } => { + assert_eq!("mydb", db_name.to_string()); + assert!(!if_not_exists); + assert_eq!(None, location); + assert_eq!(None, managed_location); + assert_eq!( + Some(ObjectName::from(vec![Ident::new("otherdb".to_string())])), + clone + ); } _ => unreachable!(), } @@ -7919,11 +7942,13 @@ fn parse_create_database_ine() { if_not_exists, location, managed_location, + clone, } => { assert_eq!("mydb", db_name.to_string()); assert!(if_not_exists); assert_eq!(None, location); assert_eq!(None, managed_location); + assert_eq!(None, clone); } _ => unreachable!(), } From 492184643a58724f1f3b2eb2f4f7aeb9cd20478c Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Mon, 21 Jul 2025 14:43:29 +0300 Subject: [PATCH 207/286] Add support for `GRANT DROP` statement (#1959) --- src/ast/mod.rs | 2 ++ src/parser/mod.rs | 2 ++ tests/sqlparser_common.rs | 3 ++- 3 files changed, 6 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index b321bb535..1798223f3 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6611,6 +6611,7 @@ pub enum Action { role: ObjectName, }, Delete, + Drop, EvolveSchema, Exec { obj_type: Option, @@ -6680,6 +6681,7 @@ impl fmt::Display for Action { } Action::DatabaseRole { role } => write!(f, "DATABASE ROLE {role}")?, Action::Delete => f.write_str("DELETE")?, + Action::Drop => f.write_str("DROP")?, Action::EvolveSchema => f.write_str("EVOLVE SCHEMA")?, Action::Exec { obj_type } => { f.write_str("EXEC")?; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 7cccde370..8d5a55da0 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -14339,6 +14339,8 @@ impl<'a> Parser<'a> { Ok(Action::Usage) } else if self.parse_keyword(Keyword::OWNERSHIP) { Ok(Action::Ownership) + } else if self.parse_keyword(Keyword::DROP) { + Ok(Action::Drop) } else { self.expected("a privilege keyword", self.peek_token())? } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index bbe202264..5d8284a46 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9342,7 +9342,7 @@ fn parse_drop_role() { #[test] fn parse_grant() { - let sql = "GRANT SELECT, INSERT, UPDATE (shape, size), USAGE, DELETE, TRUNCATE, REFERENCES, TRIGGER, CONNECT, CREATE, EXECUTE, TEMPORARY ON abc, def TO xyz, m WITH GRANT OPTION GRANTED BY jj"; + let sql = "GRANT SELECT, INSERT, UPDATE (shape, size), USAGE, DELETE, TRUNCATE, REFERENCES, TRIGGER, CONNECT, CREATE, EXECUTE, TEMPORARY, DROP ON abc, def TO xyz, m WITH GRANT OPTION GRANTED BY jj"; match verified_stmt(sql) { Statement::Grant { privileges, @@ -9380,6 +9380,7 @@ fn parse_grant() { Action::Create { obj_type: None }, Action::Execute { obj_type: None }, Action::Temporary, + Action::Drop, ], actions ); From 2ed2cbe291150d4a28f2a4f20f9c95ceead57b60 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 23 Jul 2025 18:52:08 +0300 Subject: [PATCH 208/286] Snowflake: Add support for `CREATE USER` (#1950) --- src/ast/helpers/key_value_options.rs | 28 ++++--- src/ast/mod.rs | 50 ++++++++++++ src/ast/spans.rs | 1 + src/dialect/snowflake.rs | 109 +++++++++------------------ src/parser/mod.rs | 107 +++++++++++++++++++++++++- tests/sqlparser_common.rs | 71 +++++++++++++++++ 6 files changed, 281 insertions(+), 85 deletions(-) diff --git a/src/ast/helpers/key_value_options.rs b/src/ast/helpers/key_value_options.rs index 796bfd5e3..7f1bb0fdb 100644 --- a/src/ast/helpers/key_value_options.rs +++ b/src/ast/helpers/key_value_options.rs @@ -31,11 +31,22 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; +use crate::ast::display_separated; + #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct KeyValueOptions { pub options: Vec, + pub delimiter: KeyValueOptionsDelimiter, +} + +#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum KeyValueOptionsDelimiter { + Space, + Comma, } #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] @@ -59,18 +70,11 @@ pub struct KeyValueOption { impl fmt::Display for KeyValueOptions { fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result { - if !self.options.is_empty() { - let mut first = false; - for option in &self.options { - if !first { - first = true; - } else { - f.write_str(" ")?; - } - write!(f, "{option}")?; - } - } - Ok(()) + let sep = match self.delimiter { + KeyValueOptionsDelimiter::Space => " ", + KeyValueOptionsDelimiter::Comma => ", ", + }; + write!(f, "{}", display_separated(&self.options, sep)) } } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 1798223f3..e16463495 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4355,6 +4355,11 @@ pub enum Statement { /// /// See [ReturnStatement] Return(ReturnStatement), + /// ```sql + /// CREATE [OR REPLACE] USER [IF NOT EXISTS] + /// ``` + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-user) + CreateUser(CreateUser), } /// ```sql @@ -6193,6 +6198,7 @@ impl fmt::Display for Statement { Statement::Return(r) => write!(f, "{r}"), Statement::List(command) => write!(f, "LIST {command}"), Statement::Remove(command) => write!(f, "REMOVE {command}"), + Statement::CreateUser(s) => write!(f, "{s}"), } } } @@ -10125,6 +10131,50 @@ impl fmt::Display for MemberOf { } } +/// Creates a user +/// +/// Syntax: +/// ```sql +/// CREATE [OR REPLACE] USER [IF NOT EXISTS] [OPTIONS] +/// ``` +/// +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-user) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateUser { + pub or_replace: bool, + pub if_not_exists: bool, + pub name: Ident, + pub options: KeyValueOptions, + pub with_tags: bool, + pub tags: KeyValueOptions, +} + +impl fmt::Display for CreateUser { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "CREATE")?; + if self.or_replace { + write!(f, " OR REPLACE")?; + } + write!(f, " USER")?; + if self.if_not_exists { + write!(f, " IF NOT EXISTS")?; + } + write!(f, " {}", self.name)?; + if !self.options.options.is_empty() { + write!(f, " {}", self.options)?; + } + if !self.tags.options.is_empty() { + if self.with_tags { + write!(f, " WITH")?; + } + write!(f, " TAG ({})", self.tags)?; + } + Ok(()) + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 3e82905e1..4deedca04 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -531,6 +531,7 @@ impl Spanned for Statement { Statement::Print { .. } => Span::empty(), Statement::Return { .. } => Span::empty(), Statement::List(..) | Statement::Remove(..) => Span::empty(), + Statement::CreateUser(..) => Span::empty(), } } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index baf99b84b..9786aba0d 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -17,7 +17,9 @@ #[cfg(not(feature = "std"))] use crate::alloc::string::ToString; -use crate::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionType, KeyValueOptions}; +use crate::ast::helpers::key_value_options::{ + KeyValueOption, KeyValueOptionType, KeyValueOptions, KeyValueOptionsDelimiter, +}; use crate::ast::helpers::stmt_create_table::CreateTableBuilder; use crate::ast::helpers::stmt_data_loading::{ FileStagingCommand, StageLoadSelectItem, StageLoadSelectItemKind, StageParamsObject, @@ -31,7 +33,7 @@ use crate::ast::{ use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; use crate::parser::{IsOptional, Parser, ParserError}; -use crate::tokenizer::{Token, Word}; +use crate::tokenizer::Token; #[cfg(not(feature = "std"))] use alloc::boxed::Box; #[cfg(not(feature = "std"))] @@ -516,6 +518,7 @@ fn parse_alter_session(parser: &mut Parser, set: bool) -> Result Result { let mut from_stage = None; let mut stage_params = StageParamsObject { url: None, - encryption: KeyValueOptions { options: vec![] }, + encryption: KeyValueOptions { + options: vec![], + delimiter: KeyValueOptionsDelimiter::Space, + }, endpoint: None, storage_integration: None, - credentials: KeyValueOptions { options: vec![] }, + credentials: KeyValueOptions { + options: vec![], + delimiter: KeyValueOptionsDelimiter::Space, + }, }; let mut from_query = None; let mut partition = None; @@ -944,7 +956,7 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { // FILE_FORMAT if parser.parse_keyword(Keyword::FILE_FORMAT) { parser.expect_token(&Token::Eq)?; - file_format = parse_parentheses_options(parser)?; + file_format = parser.parse_key_value_options(true, &[])?; // PARTITION BY } else if parser.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) { partition = Some(Box::new(parser.parse_expr()?)) @@ -982,14 +994,14 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { // COPY OPTIONS } else if parser.parse_keyword(Keyword::COPY_OPTIONS) { parser.expect_token(&Token::Eq)?; - copy_options = parse_parentheses_options(parser)?; + copy_options = parser.parse_key_value_options(true, &[])?; } else { match parser.next_token().token { Token::SemiColon | Token::EOF => break, Token::Comma => continue, // In `COPY INTO ` the copy options do not have a shared key // like in `COPY INTO
` - Token::Word(key) => copy_options.push(parse_option(parser, key)?), + Token::Word(key) => copy_options.push(parser.parse_key_value_option(key)?), _ => return parser.expected("another copy option, ; or EOF'", parser.peek_token()), } } @@ -1008,9 +1020,11 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { pattern, file_format: KeyValueOptions { options: file_format, + delimiter: KeyValueOptionsDelimiter::Space, }, copy_options: KeyValueOptions { options: copy_options, + delimiter: KeyValueOptionsDelimiter::Space, }, validation_mode, partition, @@ -1110,8 +1124,14 @@ fn parse_select_item_for_data_load( fn parse_stage_params(parser: &mut Parser) -> Result { let (mut url, mut storage_integration, mut endpoint) = (None, None, None); - let mut encryption: KeyValueOptions = KeyValueOptions { options: vec![] }; - let mut credentials: KeyValueOptions = KeyValueOptions { options: vec![] }; + let mut encryption: KeyValueOptions = KeyValueOptions { + options: vec![], + delimiter: KeyValueOptionsDelimiter::Space, + }; + let mut credentials: KeyValueOptions = KeyValueOptions { + options: vec![], + delimiter: KeyValueOptionsDelimiter::Space, + }; // URL if parser.parse_keyword(Keyword::URL) { @@ -1141,7 +1161,8 @@ fn parse_stage_params(parser: &mut Parser) -> Result Result { parser.advance_token(); if set { - let option = parse_option(parser, key)?; + let option = parser.parse_key_value_option(key)?; options.push(option); } else { options.push(KeyValueOption { @@ -1207,63 +1229,6 @@ fn parse_session_options( } } -/// Parses options provided within parentheses like: -/// ( ENABLE = { TRUE | FALSE } -/// [ AUTO_REFRESH = { TRUE | FALSE } ] -/// [ REFRESH_ON_CREATE = { TRUE | FALSE } ] -/// [ NOTIFICATION_INTEGRATION = '' ] ) -/// -fn parse_parentheses_options(parser: &mut Parser) -> Result, ParserError> { - let mut options: Vec = Vec::new(); - parser.expect_token(&Token::LParen)?; - loop { - match parser.next_token().token { - Token::RParen => break, - Token::Comma => continue, - Token::Word(key) => options.push(parse_option(parser, key)?), - _ => return parser.expected("another option or ')'", parser.peek_token()), - }; - } - Ok(options) -} - -/// Parses a `KEY = VALUE` construct based on the specified key -fn parse_option(parser: &mut Parser, key: Word) -> Result { - parser.expect_token(&Token::Eq)?; - if parser.parse_keyword(Keyword::TRUE) { - Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::BOOLEAN, - value: "TRUE".to_string(), - }) - } else if parser.parse_keyword(Keyword::FALSE) { - Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::BOOLEAN, - value: "FALSE".to_string(), - }) - } else { - match parser.next_token().token { - Token::SingleQuotedString(value) => Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::STRING, - value, - }), - Token::Word(word) => Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::ENUM, - value: word.value, - }), - Token::Number(n, _) => Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::NUMBER, - value: n, - }), - _ => parser.expected("expected option value", parser.peek_token()), - } - } -} - /// Parsing a property of identity or autoincrement column option /// Syntax: /// ```sql diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 8d5a55da0..5ac19d13c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -32,7 +32,12 @@ use recursion::RecursionCounter; use IsLateral::*; use IsOptional::*; -use crate::ast::helpers::stmt_create_table::{CreateTableBuilder, CreateTableConfiguration}; +use crate::ast::helpers::{ + key_value_options::{ + KeyValueOption, KeyValueOptionType, KeyValueOptions, KeyValueOptionsDelimiter, + }, + stmt_create_table::{CreateTableBuilder, CreateTableConfiguration}, +}; use crate::ast::Statement::CreatePolicy; use crate::ast::*; use crate::dialect::*; @@ -4680,6 +4685,8 @@ impl<'a> Parser<'a> { self.parse_create_macro(or_replace, temporary) } else if self.parse_keyword(Keyword::SECRET) { self.parse_create_secret(or_replace, temporary, persistent) + } else if self.parse_keyword(Keyword::USER) { + self.parse_create_user(or_replace) } else if or_replace { self.expected( "[EXTERNAL] TABLE or [MATERIALIZED] VIEW or FUNCTION after CREATE OR REPLACE", @@ -4714,6 +4721,32 @@ impl<'a> Parser<'a> { } } + fn parse_create_user(&mut self, or_replace: bool) -> Result { + let if_not_exists = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let name = self.parse_identifier()?; + let options = self.parse_key_value_options(false, &[Keyword::WITH, Keyword::TAG])?; + let with_tags = self.parse_keyword(Keyword::WITH); + let tags = if self.parse_keyword(Keyword::TAG) { + self.parse_key_value_options(true, &[])? + } else { + vec![] + }; + Ok(Statement::CreateUser(CreateUser { + or_replace, + if_not_exists, + name, + options: KeyValueOptions { + options, + delimiter: KeyValueOptionsDelimiter::Space, + }, + with_tags, + tags: KeyValueOptions { + options: tags, + delimiter: KeyValueOptionsDelimiter::Comma, + }, + })) + } + /// See [DuckDB Docs](https://duckdb.org/docs/sql/statements/create_secret.html) for more details. pub fn parse_create_secret( &mut self, @@ -16612,6 +16645,78 @@ impl<'a> Parser<'a> { pub(crate) fn in_column_definition_state(&self) -> bool { matches!(self.state, ColumnDefinition) } + + /// Parses options provided in key-value format. + /// + /// * `parenthesized` - true if the options are enclosed in parenthesis + /// * `end_words` - a list of keywords that any of them indicates the end of the options section + pub(crate) fn parse_key_value_options( + &mut self, + parenthesized: bool, + end_words: &[Keyword], + ) -> Result, ParserError> { + let mut options: Vec = Vec::new(); + if parenthesized { + self.expect_token(&Token::LParen)?; + } + loop { + match self.next_token().token { + Token::RParen => { + if parenthesized { + break; + } else { + return self.expected(" another option or EOF", self.peek_token()); + } + } + Token::EOF => break, + Token::Comma => continue, + Token::Word(w) if !end_words.contains(&w.keyword) => { + options.push(self.parse_key_value_option(w)?) + } + Token::Word(w) if end_words.contains(&w.keyword) => { + self.prev_token(); + break; + } + _ => return self.expected("another option, EOF, Comma or ')'", self.peek_token()), + }; + } + Ok(options) + } + + /// Parses a `KEY = VALUE` construct based on the specified key + pub(crate) fn parse_key_value_option( + &mut self, + key: Word, + ) -> Result { + self.expect_token(&Token::Eq)?; + match self.next_token().token { + Token::SingleQuotedString(value) => Ok(KeyValueOption { + option_name: key.value, + option_type: KeyValueOptionType::STRING, + value, + }), + Token::Word(word) + if word.keyword == Keyword::TRUE || word.keyword == Keyword::FALSE => + { + Ok(KeyValueOption { + option_name: key.value, + option_type: KeyValueOptionType::BOOLEAN, + value: word.value.to_uppercase(), + }) + } + Token::Word(word) => Ok(KeyValueOption { + option_name: key.value, + option_type: KeyValueOptionType::ENUM, + value: word.value, + }), + Token::Number(n, _) => Ok(KeyValueOption { + option_name: key.value, + option_type: KeyValueOptionType::NUMBER, + value: n, + }), + _ => self.expected("expected option value", self.peek_token()), + } + } } fn maybe_prefixed_expr(expr: Expr, prefix: Option) -> Expr { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5d8284a46..5ae12ff40 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -27,6 +27,8 @@ extern crate core; use helpers::attached_token::AttachedToken; use matches::assert_matches; +use sqlparser::ast::helpers::key_value_options::*; +use sqlparser::ast::helpers::key_value_options::{KeyValueOptions, KeyValueOptionsDelimiter}; use sqlparser::ast::SelectItem::UnnamedExpr; use sqlparser::ast::TableFactor::{Pivot, Unpivot}; use sqlparser::ast::*; @@ -16256,3 +16258,72 @@ fn parse_notnull() { // for unsupported dialects, parsing should stop at `NOT NULL` notnull_unsupported_dialects.expr_parses_to("NOT NULL NOTNULL", "NOT NULL"); } + +#[test] +fn parse_create_user() { + let create = verified_stmt("CREATE USER u1"); + match create { + Statement::CreateUser(stmt) => { + assert_eq!(stmt.name, Ident::new("u1")); + } + _ => unreachable!(), + } + verified_stmt("CREATE OR REPLACE USER u1"); + verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1"); + verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret'"); + verified_stmt( + "CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=TRUE", + ); + verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=TRUE TYPE=SERVICE TAG (t1='v1')"); + let create = verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=TRUE TYPE=SERVICE WITH TAG (t1='v1', t2='v2')"); + match create { + Statement::CreateUser(stmt) => { + assert_eq!(stmt.name, Ident::new("u1")); + assert_eq!(stmt.or_replace, true); + assert_eq!(stmt.if_not_exists, true); + assert_eq!( + stmt.options, + KeyValueOptions { + delimiter: KeyValueOptionsDelimiter::Space, + options: vec![ + KeyValueOption { + option_name: "PASSWORD".to_string(), + value: "secret".to_string(), + option_type: KeyValueOptionType::STRING + }, + KeyValueOption { + option_name: "MUST_CHANGE_PASSWORD".to_string(), + value: "TRUE".to_string(), + option_type: KeyValueOptionType::BOOLEAN + }, + KeyValueOption { + option_name: "TYPE".to_string(), + value: "SERVICE".to_string(), + option_type: KeyValueOptionType::ENUM + }, + ], + }, + ); + assert_eq!(stmt.with_tags, true); + assert_eq!( + stmt.tags, + KeyValueOptions { + delimiter: KeyValueOptionsDelimiter::Comma, + options: vec![ + KeyValueOption { + option_name: "t1".to_string(), + value: "v1".to_string(), + option_type: KeyValueOptionType::STRING + }, + KeyValueOption { + option_name: "t2".to_string(), + value: "v2".to_string(), + option_type: KeyValueOptionType::STRING + }, + ] + } + ); + } + _ => unreachable!(), + } +} From f49c30feb6d5984f96cdc3a87dbf09b6b4b3ad04 Mon Sep 17 00:00:00 2001 From: carl <44021312+achristmascarl@users.noreply.github.com> Date: Wed, 23 Jul 2025 11:53:17 -0400 Subject: [PATCH 209/286] Postgres: Support parenthesized `SET` options for `ALTER TABLE` (#1947) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 13 +++++++++++++ src/ast/spans.rs | 3 +++ src/parser/mod.rs | 13 +++++++++---- tests/sqlparser_common.rs | 28 ++++++++++++++++++++++++++++ 4 files changed, 53 insertions(+), 4 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 51e057840..92936a6f0 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -351,6 +351,16 @@ pub enum AlterTableOperation { ValidateConstraint { name: Ident, }, + /// Arbitrary parenthesized `SET` options. + /// + /// Example: + /// ```sql + /// SET (scale_factor = 0.01, threshold = 500)` + /// ``` + /// [PostgreSQL](https://www.postgresql.org/docs/current/sql-altertable.html) + SetOptionsParens { + options: Vec, + }, } /// An `ALTER Policy` (`Statement::AlterPolicy`) operation @@ -791,6 +801,9 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::ValidateConstraint { name } => { write!(f, "VALIDATE CONSTRAINT {name}") } + AlterTableOperation::SetOptionsParens { options } => { + write!(f, "SET ({})", display_comma_separated(options)) + } } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 4deedca04..91523925e 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1202,6 +1202,9 @@ impl Spanned for AlterTableOperation { AlterTableOperation::Lock { .. } => Span::empty(), AlterTableOperation::ReplicaIdentity { .. } => Span::empty(), AlterTableOperation::ValidateConstraint { name } => name.span, + AlterTableOperation::SetOptionsParens { options } => { + union_spans(options.iter().map(|i| i.span())) + } } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 5ac19d13c..058423f1b 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9045,17 +9045,22 @@ impl<'a> Parser<'a> { let name = self.parse_identifier()?; AlterTableOperation::ValidateConstraint { name } } else { - let options: Vec = + let mut options = self.parse_options_with_keywords(&[Keyword::SET, Keyword::TBLPROPERTIES])?; if !options.is_empty() { AlterTableOperation::SetTblProperties { table_properties: options, } } else { - return self.expected( - "ADD, RENAME, PARTITION, SWAP, DROP, REPLICA IDENTITY, or SET TBLPROPERTIES after ALTER TABLE", + options = self.parse_options(Keyword::SET)?; + if !options.is_empty() { + AlterTableOperation::SetOptionsParens { options } + } else { + return self.expected( + "ADD, RENAME, PARTITION, SWAP, DROP, REPLICA IDENTITY, SET, or SET TBLPROPERTIES after ALTER TABLE", self.peek_token(), - ); + ); + } } }; Ok(operation) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5ae12ff40..074d9eab6 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4728,6 +4728,34 @@ fn parse_alter_table() { } _ => unreachable!(), } + + let set_storage_parameters = "ALTER TABLE tab SET (autovacuum_vacuum_scale_factor = 0.01, autovacuum_vacuum_threshold = 500)"; + match alter_table_op(verified_stmt(set_storage_parameters)) { + AlterTableOperation::SetOptionsParens { options } => { + assert_eq!( + options, + [ + SqlOption::KeyValue { + key: Ident { + value: "autovacuum_vacuum_scale_factor".to_string(), + quote_style: None, + span: Span::empty(), + }, + value: Expr::Value(test_utils::number("0.01").with_empty_span()), + }, + SqlOption::KeyValue { + key: Ident { + value: "autovacuum_vacuum_threshold".to_string(), + quote_style: None, + span: Span::empty(), + }, + value: Expr::Value(test_utils::number("500").with_empty_span()), + } + ], + ); + } + _ => unreachable!(), + } } #[test] From 7558d35c84d14f2b91b170a5b1eec8528b4965bb Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 23 Jul 2025 18:59:20 +0300 Subject: [PATCH 210/286] Snowflake: Support IDENTIFIER for GRANT ROLE (#1957) --- src/ast/mod.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_common.rs | 2 ++ tests/sqlparser_snowflake.rs | 3 +++ 4 files changed, 7 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index e16463495..d83591364 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6654,7 +6654,7 @@ pub enum Action { Replicate, ResolveAll, Role { - role: Ident, + role: ObjectName, }, Select { columns: Option>, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 058423f1b..a876d90e5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -14357,7 +14357,7 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::REPLICATE) { Ok(Action::Replicate) } else if self.parse_keyword(Keyword::ROLE) { - let role = self.parse_identifier()?; + let role = self.parse_object_name(false)?; Ok(Action::Role { role }) } else if self.parse_keyword(Keyword::SELECT) { Ok(Action::Select { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 074d9eab6..add54a71a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9549,6 +9549,8 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON FUTURE SEQUENCES IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT USAGE ON PROCEDURE db1.sc1.foo(INT) TO ROLE role1"); verified_stmt("GRANT USAGE ON FUNCTION db1.sc1.foo(INT) TO ROLE role1"); + verified_stmt("GRANT ROLE role1 TO ROLE role2"); + verified_stmt("GRANT ROLE role1 TO USER user"); } #[test] diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index a7a633152..daa711d4c 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4500,4 +4500,7 @@ fn test_snowflake_identifier_function() { .is_err(), true ); + + snowflake().verified_stmt("GRANT ROLE IDENTIFIER('AAA') TO USER IDENTIFIER('AAA')"); + snowflake().verified_stmt("REVOKE ROLE IDENTIFIER('AAA') FROM USER IDENTIFIER('AAA')"); } From 40b187f32a4bc4b8a5fddd5c9cedc96c6c66b5af Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 23 Jul 2025 19:44:00 +0300 Subject: [PATCH 211/286] Snowflake: Numeric prefix for stage name part (#1966) --- src/dialect/snowflake.rs | 1 + tests/sqlparser_snowflake.rs | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 9786aba0d..26432b3f0 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -841,6 +841,7 @@ pub fn parse_stage_name_identifier(parser: &mut Parser) -> Result ident.push('%'), Token::Div => ident.push('/'), Token::Plus => ident.push('+'), + Token::Number(n, _) => ident.push_str(n), Token::Word(w) => ident.push_str(&w.to_string()), _ => return parser.expected("stage name identifier", parser.peek_token()), } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index daa711d4c..7edb0d069 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -2628,7 +2628,7 @@ fn test_snowflake_copy_into() { } // Test for non-ident characters in stage names - let sql = "COPY INTO a.b FROM @namespace.stage_name/x@x~x%x+"; + let sql = "COPY INTO a.b FROM @namespace.stage_name/x@x~x%x+/20250723_data"; assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { into, from_obj, .. } => { @@ -2640,7 +2640,7 @@ fn test_snowflake_copy_into() { from_obj, Some(ObjectName::from(vec![ Ident::new("@namespace"), - Ident::new("stage_name/x@x~x%x+") + Ident::new("stage_name/x@x~x%x+/20250723_data") ])) ) } From 6a5ef48921d8eea8cbd34c7b2828f4dae09cd209 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 24 Jul 2025 20:54:17 +0300 Subject: [PATCH 212/286] Snowflake: Support `GRANT CREATE SCHEMA` `GRANT .. ON ALL FUNCTIONS IN SCHEMA` (#1964) --- src/ast/mod.rs | 11 +++++++++++ src/parser/mod.rs | 11 +++++++++++ tests/sqlparser_common.rs | 2 ++ 3 files changed, 24 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d83591364..18c80aa01 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -6771,6 +6771,7 @@ pub enum ActionCreateObjectType { OrganiationListing, ReplicationGroup, Role, + Schema, Share, User, Warehouse, @@ -6792,6 +6793,7 @@ impl fmt::Display for ActionCreateObjectType { ActionCreateObjectType::OrganiationListing => write!(f, "ORGANIZATION LISTING"), ActionCreateObjectType::ReplicationGroup => write!(f, "REPLICATION GROUP"), ActionCreateObjectType::Role => write!(f, "ROLE"), + ActionCreateObjectType::Schema => write!(f, "SCHEMA"), ActionCreateObjectType::Share => write!(f, "SHARE"), ActionCreateObjectType::User => write!(f, "USER"), ActionCreateObjectType::Warehouse => write!(f, "WAREHOUSE"), @@ -7029,6 +7031,8 @@ pub enum GrantObjects { AllMaterializedViewsInSchema { schemas: Vec }, /// Grant privileges on `ALL EXTERNAL TABLES IN SCHEMA [, ...]` AllExternalTablesInSchema { schemas: Vec }, + /// Grant privileges on `ALL FUNCTIONS IN SCHEMA [, ...]` + AllFunctionsInSchema { schemas: Vec }, /// Grant privileges on `FUTURE SCHEMAS IN DATABASE [, ...]` FutureSchemasInDatabase { databases: Vec }, /// Grant privileges on `FUTURE TABLES IN SCHEMA [, ...]` @@ -7149,6 +7153,13 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::AllFunctionsInSchema { schemas } => { + write!( + f, + "ALL FUNCTIONS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } GrantObjects::FutureSchemasInDatabase { databases } => { write!( f, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a876d90e5..bf0380a9d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -14108,6 +14108,15 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllMaterializedViewsInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::FUNCTIONS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllFunctionsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[ Keyword::FUTURE, Keyword::SCHEMAS, @@ -14414,6 +14423,8 @@ impl<'a> Parser<'a> { Some(ActionCreateObjectType::Integration) } else if self.parse_keyword(Keyword::ROLE) { Some(ActionCreateObjectType::Role) + } else if self.parse_keyword(Keyword::SCHEMA) { + Some(ActionCreateObjectType::Schema) } else if self.parse_keyword(Keyword::SHARE) { Some(ActionCreateObjectType::Share) } else if self.parse_keyword(Keyword::USER) { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index add54a71a..3a14028f0 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9528,6 +9528,7 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON ALL VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT SELECT ON ALL MATERIALIZED VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT SELECT ON ALL EXTERNAL TABLES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT USAGE ON ALL FUNCTIONS IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO a:b"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO GROUP group1"); verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST"); @@ -9551,6 +9552,7 @@ fn parse_grant() { verified_stmt("GRANT USAGE ON FUNCTION db1.sc1.foo(INT) TO ROLE role1"); verified_stmt("GRANT ROLE role1 TO ROLE role2"); verified_stmt("GRANT ROLE role1 TO USER user"); + verified_stmt("GRANT CREATE SCHEMA ON DATABASE db1 TO ROLE role1"); } #[test] From 145922affe7cb9d87f62609933121f3af9d03d06 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 24 Jul 2025 21:45:51 +0300 Subject: [PATCH 213/286] Snowflake: DROP STREAM (#1973) --- src/ast/mod.rs | 2 ++ src/keywords.rs | 1 + src/parser/mod.rs | 2 ++ tests/sqlparser_common.rs | 18 ++++++++++++++++++ 4 files changed, 23 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 18c80aa01..e28c3739a 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -7835,6 +7835,7 @@ pub enum ObjectType { Stage, Type, User, + Stream, } impl fmt::Display for ObjectType { @@ -7851,6 +7852,7 @@ impl fmt::Display for ObjectType { ObjectType::Stage => "STAGE", ObjectType::Type => "TYPE", ObjectType::User => "USER", + ObjectType::Stream => "STREAM", }) } } diff --git a/src/keywords.rs b/src/keywords.rs index 7781939bc..509c1f96c 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -870,6 +870,7 @@ define_keywords!( STORAGE_SERIALIZATION_POLICY, STORED, STRAIGHT_JOIN, + STREAM, STRICT, STRING, STRUCT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index bf0380a9d..d35d7880f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -6325,6 +6325,8 @@ impl<'a> Parser<'a> { ObjectType::Type } else if self.parse_keyword(Keyword::USER) { ObjectType::User + } else if self.parse_keyword(Keyword::STREAM) { + ObjectType::Stream } else if self.parse_keyword(Keyword::FUNCTION) { return self.parse_drop_function(); } else if self.parse_keyword(Keyword::POLICY) { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 3a14028f0..e6548d3e0 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16359,3 +16359,21 @@ fn parse_create_user() { _ => unreachable!(), } } + +#[test] +fn parse_drop_stream() { + let sql = "DROP STREAM s1"; + match verified_stmt(sql) { + Statement::Drop { + names, object_type, .. + } => { + assert_eq!( + vec!["s1"], + names.iter().map(ToString::to_string).collect::>() + ); + assert_eq!(ObjectType::Stream, object_type); + } + _ => unreachable!(), + } + verified_stmt("DROP STREAM IF EXISTS s1"); +} From 865c191a535ddb3de53e01d1343511a46c77a3db Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Fri, 25 Jul 2025 22:56:12 +0800 Subject: [PATCH 214/286] fix: begin statement for bigquery (#1975) --- src/dialect/bigquery.rs | 8 ++++++++ tests/sqlparser_bigquery.rs | 10 ++++++++++ 2 files changed, 18 insertions(+) diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index d53c9db05..27fd3cca3 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -19,6 +19,7 @@ use crate::ast::Statement; use crate::dialect::Dialect; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; +use crate::tokenizer::Token; /// These keywords are disallowed as column identifiers. Such that /// `SELECT 5 AS FROM T` is rejected by BigQuery. @@ -47,6 +48,13 @@ pub struct BigQueryDialect; impl Dialect for BigQueryDialect { fn parse_statement(&self, parser: &mut Parser) -> Option> { if parser.parse_keyword(Keyword::BEGIN) { + if parser.peek_keyword(Keyword::TRANSACTION) + || parser.peek_token_ref().token == Token::SemiColon + || parser.peek_token_ref().token == Token::EOF + { + parser.prev_token(); + return None; + } return Some(parser.parse_begin_exception_end()); } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 2ba54d3e1..bc03011d8 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -2566,3 +2566,13 @@ fn test_struct_trailing_and_nested_bracket() { ) ); } + +#[test] +fn test_begin_transaction() { + bigquery().verified_stmt("BEGIN TRANSACTION"); +} + +#[test] +fn test_begin_statement() { + bigquery().verified_stmt("BEGIN"); +} From 5ec953bd78dac41690cfa3cf5219103e39b3e877 Mon Sep 17 00:00:00 2001 From: Tyler White <50381805+IndexSeek@users.noreply.github.com> Date: Sat, 26 Jul 2025 03:28:55 -0400 Subject: [PATCH 215/286] fix: update DuckDB and ClickHouse documentation links (#1978) --- src/ast/ddl.rs | 2 +- src/ast/mod.rs | 2 +- src/dialect/duckdb.rs | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 92936a6f0..231ab49ca 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -149,7 +149,7 @@ pub enum AlterTableOperation { }, /// `ATTACH PART|PARTITION ` /// Note: this is a ClickHouse-specific operation, please refer to - /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/alter/pakrtition#attach-partitionpart) + /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/alter/partition#attach-partitionpart) AttachPartition { // PART is not a short form of PARTITION, it's a separate keyword // which represents a physical file on disk and partition is a logical entity. diff --git a/src/ast/mod.rs b/src/ast/mod.rs index e28c3739a..7b401606e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -1144,7 +1144,7 @@ pub enum Expr { /// /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/functions#higher-order-functions---operator-and-lambdaparams-expr-function) /// [Databricks](https://docs.databricks.com/en/sql/language-manual/sql-ref-lambda-functions.html) - /// [DuckDb](https://duckdb.org/docs/sql/functions/lambda.html) + /// [DuckDB](https://duckdb.org/docs/stable/sql/functions/lambda) Lambda(LambdaFunction), /// Checks membership of a value in a JSON array MemberOf(MemberOf), diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index aee7ee935..37cf7d369 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -65,7 +65,7 @@ impl Dialect for DuckDbDialect { true } - /// See + /// See fn supports_lambda_functions(&self) -> bool { true } From 97a5b61a733657d2fb1dd2e3d28f8ddee050d929 Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Mon, 28 Jul 2025 21:17:51 +0800 Subject: [PATCH 216/286] feat: support export data for bigquery (#1976) --- src/ast/mod.rs | 38 ++++++ src/ast/spans.rs | 13 +- src/parser/mod.rs | 28 +++++ tests/sqlparser_bigquery.rs | 232 +++++++++++++++++++++++++++++++++++- 4 files changed, 309 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 7b401606e..00e7f86e5 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4355,6 +4355,15 @@ pub enum Statement { /// /// See [ReturnStatement] Return(ReturnStatement), + /// Export data statement + /// + /// Example: + /// ```sql + /// EXPORT DATA OPTIONS(uri='gs://bucket/folder/*', format='PARQUET', overwrite=true) AS + /// SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10 + /// ``` + /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/export-statements) + ExportData(ExportData), /// ```sql /// CREATE [OR REPLACE] USER [IF NOT EXISTS] /// ``` @@ -6198,6 +6207,7 @@ impl fmt::Display for Statement { Statement::Return(r) => write!(f, "{r}"), Statement::List(command) => write!(f, "LIST {command}"), Statement::Remove(command) => write!(f, "REMOVE {command}"), + Statement::ExportData(e) => write!(f, "{e}"), Statement::CreateUser(s) => write!(f, "{s}"), } } @@ -10144,6 +10154,34 @@ impl fmt::Display for MemberOf { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ExportData { + pub options: Vec, + pub query: Box, + pub connection: Option, +} + +impl fmt::Display for ExportData { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + if let Some(connection) = &self.connection { + write!( + f, + "EXPORT DATA WITH CONNECTION {connection} OPTIONS({}) AS {}", + display_comma_separated(&self.options), + self.query + ) + } else { + write!( + f, + "EXPORT DATA OPTIONS({}) AS {}", + display_comma_separated(&self.options), + self.query + ) + } + } +} /// Creates a user /// /// Syntax: diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 91523925e..7f96465b6 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -15,7 +15,7 @@ // specific language governing permissions and limitations // under the License. -use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions}; +use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions, ExportData}; use core::iter; use crate::tokenizer::Span; @@ -531,6 +531,17 @@ impl Spanned for Statement { Statement::Print { .. } => Span::empty(), Statement::Return { .. } => Span::empty(), Statement::List(..) | Statement::Remove(..) => Span::empty(), + Statement::ExportData(ExportData { + options, + query, + connection, + }) => union_spans( + options + .iter() + .map(|i| i.span()) + .chain(core::iter::once(query.span())) + .chain(connection.iter().map(|i| i.span())), + ), Statement::CreateUser(..) => Span::empty(), } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d35d7880f..5ea57f6f8 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -645,6 +645,10 @@ impl<'a> Parser<'a> { Keyword::COMMENT if self.dialect.supports_comment_on() => self.parse_comment(), Keyword::PRINT => self.parse_print(), Keyword::RETURN => self.parse_return(), + Keyword::EXPORT => { + self.prev_token(); + self.parse_export_data() + } _ => self.expected("an SQL statement", next_token), }, Token::LParen => { @@ -16523,6 +16527,30 @@ impl<'a> Parser<'a> { } } + /// /// Parse a `EXPORT DATA` statement. + /// + /// See [Statement::ExportData] + fn parse_export_data(&mut self) -> Result { + self.expect_keywords(&[Keyword::EXPORT, Keyword::DATA])?; + + let connection = if self.parse_keywords(&[Keyword::WITH, Keyword::CONNECTION]) { + Some(self.parse_object_name(false)?) + } else { + None + }; + self.expect_keyword(Keyword::OPTIONS)?; + self.expect_token(&Token::LParen)?; + let options = self.parse_comma_separated(|p| p.parse_sql_option())?; + self.expect_token(&Token::RParen)?; + self.expect_keyword(Keyword::AS)?; + let query = self.parse_query()?; + Ok(Statement::ExportData(ExportData { + options, + query, + connection, + })) + } + /// Consume the parser and return its underlying token buffer pub fn into_tokens(self) -> Vec { self.tokens diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index bc03011d8..10a356717 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -20,10 +20,12 @@ mod test_utils; use std::ops::Deref; +use sqlparser::ast::helpers::attached_token::AttachedToken; use sqlparser::ast::*; use sqlparser::dialect::{BigQueryDialect, GenericDialect}; +use sqlparser::keywords::Keyword; use sqlparser::parser::{ParserError, ParserOptions}; -use sqlparser::tokenizer::{Location, Span}; +use sqlparser::tokenizer::{Location, Span, Token, TokenWithSpan, Word}; use test_utils::*; #[test] @@ -2567,6 +2569,234 @@ fn test_struct_trailing_and_nested_bracket() { ); } +#[test] +fn test_export_data() { + let stmt = bigquery().verified_stmt(concat!( + "EXPORT DATA OPTIONS(", + "uri = 'gs://bucket/folder/*', ", + "format = 'PARQUET', ", + "overwrite = true", + ") AS ", + "SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10", + )); + assert_eq!( + stmt, + Statement::ExportData(ExportData { + options: vec![ + SqlOption::KeyValue { + key: Ident::new("uri"), + value: Expr::Value( + Value::SingleQuotedString("gs://bucket/folder/*".to_owned()) + .with_empty_span() + ), + }, + SqlOption::KeyValue { + key: Ident::new("format"), + value: Expr::Value( + Value::SingleQuotedString("PARQUET".to_owned()).with_empty_span() + ), + }, + SqlOption::KeyValue { + key: Ident::new("overwrite"), + value: Expr::Value(Value::Boolean(true).with_empty_span()), + }, + ], + connection: None, + query: Box::new(Query { + with: None, + body: Box::new(SetExpr::Select(Box::new(Select { + select_token: AttachedToken(TokenWithSpan::new( + Token::Word(Word { + value: "SELECT".to_string(), + quote_style: None, + keyword: Keyword::SELECT, + }), + Span::empty() + )), + distinct: None, + top: None, + top_before_distinct: false, + projection: vec![ + SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("field1"))), + SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("field2"))), + ], + exclude: None, + into: None, + from: vec![TableWithJoins { + relation: table_from_name(ObjectName::from(vec![ + Ident::new("mydataset"), + Ident::new("table1") + ])), + joins: vec![], + }], + lateral_views: vec![], + prewhere: None, + selection: None, + group_by: GroupByExpr::Expressions(vec![], vec![]), + cluster_by: vec![], + distribute_by: vec![], + sort_by: vec![], + having: None, + named_window: vec![], + qualify: None, + window_before_qualify: false, + value_table_mode: None, + connect_by: None, + flavor: SelectFlavor::Standard, + }))), + order_by: Some(OrderBy { + kind: OrderByKind::Expressions(vec![OrderByExpr { + expr: Expr::Identifier(Ident::new("field1")), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + },]), + interpolate: None, + }), + limit_clause: Some(LimitClause::LimitOffset { + limit: Some(Expr::Value(number("10").with_empty_span())), + offset: None, + limit_by: vec![], + }), + fetch: None, + locks: vec![], + for_clause: None, + settings: None, + format_clause: None, + pipe_operators: vec![], + }) + }) + ); + + let stmt = bigquery().verified_stmt(concat!( + "EXPORT DATA WITH CONNECTION myconnection.myproject.us OPTIONS(", + "uri = 'gs://bucket/folder/*', ", + "format = 'PARQUET', ", + "overwrite = true", + ") AS ", + "SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10", + )); + + assert_eq!( + stmt, + Statement::ExportData(ExportData { + options: vec![ + SqlOption::KeyValue { + key: Ident::new("uri"), + value: Expr::Value( + Value::SingleQuotedString("gs://bucket/folder/*".to_owned()) + .with_empty_span() + ), + }, + SqlOption::KeyValue { + key: Ident::new("format"), + value: Expr::Value( + Value::SingleQuotedString("PARQUET".to_owned()).with_empty_span() + ), + }, + SqlOption::KeyValue { + key: Ident::new("overwrite"), + value: Expr::Value(Value::Boolean(true).with_empty_span()), + }, + ], + connection: Some(ObjectName::from(vec![ + Ident::new("myconnection"), + Ident::new("myproject"), + Ident::new("us") + ])), + query: Box::new(Query { + with: None, + body: Box::new(SetExpr::Select(Box::new(Select { + select_token: AttachedToken(TokenWithSpan::new( + Token::Word(Word { + value: "SELECT".to_string(), + quote_style: None, + keyword: Keyword::SELECT, + }), + Span::empty() + )), + distinct: None, + top: None, + top_before_distinct: false, + projection: vec![ + SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("field1"))), + SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("field2"))), + ], + exclude: None, + into: None, + from: vec![TableWithJoins { + relation: table_from_name(ObjectName::from(vec![ + Ident::new("mydataset"), + Ident::new("table1") + ])), + joins: vec![], + }], + lateral_views: vec![], + prewhere: None, + selection: None, + group_by: GroupByExpr::Expressions(vec![], vec![]), + cluster_by: vec![], + distribute_by: vec![], + sort_by: vec![], + having: None, + named_window: vec![], + qualify: None, + window_before_qualify: false, + value_table_mode: None, + connect_by: None, + flavor: SelectFlavor::Standard, + }))), + order_by: Some(OrderBy { + kind: OrderByKind::Expressions(vec![OrderByExpr { + expr: Expr::Identifier(Ident::new("field1")), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + },]), + interpolate: None, + }), + limit_clause: Some(LimitClause::LimitOffset { + limit: Some(Expr::Value(number("10").with_empty_span())), + offset: None, + limit_by: vec![], + }), + fetch: None, + locks: vec![], + for_clause: None, + settings: None, + format_clause: None, + pipe_operators: vec![], + }) + }) + ); + + // at least one option (uri) is required + let err = bigquery() + .parse_sql_statements(concat!( + "EXPORT DATA OPTIONS() AS ", + "SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10", + )) + .unwrap_err(); + assert_eq!( + err.to_string(), + "sql parser error: Expected: identifier, found: )" + ); + + let err = bigquery() + .parse_sql_statements(concat!( + "EXPORT DATA AS SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10", + )) + .unwrap_err(); + assert_eq!( + err.to_string(), + "sql parser error: Expected: OPTIONS, found: AS" + ); +} + #[test] fn test_begin_transaction() { bigquery().verified_stmt("BEGIN TRANSACTION"); From bde269b56f5af5e16759403ae3efda03029616e8 Mon Sep 17 00:00:00 2001 From: etgarperets Date: Tue, 29 Jul 2025 13:37:04 +0300 Subject: [PATCH 217/286] Add ODBC escape syntax support for time expressions (#1953) --- src/ast/mod.rs | 58 ++++++++++--- src/ast/spans.rs | 4 +- src/parser/mod.rs | 56 +++++++++++-- tests/sqlparser_bigquery.rs | 87 +++++++++++--------- tests/sqlparser_common.rs | 150 ++++++++++++++++++++++------------ tests/sqlparser_databricks.rs | 7 +- tests/sqlparser_postgres.rs | 7 +- 7 files changed, 259 insertions(+), 110 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 00e7f86e5..42e4d3b04 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -1014,12 +1014,7 @@ pub enum Expr { /// A constant of form ` 'value'`. /// This can represent ANSI SQL `DATE`, `TIME`, and `TIMESTAMP` literals (such as `DATE '2020-01-01'`), /// as well as constants of other types (a non-standard PostgreSQL extension). - TypedString { - data_type: DataType, - /// The value of the constant. - /// Hint: you can unwrap the string value using `value.into_string()`. - value: ValueWithSpan, - }, + TypedString(TypedString), /// Scalar function call e.g. `LEFT(foo, 5)` Function(Function), /// `CASE [] WHEN THEN ... [ELSE ] END` @@ -1734,10 +1729,7 @@ impl fmt::Display for Expr { Expr::Nested(ast) => write!(f, "({ast})"), Expr::Value(v) => write!(f, "{v}"), Expr::Prefixed { prefix, value } => write!(f, "{prefix} {value}"), - Expr::TypedString { data_type, value } => { - write!(f, "{data_type}")?; - write!(f, " {value}") - } + Expr::TypedString(ts) => ts.fmt(f), Expr::Function(fun) => fun.fmt(f), Expr::Case { case_token: _, @@ -7450,6 +7442,52 @@ pub struct DropDomain { pub drop_behavior: Option, } +/// A constant of form ` 'value'`. +/// This can represent ANSI SQL `DATE`, `TIME`, and `TIMESTAMP` literals (such as `DATE '2020-01-01'`), +/// as well as constants of other types (a non-standard PostgreSQL extension). +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct TypedString { + pub data_type: DataType, + /// The value of the constant. + /// Hint: you can unwrap the string value using `value.into_string()`. + pub value: ValueWithSpan, + /// Flags whether this TypedString uses the [ODBC syntax]. + /// + /// Example: + /// ```sql + /// -- An ODBC date literal: + /// SELECT {d '2025-07-16'} + /// -- This is equivalent to the standard ANSI SQL literal: + /// SELECT DATE '2025-07-16' + /// + /// [ODBC syntax]: https://learn.microsoft.com/en-us/sql/odbc/reference/develop-app/date-time-and-timestamp-literals?view=sql-server-2017 + pub uses_odbc_syntax: bool, +} + +impl fmt::Display for TypedString { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let data_type = &self.data_type; + let value = &self.value; + match self.uses_odbc_syntax { + false => { + write!(f, "{data_type}")?; + write!(f, " {value}") + } + true => { + let prefix = match data_type { + DataType::Date => "d", + DataType::Time(..) => "t", + DataType::Timestamp(..) => "ts", + _ => "?", + }; + write!(f, "{{{prefix} {value}}}") + } + } + } +} + /// A function call #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 7f96465b6..46f7a9edf 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -15,7 +15,7 @@ // specific language governing permissions and limitations // under the License. -use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions, ExportData}; +use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions, ExportData, TypedString}; use core::iter; use crate::tokenizer::Span; @@ -1525,7 +1525,7 @@ impl Spanned for Expr { .union(&union_spans(collation.0.iter().map(|i| i.span()))), Expr::Nested(expr) => expr.span(), Expr::Value(value) => value.span(), - Expr::TypedString { value, .. } => value.span(), + Expr::TypedString(TypedString { value, .. }) => value.span(), Expr::Function(function) => function.span(), Expr::GroupingSets(vec) => { union_spans(vec.iter().flat_map(|i| i.iter().map(|k| k.span()))) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 5ea57f6f8..3b1db0f64 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1543,10 +1543,11 @@ impl<'a> Parser<'a> { // an unary negation `NOT ('a' LIKE 'b')`. To solve this, we don't accept the // `type 'string'` syntax for the custom data types at all. DataType::Custom(..) => parser_err!("dummy", loc), - data_type => Ok(Expr::TypedString { + data_type => Ok(Expr::TypedString(TypedString { data_type, value: parser.parse_value()?, - }), + uses_odbc_syntax: false, + })), } })?; @@ -1732,10 +1733,11 @@ impl<'a> Parser<'a> { } fn parse_geometric_type(&mut self, kind: GeometricTypeKind) -> Result { - Ok(Expr::TypedString { + Ok(Expr::TypedString(TypedString { data_type: DataType::GeometricType(kind), value: self.parse_value()?, - }) + uses_odbc_syntax: false, + })) } /// Try to parse an [Expr::CompoundFieldAccess] like `a.b.c` or `a.b[1].c`. @@ -2032,6 +2034,50 @@ impl<'a> Parser<'a> { }) } + /// Tries to parse the body of an [ODBC escaping sequence] + /// i.e. without the enclosing braces + /// Currently implemented: + /// Scalar Function Calls + /// Date, Time, and Timestamp Literals + /// See + fn maybe_parse_odbc_body(&mut self) -> Result, ParserError> { + // Attempt 1: Try to parse it as a function. + if let Some(expr) = self.maybe_parse_odbc_fn_body()? { + return Ok(Some(expr)); + } + // Attempt 2: Try to parse it as a Date, Time or Timestamp Literal + self.maybe_parse_odbc_body_datetime() + } + + /// Tries to parse the body of an [ODBC Date, Time, and Timestamp Literals] call. + /// + /// ```sql + /// {d '2025-07-17'} + /// {t '14:12:01'} + /// {ts '2025-07-17 14:12:01'} + /// ``` + /// + /// [ODBC Date, Time, and Timestamp Literals]: + /// https://learn.microsoft.com/en-us/sql/odbc/reference/develop-app/date-time-and-timestamp-literals?view=sql-server-2017 + fn maybe_parse_odbc_body_datetime(&mut self) -> Result, ParserError> { + self.maybe_parse(|p| { + let token = p.next_token().clone(); + let word_string = token.token.to_string(); + let data_type = match word_string.as_str() { + "t" => DataType::Time(None, TimezoneInfo::None), + "d" => DataType::Date, + "ts" => DataType::Timestamp(None, TimezoneInfo::None), + _ => return p.expected("ODBC datetime keyword (t, d, or ts)", token), + }; + let value = p.parse_value()?; + Ok(Expr::TypedString(TypedString { + data_type, + value, + uses_odbc_syntax: true, + })) + }) + } + /// Tries to parse the body of an [ODBC function] call. /// i.e. without the enclosing braces /// @@ -2786,7 +2832,7 @@ impl<'a> Parser<'a> { fn parse_lbrace_expr(&mut self) -> Result { let token = self.expect_token(&Token::LBrace)?; - if let Some(fn_expr) = self.maybe_parse_odbc_fn_body()? { + if let Some(fn_expr) = self.maybe_parse_odbc_body()? { self.expect_token(&Token::RBrace)?; return Ok(fn_expr); } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 10a356717..839ea6b8c 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -906,13 +906,14 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Datetime(None), value: ValueWithSpan { value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), span: Span::empty(), }, - }], + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Datetime(None), @@ -968,15 +969,16 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::JSON, value: ValueWithSpan { value: Value::SingleQuotedString( r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() ), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::JSON, @@ -1004,7 +1006,7 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString( @@ -1012,7 +1014,8 @@ fn parse_typed_struct_syntax_bigquery() { ), span: Span::empty(), }, - }], + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Timestamp(None, TimezoneInfo::None), @@ -1024,13 +1027,14 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Time(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("15:30:00".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Time(None, TimezoneInfo::None), @@ -1045,13 +1049,14 @@ fn parse_typed_struct_syntax_bigquery() { assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("1".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Numeric(ExactNumberInfo::None), @@ -1062,13 +1067,14 @@ fn parse_typed_struct_syntax_bigquery() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("1".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::BigNumeric(ExactNumberInfo::None), @@ -1239,13 +1245,14 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Datetime(None), value: ValueWithSpan { value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Datetime(None), @@ -1301,15 +1308,16 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::JSON, value: ValueWithSpan { value: Value::SingleQuotedString( r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() ), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::JSON, @@ -1337,15 +1345,16 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString( "2008-12-25 15:30:00 America/Los_Angeles".into() ), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Timestamp(None, TimezoneInfo::None), @@ -1357,13 +1366,14 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Time(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("15:30:00".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Time(None, TimezoneInfo::None), @@ -1378,13 +1388,14 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { assert_eq!(2, select.projection.len()); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("1".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::Numeric(ExactNumberInfo::None), @@ -1395,13 +1406,14 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { ); assert_eq!( &Expr::Struct { - values: vec![Expr::TypedString { + values: vec![Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("1".into()), span: Span::empty(), - } - }], + }, + uses_odbc_syntax: false + })], fields: vec![StructField { field_name: None, field_type: DataType::BigNumeric(ExactNumberInfo::None), @@ -2433,13 +2445,14 @@ fn test_triple_quote_typed_strings() { let expr = bigquery().verified_expr(r#"JSON """{"foo":"bar's"}""""#); assert_eq!( - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::JSON, value: ValueWithSpan { value: Value::TripleDoubleQuotedString(r#"{"foo":"bar's"}"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr ); } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index e6548d3e0..7dc6fa81a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -5914,13 +5914,14 @@ fn parse_literal_date() { let sql = "SELECT DATE '1999-01-01'"; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::Date, value: ValueWithSpan { value: Value::SingleQuotedString("1999-01-01".into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); } @@ -5930,13 +5931,14 @@ fn parse_literal_time() { let sql = "SELECT TIME '01:23:34'"; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::Time(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("01:23:34".into()), span: Span::empty(), }, - }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); } @@ -5946,13 +5948,14 @@ fn parse_literal_datetime() { let sql = "SELECT DATETIME '1999-01-01 01:23:34.45'"; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::Datetime(None), value: ValueWithSpan { value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), span: Span::empty(), }, - }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); } @@ -5962,13 +5965,14 @@ fn parse_literal_timestamp_without_time_zone() { let sql = "SELECT TIMESTAMP '1999-01-01 01:23:34'"; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("1999-01-01 01:23:34".into()), span: Span::empty(), }, - }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); @@ -5980,13 +5984,14 @@ fn parse_literal_timestamp_with_time_zone() { let sql = "SELECT TIMESTAMPTZ '1999-01-01 01:23:34Z'"; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::Tz), value: ValueWithSpan { value: Value::SingleQuotedString("1999-01-01 01:23:34Z".into()), span: Span::empty(), }, - }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); @@ -6556,7 +6561,7 @@ fn parse_json_keyword() { }'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::JSON, value: ValueWithSpan { value: Value::SingleQuotedString( @@ -6583,8 +6588,9 @@ fn parse_json_keyword() { .to_string() ), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false, + }), expr_from_projection(only(&select.projection)), ); } @@ -6593,17 +6599,23 @@ fn parse_json_keyword() { fn parse_typed_strings() { let expr = verified_expr(r#"JSON '{"foo":"bar"}'"#); assert_eq!( - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::JSON, value: ValueWithSpan { value: Value::SingleQuotedString(r#"{"foo":"bar"}"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr ); - if let Expr::TypedString { data_type, value } = expr { + if let Expr::TypedString(TypedString { + data_type, + value, + uses_odbc_syntax: false, + }) = expr + { assert_eq!(DataType::JSON, data_type); assert_eq!(r#"{"foo":"bar"}"#, value.into_string().unwrap()); } @@ -6614,13 +6626,14 @@ fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '0'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString(r#"0"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); verified_stmt("SELECT BIGNUMERIC '0'"); @@ -6628,13 +6641,14 @@ fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '123456'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString(r#"123456"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); verified_stmt("SELECT BIGNUMERIC '123456'"); @@ -6642,13 +6656,14 @@ fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '-3.14'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString(r#"-3.14"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); verified_stmt("SELECT BIGNUMERIC '-3.14'"); @@ -6656,13 +6671,14 @@ fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '-0.54321'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString(r#"-0.54321"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); verified_stmt("SELECT BIGNUMERIC '-0.54321'"); @@ -6670,13 +6686,14 @@ fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '1.23456e05'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString(r#"1.23456e05"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); verified_stmt("SELECT BIGNUMERIC '1.23456e05'"); @@ -6684,13 +6701,14 @@ fn parse_bignumeric_keyword() { let sql = r#"SELECT BIGNUMERIC '-9.876e-3'"#; let select = verified_only_select(sql); assert_eq!( - &Expr::TypedString { + &Expr::TypedString(TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString(r#"-9.876e-3"#.into()), span: Span::empty(), - } - }, + }, + uses_odbc_syntax: false + }), expr_from_projection(only(&select.projection)), ); verified_stmt("SELECT BIGNUMERIC '-9.876e-3'"); @@ -15015,83 +15033,90 @@ fn test_geometry_type() { let sql = "point '1,2'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Point), value: ValueWithSpan { value: Value::SingleQuotedString("1,2".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); let sql = "line '1,2,3,4'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Line), value: ValueWithSpan { value: Value::SingleQuotedString("1,2,3,4".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); let sql = "path '1,2,3,4'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::GeometricPath), value: ValueWithSpan { value: Value::SingleQuotedString("1,2,3,4".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); let sql = "box '1,2,3,4'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::GeometricBox), value: ValueWithSpan { value: Value::SingleQuotedString("1,2,3,4".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); let sql = "circle '1,2,3'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Circle), value: ValueWithSpan { value: Value::SingleQuotedString("1,2,3".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); let sql = "polygon '1,2,3,4'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Polygon), value: ValueWithSpan { value: Value::SingleQuotedString("1,2,3,4".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); let sql = "lseg '1,2,3,4'"; assert_eq!( all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::GeometricType(GeometricTypeKind::LineSegment), value: ValueWithSpan { value: Value::SingleQuotedString("1,2,3,4".to_string()), span: Span::empty(), }, - } + uses_odbc_syntax: false + }) ); } #[test] @@ -16291,6 +16316,31 @@ fn parse_notnull() { notnull_unsupported_dialects.expr_parses_to("NOT NULL NOTNULL", "NOT NULL"); } +#[test] +fn parse_odbc_time_date_timestamp() { + // Supported statements + let sql_d = "SELECT {d '2025-07-17'}, category_name FROM categories"; + let _ = all_dialects().verified_stmt(sql_d); + let sql_t = "SELECT {t '14:12:01'}, category_name FROM categories"; + let _ = all_dialects().verified_stmt(sql_t); + let sql_ts = "SELECT {ts '2025-07-17 14:12:01'}, category_name FROM categories"; + let _ = all_dialects().verified_stmt(sql_ts); + // Unsupported statement + let supports_dictionary = all_dialects_where(|d| d.supports_dictionary_syntax()); + let dictionary_unsupported = all_dialects_where(|d| !d.supports_dictionary_syntax()); + let sql = "SELECT {tt '14:12:01'} FROM foo"; + let res = supports_dictionary.parse_sql_statements(sql); + let res_dict = dictionary_unsupported.parse_sql_statements(sql); + assert_eq!( + ParserError::ParserError("Expected: :, found: '14:12:01'".to_string()), + res.unwrap_err() + ); + assert_eq!( + ParserError::ParserError("Expected: an expression, found: {".to_string()), + res_dict.unwrap_err() + ); +} + #[test] fn parse_create_user() { let create = verified_stmt("CREATE USER u1"); diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index a27e0699b..e01611b6f 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -327,13 +327,14 @@ fn data_type_timestamp_ntz() { // Literal assert_eq!( databricks().verified_expr("TIMESTAMP_NTZ '2025-03-29T18:52:00'"), - Expr::TypedString { + Expr::TypedString(TypedString { data_type: DataType::TimestampNtz, value: ValueWithSpan { value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()), span: Span::empty(), - } - } + }, + uses_odbc_syntax: false + }) ); // Cast diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index d163096c4..6ce4a4835 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -4723,7 +4723,7 @@ fn parse_dollar_quoted_string() { quote_style: None, span: Span::empty(), }, - } + }, ); assert_eq!( @@ -5296,13 +5296,14 @@ fn parse_at_time_zone() { // check precedence let expr = Expr::BinaryOp { left: Box::new(Expr::AtTimeZone { - timestamp: Box::new(Expr::TypedString { + timestamp: Box::new(Expr::TypedString(TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), value: ValueWithSpan { value: Value::SingleQuotedString("2001-09-28 01:00".to_string()), span: Span::empty(), }, - }), + uses_odbc_syntax: false, + })), time_zone: Box::new(Expr::Cast { kind: CastKind::DoubleColon, expr: Box::new(Expr::Value( From 15d8bfea62726464a3065765f5511af7fa1548f8 Mon Sep 17 00:00:00 2001 From: etgarperets Date: Tue, 29 Jul 2025 13:38:07 +0300 Subject: [PATCH 218/286] Add support for `SHOW CHARSET` (#1974) Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 33 +++++++++++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/parser/mod.rs | 12 ++++++++++++ tests/sqlparser_mysql.rs | 15 +++++++++++++++ 4 files changed, 61 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 42e4d3b04..751da66b3 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3696,6 +3696,12 @@ pub enum Statement { history: bool, show_options: ShowStatementOptions, }, + // ```sql + // SHOW {CHARACTER SET | CHARSET} + // ``` + // [MySQL]: + // + ShowCharset(ShowCharset), /// ```sql /// SHOW OBJECTS LIKE 'line%' IN mydb.public /// ``` @@ -5680,6 +5686,7 @@ impl fmt::Display for Statement { } Ok(()) } + Statement::ShowCharset(show_stm) => show_stm.fmt(f), Statement::StartTransaction { modes, begin: syntax_begin, @@ -9859,6 +9866,32 @@ impl fmt::Display for ShowStatementIn { } } +/// A Show Charset statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ShowCharset { + /// The statement can be written as `SHOW CHARSET` or `SHOW CHARACTER SET` + /// true means CHARSET was used and false means CHARACTER SET was used + pub is_shorthand: bool, + pub filter: Option, +} + +impl fmt::Display for ShowCharset { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "SHOW")?; + if self.is_shorthand { + write!(f, " CHARSET")?; + } else { + write!(f, " CHARACTER SET")?; + } + if self.filter.is_some() { + write!(f, " {}", self.filter.as_ref().unwrap())?; + } + Ok(()) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 46f7a9edf..8e7011f54 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -477,6 +477,7 @@ impl Spanned for Statement { Statement::ShowColumns { .. } => Span::empty(), Statement::ShowTables { .. } => Span::empty(), Statement::ShowCollation { .. } => Span::empty(), + Statement::ShowCharset { .. } => Span::empty(), Statement::Use(u) => u.span(), Statement::StartTransaction { .. } => Span::empty(), Statement::Comment { .. } => Span::empty(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 3b1db0f64..0ab0ccf46 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12595,6 +12595,10 @@ impl<'a> Parser<'a> { self.parse_show_databases(terse) } else if self.parse_keyword(Keyword::SCHEMAS) { self.parse_show_schemas(terse) + } else if self.parse_keywords(&[Keyword::CHARACTER, Keyword::SET]) { + self.parse_show_charset(false) + } else if self.parse_keyword(Keyword::CHARSET) { + self.parse_show_charset(true) } else { Ok(Statement::ShowVariable { variable: self.parse_identifiers()?, @@ -12602,6 +12606,14 @@ impl<'a> Parser<'a> { } } + fn parse_show_charset(&mut self, is_shorthand: bool) -> Result { + // parse one of keywords + Ok(Statement::ShowCharset(ShowCharset { + is_shorthand, + filter: self.parse_show_statement_filter()?, + })) + } + fn parse_show_databases(&mut self, terse: bool) -> Result { let history = self.parse_keyword(Keyword::HISTORY); let show_options = self.parse_show_stmt_options()?; diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 9068ed9c5..82b45740d 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -4143,3 +4143,18 @@ fn parse_json_member_of() { _ => panic!("Unexpected statement {stmt}"), } } + +#[test] +fn parse_show_charset() { + let res = mysql().verified_stmt("SHOW CHARACTER SET"); + assert_eq!( + res, + Statement::ShowCharset(ShowCharset { + is_shorthand: false, + filter: None + }) + ); + mysql().verified_stmt("SHOW CHARACTER SET LIKE 'utf8mb4%'"); + mysql().verified_stmt("SHOW CHARSET WHERE charset = 'utf8mb4%'"); + mysql().verified_stmt("SHOW CHARSET LIKE 'utf8mb4%'"); +} From 91273703d44802fe5ca92e850357687777cd50c0 Mon Sep 17 00:00:00 2001 From: etgarperets Date: Wed, 30 Jul 2025 13:24:16 +0300 Subject: [PATCH 219/286] Snowflake: Support `CREATE VIEW myview IF NOT EXISTS` (#1961) --- src/ast/mod.rs | 25 ++++++++++++++++++++++--- src/ast/spans.rs | 1 + src/parser/mod.rs | 12 +++++++++--- tests/sqlparser_common.rs | 24 ++++++++++++++++++++++++ 4 files changed, 56 insertions(+), 6 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 751da66b3..805e5f42b 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3255,6 +3255,17 @@ pub enum Statement { materialized: bool, /// View name name: ObjectName, + /// If `if_not_exists` is true, this flag is set to true if the view name comes before the `IF NOT EXISTS` clause. + /// Example: + /// ```sql + /// CREATE VIEW myview IF NOT EXISTS AS SELECT 1` + /// ``` + /// Otherwise, the flag is set to false if the view name comes after the clause + /// Example: + /// ```sql + /// CREATE VIEW IF NOT EXISTS myview AS SELECT 1` + /// ``` + name_before_not_exists: bool, columns: Vec, query: Box, options: CreateTableOptions, @@ -4994,6 +5005,7 @@ impl fmt::Display for Statement { temporary, to, params, + name_before_not_exists, } => { write!( f, @@ -5006,11 +5018,18 @@ impl fmt::Display for Statement { } write!( f, - "{materialized}{temporary}VIEW {if_not_exists}{name}{to}", + "{materialized}{temporary}VIEW {if_not_and_name}{to}", + if_not_and_name = if *if_not_exists { + if *name_before_not_exists { + format!("{name} IF NOT EXISTS") + } else { + format!("IF NOT EXISTS {name}") + } + } else { + format!("{name}") + }, materialized = if *materialized { "MATERIALIZED " } else { "" }, - name = name, temporary = if *temporary { "TEMPORARY " } else { "" }, - if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, to = to .as_ref() .map(|to| format!(" TO {to}")) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 8e7011f54..00313c0e8 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -400,6 +400,7 @@ impl Spanned for Statement { if_not_exists: _, temporary: _, to, + name_before_not_exists: _, params: _, } => union_spans( core::iter::once(name.span()) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0ab0ccf46..4d2aa2027 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5818,12 +5818,17 @@ impl<'a> Parser<'a> { ) -> Result { let materialized = self.parse_keyword(Keyword::MATERIALIZED); self.expect_keyword_is(Keyword::VIEW)?; - let if_not_exists = dialect_of!(self is BigQueryDialect|SQLiteDialect|GenericDialect) + let allow_unquoted_hyphen = dialect_of!(self is BigQueryDialect); + // Tries to parse IF NOT EXISTS either before name or after name + // Name before IF NOT EXISTS is supported by snowflake but undocumented + let if_not_exists_first = + self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let name = self.parse_object_name(allow_unquoted_hyphen)?; + let name_before_not_exists = !if_not_exists_first && self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let if_not_exists = if_not_exists_first || name_before_not_exists; // Many dialects support `OR ALTER` right after `CREATE`, but we don't (yet). // ANSI SQL and Postgres support RECURSIVE here, but we don't support it either. - let allow_unquoted_hyphen = dialect_of!(self is BigQueryDialect); - let name = self.parse_object_name(allow_unquoted_hyphen)?; let columns = self.parse_view_columns()?; let mut options = CreateTableOptions::None; let with_options = self.parse_options(Keyword::WITH)?; @@ -5890,6 +5895,7 @@ impl<'a> Parser<'a> { temporary, to, params: create_view_params, + name_before_not_exists, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 7dc6fa81a..ec7eec7d4 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8058,6 +8058,7 @@ fn parse_create_view() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -8126,6 +8127,7 @@ fn parse_create_view_with_columns() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); @@ -8175,6 +8177,7 @@ fn parse_create_view_temporary() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -8214,6 +8217,7 @@ fn parse_create_or_replace_view() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); @@ -8257,6 +8261,7 @@ fn parse_create_or_replace_materialized_view() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); @@ -8296,6 +8301,7 @@ fn parse_create_materialized_view() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -8335,6 +8341,7 @@ fn parse_create_materialized_view_with_cluster_by() { temporary, to, params, + name_before_not_exists: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -16427,3 +16434,20 @@ fn parse_drop_stream() { } verified_stmt("DROP STREAM IF EXISTS s1"); } + +#[test] +fn parse_create_view_if_not_exists() { + // Name after IF NOT EXISTS + let sql: &'static str = "CREATE VIEW IF NOT EXISTS v AS SELECT 1"; + let _ = all_dialects().verified_stmt(sql); + // Name before IF NOT EXISTS + let sql = "CREATE VIEW v IF NOT EXISTS AS SELECT 1"; + let _ = all_dialects().verified_stmt(sql); + // Name missing from query + let sql = "CREATE VIEW IF NOT EXISTS AS SELECT 1"; + let res = all_dialects().parse_sql_statements(sql); + assert_eq!( + ParserError::ParserError("Expected: AS, found: SELECT".to_string()), + res.unwrap_err() + ); +} From 85fa88137953259d3e55956fa9788547d0e41a6c Mon Sep 17 00:00:00 2001 From: "dependabot[bot]" <49699333+dependabot[bot]@users.noreply.github.com> Date: Wed, 30 Jul 2025 21:29:34 +0200 Subject: [PATCH 220/286] Update criterion requirement from 0.6 to 0.7 in /sqlparser_bench (#1981) Signed-off-by: dependabot[bot] Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> --- sqlparser_bench/Cargo.toml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sqlparser_bench/Cargo.toml b/sqlparser_bench/Cargo.toml index 01c59be75..4fb9af16e 100644 --- a/sqlparser_bench/Cargo.toml +++ b/sqlparser_bench/Cargo.toml @@ -26,7 +26,7 @@ edition = "2018" sqlparser = { path = "../" } [dev-dependencies] -criterion = "0.6" +criterion = "0.7" [[bench]] name = "sqlparser_bench" From 3d2db8c69b499ae304d3c66a411c37d4a005522b Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 31 Jul 2025 10:41:33 +0200 Subject: [PATCH 221/286] Snowflake: Improve support for reserved keywords for table factor (#1942) --- src/dialect/mod.rs | 14 +++--- src/dialect/snowflake.rs | 83 ++++++++++++++++++++++++++++++++++++ src/parser/mod.rs | 6 +-- tests/sqlparser_snowflake.rs | 12 ++++++ 4 files changed, 103 insertions(+), 12 deletions(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index c78b00033..34cb445cc 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -963,12 +963,6 @@ pub trait Dialect: Debug + Any { keywords::RESERVED_FOR_IDENTIFIER.contains(&kw) } - /// Returns reserved keywords when looking to parse a `TableFactor`. - /// See [Self::supports_from_trailing_commas] - fn get_reserved_keywords_for_table_factor(&self) -> &[Keyword] { - keywords::RESERVED_FOR_TABLE_FACTOR - } - /// Returns reserved keywords that may prefix a select item expression /// e.g. `SELECT CONNECT_BY_ROOT name FROM Tbl2` (Snowflake) fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] { @@ -1027,7 +1021,13 @@ pub trait Dialect: Debug + Any { explicit || self.is_column_alias(kw, parser) } - /// Returns true if the specified keyword should be parsed as a table identifier. + /// Returns true if the specified keyword should be parsed as a table factor identifier. + /// See [keywords::RESERVED_FOR_TABLE_FACTOR] + fn is_table_factor(&self, kw: &Keyword, _parser: &mut Parser) -> bool { + !keywords::RESERVED_FOR_TABLE_FACTOR.contains(kw) + } + + /// Returns true if the specified keyword should be parsed as a table factor alias. /// See [keywords::RESERVED_FOR_TABLE_ALIAS] fn is_table_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 26432b3f0..706d5198e 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -47,6 +47,82 @@ use super::keywords::RESERVED_FOR_IDENTIFIER; use sqlparser::ast::StorageSerializationPolicy; const RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR: [Keyword; 1] = [Keyword::CONNECT_BY_ROOT]; + +// See: +const RESERVED_KEYWORDS_FOR_TABLE_FACTOR: &[Keyword] = &[ + Keyword::ALL, + Keyword::ALTER, + Keyword::AND, + Keyword::ANY, + Keyword::AS, + Keyword::BETWEEN, + Keyword::BY, + Keyword::CHECK, + Keyword::COLUMN, + Keyword::CONNECT, + Keyword::CREATE, + Keyword::CROSS, + Keyword::CURRENT, + Keyword::DELETE, + Keyword::DISTINCT, + Keyword::DROP, + Keyword::ELSE, + Keyword::EXISTS, + Keyword::FOLLOWING, + Keyword::FOR, + Keyword::FROM, + Keyword::FULL, + Keyword::GRANT, + Keyword::GROUP, + Keyword::HAVING, + Keyword::ILIKE, + Keyword::IN, + Keyword::INCREMENT, + Keyword::INNER, + Keyword::INSERT, + Keyword::INTERSECT, + Keyword::INTO, + Keyword::IS, + Keyword::JOIN, + Keyword::LEFT, + Keyword::LIKE, + Keyword::MINUS, + Keyword::NATURAL, + Keyword::NOT, + Keyword::NULL, + Keyword::OF, + Keyword::ON, + Keyword::OR, + Keyword::ORDER, + Keyword::QUALIFY, + Keyword::REGEXP, + Keyword::REVOKE, + Keyword::RIGHT, + Keyword::RLIKE, + Keyword::ROW, + Keyword::ROWS, + Keyword::SAMPLE, + Keyword::SELECT, + Keyword::SET, + Keyword::SOME, + Keyword::START, + Keyword::TABLE, + Keyword::TABLESAMPLE, + Keyword::THEN, + Keyword::TO, + Keyword::TRIGGER, + Keyword::UNION, + Keyword::UNIQUE, + Keyword::UPDATE, + Keyword::USING, + Keyword::VALUES, + Keyword::WHEN, + Keyword::WHENEVER, + Keyword::WHERE, + Keyword::WINDOW, + Keyword::WITH, +]; + /// A [`Dialect`] for [Snowflake](https://www.snowflake.com/) #[derive(Debug, Default)] pub struct SnowflakeDialect; @@ -433,6 +509,13 @@ impl Dialect for SnowflakeDialect { } } + fn is_table_factor(&self, kw: &Keyword, parser: &mut Parser) -> bool { + match kw { + Keyword::LIMIT if peek_for_limit_options(parser) => false, + _ => !RESERVED_KEYWORDS_FOR_TABLE_FACTOR.contains(kw), + } + } + /// See: fn supports_timestamp_versioning(&self) -> bool { true diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 4d2aa2027..29e00b064 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4449,11 +4449,7 @@ impl<'a> Parser<'a> { self.parse_comma_separated_with_trailing_commas( Parser::parse_table_and_joins, trailing_commas, - |kw, _parser| { - self.dialect - .get_reserved_keywords_for_table_factor() - .contains(kw) - }, + |kw, parser| !self.dialect.is_table_factor(kw, parser), ) } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 7edb0d069..bd089ad7d 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3606,6 +3606,18 @@ fn test_sql_keywords_as_table_aliases() { snowflake().verified_stmt("SELECT * FROM tbl LIMIT $$$$"); } +#[test] +fn test_sql_keywords_as_table_factor() { + // LIMIT is a table factor, Snowflake does not reserve it + snowflake().one_statement_parses_to("SELECT * FROM tbl, LIMIT", "SELECT * FROM tbl, LIMIT"); + // LIMIT is not a table factor + snowflake().one_statement_parses_to("SELECT * FROM tbl, LIMIT 1", "SELECT * FROM tbl LIMIT 1"); + // ORDER is reserved + assert!(snowflake() + .parse_sql_statements("SELECT * FROM tbl, order") + .is_err()); +} + #[test] fn test_timetravel_at_before() { snowflake().verified_only_select("SELECT * FROM tbl AT(TIMESTAMP => '2024-12-15 00:00:00')"); From f5f51eb6f10012b4c4b742979b001872eaefdd36 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 1 Aug 2025 03:33:43 -0700 Subject: [PATCH 222/286] MySQL: Allow optional `SIGNED` suffix on integer data types (#1985) --- src/dialect/generic.rs | 4 ++++ src/dialect/mod.rs | 12 +++++++++++ src/dialect/mysql.rs | 4 ++++ src/parser/mod.rs | 18 ++++++++++++++++ tests/sqlparser_mysql.rs | 45 ++++++++++++++++++++++++++++++++++++++++ 5 files changed, 83 insertions(+) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index be2cc0076..de83d507e 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -183,4 +183,8 @@ impl Dialect for GenericDialect { fn supports_select_wildcard_exclude(&self) -> bool { true } + + fn supports_data_type_signed_suffix(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 34cb445cc..9003220c0 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1136,6 +1136,18 @@ pub trait Dialect: Debug + Any { fn supports_notnull_operator(&self) -> bool { false } + + /// Returns true if this dialect allows an optional `SIGNED` suffix after integer data types. + /// + /// Example: + /// ```sql + /// CREATE TABLE t (i INT(20) SIGNED); + /// ``` + /// + /// Note that this is canonicalized to `INT(20)`. + fn supports_data_type_signed_suffix(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index f7b5f574e..6cf24e14e 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -154,6 +154,10 @@ impl Dialect for MySqlDialect { fn supports_comma_separated_set_assignments(&self) -> bool { true } + + fn supports_data_type_signed_suffix(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 29e00b064..3bf036051 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9848,6 +9848,9 @@ impl<'a> Parser<'a> { if self.parse_keyword(Keyword::UNSIGNED) { Ok(DataType::TinyIntUnsigned(optional_precision?)) } else { + if dialect.supports_data_type_signed_suffix() { + let _ = self.parse_keyword(Keyword::SIGNED); + } Ok(DataType::TinyInt(optional_precision?)) } } @@ -9864,6 +9867,9 @@ impl<'a> Parser<'a> { if self.parse_keyword(Keyword::UNSIGNED) { Ok(DataType::SmallIntUnsigned(optional_precision?)) } else { + if dialect.supports_data_type_signed_suffix() { + let _ = self.parse_keyword(Keyword::SIGNED); + } Ok(DataType::SmallInt(optional_precision?)) } } @@ -9872,6 +9878,9 @@ impl<'a> Parser<'a> { if self.parse_keyword(Keyword::UNSIGNED) { Ok(DataType::MediumIntUnsigned(optional_precision?)) } else { + if dialect.supports_data_type_signed_suffix() { + let _ = self.parse_keyword(Keyword::SIGNED); + } Ok(DataType::MediumInt(optional_precision?)) } } @@ -9880,6 +9889,9 @@ impl<'a> Parser<'a> { if self.parse_keyword(Keyword::UNSIGNED) { Ok(DataType::IntUnsigned(optional_precision?)) } else { + if dialect.supports_data_type_signed_suffix() { + let _ = self.parse_keyword(Keyword::SIGNED); + } Ok(DataType::Int(optional_precision?)) } } @@ -9909,6 +9921,9 @@ impl<'a> Parser<'a> { if self.parse_keyword(Keyword::UNSIGNED) { Ok(DataType::IntegerUnsigned(optional_precision?)) } else { + if dialect.supports_data_type_signed_suffix() { + let _ = self.parse_keyword(Keyword::SIGNED); + } Ok(DataType::Integer(optional_precision?)) } } @@ -9917,6 +9932,9 @@ impl<'a> Parser<'a> { if self.parse_keyword(Keyword::UNSIGNED) { Ok(DataType::BigIntUnsigned(optional_precision?)) } else { + if dialect.supports_data_type_signed_suffix() { + let _ = self.parse_keyword(Keyword::SIGNED); + } Ok(DataType::BigInt(optional_precision?)) } } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 82b45740d..3ea0543f1 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1705,6 +1705,51 @@ fn parse_create_table_unsigned() { } } +#[test] +fn parse_signed_data_types() { + let sql = "CREATE TABLE foo (bar_tinyint TINYINT(3) SIGNED, bar_smallint SMALLINT(5) SIGNED, bar_mediumint MEDIUMINT(13) SIGNED, bar_int INT(11) SIGNED, bar_bigint BIGINT(20) SIGNED)"; + let canonical = "CREATE TABLE foo (bar_tinyint TINYINT(3), bar_smallint SMALLINT(5), bar_mediumint MEDIUMINT(13), bar_int INT(11), bar_bigint BIGINT(20))"; + match mysql().one_statement_parses_to(sql, canonical) { + Statement::CreateTable(CreateTable { name, columns, .. }) => { + assert_eq!(name.to_string(), "foo"); + assert_eq!( + vec![ + ColumnDef { + name: Ident::new("bar_tinyint"), + data_type: DataType::TinyInt(Some(3)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_smallint"), + data_type: DataType::SmallInt(Some(5)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_mediumint"), + data_type: DataType::MediumInt(Some(13)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_int"), + data_type: DataType::Int(Some(11)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_bigint"), + data_type: DataType::BigInt(Some(20)), + options: vec![], + }, + ], + columns + ); + } + _ => unreachable!(), + } + all_dialects_except(|d| d.supports_data_type_signed_suffix()) + .run_parser_method(sql, |p| p.parse_statement()) + .expect_err("SIGNED suffix should not be allowed"); +} + #[test] fn parse_simple_insert() { let sql = r"INSERT INTO tasks (title, priority) VALUES ('Test Some Inserts', 1), ('Test Entry 2', 2), ('Test Entry 3', 3)"; From 6932f4ad65f442c5b44b3445a76afd8b02dc0fdb Mon Sep 17 00:00:00 2001 From: xitep Date: Fri, 1 Aug 2025 12:40:33 +0200 Subject: [PATCH 223/286] Fix placeholder spans (#1979) --- src/ast/spans.rs | 23 +++++++++++++++++++++++ src/parser/mod.rs | 25 +++++++++++++++++++------ 2 files changed, 42 insertions(+), 6 deletions(-) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 00313c0e8..585836601 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2525,4 +2525,27 @@ pub mod tests { "CASE 1 WHEN 2 THEN 3 ELSE 4 END" ); } + + #[test] + fn test_placeholder_span() { + let sql = "\nSELECT\n :fooBar"; + let r = Parser::parse_sql(&GenericDialect, sql).unwrap(); + assert_eq!(1, r.len()); + match &r[0] { + Statement::Query(q) => { + let col = &q.body.as_select().unwrap().projection[0]; + match col { + SelectItem::UnnamedExpr(Expr::Value(ValueWithSpan { + value: Value::Placeholder(s), + span, + })) => { + assert_eq!(":fooBar", s); + assert_eq!(&Span::new((3, 3).into(), (3, 10).into()), span); + } + _ => panic!("expected unnamed expression; got {col:?}"), + } + } + stmt => panic!("expected query; got {stmt:?}"), + } + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 3bf036051..9b8fa17ac 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9636,16 +9636,21 @@ impl<'a> Parser<'a> { Token::HexStringLiteral(ref s) => ok_value(Value::HexStringLiteral(s.to_string())), Token::Placeholder(ref s) => ok_value(Value::Placeholder(s.to_string())), tok @ Token::Colon | tok @ Token::AtSign => { - // Not calling self.parse_identifier(false)? because only in placeholder we want to check numbers as idfentifies - // This because snowflake allows numbers as placeholders - let next_token = self.next_token(); + // 1. Not calling self.parse_identifier(false)? + // because only in placeholder we want to check + // numbers as idfentifies. This because snowflake + // allows numbers as placeholders + // 2. Not calling self.next_token() to enforce `tok` + // be followed immediately by a word/number, ie. + // without any whitespace in between + let next_token = self.next_token_no_skip().unwrap_or(&EOF_TOKEN).clone(); let ident = match next_token.token { Token::Word(w) => Ok(w.into_ident(next_token.span)), - Token::Number(w, false) => Ok(Ident::new(w)), + Token::Number(w, false) => Ok(Ident::with_span(next_token.span, w)), _ => self.expected("placeholder", next_token), }?; - let placeholder = tok.to_string() + &ident.value; - ok_value(Value::Placeholder(placeholder)) + Ok(Value::Placeholder(tok.to_string() + &ident.value) + .with_span(Span::new(span.start, ident.span.end))) } unexpected => self.expected( "a value", @@ -17600,4 +17605,12 @@ mod tests { canonical, ); } + + #[test] + fn test_placeholder_invalid_whitespace() { + for w in [" ", "/*invalid*/"] { + let sql = format!("\nSELECT\n :{w}fooBar"); + assert!(Parser::parse_sql(&GenericDialect, &sql).is_err()); + } + } } From ec0026d1362bb2d6a76184f2fb1a4717a65b7999 Mon Sep 17 00:00:00 2001 From: Artem Osipov <59066880+osipovartem@users.noreply.github.com> Date: Fri, 1 Aug 2025 14:56:21 +0300 Subject: [PATCH 224/286] Snowflake create database (#1939) --- src/ast/helpers/mod.rs | 1 + src/ast/helpers/stmt_create_database.rs | 324 ++++++++++++++++++++++++ src/ast/mod.rs | 147 ++++++++++- src/dialect/snowflake.rs | 144 +++++++++-- src/keywords.rs | 6 + src/parser/mod.rs | 24 ++ tests/sqlparser_common.rs | 3 + tests/sqlparser_snowflake.rs | 51 ++++ 8 files changed, 663 insertions(+), 37 deletions(-) create mode 100644 src/ast/helpers/stmt_create_database.rs diff --git a/src/ast/helpers/mod.rs b/src/ast/helpers/mod.rs index 55831220d..3efbcf7b0 100644 --- a/src/ast/helpers/mod.rs +++ b/src/ast/helpers/mod.rs @@ -16,5 +16,6 @@ // under the License. pub mod attached_token; pub mod key_value_options; +pub mod stmt_create_database; pub mod stmt_create_table; pub mod stmt_data_loading; diff --git a/src/ast/helpers/stmt_create_database.rs b/src/ast/helpers/stmt_create_database.rs new file mode 100644 index 000000000..94997bfa5 --- /dev/null +++ b/src/ast/helpers/stmt_create_database.rs @@ -0,0 +1,324 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + +#[cfg(not(feature = "std"))] +use alloc::{boxed::Box, format, string::String, vec, vec::Vec}; + +#[cfg(feature = "serde")] +use serde::{Deserialize, Serialize}; + +#[cfg(feature = "visitor")] +use sqlparser_derive::{Visit, VisitMut}; + +use crate::ast::{ + CatalogSyncNamespaceMode, ContactEntry, ObjectName, Statement, StorageSerializationPolicy, Tag, +}; +use crate::parser::ParserError; + +/// Builder for create database statement variant ([1]). +/// +/// This structure helps building and accessing a create database with more ease, without needing to: +/// - Match the enum itself a lot of times; or +/// - Moving a lot of variables around the code. +/// +/// # Example +/// ```rust +/// use sqlparser::ast::helpers::stmt_create_database::CreateDatabaseBuilder; +/// use sqlparser::ast::{ColumnDef, Ident, ObjectName}; +/// let builder = CreateDatabaseBuilder::new(ObjectName::from(vec![Ident::new("database_name")])) +/// .if_not_exists(true); +/// // You can access internal elements with ease +/// assert!(builder.if_not_exists); +/// // Convert to a statement +/// assert_eq!( +/// builder.build().to_string(), +/// "CREATE DATABASE IF NOT EXISTS database_name" +/// ) +/// ``` +/// +/// [1]: Statement::CreateDatabase +#[derive(Debug, Clone, PartialEq, Eq, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateDatabaseBuilder { + pub db_name: ObjectName, + pub if_not_exists: bool, + pub location: Option, + pub managed_location: Option, + pub or_replace: bool, + pub transient: bool, + pub clone: Option, + pub data_retention_time_in_days: Option, + pub max_data_extension_time_in_days: Option, + pub external_volume: Option, + pub catalog: Option, + pub replace_invalid_characters: Option, + pub default_ddl_collation: Option, + pub storage_serialization_policy: Option, + pub comment: Option, + pub catalog_sync: Option, + pub catalog_sync_namespace_mode: Option, + pub catalog_sync_namespace_flatten_delimiter: Option, + pub with_tags: Option>, + pub with_contacts: Option>, +} + +impl CreateDatabaseBuilder { + pub fn new(name: ObjectName) -> Self { + Self { + db_name: name, + if_not_exists: false, + location: None, + managed_location: None, + or_replace: false, + transient: false, + clone: None, + data_retention_time_in_days: None, + max_data_extension_time_in_days: None, + external_volume: None, + catalog: None, + replace_invalid_characters: None, + default_ddl_collation: None, + storage_serialization_policy: None, + comment: None, + catalog_sync: None, + catalog_sync_namespace_mode: None, + catalog_sync_namespace_flatten_delimiter: None, + with_tags: None, + with_contacts: None, + } + } + + pub fn location(mut self, location: Option) -> Self { + self.location = location; + self + } + + pub fn managed_location(mut self, managed_location: Option) -> Self { + self.managed_location = managed_location; + self + } + + pub fn or_replace(mut self, or_replace: bool) -> Self { + self.or_replace = or_replace; + self + } + + pub fn transient(mut self, transient: bool) -> Self { + self.transient = transient; + self + } + + pub fn if_not_exists(mut self, if_not_exists: bool) -> Self { + self.if_not_exists = if_not_exists; + self + } + + pub fn clone_clause(mut self, clone: Option) -> Self { + self.clone = clone; + self + } + + pub fn data_retention_time_in_days(mut self, data_retention_time_in_days: Option) -> Self { + self.data_retention_time_in_days = data_retention_time_in_days; + self + } + + pub fn max_data_extension_time_in_days( + mut self, + max_data_extension_time_in_days: Option, + ) -> Self { + self.max_data_extension_time_in_days = max_data_extension_time_in_days; + self + } + + pub fn external_volume(mut self, external_volume: Option) -> Self { + self.external_volume = external_volume; + self + } + + pub fn catalog(mut self, catalog: Option) -> Self { + self.catalog = catalog; + self + } + + pub fn replace_invalid_characters(mut self, replace_invalid_characters: Option) -> Self { + self.replace_invalid_characters = replace_invalid_characters; + self + } + + pub fn default_ddl_collation(mut self, default_ddl_collation: Option) -> Self { + self.default_ddl_collation = default_ddl_collation; + self + } + + pub fn storage_serialization_policy( + mut self, + storage_serialization_policy: Option, + ) -> Self { + self.storage_serialization_policy = storage_serialization_policy; + self + } + + pub fn comment(mut self, comment: Option) -> Self { + self.comment = comment; + self + } + + pub fn catalog_sync(mut self, catalog_sync: Option) -> Self { + self.catalog_sync = catalog_sync; + self + } + + pub fn catalog_sync_namespace_mode( + mut self, + catalog_sync_namespace_mode: Option, + ) -> Self { + self.catalog_sync_namespace_mode = catalog_sync_namespace_mode; + self + } + + pub fn catalog_sync_namespace_flatten_delimiter( + mut self, + catalog_sync_namespace_flatten_delimiter: Option, + ) -> Self { + self.catalog_sync_namespace_flatten_delimiter = catalog_sync_namespace_flatten_delimiter; + self + } + + pub fn with_tags(mut self, with_tags: Option>) -> Self { + self.with_tags = with_tags; + self + } + + pub fn with_contacts(mut self, with_contacts: Option>) -> Self { + self.with_contacts = with_contacts; + self + } + + pub fn build(self) -> Statement { + Statement::CreateDatabase { + db_name: self.db_name, + if_not_exists: self.if_not_exists, + managed_location: self.managed_location, + location: self.location, + or_replace: self.or_replace, + transient: self.transient, + clone: self.clone, + data_retention_time_in_days: self.data_retention_time_in_days, + max_data_extension_time_in_days: self.max_data_extension_time_in_days, + external_volume: self.external_volume, + catalog: self.catalog, + replace_invalid_characters: self.replace_invalid_characters, + default_ddl_collation: self.default_ddl_collation, + storage_serialization_policy: self.storage_serialization_policy, + comment: self.comment, + catalog_sync: self.catalog_sync, + catalog_sync_namespace_mode: self.catalog_sync_namespace_mode, + catalog_sync_namespace_flatten_delimiter: self.catalog_sync_namespace_flatten_delimiter, + with_tags: self.with_tags, + with_contacts: self.with_contacts, + } + } +} + +impl TryFrom for CreateDatabaseBuilder { + type Error = ParserError; + + fn try_from(stmt: Statement) -> Result { + match stmt { + Statement::CreateDatabase { + db_name, + if_not_exists, + location, + managed_location, + or_replace, + transient, + clone, + data_retention_time_in_days, + max_data_extension_time_in_days, + external_volume, + catalog, + replace_invalid_characters, + default_ddl_collation, + storage_serialization_policy, + comment, + catalog_sync, + catalog_sync_namespace_mode, + catalog_sync_namespace_flatten_delimiter, + with_tags, + with_contacts, + } => Ok(Self { + db_name, + if_not_exists, + location, + managed_location, + or_replace, + transient, + clone, + data_retention_time_in_days, + max_data_extension_time_in_days, + external_volume, + catalog, + replace_invalid_characters, + default_ddl_collation, + storage_serialization_policy, + comment, + catalog_sync, + catalog_sync_namespace_mode, + catalog_sync_namespace_flatten_delimiter, + with_tags, + with_contacts, + }), + _ => Err(ParserError::ParserError(format!( + "Expected create database statement, but received: {stmt}" + ))), + } + } +} + +#[cfg(test)] +mod tests { + use crate::ast::helpers::stmt_create_database::CreateDatabaseBuilder; + use crate::ast::{Ident, ObjectName, Statement}; + use crate::parser::ParserError; + + #[test] + pub fn test_from_valid_statement() { + let builder = CreateDatabaseBuilder::new(ObjectName::from(vec![Ident::new("db_name")])); + + let stmt = builder.clone().build(); + + assert_eq!(builder, CreateDatabaseBuilder::try_from(stmt).unwrap()); + } + + #[test] + pub fn test_from_invalid_statement() { + let stmt = Statement::Commit { + chain: false, + end: false, + modifier: None, + }; + + assert_eq!( + CreateDatabaseBuilder::try_from(stmt).unwrap_err(), + ParserError::ParserError( + "Expected create database statement, but received: COMMIT".to_owned() + ) + ); + } +} diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 805e5f42b..34112ac64 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3867,19 +3867,29 @@ pub enum Statement { /// ```sql /// CREATE DATABASE /// ``` + /// See: + /// CreateDatabase { db_name: ObjectName, if_not_exists: bool, location: Option, managed_location: Option, - /// Clones a database - /// - /// ```sql - /// CREATE DATABASE mydb CLONE otherdb - /// ``` - /// - /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-clone#databases-schemas) + or_replace: bool, + transient: bool, clone: Option, + data_retention_time_in_days: Option, + max_data_extension_time_in_days: Option, + external_volume: Option, + catalog: Option, + replace_invalid_characters: Option, + default_ddl_collation: Option, + storage_serialization_policy: Option, + comment: Option, + catalog_sync: Option, + catalog_sync_namespace_mode: Option, + catalog_sync_namespace_flatten_delimiter: Option, + with_tags: Option>, + with_contacts: Option>, }, /// ```sql /// CREATE FUNCTION @@ -4836,13 +4846,32 @@ impl fmt::Display for Statement { if_not_exists, location, managed_location, + or_replace, + transient, clone, + data_retention_time_in_days, + max_data_extension_time_in_days, + external_volume, + catalog, + replace_invalid_characters, + default_ddl_collation, + storage_serialization_policy, + comment, + catalog_sync, + catalog_sync_namespace_mode, + catalog_sync_namespace_flatten_delimiter, + with_tags, + with_contacts, } => { - write!(f, "CREATE DATABASE")?; - if *if_not_exists { - write!(f, " IF NOT EXISTS")?; - } - write!(f, " {db_name}")?; + write!( + f, + "CREATE {or_replace}{transient}DATABASE {if_not_exists}{name}", + or_replace = if *or_replace { "OR REPLACE " } else { "" }, + transient = if *transient { "TRANSIENT " } else { "" }, + if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, + name = db_name, + )?; + if let Some(l) = location { write!(f, " LOCATION '{l}'")?; } @@ -4852,6 +4881,60 @@ impl fmt::Display for Statement { if let Some(clone) = clone { write!(f, " CLONE {clone}")?; } + + if let Some(value) = data_retention_time_in_days { + write!(f, " DATA_RETENTION_TIME_IN_DAYS = {value}")?; + } + + if let Some(value) = max_data_extension_time_in_days { + write!(f, " MAX_DATA_EXTENSION_TIME_IN_DAYS = {value}")?; + } + + if let Some(vol) = external_volume { + write!(f, " EXTERNAL_VOLUME = '{vol}'")?; + } + + if let Some(cat) = catalog { + write!(f, " CATALOG = '{cat}'")?; + } + + if let Some(true) = replace_invalid_characters { + write!(f, " REPLACE_INVALID_CHARACTERS = TRUE")?; + } else if let Some(false) = replace_invalid_characters { + write!(f, " REPLACE_INVALID_CHARACTERS = FALSE")?; + } + + if let Some(collation) = default_ddl_collation { + write!(f, " DEFAULT_DDL_COLLATION = '{collation}'")?; + } + + if let Some(policy) = storage_serialization_policy { + write!(f, " STORAGE_SERIALIZATION_POLICY = {policy}")?; + } + + if let Some(comment) = comment { + write!(f, " COMMENT = '{comment}'")?; + } + + if let Some(sync) = catalog_sync { + write!(f, " CATALOG_SYNC = '{sync}'")?; + } + + if let Some(mode) = catalog_sync_namespace_mode { + write!(f, " CATALOG_SYNC_NAMESPACE_MODE = {mode}")?; + } + + if let Some(delim) = catalog_sync_namespace_flatten_delimiter { + write!(f, " CATALOG_SYNC_NAMESPACE_FLATTEN_DELIMITER = '{delim}'")?; + } + + if let Some(tags) = with_tags { + write!(f, " WITH TAG ({})", display_comma_separated(tags))?; + } + + if let Some(contacts) = with_contacts { + write!(f, " WITH CONTACT ({})", display_comma_separated(contacts))?; + } Ok(()) } Statement::CreateFunction(create_function) => create_function.fmt(f), @@ -9682,6 +9765,23 @@ impl Display for Tag { } } +/// Snowflake `WITH CONTACT ( purpose = contact [ , purpose = contact ...] )` +/// +/// +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ContactEntry { + pub purpose: String, + pub contact: String, +} + +impl Display for ContactEntry { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!(f, "{} = {}", self.purpose, self.contact) + } +} + /// Helper to indicate if a comment includes the `=` in the display form #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -10134,6 +10234,29 @@ impl Display for StorageSerializationPolicy { } } +/// Snowflake CatalogSyncNamespaceMode +/// ```sql +/// [ CATALOG_SYNC_NAMESPACE_MODE = { NEST | FLATTEN } ] +/// ``` +/// +/// +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum CatalogSyncNamespaceMode { + Nest, + Flatten, +} + +impl Display for CatalogSyncNamespaceMode { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + CatalogSyncNamespaceMode::Nest => write!(f, "NEST"), + CatalogSyncNamespaceMode::Flatten => write!(f, "FLATTEN"), + } + } +} + /// Variants of the Snowflake `COPY INTO` statement #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 706d5198e..697ba3b4f 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -20,15 +20,17 @@ use crate::alloc::string::ToString; use crate::ast::helpers::key_value_options::{ KeyValueOption, KeyValueOptionType, KeyValueOptions, KeyValueOptionsDelimiter, }; +use crate::ast::helpers::stmt_create_database::CreateDatabaseBuilder; use crate::ast::helpers::stmt_create_table::CreateTableBuilder; use crate::ast::helpers::stmt_data_loading::{ FileStagingCommand, StageLoadSelectItem, StageLoadSelectItemKind, StageParamsObject, }; use crate::ast::{ - ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, DollarQuotedString, - Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, - IdentityPropertyOrder, ObjectName, ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, - Statement, TagsColumnOption, WrappedCollection, + CatalogSyncNamespaceMode, ColumnOption, ColumnPolicy, ColumnPolicyProperty, ContactEntry, + CopyIntoSnowflakeKind, DollarQuotedString, Ident, IdentityParameters, IdentityProperty, + IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, ObjectName, + ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, Statement, StorageSerializationPolicy, + TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -44,7 +46,6 @@ use alloc::vec::Vec; use alloc::{format, vec}; use super::keywords::RESERVED_FOR_IDENTIFIER; -use sqlparser::ast::StorageSerializationPolicy; const RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR: [Keyword; 1] = [Keyword::CONNECT_BY_ROOT]; @@ -260,6 +261,8 @@ impl Dialect for SnowflakeDialect { return Some(parse_create_table( or_replace, global, temporary, volatile, transient, iceberg, parser, )); + } else if parser.parse_keyword(Keyword::DATABASE) { + return Some(parse_create_database(or_replace, transient, parser)); } else { // need to go back with the cursor let mut back = 1; @@ -678,29 +681,11 @@ pub fn parse_create_table( } Keyword::ENABLE_SCHEMA_EVOLUTION => { parser.expect_token(&Token::Eq)?; - let enable_schema_evolution = - match parser.parse_one_of_keywords(&[Keyword::TRUE, Keyword::FALSE]) { - Some(Keyword::TRUE) => true, - Some(Keyword::FALSE) => false, - _ => { - return parser.expected("TRUE or FALSE", next_token); - } - }; - - builder = builder.enable_schema_evolution(Some(enable_schema_evolution)); + builder = builder.enable_schema_evolution(Some(parser.parse_boolean_string()?)); } Keyword::CHANGE_TRACKING => { parser.expect_token(&Token::Eq)?; - let change_tracking = - match parser.parse_one_of_keywords(&[Keyword::TRUE, Keyword::FALSE]) { - Some(Keyword::TRUE) => true, - Some(Keyword::FALSE) => false, - _ => { - return parser.expected("TRUE or FALSE", next_token); - } - }; - - builder = builder.change_tracking(Some(change_tracking)); + builder = builder.change_tracking(Some(parser.parse_boolean_string()?)); } Keyword::DATA_RETENTION_TIME_IN_DAYS => { parser.expect_token(&Token::Eq)?; @@ -830,6 +815,115 @@ pub fn parse_create_table( Ok(builder.build()) } +/// Parse snowflake create database statement. +/// +pub fn parse_create_database( + or_replace: bool, + transient: bool, + parser: &mut Parser, +) -> Result { + let if_not_exists = parser.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let name = parser.parse_object_name(false)?; + + let mut builder = CreateDatabaseBuilder::new(name) + .or_replace(or_replace) + .transient(transient) + .if_not_exists(if_not_exists); + + loop { + let next_token = parser.next_token(); + match &next_token.token { + Token::Word(word) => match word.keyword { + Keyword::CLONE => { + builder = builder.clone_clause(Some(parser.parse_object_name(false)?)); + } + Keyword::DATA_RETENTION_TIME_IN_DAYS => { + parser.expect_token(&Token::Eq)?; + builder = + builder.data_retention_time_in_days(Some(parser.parse_literal_uint()?)); + } + Keyword::MAX_DATA_EXTENSION_TIME_IN_DAYS => { + parser.expect_token(&Token::Eq)?; + builder = + builder.max_data_extension_time_in_days(Some(parser.parse_literal_uint()?)); + } + Keyword::EXTERNAL_VOLUME => { + parser.expect_token(&Token::Eq)?; + builder = builder.external_volume(Some(parser.parse_literal_string()?)); + } + Keyword::CATALOG => { + parser.expect_token(&Token::Eq)?; + builder = builder.catalog(Some(parser.parse_literal_string()?)); + } + Keyword::REPLACE_INVALID_CHARACTERS => { + parser.expect_token(&Token::Eq)?; + builder = + builder.replace_invalid_characters(Some(parser.parse_boolean_string()?)); + } + Keyword::DEFAULT_DDL_COLLATION => { + parser.expect_token(&Token::Eq)?; + builder = builder.default_ddl_collation(Some(parser.parse_literal_string()?)); + } + Keyword::STORAGE_SERIALIZATION_POLICY => { + parser.expect_token(&Token::Eq)?; + let policy = parse_storage_serialization_policy(parser)?; + builder = builder.storage_serialization_policy(Some(policy)); + } + Keyword::COMMENT => { + parser.expect_token(&Token::Eq)?; + builder = builder.comment(Some(parser.parse_literal_string()?)); + } + Keyword::CATALOG_SYNC => { + parser.expect_token(&Token::Eq)?; + builder = builder.catalog_sync(Some(parser.parse_literal_string()?)); + } + Keyword::CATALOG_SYNC_NAMESPACE_FLATTEN_DELIMITER => { + parser.expect_token(&Token::Eq)?; + builder = builder.catalog_sync_namespace_flatten_delimiter(Some( + parser.parse_literal_string()?, + )); + } + Keyword::CATALOG_SYNC_NAMESPACE_MODE => { + parser.expect_token(&Token::Eq)?; + let mode = + match parser.parse_one_of_keywords(&[Keyword::NEST, Keyword::FLATTEN]) { + Some(Keyword::NEST) => CatalogSyncNamespaceMode::Nest, + Some(Keyword::FLATTEN) => CatalogSyncNamespaceMode::Flatten, + _ => { + return parser.expected("NEST or FLATTEN", next_token); + } + }; + builder = builder.catalog_sync_namespace_mode(Some(mode)); + } + Keyword::WITH => { + if parser.parse_keyword(Keyword::TAG) { + parser.expect_token(&Token::LParen)?; + let tags = parser.parse_comma_separated(Parser::parse_tag)?; + parser.expect_token(&Token::RParen)?; + builder = builder.with_tags(Some(tags)); + } else if parser.parse_keyword(Keyword::CONTACT) { + parser.expect_token(&Token::LParen)?; + let contacts = parser.parse_comma_separated(|p| { + let purpose = p.parse_identifier()?.value; + p.expect_token(&Token::Eq)?; + let contact = p.parse_identifier()?.value; + Ok(ContactEntry { purpose, contact }) + })?; + parser.expect_token(&Token::RParen)?; + builder = builder.with_contacts(Some(contacts)); + } else { + return parser.expected("TAG or CONTACT", next_token); + } + } + _ => return parser.expected("end of statement", next_token), + }, + Token::SemiColon | Token::EOF => break, + _ => return parser.expected("end of statement", next_token), + } + } + Ok(builder.build()) +} + pub fn parse_storage_serialization_policy( parser: &mut Parser, ) -> Result { diff --git a/src/keywords.rs b/src/keywords.rs index 509c1f96c..a729a525f 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -166,6 +166,8 @@ define_keywords!( CAST, CATALOG, CATALOG_SYNC, + CATALOG_SYNC_NAMESPACE_FLATTEN_DELIMITER, + CATALOG_SYNC_NAMESPACE_MODE, CATCH, CEIL, CEILING, @@ -213,6 +215,7 @@ define_keywords!( CONNECTOR, CONNECT_BY_ROOT, CONSTRAINT, + CONTACT, CONTAINS, CONTINUE, CONVERT, @@ -366,6 +369,7 @@ define_keywords!( FIRST, FIRST_VALUE, FIXEDSTRING, + FLATTEN, FLOAT, FLOAT32, FLOAT4, @@ -584,6 +588,7 @@ define_keywords!( NATURAL, NCHAR, NCLOB, + NEST, NESTED, NETWORK, NEW, @@ -756,6 +761,7 @@ define_keywords!( REPAIR, REPEATABLE, REPLACE, + REPLACE_INVALID_CHARACTERS, REPLICA, REPLICATE, REPLICATION, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9b8fa17ac..75b2cf520 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5054,7 +5054,22 @@ impl<'a> Parser<'a> { if_not_exists: ine, location, managed_location, + or_replace: false, + transient: false, clone, + data_retention_time_in_days: None, + max_data_extension_time_in_days: None, + external_volume: None, + catalog: None, + replace_invalid_characters: None, + default_ddl_collation: None, + storage_serialization_policy: None, + comment: None, + catalog_sync: None, + catalog_sync_namespace_mode: None, + catalog_sync_namespace_flatten_delimiter: None, + with_tags: None, + with_contacts: None, }) } @@ -9763,6 +9778,15 @@ impl<'a> Parser<'a> { } } + /// Parse a boolean string + pub(crate) fn parse_boolean_string(&mut self) -> Result { + match self.parse_one_of_keywords(&[Keyword::TRUE, Keyword::FALSE]) { + Some(Keyword::TRUE) => Ok(true), + Some(Keyword::FALSE) => Ok(false), + _ => self.expected("TRUE or FALSE", self.peek_token()), + } + } + /// Parse a literal unicode normalization clause pub fn parse_unicode_is_normalized(&mut self, expr: Expr) -> Result { let neg = self.parse_keyword(Keyword::NOT); diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index ec7eec7d4..72b6c4e46 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7950,6 +7950,7 @@ fn parse_create_database() { location, managed_location, clone, + .. } => { assert_eq!("mydb", db_name.to_string()); assert!(!if_not_exists); @@ -7967,6 +7968,7 @@ fn parse_create_database() { location, managed_location, clone, + .. } => { assert_eq!("mydb", db_name.to_string()); assert!(!if_not_exists); @@ -7991,6 +7993,7 @@ fn parse_create_database_ine() { location, managed_location, clone, + .. } => { assert_eq!("mydb", db_name.to_string()); assert!(if_not_exists); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index bd089ad7d..eb4bda7bd 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4516,3 +4516,54 @@ fn test_snowflake_identifier_function() { snowflake().verified_stmt("GRANT ROLE IDENTIFIER('AAA') TO USER IDENTIFIER('AAA')"); snowflake().verified_stmt("REVOKE ROLE IDENTIFIER('AAA') FROM USER IDENTIFIER('AAA')"); } + +#[test] +fn test_create_database() { + snowflake().verified_stmt("CREATE DATABASE my_db"); + snowflake().verified_stmt("CREATE OR REPLACE DATABASE my_db"); + snowflake().verified_stmt("CREATE TRANSIENT DATABASE IF NOT EXISTS my_db"); + snowflake().verified_stmt("CREATE DATABASE my_db CLONE src_db"); + snowflake().verified_stmt( + "CREATE OR REPLACE DATABASE my_db CLONE src_db DATA_RETENTION_TIME_IN_DAYS = 1", + ); + snowflake().one_statement_parses_to( + r#" + CREATE OR REPLACE TRANSIENT DATABASE IF NOT EXISTS my_db + CLONE src_db + DATA_RETENTION_TIME_IN_DAYS = 1 + MAX_DATA_EXTENSION_TIME_IN_DAYS = 5 + EXTERNAL_VOLUME = 'volume1' + CATALOG = 'my_catalog' + REPLACE_INVALID_CHARACTERS = TRUE + DEFAULT_DDL_COLLATION = 'en-ci' + STORAGE_SERIALIZATION_POLICY = COMPATIBLE + COMMENT = 'This is my database' + CATALOG_SYNC = 'sync_integration' + CATALOG_SYNC_NAMESPACE_MODE = NEST + CATALOG_SYNC_NAMESPACE_FLATTEN_DELIMITER = '/' + WITH TAG (env = 'prod', team = 'data') + WITH CONTACT (owner = 'admin', dpo = 'compliance') + "#, + "CREATE OR REPLACE TRANSIENT DATABASE IF NOT EXISTS \ + my_db CLONE src_db DATA_RETENTION_TIME_IN_DAYS = 1 MAX_DATA_EXTENSION_TIME_IN_DAYS = 5 \ + EXTERNAL_VOLUME = 'volume1' CATALOG = 'my_catalog' \ + REPLACE_INVALID_CHARACTERS = TRUE DEFAULT_DDL_COLLATION = 'en-ci' \ + STORAGE_SERIALIZATION_POLICY = COMPATIBLE COMMENT = 'This is my database' \ + CATALOG_SYNC = 'sync_integration' CATALOG_SYNC_NAMESPACE_MODE = NEST \ + CATALOG_SYNC_NAMESPACE_FLATTEN_DELIMITER = '/' \ + WITH TAG (env='prod', team='data') \ + WITH CONTACT (owner = admin, dpo = compliance)", + ); + + let err = snowflake() + .parse_sql_statements("CREATE DATABASE") + .unwrap_err() + .to_string(); + assert!(err.contains("Expected"), "Unexpected error: {err}"); + + let err = snowflake() + .parse_sql_statements("CREATE DATABASE my_db CLONE") + .unwrap_err() + .to_string(); + assert!(err.contains("Expected"), "Unexpected error: {err}"); +} From dd650b88f3fed130cc34fc3ae50435e61aa1f947 Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Fri, 1 Aug 2025 22:52:41 +0800 Subject: [PATCH 225/286] feat: support multi value columns and aliases in unpivot (#1969) --- src/ast/query.rs | 5 +- src/ast/spans.rs | 4 +- src/parser/mod.rs | 6 +- tests/sqlparser_common.rs | 159 ++++++++++++++++++++++++++++++++------ 4 files changed, 144 insertions(+), 30 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 7ffb64d9b..ea641deba 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1349,11 +1349,12 @@ pub enum TableFactor { /// ``` /// /// See . + /// See . Unpivot { table: Box, - value: Ident, + value: Expr, name: Ident, - columns: Vec, + columns: Vec, null_inclusion: Option, alias: Option, }, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 585836601..dec26566c 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2002,9 +2002,9 @@ impl Spanned for TableFactor { alias, } => union_spans( core::iter::once(table.span()) - .chain(core::iter::once(value.span)) + .chain(core::iter::once(value.span())) .chain(core::iter::once(name.span)) - .chain(columns.iter().map(|i| i.span)) + .chain(columns.iter().map(|ilist| ilist.span())) .chain(alias.as_ref().map(|alias| alias.span())), ), TableFactor::MatchRecognize { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 75b2cf520..455e0caad 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -14033,11 +14033,13 @@ impl<'a> Parser<'a> { None }; self.expect_token(&Token::LParen)?; - let value = self.parse_identifier()?; + let value = self.parse_expr()?; self.expect_keyword_is(Keyword::FOR)?; let name = self.parse_identifier()?; self.expect_keyword_is(Keyword::IN)?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_column_list_inner(Mandatory, false, |p| { + p.parse_expr_with_alias() + })?; self.expect_token(&Token::RParen)?; let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Unpivot { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 72b6c4e46..5e389aeea 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11009,20 +11009,14 @@ fn parse_unpivot_table() { index_hints: vec![], }), null_inclusion: None, - value: Ident { - value: "quantity".to_string(), - quote_style: None, - span: Span::empty(), - }, - - name: Ident { - value: "quarter".to_string(), - quote_style: None, - span: Span::empty(), - }, + value: Expr::Identifier(Ident::new("quantity")), + name: Ident::new("quarter"), columns: ["Q1", "Q2", "Q3", "Q4"] .into_iter() - .map(Ident::new) + .map(|col| ExprWithAlias { + expr: Expr::Identifier(Ident::new(col)), + alias: None, + }) .collect(), alias: Some(TableAlias { name: Ident::new("u"), @@ -11084,6 +11078,129 @@ fn parse_unpivot_table() { verified_stmt(sql_unpivot_include_nulls).to_string(), sql_unpivot_include_nulls ); + + let sql_unpivot_with_alias = concat!( + "SELECT * FROM sales AS s ", + "UNPIVOT INCLUDE NULLS ", + "(quantity FOR quarter IN ", + "(Q1 AS Quater1, Q2 AS Quater2, Q3 AS Quater3, Q4 AS Quater4)) ", + "AS u (product, quarter, quantity)" + ); + + if let Unpivot { value, columns, .. } = + &verified_only_select(sql_unpivot_with_alias).from[0].relation + { + assert_eq!( + *columns, + vec![ + ExprWithAlias { + expr: Expr::Identifier(Ident::new("Q1")), + alias: Some(Ident::new("Quater1")), + }, + ExprWithAlias { + expr: Expr::Identifier(Ident::new("Q2")), + alias: Some(Ident::new("Quater2")), + }, + ExprWithAlias { + expr: Expr::Identifier(Ident::new("Q3")), + alias: Some(Ident::new("Quater3")), + }, + ExprWithAlias { + expr: Expr::Identifier(Ident::new("Q4")), + alias: Some(Ident::new("Quater4")), + }, + ] + ); + assert_eq!(*value, Expr::Identifier(Ident::new("quantity"))); + } + + assert_eq!( + verified_stmt(sql_unpivot_with_alias).to_string(), + sql_unpivot_with_alias + ); + + let sql_unpivot_with_alias_and_multi_value = concat!( + "SELECT * FROM sales AS s ", + "UNPIVOT INCLUDE NULLS ((first_quarter, second_quarter) ", + "FOR half_of_the_year IN (", + "(Q1, Q2) AS H1, ", + "(Q3, Q4) AS H2", + "))" + ); + + if let Unpivot { value, columns, .. } = + &verified_only_select(sql_unpivot_with_alias_and_multi_value).from[0].relation + { + assert_eq!( + *columns, + vec![ + ExprWithAlias { + expr: Expr::Tuple(vec![ + Expr::Identifier(Ident::new("Q1")), + Expr::Identifier(Ident::new("Q2")), + ]), + alias: Some(Ident::new("H1")), + }, + ExprWithAlias { + expr: Expr::Tuple(vec![ + Expr::Identifier(Ident::new("Q3")), + Expr::Identifier(Ident::new("Q4")), + ]), + alias: Some(Ident::new("H2")), + }, + ] + ); + assert_eq!( + *value, + Expr::Tuple(vec![ + Expr::Identifier(Ident::new("first_quarter")), + Expr::Identifier(Ident::new("second_quarter")), + ]) + ); + } + + assert_eq!( + verified_stmt(sql_unpivot_with_alias_and_multi_value).to_string(), + sql_unpivot_with_alias_and_multi_value + ); + + let sql_unpivot_with_alias_and_multi_value_and_qualifier = concat!( + "SELECT * FROM sales AS s ", + "UNPIVOT INCLUDE NULLS ((first_quarter, second_quarter) ", + "FOR half_of_the_year IN (", + "(sales.Q1, sales.Q2) AS H1, ", + "(sales.Q3, sales.Q4) AS H2", + "))" + ); + + if let Unpivot { columns, .. } = + &verified_only_select(sql_unpivot_with_alias_and_multi_value_and_qualifier).from[0].relation + { + assert_eq!( + *columns, + vec![ + ExprWithAlias { + expr: Expr::Tuple(vec![ + Expr::CompoundIdentifier(vec![Ident::new("sales"), Ident::new("Q1"),]), + Expr::CompoundIdentifier(vec![Ident::new("sales"), Ident::new("Q2"),]), + ]), + alias: Some(Ident::new("H1")), + }, + ExprWithAlias { + expr: Expr::Tuple(vec![ + Expr::CompoundIdentifier(vec![Ident::new("sales"), Ident::new("Q3"),]), + Expr::CompoundIdentifier(vec![Ident::new("sales"), Ident::new("Q4"),]), + ]), + alias: Some(Ident::new("H2")), + }, + ] + ); + } + + assert_eq!( + verified_stmt(sql_unpivot_with_alias_and_multi_value_and_qualifier).to_string(), + sql_unpivot_with_alias_and_multi_value_and_qualifier + ); } #[test] @@ -11181,20 +11298,14 @@ fn parse_pivot_unpivot_table() { index_hints: vec![], }), null_inclusion: None, - value: Ident { - value: "population".to_string(), - quote_style: None, - span: Span::empty() - }, - - name: Ident { - value: "year".to_string(), - quote_style: None, - span: Span::empty() - }, + value: Expr::Identifier(Ident::new("population")), + name: Ident::new("year"), columns: ["population_2000", "population_2010"] .into_iter() - .map(Ident::new) + .map(|col| ExprWithAlias { + expr: Expr::Identifier(Ident::new(col)), + alias: None, + }) .collect(), alias: Some(TableAlias { name: Ident::new("u"), From c1648e79fe1452a6e574c88bd089f3224f012c8c Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 1 Aug 2025 23:05:13 -0700 Subject: [PATCH 226/286] Postgres: Support `INTERVAL` data type options (#1984) --- src/ast/data_type.rs | 64 +++++++++++++++++++++++- src/ast/mod.rs | 2 +- src/dialect/generic.rs | 4 ++ src/dialect/mod.rs | 15 ++++++ src/dialect/postgresql.rs | 7 +++ src/parser/mod.rs | 97 +++++++++++++++++++++++++++++++++++-- tests/sqlparser_bigquery.rs | 10 +++- tests/sqlparser_common.rs | 15 ++++-- tests/sqlparser_postgres.rs | 38 +++++++++++++++ 9 files changed, 239 insertions(+), 13 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 0897f2db7..b4a8af60f 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -346,7 +346,16 @@ pub enum DataType { /// [1]: https://docs.databricks.com/aws/en/sql/language-manual/data-types/timestamp-ntz-type TimestampNtz, /// Interval type. - Interval, + Interval { + /// [PostgreSQL] fields specification like `INTERVAL YEAR TO MONTH`. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/17/datatype-datetime.html + fields: Option, + /// [PostgreSQL] subsecond precision like `INTERVAL HOUR TO SECOND(3)` + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/17/datatype-datetime.html + precision: Option, + }, /// JSON type. JSON, /// Binary JSON type. @@ -635,7 +644,16 @@ impl fmt::Display for DataType { timezone, ) } - DataType::Interval => write!(f, "INTERVAL"), + DataType::Interval { fields, precision } => { + write!(f, "INTERVAL")?; + if let Some(fields) = fields { + write!(f, " {fields}")?; + } + if let Some(precision) = precision { + write!(f, "({precision})")?; + } + Ok(()) + } DataType::JSON => write!(f, "JSON"), DataType::JSONB => write!(f, "JSONB"), DataType::Regclass => write!(f, "REGCLASS"), @@ -889,6 +907,48 @@ impl fmt::Display for TimezoneInfo { } } +/// Fields for [Postgres] `INTERVAL` type. +/// +/// [Postgres]: https://www.postgresql.org/docs/17/datatype-datetime.html +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum IntervalFields { + Year, + Month, + Day, + Hour, + Minute, + Second, + YearToMonth, + DayToHour, + DayToMinute, + DayToSecond, + HourToMinute, + HourToSecond, + MinuteToSecond, +} + +impl fmt::Display for IntervalFields { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + IntervalFields::Year => write!(f, "YEAR"), + IntervalFields::Month => write!(f, "MONTH"), + IntervalFields::Day => write!(f, "DAY"), + IntervalFields::Hour => write!(f, "HOUR"), + IntervalFields::Minute => write!(f, "MINUTE"), + IntervalFields::Second => write!(f, "SECOND"), + IntervalFields::YearToMonth => write!(f, "YEAR TO MONTH"), + IntervalFields::DayToHour => write!(f, "DAY TO HOUR"), + IntervalFields::DayToMinute => write!(f, "DAY TO MINUTE"), + IntervalFields::DayToSecond => write!(f, "DAY TO SECOND"), + IntervalFields::HourToMinute => write!(f, "HOUR TO MINUTE"), + IntervalFields::HourToSecond => write!(f, "HOUR TO SECOND"), + IntervalFields::MinuteToSecond => write!(f, "MINUTE TO SECOND"), + } + } +} + /// Additional information for `NUMERIC`, `DECIMAL`, and `DEC` data types /// following the 2016 [SQL Standard]. /// diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 34112ac64..c6212f1e4 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -52,7 +52,7 @@ use crate::{ pub use self::data_type::{ ArrayElemTypeDef, BinaryLength, CharLengthUnits, CharacterLength, DataType, EnumMember, - ExactNumberInfo, StructBracketKind, TimezoneInfo, + ExactNumberInfo, IntervalFields, StructBracketKind, TimezoneInfo, }; pub use self::dcl::{ AlterRoleOperation, ResetConfig, RoleOption, SecondaryRoles, SetConfigValue, Use, diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index de83d507e..b4c3ef027 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -187,4 +187,8 @@ impl Dialect for GenericDialect { fn supports_data_type_signed_suffix(&self) -> bool { true } + + fn supports_interval_options(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 9003220c0..5f333a931 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1148,6 +1148,21 @@ pub trait Dialect: Debug + Any { fn supports_data_type_signed_suffix(&self) -> bool { false } + + /// Returns true if the dialect supports the `INTERVAL` data type with [Postgres]-style options. + /// + /// Examples: + /// ```sql + /// CREATE TABLE t (i INTERVAL YEAR TO MONTH); + /// SELECT '1 second'::INTERVAL HOUR TO SECOND(3); + /// ``` + /// + /// See [`crate::ast::DataType::Interval`] and [`crate::ast::IntervalFields`]. + /// + /// [Postgres]: https://www.postgresql.org/docs/17/datatype-datetime.html + fn supports_interval_options(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index b12abaaf7..207f4787f 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -269,4 +269,11 @@ impl Dialect for PostgreSqlDialect { fn supports_notnull_operator(&self) -> bool { true } + + /// [Postgres] supports optional field and precision options for `INTERVAL` data type. + /// + /// [Postgres]: https://www.postgresql.org/docs/17/datatype-datetime.html + fn supports_interval_options(&self) -> bool { + true + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 455e0caad..c3230a215 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1534,7 +1534,7 @@ impl<'a> Parser<'a> { let loc = self.peek_token_ref().span.start; let opt_expr = self.maybe_parse(|parser| { match parser.parse_data_type()? { - DataType::Interval => parser.parse_interval(), + DataType::Interval { .. } => parser.parse_interval(), // PostgreSQL allows almost any identifier to be used as custom data type name, // and we support that in `parse_data_type()`. But unlike Postgres we don't // have a list of globally reserved keywords (since they vary across dialects), @@ -10066,10 +10066,18 @@ impl<'a> Parser<'a> { self.parse_optional_precision()?, TimezoneInfo::Tz, )), - // Interval types can be followed by a complicated interval - // qualifier that we don't currently support. See - // parse_interval for a taste. - Keyword::INTERVAL => Ok(DataType::Interval), + Keyword::INTERVAL => { + if self.dialect.supports_interval_options() { + let fields = self.maybe_parse_optional_interval_fields()?; + let precision = self.parse_optional_precision()?; + Ok(DataType::Interval { fields, precision }) + } else { + Ok(DataType::Interval { + fields: None, + precision: None, + }) + } + } Keyword::JSON => Ok(DataType::JSON), Keyword::JSONB => Ok(DataType::JSONB), Keyword::REGCLASS => Ok(DataType::Regclass), @@ -11038,6 +11046,85 @@ impl<'a> Parser<'a> { } } + fn maybe_parse_optional_interval_fields( + &mut self, + ) -> Result, ParserError> { + match self.parse_one_of_keywords(&[ + // Can be followed by `TO` option + Keyword::YEAR, + Keyword::DAY, + Keyword::HOUR, + Keyword::MINUTE, + // No `TO` option + Keyword::MONTH, + Keyword::SECOND, + ]) { + Some(Keyword::YEAR) => { + if self.peek_keyword(Keyword::TO) { + self.expect_keyword(Keyword::TO)?; + self.expect_keyword(Keyword::MONTH)?; + Ok(Some(IntervalFields::YearToMonth)) + } else { + Ok(Some(IntervalFields::Year)) + } + } + Some(Keyword::DAY) => { + if self.peek_keyword(Keyword::TO) { + self.expect_keyword(Keyword::TO)?; + match self.expect_one_of_keywords(&[ + Keyword::HOUR, + Keyword::MINUTE, + Keyword::SECOND, + ])? { + Keyword::HOUR => Ok(Some(IntervalFields::DayToHour)), + Keyword::MINUTE => Ok(Some(IntervalFields::DayToMinute)), + Keyword::SECOND => Ok(Some(IntervalFields::DayToSecond)), + _ => { + self.prev_token(); + self.expected("HOUR, MINUTE, or SECOND", self.peek_token()) + } + } + } else { + Ok(Some(IntervalFields::Day)) + } + } + Some(Keyword::HOUR) => { + if self.peek_keyword(Keyword::TO) { + self.expect_keyword(Keyword::TO)?; + match self.expect_one_of_keywords(&[Keyword::MINUTE, Keyword::SECOND])? { + Keyword::MINUTE => Ok(Some(IntervalFields::HourToMinute)), + Keyword::SECOND => Ok(Some(IntervalFields::HourToSecond)), + _ => { + self.prev_token(); + self.expected("MINUTE or SECOND", self.peek_token()) + } + } + } else { + Ok(Some(IntervalFields::Hour)) + } + } + Some(Keyword::MINUTE) => { + if self.peek_keyword(Keyword::TO) { + self.expect_keyword(Keyword::TO)?; + self.expect_keyword(Keyword::SECOND)?; + Ok(Some(IntervalFields::MinuteToSecond)) + } else { + Ok(Some(IntervalFields::Minute)) + } + } + Some(Keyword::MONTH) => Ok(Some(IntervalFields::Month)), + Some(Keyword::SECOND) => Ok(Some(IntervalFields::Second)), + Some(_) => { + self.prev_token(); + self.expected( + "YEAR, MONTH, DAY, HOUR, MINUTE, or SECOND", + self.peek_token(), + ) + } + None => Ok(None), + } + } + /// Parse datetime64 [1] /// Syntax /// ```sql diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 839ea6b8c..5c14d140c 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -961,7 +961,10 @@ fn parse_typed_struct_syntax_bigquery() { })], fields: vec![StructField { field_name: None, - field_type: DataType::Interval, + field_type: DataType::Interval { + fields: None, + precision: None + }, options: None, }] }, @@ -1300,7 +1303,10 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { })], fields: vec![StructField { field_name: None, - field_type: DataType::Interval, + field_type: DataType::Interval { + fields: None, + precision: None + }, options: None, }] }, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 5e389aeea..a64733d67 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -12955,7 +12955,10 @@ fn test_extract_seconds_ok() { expr: Box::new(Expr::Value( (Value::SingleQuotedString("2 seconds".to_string())).with_empty_span() )), - data_type: DataType::Interval, + data_type: DataType::Interval { + fields: None, + precision: None + }, format: None, }), } @@ -12980,7 +12983,10 @@ fn test_extract_seconds_ok() { expr: Box::new(Expr::Value( (Value::SingleQuotedString("2 seconds".to_string())).with_empty_span(), )), - data_type: DataType::Interval, + data_type: DataType::Interval { + fields: None, + precision: None, + }, format: None, }), })], @@ -13034,7 +13040,10 @@ fn test_extract_seconds_single_quote_ok() { expr: Box::new(Expr::Value( (Value::SingleQuotedString("2 seconds".to_string())).with_empty_span() )), - data_type: DataType::Interval, + data_type: DataType::Interval { + fields: None, + precision: None + }, format: None, }), } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 6ce4a4835..6c00b1f11 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5332,6 +5332,44 @@ fn parse_at_time_zone() { ); } +#[test] +fn parse_interval_data_type() { + pg_and_generic().verified_stmt("CREATE TABLE t (i INTERVAL)"); + for p in 0..=6 { + pg_and_generic().verified_stmt(&format!("CREATE TABLE t (i INTERVAL({p}))")); + pg_and_generic().verified_stmt(&format!("SELECT '1 second'::INTERVAL({p})")); + pg_and_generic().verified_stmt(&format!("SELECT CAST('1 second' AS INTERVAL({p}))")); + } + let fields = [ + "YEAR", + "MONTH", + "DAY", + "HOUR", + "MINUTE", + "SECOND", + "YEAR TO MONTH", + "DAY TO HOUR", + "DAY TO MINUTE", + "DAY TO SECOND", + "HOUR TO MINUTE", + "HOUR TO SECOND", + "MINUTE TO SECOND", + ]; + for field in fields { + pg_and_generic().verified_stmt(&format!("CREATE TABLE t (i INTERVAL {field})")); + pg_and_generic().verified_stmt(&format!("SELECT '1 second'::INTERVAL {field}")); + pg_and_generic().verified_stmt(&format!("SELECT CAST('1 second' AS INTERVAL {field})")); + } + for p in 0..=6 { + for field in fields { + pg_and_generic().verified_stmt(&format!("CREATE TABLE t (i INTERVAL {field}({p}))")); + pg_and_generic().verified_stmt(&format!("SELECT '1 second'::INTERVAL {field}({p})")); + pg_and_generic() + .verified_stmt(&format!("SELECT CAST('1 second' AS INTERVAL {field}({p}))")); + } + } +} + #[test] fn parse_create_table_with_options() { let sql = "CREATE TABLE t (c INT) WITH (foo = 'bar', a = 123)"; From 35835146020f0f4782fd4e5caf05aa8b7e17e7b6 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Tue, 5 Aug 2025 01:30:42 -0700 Subject: [PATCH 227/286] MySQL: Support comma-separated `CREATE TABLE` options (#1989) --- src/dialect/mod.rs | 2 +- src/parser/mod.rs | 3 +++ tests/sqlparser_mysql.rs | 7 +++++++ 3 files changed, 11 insertions(+), 1 deletion(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 5f333a931..6ca364784 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -590,7 +590,7 @@ pub trait Dialect: Debug + Any { false } - /// Returne true if the dialect supports specifying multiple options + /// Return true if the dialect supports specifying multiple options /// in a `CREATE TABLE` statement for the structure of the new table. For example: /// `CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a` fn supports_create_table_multi_schema_info_sources(&self) -> bool { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c3230a215..39571ba10 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7708,6 +7708,9 @@ impl<'a> Parser<'a> { while let Some(option) = self.parse_plain_option()? { options.push(option); + // Some dialects support comma-separated options; it shouldn't introduce ambiguity to + // consume it for all dialects. + let _ = self.consume_token(&Token::Comma); } Ok(options) diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 3ea0543f1..cf4e24adb 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1361,6 +1361,13 @@ fn parse_create_table_gencol() { mysql_and_generic().verified_stmt("CREATE TABLE t1 (a INT, b INT AS (a * 2) STORED)"); } +#[test] +fn parse_create_table_options_comma_separated() { + let sql = "CREATE TABLE t (x INT) DEFAULT CHARSET = utf8mb4, ENGINE = InnoDB , AUTO_INCREMENT 1 DATA DIRECTORY '/var/lib/mysql/data'"; + let canonical = "CREATE TABLE t (x INT) DEFAULT CHARSET = utf8mb4 ENGINE = InnoDB AUTO_INCREMENT = 1 DATA DIRECTORY = '/var/lib/mysql/data'"; + mysql_and_generic().one_statement_parses_to(sql, canonical); +} + #[test] fn parse_quote_identifiers() { let sql = "CREATE TABLE `PRIMARY` (`BEGIN` INT PRIMARY KEY)"; From 698154d0e03ddce63e9a459795f3523db7aaef42 Mon Sep 17 00:00:00 2001 From: Marcelo Altmann Date: Wed, 6 Aug 2025 08:18:58 -0300 Subject: [PATCH 228/286] MySQL: Support `ALTER TABLE RENAME AS` (#1965) --- src/ast/ddl.rs | 41 ++++++++++++++++++++++++++++++++++----- src/ast/mod.rs | 6 +++--- src/parser/mod.rs | 9 ++++++++- tests/sqlparser_common.rs | 16 ++++++++++++++- 4 files changed, 62 insertions(+), 10 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 231ab49ca..4d1fc4628 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -33,11 +33,11 @@ use crate::ast::{ display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, Ident, IndexColumn, MySQLColumnPosition, - ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, - Value, ValueWithSpan, + ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, Spanned, + SqlOption, Tag, Value, ValueWithSpan, }; use crate::keywords::Keyword; -use crate::tokenizer::Token; +use crate::tokenizer::{Span, Token}; /// ALTER TABLE operation REPLICA IDENTITY values /// See [Postgres ALTER TABLE docs](https://www.postgresql.org/docs/current/sql-altertable.html) @@ -264,7 +264,7 @@ pub enum AlterTableOperation { }, /// `RENAME TO ` RenameTable { - table_name: ObjectName, + table_name: RenameTableNameKind, }, // CHANGE [ COLUMN ] [ ] ChangeColumn { @@ -697,7 +697,7 @@ impl fmt::Display for AlterTableOperation { new_column_name, } => write!(f, "RENAME COLUMN {old_column_name} TO {new_column_name}"), AlterTableOperation::RenameTable { table_name } => { - write!(f, "RENAME TO {table_name}") + write!(f, "RENAME {table_name}") } AlterTableOperation::ChangeColumn { old_name, @@ -2537,3 +2537,34 @@ impl fmt::Display for CreateConnector { Ok(()) } } + +/// `RenameTableNameKind` is the kind used in an `ALTER TABLE _ RENAME` statement. +/// +/// Note: [MySQL] is the only database that supports the AS keyword for this operation. +/// +/// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum RenameTableNameKind { + As(ObjectName), + To(ObjectName), +} + +impl fmt::Display for RenameTableNameKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + RenameTableNameKind::As(name) => write!(f, "AS {name}"), + RenameTableNameKind::To(name) => write!(f, "TO {name}"), + } + } +} + +impl Spanned for RenameTableNameKind { + fn span(&self) -> Span { + match self { + RenameTableNameKind::As(name) => name.span(), + RenameTableNameKind::To(name) => name.span(), + } + } +} diff --git a/src/ast/mod.rs b/src/ast/mod.rs index c6212f1e4..4a864cfa7 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -66,9 +66,9 @@ pub use self::ddl::{ Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, - Partition, ProcedureParam, ReferentialAction, ReplicaIdentity, TableConstraint, - TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, - ViewColumnDef, + Partition, ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, + TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, + UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 39571ba10..d9fe7999e 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8752,7 +8752,14 @@ impl<'a> Parser<'a> { AlterTableOperation::RenameConstraint { old_name, new_name } } else if self.parse_keyword(Keyword::TO) { let table_name = self.parse_object_name(false)?; - AlterTableOperation::RenameTable { table_name } + AlterTableOperation::RenameTable { + table_name: RenameTableNameKind::To(table_name), + } + } else if self.parse_keyword(Keyword::AS) { + let table_name = self.parse_object_name(false)?; + AlterTableOperation::RenameTable { + table_name: RenameTableNameKind::As(table_name), + } } else { let _ = self.parse_keyword(Keyword::COLUMN); // [ COLUMN ] let old_column_name = self.parse_identifier()?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a64733d67..7eff55039 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -4692,7 +4692,21 @@ fn parse_alter_table() { let rename_table = "ALTER TABLE tab RENAME TO new_tab"; match alter_table_op(verified_stmt(rename_table)) { AlterTableOperation::RenameTable { table_name } => { - assert_eq!("new_tab", table_name.to_string()); + assert_eq!( + RenameTableNameKind::To(ObjectName::from(vec![Ident::new("new_tab")])), + table_name + ); + } + _ => unreachable!(), + }; + + let rename_table_as = "ALTER TABLE tab RENAME AS new_tab"; + match alter_table_op(verified_stmt(rename_table_as)) { + AlterTableOperation::RenameTable { table_name } => { + assert_eq!( + RenameTableNameKind::As(ObjectName::from(vec![Ident::new("new_tab")])), + table_name + ); } _ => unreachable!(), }; From 183bc7c5ff93e4c3978832eaf772ff4c3d2315a4 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Thu, 7 Aug 2025 22:02:59 -0700 Subject: [PATCH 229/286] Improve MySQL `CREATE TRIGGER` parsing (#1998) --- src/ast/mod.rs | 20 ++++++++++++++++++-- src/dialect/mssql.rs | 2 ++ src/parser/mod.rs | 22 ++++++++++++++-------- tests/sqlparser_bigquery.rs | 4 ++-- tests/sqlparser_mssql.rs | 2 ++ tests/sqlparser_mysql.rs | 17 +++++++++++------ tests/sqlparser_postgres.rs | 14 +++++++++++++- 7 files changed, 62 insertions(+), 19 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 4a864cfa7..0bf412e8e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3962,6 +3962,15 @@ pub enum Statement { /// EXECUTE FUNCTION trigger_function(); /// ``` period: TriggerPeriod, + /// Whether the trigger period was specified before the target table name. + /// + /// ```sql + /// -- period_before_table == true: Postgres, MySQL, and standard SQL + /// CREATE TRIGGER t BEFORE INSERT ON table_name ...; + /// -- period_before_table == false: MSSQL + /// CREATE TRIGGER t ON table_name BEFORE INSERT ...; + /// ``` + period_before_table: bool, /// Multiple events can be specified using OR, such as `INSERT`, `UPDATE`, `DELETE`, or `TRUNCATE`. events: Vec, /// The table on which the trigger is to be created. @@ -3980,6 +3989,8 @@ pub enum Statement { condition: Option, /// Execute logic block exec_body: Option, + /// For MSSQL and dialects where statements are preceded by `AS` + statements_as: bool, /// For SQL dialects with statement(s) for a body statements: Option, /// The characteristic of the trigger, which include whether the trigger is `DEFERRABLE`, `INITIALLY DEFERRED`, or `INITIALLY IMMEDIATE`, @@ -4944,6 +4955,7 @@ impl fmt::Display for Statement { or_replace, is_constraint, name, + period_before_table, period, events, table_name, @@ -4953,6 +4965,7 @@ impl fmt::Display for Statement { condition, include_each, exec_body, + statements_as, statements, characteristics, } => { @@ -4964,7 +4977,7 @@ impl fmt::Display for Statement { is_constraint = if *is_constraint { "CONSTRAINT " } else { "" }, )?; - if exec_body.is_some() { + if *period_before_table { write!(f, "{period}")?; if !events.is_empty() { write!(f, " {}", display_separated(events, " OR "))?; @@ -5002,7 +5015,10 @@ impl fmt::Display for Statement { write!(f, " EXECUTE {exec_body}")?; } if let Some(statements) = statements { - write!(f, " AS {statements}")?; + if *statements_as { + write!(f, " AS")?; + } + write!(f, " {statements}")?; } Ok(()) } diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 36bd222b8..518dab248 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -257,6 +257,7 @@ impl MsSqlDialect { is_constraint: false, name, period, + period_before_table: false, events, table_name, referenced_table_name: None, @@ -265,6 +266,7 @@ impl MsSqlDialect { include_each: false, condition: None, exec_body: None, + statements_as: true, statements, characteristics: None, }) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d9fe7999e..57da87777 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5593,9 +5593,13 @@ impl<'a> Parser<'a> { .then(|| self.parse_expr()) .transpose()?; - self.expect_keyword_is(Keyword::EXECUTE)?; - - let exec_body = self.parse_trigger_exec_body()?; + let mut exec_body = None; + let mut statements = None; + if self.parse_keyword(Keyword::EXECUTE) { + exec_body = Some(self.parse_trigger_exec_body()?); + } else { + statements = Some(self.parse_conditional_statements(&[Keyword::END])?); + } Ok(Statement::CreateTrigger { or_alter, @@ -5603,6 +5607,7 @@ impl<'a> Parser<'a> { is_constraint, name, period, + period_before_table: true, events, table_name, referenced_table_name, @@ -5610,8 +5615,9 @@ impl<'a> Parser<'a> { trigger_object, include_each, condition, - exec_body: Some(exec_body), - statements: None, + exec_body, + statements_as: false, + statements, characteristics, }) } @@ -6537,7 +6543,7 @@ impl<'a> Parser<'a> { let args = if self.consume_token(&Token::LParen) { if self.consume_token(&Token::RParen) { - None + Some(vec![]) } else { let args = self.parse_comma_separated(Parser::parse_function_arg)?; self.expect_token(&Token::RParen)?; @@ -9307,10 +9313,10 @@ impl<'a> Parser<'a> { }), })) } else { - return self.expected_ref( + self.expected_ref( "{RENAME TO | { RENAME | ADD } VALUE}", self.peek_token_ref(), - ); + ) } } diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 5c14d140c..1ca26ae49 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -2806,9 +2806,9 @@ fn test_export_data() { ); let err = bigquery() - .parse_sql_statements(concat!( + .parse_sql_statements( "EXPORT DATA AS SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10", - )) + ) .unwrap_err(); assert_eq!( err.to_string(), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 50c6448d9..63e4eecb9 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2376,6 +2376,7 @@ fn parse_create_trigger() { is_constraint: false, name: ObjectName::from(vec![Ident::new("reminder1")]), period: TriggerPeriod::After, + period_before_table: false, events: vec![TriggerEvent::Insert, TriggerEvent::Update(vec![]),], table_name: ObjectName::from(vec![Ident::new("Sales"), Ident::new("Customer")]), referenced_table_name: None, @@ -2384,6 +2385,7 @@ fn parse_create_trigger() { include_each: false, condition: None, exec_body: None, + statements_as: true, statements: Some(ConditionalStatements::Sequence { statements: vec![Statement::RaisError { message: Box::new(Expr::Value( diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index cf4e24adb..c13f2266f 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3914,11 +3914,8 @@ fn parse_looks_like_single_line_comment() { #[test] fn parse_create_trigger() { - let sql_create_trigger = r#" - CREATE TRIGGER emp_stamp BEFORE INSERT ON emp - FOR EACH ROW EXECUTE FUNCTION emp_stamp(); - "#; - let create_stmt = mysql().one_statement_parses_to(sql_create_trigger, ""); + let sql_create_trigger = r#"CREATE TRIGGER emp_stamp BEFORE INSERT ON emp FOR EACH ROW EXECUTE FUNCTION emp_stamp()"#; + let create_stmt = mysql().verified_stmt(sql_create_trigger); assert_eq!( create_stmt, Statement::CreateTrigger { @@ -3927,6 +3924,7 @@ fn parse_create_trigger() { is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), period: TriggerPeriod::Before, + period_before_table: true, events: vec![TriggerEvent::Insert], table_name: ObjectName::from(vec![Ident::new("emp")]), referenced_table_name: None, @@ -3938,15 +3936,22 @@ fn parse_create_trigger() { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("emp_stamp")]), - args: None, + args: Some(vec![]), } }), + statements_as: false, statements: None, characteristics: None, } ); } +#[test] +fn parse_create_trigger_compound_statement() { + mysql_and_generic().verified_stmt("CREATE TRIGGER mytrigger BEFORE INSERT ON mytable FOR EACH ROW BEGIN SET NEW.a = 1; SET NEW.b = 2; END"); + mysql_and_generic().verified_stmt("CREATE TRIGGER tr AFTER INSERT ON t1 FOR EACH ROW BEGIN INSERT INTO t2 VALUES (NEW.id); END"); +} + #[test] fn parse_drop_trigger() { let sql_drop_trigger = "DROP TRIGGER emp_stamp;"; diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 6c00b1f11..997384a60 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5552,6 +5552,7 @@ fn parse_create_simple_before_insert_trigger() { is_constraint: false, name: ObjectName::from(vec![Ident::new("check_insert")]), period: TriggerPeriod::Before, + period_before_table: true, events: vec![TriggerEvent::Insert], table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, @@ -5566,6 +5567,7 @@ fn parse_create_simple_before_insert_trigger() { args: None, }, }), + statements_as: false, statements: None, characteristics: None, }; @@ -5582,6 +5584,7 @@ fn parse_create_after_update_trigger_with_condition() { is_constraint: false, name: ObjectName::from(vec![Ident::new("check_update")]), period: TriggerPeriod::After, + period_before_table: true, events: vec![TriggerEvent::Update(vec![])], table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, @@ -5603,6 +5606,7 @@ fn parse_create_after_update_trigger_with_condition() { args: None, }, }), + statements_as: false, statements: None, characteristics: None, }; @@ -5619,6 +5623,7 @@ fn parse_create_instead_of_delete_trigger() { is_constraint: false, name: ObjectName::from(vec![Ident::new("check_delete")]), period: TriggerPeriod::InsteadOf, + period_before_table: true, events: vec![TriggerEvent::Delete], table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, @@ -5633,6 +5638,7 @@ fn parse_create_instead_of_delete_trigger() { args: None, }, }), + statements_as: false, statements: None, characteristics: None, }; @@ -5649,6 +5655,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { is_constraint: true, name: ObjectName::from(vec![Ident::new("check_multiple_events")]), period: TriggerPeriod::Before, + period_before_table: true, events: vec![ TriggerEvent::Insert, TriggerEvent::Update(vec![]), @@ -5667,6 +5674,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { args: None, }, }), + statements_as: false, statements: None, characteristics: Some(ConstraintCharacteristics { deferrable: Some(true), @@ -5687,6 +5695,7 @@ fn parse_create_trigger_with_referencing() { is_constraint: false, name: ObjectName::from(vec![Ident::new("check_referencing")]), period: TriggerPeriod::Before, + period_before_table: true, events: vec![TriggerEvent::Insert], table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, @@ -5712,6 +5721,7 @@ fn parse_create_trigger_with_referencing() { args: None, }, }), + statements_as: false, statements: None, characteristics: None, }; @@ -5994,6 +6004,7 @@ fn parse_trigger_related_functions() { is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), period: TriggerPeriod::Before, + period_before_table: true, events: vec![TriggerEvent::Insert, TriggerEvent::Update(vec![])], table_name: ObjectName::from(vec![Ident::new("emp")]), referenced_table_name: None, @@ -6005,9 +6016,10 @@ fn parse_trigger_related_functions() { exec_type: TriggerExecBodyType::Function, func_desc: FunctionDesc { name: ObjectName::from(vec![Ident::new("emp_stamp")]), - args: None, + args: Some(vec![]), } }), + statements_as: false, statements: None, characteristics: None } From 18b4a14e0f14fad05d6df9adcaeaecc996f54b28 Mon Sep 17 00:00:00 2001 From: tomershaniii <65544633+tomershaniii@users.noreply.github.com> Date: Fri, 8 Aug 2025 12:22:30 +0300 Subject: [PATCH 230/286] Snowflake - support table function in table factor (regression) (#1996) --- src/dialect/snowflake.rs | 2 ++ tests/sqlparser_snowflake.rs | 6 +++++- 2 files changed, 7 insertions(+), 1 deletion(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 697ba3b4f..8830e09a0 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -515,6 +515,8 @@ impl Dialect for SnowflakeDialect { fn is_table_factor(&self, kw: &Keyword, parser: &mut Parser) -> bool { match kw { Keyword::LIMIT if peek_for_limit_options(parser) => false, + // Table function + Keyword::TABLE if matches!(parser.peek_token_ref().token, Token::LParen) => true, _ => !RESERVED_KEYWORDS_FOR_TABLE_FACTOR.contains(kw), } } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index eb4bda7bd..bd8a6d30a 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -3609,9 +3609,13 @@ fn test_sql_keywords_as_table_aliases() { #[test] fn test_sql_keywords_as_table_factor() { // LIMIT is a table factor, Snowflake does not reserve it - snowflake().one_statement_parses_to("SELECT * FROM tbl, LIMIT", "SELECT * FROM tbl, LIMIT"); + snowflake().verified_stmt("SELECT * FROM tbl, LIMIT"); // LIMIT is not a table factor snowflake().one_statement_parses_to("SELECT * FROM tbl, LIMIT 1", "SELECT * FROM tbl LIMIT 1"); + + // Table functions are table factors + snowflake().verified_stmt("SELECT 1 FROM TABLE(GENERATOR(ROWCOUNT => 10)) AS a, TABLE(GENERATOR(ROWCOUNT => 10)) AS b"); + // ORDER is reserved assert!(snowflake() .parse_sql_statements("SELECT * FROM tbl, order") From 67fca82495362fb824a459d0ce8f3e880a3d71b5 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Fri, 8 Aug 2025 02:32:20 -0700 Subject: [PATCH 231/286] Improve MySQL option parsing in index definitions (#1997) --- src/ast/ddl.rs | 502 ++++++++++++++++++++++++++- src/ast/dml.rs | 470 +------------------------ src/ast/helpers/stmt_create_table.rs | 3 +- src/ast/mod.rs | 14 +- src/ast/spans.rs | 6 +- src/parser/mod.rs | 27 ++ tests/sqlparser_common.rs | 12 +- tests/sqlparser_mysql.rs | 28 ++ tests/sqlparser_postgres.rs | 64 +++- 9 files changed, 628 insertions(+), 498 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 4d1fc4628..1c2aaf48d 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -31,14 +31,36 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, - CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, - FunctionDeterminismSpecifier, FunctionParallel, Ident, IndexColumn, MySQLColumnPosition, - ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, Spanned, - SqlOption, Tag, Value, ValueWithSpan, + CreateFunctionUsing, CreateTableOptions, DataType, Expr, FileFormat, FunctionBehavior, + FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, HiveDistributionStyle, + HiveFormat, HiveIOFormat, HiveRowFormat, Ident, MySQLColumnPosition, ObjectName, OnCommit, + OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RowAccessPolicy, + SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, Tag, Value, ValueWithSpan, + WrappedCollection, }; +use crate::display_utils::{DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; use crate::keywords::Keyword; use crate::tokenizer::{Span, Token}; +/// Index column type. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct IndexColumn { + pub column: OrderByExpr, + pub operator_class: Option, +} + +impl fmt::Display for IndexColumn { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}", self.column)?; + if let Some(operator_class) = &self.operator_class { + write!(f, " {operator_class}")?; + } + Ok(()) + } +} + /// ALTER TABLE operation REPLICA IDENTITY values /// See [Postgres ALTER TABLE docs](https://www.postgresql.org/docs/current/sql-altertable.html) #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -1103,6 +1125,8 @@ pub enum TableConstraint { index_type: Option, /// Referred column identifier list. columns: Vec, + /// Optional index options such as `USING`; see [`IndexOption`]. + index_options: Vec, }, /// MySQLs [fulltext][1] definition. Since the [`SPATIAL`][2] definition is exactly the same, /// and MySQL displays both the same way, it is part of this definition as well. @@ -1231,6 +1255,7 @@ impl fmt::Display for TableConstraint { name, index_type, columns, + index_options, } => { write!(f, "{}", if *display_as_key { "KEY" } else { "INDEX" })?; if let Some(name) = name { @@ -1240,7 +1265,9 @@ impl fmt::Display for TableConstraint { write!(f, " USING {index_type}")?; } write!(f, " ({})", display_comma_separated(columns))?; - + if !index_options.is_empty() { + write!(f, " {}", display_comma_separated(index_options))?; + } Ok(()) } Self::FulltextOrSpatial { @@ -1355,17 +1382,20 @@ impl fmt::Display for IndexType { } } -/// MySQLs index option. -/// -/// This structure used here [`MySQL` CREATE TABLE][1], [`MySQL` CREATE INDEX][2]. +/// MySQL index option, used in [`CREATE TABLE`], [`CREATE INDEX`], and [`ALTER TABLE`]. /// -/// [1]: https://dev.mysql.com/doc/refman/8.3/en/create-table.html -/// [2]: https://dev.mysql.com/doc/refman/8.3/en/create-index.html +/// [`CREATE TABLE`]: https://dev.mysql.com/doc/refman/8.4/en/create-table.html +/// [`CREATE INDEX`]: https://dev.mysql.com/doc/refman/8.4/en/create-index.html +/// [`ALTER TABLE`]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum IndexOption { + /// `USING { BTREE | HASH }`: Index type to use for the index. + /// + /// Note that we permissively parse non-MySQL index types, like `GIN`. Using(IndexType), + /// `COMMENT 'string'`: Specifies a comment for the index. Comment(String), } @@ -2294,6 +2324,458 @@ impl fmt::Display for ClusteredBy { } } +/// CREATE INDEX statement. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateIndex { + /// index name + pub name: Option, + #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] + pub table_name: ObjectName, + pub using: Option, + pub columns: Vec, + pub unique: bool, + pub concurrently: bool, + pub if_not_exists: bool, + pub include: Vec, + pub nulls_distinct: Option, + /// WITH clause: + pub with: Vec, + pub predicate: Option, + pub index_options: Vec, + /// [MySQL] allows a subset of options normally used for `ALTER TABLE`: + /// + /// - `ALGORITHM` + /// - `LOCK` + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/create-index.html + pub alter_options: Vec, +} + +impl fmt::Display for CreateIndex { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "CREATE {unique}INDEX {concurrently}{if_not_exists}", + unique = if self.unique { "UNIQUE " } else { "" }, + concurrently = if self.concurrently { + "CONCURRENTLY " + } else { + "" + }, + if_not_exists = if self.if_not_exists { + "IF NOT EXISTS " + } else { + "" + }, + )?; + if let Some(value) = &self.name { + write!(f, "{value} ")?; + } + write!(f, "ON {}", self.table_name)?; + if let Some(value) = &self.using { + write!(f, " USING {value} ")?; + } + write!(f, "({})", display_comma_separated(&self.columns))?; + if !self.include.is_empty() { + write!(f, " INCLUDE ({})", display_comma_separated(&self.include))?; + } + if let Some(value) = self.nulls_distinct { + if value { + write!(f, " NULLS DISTINCT")?; + } else { + write!(f, " NULLS NOT DISTINCT")?; + } + } + if !self.with.is_empty() { + write!(f, " WITH ({})", display_comma_separated(&self.with))?; + } + if let Some(predicate) = &self.predicate { + write!(f, " WHERE {predicate}")?; + } + if !self.index_options.is_empty() { + write!(f, " {}", display_separated(&self.index_options, " "))?; + } + if !self.alter_options.is_empty() { + write!(f, " {}", display_separated(&self.alter_options, " "))?; + } + Ok(()) + } +} + +/// CREATE TABLE statement. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateTable { + pub or_replace: bool, + pub temporary: bool, + pub external: bool, + pub global: Option, + pub if_not_exists: bool, + pub transient: bool, + pub volatile: bool, + pub iceberg: bool, + /// Table name + #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] + pub name: ObjectName, + /// Optional schema + pub columns: Vec, + pub constraints: Vec, + pub hive_distribution: HiveDistributionStyle, + pub hive_formats: Option, + pub table_options: CreateTableOptions, + pub file_format: Option, + pub location: Option, + pub query: Option>, + pub without_rowid: bool, + pub like: Option, + pub clone: Option, + // For Hive dialect, the table comment is after the column definitions without `=`, + // so the `comment` field is optional and different than the comment field in the general options list. + // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) + pub comment: Option, + pub on_commit: Option, + /// ClickHouse "ON CLUSTER" clause: + /// + pub on_cluster: Option, + /// ClickHouse "PRIMARY KEY " clause. + /// + pub primary_key: Option>, + /// ClickHouse "ORDER BY " clause. Note that omitted ORDER BY is different + /// than empty (represented as ()), the latter meaning "no sorting". + /// + pub order_by: Option>, + /// BigQuery: A partition expression for the table. + /// + pub partition_by: Option>, + /// BigQuery: Table clustering column list. + /// + /// Snowflake: Table clustering list which contains base column, expressions on base columns. + /// + pub cluster_by: Option>>, + /// Hive: Table clustering column list. + /// + pub clustered_by: Option, + /// Postgres `INHERITs` clause, which contains the list of tables from which + /// the new table inherits. + /// + /// + pub inherits: Option>, + /// SQLite "STRICT" clause. + /// if the "STRICT" table-option keyword is added to the end, after the closing ")", + /// then strict typing rules apply to that table. + pub strict: bool, + /// Snowflake "COPY GRANTS" clause + /// + pub copy_grants: bool, + /// Snowflake "ENABLE_SCHEMA_EVOLUTION" clause + /// + pub enable_schema_evolution: Option, + /// Snowflake "CHANGE_TRACKING" clause + /// + pub change_tracking: Option, + /// Snowflake "DATA_RETENTION_TIME_IN_DAYS" clause + /// + pub data_retention_time_in_days: Option, + /// Snowflake "MAX_DATA_EXTENSION_TIME_IN_DAYS" clause + /// + pub max_data_extension_time_in_days: Option, + /// Snowflake "DEFAULT_DDL_COLLATION" clause + /// + pub default_ddl_collation: Option, + /// Snowflake "WITH AGGREGATION POLICY" clause + /// + pub with_aggregation_policy: Option, + /// Snowflake "WITH ROW ACCESS POLICY" clause + /// + pub with_row_access_policy: Option, + /// Snowflake "WITH TAG" clause + /// + pub with_tags: Option>, + /// Snowflake "EXTERNAL_VOLUME" clause for Iceberg tables + /// + pub external_volume: Option, + /// Snowflake "BASE_LOCATION" clause for Iceberg tables + /// + pub base_location: Option, + /// Snowflake "CATALOG" clause for Iceberg tables + /// + pub catalog: Option, + /// Snowflake "CATALOG_SYNC" clause for Iceberg tables + /// + pub catalog_sync: Option, + /// Snowflake "STORAGE_SERIALIZATION_POLICY" clause for Iceberg tables + /// + pub storage_serialization_policy: Option, +} + +impl fmt::Display for CreateTable { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + // We want to allow the following options + // Empty column list, allowed by PostgreSQL: + // `CREATE TABLE t ()` + // No columns provided for CREATE TABLE AS: + // `CREATE TABLE t AS SELECT a from t2` + // Columns provided for CREATE TABLE AS: + // `CREATE TABLE t (a INT) AS SELECT a from t2` + write!( + f, + "CREATE {or_replace}{external}{global}{temporary}{transient}{volatile}{iceberg}TABLE {if_not_exists}{name}", + or_replace = if self.or_replace { "OR REPLACE " } else { "" }, + external = if self.external { "EXTERNAL " } else { "" }, + global = self.global + .map(|global| { + if global { + "GLOBAL " + } else { + "LOCAL " + } + }) + .unwrap_or(""), + if_not_exists = if self.if_not_exists { "IF NOT EXISTS " } else { "" }, + temporary = if self.temporary { "TEMPORARY " } else { "" }, + transient = if self.transient { "TRANSIENT " } else { "" }, + volatile = if self.volatile { "VOLATILE " } else { "" }, + // Only for Snowflake + iceberg = if self.iceberg { "ICEBERG " } else { "" }, + name = self.name, + )?; + if let Some(on_cluster) = &self.on_cluster { + write!(f, " ON CLUSTER {on_cluster}")?; + } + if !self.columns.is_empty() || !self.constraints.is_empty() { + f.write_str(" (")?; + NewLine.fmt(f)?; + Indent(DisplayCommaSeparated(&self.columns)).fmt(f)?; + if !self.columns.is_empty() && !self.constraints.is_empty() { + f.write_str(",")?; + SpaceOrNewline.fmt(f)?; + } + Indent(DisplayCommaSeparated(&self.constraints)).fmt(f)?; + NewLine.fmt(f)?; + f.write_str(")")?; + } else if self.query.is_none() && self.like.is_none() && self.clone.is_none() { + // PostgreSQL allows `CREATE TABLE t ();`, but requires empty parens + f.write_str(" ()")?; + } + + // Hive table comment should be after column definitions, please refer to: + // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) + if let Some(comment) = &self.comment { + write!(f, " COMMENT '{comment}'")?; + } + + // Only for SQLite + if self.without_rowid { + write!(f, " WITHOUT ROWID")?; + } + + // Only for Hive + if let Some(l) = &self.like { + write!(f, " LIKE {l}")?; + } + + if let Some(c) = &self.clone { + write!(f, " CLONE {c}")?; + } + + match &self.hive_distribution { + HiveDistributionStyle::PARTITIONED { columns } => { + write!(f, " PARTITIONED BY ({})", display_comma_separated(columns))?; + } + HiveDistributionStyle::SKEWED { + columns, + on, + stored_as_directories, + } => { + write!( + f, + " SKEWED BY ({})) ON ({})", + display_comma_separated(columns), + display_comma_separated(on) + )?; + if *stored_as_directories { + write!(f, " STORED AS DIRECTORIES")?; + } + } + _ => (), + } + + if let Some(clustered_by) = &self.clustered_by { + write!(f, " {clustered_by}")?; + } + + if let Some(HiveFormat { + row_format, + serde_properties, + storage, + location, + }) = &self.hive_formats + { + match row_format { + Some(HiveRowFormat::SERDE { class }) => write!(f, " ROW FORMAT SERDE '{class}'")?, + Some(HiveRowFormat::DELIMITED { delimiters }) => { + write!(f, " ROW FORMAT DELIMITED")?; + if !delimiters.is_empty() { + write!(f, " {}", display_separated(delimiters, " "))?; + } + } + None => (), + } + match storage { + Some(HiveIOFormat::IOF { + input_format, + output_format, + }) => write!( + f, + " STORED AS INPUTFORMAT {input_format} OUTPUTFORMAT {output_format}" + )?, + Some(HiveIOFormat::FileFormat { format }) if !self.external => { + write!(f, " STORED AS {format}")? + } + _ => (), + } + if let Some(serde_properties) = serde_properties.as_ref() { + write!( + f, + " WITH SERDEPROPERTIES ({})", + display_comma_separated(serde_properties) + )?; + } + if !self.external { + if let Some(loc) = location { + write!(f, " LOCATION '{loc}'")?; + } + } + } + if self.external { + if let Some(file_format) = self.file_format { + write!(f, " STORED AS {file_format}")?; + } + write!(f, " LOCATION '{}'", self.location.as_ref().unwrap())?; + } + + match &self.table_options { + options @ CreateTableOptions::With(_) + | options @ CreateTableOptions::Plain(_) + | options @ CreateTableOptions::TableProperties(_) => write!(f, " {options}")?, + _ => (), + } + + if let Some(primary_key) = &self.primary_key { + write!(f, " PRIMARY KEY {primary_key}")?; + } + if let Some(order_by) = &self.order_by { + write!(f, " ORDER BY {order_by}")?; + } + if let Some(inherits) = &self.inherits { + write!(f, " INHERITS ({})", display_comma_separated(inherits))?; + } + if let Some(partition_by) = self.partition_by.as_ref() { + write!(f, " PARTITION BY {partition_by}")?; + } + if let Some(cluster_by) = self.cluster_by.as_ref() { + write!(f, " CLUSTER BY {cluster_by}")?; + } + if let options @ CreateTableOptions::Options(_) = &self.table_options { + write!(f, " {options}")?; + } + if let Some(external_volume) = self.external_volume.as_ref() { + write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; + } + + if let Some(catalog) = self.catalog.as_ref() { + write!(f, " CATALOG = '{catalog}'")?; + } + + if self.iceberg { + if let Some(base_location) = self.base_location.as_ref() { + write!(f, " BASE_LOCATION = '{base_location}'")?; + } + } + + if let Some(catalog_sync) = self.catalog_sync.as_ref() { + write!(f, " CATALOG_SYNC = '{catalog_sync}'")?; + } + + if let Some(storage_serialization_policy) = self.storage_serialization_policy.as_ref() { + write!( + f, + " STORAGE_SERIALIZATION_POLICY = {storage_serialization_policy}" + )?; + } + + if self.copy_grants { + write!(f, " COPY GRANTS")?; + } + + if let Some(is_enabled) = self.enable_schema_evolution { + write!( + f, + " ENABLE_SCHEMA_EVOLUTION={}", + if is_enabled { "TRUE" } else { "FALSE" } + )?; + } + + if let Some(is_enabled) = self.change_tracking { + write!( + f, + " CHANGE_TRACKING={}", + if is_enabled { "TRUE" } else { "FALSE" } + )?; + } + + if let Some(data_retention_time_in_days) = self.data_retention_time_in_days { + write!( + f, + " DATA_RETENTION_TIME_IN_DAYS={data_retention_time_in_days}", + )?; + } + + if let Some(max_data_extension_time_in_days) = self.max_data_extension_time_in_days { + write!( + f, + " MAX_DATA_EXTENSION_TIME_IN_DAYS={max_data_extension_time_in_days}", + )?; + } + + if let Some(default_ddl_collation) = &self.default_ddl_collation { + write!(f, " DEFAULT_DDL_COLLATION='{default_ddl_collation}'",)?; + } + + if let Some(with_aggregation_policy) = &self.with_aggregation_policy { + write!(f, " WITH AGGREGATION POLICY {with_aggregation_policy}",)?; + } + + if let Some(row_access_policy) = &self.with_row_access_policy { + write!(f, " {row_access_policy}",)?; + } + + if let Some(tag) = &self.with_tags { + write!(f, " WITH TAG ({})", display_comma_separated(tag.as_slice()))?; + } + + if self.on_commit.is_some() { + let on_commit = match self.on_commit { + Some(OnCommit::DeleteRows) => "ON COMMIT DELETE ROWS", + Some(OnCommit::PreserveRows) => "ON COMMIT PRESERVE ROWS", + Some(OnCommit::Drop) => "ON COMMIT DROP", + None => "", + }; + write!(f, " {on_commit}")?; + } + if self.strict { + write!(f, " STRICT")?; + } + if let Some(query) = &self.query { + write!(f, " AS {query}")?; + } + Ok(()) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] diff --git a/src/ast/dml.rs b/src/ast/dml.rs index e179f5d70..63d6b86c7 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -29,476 +29,14 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use crate::display_utils::{indented_list, DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; - -pub use super::ddl::{ColumnDef, TableConstraint}; +use crate::display_utils::{indented_list, Indent, SpaceOrNewline}; use super::{ - display_comma_separated, display_separated, query::InputFormatClause, Assignment, ClusteredBy, - CommentDef, CreateTableOptions, Expr, FileFormat, FromTable, HiveDistributionStyle, HiveFormat, - HiveIOFormat, HiveRowFormat, Ident, IndexType, InsertAliases, MysqlInsertPriority, ObjectName, - OnCommit, OnInsert, OneOrManyWithParens, OrderByExpr, Query, RowAccessPolicy, SelectItem, - Setting, SqliteOnConflict, StorageSerializationPolicy, TableObject, TableWithJoins, Tag, - WrappedCollection, + display_comma_separated, query::InputFormatClause, Assignment, Expr, FromTable, Ident, + InsertAliases, MysqlInsertPriority, ObjectName, OnInsert, OrderByExpr, Query, SelectItem, + Setting, SqliteOnConflict, TableObject, TableWithJoins, }; -/// Index column type. -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct IndexColumn { - pub column: OrderByExpr, - pub operator_class: Option, -} - -impl Display for IndexColumn { - fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{}", self.column)?; - if let Some(operator_class) = &self.operator_class { - write!(f, " {operator_class}")?; - } - Ok(()) - } -} - -/// CREATE INDEX statement. -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct CreateIndex { - /// index name - pub name: Option, - #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] - pub table_name: ObjectName, - pub using: Option, - pub columns: Vec, - pub unique: bool, - pub concurrently: bool, - pub if_not_exists: bool, - pub include: Vec, - pub nulls_distinct: Option, - /// WITH clause: - pub with: Vec, - pub predicate: Option, -} - -impl Display for CreateIndex { - fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!( - f, - "CREATE {unique}INDEX {concurrently}{if_not_exists}", - unique = if self.unique { "UNIQUE " } else { "" }, - concurrently = if self.concurrently { - "CONCURRENTLY " - } else { - "" - }, - if_not_exists = if self.if_not_exists { - "IF NOT EXISTS " - } else { - "" - }, - )?; - if let Some(value) = &self.name { - write!(f, "{value} ")?; - } - write!(f, "ON {}", self.table_name)?; - if let Some(value) = &self.using { - write!(f, " USING {value} ")?; - } - write!(f, "({})", display_separated(&self.columns, ","))?; - if !self.include.is_empty() { - write!(f, " INCLUDE ({})", display_separated(&self.include, ","))?; - } - if let Some(value) = self.nulls_distinct { - if value { - write!(f, " NULLS DISTINCT")?; - } else { - write!(f, " NULLS NOT DISTINCT")?; - } - } - if !self.with.is_empty() { - write!(f, " WITH ({})", display_comma_separated(&self.with))?; - } - if let Some(predicate) = &self.predicate { - write!(f, " WHERE {predicate}")?; - } - Ok(()) - } -} - -/// CREATE TABLE statement. -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct CreateTable { - pub or_replace: bool, - pub temporary: bool, - pub external: bool, - pub global: Option, - pub if_not_exists: bool, - pub transient: bool, - pub volatile: bool, - pub iceberg: bool, - /// Table name - #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] - pub name: ObjectName, - /// Optional schema - pub columns: Vec, - pub constraints: Vec, - pub hive_distribution: HiveDistributionStyle, - pub hive_formats: Option, - pub table_options: CreateTableOptions, - pub file_format: Option, - pub location: Option, - pub query: Option>, - pub without_rowid: bool, - pub like: Option, - pub clone: Option, - // For Hive dialect, the table comment is after the column definitions without `=`, - // so the `comment` field is optional and different than the comment field in the general options list. - // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) - pub comment: Option, - pub on_commit: Option, - /// ClickHouse "ON CLUSTER" clause: - /// - pub on_cluster: Option, - /// ClickHouse "PRIMARY KEY " clause. - /// - pub primary_key: Option>, - /// ClickHouse "ORDER BY " clause. Note that omitted ORDER BY is different - /// than empty (represented as ()), the latter meaning "no sorting". - /// - pub order_by: Option>, - /// BigQuery: A partition expression for the table. - /// - pub partition_by: Option>, - /// BigQuery: Table clustering column list. - /// - /// Snowflake: Table clustering list which contains base column, expressions on base columns. - /// - pub cluster_by: Option>>, - /// Hive: Table clustering column list. - /// - pub clustered_by: Option, - /// Postgres `INHERITs` clause, which contains the list of tables from which - /// the new table inherits. - /// - /// - pub inherits: Option>, - /// SQLite "STRICT" clause. - /// if the "STRICT" table-option keyword is added to the end, after the closing ")", - /// then strict typing rules apply to that table. - pub strict: bool, - /// Snowflake "COPY GRANTS" clause - /// - pub copy_grants: bool, - /// Snowflake "ENABLE_SCHEMA_EVOLUTION" clause - /// - pub enable_schema_evolution: Option, - /// Snowflake "CHANGE_TRACKING" clause - /// - pub change_tracking: Option, - /// Snowflake "DATA_RETENTION_TIME_IN_DAYS" clause - /// - pub data_retention_time_in_days: Option, - /// Snowflake "MAX_DATA_EXTENSION_TIME_IN_DAYS" clause - /// - pub max_data_extension_time_in_days: Option, - /// Snowflake "DEFAULT_DDL_COLLATION" clause - /// - pub default_ddl_collation: Option, - /// Snowflake "WITH AGGREGATION POLICY" clause - /// - pub with_aggregation_policy: Option, - /// Snowflake "WITH ROW ACCESS POLICY" clause - /// - pub with_row_access_policy: Option, - /// Snowflake "WITH TAG" clause - /// - pub with_tags: Option>, - /// Snowflake "EXTERNAL_VOLUME" clause for Iceberg tables - /// - pub external_volume: Option, - /// Snowflake "BASE_LOCATION" clause for Iceberg tables - /// - pub base_location: Option, - /// Snowflake "CATALOG" clause for Iceberg tables - /// - pub catalog: Option, - /// Snowflake "CATALOG_SYNC" clause for Iceberg tables - /// - pub catalog_sync: Option, - /// Snowflake "STORAGE_SERIALIZATION_POLICY" clause for Iceberg tables - /// - pub storage_serialization_policy: Option, -} - -impl Display for CreateTable { - fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - // We want to allow the following options - // Empty column list, allowed by PostgreSQL: - // `CREATE TABLE t ()` - // No columns provided for CREATE TABLE AS: - // `CREATE TABLE t AS SELECT a from t2` - // Columns provided for CREATE TABLE AS: - // `CREATE TABLE t (a INT) AS SELECT a from t2` - write!( - f, - "CREATE {or_replace}{external}{global}{temporary}{transient}{volatile}{iceberg}TABLE {if_not_exists}{name}", - or_replace = if self.or_replace { "OR REPLACE " } else { "" }, - external = if self.external { "EXTERNAL " } else { "" }, - global = self.global - .map(|global| { - if global { - "GLOBAL " - } else { - "LOCAL " - } - }) - .unwrap_or(""), - if_not_exists = if self.if_not_exists { "IF NOT EXISTS " } else { "" }, - temporary = if self.temporary { "TEMPORARY " } else { "" }, - transient = if self.transient { "TRANSIENT " } else { "" }, - volatile = if self.volatile { "VOLATILE " } else { "" }, - // Only for Snowflake - iceberg = if self.iceberg { "ICEBERG " } else { "" }, - name = self.name, - )?; - if let Some(on_cluster) = &self.on_cluster { - write!(f, " ON CLUSTER {on_cluster}")?; - } - if !self.columns.is_empty() || !self.constraints.is_empty() { - f.write_str(" (")?; - NewLine.fmt(f)?; - Indent(DisplayCommaSeparated(&self.columns)).fmt(f)?; - if !self.columns.is_empty() && !self.constraints.is_empty() { - f.write_str(",")?; - SpaceOrNewline.fmt(f)?; - } - Indent(DisplayCommaSeparated(&self.constraints)).fmt(f)?; - NewLine.fmt(f)?; - f.write_str(")")?; - } else if self.query.is_none() && self.like.is_none() && self.clone.is_none() { - // PostgreSQL allows `CREATE TABLE t ();`, but requires empty parens - f.write_str(" ()")?; - } - - // Hive table comment should be after column definitions, please refer to: - // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) - if let Some(comment) = &self.comment { - write!(f, " COMMENT '{comment}'")?; - } - - // Only for SQLite - if self.without_rowid { - write!(f, " WITHOUT ROWID")?; - } - - // Only for Hive - if let Some(l) = &self.like { - write!(f, " LIKE {l}")?; - } - - if let Some(c) = &self.clone { - write!(f, " CLONE {c}")?; - } - - match &self.hive_distribution { - HiveDistributionStyle::PARTITIONED { columns } => { - write!(f, " PARTITIONED BY ({})", display_comma_separated(columns))?; - } - HiveDistributionStyle::SKEWED { - columns, - on, - stored_as_directories, - } => { - write!( - f, - " SKEWED BY ({})) ON ({})", - display_comma_separated(columns), - display_comma_separated(on) - )?; - if *stored_as_directories { - write!(f, " STORED AS DIRECTORIES")?; - } - } - _ => (), - } - - if let Some(clustered_by) = &self.clustered_by { - write!(f, " {clustered_by}")?; - } - - if let Some(HiveFormat { - row_format, - serde_properties, - storage, - location, - }) = &self.hive_formats - { - match row_format { - Some(HiveRowFormat::SERDE { class }) => write!(f, " ROW FORMAT SERDE '{class}'")?, - Some(HiveRowFormat::DELIMITED { delimiters }) => { - write!(f, " ROW FORMAT DELIMITED")?; - if !delimiters.is_empty() { - write!(f, " {}", display_separated(delimiters, " "))?; - } - } - None => (), - } - match storage { - Some(HiveIOFormat::IOF { - input_format, - output_format, - }) => write!( - f, - " STORED AS INPUTFORMAT {input_format} OUTPUTFORMAT {output_format}" - )?, - Some(HiveIOFormat::FileFormat { format }) if !self.external => { - write!(f, " STORED AS {format}")? - } - _ => (), - } - if let Some(serde_properties) = serde_properties.as_ref() { - write!( - f, - " WITH SERDEPROPERTIES ({})", - display_comma_separated(serde_properties) - )?; - } - if !self.external { - if let Some(loc) = location { - write!(f, " LOCATION '{loc}'")?; - } - } - } - if self.external { - if let Some(file_format) = self.file_format { - write!(f, " STORED AS {file_format}")?; - } - write!(f, " LOCATION '{}'", self.location.as_ref().unwrap())?; - } - - match &self.table_options { - options @ CreateTableOptions::With(_) - | options @ CreateTableOptions::Plain(_) - | options @ CreateTableOptions::TableProperties(_) => write!(f, " {options}")?, - _ => (), - } - - if let Some(primary_key) = &self.primary_key { - write!(f, " PRIMARY KEY {primary_key}")?; - } - if let Some(order_by) = &self.order_by { - write!(f, " ORDER BY {order_by}")?; - } - if let Some(inherits) = &self.inherits { - write!(f, " INHERITS ({})", display_comma_separated(inherits))?; - } - if let Some(partition_by) = self.partition_by.as_ref() { - write!(f, " PARTITION BY {partition_by}")?; - } - if let Some(cluster_by) = self.cluster_by.as_ref() { - write!(f, " CLUSTER BY {cluster_by}")?; - } - if let options @ CreateTableOptions::Options(_) = &self.table_options { - write!(f, " {options}")?; - } - if let Some(external_volume) = self.external_volume.as_ref() { - write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; - } - - if let Some(catalog) = self.catalog.as_ref() { - write!(f, " CATALOG = '{catalog}'")?; - } - - if self.iceberg { - if let Some(base_location) = self.base_location.as_ref() { - write!(f, " BASE_LOCATION = '{base_location}'")?; - } - } - - if let Some(catalog_sync) = self.catalog_sync.as_ref() { - write!(f, " CATALOG_SYNC = '{catalog_sync}'")?; - } - - if let Some(storage_serialization_policy) = self.storage_serialization_policy.as_ref() { - write!( - f, - " STORAGE_SERIALIZATION_POLICY = {storage_serialization_policy}" - )?; - } - - if self.copy_grants { - write!(f, " COPY GRANTS")?; - } - - if let Some(is_enabled) = self.enable_schema_evolution { - write!( - f, - " ENABLE_SCHEMA_EVOLUTION={}", - if is_enabled { "TRUE" } else { "FALSE" } - )?; - } - - if let Some(is_enabled) = self.change_tracking { - write!( - f, - " CHANGE_TRACKING={}", - if is_enabled { "TRUE" } else { "FALSE" } - )?; - } - - if let Some(data_retention_time_in_days) = self.data_retention_time_in_days { - write!( - f, - " DATA_RETENTION_TIME_IN_DAYS={data_retention_time_in_days}", - )?; - } - - if let Some(max_data_extension_time_in_days) = self.max_data_extension_time_in_days { - write!( - f, - " MAX_DATA_EXTENSION_TIME_IN_DAYS={max_data_extension_time_in_days}", - )?; - } - - if let Some(default_ddl_collation) = &self.default_ddl_collation { - write!(f, " DEFAULT_DDL_COLLATION='{default_ddl_collation}'",)?; - } - - if let Some(with_aggregation_policy) = &self.with_aggregation_policy { - write!(f, " WITH AGGREGATION POLICY {with_aggregation_policy}",)?; - } - - if let Some(row_access_policy) = &self.with_row_access_policy { - write!(f, " {row_access_policy}",)?; - } - - if let Some(tag) = &self.with_tags { - write!(f, " WITH TAG ({})", display_comma_separated(tag.as_slice()))?; - } - - if self.on_commit.is_some() { - let on_commit = match self.on_commit { - Some(OnCommit::DeleteRows) => "ON COMMIT DELETE ROWS", - Some(OnCommit::PreserveRows) => "ON COMMIT PRESERVE ROWS", - Some(OnCommit::Drop) => "ON COMMIT DROP", - None => "", - }; - write!(f, " {on_commit}")?; - } - if self.strict { - write!(f, " STRICT")?; - } - if let Some(query) = &self.query { - write!(f, " AS {query}")?; - } - Ok(()) - } -} - /// INSERT statement. #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index 60b8fb2a0..c727276d3 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -24,9 +24,8 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use super::super::dml::CreateTable; use crate::ast::{ - ClusteredBy, ColumnDef, CommentDef, CreateTableOptions, Expr, FileFormat, + ClusteredBy, ColumnDef, CommentDef, CreateTable, CreateTableOptions, Expr, FileFormat, HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, OneOrManyWithParens, Query, RowAccessPolicy, Statement, StorageSerializationPolicy, TableConstraint, Tag, WrappedCollection, diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 0bf412e8e..d21f83574 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -63,14 +63,14 @@ pub use self::ddl::{ AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, - Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, - IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, - IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, - Partition, ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, - TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, - UserDefinedTypeRepresentation, ViewColumnDef, + CreateIndex, CreateTable, Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, + GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, + IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, IndexOption, IndexType, + KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, + RenameTableNameKind, ReplicaIdentity, TableConstraint, TagsColumnOption, + UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; -pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; +pub use self::dml::{Delete, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; pub use self::query::{ AfterMatchSkip, ConnectBy, Cte, CteAsMaterialized, Distinct, EmptyMatchesMode, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index dec26566c..e6e601c33 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -713,6 +713,7 @@ impl Spanned for TableConstraint { name, index_type: _, columns, + index_options: _, } => union_spans( name.iter() .map(|i| i.span) @@ -747,6 +748,8 @@ impl Spanned for CreateIndex { nulls_distinct: _, // bool with, predicate, + index_options: _, + alter_options, } = self; union_spans( @@ -756,7 +759,8 @@ impl Spanned for CreateIndex { .chain(columns.iter().map(|i| i.column.span())) .chain(include.iter().map(|i| i.span)) .chain(with.iter().map(|i| i.span())) - .chain(predicate.iter().map(|i| i.span())), + .chain(predicate.iter().map(|i| i.span())) + .chain(alter_options.iter().map(|i| i.span())), ) } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 57da87777..78ce6de6d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7083,6 +7083,22 @@ impl<'a> Parser<'a> { None }; + // MySQL options (including the modern style of `USING` after the column list instead of + // before, which is deprecated) shouldn't conflict with other preceding options (e.g. `WITH + // PARSER` won't be caught by the above `WITH` clause parsing because MySQL doesn't set that + // support flag). This is probably invalid syntax for other dialects, but it is simpler to + // parse it anyway (as we do inside `ALTER TABLE` and `CREATE TABLE` parsing). + let index_options = self.parse_index_options()?; + + // MySQL allows `ALGORITHM` and `LOCK` options. Unlike in `ALTER TABLE`, they need not be comma separated. + let mut alter_options = Vec::new(); + while self + .peek_one_of_keywords(&[Keyword::ALGORITHM, Keyword::LOCK]) + .is_some() + { + alter_options.push(self.parse_alter_table_operation()?) + } + Ok(Statement::CreateIndex(CreateIndex { name: index_name, table_name, @@ -7095,6 +7111,8 @@ impl<'a> Parser<'a> { nulls_distinct, with, predicate, + index_options, + alter_options, })) } @@ -8407,12 +8425,14 @@ impl<'a> Parser<'a> { let index_type = self.parse_optional_using_then_index_type()?; let columns = self.parse_parenthesized_index_column_list()?; + let index_options = self.parse_index_options()?; Ok(Some(TableConstraint::Index { display_as_key, name, index_type, columns, + index_options, })) } Token::Word(w) @@ -17475,6 +17495,7 @@ mod tests { name: None, index_type: None, columns: vec![mk_expected_col("c1")], + index_options: vec![], } ); @@ -17486,6 +17507,7 @@ mod tests { name: None, index_type: None, columns: vec![mk_expected_col("c1")], + index_options: vec![], } ); @@ -17497,6 +17519,7 @@ mod tests { name: Some(Ident::with_quote('\'', "index")), index_type: None, columns: vec![mk_expected_col("c1"), mk_expected_col("c2")], + index_options: vec![], } ); @@ -17508,6 +17531,7 @@ mod tests { name: None, index_type: Some(IndexType::BTree), columns: vec![mk_expected_col("c1")], + index_options: vec![], } ); @@ -17519,6 +17543,7 @@ mod tests { name: None, index_type: Some(IndexType::Hash), columns: vec![mk_expected_col("c1")], + index_options: vec![], } ); @@ -17530,6 +17555,7 @@ mod tests { name: Some(Ident::new("idx_name")), index_type: Some(IndexType::BTree), columns: vec![mk_expected_col("c1")], + index_options: vec![], } ); @@ -17541,6 +17567,7 @@ mod tests { name: Some(Ident::new("idx_name")), index_type: Some(IndexType::Hash), columns: vec![mk_expected_col("c1")], + index_options: vec![], } ); } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 7eff55039..d71449d92 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9189,7 +9189,7 @@ fn ensure_multiple_dialects_are_tested() { #[test] fn parse_create_index() { - let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test(name,age DESC)"; + let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test(name, age DESC)"; let indexed_columns: Vec = vec![ IndexColumn { operator_class: None, @@ -9235,7 +9235,7 @@ fn parse_create_index() { #[test] fn test_create_index_with_using_function() { - let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test USING BTREE (name,age DESC)"; + let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test USING BTREE (name, age DESC)"; let indexed_columns: Vec = vec![ IndexColumn { operator_class: None, @@ -9273,6 +9273,8 @@ fn test_create_index_with_using_function() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq!("idx_name", name.to_string()); assert_eq!("test", table_name.to_string()); @@ -9283,6 +9285,8 @@ fn test_create_index_with_using_function() { assert!(if_not_exists); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -9324,6 +9328,8 @@ fn test_create_index_with_with_clause() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { pretty_assertions::assert_eq!("title_idx", name.to_string()); pretty_assertions::assert_eq!("films", table_name.to_string()); @@ -9333,6 +9339,8 @@ fn test_create_index_with_with_clause() { assert!(!if_not_exists); assert!(include.is_empty()); pretty_assertions::assert_eq!(with_parameters, with); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index c13f2266f..2b8d4b5ee 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -4215,3 +4215,31 @@ fn parse_show_charset() { mysql().verified_stmt("SHOW CHARSET WHERE charset = 'utf8mb4%'"); mysql().verified_stmt("SHOW CHARSET LIKE 'utf8mb4%'"); } + +#[test] +fn test_ddl_with_index_using() { + let columns = "(name, age DESC)"; + let using = "USING BTREE"; + + for sql in [ + format!("CREATE INDEX idx_name ON test {using} {columns}"), + format!("CREATE TABLE foo (name VARCHAR(255), age INT, KEY idx_name {using} {columns})"), + format!("ALTER TABLE foo ADD KEY idx_name {using} {columns}"), + format!("CREATE INDEX idx_name ON test{columns} {using}"), + format!("CREATE TABLE foo (name VARCHAR(255), age INT, KEY idx_name {columns} {using})"), + format!("ALTER TABLE foo ADD KEY idx_name {columns} {using}"), + ] { + mysql_and_generic().verified_stmt(&sql); + } +} + +#[test] +fn test_create_index_options() { + mysql_and_generic() + .verified_stmt("CREATE INDEX idx_name ON t(c1, c2) USING HASH LOCK = SHARED"); + mysql_and_generic() + .verified_stmt("CREATE INDEX idx_name ON t(c1, c2) USING BTREE ALGORITHM = INPLACE"); + mysql_and_generic().verified_stmt( + "CREATE INDEX idx_name ON t(c1, c2) USING BTREE LOCK = EXCLUSIVE ALGORITHM = DEFAULT", + ); +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 997384a60..a7c9779b1 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2486,7 +2486,7 @@ fn parse_array_multi_subscript() { #[test] fn parse_create_index() { - let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1,col2)"; + let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1, col2)"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2500,6 +2500,8 @@ fn parse_create_index() { include, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["my_index"], &name); assert_eq_vec(&["my_table"], &table_name); @@ -2510,6 +2512,8 @@ fn parse_create_index() { assert_eq_vec(&["col1", "col2"], &columns); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2517,7 +2521,7 @@ fn parse_create_index() { #[test] fn parse_create_anonymous_index() { - let sql = "CREATE INDEX ON my_table(col1,col2)"; + let sql = "CREATE INDEX ON my_table(col1, col2)"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name, @@ -2531,6 +2535,8 @@ fn parse_create_anonymous_index() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq!(None, name); assert_eq_vec(&["my_table"], &table_name); @@ -2541,6 +2547,8 @@ fn parse_create_anonymous_index() { assert_eq_vec(&["col1", "col2"], &columns); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2577,7 +2585,7 @@ fn parse_create_indices_with_operator_classes() { .unwrap_or_default() ); let multi_column_sql_statement = format!( - "CREATE INDEX the_index_name ON users USING {expected_index_type} (column_name,concat_users_name(first_name, last_name){})", + "CREATE INDEX the_index_name ON users USING {expected_index_type} (column_name, concat_users_name(first_name, last_name){})", expected_operator_class.as_ref().map(|oc| format!(" {oc}")) .unwrap_or_default() ); @@ -2639,6 +2647,8 @@ fn parse_create_indices_with_operator_classes() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["the_index_name"], &name); assert_eq_vec(&["users"], &table_name); @@ -2646,6 +2656,8 @@ fn parse_create_indices_with_operator_classes() { assert_eq!(expected_function_column, columns[0],); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2663,6 +2675,8 @@ fn parse_create_indices_with_operator_classes() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["the_index_name"], &name); assert_eq_vec(&["users"], &table_name); @@ -2688,6 +2702,8 @@ fn parse_create_indices_with_operator_classes() { assert_eq!(expected_function_column, columns[1],); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2698,7 +2714,7 @@ fn parse_create_indices_with_operator_classes() { #[test] fn parse_create_bloom() { let sql = - "CREATE INDEX bloomidx ON tbloom USING BLOOM (i1,i2,i3) WITH (length = 80, col1 = 2, col2 = 2, col3 = 4)"; + "CREATE INDEX bloomidx ON tbloom USING BLOOM (i1, i2, i3) WITH (length = 80, col1 = 2, col2 = 2, col3 = 4)"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2712,6 +2728,8 @@ fn parse_create_bloom() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["bloomidx"], &name); assert_eq_vec(&["tbloom"], &table_name); @@ -2743,6 +2761,8 @@ fn parse_create_bloom() { ], with ); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2764,6 +2784,8 @@ fn parse_create_brin() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["brin_sensor_data_recorded_at"], &name); assert_eq_vec(&["sensor_data"], &table_name); @@ -2771,6 +2793,8 @@ fn parse_create_brin() { assert_eq_vec(&["recorded_at"], &columns); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2813,7 +2837,7 @@ fn parse_create_table_with_empty_inherits_fails() { #[test] fn parse_create_index_concurrently() { - let sql = "CREATE INDEX CONCURRENTLY IF NOT EXISTS my_index ON my_table(col1,col2)"; + let sql = "CREATE INDEX CONCURRENTLY IF NOT EXISTS my_index ON my_table(col1, col2)"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2827,6 +2851,8 @@ fn parse_create_index_concurrently() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["my_index"], &name); assert_eq_vec(&["my_table"], &table_name); @@ -2837,6 +2863,8 @@ fn parse_create_index_concurrently() { assert_eq_vec(&["col1", "col2"], &columns); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2844,7 +2872,7 @@ fn parse_create_index_concurrently() { #[test] fn parse_create_index_with_predicate() { - let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1,col2) WHERE col3 IS NULL"; + let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1, col2) WHERE col3 IS NULL"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2858,6 +2886,8 @@ fn parse_create_index_with_predicate() { nulls_distinct: None, with, predicate: Some(_), + index_options, + alter_options, }) => { assert_eq_vec(&["my_index"], &name); assert_eq_vec(&["my_table"], &table_name); @@ -2868,6 +2898,8 @@ fn parse_create_index_with_predicate() { assert_eq_vec(&["col1", "col2"], &columns); assert!(include.is_empty()); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2875,7 +2907,7 @@ fn parse_create_index_with_predicate() { #[test] fn parse_create_index_with_include() { - let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1,col2) INCLUDE (col3)"; + let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1, col2) INCLUDE (col3, col4)"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2889,6 +2921,8 @@ fn parse_create_index_with_include() { nulls_distinct: None, with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["my_index"], &name); assert_eq_vec(&["my_table"], &table_name); @@ -2897,8 +2931,10 @@ fn parse_create_index_with_include() { assert!(!concurrently); assert!(if_not_exists); assert_eq_vec(&["col1", "col2"], &columns); - assert_eq_vec(&["col3"], &include); + assert_eq_vec(&["col3", "col4"], &include); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } @@ -2906,7 +2942,7 @@ fn parse_create_index_with_include() { #[test] fn parse_create_index_with_nulls_distinct() { - let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1,col2) NULLS NOT DISTINCT"; + let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1, col2) NULLS NOT DISTINCT"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2920,6 +2956,8 @@ fn parse_create_index_with_nulls_distinct() { nulls_distinct: Some(nulls_distinct), with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["my_index"], &name); assert_eq_vec(&["my_table"], &table_name); @@ -2931,11 +2969,13 @@ fn parse_create_index_with_nulls_distinct() { assert!(include.is_empty()); assert!(!nulls_distinct); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } - let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1,col2) NULLS DISTINCT"; + let sql = "CREATE INDEX IF NOT EXISTS my_index ON my_table(col1, col2) NULLS DISTINCT"; match pg().verified_stmt(sql) { Statement::CreateIndex(CreateIndex { name: Some(ObjectName(name)), @@ -2949,6 +2989,8 @@ fn parse_create_index_with_nulls_distinct() { nulls_distinct: Some(nulls_distinct), with, predicate: None, + index_options, + alter_options, }) => { assert_eq_vec(&["my_index"], &name); assert_eq_vec(&["my_table"], &table_name); @@ -2960,6 +3002,8 @@ fn parse_create_index_with_nulls_distinct() { assert!(include.is_empty()); assert!(nulls_distinct); assert!(with.is_empty()); + assert!(index_options.is_empty()); + assert!(alter_options.is_empty()); } _ => unreachable!(), } From 27544f934372858500aebfce0c63e070c6d84ca0 Mon Sep 17 00:00:00 2001 From: Sven Sauleau Date: Fri, 8 Aug 2025 12:19:35 +0200 Subject: [PATCH 232/286] Add support for `UPDATE ... LIMIT ...` (#1991) --- src/ast/mod.rs | 7 +++++++ src/ast/spans.rs | 1 + src/parser/mod.rs | 6 ++++++ tests/sqlparser_common.rs | 2 ++ tests/sqlparser_mysql.rs | 1 + tests/sqlparser_sqlite.rs | 20 +++++++++++++++++++- 6 files changed, 36 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index d21f83574..93f44e1b6 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3238,6 +3238,8 @@ pub enum Statement { returning: Option>, /// SQLite-specific conflict resolution clause or: Option, + /// LIMIT + limit: Option, }, /// ```sql /// DELETE @@ -4810,6 +4812,7 @@ impl fmt::Display for Statement { selection, returning, or, + limit, } => { f.write_str("UPDATE ")?; if let Some(or) = or { @@ -4843,6 +4846,10 @@ impl fmt::Display for Statement { f.write_str("RETURNING")?; indented_list(f, returning)?; } + if let Some(limit) = limit { + SpaceOrNewline.fmt(f)?; + write!(f, "LIMIT {limit}")?; + } Ok(()) } Statement::Delete(delete) => delete.fmt(f), diff --git a/src/ast/spans.rs b/src/ast/spans.rs index e6e601c33..9705a0076 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -378,6 +378,7 @@ impl Spanned for Statement { selection, returning, or: _, + limit: _, } => union_spans( core::iter::once(table.span()) .chain(assignments.iter().map(|i| i.span())) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 78ce6de6d..d7a07d385 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15152,6 +15152,11 @@ impl<'a> Parser<'a> { } else { None }; + let limit = if self.parse_keyword(Keyword::LIMIT) { + Some(self.parse_expr()?) + } else { + None + }; Ok(Statement::Update { table, assignments, @@ -15159,6 +15164,7 @@ impl<'a> Parser<'a> { selection, returning, or, + limit, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index d71449d92..06d9cf14b 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -515,6 +515,7 @@ fn parse_update_set_from() { }), returning: None, or: None, + limit: None } ); @@ -533,6 +534,7 @@ fn parse_update_with_table_alias() { selection, returning, or: None, + limit: None, } => { assert_eq!( TableWithJoins { diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 2b8d4b5ee..0dcf7c0d3 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2516,6 +2516,7 @@ fn parse_update_with_joins() { selection, returning, or: None, + limit: None, } => { assert_eq!( TableWithJoins { diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 06496f0ce..114aca03a 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -485,7 +485,8 @@ fn parse_update_tuple_row_values() { joins: vec![], }, from: None, - returning: None + returning: None, + limit: None } ); } @@ -592,6 +593,23 @@ fn test_regexp_operator() { sqlite().verified_only_select(r#"SELECT count(*) FROM messages WHERE msg_text REGEXP '\d+'"#); } +#[test] +fn test_update_delete_limit() { + match sqlite().verified_stmt("UPDATE foo SET bar = 1 LIMIT 99") { + Statement::Update { limit, .. } => { + assert_eq!(limit, Some(Expr::value(number("99")))); + } + _ => unreachable!(), + } + + match sqlite().verified_stmt("DELETE FROM foo LIMIT 99") { + Statement::Delete(Delete { limit, .. }) => { + assert_eq!(limit, Some(Expr::value(number("99")))); + } + _ => unreachable!(), + } +} + fn sqlite() -> TestedDialects { TestedDialects::new(vec![Box::new(SQLiteDialect {})]) } From 356308b348a57f8e243428866c2163cbb270d86c Mon Sep 17 00:00:00 2001 From: Tyler White <50381805+IndexSeek@users.noreply.github.com> Date: Sat, 9 Aug 2025 02:31:58 -0400 Subject: [PATCH 233/286] feat: Include end token in `ALTER TABLE` statement (#1999) Co-authored-by: Ifeanyi Ubah --- src/ast/mod.rs | 3 +++ src/ast/spans.rs | 20 +++++++++++++++++++- src/parser/mod.rs | 7 +++++++ src/test_utils.rs | 1 + tests/sqlparser_mysql.rs | 1 + 5 files changed, 31 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 93f44e1b6..5b50d020c 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3384,6 +3384,8 @@ pub enum Statement { /// Snowflake "ICEBERG" clause for Iceberg tables /// iceberg: bool, + /// Token that represents the end of the statement (semicolon or EOF) + end_token: AttachedToken, }, /// ```sql /// ALTER INDEX @@ -5442,6 +5444,7 @@ impl fmt::Display for Statement { location, on_cluster, iceberg, + end_token: _, } => { if *iceberg { write!(f, "ALTER ICEBERG TABLE ")?; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 9705a0076..e17090268 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -435,10 +435,12 @@ impl Spanned for Statement { location: _, on_cluster, iceberg: _, + end_token, } => union_spans( core::iter::once(name.span()) .chain(operations.iter().map(|i| i.span())) - .chain(on_cluster.iter().map(|i| i.span)), + .chain(on_cluster.iter().map(|i| i.span)) + .chain(core::iter::once(end_token.0.span)), ), Statement::AlterIndex { name, operation } => name.span().union(&operation.span()), Statement::AlterView { @@ -2553,4 +2555,20 @@ pub mod tests { stmt => panic!("expected query; got {stmt:?}"), } } + + #[test] + fn test_alter_table_multiline_span() { + let sql = r#"-- foo +ALTER TABLE users + ADD COLUMN foo + varchar; -- hi there"#; + + let r = Parser::parse_sql(&crate::dialect::PostgreSqlDialect {}, sql).unwrap(); + assert_eq!(1, r.len()); + + let stmt_span = r[0].span(); + + assert_eq!(stmt_span.start, (2, 13).into()); + assert_eq!(stmt_span.end, (4, 11).into()); + } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d7a07d385..d6ace1eb6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9263,6 +9263,12 @@ impl<'a> Parser<'a> { }); } + let end_token = if self.peek_token_ref().token == Token::SemiColon { + self.peek_token_ref().clone() + } else { + self.get_current_token().clone() + }; + Ok(Statement::AlterTable { name: table_name, if_exists, @@ -9271,6 +9277,7 @@ impl<'a> Parser<'a> { location, on_cluster, iceberg, + end_token: AttachedToken(end_token), }) } diff --git a/src/test_utils.rs b/src/test_utils.rs index 654f2723e..ab2cf89b2 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -351,6 +351,7 @@ pub fn alter_table_op_with_name(stmt: Statement, expected_name: &str) -> AlterTa on_cluster: _, location: _, iceberg, + end_token: _, } => { assert_eq!(name.to_string(), expected_name); assert!(!if_exists); diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 0dcf7c0d3..c47750b59 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2643,6 +2643,7 @@ fn parse_alter_table_add_column() { iceberg, location: _, on_cluster: _, + end_token: _, } => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); From b2f977384c469477a06dd1d61489877c12fb9798 Mon Sep 17 00:00:00 2001 From: Tyler White <50381805+IndexSeek@users.noreply.github.com> Date: Sat, 9 Aug 2025 02:33:34 -0400 Subject: [PATCH 234/286] Postgres: enhance NUMERIC/DECIMAL parsing to support negative scale (#1990) Co-authored-by: Ifeanyi Ubah --- src/ast/data_type.rs | 2 +- src/parser/mod.rs | 79 +++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 76 insertions(+), 5 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index b4a8af60f..dde9e0b7e 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -962,7 +962,7 @@ pub enum ExactNumberInfo { /// Only precision information, e.g. `DECIMAL(10)` Precision(u64), /// Precision and scale information, e.g. `DECIMAL(10,2)` - PrecisionAndScale(u64, u64), + PrecisionAndScale(u64, i64), } impl fmt::Display for ExactNumberInfo { diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d6ace1eb6..2c5f1b9e6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11256,7 +11256,7 @@ impl<'a> Parser<'a> { if self.consume_token(&Token::LParen) { let precision = self.parse_literal_uint()?; let scale = if self.consume_token(&Token::Comma) { - Some(self.parse_literal_uint()?) + Some(self.parse_signed_integer()?) } else { None }; @@ -11272,6 +11272,27 @@ impl<'a> Parser<'a> { } } + /// Parse an optionally signed integer literal. + fn parse_signed_integer(&mut self) -> Result { + let is_negative = self.consume_token(&Token::Minus); + + if !is_negative { + let _ = self.consume_token(&Token::Plus); + } + + let current_token = self.peek_token_ref(); + match ¤t_token.token { + Token::Number(s, _) => { + let s = s.clone(); + let span_start = current_token.span.start; + self.advance_token(); + let value = Self::parse::(s, span_start)?; + Ok(if is_negative { -value } else { value }) + } + _ => self.expected_ref("number", current_token), + } + } + pub fn parse_optional_type_modifiers(&mut self) -> Result>, ParserError> { if self.consume_token(&Token::LParen) { let mut modifiers = Vec::new(); @@ -17118,7 +17139,7 @@ mod tests { use crate::ast::{ CharLengthUnits, CharacterLength, DataType, ExactNumberInfo, ObjectName, TimezoneInfo, }; - use crate::dialect::{AnsiDialect, GenericDialect}; + use crate::dialect::{AnsiDialect, GenericDialect, PostgreSqlDialect}; use crate::test_utils::TestedDialects; macro_rules! test_parse_data_type { @@ -17324,8 +17345,11 @@ mod tests { #[test] fn test_ansii_exact_numeric_types() { // Exact numeric types: - let dialect = - TestedDialects::new(vec![Box::new(GenericDialect {}), Box::new(AnsiDialect {})]); + let dialect = TestedDialects::new(vec![ + Box::new(GenericDialect {}), + Box::new(AnsiDialect {}), + Box::new(PostgreSqlDialect {}), + ]); test_parse_data_type!(dialect, "NUMERIC", DataType::Numeric(ExactNumberInfo::None)); @@ -17368,6 +17392,53 @@ mod tests { "DEC(2,10)", DataType::Dec(ExactNumberInfo::PrecisionAndScale(2, 10)) ); + + // Test negative scale values. + test_parse_data_type!( + dialect, + "NUMERIC(10,-2)", + DataType::Numeric(ExactNumberInfo::PrecisionAndScale(10, -2)) + ); + + test_parse_data_type!( + dialect, + "DECIMAL(1000,-10)", + DataType::Decimal(ExactNumberInfo::PrecisionAndScale(1000, -10)) + ); + + test_parse_data_type!( + dialect, + "DEC(5,-1000)", + DataType::Dec(ExactNumberInfo::PrecisionAndScale(5, -1000)) + ); + + test_parse_data_type!( + dialect, + "NUMERIC(10,-5)", + DataType::Numeric(ExactNumberInfo::PrecisionAndScale(10, -5)) + ); + + test_parse_data_type!( + dialect, + "DECIMAL(20,-10)", + DataType::Decimal(ExactNumberInfo::PrecisionAndScale(20, -10)) + ); + + test_parse_data_type!( + dialect, + "DEC(5,-2)", + DataType::Dec(ExactNumberInfo::PrecisionAndScale(5, -2)) + ); + + dialect.run_parser_method("NUMERIC(10,+5)", |parser| { + let data_type = parser.parse_data_type().unwrap(); + assert_eq!( + DataType::Numeric(ExactNumberInfo::PrecisionAndScale(10, 5)), + data_type + ); + // Note: Explicit '+' sign is not preserved in output, which is correct + assert_eq!("NUMERIC(10,5)", data_type.to_string()); + }); } #[test] From 60a5c8d42aa97500654b04ab28b4cd5f5de9cfd8 Mon Sep 17 00:00:00 2001 From: Michael Victor Zink Date: Wed, 13 Aug 2025 02:05:15 -0700 Subject: [PATCH 235/286] Fix column definition `COLLATE` parsing (#1986) --- src/dialect/postgresql.rs | 6 ++++- src/parser/mod.rs | 27 ++----------------- tests/sqlparser_common.rs | 57 +++++++++++++++++++++++++++++++++++++++ 3 files changed, 64 insertions(+), 26 deletions(-) diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 207f4787f..e861cc515 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -110,7 +110,11 @@ impl Dialect for PostgreSqlDialect { // we only return some custom value here when the behaviour (not merely the numeric value) differs // from the default implementation match token.token { - Token::Word(w) if w.keyword == Keyword::COLLATE => Some(Ok(COLLATE_PREC)), + Token::Word(w) + if w.keyword == Keyword::COLLATE && !parser.in_column_definition_state() => + { + Some(Ok(COLLATE_PREC)) + } Token::LBracket => Some(Ok(BRACKET_PREC)), Token::Arrow | Token::LongArrow diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2c5f1b9e6..3d4762541 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1722,7 +1722,7 @@ impl<'a> Parser<'a> { _ => self.expected_at("an expression", next_token_index), }?; - if self.parse_keyword(Keyword::COLLATE) { + if !self.in_column_definition_state() && self.parse_keyword(Keyword::COLLATE) { Ok(Expr::Collate { expr: Box::new(expr), collation: self.parse_object_name(false)?, @@ -3372,6 +3372,7 @@ impl<'a> Parser<'a> { self.advance_token(); let tok = self.get_current_token(); + debug!("infix: {tok:?}"); let tok_index = self.get_current_index(); let span = tok.span; let regular_binary_operator = match &tok.token { @@ -17822,30 +17823,6 @@ mod tests { assert!(Parser::parse_sql(&MySqlDialect {}, sql).is_err()); } - #[test] - fn test_parse_not_null_in_column_options() { - let canonical = concat!( - "CREATE TABLE foo (", - "abc INT DEFAULT (42 IS NOT NULL) NOT NULL,", - " def INT,", - " def_null BOOL GENERATED ALWAYS AS (def IS NOT NULL) STORED,", - " CHECK (abc IS NOT NULL)", - ")" - ); - all_dialects().verified_stmt(canonical); - all_dialects().one_statement_parses_to( - concat!( - "CREATE TABLE foo (", - "abc INT DEFAULT (42 NOT NULL) NOT NULL,", - " def INT,", - " def_null BOOL GENERATED ALWAYS AS (def NOT NULL) STORED,", - " CHECK (abc NOT NULL)", - ")" - ), - canonical, - ); - } - #[test] fn test_placeholder_invalid_whitespace() { for w in [" ", "/*invalid*/"] { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 06d9cf14b..4b9d748fc 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16598,3 +16598,60 @@ fn parse_create_view_if_not_exists() { res.unwrap_err() ); } + +#[test] +fn test_parse_not_null_in_column_options() { + let canonical = concat!( + "CREATE TABLE foo (", + "abc INT DEFAULT (42 IS NOT NULL) NOT NULL,", + " def INT,", + " def_null BOOL GENERATED ALWAYS AS (def IS NOT NULL) STORED,", + " CHECK (abc IS NOT NULL)", + ")" + ); + all_dialects().verified_stmt(canonical); + all_dialects().one_statement_parses_to( + concat!( + "CREATE TABLE foo (", + "abc INT DEFAULT (42 NOT NULL) NOT NULL,", + " def INT,", + " def_null BOOL GENERATED ALWAYS AS (def NOT NULL) STORED,", + " CHECK (abc NOT NULL)", + ")" + ), + canonical, + ); +} + +#[test] +fn test_parse_default_with_collate_column_option() { + let sql = "CREATE TABLE foo (abc TEXT DEFAULT 'foo' COLLATE 'en_US')"; + let stmt = all_dialects().verified_stmt(sql); + if let Statement::CreateTable(CreateTable { mut columns, .. }) = stmt { + let mut column = columns.pop().unwrap(); + assert_eq!(&column.name.value, "abc"); + assert_eq!(column.data_type, DataType::Text); + let collate_option = column.options.pop().unwrap(); + if let ColumnOptionDef { + name: None, + option: ColumnOption::Collation(collate), + } = collate_option + { + assert_eq!(collate.to_string(), "'en_US'"); + } else { + panic!("Expected collate column option, got {collate_option}"); + } + let default_option = column.options.pop().unwrap(); + if let ColumnOptionDef { + name: None, + option: ColumnOption::Default(Expr::Value(value)), + } = default_option + { + assert_eq!(value.to_string(), "'foo'"); + } else { + panic!("Expected default column option, got {default_option}"); + } + } else { + panic!("Expected create table statement"); + } +} From b660a3b1ea41ec8c48fcb4341b6112f043d7cd57 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Sat, 16 Aug 2025 08:34:23 +0200 Subject: [PATCH 236/286] Redshift: `CREATE TABLE ... (LIKE ..)` (#1967) --- src/ast/ddl.rs | 21 ++++---- src/ast/helpers/stmt_create_table.rs | 12 ++--- src/ast/mod.rs | 56 +++++++++++++++++++++ src/ast/spans.rs | 3 +- src/dialect/mod.rs | 19 +++++++ src/dialect/redshift.rs | 4 ++ src/dialect/snowflake.rs | 17 ++++--- src/keywords.rs | 3 ++ src/parser/mod.rs | 44 +++++++++++++++-- tests/sqlparser_common.rs | 74 ++++++++++++++++++++++++++++ 10 files changed, 224 insertions(+), 29 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 1c2aaf48d..65dfd6c56 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -31,12 +31,12 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, - CreateFunctionUsing, CreateTableOptions, DataType, Expr, FileFormat, FunctionBehavior, - FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, HiveDistributionStyle, - HiveFormat, HiveIOFormat, HiveRowFormat, Ident, MySQLColumnPosition, ObjectName, OnCommit, - OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RowAccessPolicy, - SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, Tag, Value, ValueWithSpan, - WrappedCollection, + CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, DataType, Expr, FileFormat, + FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, + HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, MySQLColumnPosition, + ObjectName, OnCommit, OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, + Query, RowAccessPolicy, SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, Tag, + Value, ValueWithSpan, WrappedCollection, }; use crate::display_utils::{DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; use crate::keywords::Keyword; @@ -2430,7 +2430,7 @@ pub struct CreateTable { pub location: Option, pub query: Option>, pub without_rowid: bool, - pub like: Option, + pub like: Option, pub clone: Option, // For Hive dialect, the table comment is after the column definitions without `=`, // so the `comment` field is optional and different than the comment field in the general options list. @@ -2559,6 +2559,8 @@ impl fmt::Display for CreateTable { } else if self.query.is_none() && self.like.is_none() && self.clone.is_none() { // PostgreSQL allows `CREATE TABLE t ();`, but requires empty parens f.write_str(" ()")?; + } else if let Some(CreateTableLikeKind::Parenthesized(like_in_columns_list)) = &self.like { + write!(f, " ({like_in_columns_list})")?; } // Hive table comment should be after column definitions, please refer to: @@ -2572,9 +2574,8 @@ impl fmt::Display for CreateTable { write!(f, " WITHOUT ROWID")?; } - // Only for Hive - if let Some(l) = &self.like { - write!(f, " LIKE {l}")?; + if let Some(CreateTableLikeKind::Plain(like)) = &self.like { + write!(f, " {like}")?; } if let Some(c) = &self.clone { diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index c727276d3..9e9d229ef 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -25,10 +25,10 @@ use serde::{Deserialize, Serialize}; use sqlparser_derive::{Visit, VisitMut}; use crate::ast::{ - ClusteredBy, ColumnDef, CommentDef, CreateTable, CreateTableOptions, Expr, FileFormat, - HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, OneOrManyWithParens, Query, - RowAccessPolicy, Statement, StorageSerializationPolicy, TableConstraint, Tag, - WrappedCollection, + ClusteredBy, ColumnDef, CommentDef, CreateTable, CreateTableLikeKind, CreateTableOptions, Expr, + FileFormat, HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, + OneOrManyWithParens, Query, RowAccessPolicy, Statement, StorageSerializationPolicy, + TableConstraint, Tag, WrappedCollection, }; use crate::parser::ParserError; @@ -81,7 +81,7 @@ pub struct CreateTableBuilder { pub location: Option, pub query: Option>, pub without_rowid: bool, - pub like: Option, + pub like: Option, pub clone: Option, pub comment: Option, pub on_commit: Option, @@ -237,7 +237,7 @@ impl CreateTableBuilder { self } - pub fn like(mut self, like: Option) -> Self { + pub fn like(mut self, like: Option) -> Self { self.like = like; self } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 5b50d020c..a30e24239 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -10465,6 +10465,62 @@ impl fmt::Display for CreateUser { } } +/// Specifies how to create a new table based on an existing table's schema. +/// '''sql +/// CREATE TABLE new LIKE old ... +/// ''' +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum CreateTableLikeKind { + /// '''sql + /// CREATE TABLE new (LIKE old ...) + /// ''' + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_CREATE_TABLE_NEW.html) + Parenthesized(CreateTableLike), + /// '''sql + /// CREATE TABLE new LIKE old ... + /// ''' + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-table#label-create-table-like) + /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_table_like) + Plain(CreateTableLike), +} + +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum CreateTableLikeDefaults { + Including, + Excluding, +} + +impl fmt::Display for CreateTableLikeDefaults { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + CreateTableLikeDefaults::Including => write!(f, "INCLUDING DEFAULTS"), + CreateTableLikeDefaults::Excluding => write!(f, "EXCLUDING DEFAULTS"), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateTableLike { + pub name: ObjectName, + pub defaults: Option, +} + +impl fmt::Display for CreateTableLike { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "LIKE {}", self.name)?; + if let Some(defaults) = &self.defaults { + write!(f, " {defaults}")?; + } + Ok(()) + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index e17090268..da47b0f85 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -592,7 +592,7 @@ impl Spanned for CreateTable { location: _, // string, no span query, without_rowid: _, // bool - like, + like: _, clone, comment: _, // todo, no span on_commit: _, @@ -627,7 +627,6 @@ impl Spanned for CreateTable { .chain(columns.iter().map(|i| i.span())) .chain(constraints.iter().map(|i| i.span())) .chain(query.iter().map(|i| i.span())) - .chain(like.iter().map(|i| i.span())) .chain(clone.iter().map(|i| i.span())), ) } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 6ca364784..7cf9d4fd1 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1163,6 +1163,25 @@ pub trait Dialect: Debug + Any { fn supports_interval_options(&self) -> bool { false } + + /// Returns true if the dialect supports specifying which table to copy + /// the schema from inside parenthesis. + /// + /// Not parenthesized: + /// '''sql + /// CREATE TABLE new LIKE old ... + /// ''' + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-table#label-create-table-like) + /// [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_table_like) + /// + /// Parenthesized: + /// '''sql + /// CREATE TABLE new (LIKE old ...) + /// ''' + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_CREATE_TABLE_NEW.html) + fn supports_create_table_like_parenthesized(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index 68e025d18..1cd6098a6 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -139,4 +139,8 @@ impl Dialect for RedshiftSqlDialect { fn supports_select_exclude(&self) -> bool { true } + + fn supports_create_table_like_parenthesized(&self) -> bool { + true + } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 8830e09a0..7ef4de9c2 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -27,10 +27,10 @@ use crate::ast::helpers::stmt_data_loading::{ }; use crate::ast::{ CatalogSyncNamespaceMode, ColumnOption, ColumnPolicy, ColumnPolicyProperty, ContactEntry, - CopyIntoSnowflakeKind, DollarQuotedString, Ident, IdentityParameters, IdentityProperty, - IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, ObjectName, - ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, Statement, StorageSerializationPolicy, - TagsColumnOption, WrappedCollection, + CopyIntoSnowflakeKind, CreateTableLikeKind, DollarQuotedString, Ident, IdentityParameters, + IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, + ObjectName, ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, Statement, + StorageSerializationPolicy, TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -668,8 +668,13 @@ pub fn parse_create_table( builder = builder.clone_clause(clone); } Keyword::LIKE => { - let like = parser.parse_object_name(false).ok(); - builder = builder.like(like); + let name = parser.parse_object_name(false)?; + builder = builder.like(Some(CreateTableLikeKind::Plain( + crate::ast::CreateTableLike { + name, + defaults: None, + }, + ))); } Keyword::CLUSTER => { parser.expect_keyword_is(Keyword::BY)?; diff --git a/src/keywords.rs b/src/keywords.rs index a729a525f..659bc0439 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -268,6 +268,7 @@ define_keywords!( DECLARE, DEDUPLICATE, DEFAULT, + DEFAULTS, DEFAULT_DDL_COLLATION, DEFERRABLE, DEFERRED, @@ -339,6 +340,7 @@ define_keywords!( EXCEPTION, EXCHANGE, EXCLUDE, + EXCLUDING, EXCLUSIVE, EXEC, EXECUTE, @@ -441,6 +443,7 @@ define_keywords!( IN, INCLUDE, INCLUDE_NULL_VALUES, + INCLUDING, INCREMENT, INDEX, INDICATOR, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 3d4762541..9cddf8272 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7347,11 +7347,7 @@ impl<'a> Parser<'a> { // Clickhouse has `ON CLUSTER 'cluster'` syntax for DDLs let on_cluster = self.parse_optional_on_cluster()?; - let like = if self.parse_keyword(Keyword::LIKE) || self.parse_keyword(Keyword::ILIKE) { - self.parse_object_name(allow_unquoted_hyphen).ok() - } else { - None - }; + let like = self.maybe_parse_create_table_like(allow_unquoted_hyphen)?; let clone = if self.parse_keyword(Keyword::CLONE) { self.parse_object_name(allow_unquoted_hyphen).ok() @@ -7455,6 +7451,44 @@ impl<'a> Parser<'a> { .build()) } + fn maybe_parse_create_table_like( + &mut self, + allow_unquoted_hyphen: bool, + ) -> Result, ParserError> { + let like = if self.dialect.supports_create_table_like_parenthesized() + && self.consume_token(&Token::LParen) + { + if self.parse_keyword(Keyword::LIKE) { + let name = self.parse_object_name(allow_unquoted_hyphen)?; + let defaults = if self.parse_keywords(&[Keyword::INCLUDING, Keyword::DEFAULTS]) { + Some(CreateTableLikeDefaults::Including) + } else if self.parse_keywords(&[Keyword::EXCLUDING, Keyword::DEFAULTS]) { + Some(CreateTableLikeDefaults::Excluding) + } else { + None + }; + self.expect_token(&Token::RParen)?; + Some(CreateTableLikeKind::Parenthesized(CreateTableLike { + name, + defaults, + })) + } else { + // Rollback the '(' it's probably the columns list + self.prev_token(); + None + } + } else if self.parse_keyword(Keyword::LIKE) || self.parse_keyword(Keyword::ILIKE) { + let name = self.parse_object_name(allow_unquoted_hyphen)?; + Some(CreateTableLikeKind::Plain(CreateTableLike { + name, + defaults: None, + })) + } else { + None + }; + Ok(like) + } + pub(crate) fn parse_create_table_on_commit(&mut self) -> Result { if self.parse_keywords(&[Keyword::DELETE, Keyword::ROWS]) { Ok(OnCommit::DeleteRows) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 4b9d748fc..53b0d2036 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16655,3 +16655,77 @@ fn test_parse_default_with_collate_column_option() { panic!("Expected create table statement"); } } + +#[test] +fn parse_create_table_like() { + let dialects = all_dialects_except(|d| d.supports_create_table_like_parenthesized()); + let sql = "CREATE TABLE new LIKE old"; + match dialects.verified_stmt(sql) { + Statement::CreateTable(stmt) => { + assert_eq!( + stmt.name, + ObjectName::from(vec![Ident::new("new".to_string())]) + ); + assert_eq!( + stmt.like, + Some(CreateTableLikeKind::Plain(CreateTableLike { + name: ObjectName::from(vec![Ident::new("old".to_string())]), + defaults: None, + })) + ) + } + _ => unreachable!(), + } + let dialects = all_dialects_where(|d| d.supports_create_table_like_parenthesized()); + let sql = "CREATE TABLE new (LIKE old)"; + match dialects.verified_stmt(sql) { + Statement::CreateTable(stmt) => { + assert_eq!( + stmt.name, + ObjectName::from(vec![Ident::new("new".to_string())]) + ); + assert_eq!( + stmt.like, + Some(CreateTableLikeKind::Parenthesized(CreateTableLike { + name: ObjectName::from(vec![Ident::new("old".to_string())]), + defaults: None, + })) + ) + } + _ => unreachable!(), + } + let sql = "CREATE TABLE new (LIKE old INCLUDING DEFAULTS)"; + match dialects.verified_stmt(sql) { + Statement::CreateTable(stmt) => { + assert_eq!( + stmt.name, + ObjectName::from(vec![Ident::new("new".to_string())]) + ); + assert_eq!( + stmt.like, + Some(CreateTableLikeKind::Parenthesized(CreateTableLike { + name: ObjectName::from(vec![Ident::new("old".to_string())]), + defaults: Some(CreateTableLikeDefaults::Including), + })) + ) + } + _ => unreachable!(), + } + let sql = "CREATE TABLE new (LIKE old EXCLUDING DEFAULTS)"; + match dialects.verified_stmt(sql) { + Statement::CreateTable(stmt) => { + assert_eq!( + stmt.name, + ObjectName::from(vec![Ident::new("new".to_string())]) + ); + assert_eq!( + stmt.like, + Some(CreateTableLikeKind::Parenthesized(CreateTableLike { + name: ObjectName::from(vec![Ident::new("old".to_string())]), + defaults: Some(CreateTableLikeDefaults::Excluding), + })) + ) + } + _ => unreachable!(), + } +} From 12c0878a10ae33ec3b98d48d3d83859fce0aea13 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Tue, 19 Aug 2025 21:47:41 +0300 Subject: [PATCH 237/286] Add drop behavior to `DROP PRIMARY/FOREIGN KEY` (#2002) --- src/ast/ddl.rs | 72 ++++++++++++++++++++++-------------- src/ast/spans.rs | 4 +- src/parser/mod.rs | 9 ++++- tests/sqlparser_mysql.rs | 6 ++- tests/sqlparser_snowflake.rs | 10 +++++ 5 files changed, 67 insertions(+), 34 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 65dfd6c56..1336833a1 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -200,17 +200,18 @@ pub enum AlterTableOperation { }, /// `DROP PRIMARY KEY` /// - /// Note: this is a [MySQL]-specific operation. - /// - /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html - DropPrimaryKey, + /// [MySQL](https://dev.mysql.com/doc/refman/8.4/en/alter-table.html) + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/constraints-drop) + DropPrimaryKey { + drop_behavior: Option, + }, /// `DROP FOREIGN KEY ` /// - /// Note: this is a [MySQL]-specific operation. - /// - /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/alter-table.html + /// [MySQL](https://dev.mysql.com/doc/refman/8.4/en/alter-table.html) + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/constraints-drop) DropForeignKey { name: Ident, + drop_behavior: Option, }, /// `DROP INDEX ` /// @@ -648,36 +649,51 @@ impl fmt::Display for AlterTableOperation { } => { write!( f, - "DROP CONSTRAINT {}{}{}", + "DROP CONSTRAINT {}{}", if *if_exists { "IF EXISTS " } else { "" }, - name, - match drop_behavior { - None => "", - Some(DropBehavior::Restrict) => " RESTRICT", - Some(DropBehavior::Cascade) => " CASCADE", - } - ) + name + )?; + if let Some(drop_behavior) = drop_behavior { + write!(f, " {drop_behavior}")?; + } + Ok(()) + } + AlterTableOperation::DropPrimaryKey { drop_behavior } => { + write!(f, "DROP PRIMARY KEY")?; + if let Some(drop_behavior) = drop_behavior { + write!(f, " {drop_behavior}")?; + } + Ok(()) + } + AlterTableOperation::DropForeignKey { + name, + drop_behavior, + } => { + write!(f, "DROP FOREIGN KEY {name}")?; + if let Some(drop_behavior) = drop_behavior { + write!(f, " {drop_behavior}")?; + } + Ok(()) } - AlterTableOperation::DropPrimaryKey => write!(f, "DROP PRIMARY KEY"), - AlterTableOperation::DropForeignKey { name } => write!(f, "DROP FOREIGN KEY {name}"), AlterTableOperation::DropIndex { name } => write!(f, "DROP INDEX {name}"), AlterTableOperation::DropColumn { has_column_keyword, column_names: column_name, if_exists, drop_behavior, - } => write!( - f, - "DROP {}{}{}{}", - if *has_column_keyword { "COLUMN " } else { "" }, - if *if_exists { "IF EXISTS " } else { "" }, - display_comma_separated(column_name), - match drop_behavior { - None => "", - Some(DropBehavior::Restrict) => " RESTRICT", - Some(DropBehavior::Cascade) => " CASCADE", + } => { + write!( + f, + "DROP {}{}{}", + if *has_column_keyword { "COLUMN " } else { "" }, + if *if_exists { "IF EXISTS " } else { "" }, + display_comma_separated(column_name), + )?; + if let Some(drop_behavior) = drop_behavior { + write!(f, " {drop_behavior}")?; } - ), + Ok(()) + } AlterTableOperation::AttachPartition { partition } => { write!(f, "ATTACH {partition}") } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index da47b0f85..5d94e2dcc 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1150,8 +1150,8 @@ impl Spanned for AlterTableOperation { } => partition .span() .union_opt(&with_name.as_ref().map(|n| n.span)), - AlterTableOperation::DropPrimaryKey => Span::empty(), - AlterTableOperation::DropForeignKey { name } => name.span, + AlterTableOperation::DropPrimaryKey { .. } => Span::empty(), + AlterTableOperation::DropForeignKey { name, .. } => name.span, AlterTableOperation::DropIndex { name } => name.span, AlterTableOperation::EnableAlwaysRule { name } => name.span, AlterTableOperation::EnableAlwaysTrigger { name } => name.span, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9cddf8272..5480112e1 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8930,10 +8930,15 @@ impl<'a> Parser<'a> { drop_behavior, } } else if self.parse_keywords(&[Keyword::PRIMARY, Keyword::KEY]) { - AlterTableOperation::DropPrimaryKey + let drop_behavior = self.parse_optional_drop_behavior(); + AlterTableOperation::DropPrimaryKey { drop_behavior } } else if self.parse_keywords(&[Keyword::FOREIGN, Keyword::KEY]) { let name = self.parse_identifier()?; - AlterTableOperation::DropForeignKey { name } + let drop_behavior = self.parse_optional_drop_behavior(); + AlterTableOperation::DropForeignKey { + name, + drop_behavior, + } } else if self.parse_keyword(Keyword::INDEX) { let name = self.parse_identifier()?; AlterTableOperation::DropIndex { name } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index c47750b59..184532035 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2752,7 +2752,9 @@ fn parse_alter_table_add_columns() { fn parse_alter_table_drop_primary_key() { assert_matches!( alter_table_op(mysql_and_generic().verified_stmt("ALTER TABLE tab DROP PRIMARY KEY")), - AlterTableOperation::DropPrimaryKey + AlterTableOperation::DropPrimaryKey { + drop_behavior: None + } ); } @@ -2762,7 +2764,7 @@ fn parse_alter_table_drop_foreign_key() { alter_table_op( mysql_and_generic().verified_stmt("ALTER TABLE tab DROP FOREIGN KEY foo_ibfk_1") ), - AlterTableOperation::DropForeignKey { name } if name.value == "foo_ibfk_1" + AlterTableOperation::DropForeignKey { name, .. } if name.value == "foo_ibfk_1" ); } diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index bd8a6d30a..55f6ab287 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4571,3 +4571,13 @@ fn test_create_database() { .to_string(); assert!(err.contains("Expected"), "Unexpected error: {err}"); } + +#[test] +fn test_drop_constraints() { + snowflake().verified_stmt("ALTER TABLE tbl DROP PRIMARY KEY"); + snowflake().verified_stmt("ALTER TABLE tbl DROP FOREIGN KEY k1"); + snowflake().verified_stmt("ALTER TABLE tbl DROP CONSTRAINT c1"); + snowflake().verified_stmt("ALTER TABLE tbl DROP PRIMARY KEY CASCADE"); + snowflake().verified_stmt("ALTER TABLE tbl DROP FOREIGN KEY k1 RESTRICT"); + snowflake().verified_stmt("ALTER TABLE tbl DROP CONSTRAINT c1 CASCADE"); +} From 3b5242821e184998f27b7a7f5d83154896b8e3d7 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 20 Aug 2025 11:57:36 +0300 Subject: [PATCH 238/286] Redshift: Add support for IAM_ROLE and IGNOREHEADER COPY options (#1968) --- src/ast/mod.rs | 36 ++++++++++++++++++++++++++++++++++- src/keywords.rs | 2 ++ src/parser/mod.rs | 17 +++++++++++++++++ tests/sqlparser_common.rs | 40 +++++++++++++++++++++++++++++++++++++++ 4 files changed, 94 insertions(+), 1 deletion(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index a30e24239..8c026ce88 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -8772,6 +8772,10 @@ pub enum CopyLegacyOption { Null(String), /// CSV ... Csv(Vec), + /// IAM_ROLE { DEFAULT | 'arn:aws:iam::123456789:role/role1' } + IamRole(IamRoleKind), + /// IGNOREHEADER \[ AS \] number_rows + IgnoreHeader(u64), } impl fmt::Display for CopyLegacyOption { @@ -8781,7 +8785,37 @@ impl fmt::Display for CopyLegacyOption { Binary => write!(f, "BINARY"), Delimiter(char) => write!(f, "DELIMITER '{char}'"), Null(string) => write!(f, "NULL '{}'", value::escape_single_quote_string(string)), - Csv(opts) => write!(f, "CSV {}", display_separated(opts, " ")), + Csv(opts) => { + write!(f, "CSV")?; + if !opts.is_empty() { + write!(f, " {}", display_separated(opts, " "))?; + } + Ok(()) + } + IamRole(role) => write!(f, "IAM_ROLE {role}"), + IgnoreHeader(num_rows) => write!(f, "IGNOREHEADER {num_rows}"), + } + } +} + +/// An `IAM_ROLE` option in the AWS ecosystem +/// +/// [Redshift COPY](https://docs.aws.amazon.com/redshift/latest/dg/copy-parameters-authorization.html#copy-iam-role) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum IamRoleKind { + /// Default role + Default, + /// Specific role ARN, for example: `arn:aws:iam::123456789:role/role1` + Arn(String), +} + +impl fmt::Display for IamRoleKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + IamRoleKind::Default => write!(f, "DEFAULT"), + IamRoleKind::Arn(arn) => write!(f, "'{arn}'"), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 659bc0439..1d94ae23a 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -429,12 +429,14 @@ define_keywords!( HOUR, HOURS, HUGEINT, + IAM_ROLE, ICEBERG, ID, IDENTITY, IDENTITY_INSERT, IF, IGNORE, + IGNOREHEADER, ILIKE, IMMEDIATE, IMMUTABLE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 5480112e1..1afd57137 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9548,6 +9548,8 @@ impl<'a> Parser<'a> { Keyword::DELIMITER, Keyword::NULL, Keyword::CSV, + Keyword::IAM_ROLE, + Keyword::IGNOREHEADER, ]) { Some(Keyword::BINARY) => CopyLegacyOption::Binary, Some(Keyword::DELIMITER) => { @@ -9567,11 +9569,26 @@ impl<'a> Parser<'a> { } opts }), + Some(Keyword::IAM_ROLE) => CopyLegacyOption::IamRole(self.parse_iam_role_kind()?), + Some(Keyword::IGNOREHEADER) => { + let _ = self.parse_keyword(Keyword::AS); + let num_rows = self.parse_literal_uint()?; + CopyLegacyOption::IgnoreHeader(num_rows) + } _ => self.expected("option", self.peek_token())?, }; Ok(ret) } + fn parse_iam_role_kind(&mut self) -> Result { + if self.parse_keyword(Keyword::DEFAULT) { + Ok(IamRoleKind::Default) + } else { + let arn = self.parse_literal_string()?; + Ok(IamRoleKind::Arn(arn)) + } + } + fn parse_copy_legacy_csv_option(&mut self) -> Result { let ret = match self.parse_one_of_keywords(&[ Keyword::HEADER, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 53b0d2036..3bf7fef99 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16729,3 +16729,43 @@ fn parse_create_table_like() { _ => unreachable!(), } } + +#[test] +fn parse_copy_options() { + let copy = verified_stmt( + r#"COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' IAM_ROLE 'arn:aws:iam::123456789:role/role1' CSV IGNOREHEADER 1"#, + ); + match copy { + Statement::Copy { legacy_options, .. } => { + assert_eq!( + legacy_options, + vec![ + CopyLegacyOption::IamRole(IamRoleKind::Arn( + "arn:aws:iam::123456789:role/role1".to_string() + )), + CopyLegacyOption::Csv(vec![]), + CopyLegacyOption::IgnoreHeader(1), + ] + ); + } + _ => unreachable!(), + } + + let copy = one_statement_parses_to( + r#"COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' IAM_ROLE DEFAULT CSV IGNOREHEADER AS 1"#, + r#"COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' IAM_ROLE DEFAULT CSV IGNOREHEADER 1"#, + ); + match copy { + Statement::Copy { legacy_options, .. } => { + assert_eq!( + legacy_options, + vec![ + CopyLegacyOption::IamRole(IamRoleKind::Default), + CopyLegacyOption::Csv(vec![]), + CopyLegacyOption::IgnoreHeader(1), + ] + ); + } + _ => unreachable!(), + } +} From 4b8797e8f31c8e3450768bf2c88f80457abfc898 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 20 Aug 2025 12:08:22 +0300 Subject: [PATCH 239/286] Snowflake: Add support for `CREATE DYNAMIC TABLE` (#1960) --- src/ast/ddl.rs | 63 +++++++++++--- src/ast/helpers/stmt_create_table.rs | 90 +++++++++++++++----- src/ast/mod.rs | 42 ++++++++++ src/ast/query.rs | 6 +- src/ast/spans.rs | 7 ++ src/dialect/snowflake.rs | 66 +++++++++++---- src/keywords.rs | 8 ++ tests/sqlparser_duckdb.rs | 9 +- tests/sqlparser_mssql.rs | 18 +++- tests/sqlparser_postgres.rs | 9 +- tests/sqlparser_snowflake.rs | 120 +++++++++++++++++---------- 11 files changed, 342 insertions(+), 96 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 1336833a1..9e6e4517b 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -33,10 +33,11 @@ use crate::ast::{ display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, DataType, Expr, FileFormat, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, - HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, MySQLColumnPosition, - ObjectName, OnCommit, OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, - Query, RowAccessPolicy, SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, Tag, - Value, ValueWithSpan, WrappedCollection, + HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, InitializeKind, + MySQLColumnPosition, ObjectName, OnCommit, OneOrManyWithParens, OperateFunctionArg, + OrderByExpr, ProjectionSelect, Query, RefreshModeKind, RowAccessPolicy, SequenceOptions, + Spanned, SqlOption, StorageSerializationPolicy, TableVersion, Tag, Value, ValueWithSpan, + WrappedCollection, }; use crate::display_utils::{DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; use crate::keywords::Keyword; @@ -2428,6 +2429,7 @@ pub struct CreateTable { pub or_replace: bool, pub temporary: bool, pub external: bool, + pub dynamic: bool, pub global: Option, pub if_not_exists: bool, pub transient: bool, @@ -2448,6 +2450,7 @@ pub struct CreateTable { pub without_rowid: bool, pub like: Option, pub clone: Option, + pub version: Option, // For Hive dialect, the table comment is after the column definitions without `=`, // so the `comment` field is optional and different than the comment field in the general options list. // [Hive](https://cwiki.apache.org/confluence/display/Hive/LanguageManual+DDL#LanguageManualDDL-CreateTable) @@ -2525,6 +2528,21 @@ pub struct CreateTable { /// Snowflake "STORAGE_SERIALIZATION_POLICY" clause for Iceberg tables /// pub storage_serialization_policy: Option, + /// Snowflake "TARGET_LAG" clause for dybamic tables + /// + pub target_lag: Option, + /// Snowflake "WAREHOUSE" clause for dybamic tables + /// + pub warehouse: Option, + /// Snowflake "REFRESH_MODE" clause for dybamic tables + /// + pub refresh_mode: Option, + /// Snowflake "INITIALIZE" clause for dybamic tables + /// + pub initialize: Option, + /// Snowflake "REQUIRE USER" clause for dybamic tables + /// + pub require_user: bool, } impl fmt::Display for CreateTable { @@ -2538,7 +2556,7 @@ impl fmt::Display for CreateTable { // `CREATE TABLE t (a INT) AS SELECT a from t2` write!( f, - "CREATE {or_replace}{external}{global}{temporary}{transient}{volatile}{iceberg}TABLE {if_not_exists}{name}", + "CREATE {or_replace}{external}{global}{temporary}{transient}{volatile}{dynamic}{iceberg}TABLE {if_not_exists}{name}", or_replace = if self.or_replace { "OR REPLACE " } else { "" }, external = if self.external { "EXTERNAL " } else { "" }, global = self.global @@ -2556,6 +2574,7 @@ impl fmt::Display for CreateTable { volatile = if self.volatile { "VOLATILE " } else { "" }, // Only for Snowflake iceberg = if self.iceberg { "ICEBERG " } else { "" }, + dynamic = if self.dynamic { "DYNAMIC " } else { "" }, name = self.name, )?; if let Some(on_cluster) = &self.on_cluster { @@ -2598,6 +2617,10 @@ impl fmt::Display for CreateTable { write!(f, " CLONE {c}")?; } + if let Some(version) = &self.version { + write!(f, " {version}")?; + } + match &self.hive_distribution { HiveDistributionStyle::PARTITIONED { columns } => { write!(f, " PARTITIONED BY ({})", display_comma_separated(columns))?; @@ -2700,27 +2723,27 @@ impl fmt::Display for CreateTable { write!(f, " {options}")?; } if let Some(external_volume) = self.external_volume.as_ref() { - write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; + write!(f, " EXTERNAL_VOLUME='{external_volume}'")?; } if let Some(catalog) = self.catalog.as_ref() { - write!(f, " CATALOG = '{catalog}'")?; + write!(f, " CATALOG='{catalog}'")?; } if self.iceberg { if let Some(base_location) = self.base_location.as_ref() { - write!(f, " BASE_LOCATION = '{base_location}'")?; + write!(f, " BASE_LOCATION='{base_location}'")?; } } if let Some(catalog_sync) = self.catalog_sync.as_ref() { - write!(f, " CATALOG_SYNC = '{catalog_sync}'")?; + write!(f, " CATALOG_SYNC='{catalog_sync}'")?; } if let Some(storage_serialization_policy) = self.storage_serialization_policy.as_ref() { write!( f, - " STORAGE_SERIALIZATION_POLICY = {storage_serialization_policy}" + " STORAGE_SERIALIZATION_POLICY={storage_serialization_policy}" )?; } @@ -2774,6 +2797,26 @@ impl fmt::Display for CreateTable { write!(f, " WITH TAG ({})", display_comma_separated(tag.as_slice()))?; } + if let Some(target_lag) = &self.target_lag { + write!(f, " TARGET_LAG='{target_lag}'")?; + } + + if let Some(warehouse) = &self.warehouse { + write!(f, " WAREHOUSE={warehouse}")?; + } + + if let Some(refresh_mode) = &self.refresh_mode { + write!(f, " REFRESH_MODE={refresh_mode}")?; + } + + if let Some(initialize) = &self.initialize { + write!(f, " INITIALIZE={initialize}")?; + } + + if self.require_user { + write!(f, " REQUIRE USER")?; + } + if self.on_commit.is_some() { let on_commit = match self.on_commit { Some(OnCommit::DeleteRows) => "ON COMMIT DELETE ROWS", diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index 9e9d229ef..fc0002697 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -26,9 +26,9 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::{ ClusteredBy, ColumnDef, CommentDef, CreateTable, CreateTableLikeKind, CreateTableOptions, Expr, - FileFormat, HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, - OneOrManyWithParens, Query, RowAccessPolicy, Statement, StorageSerializationPolicy, - TableConstraint, Tag, WrappedCollection, + FileFormat, HiveDistributionStyle, HiveFormat, Ident, InitializeKind, ObjectName, OnCommit, + OneOrManyWithParens, Query, RefreshModeKind, RowAccessPolicy, Statement, + StorageSerializationPolicy, TableConstraint, TableVersion, Tag, WrappedCollection, }; use crate::parser::ParserError; @@ -72,6 +72,7 @@ pub struct CreateTableBuilder { pub transient: bool, pub volatile: bool, pub iceberg: bool, + pub dynamic: bool, pub name: ObjectName, pub columns: Vec, pub constraints: Vec, @@ -83,6 +84,7 @@ pub struct CreateTableBuilder { pub without_rowid: bool, pub like: Option, pub clone: Option, + pub version: Option, pub comment: Option, pub on_commit: Option, pub on_cluster: Option, @@ -108,6 +110,11 @@ pub struct CreateTableBuilder { pub catalog_sync: Option, pub storage_serialization_policy: Option, pub table_options: CreateTableOptions, + pub target_lag: Option, + pub warehouse: Option, + pub refresh_mode: Option, + pub initialize: Option, + pub require_user: bool, } impl CreateTableBuilder { @@ -121,6 +128,7 @@ impl CreateTableBuilder { transient: false, volatile: false, iceberg: false, + dynamic: false, name, columns: vec![], constraints: vec![], @@ -132,6 +140,7 @@ impl CreateTableBuilder { without_rowid: false, like: None, clone: None, + version: None, comment: None, on_commit: None, on_cluster: None, @@ -157,6 +166,11 @@ impl CreateTableBuilder { catalog_sync: None, storage_serialization_policy: None, table_options: CreateTableOptions::None, + target_lag: None, + warehouse: None, + refresh_mode: None, + initialize: None, + require_user: false, } } pub fn or_replace(mut self, or_replace: bool) -> Self { @@ -199,6 +213,11 @@ impl CreateTableBuilder { self } + pub fn dynamic(mut self, dynamic: bool) -> Self { + self.dynamic = dynamic; + self + } + pub fn columns(mut self, columns: Vec) -> Self { self.columns = columns; self @@ -248,6 +267,11 @@ impl CreateTableBuilder { self } + pub fn version(mut self, version: Option) -> Self { + self.version = version; + self + } + pub fn comment_after_column_def(mut self, comment: Option) -> Self { self.comment = comment; self @@ -382,24 +406,29 @@ impl CreateTableBuilder { self } - /// Returns true if the statement has exactly one source of info on the schema of the new table. - /// This is Snowflake-specific, some dialects allow more than one source. - pub(crate) fn validate_schema_info(&self) -> bool { - let mut sources = 0; - if !self.columns.is_empty() { - sources += 1; - } - if self.query.is_some() { - sources += 1; - } - if self.like.is_some() { - sources += 1; - } - if self.clone.is_some() { - sources += 1; - } + pub fn target_lag(mut self, target_lag: Option) -> Self { + self.target_lag = target_lag; + self + } + + pub fn warehouse(mut self, warehouse: Option) -> Self { + self.warehouse = warehouse; + self + } - sources == 1 + pub fn refresh_mode(mut self, refresh_mode: Option) -> Self { + self.refresh_mode = refresh_mode; + self + } + + pub fn initialize(mut self, initialize: Option) -> Self { + self.initialize = initialize; + self + } + + pub fn require_user(mut self, require_user: bool) -> Self { + self.require_user = require_user; + self } pub fn build(self) -> Statement { @@ -412,6 +441,7 @@ impl CreateTableBuilder { transient: self.transient, volatile: self.volatile, iceberg: self.iceberg, + dynamic: self.dynamic, name: self.name, columns: self.columns, constraints: self.constraints, @@ -423,6 +453,7 @@ impl CreateTableBuilder { without_rowid: self.without_rowid, like: self.like, clone: self.clone, + version: self.version, comment: self.comment, on_commit: self.on_commit, on_cluster: self.on_cluster, @@ -448,6 +479,11 @@ impl CreateTableBuilder { catalog_sync: self.catalog_sync, storage_serialization_policy: self.storage_serialization_policy, table_options: self.table_options, + target_lag: self.target_lag, + warehouse: self.warehouse, + refresh_mode: self.refresh_mode, + initialize: self.initialize, + require_user: self.require_user, }) } } @@ -468,6 +504,7 @@ impl TryFrom for CreateTableBuilder { transient, volatile, iceberg, + dynamic, name, columns, constraints, @@ -479,6 +516,7 @@ impl TryFrom for CreateTableBuilder { without_rowid, like, clone, + version, comment, on_commit, on_cluster, @@ -504,6 +542,11 @@ impl TryFrom for CreateTableBuilder { catalog_sync, storage_serialization_policy, table_options, + target_lag, + warehouse, + refresh_mode, + initialize, + require_user, }) => Ok(Self { or_replace, temporary, @@ -511,6 +554,7 @@ impl TryFrom for CreateTableBuilder { global, if_not_exists, transient, + dynamic, name, columns, constraints, @@ -522,6 +566,7 @@ impl TryFrom for CreateTableBuilder { without_rowid, like, clone, + version, comment, on_commit, on_cluster, @@ -549,6 +594,11 @@ impl TryFrom for CreateTableBuilder { catalog_sync, storage_serialization_policy, table_options, + target_lag, + warehouse, + refresh_mode, + initialize, + require_user, }), _ => Err(ParserError::ParserError(format!( "Expected create table statement, but received: {stmt}" diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8c026ce88..35a864042 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -10555,6 +10555,48 @@ impl fmt::Display for CreateTableLike { } } +/// Specifies the refresh mode for the dynamic table. +/// +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-dynamic-table) +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum RefreshModeKind { + Auto, + Full, + Incremental, +} + +impl fmt::Display for RefreshModeKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + RefreshModeKind::Auto => write!(f, "AUTO"), + RefreshModeKind::Full => write!(f, "FULL"), + RefreshModeKind::Incremental => write!(f, "INCREMENTAL"), + } + } +} + +/// Specifies the behavior of the initial refresh of the dynamic table. +/// +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-dynamic-table) +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum InitializeKind { + OnCreate, + OnSchedule, +} + +impl fmt::Display for InitializeKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + InitializeKind::OnCreate => write!(f, "ON_CREATE"), + InitializeKind::OnSchedule => write!(f, "ON_SCHEDULE"), + } + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/ast/query.rs b/src/ast/query.rs index ea641deba..3e06a00ac 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1884,7 +1884,7 @@ impl fmt::Display for TableFactor { write!(f, " WITH ({})", display_comma_separated(with_hints))?; } if let Some(version) = version { - write!(f, "{version}")?; + write!(f, " {version}")?; } if let Some(TableSampleKind::AfterTableAlias(sample)) = sample { write!(f, " {sample}")?; @@ -2179,8 +2179,8 @@ pub enum TableVersion { impl Display for TableVersion { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - TableVersion::ForSystemTimeAsOf(e) => write!(f, " FOR SYSTEM_TIME AS OF {e}")?, - TableVersion::Function(func) => write!(f, " {func}")?, + TableVersion::ForSystemTimeAsOf(e) => write!(f, "FOR SYSTEM_TIME AS OF {e}")?, + TableVersion::Function(func) => write!(f, "{func}")?, } Ok(()) } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 5d94e2dcc..fc5b762ae 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -579,6 +579,7 @@ impl Spanned for CreateTable { temporary: _, // bool external: _, // bool global: _, // bool + dynamic: _, // bool if_not_exists: _, // bool transient: _, // bool volatile: _, // bool @@ -619,6 +620,12 @@ impl Spanned for CreateTable { catalog_sync: _, // todo, Snowflake specific storage_serialization_policy: _, table_options, + target_lag: _, + warehouse: _, + version: _, + refresh_mode: _, + initialize: _, + require_user: _, } = self; union_spans( diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 7ef4de9c2..46c72a799 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -29,8 +29,8 @@ use crate::ast::{ CatalogSyncNamespaceMode, ColumnOption, ColumnPolicy, ColumnPolicyProperty, ContactEntry, CopyIntoSnowflakeKind, CreateTableLikeKind, DollarQuotedString, Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, - ObjectName, ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, Statement, - StorageSerializationPolicy, TagsColumnOption, WrappedCollection, + InitializeKind, ObjectName, ObjectNamePart, RefreshModeKind, RowAccessPolicy, ShowObjects, + SqlOption, Statement, StorageSerializationPolicy, TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -235,6 +235,8 @@ impl Dialect for SnowflakeDialect { _ => None, }; + let dynamic = parser.parse_keyword(Keyword::DYNAMIC); + let mut temporary = false; let mut volatile = false; let mut transient = false; @@ -259,7 +261,7 @@ impl Dialect for SnowflakeDialect { return Some(parse_create_stage(or_replace, temporary, parser)); } else if parser.parse_keyword(Keyword::TABLE) { return Some(parse_create_table( - or_replace, global, temporary, volatile, transient, iceberg, parser, + or_replace, global, temporary, volatile, transient, iceberg, dynamic, parser, )); } else if parser.parse_keyword(Keyword::DATABASE) { return Some(parse_create_database(or_replace, transient, parser)); @@ -614,6 +616,7 @@ fn parse_alter_session(parser: &mut Parser, set: bool) -> Result /// +#[allow(clippy::too_many_arguments)] pub fn parse_create_table( or_replace: bool, global: Option, @@ -621,6 +624,7 @@ pub fn parse_create_table( volatile: bool, transient: bool, iceberg: bool, + dynamic: bool, parser: &mut Parser, ) -> Result { let if_not_exists = parser.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); @@ -634,6 +638,7 @@ pub fn parse_create_table( .volatile(volatile) .iceberg(iceberg) .global(global) + .dynamic(dynamic) .hive_formats(Some(Default::default())); // Snowflake does not enforce order of the parameters in the statement. The parser needs to @@ -772,6 +777,49 @@ pub fn parse_create_table( Keyword::IF if parser.parse_keywords(&[Keyword::NOT, Keyword::EXISTS]) => { builder = builder.if_not_exists(true); } + Keyword::TARGET_LAG => { + parser.expect_token(&Token::Eq)?; + let target_lag = parser.parse_literal_string()?; + builder = builder.target_lag(Some(target_lag)); + } + Keyword::WAREHOUSE => { + parser.expect_token(&Token::Eq)?; + let warehouse = parser.parse_identifier()?; + builder = builder.warehouse(Some(warehouse)); + } + Keyword::AT | Keyword::BEFORE => { + parser.prev_token(); + let version = parser.maybe_parse_table_version()?; + builder = builder.version(version); + } + Keyword::REFRESH_MODE => { + parser.expect_token(&Token::Eq)?; + let refresh_mode = match parser.parse_one_of_keywords(&[ + Keyword::AUTO, + Keyword::FULL, + Keyword::INCREMENTAL, + ]) { + Some(Keyword::AUTO) => Some(RefreshModeKind::Auto), + Some(Keyword::FULL) => Some(RefreshModeKind::Full), + Some(Keyword::INCREMENTAL) => Some(RefreshModeKind::Incremental), + _ => return parser.expected("AUTO, FULL or INCREMENTAL", next_token), + }; + builder = builder.refresh_mode(refresh_mode); + } + Keyword::INITIALIZE => { + parser.expect_token(&Token::Eq)?; + let initialize = match parser + .parse_one_of_keywords(&[Keyword::ON_CREATE, Keyword::ON_SCHEDULE]) + { + Some(Keyword::ON_CREATE) => Some(InitializeKind::OnCreate), + Some(Keyword::ON_SCHEDULE) => Some(InitializeKind::OnSchedule), + _ => return parser.expected("ON_CREATE or ON_SCHEDULE", next_token), + }; + builder = builder.initialize(initialize); + } + Keyword::REQUIRE if parser.parse_keyword(Keyword::USER) => { + builder = builder.require_user(true); + } _ => { return parser.expected("end of statement", next_token); } @@ -782,21 +830,9 @@ pub fn parse_create_table( builder = builder.columns(columns).constraints(constraints); } Token::EOF => { - if !builder.validate_schema_info() { - return Err(ParserError::ParserError( - "unexpected end of input".to_string(), - )); - } - break; } Token::SemiColon => { - if !builder.validate_schema_info() { - return Err(ParserError::ParserError( - "unexpected end of input".to_string(), - )); - } - parser.prev_token(); break; } diff --git a/src/keywords.rs b/src/keywords.rs index 1d94ae23a..245b14373 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -300,6 +300,7 @@ define_keywords!( DOMAIN, DOUBLE, DOW, + DOWNSTREAM, DOY, DROP, DRY, @@ -447,10 +448,12 @@ define_keywords!( INCLUDE_NULL_VALUES, INCLUDING, INCREMENT, + INCREMENTAL, INDEX, INDICATOR, INHERIT, INHERITS, + INITIALIZE, INITIALLY, INNER, INOUT, @@ -643,6 +646,8 @@ define_keywords!( ON, ONE, ONLY, + ON_CREATE, + ON_SCHEDULE, OPEN, OPENJSON, OPERATE, @@ -744,6 +749,7 @@ define_keywords!( REF, REFERENCES, REFERENCING, + REFRESH_MODE, REGCLASS, REGEXP, REGR_AVGX, @@ -770,6 +776,7 @@ define_keywords!( REPLICA, REPLICATE, REPLICATION, + REQUIRE, RESET, RESOLVE, RESOURCE, @@ -907,6 +914,7 @@ define_keywords!( TABLESPACE, TAG, TARGET, + TARGET_LAG, TASK, TBLPROPERTIES, TEMP, diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index fe14b7ba5..5bad73365 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -700,6 +700,7 @@ fn test_duckdb_union_datatype() { transient: Default::default(), volatile: Default::default(), iceberg: Default::default(), + dynamic: Default::default(), name: ObjectName::from(vec!["tbl1".into()]), columns: vec![ ColumnDef { @@ -774,7 +775,13 @@ fn test_duckdb_union_datatype() { catalog: Default::default(), catalog_sync: Default::default(), storage_serialization_policy: Default::default(), - table_options: CreateTableOptions::None + table_options: CreateTableOptions::None, + target_lag: None, + warehouse: None, + version: None, + refresh_mode: None, + initialize: None, + require_user: Default::default(), }), stmt ); diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 63e4eecb9..a1e05d030 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -1848,6 +1848,7 @@ fn parse_create_table_with_valid_options() { temporary: false, external: false, global: None, + dynamic: false, if_not_exists: false, transient: false, volatile: false, @@ -1924,7 +1925,13 @@ fn parse_create_table_with_valid_options() { catalog: None, catalog_sync: None, storage_serialization_policy: None, - table_options: CreateTableOptions::With(with_options) + table_options: CreateTableOptions::With(with_options), + target_lag: None, + warehouse: None, + version: None, + refresh_mode: None, + initialize: None, + require_user: false, }) ); } @@ -2031,6 +2038,7 @@ fn parse_create_table_with_identity_column() { temporary: false, external: false, global: None, + dynamic: false, if_not_exists: false, transient: false, volatile: false, @@ -2088,7 +2096,13 @@ fn parse_create_table_with_identity_column() { catalog: None, catalog_sync: None, storage_serialization_policy: None, - table_options: CreateTableOptions::None + table_options: CreateTableOptions::None, + target_lag: None, + warehouse: None, + version: None, + refresh_mode: None, + initialize: None, + require_user: false, }), ); } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index a7c9779b1..43b30aade 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5932,6 +5932,7 @@ fn parse_trigger_related_functions() { temporary: false, external: false, global: None, + dynamic: false, if_not_exists: false, transient: false, volatile: false, @@ -5997,7 +5998,13 @@ fn parse_trigger_related_functions() { catalog: None, catalog_sync: None, storage_serialization_policy: None, - table_options: CreateTableOptions::None + table_options: CreateTableOptions::None, + target_lag: None, + warehouse: None, + version: None, + refresh_mode: None, + initialize: None, + require_user: false, } ); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 55f6ab287..0448e0c46 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -528,23 +528,6 @@ fn test_snowflake_create_table_comment() { } } -#[test] -fn test_snowflake_create_table_incomplete_statement() { - assert_eq!( - snowflake().parse_sql_statements("CREATE TABLE my_table"), - Err(ParserError::ParserError( - "unexpected end of input".to_string() - )) - ); - - assert_eq!( - snowflake().parse_sql_statements("CREATE TABLE my_table; (c int)"), - Err(ParserError::ParserError( - "unexpected end of input".to_string() - )) - ); -} - #[test] fn test_snowflake_single_line_tokenize() { let sql = "CREATE TABLE# this is a comment \ntable_1"; @@ -923,8 +906,8 @@ fn test_snowflake_create_table_with_several_column_options() { #[test] fn test_snowflake_create_iceberg_table_all_options() { match snowflake().verified_stmt("CREATE ICEBERG TABLE my_table (a INT, b INT) \ - CLUSTER BY (a, b) EXTERNAL_VOLUME = 'volume' CATALOG = 'SNOWFLAKE' BASE_LOCATION = 'relative/path' CATALOG_SYNC = 'OPEN_CATALOG' \ - STORAGE_SERIALIZATION_POLICY = COMPATIBLE COPY GRANTS CHANGE_TRACKING=TRUE DATA_RETENTION_TIME_IN_DAYS=5 MAX_DATA_EXTENSION_TIME_IN_DAYS=10 \ + CLUSTER BY (a, b) EXTERNAL_VOLUME='volume' CATALOG='SNOWFLAKE' BASE_LOCATION='relative/path' CATALOG_SYNC='OPEN_CATALOG' \ + STORAGE_SERIALIZATION_POLICY=COMPATIBLE COPY GRANTS CHANGE_TRACKING=TRUE DATA_RETENTION_TIME_IN_DAYS=5 MAX_DATA_EXTENSION_TIME_IN_DAYS=10 \ WITH AGGREGATION POLICY policy_name WITH ROW ACCESS POLICY policy_name ON (a) WITH TAG (A='TAG A', B='TAG B')") { Statement::CreateTable(CreateTable { name, cluster_by, base_location, @@ -972,7 +955,7 @@ fn test_snowflake_create_iceberg_table_all_options() { #[test] fn test_snowflake_create_iceberg_table() { match snowflake() - .verified_stmt("CREATE ICEBERG TABLE my_table (a INT) BASE_LOCATION = 'relative_path'") + .verified_stmt("CREATE ICEBERG TABLE my_table (a INT) BASE_LOCATION='relative_path'") { Statement::CreateTable(CreateTable { name, @@ -1019,27 +1002,6 @@ fn test_snowflake_create_table_trailing_options() { .unwrap(); } -#[test] -fn test_snowflake_create_table_valid_schema_info() { - // Validate there's exactly one source of information on the schema of the new table - assert_eq!( - snowflake() - .parse_sql_statements("CREATE TABLE dst") - .is_err(), - true - ); - assert_eq!( - snowflake().parse_sql_statements("CREATE OR REPLACE TEMP TABLE dst LIKE src AS (SELECT * FROM CUSTOMERS) ON COMMIT PRESERVE ROWS").is_err(), - true - ); - assert_eq!( - snowflake() - .parse_sql_statements("CREATE OR REPLACE TEMP TABLE dst CLONE customers LIKE customer2") - .is_err(), - true - ); -} - #[test] fn parse_sf_create_or_replace_view_with_comment_missing_equal() { assert!(snowflake_and_generic() @@ -1104,6 +1066,79 @@ fn parse_sf_create_table_or_view_with_dollar_quoted_comment() { ); } +#[test] +fn parse_create_dynamic_table() { + snowflake().verified_stmt(r#"CREATE OR REPLACE DYNAMIC TABLE my_dynamic_table TARGET_LAG='20 minutes' WAREHOUSE=mywh AS SELECT product_id, product_name FROM staging_table"#); + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC ICEBERG TABLE my_dynamic_table (date TIMESTAMP_NTZ, id NUMBER, content STRING)", + " EXTERNAL_VOLUME='my_external_volume'", + " CATALOG='SNOWFLAKE'", + " BASE_LOCATION='my_iceberg_table'", + " TARGET_LAG='20 minutes'", + " WAREHOUSE=mywh", + " AS SELECT product_id, product_name FROM staging_table" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_dynamic_table (date TIMESTAMP_NTZ, id NUMBER, content VARIANT)", + " CLUSTER BY (date, id)", + " TARGET_LAG='20 minutes'", + " WAREHOUSE=mywh", + " AS SELECT product_id, product_name FROM staging_table" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_cloned_dynamic_table", + " CLONE my_dynamic_table", + " AT(TIMESTAMP => TO_TIMESTAMP_TZ('04/05/2013 01:02:03', 'mm/dd/yyyy hh24:mi:ss'))" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_cloned_dynamic_table", + " CLONE my_dynamic_table", + " BEFORE(OFFSET => TO_TIMESTAMP_TZ('04/05/2013 01:02:03', 'mm/dd/yyyy hh24:mi:ss'))" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_dynamic_table", + " TARGET_LAG='DOWNSTREAM'", + " WAREHOUSE=mywh", + " INITIALIZE=ON_SCHEDULE", + " REQUIRE USER", + " AS SELECT product_id, product_name FROM staging_table" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_dynamic_table", + " TARGET_LAG='DOWNSTREAM'", + " WAREHOUSE=mywh", + " REFRESH_MODE=AUTO", + " INITIALIZE=ON_SCHEDULE", + " REQUIRE USER", + " AS SELECT product_id, product_name FROM staging_table" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_dynamic_table", + " TARGET_LAG='DOWNSTREAM'", + " WAREHOUSE=mywh", + " REFRESH_MODE=FULL", + " INITIALIZE=ON_SCHEDULE", + " REQUIRE USER", + " AS SELECT product_id, product_name FROM staging_table" + )); + + snowflake().verified_stmt(concat!( + "CREATE DYNAMIC TABLE my_dynamic_table", + " TARGET_LAG='DOWNSTREAM'", + " WAREHOUSE=mywh", + " REFRESH_MODE=INCREMENTAL", + " INITIALIZE=ON_SCHEDULE", + " REQUIRE USER", + " AS SELECT product_id, product_name FROM staging_table" + )); +} + #[test] fn test_sf_derived_table_in_parenthesis() { // Nesting a subquery in an extra set of parentheses is non-standard, @@ -4516,9 +4551,6 @@ fn test_snowflake_identifier_function() { .is_err(), true ); - - snowflake().verified_stmt("GRANT ROLE IDENTIFIER('AAA') TO USER IDENTIFIER('AAA')"); - snowflake().verified_stmt("REVOKE ROLE IDENTIFIER('AAA') FROM USER IDENTIFIER('AAA')"); } #[test] From 56848b03dd8fa88b002ac146411a5cb6d619d733 Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Wed, 20 Aug 2025 19:09:37 +0800 Subject: [PATCH 240/286] feat: support multiple value for pivot (#1970) --- src/ast/query.rs | 11 ++++-- src/ast/spans.rs | 2 +- src/ast/visitor.rs | 2 ++ src/parser/mod.rs | 20 ++++++++++- tests/sqlparser_common.rs | 76 +++++++++++++++++++++++++++++++++++++-- 5 files changed, 104 insertions(+), 7 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 3e06a00ac..781157069 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1336,7 +1336,7 @@ pub enum TableFactor { Pivot { table: Box, aggregate_functions: Vec, // Function expression - value_column: Vec, + value_column: Vec, value_source: PivotValueSource, default_on_null: Option, alias: Option, @@ -2011,10 +2011,15 @@ impl fmt::Display for TableFactor { } => { write!( f, - "{table} PIVOT({} FOR {} IN ({value_source})", + "{table} PIVOT({} FOR ", display_comma_separated(aggregate_functions), - Expr::CompoundIdentifier(value_column.to_vec()), )?; + if value_column.len() == 1 { + write!(f, "{}", value_column[0])?; + } else { + write!(f, "({})", display_comma_separated(value_column))?; + } + write!(f, " IN ({value_source})")?; if let Some(expr) = default_on_null { write!(f, " DEFAULT ON NULL ({expr})")?; } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index fc5b762ae..f8a6e9030 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2001,7 +2001,7 @@ impl Spanned for TableFactor { } => union_spans( core::iter::once(table.span()) .chain(aggregate_functions.iter().map(|i| i.span())) - .chain(value_column.iter().map(|i| i.span)) + .chain(value_column.iter().map(|i| i.span())) .chain(core::iter::once(value_source.span())) .chain(default_on_null.as_ref().map(|i| i.span())) .chain(alias.as_ref().map(|i| i.span())), diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index 8e0a3139a..7840f0e14 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -884,6 +884,8 @@ mod tests { "PRE: EXPR: a.amount", "POST: EXPR: a.amount", "POST: EXPR: SUM(a.amount)", + "PRE: EXPR: a.MONTH", + "POST: EXPR: a.MONTH", "PRE: EXPR: 'JAN'", "POST: EXPR: 'JAN'", "PRE: EXPR: 'FEB'", diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 1afd57137..e5af9cecb 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11067,6 +11067,18 @@ impl<'a> Parser<'a> { self.parse_parenthesized_column_list_inner(optional, allow_empty, |p| p.parse_identifier()) } + pub fn parse_parenthesized_compound_identifier_list( + &mut self, + optional: IsOptional, + allow_empty: bool, + ) -> Result, ParserError> { + self.parse_parenthesized_column_list_inner(optional, allow_empty, |p| { + Ok(Expr::CompoundIdentifier( + p.parse_period_separated(|p| p.parse_identifier())?, + )) + }) + } + /// Parses a parenthesized comma-separated list of index columns, which can be arbitrary /// expressions with ordering information (and an opclass in some dialects). fn parse_parenthesized_index_column_list(&mut self) -> Result, ParserError> { @@ -14187,7 +14199,13 @@ impl<'a> Parser<'a> { self.expect_token(&Token::LParen)?; let aggregate_functions = self.parse_comma_separated(Self::parse_aliased_function_call)?; self.expect_keyword_is(Keyword::FOR)?; - let value_column = self.parse_period_separated(|p| p.parse_identifier())?; + let value_column = if self.peek_token_ref().token == Token::LParen { + self.parse_parenthesized_column_list_inner(Mandatory, false, |p| { + p.parse_subexpr(self.dialect.prec_value(Precedence::Between)) + })? + } else { + vec![self.parse_subexpr(self.dialect.prec_value(Precedence::Between))?] + }; self.expect_keyword_is(Keyword::IN)?; self.expect_token(&Token::LParen)?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 3bf7fef99..ca2289616 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -10961,7 +10961,10 @@ fn parse_pivot_table() { expected_function("b", Some("t")), expected_function("c", Some("u")), ], - value_column: vec![Ident::new("a"), Ident::new("MONTH")], + value_column: vec![Expr::CompoundIdentifier(vec![ + Ident::new("a"), + Ident::new("MONTH") + ])], value_source: PivotValueSource::List(vec![ ExprWithAlias { expr: Expr::value(number("1")), @@ -11008,6 +11011,75 @@ fn parse_pivot_table() { verified_stmt(sql_without_table_alias).to_string(), sql_without_table_alias ); + + let multiple_value_columns_sql = concat!( + "SELECT * FROM person ", + "PIVOT(", + "SUM(age) AS a, AVG(class) AS c ", + "FOR (name, age) IN (('John', 30) AS c1, ('Mike', 40) AS c2))", + ); + + assert_eq!( + verified_only_select(multiple_value_columns_sql).from[0].relation, + Pivot { + table: Box::new(TableFactor::Table { + name: ObjectName::from(vec![Ident::new("person")]), + alias: None, + args: None, + with_hints: vec![], + version: None, + partitions: vec![], + with_ordinality: false, + json_path: None, + sample: None, + index_hints: vec![], + }), + aggregate_functions: vec![ + ExprWithAlias { + expr: call("SUM", [Expr::Identifier(Ident::new("age"))]), + alias: Some(Ident::new("a")) + }, + ExprWithAlias { + expr: call("AVG", [Expr::Identifier(Ident::new("class"))]), + alias: Some(Ident::new("c")) + }, + ], + value_column: vec![ + Expr::Identifier(Ident::new("name")), + Expr::Identifier(Ident::new("age")), + ], + value_source: PivotValueSource::List(vec![ + ExprWithAlias { + expr: Expr::Tuple(vec![ + Expr::Value( + (Value::SingleQuotedString("John".to_string())).with_empty_span() + ), + Expr::Value( + (Value::Number("30".parse().unwrap(), false)).with_empty_span() + ), + ]), + alias: Some(Ident::new("c1")) + }, + ExprWithAlias { + expr: Expr::Tuple(vec![ + Expr::Value( + (Value::SingleQuotedString("Mike".to_string())).with_empty_span() + ), + Expr::Value( + (Value::Number("40".parse().unwrap(), false)).with_empty_span() + ), + ]), + alias: Some(Ident::new("c2")) + }, + ]), + default_on_null: None, + alias: None, + } + ); + assert_eq!( + verified_stmt(multiple_value_columns_sql).to_string(), + multiple_value_columns_sql + ); } #[test] @@ -11340,7 +11412,7 @@ fn parse_pivot_unpivot_table() { expr: call("sum", [Expr::Identifier(Ident::new("population"))]), alias: None }], - value_column: vec![Ident::new("year")], + value_column: vec![Expr::Identifier(Ident::new("year"))], value_source: PivotValueSource::List(vec![ ExprWithAlias { expr: Expr::Value( From cb7a51e85f434d10d5b3ac79cfb09533cca004f7 Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Wed, 20 Aug 2025 19:45:42 +0800 Subject: [PATCH 241/286] feat: Add `ALTER SCHEMA` support (#1980) --- src/ast/ddl.rs | 66 +++++++++++++++++++++++++++++++++++++ src/ast/mod.rs | 27 +++++++++------ src/ast/spans.rs | 30 ++++++++++++++++- src/parser/mod.rs | 40 ++++++++++++++++++++++ tests/sqlparser_bigquery.rs | 12 +++++++ 5 files changed, 164 insertions(+), 11 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 9e6e4517b..db1cb61f6 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -3080,6 +3080,48 @@ impl fmt::Display for CreateConnector { } } +/// An `ALTER SCHEMA` (`Statement::AlterSchema`) operation. +/// +/// See [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#alter_schema_collate_statement) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterSchemaOperation { + SetDefaultCollate { + collate: Expr, + }, + AddReplica { + replica: Ident, + options: Option>, + }, + DropReplica { + replica: Ident, + }, + SetOptionsParens { + options: Vec, + }, +} + +impl fmt::Display for AlterSchemaOperation { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + AlterSchemaOperation::SetDefaultCollate { collate } => { + write!(f, "SET DEFAULT COLLATE {collate}") + } + AlterSchemaOperation::AddReplica { replica, options } => { + write!(f, "ADD REPLICA {replica}")?; + if let Some(options) = options { + write!(f, " OPTIONS ({})", display_comma_separated(options))?; + } + Ok(()) + } + AlterSchemaOperation::DropReplica { replica } => write!(f, "DROP REPLICA {replica}"), + AlterSchemaOperation::SetOptionsParens { options } => { + write!(f, "SET OPTIONS ({})", display_comma_separated(options)) + } + } + } +} /// `RenameTableNameKind` is the kind used in an `ALTER TABLE _ RENAME` statement. /// /// Note: [MySQL] is the only database that supports the AS keyword for this operation. @@ -3102,6 +3144,30 @@ impl fmt::Display for RenameTableNameKind { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterSchema { + pub name: ObjectName, + pub if_exists: bool, + pub operations: Vec, +} + +impl fmt::Display for AlterSchema { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "ALTER SCHEMA ")?; + if self.if_exists { + write!(f, "IF EXISTS ")?; + } + write!(f, "{}", self.name)?; + for operation in &self.operations { + write!(f, " {operation}")?; + } + + Ok(()) + } +} + impl Spanned for RenameTableNameKind { fn span(&self) -> Span { match self { diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 35a864042..7576123b6 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -59,16 +59,17 @@ pub use self::dcl::{ }; pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, - AlterTableAlgorithm, AlterTableLock, AlterTableOperation, AlterType, AlterTypeAddValue, - AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, - ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, - ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, - CreateIndex, CreateTable, Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, - GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, - IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, IndexOption, IndexType, - KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, - RenameTableNameKind, ReplicaIdentity, TableConstraint, TagsColumnOption, - UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, + AlterSchema, AlterSchemaOperation, AlterTableAlgorithm, AlterTableLock, AlterTableOperation, + AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, + AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, + ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, + CreateFunction, CreateIndex, CreateTable, Deduplicate, DeferrableInitial, DropBehavior, + GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, + IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, + IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, + ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, TableConstraint, + TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, + ViewColumnDef, }; pub use self::dml::{Delete, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; @@ -3388,6 +3389,11 @@ pub enum Statement { end_token: AttachedToken, }, /// ```sql + /// ALTER SCHEMA + /// ``` + /// See [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#alter_schema_collate_statement) + AlterSchema(AlterSchema), + /// ```sql /// ALTER INDEX /// ``` AlterIndex { @@ -6336,6 +6342,7 @@ impl fmt::Display for Statement { Statement::Remove(command) => write!(f, "REMOVE {command}"), Statement::ExportData(e) => write!(f, "{e}"), Statement::CreateUser(s) => write!(f, "{s}"), + Statement::AlterSchema(s) => write!(f, "{s}"), } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index f8a6e9030..8eb22c8e3 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -15,7 +15,10 @@ // specific language governing permissions and limitations // under the License. -use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions, ExportData, TypedString}; +use crate::ast::{ + ddl::AlterSchema, query::SelectItemQualifiedWildcardKind, AlterSchemaOperation, ColumnOptions, + ExportData, TypedString, +}; use core::iter; use crate::tokenizer::Span; @@ -548,6 +551,7 @@ impl Spanned for Statement { .chain(connection.iter().map(|i| i.span())), ), Statement::CreateUser(..) => Span::empty(), + Statement::AlterSchema(s) => s.span(), } } } @@ -2387,6 +2391,30 @@ impl Spanned for OpenStatement { } } +impl Spanned for AlterSchemaOperation { + fn span(&self) -> Span { + match self { + AlterSchemaOperation::SetDefaultCollate { collate } => collate.span(), + AlterSchemaOperation::AddReplica { replica, options } => union_spans( + core::iter::once(replica.span) + .chain(options.iter().flat_map(|i| i.iter().map(|i| i.span()))), + ), + AlterSchemaOperation::DropReplica { replica } => replica.span, + AlterSchemaOperation::SetOptionsParens { options } => { + union_spans(options.iter().map(|i| i.span())) + } + } + } +} + +impl Spanned for AlterSchema { + fn span(&self) -> Span { + union_spans( + core::iter::once(self.name.span()).chain(self.operations.iter().map(|i| i.span())), + ) + } +} + #[cfg(test)] pub mod tests { use crate::dialect::{Dialect, GenericDialect, SnowflakeDialect}; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e5af9cecb..57a982862 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9247,8 +9247,14 @@ impl<'a> Parser<'a> { Keyword::POLICY, Keyword::CONNECTOR, Keyword::ICEBERG, + Keyword::SCHEMA, ])?; match object_type { + Keyword::SCHEMA => { + self.prev_token(); + self.prev_token(); + self.parse_alter_schema() + } Keyword::VIEW => self.parse_alter_view(), Keyword::TYPE => self.parse_alter_type(), Keyword::TABLE => self.parse_alter_table(false), @@ -9387,6 +9393,40 @@ impl<'a> Parser<'a> { } } + // Parse a [Statement::AlterSchema] + // ALTER SCHEMA [ IF EXISTS ] schema_name + pub fn parse_alter_schema(&mut self) -> Result { + self.expect_keywords(&[Keyword::ALTER, Keyword::SCHEMA])?; + let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); + let name = self.parse_object_name(false)?; + let operation = if self.parse_keywords(&[Keyword::SET, Keyword::OPTIONS]) { + self.prev_token(); + let options = self.parse_options(Keyword::OPTIONS)?; + AlterSchemaOperation::SetOptionsParens { options } + } else if self.parse_keywords(&[Keyword::SET, Keyword::DEFAULT, Keyword::COLLATE]) { + let collate = self.parse_expr()?; + AlterSchemaOperation::SetDefaultCollate { collate } + } else if self.parse_keywords(&[Keyword::ADD, Keyword::REPLICA]) { + let replica = self.parse_identifier()?; + let options = if self.peek_keyword(Keyword::OPTIONS) { + Some(self.parse_options(Keyword::OPTIONS)?) + } else { + None + }; + AlterSchemaOperation::AddReplica { replica, options } + } else if self.parse_keywords(&[Keyword::DROP, Keyword::REPLICA]) { + let replica = self.parse_identifier()?; + AlterSchemaOperation::DropReplica { replica } + } else { + return self.expected_ref("ALTER SCHEMA operation", self.peek_token_ref()); + }; + Ok(Statement::AlterSchema(AlterSchema { + name, + if_exists, + operations: vec![operation], + })) + } + /// Parse a `CALL procedure_name(arg1, arg2, ...)` /// or `CALL procedure_name` statement pub fn parse_call(&mut self) -> Result { diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 1ca26ae49..4f0cfa3e8 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -2825,3 +2825,15 @@ fn test_begin_transaction() { fn test_begin_statement() { bigquery().verified_stmt("BEGIN"); } + +#[test] +fn test_alter_schema() { + bigquery_and_generic().verified_stmt("ALTER SCHEMA mydataset SET DEFAULT COLLATE 'und:ci'"); + bigquery_and_generic().verified_stmt("ALTER SCHEMA mydataset ADD REPLICA 'us'"); + bigquery_and_generic() + .verified_stmt("ALTER SCHEMA mydataset ADD REPLICA 'us' OPTIONS (location = 'us')"); + bigquery_and_generic().verified_stmt("ALTER SCHEMA mydataset DROP REPLICA 'us'"); + bigquery_and_generic().verified_stmt("ALTER SCHEMA mydataset SET OPTIONS (location = 'us')"); + bigquery_and_generic() + .verified_stmt("ALTER SCHEMA IF EXISTS mydataset SET OPTIONS (location = 'us')"); +} From e9eee00ed928430238cb72e7c31b82ec7c202920 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 21 Aug 2025 16:29:01 +0300 Subject: [PATCH 242/286] Add support for VACUUM in Redshift (#2005) --- src/ast/mod.rs | 52 +++++++++++++++++++++++++++++++++++++ src/ast/spans.rs | 1 + src/keywords.rs | 2 ++ src/parser/mod.rs | 38 +++++++++++++++++++++++++++ tests/sqlparser_redshift.rs | 45 ++++++++++++++++++++++++++++++++ 5 files changed, 138 insertions(+) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 7576123b6..cb6058a0e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4409,6 +4409,13 @@ pub enum Statement { /// ``` /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-user) CreateUser(CreateUser), + /// Re-sorts rows and reclaims space in either a specified table or all tables in the current database + /// + /// ```sql + /// VACUUM tbl + /// ``` + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_VACUUM_command.html) + Vacuum(VacuumStatement), } /// ```sql @@ -6343,6 +6350,7 @@ impl fmt::Display for Statement { Statement::ExportData(e) => write!(f, "{e}"), Statement::CreateUser(s) => write!(f, "{s}"), Statement::AlterSchema(s) => write!(f, "{s}"), + Statement::Vacuum(s) => write!(f, "{s}"), } } } @@ -10604,6 +10612,50 @@ impl fmt::Display for InitializeKind { } } +/// Re-sorts rows and reclaims space in either a specified table or all tables in the current database +/// +/// '''sql +/// VACUUM [ FULL | SORT ONLY | DELETE ONLY | REINDEX | RECLUSTER ] [ \[ table_name \] [ TO threshold PERCENT ] \[ BOOST \] ] +/// ''' +/// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_VACUUM_command.html) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct VacuumStatement { + pub full: bool, + pub sort_only: bool, + pub delete_only: bool, + pub reindex: bool, + pub recluster: bool, + pub table_name: Option, + pub threshold: Option, + pub boost: bool, +} + +impl fmt::Display for VacuumStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "VACUUM{}{}{}{}{}", + if self.full { " FULL" } else { "" }, + if self.sort_only { " SORT ONLY" } else { "" }, + if self.delete_only { " DELETE ONLY" } else { "" }, + if self.reindex { " REINDEX" } else { "" }, + if self.recluster { " RECLUSTER" } else { "" }, + )?; + if let Some(table_name) = &self.table_name { + write!(f, " {table_name}")?; + } + if let Some(threshold) = &self.threshold { + write!(f, " TO {threshold} PERCENT")?; + } + if self.boost { + write!(f, " BOOST")?; + } + Ok(()) + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 8eb22c8e3..7f0175828 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -552,6 +552,7 @@ impl Spanned for Statement { ), Statement::CreateUser(..) => Span::empty(), Statement::AlterSchema(s) => s.span(), + Statement::Vacuum(..) => Span::empty(), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 245b14373..126871302 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -144,6 +144,7 @@ define_keywords!( BLOOMFILTER, BOOL, BOOLEAN, + BOOST, BOTH, BOX, BRIN, @@ -761,6 +762,7 @@ define_keywords!( REGR_SXX, REGR_SXY, REGR_SYY, + REINDEX, RELATIVE, RELAY, RELEASE, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 57a982862..52db37b7d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -649,6 +649,10 @@ impl<'a> Parser<'a> { self.prev_token(); self.parse_export_data() } + Keyword::VACUUM => { + self.prev_token(); + self.parse_vacuum() + } _ => self.expected("an SQL statement", next_token), }, Token::LParen => { @@ -16932,6 +16936,40 @@ impl<'a> Parser<'a> { })) } + fn parse_vacuum(&mut self) -> Result { + self.expect_keyword(Keyword::VACUUM)?; + let full = self.parse_keyword(Keyword::FULL); + let sort_only = self.parse_keywords(&[Keyword::SORT, Keyword::ONLY]); + let delete_only = self.parse_keywords(&[Keyword::DELETE, Keyword::ONLY]); + let reindex = self.parse_keyword(Keyword::REINDEX); + let recluster = self.parse_keyword(Keyword::RECLUSTER); + let (table_name, threshold, boost) = + match self.maybe_parse(|p| p.parse_object_name(false))? { + Some(table_name) => { + let threshold = if self.parse_keyword(Keyword::TO) { + let value = self.parse_value()?; + self.expect_keyword(Keyword::PERCENT)?; + Some(value.value) + } else { + None + }; + let boost = self.parse_keyword(Keyword::BOOST); + (Some(table_name), threshold, boost) + } + _ => (None, None, false), + }; + Ok(Statement::Vacuum(VacuumStatement { + full, + sort_only, + delete_only, + reindex, + recluster, + table_name, + threshold, + boost, + })) + } + /// Consume the parser and return its underlying token buffer pub fn into_tokens(self) -> Vec { self.tokens diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index d539adf66..90652ff48 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -407,3 +407,48 @@ fn parse_string_literal_backslash_escape() { fn parse_utf8_multibyte_idents() { redshift().verified_stmt("SELECT 🚀.city AS 🎸 FROM customers AS 🚀"); } + +#[test] +fn parse_vacuum() { + let stmt = redshift().verified_stmt("VACUUM FULL"); + match stmt { + Statement::Vacuum(v) => { + assert!(v.full); + assert_eq!(v.table_name, None); + } + _ => unreachable!(), + } + let stmt = redshift().verified_stmt("VACUUM tbl"); + match stmt { + Statement::Vacuum(v) => { + assert_eq!( + v.table_name, + Some(ObjectName::from(vec![Ident::new("tbl"),])) + ); + } + _ => unreachable!(), + } + let stmt = redshift().verified_stmt( + "VACUUM FULL SORT ONLY DELETE ONLY REINDEX RECLUSTER db1.sc1.tbl1 TO 20 PERCENT BOOST", + ); + match stmt { + Statement::Vacuum(v) => { + assert!(v.full); + assert!(v.sort_only); + assert!(v.delete_only); + assert!(v.reindex); + assert!(v.recluster); + assert_eq!( + v.table_name, + Some(ObjectName::from(vec![ + Ident::new("db1"), + Ident::new("sc1"), + Ident::new("tbl1"), + ])) + ); + assert_eq!(v.threshold, Some(number("20"))); + assert!(v.boost); + } + _ => unreachable!(), + } +} From 6e80e5c23765bf0042c456d1ffa7e67fe6120cc2 Mon Sep 17 00:00:00 2001 From: Simon Sawert Date: Thu, 21 Aug 2025 17:05:25 +0200 Subject: [PATCH 243/286] Add support for `SEMANTIC_VIEW` table factor (#2009) --- src/ast/query.rs | 59 ++++++++++++++++++ src/ast/spans.rs | 17 ++++++ src/dialect/mod.rs | 14 +++++ src/dialect/snowflake.rs | 4 ++ src/keywords.rs | 4 ++ src/parser/mod.rs | 90 +++++++++++++++++++++++++++- tests/sqlparser_common.rs | 123 ++++++++++++++++++++++++++++++++++++++ 7 files changed, 308 insertions(+), 3 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 781157069..2ef456b1f 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1410,6 +1410,31 @@ pub enum TableFactor { /// The alias for the table. alias: Option, }, + /// Snowflake's SEMANTIC_VIEW function for semantic models. + /// + /// + /// + /// ```sql + /// SELECT * FROM SEMANTIC_VIEW( + /// tpch_analysis + /// DIMENSIONS customer.customer_market_segment + /// METRICS orders.order_average_value + /// ); + /// ``` + SemanticView { + /// The name of the semantic model + name: ObjectName, + /// List of dimensions or expression referring to dimensions (e.g. DATE_PART('year', col)) + dimensions: Vec, + /// List of metrics (references to objects like orders.value, value, orders.*) + metrics: Vec, + /// List of facts or expressions referring to facts or dimensions. + facts: Vec, + /// WHERE clause for filtering + where_clause: Option, + /// The alias for the table + alias: Option, + }, } /// The table sample modifier options @@ -2112,6 +2137,40 @@ impl fmt::Display for TableFactor { } Ok(()) } + TableFactor::SemanticView { + name, + dimensions, + metrics, + facts, + where_clause, + alias, + } => { + write!(f, "SEMANTIC_VIEW({name}")?; + + if !dimensions.is_empty() { + write!(f, " DIMENSIONS {}", display_comma_separated(dimensions))?; + } + + if !metrics.is_empty() { + write!(f, " METRICS {}", display_comma_separated(metrics))?; + } + + if !facts.is_empty() { + write!(f, " FACTS {}", display_comma_separated(facts))?; + } + + if let Some(where_clause) = where_clause { + write!(f, " WHERE {where_clause}")?; + } + + write!(f, ")")?; + + if let Some(alias) = alias { + write!(f, " AS {alias}")?; + } + + Ok(()) + } } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 7f0175828..add6c3904 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2044,6 +2044,23 @@ impl Spanned for TableFactor { .chain(symbols.iter().map(|i| i.span())) .chain(alias.as_ref().map(|i| i.span())), ), + TableFactor::SemanticView { + name, + dimensions, + metrics, + facts, + where_clause, + alias, + } => union_spans( + name.0 + .iter() + .map(|i| i.span()) + .chain(dimensions.iter().map(|d| d.span())) + .chain(metrics.iter().map(|m| m.span())) + .chain(facts.iter().map(|f| f.span())) + .chain(where_clause.as_ref().map(|e| e.span())) + .chain(alias.as_ref().map(|a| a.span())), + ), TableFactor::OpenJsonTable { .. } => Span::empty(), } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 7cf9d4fd1..f91209722 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -1182,6 +1182,20 @@ pub trait Dialect: Debug + Any { fn supports_create_table_like_parenthesized(&self) -> bool { false } + + /// Returns true if the dialect supports `SEMANTIC_VIEW()` table functions. + /// + /// ```sql + /// SELECT * FROM SEMANTIC_VIEW( + /// model_name + /// DIMENSIONS customer.name, customer.region + /// METRICS orders.revenue, orders.count + /// WHERE customer.active = true + /// ) + /// ``` + fn supports_semantic_view_table_factor(&self) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 46c72a799..07ef8317a 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -566,6 +566,10 @@ impl Dialect for SnowflakeDialect { fn supports_select_wildcard_exclude(&self) -> bool { true } + + fn supports_semantic_view_table_factor(&self) -> bool { + true + } } // Peeks ahead to identify tokens that are expected after diff --git a/src/keywords.rs b/src/keywords.rs index 126871302..988f375c0 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -290,6 +290,7 @@ define_keywords!( DETACH, DETAIL, DETERMINISTIC, + DIMENSIONS, DIRECTORY, DISABLE, DISCARD, @@ -359,6 +360,7 @@ define_keywords!( EXTERNAL, EXTERNAL_VOLUME, EXTRACT, + FACTS, FAIL, FAILOVER, FALSE, @@ -566,6 +568,7 @@ define_keywords!( METADATA, METHOD, METRIC, + METRICS, MICROSECOND, MICROSECONDS, MILLENIUM, @@ -828,6 +831,7 @@ define_keywords!( SECURITY, SEED, SELECT, + SEMANTIC_VIEW, SEMI, SENSITIVE, SEPARATOR, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 52db37b7d..6179b8345 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4245,6 +4245,18 @@ impl<'a> Parser<'a> { /// not be efficient as it does a loop on the tokens with `peek_nth_token` /// each time. pub fn parse_keyword_with_tokens(&mut self, expected: Keyword, tokens: &[Token]) -> bool { + self.keyword_with_tokens(expected, tokens, true) + } + + /// Peeks to see if the current token is the `expected` keyword followed by specified tokens + /// without consuming them. + /// + /// See [Self::parse_keyword_with_tokens] for details. + pub(crate) fn peek_keyword_with_tokens(&mut self, expected: Keyword, tokens: &[Token]) -> bool { + self.keyword_with_tokens(expected, tokens, false) + } + + fn keyword_with_tokens(&mut self, expected: Keyword, tokens: &[Token], consume: bool) -> bool { match &self.peek_token_ref().token { Token::Word(w) if expected == w.keyword => { for (idx, token) in tokens.iter().enumerate() { @@ -4252,10 +4264,13 @@ impl<'a> Parser<'a> { return false; } } - // consume all tokens - for _ in 0..(tokens.len() + 1) { - self.advance_token(); + + if consume { + for _ in 0..(tokens.len() + 1) { + self.advance_token(); + } } + true } _ => false, @@ -13397,6 +13412,7 @@ impl<'a> Parser<'a> { | TableFactor::Pivot { alias, .. } | TableFactor::Unpivot { alias, .. } | TableFactor::MatchRecognize { alias, .. } + | TableFactor::SemanticView { alias, .. } | TableFactor::NestedJoin { alias, .. } => { // but not `FROM (mytable AS alias1) AS alias2`. if let Some(inner_alias) = alias { @@ -13511,6 +13527,10 @@ impl<'a> Parser<'a> { } else if self.parse_keyword_with_tokens(Keyword::XMLTABLE, &[Token::LParen]) { self.prev_token(); self.parse_xml_table_factor() + } else if self.dialect.supports_semantic_view_table_factor() + && self.peek_keyword_with_tokens(Keyword::SEMANTIC_VIEW, &[Token::LParen]) + { + self.parse_semantic_view_table_factor() } else { let name = self.parse_object_name(true)?; @@ -13842,6 +13862,70 @@ impl<'a> Parser<'a> { Ok(XmlPassingClause { arguments }) } + /// Parse a [TableFactor::SemanticView] + fn parse_semantic_view_table_factor(&mut self) -> Result { + self.expect_keyword(Keyword::SEMANTIC_VIEW)?; + self.expect_token(&Token::LParen)?; + + let name = self.parse_object_name(true)?; + + // Parse DIMENSIONS, METRICS, FACTS and WHERE clauses in flexible order + let mut dimensions = Vec::new(); + let mut metrics = Vec::new(); + let mut facts = Vec::new(); + let mut where_clause = None; + + while self.peek_token().token != Token::RParen { + if self.parse_keyword(Keyword::DIMENSIONS) { + if !dimensions.is_empty() { + return Err(ParserError::ParserError( + "DIMENSIONS clause can only be specified once".to_string(), + )); + } + dimensions = self.parse_comma_separated(Parser::parse_expr)?; + } else if self.parse_keyword(Keyword::METRICS) { + if !metrics.is_empty() { + return Err(ParserError::ParserError( + "METRICS clause can only be specified once".to_string(), + )); + } + metrics = self.parse_comma_separated(|parser| parser.parse_object_name(true))?; + } else if self.parse_keyword(Keyword::FACTS) { + if !facts.is_empty() { + return Err(ParserError::ParserError( + "FACTS clause can only be specified once".to_string(), + )); + } + facts = self.parse_comma_separated(Parser::parse_expr)?; + } else if self.parse_keyword(Keyword::WHERE) { + if where_clause.is_some() { + return Err(ParserError::ParserError( + "WHERE clause can only be specified once".to_string(), + )); + } + where_clause = Some(self.parse_expr()?); + } else { + return parser_err!( + "Expected one of DIMENSIONS, METRICS, FACTS or WHERE", + self.peek_token().span.start + )?; + } + } + + self.expect_token(&Token::RParen)?; + + let alias = self.maybe_parse_table_alias()?; + + Ok(TableFactor::SemanticView { + name, + dimensions, + metrics, + facts, + where_clause, + alias, + }) + } + fn parse_match_recognize(&mut self, table: TableFactor) -> Result { self.expect_token(&Token::LParen)?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index ca2289616..f7a0b1d1f 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16841,3 +16841,126 @@ fn parse_copy_options() { _ => unreachable!(), } } + +#[test] +fn test_parse_semantic_view_table_factor() { + let dialects = all_dialects_where(|d| d.supports_semantic_view_table_factor()); + + let valid_sqls = [ + ("SELECT * FROM SEMANTIC_VIEW(model)", None), + ( + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1, dim2)", + None, + ), + ("SELECT * FROM SEMANTIC_VIEW(a.b METRICS c.d, c.e)", None), + ( + "SELECT * FROM SEMANTIC_VIEW(model FACTS fact1, fact2)", + None, + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model FACTS DATE_PART('year', col))", + None, + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 METRICS met1)", + None, + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 WHERE x > 0)", + None, + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1) AS sv", + None, + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS DATE_PART('year', col))", + None, + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model METRICS orders.col, orders.col2)", + None, + ), + // We can parse in any order but will always produce a result in a fixed order. + ( + "SELECT * FROM SEMANTIC_VIEW(model WHERE x > 0 DIMENSIONS dim1)", + Some("SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 WHERE x > 0)"), + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model METRICS met1 DIMENSIONS dim1)", + Some("SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 METRICS met1)"), + ), + ( + "SELECT * FROM SEMANTIC_VIEW(model FACTS fact1 DIMENSIONS dim1)", + Some("SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 FACTS fact1)"), + ), + ]; + + for (input_sql, expected_sql) in valid_sqls { + if let Some(expected) = expected_sql { + // Test that non-canonical order gets normalized + let parsed = dialects.parse_sql_statements(input_sql).unwrap(); + let formatted = parsed[0].to_string(); + assert_eq!(formatted, expected); + } else { + dialects.verified_stmt(input_sql); + } + } + + let invalid_sqls = [ + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 INVALID inv1)", + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 DIMENSIONS dim2)", + "SELECT * FROM SEMANTIC_VIEW(model METRICS SUM(met1.avg))", + ]; + + for sql in invalid_sqls { + let result = dialects.parse_sql_statements(sql); + assert!(result.is_err(), "Expected error for invalid SQL: {}", sql); + } + + let ast_sql = r#"SELECT * FROM SEMANTIC_VIEW( + my_model + DIMENSIONS DATE_PART('year', date_col), region_name + METRICS orders.revenue, orders.count + WHERE active = true + ) AS model_alias"#; + + let stmt = dialects.parse_sql_statements(ast_sql).unwrap(); + match &stmt[0] { + Statement::Query(q) => { + if let SetExpr::Select(select) = q.body.as_ref() { + if let Some(TableWithJoins { relation, .. }) = select.from.first() { + match relation { + TableFactor::SemanticView { + name, + dimensions, + metrics, + facts, + where_clause, + alias, + } => { + assert_eq!(name.to_string(), "my_model"); + assert_eq!(dimensions.len(), 2); + assert_eq!(dimensions[0].to_string(), "DATE_PART('year', date_col)"); + assert_eq!(dimensions[1].to_string(), "region_name"); + assert_eq!(metrics.len(), 2); + assert_eq!(metrics[0].to_string(), "orders.revenue"); + assert_eq!(metrics[1].to_string(), "orders.count"); + assert!(facts.is_empty()); + assert!(where_clause.is_some()); + assert_eq!(where_clause.as_ref().unwrap().to_string(), "active = true"); + assert!(alias.is_some()); + assert_eq!(alias.as_ref().unwrap().name.value, "model_alias"); + } + _ => panic!("Expected SemanticView table factor"), + } + } else { + panic!("Expected table in FROM clause"); + } + } else { + panic!("Expected SELECT statement"); + } + } + _ => panic!("Expected Query statement"), + } +} From 5d5c90c77fd5fd0e25250c043ef91632cba4ca8d Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 21 Aug 2025 19:45:32 +0300 Subject: [PATCH 244/286] Redshift: Add more copy options (#2008) --- src/ast/mod.rs | 54 +++++++++++++++++++++++++++++----- src/keywords.rs | 7 +++++ src/parser/mod.rs | 61 +++++++++++++++++++++++++++++++++------ tests/sqlparser_common.rs | 32 ++++++++++++++++++++ 4 files changed, 138 insertions(+), 16 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index cb6058a0e..cd9378575 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -8774,32 +8774,54 @@ impl fmt::Display for CopyOption { /// An option in `COPY` statement before PostgreSQL version 9.0. /// -/// +/// [PostgreSQL](https://www.postgresql.org/docs/8.4/sql-copy.html) +/// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_COPY-alphabetical-parm-list.html) #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum CopyLegacyOption { + /// ACCEPTANYDATE + AcceptAnyDate, + /// ACCEPTINVCHARS + AcceptInvChars(Option), /// BINARY Binary, - /// DELIMITER \[ AS \] 'delimiter_character' - Delimiter(char), - /// NULL \[ AS \] 'null_string' - Null(String), + /// BLANKSASNULL + BlankAsNull, /// CSV ... Csv(Vec), + /// DATEFORMAT \[ AS \] {'dateformat_string' | 'auto' } + DateFormat(Option), + /// DELIMITER \[ AS \] 'delimiter_character' + Delimiter(char), + /// EMPTYASNULL + EmptyAsNull, /// IAM_ROLE { DEFAULT | 'arn:aws:iam::123456789:role/role1' } IamRole(IamRoleKind), /// IGNOREHEADER \[ AS \] number_rows IgnoreHeader(u64), + /// NULL \[ AS \] 'null_string' + Null(String), + /// TIMEFORMAT \[ AS \] {'timeformat_string' | 'auto' | 'epochsecs' | 'epochmillisecs' } + TimeFormat(Option), + /// TRUNCATECOLUMNS + TruncateColumns, } impl fmt::Display for CopyLegacyOption { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { use CopyLegacyOption::*; match self { + AcceptAnyDate => write!(f, "ACCEPTANYDATE"), + AcceptInvChars(ch) => { + write!(f, "ACCEPTINVCHARS")?; + if let Some(ch) = ch { + write!(f, " '{}'", value::escape_single_quote_string(ch))?; + } + Ok(()) + } Binary => write!(f, "BINARY"), - Delimiter(char) => write!(f, "DELIMITER '{char}'"), - Null(string) => write!(f, "NULL '{}'", value::escape_single_quote_string(string)), + BlankAsNull => write!(f, "BLANKSASNULL"), Csv(opts) => { write!(f, "CSV")?; if !opts.is_empty() { @@ -8807,8 +8829,26 @@ impl fmt::Display for CopyLegacyOption { } Ok(()) } + DateFormat(fmt) => { + write!(f, "DATEFORMAT")?; + if let Some(fmt) = fmt { + write!(f, " '{}'", value::escape_single_quote_string(fmt))?; + } + Ok(()) + } + Delimiter(char) => write!(f, "DELIMITER '{char}'"), + EmptyAsNull => write!(f, "EMPTYASNULL"), IamRole(role) => write!(f, "IAM_ROLE {role}"), IgnoreHeader(num_rows) => write!(f, "IGNOREHEADER {num_rows}"), + Null(string) => write!(f, "NULL '{}'", value::escape_single_quote_string(string)), + TimeFormat(fmt) => { + write!(f, "TIMEFORMAT")?; + if let Some(fmt) = fmt { + write!(f, " '{}'", value::escape_single_quote_string(fmt))?; + } + Ok(()) + } + TruncateColumns => write!(f, "TRUNCATECOLUMNS"), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 988f375c0..d78a6c178 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -76,6 +76,8 @@ define_keywords!( ABS, ABSENT, ABSOLUTE, + ACCEPTANYDATE, + ACCEPTINVCHARS, ACCESS, ACCOUNT, ACTION, @@ -138,6 +140,7 @@ define_keywords!( BIND, BINDING, BIT, + BLANKSASNULL, BLOB, BLOCK, BLOOM, @@ -255,6 +258,7 @@ define_keywords!( DATA_RETENTION_TIME_IN_DAYS, DATE, DATE32, + DATEFORMAT, DATETIME, DATETIME64, DAY, @@ -314,6 +318,7 @@ define_keywords!( ELSE, ELSEIF, EMPTY, + EMPTYASNULL, ENABLE, ENABLE_SCHEMA_EVOLUTION, ENCODING, @@ -933,6 +938,7 @@ define_keywords!( THEN, TIES, TIME, + TIMEFORMAT, TIMESTAMP, TIMESTAMPTZ, TIMESTAMP_NTZ, @@ -961,6 +967,7 @@ define_keywords!( TRIM_ARRAY, TRUE, TRUNCATE, + TRUNCATECOLUMNS, TRY, TRY_CAST, TRY_CONVERT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6179b8345..c4c72e9c4 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9602,23 +9602,38 @@ impl<'a> Parser<'a> { } fn parse_copy_legacy_option(&mut self) -> Result { + // FORMAT \[ AS \] is optional + if self.parse_keyword(Keyword::FORMAT) { + let _ = self.parse_keyword(Keyword::AS); + } + let ret = match self.parse_one_of_keywords(&[ + Keyword::ACCEPTANYDATE, + Keyword::ACCEPTINVCHARS, Keyword::BINARY, - Keyword::DELIMITER, - Keyword::NULL, + Keyword::BLANKSASNULL, Keyword::CSV, + Keyword::DATEFORMAT, + Keyword::DELIMITER, + Keyword::EMPTYASNULL, Keyword::IAM_ROLE, Keyword::IGNOREHEADER, + Keyword::NULL, + Keyword::TIMEFORMAT, + Keyword::TRUNCATECOLUMNS, ]) { - Some(Keyword::BINARY) => CopyLegacyOption::Binary, - Some(Keyword::DELIMITER) => { + Some(Keyword::ACCEPTANYDATE) => CopyLegacyOption::AcceptAnyDate, + Some(Keyword::ACCEPTINVCHARS) => { let _ = self.parse_keyword(Keyword::AS); // [ AS ] - CopyLegacyOption::Delimiter(self.parse_literal_char()?) - } - Some(Keyword::NULL) => { - let _ = self.parse_keyword(Keyword::AS); // [ AS ] - CopyLegacyOption::Null(self.parse_literal_string()?) + let ch = if matches!(self.peek_token().token, Token::SingleQuotedString(_)) { + Some(self.parse_literal_string()?) + } else { + None + }; + CopyLegacyOption::AcceptInvChars(ch) } + Some(Keyword::BINARY) => CopyLegacyOption::Binary, + Some(Keyword::BLANKSASNULL) => CopyLegacyOption::BlankAsNull, Some(Keyword::CSV) => CopyLegacyOption::Csv({ let mut opts = vec![]; while let Some(opt) = @@ -9628,12 +9643,40 @@ impl<'a> Parser<'a> { } opts }), + Some(Keyword::DATEFORMAT) => { + let _ = self.parse_keyword(Keyword::AS); + let fmt = if matches!(self.peek_token().token, Token::SingleQuotedString(_)) { + Some(self.parse_literal_string()?) + } else { + None + }; + CopyLegacyOption::DateFormat(fmt) + } + Some(Keyword::DELIMITER) => { + let _ = self.parse_keyword(Keyword::AS); + CopyLegacyOption::Delimiter(self.parse_literal_char()?) + } + Some(Keyword::EMPTYASNULL) => CopyLegacyOption::EmptyAsNull, Some(Keyword::IAM_ROLE) => CopyLegacyOption::IamRole(self.parse_iam_role_kind()?), Some(Keyword::IGNOREHEADER) => { let _ = self.parse_keyword(Keyword::AS); let num_rows = self.parse_literal_uint()?; CopyLegacyOption::IgnoreHeader(num_rows) } + Some(Keyword::NULL) => { + let _ = self.parse_keyword(Keyword::AS); + CopyLegacyOption::Null(self.parse_literal_string()?) + } + Some(Keyword::TIMEFORMAT) => { + let _ = self.parse_keyword(Keyword::AS); + let fmt = if matches!(self.peek_token().token, Token::SingleQuotedString(_)) { + Some(self.parse_literal_string()?) + } else { + None + }; + CopyLegacyOption::TimeFormat(fmt) + } + Some(Keyword::TRUNCATECOLUMNS) => CopyLegacyOption::TruncateColumns, _ => self.expected("option", self.peek_token())?, }; Ok(ret) diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index f7a0b1d1f..54ad1732b 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16840,6 +16840,38 @@ fn parse_copy_options() { } _ => unreachable!(), } + one_statement_parses_to( + concat!( + "COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' ", + "ACCEPTANYDATE ", + "ACCEPTINVCHARS AS '*' ", + "BLANKSASNULL ", + "CSV ", + "DATEFORMAT AS 'DD-MM-YYYY' ", + "EMPTYASNULL ", + "IAM_ROLE DEFAULT ", + "IGNOREHEADER AS 1 ", + "TIMEFORMAT AS 'auto' ", + "TRUNCATECOLUMNS", + ), + concat!( + "COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' ", + "ACCEPTANYDATE ", + "ACCEPTINVCHARS '*' ", + "BLANKSASNULL ", + "CSV ", + "DATEFORMAT 'DD-MM-YYYY' ", + "EMPTYASNULL ", + "IAM_ROLE DEFAULT ", + "IGNOREHEADER 1 ", + "TIMEFORMAT 'auto' ", + "TRUNCATECOLUMNS", + ), + ); + one_statement_parses_to( + "COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' FORMAT AS CSV", + "COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' CSV", + ); } #[test] From 376f47e3d14c3ebf3cdaef750b22948ccc549add Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Fri, 22 Aug 2025 12:32:09 +0200 Subject: [PATCH 245/286] feat: MERGE statements: add RETURNING and OUTPUT without INTO (#2011) --- src/ast/mod.rs | 40 +++++++++++++++++++++----------- src/ast/query.rs | 2 ++ src/ast/spans.rs | 1 + src/parser/mod.rs | 49 ++++++++++++++++++++++++++++++--------- tests/sqlparser_common.rs | 31 +++++++++++++++++++++++++ 5 files changed, 98 insertions(+), 25 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index cd9378575..71cb6c5ee 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -9107,24 +9107,36 @@ impl Display for MergeClause { #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct OutputClause { - pub select_items: Vec, - pub into_table: SelectInto, +pub enum OutputClause { + Output { + select_items: Vec, + into_table: Option, + }, + Returning { + select_items: Vec, + }, } impl fmt::Display for OutputClause { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - let OutputClause { - select_items, - into_table, - } = self; - - write!( - f, - "OUTPUT {} {}", - display_comma_separated(select_items), - into_table - ) + match self { + OutputClause::Output { + select_items, + into_table, + } => { + f.write_str("OUTPUT ")?; + display_comma_separated(select_items).fmt(f)?; + if let Some(into_table) = into_table { + f.write_str(" ")?; + into_table.fmt(f)?; + } + Ok(()) + } + OutputClause::Returning { select_items } => { + f.write_str("RETURNING ")?; + display_comma_separated(select_items).fmt(f) + } + } } } diff --git a/src/ast/query.rs b/src/ast/query.rs index 2ef456b1f..967af85c2 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -161,6 +161,7 @@ pub enum SetExpr { Insert(Statement), Update(Statement), Delete(Statement), + Merge(Statement), Table(Box
), } @@ -188,6 +189,7 @@ impl fmt::Display for SetExpr { SetExpr::Insert(v) => v.fmt(f), SetExpr::Update(v) => v.fmt(f), SetExpr::Delete(v) => v.fmt(f), + SetExpr::Merge(v) => v.fmt(f), SetExpr::Table(t) => t.fmt(f), SetExpr::SetOperation { left, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index add6c3904..39761751f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -214,6 +214,7 @@ impl Spanned for SetExpr { SetExpr::Table(_) => Span::empty(), SetExpr::Update(statement) => statement.span(), SetExpr::Delete(statement) => statement.span(), + SetExpr::Merge(statement) => statement.span(), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index c4c72e9c4..6c559eed4 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -11508,6 +11508,13 @@ impl<'a> Parser<'a> { Ok(Box::new(SetExpr::Delete(self.parse_delete()?))) } + /// Parse a MERGE statement, returning a `Box`ed SetExpr + /// + /// This is used to reduce the size of the stack frames in debug builds + fn parse_merge_setexpr_boxed(&mut self) -> Result, ParserError> { + Ok(Box::new(SetExpr::Merge(self.parse_merge()?))) + } + pub fn parse_delete(&mut self) -> Result { let (tables, with_from_keyword) = if !self.parse_keyword(Keyword::FROM) { // `FROM` keyword is optional in BigQuery SQL. @@ -11719,6 +11726,20 @@ impl<'a> Parser<'a> { pipe_operators: vec![], } .into()) + } else if self.parse_keyword(Keyword::MERGE) { + Ok(Query { + with, + body: self.parse_merge_setexpr_boxed()?, + limit_clause: None, + order_by: None, + fetch: None, + locks: vec![], + for_clause: None, + settings: None, + format_clause: None, + pipe_operators: vec![], + } + .into()) } else { let body = self.parse_query_body(self.dialect.prec_unknown())?; @@ -16571,15 +16592,22 @@ impl<'a> Parser<'a> { Ok(clauses) } - fn parse_output(&mut self) -> Result { - self.expect_keyword_is(Keyword::OUTPUT)?; + fn parse_output(&mut self, start_keyword: Keyword) -> Result { let select_items = self.parse_projection()?; - self.expect_keyword_is(Keyword::INTO)?; - let into_table = self.parse_select_into()?; + let into_table = if start_keyword == Keyword::OUTPUT && self.peek_keyword(Keyword::INTO) { + self.expect_keyword_is(Keyword::INTO)?; + Some(self.parse_select_into()?) + } else { + None + }; - Ok(OutputClause { - select_items, - into_table, + Ok(if start_keyword == Keyword::OUTPUT { + OutputClause::Output { + select_items, + into_table, + } + } else { + OutputClause::Returning { select_items } }) } @@ -16609,10 +16637,9 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::ON)?; let on = self.parse_expr()?; let clauses = self.parse_merge_clauses()?; - let output = if self.peek_keyword(Keyword::OUTPUT) { - Some(self.parse_output()?) - } else { - None + let output = match self.parse_one_of_keywords(&[Keyword::OUTPUT, Keyword::RETURNING]) { + Some(start_keyword) => Some(self.parse_output(start_keyword)?), + None => None, }; Ok(Statement::Merge { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 54ad1732b..8b99bb1dc 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -9902,6 +9902,29 @@ fn parse_merge() { verified_stmt(sql); } +#[test] +fn test_merge_in_cte() { + verified_only_select( + "WITH x AS (\ + MERGE INTO t USING (VALUES (1)) ON 1 = 1 \ + WHEN MATCHED THEN DELETE \ + RETURNING *\ + ) SELECT * FROM x", + ); +} + +#[test] +fn test_merge_with_returning() { + let sql = "MERGE INTO wines AS w \ + USING wine_stock_changes AS s \ + ON s.winename = w.winename \ + WHEN NOT MATCHED AND s.stock_delta > 0 THEN INSERT VALUES (s.winename, s.stock_delta) \ + WHEN MATCHED AND w.stock + s.stock_delta > 0 THEN UPDATE SET stock = w.stock + s.stock_delta \ + WHEN MATCHED THEN DELETE \ + RETURNING merge_action(), w.*"; + verified_stmt(sql); +} + #[test] fn test_merge_with_output() { let sql = "MERGE INTO target_table USING source_table \ @@ -9915,6 +9938,14 @@ fn test_merge_with_output() { verified_stmt(sql); } +#[test] +fn test_merge_with_output_without_into() { + let sql = "MERGE INTO a USING b ON a.id = b.id \ + WHEN MATCHED THEN DELETE \ + OUTPUT inserted.*"; + verified_stmt(sql); +} + #[test] fn test_merge_into_using_table() { let sql = "MERGE INTO target_table USING source_table \ From 779dcf91f62bc72c89ee0815725fc626c2204d43 Mon Sep 17 00:00:00 2001 From: Simon Vandel Sillesen Date: Sat, 23 Aug 2025 09:11:29 +0200 Subject: [PATCH 246/286] `GenericDialect`: Support pipe operator (#2012) --- src/dialect/generic.rs | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index b4c3ef027..dffc5b527 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -64,6 +64,10 @@ impl Dialect for GenericDialect { true } + fn supports_pipe_operator(&self) -> bool { + true + } + fn supports_start_transaction_modifier(&self) -> bool { true } From cffff309610247dc278e5c1ef6ec7e643e87ab08 Mon Sep 17 00:00:00 2001 From: Denys Tsomenko Date: Tue, 26 Aug 2025 22:20:13 +0300 Subject: [PATCH 247/286] Add SECURE keyword for views in Snowflake (#2004) --- src/ast/mod.rs | 7 ++++++- src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 9 +++++++-- tests/sqlparser_common.rs | 7 +++++++ tests/sqlparser_snowflake.rs | 27 +++++++++++++++++++++++++++ 6 files changed, 49 insertions(+), 3 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 71cb6c5ee..a4371566e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -3256,6 +3256,9 @@ pub enum Statement { or_alter: bool, or_replace: bool, materialized: bool, + /// Snowflake: SECURE view modifier + /// + secure: bool, /// View name name: ObjectName, /// If `if_not_exists` is true, this flag is set to true if the view name comes before the `IF NOT EXISTS` clause. @@ -5118,6 +5121,7 @@ impl fmt::Display for Statement { columns, query, materialized, + secure, options, cluster_by, comment, @@ -5139,7 +5143,7 @@ impl fmt::Display for Statement { } write!( f, - "{materialized}{temporary}VIEW {if_not_and_name}{to}", + "{secure}{materialized}{temporary}VIEW {if_not_and_name}{to}", if_not_and_name = if *if_not_exists { if *name_before_not_exists { format!("{name} IF NOT EXISTS") @@ -5149,6 +5153,7 @@ impl fmt::Display for Statement { } else { format!("{name}") }, + secure = if *secure { "SECURE " } else { "" }, materialized = if *materialized { "MATERIALIZED " } else { "" }, temporary = if *temporary { "TEMPORARY " } else { "" }, to = to diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 39761751f..6978b627f 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -395,6 +395,7 @@ impl Spanned for Statement { or_alter: _, or_replace: _, materialized: _, + secure: _, name, columns, query, diff --git a/src/keywords.rs b/src/keywords.rs index d78a6c178..eb27d1c4e 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -833,6 +833,7 @@ define_keywords!( SECONDARY_ENGINE_ATTRIBUTE, SECONDS, SECRET, + SECURE, SECURITY, SEED, SELECT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 6c559eed4..486664e20 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4732,8 +4732,11 @@ impl<'a> Parser<'a> { let create_view_params = self.parse_create_view_params()?; if self.parse_keyword(Keyword::TABLE) { self.parse_create_table(or_replace, temporary, global, transient) - } else if self.parse_keyword(Keyword::MATERIALIZED) || self.parse_keyword(Keyword::VIEW) { - self.prev_token(); + } else if self.peek_keyword(Keyword::MATERIALIZED) + || self.peek_keyword(Keyword::VIEW) + || self.peek_keywords(&[Keyword::SECURE, Keyword::MATERIALIZED, Keyword::VIEW]) + || self.peek_keywords(&[Keyword::SECURE, Keyword::VIEW]) + { self.parse_create_view(or_alter, or_replace, temporary, create_view_params) } else if self.parse_keyword(Keyword::POLICY) { self.parse_create_policy() @@ -5853,6 +5856,7 @@ impl<'a> Parser<'a> { temporary: bool, create_view_params: Option, ) -> Result { + let secure = self.parse_keyword(Keyword::SECURE); let materialized = self.parse_keyword(Keyword::MATERIALIZED); self.expect_keyword_is(Keyword::VIEW)?; let allow_unquoted_hyphen = dialect_of!(self is BigQueryDialect); @@ -5923,6 +5927,7 @@ impl<'a> Parser<'a> { columns, query, materialized, + secure, or_replace, options, cluster_by, diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 8b99bb1dc..2e2d9bfd0 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -8078,6 +8078,7 @@ fn parse_create_view() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -8147,6 +8148,7 @@ fn parse_create_view_with_columns() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); @@ -8197,6 +8199,7 @@ fn parse_create_view_temporary() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -8237,6 +8240,7 @@ fn parse_create_or_replace_view() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); @@ -8281,6 +8285,7 @@ fn parse_create_or_replace_materialized_view() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); @@ -8321,6 +8326,7 @@ fn parse_create_materialized_view() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); @@ -8361,6 +8367,7 @@ fn parse_create_materialized_view_with_cluster_by() { to, params, name_before_not_exists: _, + secure: _, } => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 0448e0c46..75e649fe6 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -44,6 +44,33 @@ fn test_snowflake_create_table() { } } +#[test] +fn parse_sf_create_secure_view_and_materialized_view() { + for sql in [ + "CREATE SECURE VIEW v AS SELECT 1", + "CREATE SECURE MATERIALIZED VIEW v AS SELECT 1", + "CREATE OR REPLACE SECURE VIEW v AS SELECT 1", + "CREATE OR REPLACE SECURE MATERIALIZED VIEW v AS SELECT 1", + ] { + match snowflake().verified_stmt(sql) { + Statement::CreateView { + secure, + materialized, + .. + } => { + assert!(secure); + if sql.contains("MATERIALIZED") { + assert!(materialized); + } else { + assert!(!materialized); + } + } + _ => unreachable!(), + } + assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); + } +} + #[test] fn test_snowflake_create_or_replace_table() { let sql = "CREATE OR REPLACE TABLE my_table (a number)"; From 9f515bf8c3ad5a1f566e751c54fca2bf15a17d03 Mon Sep 17 00:00:00 2001 From: Adam Johnson Date: Tue, 26 Aug 2025 20:22:26 +0100 Subject: [PATCH 248/286] Add support for PostgreSQL JSON function 'RETURNING' clauses (#2001) --- src/ast/mod.rs | 33 +++++++++++++-- src/ast/spans.rs | 1 + src/parser/mod.rs | 29 +++++++++++-- tests/sqlparser_postgres.rs | 83 ++++++++++++++++++++++++++++++++++++- 4 files changed, 139 insertions(+), 7 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index a4371566e..615ceb1ec 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -7822,11 +7822,16 @@ pub enum FunctionArgumentClause { /// /// [`GROUP_CONCAT`]: https://dev.mysql.com/doc/refman/8.0/en/aggregate-functions.html#function_group-concat Separator(Value), - /// The json-null-clause to the [`JSON_ARRAY`]/[`JSON_OBJECT`] function in MSSQL. + /// The `ON NULL` clause for some JSON functions. /// - /// [`JSON_ARRAY`]: - /// [`JSON_OBJECT`]: + /// [MSSQL `JSON_ARRAY`](https://learn.microsoft.com/en-us/sql/t-sql/functions/json-array-transact-sql?view=sql-server-ver16) + /// [MSSQL `JSON_OBJECT`](https://learn.microsoft.com/en-us/sql/t-sql/functions/json-object-transact-sql?view=sql-server-ver16>) + /// [PostgreSQL JSON functions](https://www.postgresql.org/docs/current/functions-json.html#FUNCTIONS-JSON-PROCESSING) JsonNullClause(JsonNullClause), + /// The `RETURNING` clause for some JSON functions in PostgreSQL + /// + /// [`JSON_OBJECT`](https://www.postgresql.org/docs/current/functions-json.html#:~:text=json_object) + JsonReturningClause(JsonReturningClause), } impl fmt::Display for FunctionArgumentClause { @@ -7843,6 +7848,9 @@ impl fmt::Display for FunctionArgumentClause { FunctionArgumentClause::Having(bound) => write!(f, "{bound}"), FunctionArgumentClause::Separator(sep) => write!(f, "SEPARATOR {sep}"), FunctionArgumentClause::JsonNullClause(null_clause) => write!(f, "{null_clause}"), + FunctionArgumentClause::JsonReturningClause(returning_clause) => { + write!(f, "{returning_clause}") + } } } } @@ -10177,6 +10185,25 @@ impl Display for JsonNullClause { } } +/// PostgreSQL JSON function RETURNING clause +/// +/// Example: +/// ```sql +/// JSON_OBJECT('a': 1 RETURNING jsonb) +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct JsonReturningClause { + pub data_type: DataType, +} + +impl Display for JsonReturningClause { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!(f, "RETURNING {}", self.data_type) + } +} + /// rename object definition #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 6978b627f..5d3694be8 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1792,6 +1792,7 @@ impl Spanned for FunctionArgumentClause { FunctionArgumentClause::Having(HavingBound(_kind, expr)) => expr.span(), FunctionArgumentClause::Separator(value) => value.span(), FunctionArgumentClause::JsonNullClause(_) => Span::empty(), + FunctionArgumentClause::JsonReturningClause(_) => Span::empty(), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 486664e20..761bc3125 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -15578,7 +15578,7 @@ impl<'a> Parser<'a> { Ok(TableFunctionArgs { args, settings }) } - /// Parses a potentially empty list of arguments to a window function + /// Parses a potentially empty list of arguments to a function /// (including the closing parenthesis). /// /// Examples: @@ -15589,11 +15589,18 @@ impl<'a> Parser<'a> { fn parse_function_argument_list(&mut self) -> Result { let mut clauses = vec![]; - // For MSSQL empty argument list with json-null-clause case, e.g. `JSON_ARRAY(NULL ON NULL)` + // Handle clauses that may exist with an empty argument list + if let Some(null_clause) = self.parse_json_null_clause() { clauses.push(FunctionArgumentClause::JsonNullClause(null_clause)); } + if let Some(json_returning_clause) = self.maybe_parse_json_returning_clause()? { + clauses.push(FunctionArgumentClause::JsonReturningClause( + json_returning_clause, + )); + } + if self.consume_token(&Token::RParen) { return Ok(FunctionArgumentList { duplicate_treatment: None, @@ -15649,6 +15656,12 @@ impl<'a> Parser<'a> { clauses.push(FunctionArgumentClause::JsonNullClause(null_clause)); } + if let Some(json_returning_clause) = self.maybe_parse_json_returning_clause()? { + clauses.push(FunctionArgumentClause::JsonReturningClause( + json_returning_clause, + )); + } + self.expect_token(&Token::RParen)?; Ok(FunctionArgumentList { duplicate_treatment, @@ -15657,7 +15670,6 @@ impl<'a> Parser<'a> { }) } - /// Parses MSSQL's json-null-clause fn parse_json_null_clause(&mut self) -> Option { if self.parse_keywords(&[Keyword::ABSENT, Keyword::ON, Keyword::NULL]) { Some(JsonNullClause::AbsentOnNull) @@ -15668,6 +15680,17 @@ impl<'a> Parser<'a> { } } + fn maybe_parse_json_returning_clause( + &mut self, + ) -> Result, ParserError> { + if self.parse_keyword(Keyword::RETURNING) { + let data_type = self.parse_data_type()?; + Ok(Some(JsonReturningClause { data_type })) + } else { + Ok(None) + } + } + fn parse_duplicate_treatment(&mut self) -> Result, ParserError> { let loc = self.peek_token().span.start; match ( diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 43b30aade..287b828a8 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -3351,7 +3351,31 @@ fn test_json() { } #[test] -fn test_fn_arg_with_value_operator() { +fn json_object_colon_syntax() { + match pg().verified_expr("JSON_OBJECT('name' : 'value')") { + Expr::Function(Function { + args: FunctionArguments::List(FunctionArgumentList { args, .. }), + .. + }) => { + assert!( + matches!( + &args[..], + &[FunctionArg::ExprNamed { + operator: FunctionArgOperator::Colon, + .. + }] + ), + "Invalid function argument: {args:?}" + ); + } + other => panic!( + "Expected: JSON_OBJECT('name' : 'value') to be parsed as a function, but got {other:?}" + ), + } +} + +#[test] +fn json_object_value_syntax() { match pg().verified_expr("JSON_OBJECT('name' VALUE 'value')") { Expr::Function(Function { args: FunctionArguments::List(FunctionArgumentList { args, .. }), .. }) => { assert!(matches!( @@ -3363,6 +3387,63 @@ fn test_fn_arg_with_value_operator() { } } +#[test] +fn parse_json_object() { + let sql = "JSON_OBJECT('name' VALUE 'value' NULL ON NULL)"; + let expr = pg().verified_expr(sql); + assert!( + matches!( + expr.clone(), + Expr::Function(Function { + name: ObjectName(parts), + args: FunctionArguments::List(FunctionArgumentList { args, clauses, .. }), + .. + }) if parts == vec![ObjectNamePart::Identifier(Ident::new("JSON_OBJECT"))] + && matches!( + &args[..], + &[FunctionArg::ExprNamed { operator: FunctionArgOperator::Value, .. }] + ) + && clauses == vec![FunctionArgumentClause::JsonNullClause(JsonNullClause::NullOnNull)] + ), + "Failed to parse JSON_OBJECT with expected structure, got: {expr:?}" + ); + + let sql = "JSON_OBJECT('name' VALUE 'value' RETURNING JSONB)"; + let expr = pg().verified_expr(sql); + assert!( + matches!( + expr.clone(), + Expr::Function(Function { + name: ObjectName(parts), + args: FunctionArguments::List(FunctionArgumentList { args, clauses, .. }), + .. + }) if parts == vec![ObjectNamePart::Identifier(Ident::new("JSON_OBJECT"))] + && matches!( + &args[..], + &[FunctionArg::ExprNamed { operator: FunctionArgOperator::Value, .. }] + ) + && clauses == vec![FunctionArgumentClause::JsonReturningClause(JsonReturningClause { data_type: DataType::JSONB })] + ), + "Failed to parse JSON_OBJECT with expected structure, got: {expr:?}" + ); + + let sql = "JSON_OBJECT(RETURNING JSONB)"; + let expr = pg().verified_expr(sql); + assert!( + matches!( + expr.clone(), + Expr::Function(Function { + name: ObjectName(parts), + args: FunctionArguments::List(FunctionArgumentList { args, clauses, .. }), + .. + }) if parts == vec![ObjectNamePart::Identifier(Ident::new("JSON_OBJECT"))] + && args.is_empty() + && clauses == vec![FunctionArgumentClause::JsonReturningClause(JsonReturningClause { data_type: DataType::JSONB })] + ), + "Failed to parse JSON_OBJECT with expected structure, got: {expr:?}" + ); +} + #[test] fn parse_json_table_is_not_reserved() { // JSON_TABLE is not a reserved keyword in PostgreSQL, even though it is in SQL:2023 From bc478b0531c855cd6f53ff90c0aa964787090e69 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Tue, 26 Aug 2025 22:25:10 +0300 Subject: [PATCH 249/286] Snowflake: Minus char in stage name (#2014) --- src/dialect/snowflake.rs | 1 + tests/sqlparser_snowflake.rs | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 07ef8317a..3bb36010b 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -1065,6 +1065,7 @@ pub fn parse_stage_name_identifier(parser: &mut Parser) -> Result ident.push('%'), Token::Div => ident.push('/'), Token::Plus => ident.push('+'), + Token::Minus => ident.push('-'), Token::Number(n, _) => ident.push_str(n), Token::Word(w) => ident.push_str(&w.to_string()), _ => return parser.expected("stage name identifier", parser.peek_token()), diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 75e649fe6..7c9e5261b 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -2690,7 +2690,7 @@ fn test_snowflake_copy_into() { } // Test for non-ident characters in stage names - let sql = "COPY INTO a.b FROM @namespace.stage_name/x@x~x%x+/20250723_data"; + let sql = "COPY INTO a.b FROM @namespace.stage_name/x@x~x%x+/20250723_data-x"; assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { into, from_obj, .. } => { @@ -2702,7 +2702,7 @@ fn test_snowflake_copy_into() { from_obj, Some(ObjectName::from(vec![ Ident::new("@namespace"), - Ident::new("stage_name/x@x~x%x+/20250723_data") + Ident::new("stage_name/x@x~x%x+/20250723_data-x") ])) ) } From ee707c72a7a99cf84ea107f201ed7d58c4a75ed3 Mon Sep 17 00:00:00 2001 From: Simon Sawert Date: Thu, 28 Aug 2025 12:18:03 +0200 Subject: [PATCH 250/286] Support wildcard metrics for `SEMANTIC_VIEW` (#2016) --- src/ast/query.rs | 2 +- src/parser/mod.rs | 7 +++++-- tests/sqlparser_common.rs | 2 +- 3 files changed, 7 insertions(+), 4 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 967af85c2..ffc9ce666 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -1429,7 +1429,7 @@ pub enum TableFactor { /// List of dimensions or expression referring to dimensions (e.g. DATE_PART('year', col)) dimensions: Vec, /// List of metrics (references to objects like orders.value, value, orders.*) - metrics: Vec, + metrics: Vec, /// List of facts or expressions referring to facts or dimensions. facts: Vec, /// WHERE clause for filtering diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 761bc3125..a563e174a 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13958,7 +13958,7 @@ impl<'a> Parser<'a> { "METRICS clause can only be specified once".to_string(), )); } - metrics = self.parse_comma_separated(|parser| parser.parse_object_name(true))?; + metrics = self.parse_comma_separated(Parser::parse_wildcard_expr)?; } else if self.parse_keyword(Keyword::FACTS) { if !facts.is_empty() { return Err(ParserError::ParserError( @@ -13975,7 +13975,10 @@ impl<'a> Parser<'a> { where_clause = Some(self.parse_expr()?); } else { return parser_err!( - "Expected one of DIMENSIONS, METRICS, FACTS or WHERE", + format!( + "Expected one of DIMENSIONS, METRICS, FACTS or WHERE, got {}", + self.peek_token().token + ), self.peek_token().span.start )?; } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 2e2d9bfd0..516df5be2 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16951,6 +16951,7 @@ fn test_parse_semantic_view_table_factor() { "SELECT * FROM SEMANTIC_VIEW(model METRICS orders.col, orders.col2)", None, ), + ("SELECT * FROM SEMANTIC_VIEW(model METRICS orders.*)", None), // We can parse in any order but will always produce a result in a fixed order. ( "SELECT * FROM SEMANTIC_VIEW(model WHERE x > 0 DIMENSIONS dim1)", @@ -16980,7 +16981,6 @@ fn test_parse_semantic_view_table_factor() { let invalid_sqls = [ "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 INVALID inv1)", "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim1 DIMENSIONS dim2)", - "SELECT * FROM SEMANTIC_VIEW(model METRICS SUM(met1.avg))", ]; for sql in invalid_sqls { From 9b6f6de056f8b26df0fa781e8898be89bc47bcd9 Mon Sep 17 00:00:00 2001 From: Simon Sawert Date: Fri, 29 Aug 2025 09:28:51 +0200 Subject: [PATCH 251/286] Allow wilrdacrd for all `SEMANTIC_VIEW` types (#2017) --- src/parser/mod.rs | 4 ++-- tests/sqlparser_common.rs | 5 +++++ 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index a563e174a..f51ddc4dd 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13951,7 +13951,7 @@ impl<'a> Parser<'a> { "DIMENSIONS clause can only be specified once".to_string(), )); } - dimensions = self.parse_comma_separated(Parser::parse_expr)?; + dimensions = self.parse_comma_separated(Parser::parse_wildcard_expr)?; } else if self.parse_keyword(Keyword::METRICS) { if !metrics.is_empty() { return Err(ParserError::ParserError( @@ -13965,7 +13965,7 @@ impl<'a> Parser<'a> { "FACTS clause can only be specified once".to_string(), )); } - facts = self.parse_comma_separated(Parser::parse_expr)?; + facts = self.parse_comma_separated(Parser::parse_wildcard_expr)?; } else if self.parse_keyword(Keyword::WHERE) { if where_clause.is_some() { return Err(ParserError::ParserError( diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 516df5be2..1502ee22c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16952,6 +16952,11 @@ fn test_parse_semantic_view_table_factor() { None, ), ("SELECT * FROM SEMANTIC_VIEW(model METRICS orders.*)", None), + ("SELECT * FROM SEMANTIC_VIEW(model FACTS fact.*)", None), + ( + "SELECT * FROM SEMANTIC_VIEW(model DIMENSIONS dim.* METRICS orders.*)", + None, + ), // We can parse in any order but will always produce a result in a fixed order. ( "SELECT * FROM SEMANTIC_VIEW(model WHERE x > 0 DIMENSIONS dim1)", From eca65742bad548ccd3987cdc7b9d854b81702831 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Tue, 2 Sep 2025 12:21:04 +0200 Subject: [PATCH 252/286] Redshift: Add support for `UNLOAD` statement (#2013) --- src/ast/mod.rs | 174 ++++++++++++++++++++++++++++++++++++-- src/keywords.rs | 16 +++- src/parser/mod.rs | 129 ++++++++++++++++++++++++++-- tests/sqlparser_common.rs | 119 +++++++++++++++++++++++++- 4 files changed, 420 insertions(+), 18 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 615ceb1ec..5b8e23259 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4294,15 +4294,24 @@ pub enum Statement { /// ``` /// Note: this is a MySQL-specific statement. See UnlockTables, + /// Unloads the result of a query to file + /// + /// [Athena](https://docs.aws.amazon.com/athena/latest/ug/unload.html): /// ```sql /// UNLOAD(statement) TO [ WITH options ] /// ``` - /// See Redshift and - // Athena + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_UNLOAD.html): + /// ```sql + /// UNLOAD('statement') TO [ OPTIONS ] + /// ``` Unload { - query: Box, + query: Option>, + query_text: Option, to: Ident, + auth: Option, with: Vec, + options: Vec, }, /// ```sql /// OPTIMIZE TABLE [db.]name [ON CLUSTER cluster] [PARTITION partition | PARTITION ID 'partition_id'] [FINAL] [DEDUPLICATE [BY expression]] @@ -6282,13 +6291,31 @@ impl fmt::Display for Statement { Statement::UnlockTables => { write!(f, "UNLOCK TABLES") } - Statement::Unload { query, to, with } => { - write!(f, "UNLOAD({query}) TO {to}")?; - + Statement::Unload { + query, + query_text, + to, + auth, + with, + options, + } => { + write!(f, "UNLOAD(")?; + if let Some(query) = query { + write!(f, "{query}")?; + } + if let Some(query_text) = query_text { + write!(f, "'{query_text}'")?; + } + write!(f, ") TO {to}")?; + if let Some(auth) = auth { + write!(f, " IAM_ROLE {auth}")?; + } if !with.is_empty() { write!(f, " WITH ({})", display_comma_separated(with))?; } - + if !options.is_empty() { + write!(f, " {}", display_separated(options, " "))?; + } Ok(()) } Statement::OptimizeTable { @@ -8797,10 +8824,18 @@ pub enum CopyLegacyOption { AcceptAnyDate, /// ACCEPTINVCHARS AcceptInvChars(Option), + /// ADDQUOTES + AddQuotes, + /// ALLOWOVERWRITE + AllowOverwrite, /// BINARY Binary, /// BLANKSASNULL BlankAsNull, + /// BZIP2 + Bzip2, + /// CLEANPATH + CleanPath, /// CSV ... Csv(Vec), /// DATEFORMAT \[ AS \] {'dateformat_string' | 'auto' } @@ -8809,16 +8844,46 @@ pub enum CopyLegacyOption { Delimiter(char), /// EMPTYASNULL EmptyAsNull, + /// ENCRYPTED \[ AUTO \] + Encrypted { auto: bool }, + /// ESCAPE + Escape, + /// EXTENSION 'extension-name' + Extension(String), + /// FIXEDWIDTH \[ AS \] 'fixedwidth-spec' + FixedWidth(String), + /// GZIP + Gzip, + /// HEADER + Header, /// IAM_ROLE { DEFAULT | 'arn:aws:iam::123456789:role/role1' } IamRole(IamRoleKind), /// IGNOREHEADER \[ AS \] number_rows IgnoreHeader(u64), + /// JSON + Json, + /// MANIFEST \[ VERBOSE \] + Manifest { verbose: bool }, + /// MAXFILESIZE \[ AS \] max-size \[ MB | GB \] + MaxFileSize(FileSize), /// NULL \[ AS \] 'null_string' Null(String), + /// PARALLEL [ { ON | TRUE } | { OFF | FALSE } ] + Parallel(Option), + /// PARQUET + Parquet, + /// PARTITION BY ( column_name [, ... ] ) \[ INCLUDE \] + PartitionBy(UnloadPartitionBy), + /// REGION \[ AS \] 'aws-region' } + Region(String), + /// ROWGROUPSIZE \[ AS \] size \[ MB | GB \] + RowGroupSize(FileSize), /// TIMEFORMAT \[ AS \] {'timeformat_string' | 'auto' | 'epochsecs' | 'epochmillisecs' } TimeFormat(Option), /// TRUNCATECOLUMNS TruncateColumns, + /// ZSTD + Zstd, } impl fmt::Display for CopyLegacyOption { @@ -8833,8 +8898,12 @@ impl fmt::Display for CopyLegacyOption { } Ok(()) } + AddQuotes => write!(f, "ADDQUOTES"), + AllowOverwrite => write!(f, "ALLOWOVERWRITE"), Binary => write!(f, "BINARY"), BlankAsNull => write!(f, "BLANKSASNULL"), + Bzip2 => write!(f, "BZIP2"), + CleanPath => write!(f, "CLEANPATH"), Csv(opts) => { write!(f, "CSV")?; if !opts.is_empty() { @@ -8851,9 +8920,37 @@ impl fmt::Display for CopyLegacyOption { } Delimiter(char) => write!(f, "DELIMITER '{char}'"), EmptyAsNull => write!(f, "EMPTYASNULL"), + Encrypted { auto } => write!(f, "ENCRYPTED{}", if *auto { " AUTO" } else { "" }), + Escape => write!(f, "ESCAPE"), + Extension(ext) => write!(f, "EXTENSION '{}'", value::escape_single_quote_string(ext)), + FixedWidth(spec) => write!( + f, + "FIXEDWIDTH '{}'", + value::escape_single_quote_string(spec) + ), + Gzip => write!(f, "GZIP"), + Header => write!(f, "HEADER"), IamRole(role) => write!(f, "IAM_ROLE {role}"), IgnoreHeader(num_rows) => write!(f, "IGNOREHEADER {num_rows}"), + Json => write!(f, "JSON"), + Manifest { verbose } => write!(f, "MANIFEST{}", if *verbose { " VERBOSE" } else { "" }), + MaxFileSize(file_size) => write!(f, "MAXFILESIZE {file_size}"), Null(string) => write!(f, "NULL '{}'", value::escape_single_quote_string(string)), + Parallel(enabled) => { + write!( + f, + "PARALLEL{}", + match enabled { + Some(true) => " TRUE", + Some(false) => " FALSE", + _ => "", + } + ) + } + Parquet => write!(f, "PARQUET"), + PartitionBy(p) => write!(f, "{p}"), + Region(region) => write!(f, "REGION '{}'", value::escape_single_quote_string(region)), + RowGroupSize(file_size) => write!(f, "ROWGROUPSIZE {file_size}"), TimeFormat(fmt) => { write!(f, "TIMEFORMAT")?; if let Some(fmt) = fmt { @@ -8862,10 +8959,73 @@ impl fmt::Display for CopyLegacyOption { Ok(()) } TruncateColumns => write!(f, "TRUNCATECOLUMNS"), + Zstd => write!(f, "ZSTD"), + } + } +} + +/// ```sql +/// SIZE \[ MB | GB \] +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct FileSize { + pub size: Value, + pub unit: Option, +} + +impl fmt::Display for FileSize { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}", self.size)?; + if let Some(unit) = &self.unit { + write!(f, " {unit}")?; + } + Ok(()) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum FileSizeUnit { + MB, + GB, +} + +impl fmt::Display for FileSizeUnit { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + FileSizeUnit::MB => write!(f, "MB"), + FileSizeUnit::GB => write!(f, "GB"), } } } +/// Specifies the partition keys for the unload operation +/// +/// ```sql +/// PARTITION BY ( column_name [, ... ] ) [ INCLUDE ] +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct UnloadPartitionBy { + pub columns: Vec, + pub include: bool, +} + +impl fmt::Display for UnloadPartitionBy { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "PARTITION BY ({}){}", + display_comma_separated(&self.columns), + if self.include { " INCLUDE" } else { "" } + ) + } +} + /// An `IAM_ROLE` option in the AWS ecosystem /// /// [Redshift COPY](https://docs.aws.amazon.com/redshift/latest/dg/copy-parameters-authorization.html#copy-iam-role) diff --git a/src/keywords.rs b/src/keywords.rs index eb27d1c4e..e955cd97b 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -82,6 +82,7 @@ define_keywords!( ACCOUNT, ACTION, ADD, + ADDQUOTES, ADMIN, AFTER, AGAINST, @@ -92,6 +93,7 @@ define_keywords!( ALIAS, ALL, ALLOCATE, + ALLOWOVERWRITE, ALTER, ALWAYS, ANALYZE, @@ -159,6 +161,7 @@ define_keywords!( BYPASSRLS, BYTEA, BYTES, + BZIP2, CACHE, CALL, CALLED, @@ -190,6 +193,7 @@ define_keywords!( CHECK, CHECKSUM, CIRCLE, + CLEANPATH, CLEAR, CLOB, CLONE, @@ -322,6 +326,7 @@ define_keywords!( ENABLE, ENABLE_SCHEMA_EVOLUTION, ENCODING, + ENCRYPTED, ENCRYPTION, END, END_EXEC = "END-EXEC", @@ -380,6 +385,7 @@ define_keywords!( FIRST, FIRST_VALUE, FIXEDSTRING, + FIXEDWIDTH, FLATTEN, FLOAT, FLOAT32, @@ -411,6 +417,7 @@ define_keywords!( FUNCTIONS, FUSION, FUTURE, + GB, GENERAL, GENERATE, GENERATED, @@ -426,6 +433,7 @@ define_keywords!( GROUP, GROUPING, GROUPS, + GZIP, HASH, HAVING, HEADER, @@ -550,6 +558,7 @@ define_keywords!( MANAGE, MANAGED, MANAGEDLOCATION, + MANIFEST, MAP, MASKING, MATCH, @@ -560,9 +569,11 @@ define_keywords!( MATERIALIZE, MATERIALIZED, MAX, + MAXFILESIZE, MAXVALUE, MAX_DATA_EXTENSION_TIME_IN_DAYS, MAX_ROWS, + MB, MEASURES, MEDIUMBLOB, MEDIUMINT, @@ -761,6 +772,7 @@ define_keywords!( REFRESH_MODE, REGCLASS, REGEXP, + REGION, REGR_AVGX, REGR_AVGY, REGR_COUNT, @@ -813,6 +825,7 @@ define_keywords!( ROLLUP, ROOT, ROW, + ROWGROUPSIZE, ROWID, ROWS, ROW_FORMAT, @@ -1062,7 +1075,8 @@ define_keywords!( YEAR, YEARS, ZONE, - ZORDER + ZORDER, + ZSTD ); /// These keywords can't be used as a table alias, so that `FROM table_name alias` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f51ddc4dd..0afab28e6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -630,7 +630,10 @@ impl<'a> Parser<'a> { Keyword::NOTIFY if self.dialect.supports_listen_notify() => self.parse_notify(), // `PRAGMA` is sqlite specific https://www.sqlite.org/pragma.html Keyword::PRAGMA => self.parse_pragma(), - Keyword::UNLOAD => self.parse_unload(), + Keyword::UNLOAD => { + self.prev_token(); + self.parse_unload() + } Keyword::RENAME => self.parse_rename(), // `INSTALL` is duckdb specific https://duckdb.org/docs/extensions/overview Keyword::INSTALL if dialect_of!(self is DuckDbDialect | GenericDialect) => { @@ -9615,17 +9618,36 @@ impl<'a> Parser<'a> { let ret = match self.parse_one_of_keywords(&[ Keyword::ACCEPTANYDATE, Keyword::ACCEPTINVCHARS, + Keyword::ADDQUOTES, + Keyword::ALLOWOVERWRITE, Keyword::BINARY, Keyword::BLANKSASNULL, + Keyword::BZIP2, + Keyword::CLEANPATH, Keyword::CSV, Keyword::DATEFORMAT, Keyword::DELIMITER, Keyword::EMPTYASNULL, + Keyword::ENCRYPTED, + Keyword::ESCAPE, + Keyword::EXTENSION, + Keyword::FIXEDWIDTH, + Keyword::GZIP, + Keyword::HEADER, Keyword::IAM_ROLE, Keyword::IGNOREHEADER, + Keyword::JSON, + Keyword::MANIFEST, + Keyword::MAXFILESIZE, Keyword::NULL, + Keyword::PARALLEL, + Keyword::PARQUET, + Keyword::PARTITION, + Keyword::REGION, + Keyword::ROWGROUPSIZE, Keyword::TIMEFORMAT, Keyword::TRUNCATECOLUMNS, + Keyword::ZSTD, ]) { Some(Keyword::ACCEPTANYDATE) => CopyLegacyOption::AcceptAnyDate, Some(Keyword::ACCEPTINVCHARS) => { @@ -9637,8 +9659,12 @@ impl<'a> Parser<'a> { }; CopyLegacyOption::AcceptInvChars(ch) } + Some(Keyword::ADDQUOTES) => CopyLegacyOption::AddQuotes, + Some(Keyword::ALLOWOVERWRITE) => CopyLegacyOption::AllowOverwrite, Some(Keyword::BINARY) => CopyLegacyOption::Binary, Some(Keyword::BLANKSASNULL) => CopyLegacyOption::BlankAsNull, + Some(Keyword::BZIP2) => CopyLegacyOption::Bzip2, + Some(Keyword::CLEANPATH) => CopyLegacyOption::CleanPath, Some(Keyword::CSV) => CopyLegacyOption::Csv({ let mut opts = vec![]; while let Some(opt) = @@ -9662,16 +9688,76 @@ impl<'a> Parser<'a> { CopyLegacyOption::Delimiter(self.parse_literal_char()?) } Some(Keyword::EMPTYASNULL) => CopyLegacyOption::EmptyAsNull, + Some(Keyword::ENCRYPTED) => { + let auto = self.parse_keyword(Keyword::AUTO); + CopyLegacyOption::Encrypted { auto } + } + Some(Keyword::ESCAPE) => CopyLegacyOption::Escape, + Some(Keyword::EXTENSION) => { + let ext = self.parse_literal_string()?; + CopyLegacyOption::Extension(ext) + } + Some(Keyword::FIXEDWIDTH) => { + let spec = self.parse_literal_string()?; + CopyLegacyOption::FixedWidth(spec) + } + Some(Keyword::GZIP) => CopyLegacyOption::Gzip, + Some(Keyword::HEADER) => CopyLegacyOption::Header, Some(Keyword::IAM_ROLE) => CopyLegacyOption::IamRole(self.parse_iam_role_kind()?), Some(Keyword::IGNOREHEADER) => { let _ = self.parse_keyword(Keyword::AS); let num_rows = self.parse_literal_uint()?; CopyLegacyOption::IgnoreHeader(num_rows) } + Some(Keyword::JSON) => CopyLegacyOption::Json, + Some(Keyword::MANIFEST) => { + let verbose = self.parse_keyword(Keyword::VERBOSE); + CopyLegacyOption::Manifest { verbose } + } + Some(Keyword::MAXFILESIZE) => { + let _ = self.parse_keyword(Keyword::AS); + let size = self.parse_number_value()?.value; + let unit = match self.parse_one_of_keywords(&[Keyword::MB, Keyword::GB]) { + Some(Keyword::MB) => Some(FileSizeUnit::MB), + Some(Keyword::GB) => Some(FileSizeUnit::GB), + _ => None, + }; + CopyLegacyOption::MaxFileSize(FileSize { size, unit }) + } Some(Keyword::NULL) => { let _ = self.parse_keyword(Keyword::AS); CopyLegacyOption::Null(self.parse_literal_string()?) } + Some(Keyword::PARALLEL) => { + let enabled = match self.parse_one_of_keywords(&[ + Keyword::TRUE, + Keyword::FALSE, + Keyword::ON, + Keyword::OFF, + ]) { + Some(Keyword::TRUE) | Some(Keyword::ON) => Some(true), + Some(Keyword::FALSE) | Some(Keyword::OFF) => Some(false), + _ => None, + }; + CopyLegacyOption::Parallel(enabled) + } + Some(Keyword::PARQUET) => CopyLegacyOption::Parquet, + Some(Keyword::PARTITION) => { + self.expect_keyword(Keyword::BY)?; + let columns = self.parse_parenthesized_column_list(IsOptional::Mandatory, false)?; + let include = self.parse_keyword(Keyword::INCLUDE); + CopyLegacyOption::PartitionBy(UnloadPartitionBy { columns, include }) + } + Some(Keyword::REGION) => { + let _ = self.parse_keyword(Keyword::AS); + let region = self.parse_literal_string()?; + CopyLegacyOption::Region(region) + } + Some(Keyword::ROWGROUPSIZE) => { + let _ = self.parse_keyword(Keyword::AS); + let file_size = self.parse_file_size()?; + CopyLegacyOption::RowGroupSize(file_size) + } Some(Keyword::TIMEFORMAT) => { let _ = self.parse_keyword(Keyword::AS); let fmt = if matches!(self.peek_token().token, Token::SingleQuotedString(_)) { @@ -9682,11 +9768,26 @@ impl<'a> Parser<'a> { CopyLegacyOption::TimeFormat(fmt) } Some(Keyword::TRUNCATECOLUMNS) => CopyLegacyOption::TruncateColumns, + Some(Keyword::ZSTD) => CopyLegacyOption::Zstd, _ => self.expected("option", self.peek_token())?, }; Ok(ret) } + fn parse_file_size(&mut self) -> Result { + let size = self.parse_number_value()?.value; + let unit = self.maybe_parse_file_size_unit(); + Ok(FileSize { size, unit }) + } + + fn maybe_parse_file_size_unit(&mut self) -> Option { + match self.parse_one_of_keywords(&[Keyword::MB, Keyword::GB]) { + Some(Keyword::MB) => Some(FileSizeUnit::MB), + Some(Keyword::GB) => Some(FileSizeUnit::GB), + _ => None, + } + } + fn parse_iam_role_kind(&mut self) -> Result { if self.parse_keyword(Keyword::DEFAULT) { Ok(IamRoleKind::Default) @@ -16508,19 +16609,35 @@ impl<'a> Parser<'a> { } pub fn parse_unload(&mut self) -> Result { + self.expect_keyword(Keyword::UNLOAD)?; self.expect_token(&Token::LParen)?; - let query = self.parse_query()?; + let (query, query_text) = if matches!(self.peek_token().token, Token::SingleQuotedString(_)) + { + (None, Some(self.parse_literal_string()?)) + } else { + (Some(self.parse_query()?), None) + }; self.expect_token(&Token::RParen)?; self.expect_keyword_is(Keyword::TO)?; let to = self.parse_identifier()?; - - let with_options = self.parse_options(Keyword::WITH)?; - + let auth = if self.parse_keyword(Keyword::IAM_ROLE) { + Some(self.parse_iam_role_kind()?) + } else { + None + }; + let with = self.parse_options(Keyword::WITH)?; + let mut options = vec![]; + while let Some(opt) = self.maybe_parse(|parser| parser.parse_copy_legacy_option())? { + options.push(opt); + } Ok(Statement::Unload { query, + query_text, to, - with: with_options, + auth, + with, + options, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 1502ee22c..0ed61c991 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -11894,7 +11894,7 @@ fn parse_unload() { assert_eq!( unload, Statement::Unload { - query: Box::new(Query { + query: Some(Box::new(Query { body: Box::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), distinct: None, @@ -11931,7 +11931,7 @@ fn parse_unload() { settings: None, format_clause: None, pipe_operators: vec![], - }), + })), to: Ident { value: "s3://...".to_string(), quote_style: Some('\''), @@ -11946,9 +11946,120 @@ fn parse_unload() { value: Expr::Value( (Value::SingleQuotedString("AVRO".to_string())).with_empty_span() ) - }] + }], + query_text: None, + auth: None, + options: vec![], } ); + + one_statement_parses_to( + concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "FORMAT AS CSV ", + "FORMAT AS PARQUET ", + "FORMAT AS JSON ", + "MAXFILESIZE AS 10 MB ", + "ROWGROUPSIZE AS 10 MB ", + "PARALLEL ON ", + "PARALLEL OFF ", + "REGION AS 'us-east-1'" + ), + concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "CSV ", + "PARQUET ", + "JSON ", + "MAXFILESIZE 10 MB ", + "ROWGROUPSIZE 10 MB ", + "PARALLEL TRUE ", + "PARALLEL FALSE ", + "REGION 'us-east-1'" + ), + ); + + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3)", + )); + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3) INCLUDE", + )); + + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3) INCLUDE ", + "MANIFEST" + )); + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3) INCLUDE ", + "MANIFEST VERBOSE" + )); + + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3) INCLUDE ", + "MANIFEST VERBOSE ", + "HEADER ", + "FIXEDWIDTH 'col1:1,col2:2' ", + "ENCRYPTED" + )); + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3) INCLUDE ", + "MANIFEST VERBOSE ", + "HEADER ", + "FIXEDWIDTH 'col1:1,col2:2' ", + "ENCRYPTED AUTO" + )); + + verified_stmt(concat!( + "UNLOAD('SELECT 1') ", + "TO 's3://...' ", + "IAM_ROLE 'arn:aws:iam::123456789:role/role1' ", + "PARTITION BY (c1, c2, c3) INCLUDE ", + "MANIFEST VERBOSE ", + "HEADER ", + "FIXEDWIDTH 'col1:1,col2:2' ", + "ENCRYPTED AUTO ", + "BZIP2 ", + "GZIP ", + "ZSTD ", + "ADDQUOTES ", + "NULL 'nil' ", + "ESCAPE ", + "ALLOWOVERWRITE ", + "CLEANPATH ", + "PARALLEL ", + "PARALLEL TRUE ", + "PARALLEL FALSE ", + "MAXFILESIZE 10 ", + "MAXFILESIZE 10 MB ", + "MAXFILESIZE 10 GB ", + "ROWGROUPSIZE 10 ", + "ROWGROUPSIZE 10 MB ", + "ROWGROUPSIZE 10 GB ", + "REGION 'us-east-1' ", + "EXTENSION 'ext1'" + )); } #[test] @@ -16990,7 +17101,7 @@ fn test_parse_semantic_view_table_factor() { for sql in invalid_sqls { let result = dialects.parse_sql_statements(sql); - assert!(result.is_err(), "Expected error for invalid SQL: {}", sql); + assert!(result.is_err(), "Expected error for invalid SQL: {sql}"); } let ast_sql = r#"SELECT * FROM SEMANTIC_VIEW( From cff28334be9de19ddfa04b772301136877779bd9 Mon Sep 17 00:00:00 2001 From: etgarperets Date: Tue, 2 Sep 2025 14:20:34 +0300 Subject: [PATCH 253/286] Add support for string literal concatenation (#2003) Co-authored-by: Ifeanyi Ubah --- src/dialect/mod.rs | 6 ++++++ src/dialect/mysql.rs | 5 +++++ src/parser/mod.rs | 20 ++++++++++++++++++-- tests/sqlparser_common.rs | 10 ++++++++++ 4 files changed, 39 insertions(+), 2 deletions(-) diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index f91209722..3c7be8f7e 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -476,6 +476,12 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports concatenating of string literal + /// Example: `SELECT 'Hello ' "world" => SELECT 'Hello world'` + fn supports_string_literal_concatenation(&self) -> bool { + false + } + /// Does the dialect support trailing commas in the projection list? fn supports_projection_trailing_commas(&self) -> bool { self.supports_trailing_commas() diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 6cf24e14e..8c63bfda5 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -71,6 +71,11 @@ impl Dialect for MySqlDialect { true } + /// see + fn supports_string_literal_concatenation(&self) -> bool { + true + } + fn ignores_wildcard_escapes(&self) -> bool { true } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 0afab28e6..54db367d8 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9914,8 +9914,12 @@ impl<'a> Parser<'a> { // bigdecimal feature is enabled, and is otherwise a no-op // (i.e., it returns the input string). Token::Number(n, l) => ok_value(Value::Number(Self::parse(n, span.start)?, l)), - Token::SingleQuotedString(ref s) => ok_value(Value::SingleQuotedString(s.to_string())), - Token::DoubleQuotedString(ref s) => ok_value(Value::DoubleQuotedString(s.to_string())), + Token::SingleQuotedString(ref s) => ok_value(Value::SingleQuotedString( + self.maybe_concat_string_literal(s.to_string()), + )), + Token::DoubleQuotedString(ref s) => ok_value(Value::DoubleQuotedString( + self.maybe_concat_string_literal(s.to_string()), + )), Token::TripleSingleQuotedString(ref s) => { ok_value(Value::TripleSingleQuotedString(s.to_string())) } @@ -9985,6 +9989,18 @@ impl<'a> Parser<'a> { } } + fn maybe_concat_string_literal(&mut self, mut str: String) -> String { + if self.dialect.supports_string_literal_concatenation() { + while let Token::SingleQuotedString(ref s) | Token::DoubleQuotedString(ref s) = + self.peek_token_ref().token + { + str.push_str(s.clone().as_str()); + self.advance_token(); + } + } + str + } + /// Parse an unsigned numeric literal pub fn parse_number_value(&mut self) -> Result { let value_wrapper = self.parse_value()?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 0ed61c991..aa47c0f7a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -17150,3 +17150,13 @@ fn test_parse_semantic_view_table_factor() { _ => panic!("Expected Query statement"), } } + +#[test] +fn parse_adjacent_string_literal_concatenation() { + let sql = r#"SELECT 'M' "y" 'S' "q" 'l'"#; + let dialects = all_dialects_where(|d| d.supports_string_literal_concatenation()); + dialects.one_statement_parses_to(sql, r"SELECT 'MySql'"); + + let sql = "SELECT * FROM t WHERE col = 'Hello' \n ' ' \t 'World!'"; + dialects.one_statement_parses_to(sql, r"SELECT * FROM t WHERE col = 'Hello World!'"); +} From dd3342e4176c4e6ffceb70a2e7289bed7b21a909 Mon Sep 17 00:00:00 2001 From: Dmitrii Blaginin Date: Thu, 4 Sep 2025 21:21:09 +0100 Subject: [PATCH 254/286] Merge queue Infrastructure in sqlparser-rs (#2007) --- .asf.yaml | 14 ++++++++++++++ .github/workflows/license.yml | 1 + .github/workflows/rust.yml | 11 +++++++++-- src/keywords.rs | 2 +- 4 files changed, 25 insertions(+), 3 deletions(-) diff --git a/.asf.yaml b/.asf.yaml index 8ad476f12..0534705bc 100644 --- a/.asf.yaml +++ b/.asf.yaml @@ -36,3 +36,17 @@ github: rebase: false features: issues: true + protected_branches: + main: + required_status_checks: + contexts: + - "codestyle" + - "lint" + - "benchmark-lint" + - "compile" + - "docs" + - "compile-no-std" + - "test (stable)" + - "test (beta)" + - "test (nightly)" + - "Release Audit Tool (RAT)" diff --git a/.github/workflows/license.yml b/.github/workflows/license.yml index c851bff31..f4524f6b2 100644 --- a/.github/workflows/license.yml +++ b/.github/workflows/license.yml @@ -23,6 +23,7 @@ on: branches: - main pull_request: + merge_group: jobs: diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml index 3abf9d387..f96de9472 100644 --- a/.github/workflows/rust.yml +++ b/.github/workflows/rust.yml @@ -17,13 +17,20 @@ name: Rust -on: [push, pull_request] +on: + push: + # When PR is in the Merge Queue, GitHub will create a temporary branch - but we already have a + # CI running because of `merge_group` + # https://github.com/orgs/community/discussions/15254 + branches-ignore: + - 'gh-readonly-queue/**' + pull_request: + merge_group: permissions: contents: read jobs: - codestyle: runs-on: ubuntu-latest steps: diff --git a/src/keywords.rs b/src/keywords.rs index e955cd97b..3358e2845 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -15,7 +15,7 @@ // specific language governing permissions and limitations // under the License. -//! This module defines +//! This module defines: //! 1) a list of constants for every keyword //! 2) an `ALL_KEYWORDS` array with every keyword in it //! This is not a list of *reserved* keywords: some of these can be From befc95e8e8c98bf35f4d19e648cc01dc8af68672 Mon Sep 17 00:00:00 2001 From: Dmitrii Blaginin Date: Fri, 5 Sep 2025 08:49:05 +0100 Subject: [PATCH 255/286] MQ Test (#2019) --- .github/workflows/rust.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/rust.yml b/.github/workflows/rust.yml index f96de9472..2e1da6941 100644 --- a/.github/workflows/rust.yml +++ b/.github/workflows/rust.yml @@ -21,7 +21,7 @@ on: push: # When PR is in the Merge Queue, GitHub will create a temporary branch - but we already have a # CI running because of `merge_group` - # https://github.com/orgs/community/discussions/15254 + # See also: https://github.com/orgs/community/discussions/15254 branches-ignore: - 'gh-readonly-queue/**' pull_request: From b8539a52af7419b86d9c2fc51ba247d022f7a6f8 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Tue, 9 Sep 2025 07:39:56 +0200 Subject: [PATCH 256/286] Added derive trait `Copy` to `OrderByOptions` struct (#2021) --- src/ast/query.rs | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index ffc9ce666..151b94e82 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2569,7 +2569,7 @@ impl fmt::Display for InterpolateExpr { } } -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct OrderByOptions { From 280f51889f583a239ea2d65cfb86203577392396 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Mon, 15 Sep 2025 09:26:46 +0200 Subject: [PATCH 257/286] Moved `CreateTrigger` and `DropTrigger` out of `Statement` enum (#2026) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 237 ++++++++++++++++++++++++++++++++++-- src/ast/mod.rs | 213 ++------------------------------ src/dialect/mssql.rs | 8 +- src/parser/mod.rs | 8 +- tests/sqlparser_mssql.rs | 8 +- tests/sqlparser_mysql.rs | 8 +- tests/sqlparser_postgres.rs | 32 ++--- 7 files changed, 270 insertions(+), 244 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index db1cb61f6..a15474755 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -20,7 +20,7 @@ #[cfg(not(feature = "std"))] use alloc::{boxed::Box, string::String, vec::Vec}; -use core::fmt::{self, Write}; +use core::fmt::{self, Display, Write}; #[cfg(feature = "serde")] use serde::{Deserialize, Serialize}; @@ -30,14 +30,15 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, - CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, DataType, Expr, FileFormat, - FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, - HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, InitializeKind, - MySQLColumnPosition, ObjectName, OnCommit, OneOrManyWithParens, OperateFunctionArg, - OrderByExpr, ProjectionSelect, Query, RefreshModeKind, RowAccessPolicy, SequenceOptions, - Spanned, SqlOption, StorageSerializationPolicy, TableVersion, Tag, Value, ValueWithSpan, - WrappedCollection, + display_comma_separated, display_separated, ArgMode, CommentDef, ConditionalStatements, + CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, DataType, + Expr, FileFormat, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, + FunctionParallel, HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, + InitializeKind, MySQLColumnPosition, ObjectName, OnCommit, OneOrManyWithParens, + OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RefreshModeKind, RowAccessPolicy, + SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, TableVersion, Tag, + TriggerEvent, TriggerExecBody, TriggerObject, TriggerPeriod, TriggerReferencing, Value, + ValueWithSpan, WrappedCollection, }; use crate::display_utils::{DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; use crate::keywords::Keyword; @@ -3176,3 +3177,221 @@ impl Spanned for RenameTableNameKind { } } } + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// CREATE TRIGGER +/// +/// Examples: +/// +/// ```sql +/// CREATE TRIGGER trigger_name +/// BEFORE INSERT ON table_name +/// FOR EACH ROW +/// EXECUTE FUNCTION trigger_function(); +/// ``` +/// +/// Postgres: +/// SQL Server: +pub struct CreateTrigger { + /// True if this is a `CREATE OR ALTER TRIGGER` statement + /// + /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-trigger-transact-sql?view=sql-server-ver16#arguments) + pub or_alter: bool, + /// The `OR REPLACE` clause is used to re-create the trigger if it already exists. + /// + /// Example: + /// ```sql + /// CREATE OR REPLACE TRIGGER trigger_name + /// AFTER INSERT ON table_name + /// FOR EACH ROW + /// EXECUTE FUNCTION trigger_function(); + /// ``` + pub or_replace: bool, + /// The `CONSTRAINT` keyword is used to create a trigger as a constraint. + pub is_constraint: bool, + /// The name of the trigger to be created. + pub name: ObjectName, + /// Determines whether the function is called before, after, or instead of the event. + /// + /// Example of BEFORE: + /// + /// ```sql + /// CREATE TRIGGER trigger_name + /// BEFORE INSERT ON table_name + /// FOR EACH ROW + /// EXECUTE FUNCTION trigger_function(); + /// ``` + /// + /// Example of AFTER: + /// + /// ```sql + /// CREATE TRIGGER trigger_name + /// AFTER INSERT ON table_name + /// FOR EACH ROW + /// EXECUTE FUNCTION trigger_function(); + /// ``` + /// + /// Example of INSTEAD OF: + /// + /// ```sql + /// CREATE TRIGGER trigger_name + /// INSTEAD OF INSERT ON table_name + /// FOR EACH ROW + /// EXECUTE FUNCTION trigger_function(); + /// ``` + pub period: TriggerPeriod, + /// Whether the trigger period was specified before the target table name. + /// + /// ```sql + /// -- period_before_table == true: Postgres, MySQL, and standard SQL + /// CREATE TRIGGER t BEFORE INSERT ON table_name ...; + /// -- period_before_table == false: MSSQL + /// CREATE TRIGGER t ON table_name BEFORE INSERT ...; + /// ``` + pub period_before_table: bool, + /// Multiple events can be specified using OR, such as `INSERT`, `UPDATE`, `DELETE`, or `TRUNCATE`. + pub events: Vec, + /// The table on which the trigger is to be created. + pub table_name: ObjectName, + /// The optional referenced table name that can be referenced via + /// the `FROM` keyword. + pub referenced_table_name: Option, + /// This keyword immediately precedes the declaration of one or two relation names that provide access to the transition relations of the triggering statement. + pub referencing: Vec, + /// This specifies whether the trigger function should be fired once for + /// every row affected by the trigger event, or just once per SQL statement. + pub trigger_object: TriggerObject, + /// Whether to include the `EACH` term of the `FOR EACH`, as it is optional syntax. + pub include_each: bool, + /// Triggering conditions + pub condition: Option, + /// Execute logic block + pub exec_body: Option, + /// For MSSQL and dialects where statements are preceded by `AS` + pub statements_as: bool, + /// For SQL dialects with statement(s) for a body + pub statements: Option, + /// The characteristic of the trigger, which include whether the trigger is `DEFERRABLE`, `INITIALLY DEFERRED`, or `INITIALLY IMMEDIATE`, + pub characteristics: Option, +} + +impl Display for CreateTrigger { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + let CreateTrigger { + or_alter, + or_replace, + is_constraint, + name, + period_before_table, + period, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + include_each, + exec_body, + statements_as, + statements, + characteristics, + } = self; + write!( + f, + "CREATE {or_alter}{or_replace}{is_constraint}TRIGGER {name} ", + or_alter = if *or_alter { "OR ALTER " } else { "" }, + or_replace = if *or_replace { "OR REPLACE " } else { "" }, + is_constraint = if *is_constraint { "CONSTRAINT " } else { "" }, + )?; + + if *period_before_table { + write!(f, "{period}")?; + if !events.is_empty() { + write!(f, " {}", display_separated(events, " OR "))?; + } + write!(f, " ON {table_name}")?; + } else { + write!(f, "ON {table_name}")?; + write!(f, " {period}")?; + if !events.is_empty() { + write!(f, " {}", display_separated(events, ", "))?; + } + } + + if let Some(referenced_table_name) = referenced_table_name { + write!(f, " FROM {referenced_table_name}")?; + } + + if let Some(characteristics) = characteristics { + write!(f, " {characteristics}")?; + } + + if !referencing.is_empty() { + write!(f, " REFERENCING {}", display_separated(referencing, " "))?; + } + + if *include_each { + write!(f, " FOR EACH {trigger_object}")?; + } else if exec_body.is_some() { + write!(f, " FOR {trigger_object}")?; + } + if let Some(condition) = condition { + write!(f, " WHEN {condition}")?; + } + if let Some(exec_body) = exec_body { + write!(f, " EXECUTE {exec_body}")?; + } + if let Some(statements) = statements { + if *statements_as { + write!(f, " AS")?; + } + write!(f, " {statements}")?; + } + Ok(()) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// DROP TRIGGER +/// +/// ```sql +/// DROP TRIGGER [ IF EXISTS ] name ON table_name [ CASCADE | RESTRICT ] +/// ``` +/// +pub struct DropTrigger { + /// Whether to include the `IF EXISTS` clause. + pub if_exists: bool, + /// The name of the trigger to be dropped. + pub trigger_name: ObjectName, + /// The name of the table from which the trigger is to be dropped. + pub table_name: Option, + /// `CASCADE` or `RESTRICT` + pub option: Option, +} + +impl fmt::Display for DropTrigger { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let DropTrigger { + if_exists, + trigger_name, + table_name, + option, + } = self; + write!(f, "DROP TRIGGER")?; + if *if_exists { + write!(f, " IF EXISTS")?; + } + match &table_name { + Some(table_name) => write!(f, " {trigger_name} ON {table_name}")?, + None => write!(f, " {trigger_name}")?, + }; + if let Some(option) = option { + write!(f, " {option}")?; + } + Ok(()) + } +} diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 5b8e23259..6efe1f748 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -63,10 +63,10 @@ pub use self::ddl::{ AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, - CreateFunction, CreateIndex, CreateTable, Deduplicate, DeferrableInitial, DropBehavior, - GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, - IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, - IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, + CreateFunction, CreateIndex, CreateTable, CreateTrigger, Deduplicate, DeferrableInitial, + DropBehavior, DropTrigger, GeneratedAs, GeneratedExpressionMode, IdentityParameters, + IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, + IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, @@ -3914,114 +3914,10 @@ pub enum Statement { /// 3. [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#create_function_statement) /// 4. [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-function-transact-sql) CreateFunction(CreateFunction), - /// CREATE TRIGGER - /// - /// Examples: - /// - /// ```sql - /// CREATE TRIGGER trigger_name - /// BEFORE INSERT ON table_name - /// FOR EACH ROW - /// EXECUTE FUNCTION trigger_function(); - /// ``` - /// - /// Postgres: - /// SQL Server: - CreateTrigger { - /// True if this is a `CREATE OR ALTER TRIGGER` statement - /// - /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-trigger-transact-sql?view=sql-server-ver16#arguments) - or_alter: bool, - /// The `OR REPLACE` clause is used to re-create the trigger if it already exists. - /// - /// Example: - /// ```sql - /// CREATE OR REPLACE TRIGGER trigger_name - /// AFTER INSERT ON table_name - /// FOR EACH ROW - /// EXECUTE FUNCTION trigger_function(); - /// ``` - or_replace: bool, - /// The `CONSTRAINT` keyword is used to create a trigger as a constraint. - is_constraint: bool, - /// The name of the trigger to be created. - name: ObjectName, - /// Determines whether the function is called before, after, or instead of the event. - /// - /// Example of BEFORE: - /// - /// ```sql - /// CREATE TRIGGER trigger_name - /// BEFORE INSERT ON table_name - /// FOR EACH ROW - /// EXECUTE FUNCTION trigger_function(); - /// ``` - /// - /// Example of AFTER: - /// - /// ```sql - /// CREATE TRIGGER trigger_name - /// AFTER INSERT ON table_name - /// FOR EACH ROW - /// EXECUTE FUNCTION trigger_function(); - /// ``` - /// - /// Example of INSTEAD OF: - /// - /// ```sql - /// CREATE TRIGGER trigger_name - /// INSTEAD OF INSERT ON table_name - /// FOR EACH ROW - /// EXECUTE FUNCTION trigger_function(); - /// ``` - period: TriggerPeriod, - /// Whether the trigger period was specified before the target table name. - /// - /// ```sql - /// -- period_before_table == true: Postgres, MySQL, and standard SQL - /// CREATE TRIGGER t BEFORE INSERT ON table_name ...; - /// -- period_before_table == false: MSSQL - /// CREATE TRIGGER t ON table_name BEFORE INSERT ...; - /// ``` - period_before_table: bool, - /// Multiple events can be specified using OR, such as `INSERT`, `UPDATE`, `DELETE`, or `TRUNCATE`. - events: Vec, - /// The table on which the trigger is to be created. - table_name: ObjectName, - /// The optional referenced table name that can be referenced via - /// the `FROM` keyword. - referenced_table_name: Option, - /// This keyword immediately precedes the declaration of one or two relation names that provide access to the transition relations of the triggering statement. - referencing: Vec, - /// This specifies whether the trigger function should be fired once for - /// every row affected by the trigger event, or just once per SQL statement. - trigger_object: TriggerObject, - /// Whether to include the `EACH` term of the `FOR EACH`, as it is optional syntax. - include_each: bool, - /// Triggering conditions - condition: Option, - /// Execute logic block - exec_body: Option, - /// For MSSQL and dialects where statements are preceded by `AS` - statements_as: bool, - /// For SQL dialects with statement(s) for a body - statements: Option, - /// The characteristic of the trigger, which include whether the trigger is `DEFERRABLE`, `INITIALLY DEFERRED`, or `INITIALLY IMMEDIATE`, - characteristics: Option, - }, - /// DROP TRIGGER - /// - /// ```sql - /// DROP TRIGGER [ IF EXISTS ] name ON table_name [ CASCADE | RESTRICT ] - /// ``` - /// - DropTrigger { - if_exists: bool, - trigger_name: ObjectName, - table_name: Option, - /// `CASCADE` or `RESTRICT` - option: Option, - }, + /// CREATE TRIGGER statement. See struct [CreateTrigger] for details. + CreateTrigger(CreateTrigger), + /// DROP TRIGGER statement. See struct [DropTrigger] for details. + DropTrigger(DropTrigger), /// ```sql /// CREATE PROCEDURE /// ``` @@ -4984,97 +4880,8 @@ impl fmt::Display for Statement { } Statement::CreateFunction(create_function) => create_function.fmt(f), Statement::CreateDomain(create_domain) => create_domain.fmt(f), - Statement::CreateTrigger { - or_alter, - or_replace, - is_constraint, - name, - period_before_table, - period, - events, - table_name, - referenced_table_name, - referencing, - trigger_object, - condition, - include_each, - exec_body, - statements_as, - statements, - characteristics, - } => { - write!( - f, - "CREATE {or_alter}{or_replace}{is_constraint}TRIGGER {name} ", - or_alter = if *or_alter { "OR ALTER " } else { "" }, - or_replace = if *or_replace { "OR REPLACE " } else { "" }, - is_constraint = if *is_constraint { "CONSTRAINT " } else { "" }, - )?; - - if *period_before_table { - write!(f, "{period}")?; - if !events.is_empty() { - write!(f, " {}", display_separated(events, " OR "))?; - } - write!(f, " ON {table_name}")?; - } else { - write!(f, "ON {table_name}")?; - write!(f, " {period}")?; - if !events.is_empty() { - write!(f, " {}", display_separated(events, ", "))?; - } - } - - if let Some(referenced_table_name) = referenced_table_name { - write!(f, " FROM {referenced_table_name}")?; - } - - if let Some(characteristics) = characteristics { - write!(f, " {characteristics}")?; - } - - if !referencing.is_empty() { - write!(f, " REFERENCING {}", display_separated(referencing, " "))?; - } - - if *include_each { - write!(f, " FOR EACH {trigger_object}")?; - } else if exec_body.is_some() { - write!(f, " FOR {trigger_object}")?; - } - if let Some(condition) = condition { - write!(f, " WHEN {condition}")?; - } - if let Some(exec_body) = exec_body { - write!(f, " EXECUTE {exec_body}")?; - } - if let Some(statements) = statements { - if *statements_as { - write!(f, " AS")?; - } - write!(f, " {statements}")?; - } - Ok(()) - } - Statement::DropTrigger { - if_exists, - trigger_name, - table_name, - option, - } => { - write!(f, "DROP TRIGGER")?; - if *if_exists { - write!(f, " IF EXISTS")?; - } - match &table_name { - Some(table_name) => write!(f, " {trigger_name} ON {table_name}")?, - None => write!(f, " {trigger_name}")?, - }; - if let Some(option) = option { - write!(f, " {option}")?; - } - Ok(()) - } + Statement::CreateTrigger(create_trigger) => create_trigger.fmt(f), + Statement::DropTrigger(drop_trigger) => drop_trigger.fmt(f), Statement::CreateProcedure { name, or_alter, diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 518dab248..15946d5db 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -17,8 +17,8 @@ use crate::ast::helpers::attached_token::AttachedToken; use crate::ast::{ - BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, GranteesType, - IfStatement, Statement, TriggerObject, + BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, CreateTrigger, + GranteesType, IfStatement, Statement, TriggerObject, }; use crate::dialect::Dialect; use crate::keywords::{self, Keyword}; @@ -251,7 +251,7 @@ impl MsSqlDialect { parser.expect_keyword_is(Keyword::AS)?; let statements = Some(parser.parse_conditional_statements(&[Keyword::END])?); - Ok(Statement::CreateTrigger { + Ok(Statement::CreateTrigger(CreateTrigger { or_alter, or_replace: false, is_constraint: false, @@ -269,7 +269,7 @@ impl MsSqlDialect { statements_as: true, statements, characteristics: None, - }) + })) } /// Parse a sequence of statements, optionally separated by semicolon. diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 54db367d8..1bc2a3d6f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5564,12 +5564,12 @@ impl<'a> Parser<'a> { Keyword::RESTRICT => ReferentialAction::Restrict, _ => unreachable!(), }); - Ok(Statement::DropTrigger { + Ok(Statement::DropTrigger(DropTrigger { if_exists, trigger_name, table_name, option, - }) + })) } pub fn parse_create_trigger( @@ -5627,7 +5627,7 @@ impl<'a> Parser<'a> { statements = Some(self.parse_conditional_statements(&[Keyword::END])?); } - Ok(Statement::CreateTrigger { + Ok(Statement::CreateTrigger(CreateTrigger { or_alter, or_replace, is_constraint, @@ -5645,7 +5645,7 @@ impl<'a> Parser<'a> { statements_as: false, statements, characteristics, - }) + })) } pub fn parse_trigger_period(&mut self) -> Result { diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index a1e05d030..b1ad422ec 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2384,7 +2384,7 @@ fn parse_create_trigger() { let create_stmt = ms().verified_stmt(create_trigger); assert_eq!( create_stmt, - Statement::CreateTrigger { + Statement::CreateTrigger(CreateTrigger { or_alter: true, or_replace: false, is_constraint: false, @@ -2417,7 +2417,7 @@ fn parse_create_trigger() { }], }), characteristics: None, - } + }) ); let multi_statement_as_trigger = "\ @@ -2476,12 +2476,12 @@ fn parse_drop_trigger() { let drop_stmt = ms().one_statement_parses_to(sql_drop_trigger, ""); assert_eq!( drop_stmt, - Statement::DropTrigger { + Statement::DropTrigger(DropTrigger { if_exists: false, trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), table_name: None, option: None, - } + }) ); } diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 184532035..0857bae3b 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -3922,7 +3922,7 @@ fn parse_create_trigger() { let create_stmt = mysql().verified_stmt(sql_create_trigger); assert_eq!( create_stmt, - Statement::CreateTrigger { + Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: false, @@ -3946,7 +3946,7 @@ fn parse_create_trigger() { statements_as: false, statements: None, characteristics: None, - } + }) ); } @@ -3962,12 +3962,12 @@ fn parse_drop_trigger() { let drop_stmt = mysql().one_statement_parses_to(sql_drop_trigger, ""); assert_eq!( drop_stmt, - Statement::DropTrigger { + Statement::DropTrigger(DropTrigger { if_exists: false, trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), table_name: None, option: None, - } + }) ); } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 287b828a8..5b95bb300 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5671,7 +5671,7 @@ fn parse_create_domain() { #[test] fn parse_create_simple_before_insert_trigger() { let sql = "CREATE TRIGGER check_insert BEFORE INSERT ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_insert"; - let expected = Statement::CreateTrigger { + let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: false, @@ -5695,7 +5695,7 @@ fn parse_create_simple_before_insert_trigger() { statements_as: false, statements: None, characteristics: None, - }; + }); assert_eq!(pg().verified_stmt(sql), expected); } @@ -5703,7 +5703,7 @@ fn parse_create_simple_before_insert_trigger() { #[test] fn parse_create_after_update_trigger_with_condition() { let sql = "CREATE TRIGGER check_update AFTER UPDATE ON accounts FOR EACH ROW WHEN (NEW.balance > 10000) EXECUTE FUNCTION check_account_update"; - let expected = Statement::CreateTrigger { + let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: false, @@ -5734,7 +5734,7 @@ fn parse_create_after_update_trigger_with_condition() { statements_as: false, statements: None, characteristics: None, - }; + }); assert_eq!(pg().verified_stmt(sql), expected); } @@ -5742,7 +5742,7 @@ fn parse_create_after_update_trigger_with_condition() { #[test] fn parse_create_instead_of_delete_trigger() { let sql = "CREATE TRIGGER check_delete INSTEAD OF DELETE ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_deletes"; - let expected = Statement::CreateTrigger { + let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: false, @@ -5766,7 +5766,7 @@ fn parse_create_instead_of_delete_trigger() { statements_as: false, statements: None, characteristics: None, - }; + }); assert_eq!(pg().verified_stmt(sql), expected); } @@ -5774,7 +5774,7 @@ fn parse_create_instead_of_delete_trigger() { #[test] fn parse_create_trigger_with_multiple_events_and_deferrable() { let sql = "CREATE CONSTRAINT TRIGGER check_multiple_events BEFORE INSERT OR UPDATE OR DELETE ON accounts DEFERRABLE INITIALLY DEFERRED FOR EACH ROW EXECUTE FUNCTION check_account_changes"; - let expected = Statement::CreateTrigger { + let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: true, @@ -5806,7 +5806,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { initially: Some(DeferrableInitial::Deferred), enforced: None, }), - }; + }); assert_eq!(pg().verified_stmt(sql), expected); } @@ -5814,7 +5814,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { #[test] fn parse_create_trigger_with_referencing() { let sql = "CREATE TRIGGER check_referencing BEFORE INSERT ON accounts REFERENCING NEW TABLE AS new_accounts OLD TABLE AS old_accounts FOR EACH ROW EXECUTE FUNCTION check_account_referencing"; - let expected = Statement::CreateTrigger { + let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: false, @@ -5849,7 +5849,7 @@ fn parse_create_trigger_with_referencing() { statements_as: false, statements: None, characteristics: None, - }; + }); assert_eq!(pg().verified_stmt(sql), expected); } @@ -5901,12 +5901,12 @@ fn parse_drop_trigger() { ); assert_eq!( pg().verified_stmt(sql), - Statement::DropTrigger { + Statement::DropTrigger(DropTrigger { if_exists, trigger_name: ObjectName::from(vec![Ident::new("check_update")]), table_name: Some(ObjectName::from(vec![Ident::new("table_name")])), option - } + }) ); } } @@ -6130,7 +6130,7 @@ fn parse_trigger_related_functions() { assert_eq!( create_trigger, - Statement::CreateTrigger { + Statement::CreateTrigger(CreateTrigger { or_alter: false, or_replace: false, is_constraint: false, @@ -6154,18 +6154,18 @@ fn parse_trigger_related_functions() { statements_as: false, statements: None, characteristics: None - } + }) ); // Check the fourth statement assert_eq!( drop_trigger, - Statement::DropTrigger { + Statement::DropTrigger(DropTrigger { if_exists: false, trigger_name: ObjectName::from(vec![Ident::new("emp_stamp")]), table_name: Some(ObjectName::from(vec![Ident::new("emp")])), option: None - } + }) ); } From e3fbfd91b29706fcc7b4847f13ba71946a68e1f0 Mon Sep 17 00:00:00 2001 From: Sidney Cammeresi Date: Mon, 15 Sep 2025 00:33:17 -0700 Subject: [PATCH 258/286] MySQL: Support `CROSS JOIN` constraint (#2025) --- src/ast/query.rs | 9 +++++++-- src/ast/spans.rs | 2 +- src/dialect/mod.rs | 5 +++++ src/dialect/mysql.rs | 4 ++++ src/parser/mod.rs | 13 +++++++++++-- tests/sqlparser_common.rs | 35 ++++++++++++++++++++++++++++++++++- 6 files changed, 62 insertions(+), 6 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 151b94e82..3b414cc6e 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2333,7 +2333,11 @@ impl fmt::Display for Join { self.relation, suffix(constraint) )), - JoinOperator::CrossJoin => f.write_fmt(format_args!("CROSS JOIN {}", self.relation)), + JoinOperator::CrossJoin(constraint) => f.write_fmt(format_args!( + "CROSS JOIN {}{}", + self.relation, + suffix(constraint) + )), JoinOperator::Semi(constraint) => f.write_fmt(format_args!( "{}SEMI JOIN {}{}", prefix(constraint), @@ -2400,7 +2404,8 @@ pub enum JoinOperator { Right(JoinConstraint), RightOuter(JoinConstraint), FullOuter(JoinConstraint), - CrossJoin, + /// CROSS (constraint is non-standard) + CrossJoin(JoinConstraint), /// SEMI (non-standard) Semi(JoinConstraint), /// LEFT SEMI (non-standard) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 5d3694be8..4e15edd88 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -2226,7 +2226,7 @@ impl Spanned for JoinOperator { JoinOperator::Right(join_constraint) => join_constraint.span(), JoinOperator::RightOuter(join_constraint) => join_constraint.span(), JoinOperator::FullOuter(join_constraint) => join_constraint.span(), - JoinOperator::CrossJoin => Span::empty(), + JoinOperator::CrossJoin(join_constraint) => join_constraint.span(), JoinOperator::LeftSemi(join_constraint) => join_constraint.span(), JoinOperator::RightSemi(join_constraint) => join_constraint.span(), JoinOperator::LeftAnti(join_constraint) => join_constraint.span(), diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index 3c7be8f7e..ef4e1cdde 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -311,6 +311,11 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports a join specification on CROSS JOIN. + fn supports_cross_join_constraint(&self) -> bool { + false + } + /// Returns true if the dialect supports CONNECT BY. fn supports_connect_by(&self) -> bool { false diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index 8c63bfda5..8d2a5ad4b 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -163,6 +163,10 @@ impl Dialect for MySqlDialect { fn supports_data_type_signed_suffix(&self) -> bool { true } + + fn supports_cross_join_constraint(&self) -> bool { + true + } } /// `LOCK TABLES` diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 1bc2a3d6f..819819f9d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -13317,15 +13317,24 @@ impl<'a> Parser<'a> { let global = self.parse_keyword(Keyword::GLOBAL); let join = if self.parse_keyword(Keyword::CROSS) { let join_operator = if self.parse_keyword(Keyword::JOIN) { - JoinOperator::CrossJoin + JoinOperator::CrossJoin(JoinConstraint::None) } else if self.parse_keyword(Keyword::APPLY) { // MSSQL extension, similar to CROSS JOIN LATERAL JoinOperator::CrossApply } else { return self.expected("JOIN or APPLY after CROSS", self.peek_token()); }; + let relation = self.parse_table_factor()?; + let join_operator = if matches!(join_operator, JoinOperator::CrossJoin(_)) + && self.dialect.supports_cross_join_constraint() + { + let constraint = self.parse_join_constraint(false)?; + JoinOperator::CrossJoin(constraint) + } else { + join_operator + }; Join { - relation: self.parse_table_factor()?, + relation, global, join_operator, } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index aa47c0f7a..720c1e492 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -7131,12 +7131,45 @@ fn parse_cross_join() { Join { relation: table_from_name(ObjectName::from(vec![Ident::new("t2")])), global: false, - join_operator: JoinOperator::CrossJoin, + join_operator: JoinOperator::CrossJoin(JoinConstraint::None), }, only(only(select.from).joins), ); } +#[test] +fn parse_cross_join_constraint() { + fn join_with_constraint(constraint: JoinConstraint) -> Join { + Join { + relation: table_from_name(ObjectName::from(vec![Ident::new("t2")])), + global: false, + join_operator: JoinOperator::CrossJoin(constraint), + } + } + + fn test_constraint(sql: &str, constraint: JoinConstraint) { + let dialect = all_dialects_where(|d| d.supports_cross_join_constraint()); + let select = dialect.verified_only_select(sql); + assert_eq!( + join_with_constraint(constraint), + only(only(select.from).joins), + ); + } + + test_constraint( + "SELECT * FROM t1 CROSS JOIN t2 ON a = b", + JoinConstraint::On(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("a"))), + op: BinaryOperator::Eq, + right: Box::new(Expr::Identifier(Ident::new("b"))), + }), + ); + test_constraint( + "SELECT * FROM t1 CROSS JOIN t2 USING(a)", + JoinConstraint::Using(vec![ObjectName::from(vec![Ident::new("a")])]), + ); +} + #[test] fn parse_joins_on() { fn join_with_constraint( From 70215614747d17ed47234fc0bab85fecfcda1c9a Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Tue, 16 Sep 2025 20:21:55 +0200 Subject: [PATCH 259/286] Implemented the `From` method for all clear variants in Statement (#2028) --- src/ast/helpers/stmt_create_table.rs | 5 +- src/ast/mod.rs | 207 +++++++++++++++++++++++++-- src/dialect/mssql.rs | 10 +- 3 files changed, 208 insertions(+), 14 deletions(-) diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index fc0002697..fe950c909 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -432,7 +432,7 @@ impl CreateTableBuilder { } pub fn build(self) -> Statement { - Statement::CreateTable(CreateTable { + CreateTable { or_replace: self.or_replace, temporary: self.temporary, external: self.external, @@ -484,7 +484,8 @@ impl CreateTableBuilder { refresh_mode: self.refresh_mode, initialize: self.initialize, require_user: self.require_user, - }) + } + .into() } } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 6efe1f748..8df636f8e 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -21,6 +21,7 @@ use alloc::{ boxed::Box, format, string::{String, ToString}, + vec, vec::Vec, }; use helpers::{ @@ -3029,14 +3030,6 @@ impl Display for Set { } } -/// Convert a `Set` into a `Statement`. -/// Convenience function, instead of writing `Statement::Set(Set::Set...{...})` -impl From for Statement { - fn from(set: Set) -> Self { - Statement::Set(set) - } -} - /// A representation of a `WHEN` arm with all the identifiers catched and the statements to execute /// for the arm. /// @@ -10707,6 +10700,204 @@ impl fmt::Display for VacuumStatement { } } +impl From for Statement { + fn from(s: Set) -> Self { + Self::Set(s) + } +} + +impl From for Statement { + fn from(q: Query) -> Self { + Box::new(q).into() + } +} + +impl From> for Statement { + fn from(q: Box) -> Self { + Self::Query(q) + } +} + +impl From for Statement { + fn from(i: Insert) -> Self { + Self::Insert(i) + } +} + +impl From for Statement { + fn from(c: CaseStatement) -> Self { + Self::Case(c) + } +} + +impl From for Statement { + fn from(i: IfStatement) -> Self { + Self::If(i) + } +} + +impl From for Statement { + fn from(w: WhileStatement) -> Self { + Self::While(w) + } +} + +impl From for Statement { + fn from(r: RaiseStatement) -> Self { + Self::Raise(r) + } +} + +impl From for Statement { + fn from(f: Function) -> Self { + Self::Call(f) + } +} + +impl From for Statement { + fn from(o: OpenStatement) -> Self { + Self::Open(o) + } +} + +impl From for Statement { + fn from(d: Delete) -> Self { + Self::Delete(d) + } +} + +impl From for Statement { + fn from(c: CreateTable) -> Self { + Self::CreateTable(c) + } +} + +impl From for Statement { + fn from(c: CreateIndex) -> Self { + Self::CreateIndex(c) + } +} + +impl From for Statement { + fn from(c: CreateServerStatement) -> Self { + Self::CreateServer(c) + } +} + +impl From for Statement { + fn from(c: CreateConnector) -> Self { + Self::CreateConnector(c) + } +} + +impl From for Statement { + fn from(a: AlterSchema) -> Self { + Self::AlterSchema(a) + } +} + +impl From for Statement { + fn from(a: AlterType) -> Self { + Self::AlterType(a) + } +} + +impl From for Statement { + fn from(d: DropDomain) -> Self { + Self::DropDomain(d) + } +} + +impl From for Statement { + fn from(s: ShowCharset) -> Self { + Self::ShowCharset(s) + } +} + +impl From for Statement { + fn from(s: ShowObjects) -> Self { + Self::ShowObjects(s) + } +} + +impl From for Statement { + fn from(u: Use) -> Self { + Self::Use(u) + } +} + +impl From for Statement { + fn from(c: CreateFunction) -> Self { + Self::CreateFunction(c) + } +} + +impl From for Statement { + fn from(c: CreateTrigger) -> Self { + Self::CreateTrigger(c) + } +} + +impl From for Statement { + fn from(d: DropTrigger) -> Self { + Self::DropTrigger(d) + } +} + +impl From for Statement { + fn from(d: DenyStatement) -> Self { + Self::Deny(d) + } +} + +impl From for Statement { + fn from(c: CreateDomain) -> Self { + Self::CreateDomain(c) + } +} + +impl From for Statement { + fn from(r: RenameTable) -> Self { + vec![r].into() + } +} + +impl From> for Statement { + fn from(r: Vec) -> Self { + Self::RenameTable(r) + } +} + +impl From for Statement { + fn from(p: PrintStatement) -> Self { + Self::Print(p) + } +} + +impl From for Statement { + fn from(r: ReturnStatement) -> Self { + Self::Return(r) + } +} + +impl From for Statement { + fn from(e: ExportData) -> Self { + Self::ExportData(e) + } +} + +impl From for Statement { + fn from(c: CreateUser) -> Self { + Self::CreateUser(c) + } +} + +impl From for Statement { + fn from(v: VacuumStatement) -> Self { + Self::Vacuum(v) + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 15946d5db..4fcc0e4b6 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -226,12 +226,13 @@ impl MsSqlDialect { parser.prev_token(); } - Ok(Statement::If(IfStatement { + Ok(IfStatement { if_block, else_block, elseif_blocks: Vec::new(), end_token: None, - })) + } + .into()) } /// Parse `CREATE TRIGGER` for [MsSql] @@ -251,7 +252,7 @@ impl MsSqlDialect { parser.expect_keyword_is(Keyword::AS)?; let statements = Some(parser.parse_conditional_statements(&[Keyword::END])?); - Ok(Statement::CreateTrigger(CreateTrigger { + Ok(CreateTrigger { or_alter, or_replace: false, is_constraint: false, @@ -269,7 +270,8 @@ impl MsSqlDialect { statements_as: true, statements, characteristics: None, - })) + } + .into()) } /// Parse a sequence of statements, optionally separated by semicolon. From f642dd573cbf700b8591eeb9e1f6bf3ed8f98659 Mon Sep 17 00:00:00 2001 From: Ryan Schneider Date: Tue, 16 Sep 2025 11:25:24 -0700 Subject: [PATCH 260/286] DuckDB: Allow quoted date parts in EXTRACT (#2030) --- src/dialect/duckdb.rs | 5 +++++ tests/sqlparser_duckdb.rs | 6 ++++++ 2 files changed, 11 insertions(+) diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index 37cf7d369..f08d827b9 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -70,6 +70,11 @@ impl Dialect for DuckDbDialect { true } + /// Returns true if this dialect allows the `EXTRACT` function to use single quotes in the part being extracted. + fn allow_extract_single_quotes(&self) -> bool { + true + } + // DuckDB is compatible with PostgreSQL syntax for this statement, // although not all features may be implemented. fn supports_explain_with_utility_options(&self) -> bool { diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 5bad73365..0f8051955 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -869,3 +869,9 @@ fn test_duckdb_trim() { duckdb().parse_sql_statements(error_sql).unwrap_err() ); } + +#[test] +fn parse_extract_single_quotes() { + let sql = "SELECT EXTRACT('month' FROM my_timestamp) FROM my_table"; + duckdb().verified_stmt(sql); +} From ea7f9026f76ff67f45a414d44000aa41cd35ea15 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Fri, 19 Sep 2025 11:04:56 +0300 Subject: [PATCH 261/286] MySQL: Add support for unsigned numeric types (#2031) --- src/ast/data_type.rs | 47 ++++++++++++++++++-- src/parser/mod.rs | 58 +++++++++++++++++++------ tests/sqlparser_mysql.rs | 92 ++++++++++++++++++++++++++++++++++++++++ 3 files changed, 182 insertions(+), 15 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index dde9e0b7e..f8523e844 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -131,6 +131,11 @@ pub enum DataType { /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type Decimal(ExactNumberInfo), + /// [MySQL] unsigned decimal with optional precision and scale, e.g. DECIMAL UNSIGNED or DECIMAL(10,2) UNSIGNED. + /// Note: Using UNSIGNED with DECIMAL is deprecated in recent versions of MySQL. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/numeric-type-syntax.html + DecimalUnsigned(ExactNumberInfo), /// [BigNumeric] type used in BigQuery. /// /// [BigNumeric]: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#bignumeric_literals @@ -143,8 +148,19 @@ pub enum DataType { /// /// [1]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#exact-numeric-type Dec(ExactNumberInfo), - /// Floating point with optional precision, e.g. FLOAT(8). - Float(Option), + /// [MySQL] unsigned decimal (DEC alias) with optional precision and scale, e.g. DEC UNSIGNED or DEC(10,2) UNSIGNED. + /// Note: Using UNSIGNED with DEC is deprecated in recent versions of MySQL. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/numeric-type-syntax.html + DecUnsigned(ExactNumberInfo), + /// Floating point with optional precision and scale, e.g. FLOAT, FLOAT(8), or FLOAT(8,2). + Float(ExactNumberInfo), + /// [MySQL] unsigned floating point with optional precision and scale, e.g. + /// FLOAT UNSIGNED, FLOAT(10) UNSIGNED or FLOAT(10,2) UNSIGNED. + /// Note: Using UNSIGNED with FLOAT is deprecated in recent versions of MySQL. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/numeric-type-syntax.html + FloatUnsigned(ExactNumberInfo), /// Tiny integer with optional display width, e.g. TINYINT or TINYINT(3). TinyInt(Option), /// Unsigned tiny integer with optional display width, @@ -302,17 +318,32 @@ pub enum DataType { Float64, /// Floating point, e.g. REAL. Real, + /// [MySQL] unsigned real, e.g. REAL UNSIGNED. + /// Note: Using UNSIGNED with REAL is deprecated in recent versions of MySQL. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/numeric-type-syntax.html + RealUnsigned, /// Float8 is an alias for Double in [PostgreSQL]. /// /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html Float8, /// Double Double(ExactNumberInfo), + /// [MySQL] unsigned double precision with optional precision, e.g. DOUBLE UNSIGNED or DOUBLE(10,2) UNSIGNED. + /// Note: Using UNSIGNED with DOUBLE is deprecated in recent versions of MySQL. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/numeric-type-syntax.html + DoubleUnsigned(ExactNumberInfo), /// Double Precision, see [SQL Standard], [PostgreSQL]. /// /// [SQL Standard]: https://jakewheat.github.io/sql-overview/sql-2016-foundation-grammar.html#approximate-numeric-type /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype-numeric.html DoublePrecision, + /// [MySQL] unsigned double precision, e.g. DOUBLE PRECISION UNSIGNED. + /// Note: Using UNSIGNED with DOUBLE PRECISION is deprecated in recent versions of MySQL. + /// + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/numeric-type-syntax.html + DoublePrecisionUnsigned, /// Bool is an alias for Boolean, see [PostgreSQL]. /// /// [PostgreSQL]: https://www.postgresql.org/docs/current/datatype.html @@ -497,12 +528,19 @@ impl fmt::Display for DataType { DataType::Decimal(info) => { write!(f, "DECIMAL{info}") } + DataType::DecimalUnsigned(info) => { + write!(f, "DECIMAL{info} UNSIGNED") + } DataType::Dec(info) => { write!(f, "DEC{info}") } + DataType::DecUnsigned(info) => { + write!(f, "DEC{info} UNSIGNED") + } DataType::BigNumeric(info) => write!(f, "BIGNUMERIC{info}"), DataType::BigDecimal(info) => write!(f, "BIGDECIMAL{info}"), - DataType::Float(size) => format_type_with_optional_length(f, "FLOAT", size, false), + DataType::Float(info) => write!(f, "FLOAT{info}"), + DataType::FloatUnsigned(info) => write!(f, "FLOAT{info} UNSIGNED"), DataType::TinyInt(zerofill) => { format_type_with_optional_length(f, "TINYINT", zerofill, false) } @@ -616,12 +654,15 @@ impl fmt::Display for DataType { write!(f, "UNSIGNED INTEGER") } DataType::Real => write!(f, "REAL"), + DataType::RealUnsigned => write!(f, "REAL UNSIGNED"), DataType::Float4 => write!(f, "FLOAT4"), DataType::Float32 => write!(f, "Float32"), DataType::Float64 => write!(f, "FLOAT64"), DataType::Double(info) => write!(f, "DOUBLE{info}"), + DataType::DoubleUnsigned(info) => write!(f, "DOUBLE{info} UNSIGNED"), DataType::Float8 => write!(f, "FLOAT8"), DataType::DoublePrecision => write!(f, "DOUBLE PRECISION"), + DataType::DoublePrecisionUnsigned => write!(f, "DOUBLE PRECISION UNSIGNED"), DataType::Bool => write!(f, "BOOL"), DataType::Boolean => write!(f, "BOOLEAN"), DataType::Date => write!(f, "DATE"), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 819819f9d..e121066f2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10181,19 +10181,41 @@ impl<'a> Parser<'a> { Token::Word(w) => match w.keyword { Keyword::BOOLEAN => Ok(DataType::Boolean), Keyword::BOOL => Ok(DataType::Bool), - Keyword::FLOAT => Ok(DataType::Float(self.parse_optional_precision()?)), - Keyword::REAL => Ok(DataType::Real), + Keyword::FLOAT => { + let precision = self.parse_exact_number_optional_precision_scale()?; + + if self.parse_keyword(Keyword::UNSIGNED) { + Ok(DataType::FloatUnsigned(precision)) + } else { + Ok(DataType::Float(precision)) + } + } + Keyword::REAL => { + if self.parse_keyword(Keyword::UNSIGNED) { + Ok(DataType::RealUnsigned) + } else { + Ok(DataType::Real) + } + } Keyword::FLOAT4 => Ok(DataType::Float4), Keyword::FLOAT32 => Ok(DataType::Float32), Keyword::FLOAT64 => Ok(DataType::Float64), Keyword::FLOAT8 => Ok(DataType::Float8), Keyword::DOUBLE => { if self.parse_keyword(Keyword::PRECISION) { - Ok(DataType::DoublePrecision) + if self.parse_keyword(Keyword::UNSIGNED) { + Ok(DataType::DoublePrecisionUnsigned) + } else { + Ok(DataType::DoublePrecision) + } } else { - Ok(DataType::Double( - self.parse_exact_number_optional_precision_scale()?, - )) + let precision = self.parse_exact_number_optional_precision_scale()?; + + if self.parse_keyword(Keyword::UNSIGNED) { + Ok(DataType::DoubleUnsigned(precision)) + } else { + Ok(DataType::Double(precision)) + } } } Keyword::TINYINT => { @@ -10420,12 +10442,24 @@ impl<'a> Parser<'a> { Keyword::NUMERIC => Ok(DataType::Numeric( self.parse_exact_number_optional_precision_scale()?, )), - Keyword::DECIMAL => Ok(DataType::Decimal( - self.parse_exact_number_optional_precision_scale()?, - )), - Keyword::DEC => Ok(DataType::Dec( - self.parse_exact_number_optional_precision_scale()?, - )), + Keyword::DECIMAL => { + let precision = self.parse_exact_number_optional_precision_scale()?; + + if self.parse_keyword(Keyword::UNSIGNED) { + Ok(DataType::DecimalUnsigned(precision)) + } else { + Ok(DataType::Decimal(precision)) + } + } + Keyword::DEC => { + let precision = self.parse_exact_number_optional_precision_scale()?; + + if self.parse_keyword(Keyword::UNSIGNED) { + Ok(DataType::DecUnsigned(precision)) + } else { + Ok(DataType::Dec(precision)) + } + } Keyword::BIGNUMERIC => Ok(DataType::BigNumeric( self.parse_exact_number_optional_precision_scale()?, )), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 0857bae3b..5d75aa508 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1757,6 +1757,98 @@ fn parse_signed_data_types() { .expect_err("SIGNED suffix should not be allowed"); } +#[test] +fn parse_deprecated_mysql_unsigned_data_types() { + let sql = "CREATE TABLE foo (bar_decimal DECIMAL UNSIGNED, bar_decimal_prec DECIMAL(10) UNSIGNED, bar_decimal_scale DECIMAL(10,2) UNSIGNED, bar_dec DEC UNSIGNED, bar_dec_prec DEC(10) UNSIGNED, bar_dec_scale DEC(10,2) UNSIGNED, bar_float FLOAT UNSIGNED, bar_float_prec FLOAT(10) UNSIGNED, bar_float_scale FLOAT(10,2) UNSIGNED, bar_double DOUBLE UNSIGNED, bar_double_prec DOUBLE(10) UNSIGNED, bar_double_scale DOUBLE(10,2) UNSIGNED, bar_real REAL UNSIGNED, bar_double_precision DOUBLE PRECISION UNSIGNED)"; + match mysql().verified_stmt(sql) { + Statement::CreateTable(CreateTable { name, columns, .. }) => { + assert_eq!(name.to_string(), "foo"); + assert_eq!( + vec![ + ColumnDef { + name: Ident::new("bar_decimal"), + data_type: DataType::DecimalUnsigned(ExactNumberInfo::None), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_decimal_prec"), + data_type: DataType::DecimalUnsigned(ExactNumberInfo::Precision(10)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_decimal_scale"), + data_type: DataType::DecimalUnsigned(ExactNumberInfo::PrecisionAndScale( + 10, 2 + )), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_dec"), + data_type: DataType::DecUnsigned(ExactNumberInfo::None), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_dec_prec"), + data_type: DataType::DecUnsigned(ExactNumberInfo::Precision(10)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_dec_scale"), + data_type: DataType::DecUnsigned(ExactNumberInfo::PrecisionAndScale(10, 2)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_float"), + data_type: DataType::FloatUnsigned(ExactNumberInfo::None), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_float_prec"), + data_type: DataType::FloatUnsigned(ExactNumberInfo::Precision(10)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_float_scale"), + data_type: DataType::FloatUnsigned(ExactNumberInfo::PrecisionAndScale( + 10, 2 + )), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_double"), + data_type: DataType::DoubleUnsigned(ExactNumberInfo::None), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_double_prec"), + data_type: DataType::DoubleUnsigned(ExactNumberInfo::Precision(10)), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_double_scale"), + data_type: DataType::DoubleUnsigned(ExactNumberInfo::PrecisionAndScale( + 10, 2 + )), + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_real"), + data_type: DataType::RealUnsigned, + options: vec![], + }, + ColumnDef { + name: Ident::new("bar_double_precision"), + data_type: DataType::DoublePrecisionUnsigned, + options: vec![], + }, + ], + columns + ); + } + _ => unreachable!(), + } +} + #[test] fn parse_simple_insert() { let sql = r"INSERT INTO tasks (title, priority) VALUES ('Test Some Inserts', 1), ('Test Entry 2', 2), ('Test Entry 3', 3)"; From 0b723147b678a9e84ebb2c46245e80950245a314 Mon Sep 17 00:00:00 2001 From: Chen Chongchen Date: Fri, 19 Sep 2025 16:37:27 +0800 Subject: [PATCH 262/286] feat: support postgres alter schema (#2038) --- src/ast/ddl.rs | 9 ++++++ src/ast/spans.rs | 10 +++++- src/parser/mod.rs | 6 ++++ tests/sqlparser_postgres.rs | 63 +++++++++++++++++++++++++++++++++++++ 4 files changed, 87 insertions(+), 1 deletion(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index a15474755..f998143c8 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -3084,6 +3084,7 @@ impl fmt::Display for CreateConnector { /// An `ALTER SCHEMA` (`Statement::AlterSchema`) operation. /// /// See [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#alter_schema_collate_statement) +/// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-alterschema.html) #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -3101,6 +3102,12 @@ pub enum AlterSchemaOperation { SetOptionsParens { options: Vec, }, + Rename { + name: ObjectName, + }, + OwnerTo { + owner: Owner, + }, } impl fmt::Display for AlterSchemaOperation { @@ -3120,6 +3127,8 @@ impl fmt::Display for AlterSchemaOperation { AlterSchemaOperation::SetOptionsParens { options } => { write!(f, "SET OPTIONS ({})", display_comma_separated(options)) } + AlterSchemaOperation::Rename { name } => write!(f, "RENAME TO {name}"), + AlterSchemaOperation::OwnerTo { owner } => write!(f, "OWNER TO {owner}"), } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 4e15edd88..0c205b825 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -17,7 +17,7 @@ use crate::ast::{ ddl::AlterSchema, query::SelectItemQualifiedWildcardKind, AlterSchemaOperation, ColumnOptions, - ExportData, TypedString, + ExportData, Owner, TypedString, }; use core::iter; @@ -2424,6 +2424,14 @@ impl Spanned for AlterSchemaOperation { AlterSchemaOperation::SetOptionsParens { options } => { union_spans(options.iter().map(|i| i.span())) } + AlterSchemaOperation::Rename { name } => name.span(), + AlterSchemaOperation::OwnerTo { owner } => { + if let Owner::Ident(ident) = owner { + ident.span + } else { + Span::empty() + } + } } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e121066f2..d661efd4d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9444,6 +9444,12 @@ impl<'a> Parser<'a> { } else if self.parse_keywords(&[Keyword::DROP, Keyword::REPLICA]) { let replica = self.parse_identifier()?; AlterSchemaOperation::DropReplica { replica } + } else if self.parse_keywords(&[Keyword::RENAME, Keyword::TO]) { + let new_name = self.parse_object_name(false)?; + AlterSchemaOperation::Rename { name: new_name } + } else if self.parse_keywords(&[Keyword::OWNER, Keyword::TO]) { + let owner = self.parse_owner()?; + AlterSchemaOperation::OwnerTo { owner } } else { return self.expected_ref("ALTER SCHEMA operation", self.peek_token_ref()); }; diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 5b95bb300..196a82f54 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -6576,3 +6576,66 @@ fn parse_create_server() { assert_eq!(stmt, expected); } } + +#[test] +fn parse_alter_schema() { + match pg_and_generic().verified_stmt("ALTER SCHEMA foo RENAME TO bar") { + Statement::AlterSchema(AlterSchema { operations, .. }) => { + assert_eq!( + operations, + vec![AlterSchemaOperation::Rename { + name: ObjectName::from(vec!["bar".into()]) + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("ALTER SCHEMA foo OWNER TO bar") { + Statement::AlterSchema(AlterSchema { operations, .. }) => { + assert_eq!( + operations, + vec![AlterSchemaOperation::OwnerTo { + owner: Owner::Ident("bar".into()) + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("ALTER SCHEMA foo OWNER TO CURRENT_ROLE") { + Statement::AlterSchema(AlterSchema { operations, .. }) => { + assert_eq!( + operations, + vec![AlterSchemaOperation::OwnerTo { + owner: Owner::CurrentRole + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("ALTER SCHEMA foo OWNER TO CURRENT_USER") { + Statement::AlterSchema(AlterSchema { operations, .. }) => { + assert_eq!( + operations, + vec![AlterSchemaOperation::OwnerTo { + owner: Owner::CurrentUser + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("ALTER SCHEMA foo OWNER TO SESSION_USER") { + Statement::AlterSchema(AlterSchema { operations, .. }) => { + assert_eq!( + operations, + vec![AlterSchemaOperation::OwnerTo { + owner: Owner::SessionUser + }] + ); + } + _ => unreachable!(), + } +} From 3c61db5dc09871405f04ea715f0a0800072568cc Mon Sep 17 00:00:00 2001 From: Andrew Lamb Date: Fri, 19 Sep 2025 05:51:30 -0700 Subject: [PATCH 263/286] Prepare for 0.59.0: Changelog and update version (#2039) --- Cargo.toml | 2 +- changelog/0.59.0.md | 122 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 123 insertions(+), 1 deletion(-) create mode 100644 changelog/0.59.0.md diff --git a/Cargo.toml b/Cargo.toml index 2e48f9d0d..48db33b3f 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.58.0" +version = "0.59.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/changelog/0.59.0.md b/changelog/0.59.0.md new file mode 100644 index 000000000..a3b14f1d8 --- /dev/null +++ b/changelog/0.59.0.md @@ -0,0 +1,122 @@ + + +# sqlparser-rs 0.59.0 Changelog + +This release consists of 59 commits from 22 contributors. See credits at the end of this changelog for more information. + +**Implemented enhancements:** + +- feat: support export data for bigquery [#1976](https://github.com/apache/datafusion-sqlparser-rs/pull/1976) (chenkovsky) +- feat: support multi value columns and aliases in unpivot [#1969](https://github.com/apache/datafusion-sqlparser-rs/pull/1969) (chenkovsky) +- feat: Include end token in `ALTER TABLE` statement [#1999](https://github.com/apache/datafusion-sqlparser-rs/pull/1999) (IndexSeek) +- feat: support multiple value for pivot [#1970](https://github.com/apache/datafusion-sqlparser-rs/pull/1970) (chenkovsky) +- feat: Add `ALTER SCHEMA` support [#1980](https://github.com/apache/datafusion-sqlparser-rs/pull/1980) (chenkovsky) +- feat: MERGE statements: add RETURNING and OUTPUT without INTO [#2011](https://github.com/apache/datafusion-sqlparser-rs/pull/2011) (lovasoa) +- feat: support postgres alter schema [#2038](https://github.com/apache/datafusion-sqlparser-rs/pull/2038) (chenkovsky) + +**Fixed bugs:** + +- fix: begin statement for bigquery [#1975](https://github.com/apache/datafusion-sqlparser-rs/pull/1975) (chenkovsky) +- fix: update DuckDB and ClickHouse documentation links [#1978](https://github.com/apache/datafusion-sqlparser-rs/pull/1978) (IndexSeek) + +**Other:** + +- MySQL: Support `EXPLAIN ANALYZE` format variants [#1945](https://github.com/apache/datafusion-sqlparser-rs/pull/1945) (yoavcloud) +- Add support for `NOT NULL` and `NOTNULL` expressions [#1927](https://github.com/apache/datafusion-sqlparser-rs/pull/1927) (ryanschneider) +- Snowflake: Support `CLONE` option in `CREATE DATABASE/SCHEMA` statements [#1958](https://github.com/apache/datafusion-sqlparser-rs/pull/1958) (yoavcloud) +- Add support for `GRANT DROP` statement [#1959](https://github.com/apache/datafusion-sqlparser-rs/pull/1959) (yoavcloud) +- Snowflake: Add support for `CREATE USER` [#1950](https://github.com/apache/datafusion-sqlparser-rs/pull/1950) (yoavcloud) +- Postgres: Support parenthesized `SET` options for `ALTER TABLE` [#1947](https://github.com/apache/datafusion-sqlparser-rs/pull/1947) (achristmascarl) +- Snowflake: Support IDENTIFIER for GRANT ROLE [#1957](https://github.com/apache/datafusion-sqlparser-rs/pull/1957) (yoavcloud) +- Snowflake: Numeric prefix for stage name part [#1966](https://github.com/apache/datafusion-sqlparser-rs/pull/1966) (yoavcloud) +- Snowflake: Support `GRANT CREATE SCHEMA` `GRANT .. ON ALL FUNCTIONS IN SCHEMA` [#1964](https://github.com/apache/datafusion-sqlparser-rs/pull/1964) (yoavcloud) +- Snowflake: DROP STREAM [#1973](https://github.com/apache/datafusion-sqlparser-rs/pull/1973) (yoavcloud) +- Add ODBC escape syntax support for time expressions [#1953](https://github.com/apache/datafusion-sqlparser-rs/pull/1953) (etgarperets) +- Add support for `SHOW CHARSET` [#1974](https://github.com/apache/datafusion-sqlparser-rs/pull/1974) (etgarperets) +- Snowflake: Support `CREATE VIEW myview IF NOT EXISTS` [#1961](https://github.com/apache/datafusion-sqlparser-rs/pull/1961) (etgarperets) +- Update criterion requirement from 0.6 to 0.7 in /sqlparser_bench [#1981](https://github.com/apache/datafusion-sqlparser-rs/pull/1981) (dependabot[bot]) +- Snowflake: Improve support for reserved keywords for table factor [#1942](https://github.com/apache/datafusion-sqlparser-rs/pull/1942) (yoavcloud) +- MySQL: Allow optional `SIGNED` suffix on integer data types [#1985](https://github.com/apache/datafusion-sqlparser-rs/pull/1985) (mvzink) +- Fix placeholder spans [#1979](https://github.com/apache/datafusion-sqlparser-rs/pull/1979) (xitep) +- Snowflake create database [#1939](https://github.com/apache/datafusion-sqlparser-rs/pull/1939) (osipovartem) +- Postgres: Support `INTERVAL` data type options [#1984](https://github.com/apache/datafusion-sqlparser-rs/pull/1984) (mvzink) +- MySQL: Support comma-separated `CREATE TABLE` options [#1989](https://github.com/apache/datafusion-sqlparser-rs/pull/1989) (mvzink) +- MySQL: Support `ALTER TABLE RENAME AS` [#1965](https://github.com/apache/datafusion-sqlparser-rs/pull/1965) (altmannmarcelo) +- Improve MySQL `CREATE TRIGGER` parsing [#1998](https://github.com/apache/datafusion-sqlparser-rs/pull/1998) (mvzink) +- Snowflake - support table function in table factor (regression) [#1996](https://github.com/apache/datafusion-sqlparser-rs/pull/1996) (tomershaniii) +- Improve MySQL option parsing in index definitions [#1997](https://github.com/apache/datafusion-sqlparser-rs/pull/1997) (mvzink) +- Add support for `UPDATE ... LIMIT ...` [#1991](https://github.com/apache/datafusion-sqlparser-rs/pull/1991) (xtuc) +- Postgres: enhance NUMERIC/DECIMAL parsing to support negative scale [#1990](https://github.com/apache/datafusion-sqlparser-rs/pull/1990) (IndexSeek) +- Fix column definition `COLLATE` parsing [#1986](https://github.com/apache/datafusion-sqlparser-rs/pull/1986) (mvzink) +- Redshift: CREATE TABLE ... (LIKE ..) [#1967](https://github.com/apache/datafusion-sqlparser-rs/pull/1967) (yoavcloud) +- Add drop behavior to `DROP PRIMARY/FOREIGN KEY` [#2002](https://github.com/apache/datafusion-sqlparser-rs/pull/2002) (yoavcloud) +- Redshift: Add support for IAM_ROLE and IGNOREHEADER COPY options [#1968](https://github.com/apache/datafusion-sqlparser-rs/pull/1968) (yoavcloud) +- Snowflake: Add support for `CREATE DYNAMIC TABLE` [#1960](https://github.com/apache/datafusion-sqlparser-rs/pull/1960) (yoavcloud) +- Add support for VACUUM in Redshift [#2005](https://github.com/apache/datafusion-sqlparser-rs/pull/2005) (yoavcloud) +- Add support for `SEMANTIC_VIEW` table factor [#2009](https://github.com/apache/datafusion-sqlparser-rs/pull/2009) (bombsimon) +- Redshift: Add more copy options [#2008](https://github.com/apache/datafusion-sqlparser-rs/pull/2008) (yoavcloud) +- `GenericDialect`: Support pipe operator [#2012](https://github.com/apache/datafusion-sqlparser-rs/pull/2012) (simonvandel) +- Add SECURE keyword for views in Snowflake [#2004](https://github.com/apache/datafusion-sqlparser-rs/pull/2004) (Vedin) +- Add support for PostgreSQL JSON function 'RETURNING' clauses [#2001](https://github.com/apache/datafusion-sqlparser-rs/pull/2001) (adamchainz) +- Snowflake: Minus char in stage name [#2014](https://github.com/apache/datafusion-sqlparser-rs/pull/2014) (yoavcloud) +- Support wildcard metrics for `SEMANTIC_VIEW` [#2016](https://github.com/apache/datafusion-sqlparser-rs/pull/2016) (bombsimon) +- Allow wilrdacrd for all `SEMANTIC_VIEW` types [#2017](https://github.com/apache/datafusion-sqlparser-rs/pull/2017) (bombsimon) +- Redshift: UNLOAD [#2013](https://github.com/apache/datafusion-sqlparser-rs/pull/2013) (yoavcloud) +- Add support for string literal concatenation [#2003](https://github.com/apache/datafusion-sqlparser-rs/pull/2003) (etgarperets) +- Enable merge queue in sqlparser-rs [#2007](https://github.com/apache/datafusion-sqlparser-rs/pull/2007) (blaginin) +- Merge Queue Test [#2019](https://github.com/apache/datafusion-sqlparser-rs/pull/2019) (blaginin) +- Added derive trait `Copy` to `OrderByOptions` struct [#2021](https://github.com/apache/datafusion-sqlparser-rs/pull/2021) (LucaCappelletti94) +- Moved `CreateTrigger` and `DropTrigger` out of `Statement` enum [#2026](https://github.com/apache/datafusion-sqlparser-rs/pull/2026) (LucaCappelletti94) +- MySQL: Support `CROSS JOIN` constraint [#2025](https://github.com/apache/datafusion-sqlparser-rs/pull/2025) (rs-sac) +- Implemented the `From` method for all clear variants in Statement [#2028](https://github.com/apache/datafusion-sqlparser-rs/pull/2028) (LucaCappelletti94) +- DuckDB: Allow quoted date parts in EXTRACT [#2030](https://github.com/apache/datafusion-sqlparser-rs/pull/2030) (ryanschneider) +- MySQL: Add support for unsigned numeric types [#2031](https://github.com/apache/datafusion-sqlparser-rs/pull/2031) (MohamedAbdeen21) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 17 Yoav Cohen + 6 Chen Chongchen + 6 Michael Victor Zink + 4 etgarperets + 3 Luca Cappelletti + 3 Simon Sawert + 3 Tyler White + 2 Dmitrii Blaginin + 2 Ryan Schneider + 1 Adam Johnson + 1 Artem Osipov + 1 Denys Tsomenko + 1 Marcelo Altmann + 1 Mohamed Abdeen + 1 Ophir LOJKINE + 1 Sidney Cammeresi + 1 Simon Vandel Sillesen + 1 Sven Sauleau + 1 carl + 1 dependabot[bot] + 1 tomershaniii + 1 xitep +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + From a15c70dff98b2742179b95dca048d84281416466 Mon Sep 17 00:00:00 2001 From: Marcelo Altmann Date: Tue, 23 Sep 2025 08:07:50 -0300 Subject: [PATCH 264/286] Add support for INVISIBLE columns in MySQL (#2033) --- src/ast/ddl.rs | 10 +++++++ src/ast/spans.rs | 1 + src/keywords.rs | 1 + src/parser/mod.rs | 2 ++ tests/sqlparser_common.rs | 55 ++++++++++++++++++++++++++++++++++++++- 5 files changed, 68 insertions(+), 1 deletion(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index f998143c8..4c145f91e 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1911,6 +1911,13 @@ pub enum ColumnOption { /// ``` /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/creating-spatial-indexes.html Srid(Box), + /// MySQL specific: Column is invisible via SELECT * + /// Syntax: + /// ```sql + /// CREATE TABLE t (foo INT, bar INT INVISIBLE); + /// ``` + /// [MySQL]: https://dev.mysql.com/doc/refman/8.4/en/invisible-columns.html + Invisible, } impl fmt::Display for ColumnOption { @@ -2029,6 +2036,9 @@ impl fmt::Display for ColumnOption { Srid(srid) => { write!(f, "SRID {srid}") } + Invisible => { + write!(f, "INVISIBLE") + } } } } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 0c205b825..5913fe164 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -918,6 +918,7 @@ impl Spanned for ColumnOption { ColumnOption::Policy(..) => Span::empty(), ColumnOption::Tags(..) => Span::empty(), ColumnOption::Srid(..) => Span::empty(), + ColumnOption::Invisible => Span::empty(), } } } diff --git a/src/keywords.rs b/src/keywords.rs index 3358e2845..e0590b69d 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -499,6 +499,7 @@ define_keywords!( INTERSECTION, INTERVAL, INTO, + INVISIBLE, INVOKER, IO, IS, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d661efd4d..f47d4b9b5 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8144,6 +8144,8 @@ impl<'a> Parser<'a> { Keyword::REPLACE, ])?, ))) + } else if self.parse_keyword(Keyword::INVISIBLE) { + Ok(Some(ColumnOption::Invisible)) } else { Ok(None) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 720c1e492..f46abc7d9 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -17138,7 +17138,7 @@ fn test_parse_semantic_view_table_factor() { } let ast_sql = r#"SELECT * FROM SEMANTIC_VIEW( - my_model + my_model DIMENSIONS DATE_PART('year', date_col), region_name METRICS orders.revenue, orders.count WHERE active = true @@ -17193,3 +17193,56 @@ fn parse_adjacent_string_literal_concatenation() { let sql = "SELECT * FROM t WHERE col = 'Hello' \n ' ' \t 'World!'"; dialects.one_statement_parses_to(sql, r"SELECT * FROM t WHERE col = 'Hello World!'"); } + +#[test] +fn parse_invisible_column() { + let sql = r#"CREATE TABLE t (foo INT, bar INT INVISIBLE)"#; + let stmt = verified_stmt(sql); + match stmt { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ + ColumnDef { + name: "foo".into(), + data_type: DataType::Int(None), + options: vec![] + }, + ColumnDef { + name: "bar".into(), + data_type: DataType::Int(None), + options: vec![ColumnOptionDef { + name: None, + option: ColumnOption::Invisible + }] + } + ] + ); + } + _ => panic!("Unexpected statement {stmt}"), + } + + let sql = r#"ALTER TABLE t ADD COLUMN bar INT INVISIBLE"#; + let stmt = verified_stmt(sql); + match stmt { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::AddColumn { + column_keyword: true, + if_not_exists: false, + column_def: ColumnDef { + name: "bar".into(), + data_type: DataType::Int(None), + options: vec![ColumnOptionDef { + name: None, + option: ColumnOption::Invisible + }] + }, + column_position: None + }] + ); + } + _ => panic!("Unexpected statement {stmt}"), + } +} From 54a24e76a928c9b79a0b855fb4c6cbdc543b4a6d Mon Sep 17 00:00:00 2001 From: Ophir LOJKINE Date: Tue, 23 Sep 2025 13:34:24 +0200 Subject: [PATCH 265/286] Link to actual change logs in CHANGELOG.md (#2040) --- CHANGELOG.md | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index a5511a053..b0c3b5130 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -28,9 +28,4 @@ technically be breaking and thus will result in a `0.(N+1)` version. - Unreleased: Check https://github.com/sqlparser-rs/sqlparser-rs/commits/main for undocumented changes. -- `0.56.0`: [changelog/0.56.0.md](changelog/0.56.0.md) -- `0.55.0`: [changelog/0.55.0.md](changelog/0.55.0.md) -- `0.54.0`: [changelog/0.54.0.md](changelog/0.54.0.md) -- `0.53.0`: [changelog/0.53.0.md](changelog/0.53.0.md) -- `0.52.0`: [changelog/0.52.0.md](changelog/0.52.0.md) -- `0.51.0` and earlier: [changelog/0.51.0-pre.md](changelog/0.51.0-pre.md) +- Past releases: See https://github.com/apache/datafusion-sqlparser-rs/tree/main/changelog From a43083897468946d03b219f6bba31389e4bb4543 Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Thu, 25 Sep 2025 21:59:11 +0200 Subject: [PATCH 266/286] Snowflake: ALTER USER and KeyValueOptions Refactoring (#2035) --- src/ast/dml.rs | 7 +- src/ast/helpers/key_value_options.rs | 45 +++-- src/ast/helpers/stmt_create_database.rs | 2 +- src/ast/mod.rs | 217 +++++++++++++++++++++ src/ast/spans.rs | 1 + src/dialect/snowflake.rs | 25 ++- src/keywords.rs | 12 ++ src/parser/alter.rs | 192 ++++++++++++++++++- src/parser/mod.rs | 85 ++++++--- tests/sqlparser_common.rs | 239 ++++++++++++++++++++++-- tests/sqlparser_snowflake.rs | 119 ++++++------ 11 files changed, 809 insertions(+), 135 deletions(-) diff --git a/src/ast/dml.rs b/src/ast/dml.rs index 63d6b86c7..e4d99bcfc 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -16,12 +16,7 @@ // under the License. #[cfg(not(feature = "std"))] -use alloc::{ - boxed::Box, - format, - string::{String, ToString}, - vec::Vec, -}; +use alloc::{boxed::Box, format, string::ToString, vec::Vec}; use core::fmt::{self, Display}; #[cfg(feature = "serde")] diff --git a/src/ast/helpers/key_value_options.rs b/src/ast/helpers/key_value_options.rs index 7f1bb0fdb..745c3a65a 100644 --- a/src/ast/helpers/key_value_options.rs +++ b/src/ast/helpers/key_value_options.rs @@ -19,9 +19,7 @@ //! See [this page](https://docs.snowflake.com/en/sql-reference/commands-data-loading) for more details. #[cfg(not(feature = "std"))] -use alloc::string::String; -#[cfg(not(feature = "std"))] -use alloc::vec::Vec; +use alloc::{boxed::Box, string::String, vec::Vec}; use core::fmt; use core::fmt::Formatter; @@ -31,7 +29,7 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use crate::ast::display_separated; +use crate::ast::{display_comma_separated, display_separated, Value}; #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -52,20 +50,23 @@ pub enum KeyValueOptionsDelimiter { #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub enum KeyValueOptionType { - STRING, - BOOLEAN, - ENUM, - NUMBER, +pub struct KeyValueOption { + pub option_name: String, + pub option_value: KeyValueOptionKind, } +/// An option can have a single value, multiple values or a nested list of values. +/// +/// A value can be numeric, boolean, etc. Enum-style values are represented +/// as Value::Placeholder. For example: MFA_METHOD=SMS will be represented as +/// `Value::Placeholder("SMS".to_string)`. #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub struct KeyValueOption { - pub option_name: String, - pub option_type: KeyValueOptionType, - pub value: String, +pub enum KeyValueOptionKind { + Single(Value), + Multi(Vec), + KeyValueOptions(Box), } impl fmt::Display for KeyValueOptions { @@ -80,12 +81,20 @@ impl fmt::Display for KeyValueOptions { impl fmt::Display for KeyValueOption { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - match self.option_type { - KeyValueOptionType::STRING => { - write!(f, "{}='{}'", self.option_name, self.value)?; + match &self.option_value { + KeyValueOptionKind::Single(value) => { + write!(f, "{}={value}", self.option_name)?; + } + KeyValueOptionKind::Multi(values) => { + write!( + f, + "{}=({})", + self.option_name, + display_comma_separated(values) + )?; } - KeyValueOptionType::ENUM | KeyValueOptionType::BOOLEAN | KeyValueOptionType::NUMBER => { - write!(f, "{}={}", self.option_name, self.value)?; + KeyValueOptionKind::KeyValueOptions(options) => { + write!(f, "{}=({options})", self.option_name)?; } } Ok(()) diff --git a/src/ast/helpers/stmt_create_database.rs b/src/ast/helpers/stmt_create_database.rs index 94997bfa5..58a7b0906 100644 --- a/src/ast/helpers/stmt_create_database.rs +++ b/src/ast/helpers/stmt_create_database.rs @@ -16,7 +16,7 @@ // under the License. #[cfg(not(feature = "std"))] -use alloc::{boxed::Box, format, string::String, vec, vec::Vec}; +use alloc::{format, string::String, vec::Vec}; #[cfg(feature = "serde")] use serde::{Deserialize, Serialize}; diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 8df636f8e..4c1743feb 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -4310,6 +4310,11 @@ pub enum Statement { /// ``` /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/create-user) CreateUser(CreateUser), + /// ```sql + /// ALTER USER \[ IF EXISTS \] \[ \] + /// ``` + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/alter-user) + AlterUser(AlterUser), /// Re-sorts rows and reclaims space in either a specified table or all tables in the current database /// /// ```sql @@ -6183,6 +6188,7 @@ impl fmt::Display for Statement { Statement::CreateUser(s) => write!(f, "{s}"), Statement::AlterSchema(s) => write!(f, "{s}"), Statement::Vacuum(s) => write!(f, "{s}"), + Statement::AlterUser(s) => write!(f, "{s}"), } } } @@ -10558,6 +10564,217 @@ impl fmt::Display for CreateUser { } } +/// Modifies the properties of a user +/// +/// Syntax: +/// ```sql +/// ALTER USER [ IF EXISTS ] [ ] [ OPTIONS ] +/// ``` +/// +/// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/alter-user) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterUser { + pub if_exists: bool, + pub name: Ident, + /// The following fields are Snowflake-specific: + pub rename_to: Option, + pub reset_password: bool, + pub abort_all_queries: bool, + pub add_role_delegation: Option, + pub remove_role_delegation: Option, + pub enroll_mfa: bool, + pub set_default_mfa_method: Option, + pub remove_mfa_method: Option, + pub modify_mfa_method: Option, + pub add_mfa_method_otp: Option, + pub set_policy: Option, + pub unset_policy: Option, + pub set_tag: KeyValueOptions, + pub unset_tag: Vec, + pub set_props: KeyValueOptions, + pub unset_props: Vec, +} + +/// ```sql +/// ALTER USER [ IF EXISTS ] [ ] ADD DELEGATED AUTHORIZATION OF ROLE TO SECURITY INTEGRATION +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterUserAddRoleDelegation { + pub role: Ident, + pub integration: Ident, +} + +/// ```sql +/// ALTER USER [ IF EXISTS ] [ ] REMOVE DELEGATED { AUTHORIZATION OF ROLE | AUTHORIZATIONS } FROM SECURITY INTEGRATION +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterUserRemoveRoleDelegation { + pub role: Option, + pub integration: Ident, +} + +/// ```sql +/// ADD MFA METHOD OTP [ COUNT = number ] +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterUserAddMfaMethodOtp { + pub count: Option, +} + +/// ```sql +/// ALTER USER [ IF EXISTS ] [ ] MODIFY MFA METHOD SET COMMENT = '' +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterUserModifyMfaMethod { + pub method: MfaMethodKind, + pub comment: String, +} + +/// Types of MFA methods +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum MfaMethodKind { + PassKey, + Totp, + Duo, +} + +impl fmt::Display for MfaMethodKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + MfaMethodKind::PassKey => write!(f, "PASSKEY"), + MfaMethodKind::Totp => write!(f, "TOTP"), + MfaMethodKind::Duo => write!(f, "DUO"), + } + } +} + +/// ```sql +/// ALTER USER [ IF EXISTS ] [ ] SET { AUTHENTICATION | PASSWORD | SESSION } POLICY +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterUserSetPolicy { + pub policy_kind: UserPolicyKind, + pub policy: Ident, +} + +/// Types of user-based policies +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum UserPolicyKind { + Authentication, + Password, + Session, +} + +impl fmt::Display for UserPolicyKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + UserPolicyKind::Authentication => write!(f, "AUTHENTICATION"), + UserPolicyKind::Password => write!(f, "PASSWORD"), + UserPolicyKind::Session => write!(f, "SESSION"), + } + } +} + +impl fmt::Display for AlterUser { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "ALTER")?; + write!(f, " USER")?; + if self.if_exists { + write!(f, " IF EXISTS")?; + } + write!(f, " {}", self.name)?; + if let Some(new_name) = &self.rename_to { + write!(f, " RENAME TO {new_name}")?; + } + if self.reset_password { + write!(f, " RESET PASSWORD")?; + } + if self.abort_all_queries { + write!(f, " ABORT ALL QUERIES")?; + } + if let Some(role_delegation) = &self.add_role_delegation { + let role = &role_delegation.role; + let integration = &role_delegation.integration; + write!( + f, + " ADD DELEGATED AUTHORIZATION OF ROLE {role} TO SECURITY INTEGRATION {integration}" + )?; + } + if let Some(role_delegation) = &self.remove_role_delegation { + write!(f, " REMOVE DELEGATED")?; + match &role_delegation.role { + Some(role) => write!(f, " AUTHORIZATION OF ROLE {role}")?, + None => write!(f, " AUTHORIZATIONS")?, + } + let integration = &role_delegation.integration; + write!(f, " FROM SECURITY INTEGRATION {integration}")?; + } + if self.enroll_mfa { + write!(f, " ENROLL MFA")?; + } + if let Some(method) = &self.set_default_mfa_method { + write!(f, " SET DEFAULT_MFA_METHOD {method}")? + } + if let Some(method) = &self.remove_mfa_method { + write!(f, " REMOVE MFA METHOD {method}")?; + } + if let Some(modify) = &self.modify_mfa_method { + let method = &modify.method; + let comment = &modify.comment; + write!( + f, + " MODIFY MFA METHOD {method} SET COMMENT '{}'", + value::escape_single_quote_string(comment) + )?; + } + if let Some(add_mfa_method_otp) = &self.add_mfa_method_otp { + write!(f, " ADD MFA METHOD OTP")?; + if let Some(count) = &add_mfa_method_otp.count { + write!(f, " COUNT = {count}")?; + } + } + if let Some(policy) = &self.set_policy { + let policy_kind = &policy.policy_kind; + let name = &policy.policy; + write!(f, " SET {policy_kind} POLICY {name}")?; + } + if let Some(policy_kind) = &self.unset_policy { + write!(f, " UNSET {policy_kind} POLICY")?; + } + if !self.set_tag.options.is_empty() { + write!(f, " SET TAG {}", self.set_tag)?; + } + if !self.unset_tag.is_empty() { + write!(f, " UNSET TAG {}", display_comma_separated(&self.unset_tag))?; + } + let has_props = !self.set_props.options.is_empty(); + if has_props { + write!(f, " SET")?; + write!(f, " {}", &self.set_props)?; + } + if !self.unset_props.is_empty() { + write!(f, " UNSET {}", display_comma_separated(&self.unset_props))?; + } + Ok(()) + } +} + /// Specifies how to create a new table based on an existing table's schema. /// '''sql /// CREATE TABLE new LIKE old ... diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 5913fe164..4c53e55ce 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -555,6 +555,7 @@ impl Spanned for Statement { Statement::CreateUser(..) => Span::empty(), Statement::AlterSchema(s) => s.span(), Statement::Vacuum(..) => Span::empty(), + Statement::AlterUser(..) => Span::empty(), } } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 3bb36010b..825fd45f0 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -18,7 +18,7 @@ #[cfg(not(feature = "std"))] use crate::alloc::string::ToString; use crate::ast::helpers::key_value_options::{ - KeyValueOption, KeyValueOptionType, KeyValueOptions, KeyValueOptionsDelimiter, + KeyValueOption, KeyValueOptionKind, KeyValueOptions, KeyValueOptionsDelimiter, }; use crate::ast::helpers::stmt_create_database::CreateDatabaseBuilder; use crate::ast::helpers::stmt_create_table::CreateTableBuilder; @@ -30,7 +30,7 @@ use crate::ast::{ CopyIntoSnowflakeKind, CreateTableLikeKind, DollarQuotedString, Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, InitializeKind, ObjectName, ObjectNamePart, RefreshModeKind, RowAccessPolicy, ShowObjects, - SqlOption, Statement, StorageSerializationPolicy, TagsColumnOption, WrappedCollection, + SqlOption, Statement, StorageSerializationPolicy, TagsColumnOption, Value, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -1004,19 +1004,19 @@ pub fn parse_create_stage( // [ directoryTableParams ] if parser.parse_keyword(Keyword::DIRECTORY) { parser.expect_token(&Token::Eq)?; - directory_table_params = parser.parse_key_value_options(true, &[])?; + directory_table_params = parser.parse_key_value_options(true, &[])?.options; } // [ file_format] if parser.parse_keyword(Keyword::FILE_FORMAT) { parser.expect_token(&Token::Eq)?; - file_format = parser.parse_key_value_options(true, &[])?; + file_format = parser.parse_key_value_options(true, &[])?.options; } // [ copy_options ] if parser.parse_keyword(Keyword::COPY_OPTIONS) { parser.expect_token(&Token::Eq)?; - copy_options = parser.parse_key_value_options(true, &[])?; + copy_options = parser.parse_key_value_options(true, &[])?.options; } // [ comment ] @@ -1182,7 +1182,7 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { // FILE_FORMAT if parser.parse_keyword(Keyword::FILE_FORMAT) { parser.expect_token(&Token::Eq)?; - file_format = parser.parse_key_value_options(true, &[])?; + file_format = parser.parse_key_value_options(true, &[])?.options; // PARTITION BY } else if parser.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) { partition = Some(Box::new(parser.parse_expr()?)) @@ -1220,14 +1220,14 @@ pub fn parse_copy_into(parser: &mut Parser) -> Result { // COPY OPTIONS } else if parser.parse_keyword(Keyword::COPY_OPTIONS) { parser.expect_token(&Token::Eq)?; - copy_options = parser.parse_key_value_options(true, &[])?; + copy_options = parser.parse_key_value_options(true, &[])?.options; } else { match parser.next_token().token { Token::SemiColon | Token::EOF => break, Token::Comma => continue, // In `COPY INTO ` the copy options do not have a shared key // like in `COPY INTO
` - Token::Word(key) => copy_options.push(parser.parse_key_value_option(key)?), + Token::Word(key) => copy_options.push(parser.parse_key_value_option(&key)?), _ => return parser.expected("another copy option, ; or EOF'", parser.peek_token()), } } @@ -1387,7 +1387,7 @@ fn parse_stage_params(parser: &mut Parser) -> Result Result { parser.advance_token(); if set { - let option = parser.parse_key_value_option(key)?; + let option = parser.parse_key_value_option(&key)?; options.push(option); } else { options.push(KeyValueOption { option_name: key.value, - option_type: KeyValueOptionType::STRING, - value: empty(), + option_value: KeyValueOptionKind::Single(Value::Placeholder(empty())), }); } } diff --git a/src/keywords.rs b/src/keywords.rs index e0590b69d..3c9855222 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -119,6 +119,7 @@ define_keywords!( AUDIT, AUTHENTICATION, AUTHORIZATION, + AUTHORIZATIONS, AUTO, AUTOEXTEND_SIZE, AUTOINCREMENT, @@ -279,6 +280,8 @@ define_keywords!( DEFAULT, DEFAULTS, DEFAULT_DDL_COLLATION, + DEFAULT_MFA_METHOD, + DEFAULT_SECONDARY_ROLES, DEFERRABLE, DEFERRED, DEFINE, @@ -286,6 +289,7 @@ define_keywords!( DEFINER, DELAYED, DELAY_KEY_WRITE, + DELEGATED, DELETE, DELIMITED, DELIMITER, @@ -314,6 +318,7 @@ define_keywords!( DOY, DROP, DRY, + DUO, DUPLICATE, DYNAMIC, EACH, @@ -336,6 +341,7 @@ define_keywords!( ENFORCED, ENGINE, ENGINE_ATTRIBUTE, + ENROLL, ENUM, ENUM16, ENUM8, @@ -586,6 +592,7 @@ define_keywords!( METHOD, METRIC, METRICS, + MFA, MICROSECOND, MICROSECONDS, MILLENIUM, @@ -685,6 +692,7 @@ define_keywords!( ORDINALITY, ORGANIZATION, OTHER, + OTP, OUT, OUTER, OUTPUT, @@ -709,6 +717,7 @@ define_keywords!( PARTITIONED, PARTITIONS, PASSING, + PASSKEY, PASSWORD, PAST, PATH, @@ -753,6 +762,7 @@ define_keywords!( PURGE, QUALIFY, QUARTER, + QUERIES, QUERY, QUOTE, RAISE, @@ -969,6 +979,7 @@ define_keywords!( TO, TOP, TOTALS, + TOTP, TRACE, TRAILING, TRANSACTION, @@ -1067,6 +1078,7 @@ define_keywords!( WITHOUT, WITHOUT_ARRAY_WRAPPER, WORK, + WORKLOAD_IDENTITY, WRAPPER, WRITE, XML, diff --git a/src/parser/alter.rs b/src/parser/alter.rs index bff462ee0..b3e3c99e6 100644 --- a/src/parser/alter.rs +++ b/src/parser/alter.rs @@ -13,13 +13,16 @@ //! SQL Parser for ALTER #[cfg(not(feature = "std"))] -use alloc::vec; +use alloc::{string::ToString, vec}; use super::{Parser, ParserError}; use crate::{ ast::{ - AlterConnectorOwner, AlterPolicyOperation, AlterRoleOperation, Expr, Password, ResetConfig, - RoleOption, SetConfigValue, Statement, + helpers::key_value_options::{KeyValueOptions, KeyValueOptionsDelimiter}, + AlterConnectorOwner, AlterPolicyOperation, AlterRoleOperation, AlterUser, + AlterUserAddMfaMethodOtp, AlterUserAddRoleDelegation, AlterUserModifyMfaMethod, + AlterUserRemoveRoleDelegation, AlterUserSetPolicy, Expr, MfaMethodKind, Password, + ResetConfig, RoleOption, SetConfigValue, Statement, UserPolicyKind, }, dialect::{MsSqlDialect, PostgreSqlDialect}, keywords::Keyword, @@ -140,6 +143,189 @@ impl Parser<'_> { }) } + /// Parse an `ALTER USER` statement + /// ```sql + /// ALTER USER [ IF EXISTS ] [ ] [ OPTIONS ] + /// ``` + pub fn parse_alter_user(&mut self) -> Result { + let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); + let name = self.parse_identifier()?; + let rename_to = if self.parse_keywords(&[Keyword::RENAME, Keyword::TO]) { + Some(self.parse_identifier()?) + } else { + None + }; + let reset_password = self.parse_keywords(&[Keyword::RESET, Keyword::PASSWORD]); + let abort_all_queries = + self.parse_keywords(&[Keyword::ABORT, Keyword::ALL, Keyword::QUERIES]); + let add_role_delegation = if self.parse_keywords(&[ + Keyword::ADD, + Keyword::DELEGATED, + Keyword::AUTHORIZATION, + Keyword::OF, + Keyword::ROLE, + ]) { + let role = self.parse_identifier()?; + self.expect_keywords(&[Keyword::TO, Keyword::SECURITY, Keyword::INTEGRATION])?; + let integration = self.parse_identifier()?; + Some(AlterUserAddRoleDelegation { role, integration }) + } else { + None + }; + let remove_role_delegation = if self.parse_keywords(&[Keyword::REMOVE, Keyword::DELEGATED]) + { + let role = if self.parse_keywords(&[Keyword::AUTHORIZATION, Keyword::OF, Keyword::ROLE]) + { + Some(self.parse_identifier()?) + } else if self.parse_keyword(Keyword::AUTHORIZATIONS) { + None + } else { + return self.expected( + "REMOVE DELEGATED AUTHORIZATION OF ROLE | REMOVE DELEGATED AUTHORIZATIONS", + self.peek_token(), + ); + }; + self.expect_keywords(&[Keyword::FROM, Keyword::SECURITY, Keyword::INTEGRATION])?; + let integration = self.parse_identifier()?; + Some(AlterUserRemoveRoleDelegation { role, integration }) + } else { + None + }; + let enroll_mfa = self.parse_keywords(&[Keyword::ENROLL, Keyword::MFA]); + let set_default_mfa_method = + if self.parse_keywords(&[Keyword::SET, Keyword::DEFAULT_MFA_METHOD]) { + Some(self.parse_mfa_method()?) + } else { + None + }; + let remove_mfa_method = + if self.parse_keywords(&[Keyword::REMOVE, Keyword::MFA, Keyword::METHOD]) { + Some(self.parse_mfa_method()?) + } else { + None + }; + let modify_mfa_method = + if self.parse_keywords(&[Keyword::MODIFY, Keyword::MFA, Keyword::METHOD]) { + let method = self.parse_mfa_method()?; + self.expect_keywords(&[Keyword::SET, Keyword::COMMENT])?; + let comment = self.parse_literal_string()?; + Some(AlterUserModifyMfaMethod { method, comment }) + } else { + None + }; + let add_mfa_method_otp = + if self.parse_keywords(&[Keyword::ADD, Keyword::MFA, Keyword::METHOD, Keyword::OTP]) { + let count = if self.parse_keyword(Keyword::COUNT) { + self.expect_token(&Token::Eq)?; + Some(self.parse_value()?.into()) + } else { + None + }; + Some(AlterUserAddMfaMethodOtp { count }) + } else { + None + }; + let set_policy = + if self.parse_keywords(&[Keyword::SET, Keyword::AUTHENTICATION, Keyword::POLICY]) { + Some(AlterUserSetPolicy { + policy_kind: UserPolicyKind::Authentication, + policy: self.parse_identifier()?, + }) + } else if self.parse_keywords(&[Keyword::SET, Keyword::PASSWORD, Keyword::POLICY]) { + Some(AlterUserSetPolicy { + policy_kind: UserPolicyKind::Password, + policy: self.parse_identifier()?, + }) + } else if self.parse_keywords(&[Keyword::SET, Keyword::SESSION, Keyword::POLICY]) { + Some(AlterUserSetPolicy { + policy_kind: UserPolicyKind::Session, + policy: self.parse_identifier()?, + }) + } else { + None + }; + + let unset_policy = + if self.parse_keywords(&[Keyword::UNSET, Keyword::AUTHENTICATION, Keyword::POLICY]) { + Some(UserPolicyKind::Authentication) + } else if self.parse_keywords(&[Keyword::UNSET, Keyword::PASSWORD, Keyword::POLICY]) { + Some(UserPolicyKind::Password) + } else if self.parse_keywords(&[Keyword::UNSET, Keyword::SESSION, Keyword::POLICY]) { + Some(UserPolicyKind::Session) + } else { + None + }; + + let set_tag = if self.parse_keywords(&[Keyword::SET, Keyword::TAG]) { + self.parse_key_value_options(false, &[])? + } else { + KeyValueOptions { + delimiter: KeyValueOptionsDelimiter::Comma, + options: vec![], + } + }; + + let unset_tag = if self.parse_keywords(&[Keyword::UNSET, Keyword::TAG]) { + self.parse_comma_separated(Parser::parse_identifier)? + .iter() + .map(|i| i.to_string()) + .collect() + } else { + vec![] + }; + + let set_props = if self.parse_keyword(Keyword::SET) { + self.parse_key_value_options(false, &[])? + } else { + KeyValueOptions { + delimiter: KeyValueOptionsDelimiter::Comma, + options: vec![], + } + }; + + let unset_props = if self.parse_keyword(Keyword::UNSET) { + self.parse_comma_separated(Parser::parse_identifier)? + .iter() + .map(|i| i.to_string()) + .collect() + } else { + vec![] + }; + + Ok(Statement::AlterUser(AlterUser { + if_exists, + name, + rename_to, + reset_password, + abort_all_queries, + add_role_delegation, + remove_role_delegation, + enroll_mfa, + set_default_mfa_method, + remove_mfa_method, + modify_mfa_method, + add_mfa_method_otp, + set_policy, + unset_policy, + set_tag, + unset_tag, + set_props, + unset_props, + })) + } + + fn parse_mfa_method(&mut self) -> Result { + if self.parse_keyword(Keyword::PASSKEY) { + Ok(MfaMethodKind::PassKey) + } else if self.parse_keyword(Keyword::TOTP) { + Ok(MfaMethodKind::Totp) + } else if self.parse_keyword(Keyword::DUO) { + Ok(MfaMethodKind::Duo) + } else { + self.expected("PASSKEY, TOTP or DUO", self.peek_token()) + } + } + fn parse_mssql_alter_role(&mut self) -> Result { let role_name = self.parse_identifier()?; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f47d4b9b5..66089be7c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -34,7 +34,7 @@ use IsOptional::*; use crate::ast::helpers::{ key_value_options::{ - KeyValueOption, KeyValueOptionType, KeyValueOptions, KeyValueOptionsDelimiter, + KeyValueOption, KeyValueOptionKind, KeyValueOptions, KeyValueOptionsDelimiter, }, stmt_create_table::{CreateTableBuilder, CreateTableConfiguration}, }; @@ -4796,10 +4796,12 @@ impl<'a> Parser<'a> { fn parse_create_user(&mut self, or_replace: bool) -> Result { let if_not_exists = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); let name = self.parse_identifier()?; - let options = self.parse_key_value_options(false, &[Keyword::WITH, Keyword::TAG])?; + let options = self + .parse_key_value_options(false, &[Keyword::WITH, Keyword::TAG])? + .options; let with_tags = self.parse_keyword(Keyword::WITH); let tags = if self.parse_keyword(Keyword::TAG) { - self.parse_key_value_options(true, &[])? + self.parse_key_value_options(true, &[])?.options } else { vec![] }; @@ -9277,6 +9279,7 @@ impl<'a> Parser<'a> { Keyword::CONNECTOR, Keyword::ICEBERG, Keyword::SCHEMA, + Keyword::USER, ])?; match object_type { Keyword::SCHEMA => { @@ -9312,6 +9315,7 @@ impl<'a> Parser<'a> { Keyword::ROLE => self.parse_alter_role(), Keyword::POLICY => self.parse_alter_policy(), Keyword::CONNECTOR => self.parse_alter_connector(), + Keyword::USER => self.parse_alter_user(), // unreachable because expect_one_of_keywords used above _ => unreachable!(), } @@ -17488,8 +17492,9 @@ impl<'a> Parser<'a> { &mut self, parenthesized: bool, end_words: &[Keyword], - ) -> Result, ParserError> { + ) -> Result { let mut options: Vec = Vec::new(); + let mut delimiter = KeyValueOptionsDelimiter::Space; if parenthesized { self.expect_token(&Token::LParen)?; } @@ -17503,9 +17508,12 @@ impl<'a> Parser<'a> { } } Token::EOF => break, - Token::Comma => continue, + Token::Comma => { + delimiter = KeyValueOptionsDelimiter::Comma; + continue; + } Token::Word(w) if !end_words.contains(&w.keyword) => { - options.push(self.parse_key_value_option(w)?) + options.push(self.parse_key_value_option(&w)?) } Token::Word(w) if end_words.contains(&w.keyword) => { self.prev_token(); @@ -17514,40 +17522,67 @@ impl<'a> Parser<'a> { _ => return self.expected("another option, EOF, Comma or ')'", self.peek_token()), }; } - Ok(options) + + Ok(KeyValueOptions { delimiter, options }) } /// Parses a `KEY = VALUE` construct based on the specified key pub(crate) fn parse_key_value_option( &mut self, - key: Word, + key: &Word, ) -> Result { self.expect_token(&Token::Eq)?; - match self.next_token().token { - Token::SingleQuotedString(value) => Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::STRING, - value, + match self.peek_token().token { + Token::SingleQuotedString(_) => Ok(KeyValueOption { + option_name: key.value.clone(), + option_value: KeyValueOptionKind::Single(self.parse_value()?.into()), }), Token::Word(word) if word.keyword == Keyword::TRUE || word.keyword == Keyword::FALSE => { Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::BOOLEAN, - value: word.value.to_uppercase(), + option_name: key.value.clone(), + option_value: KeyValueOptionKind::Single(self.parse_value()?.into()), }) } - Token::Word(word) => Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::ENUM, - value: word.value, - }), - Token::Number(n, _) => Ok(KeyValueOption { - option_name: key.value, - option_type: KeyValueOptionType::NUMBER, - value: n, + Token::Number(..) => Ok(KeyValueOption { + option_name: key.value.clone(), + option_value: KeyValueOptionKind::Single(self.parse_value()?.into()), }), + Token::Word(word) => { + self.next_token(); + Ok(KeyValueOption { + option_name: key.value.clone(), + option_value: KeyValueOptionKind::Single(Value::Placeholder( + word.value.clone(), + )), + }) + } + Token::LParen => { + // Can be a list of values or a list of key value properties. + // Try to parse a list of values and if that fails, try to parse + // a list of key-value properties. + match self.maybe_parse(|parser| { + parser.expect_token(&Token::LParen)?; + let values = parser.parse_comma_separated0(|p| p.parse_value(), Token::RParen); + parser.expect_token(&Token::RParen)?; + values + })? { + Some(values) => { + let values = values.into_iter().map(|v| v.value).collect(); + Ok(KeyValueOption { + option_name: key.value.clone(), + option_value: KeyValueOptionKind::Multi(values), + }) + } + None => Ok(KeyValueOption { + option_name: key.value.clone(), + option_value: KeyValueOptionKind::KeyValueOptions(Box::new( + self.parse_key_value_options(true, &[])?, + )), + }), + } + } _ => self.expected("expected option value", self.peek_token()), } } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index f46abc7d9..b9434581a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16761,11 +16761,13 @@ fn parse_create_user() { verified_stmt("CREATE OR REPLACE USER u1"); verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1"); verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret'"); - verified_stmt( + let dialects = all_dialects_where(|d| d.supports_boolean_literals()); + dialects.one_statement_parses_to( "CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=TRUE", + "CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=true", ); - verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=TRUE TYPE=SERVICE TAG (t1='v1')"); - let create = verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=TRUE TYPE=SERVICE WITH TAG (t1='v1', t2='v2')"); + dialects.verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=true TYPE=SERVICE TAG (t1='v1')"); + let create = dialects.verified_stmt("CREATE OR REPLACE USER IF NOT EXISTS u1 PASSWORD='secret' MUST_CHANGE_PASSWORD=false TYPE=SERVICE WITH TAG (t1='v1', t2='v2')"); match create { Statement::CreateUser(stmt) => { assert_eq!(stmt.name, Ident::new("u1")); @@ -16778,18 +16780,19 @@ fn parse_create_user() { options: vec![ KeyValueOption { option_name: "PASSWORD".to_string(), - value: "secret".to_string(), - option_type: KeyValueOptionType::STRING + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "secret".to_string() + )), }, KeyValueOption { option_name: "MUST_CHANGE_PASSWORD".to_string(), - value: "TRUE".to_string(), - option_type: KeyValueOptionType::BOOLEAN + option_value: KeyValueOptionKind::Single(Value::Boolean(false)), }, KeyValueOption { option_name: "TYPE".to_string(), - value: "SERVICE".to_string(), - option_type: KeyValueOptionType::ENUM + option_value: KeyValueOptionKind::Single(Value::Placeholder( + "SERVICE".to_string() + )), }, ], }, @@ -16802,13 +16805,15 @@ fn parse_create_user() { options: vec![ KeyValueOption { option_name: "t1".to_string(), - value: "v1".to_string(), - option_type: KeyValueOptionType::STRING + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "v1".to_string() + )), }, KeyValueOption { option_name: "t2".to_string(), - value: "v2".to_string(), - option_type: KeyValueOptionType::STRING + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "v2".to_string() + )), }, ] } @@ -17246,3 +17251,211 @@ fn parse_invisible_column() { _ => panic!("Unexpected statement {stmt}"), } } + +#[test] +fn test_parse_alter_user() { + verified_stmt("ALTER USER u1"); + verified_stmt("ALTER USER IF EXISTS u1"); + let stmt = verified_stmt("ALTER USER IF EXISTS u1 RENAME TO u2"); + match stmt { + Statement::AlterUser(alter) => { + assert!(alter.if_exists); + assert_eq!(alter.name, Ident::new("u1")); + assert_eq!(alter.rename_to, Some(Ident::new("u2"))); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER IF EXISTS u1 RESET PASSWORD"); + verified_stmt("ALTER USER IF EXISTS u1 ABORT ALL QUERIES"); + verified_stmt( + "ALTER USER IF EXISTS u1 ADD DELEGATED AUTHORIZATION OF ROLE r1 TO SECURITY INTEGRATION i1", + ); + verified_stmt("ALTER USER IF EXISTS u1 REMOVE DELEGATED AUTHORIZATION OF ROLE r1 FROM SECURITY INTEGRATION i1"); + verified_stmt( + "ALTER USER IF EXISTS u1 REMOVE DELEGATED AUTHORIZATIONS FROM SECURITY INTEGRATION i1", + ); + verified_stmt("ALTER USER IF EXISTS u1 ENROLL MFA"); + let stmt = verified_stmt("ALTER USER u1 SET DEFAULT_MFA_METHOD PASSKEY"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!(alter.set_default_mfa_method, Some(MfaMethodKind::PassKey)) + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 SET DEFAULT_MFA_METHOD TOTP"); + verified_stmt("ALTER USER u1 SET DEFAULT_MFA_METHOD DUO"); + let stmt = verified_stmt("ALTER USER u1 REMOVE MFA METHOD PASSKEY"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!(alter.remove_mfa_method, Some(MfaMethodKind::PassKey)) + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 REMOVE MFA METHOD TOTP"); + verified_stmt("ALTER USER u1 REMOVE MFA METHOD DUO"); + let stmt = verified_stmt("ALTER USER u1 MODIFY MFA METHOD PASSKEY SET COMMENT 'abc'"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!( + alter.modify_mfa_method, + Some(AlterUserModifyMfaMethod { + method: MfaMethodKind::PassKey, + comment: "abc".to_string() + }) + ); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 ADD MFA METHOD OTP"); + verified_stmt("ALTER USER u1 ADD MFA METHOD OTP COUNT = 8"); + + let stmt = verified_stmt("ALTER USER u1 SET AUTHENTICATION POLICY p1"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!( + alter.set_policy, + Some(AlterUserSetPolicy { + policy_kind: UserPolicyKind::Authentication, + policy: Ident::new("p1") + }) + ); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 SET PASSWORD POLICY p1"); + verified_stmt("ALTER USER u1 SET SESSION POLICY p1"); + let stmt = verified_stmt("ALTER USER u1 UNSET AUTHENTICATION POLICY"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!(alter.unset_policy, Some(UserPolicyKind::Authentication)); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 UNSET PASSWORD POLICY"); + verified_stmt("ALTER USER u1 UNSET SESSION POLICY"); + + let stmt = verified_stmt("ALTER USER u1 SET TAG k1='v1'"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!( + alter.set_tag.options, + vec![KeyValueOption { + option_name: "k1".to_string(), + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "v1".to_string() + )), + },] + ); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 SET TAG k1='v1', k2='v2'"); + let stmt = verified_stmt("ALTER USER u1 UNSET TAG k1"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!(alter.unset_tag, vec!["k1".to_string()]); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 UNSET TAG k1, k2, k3"); + + let dialects = all_dialects_where(|d| d.supports_boolean_literals()); + dialects.one_statement_parses_to( + "ALTER USER u1 SET PASSWORD='secret', MUST_CHANGE_PASSWORD=TRUE, MINS_TO_UNLOCK=10", + "ALTER USER u1 SET PASSWORD='secret', MUST_CHANGE_PASSWORD=true, MINS_TO_UNLOCK=10", + ); + + let stmt = dialects.verified_stmt( + "ALTER USER u1 SET PASSWORD='secret', MUST_CHANGE_PASSWORD=true, MINS_TO_UNLOCK=10", + ); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!( + alter.set_props, + KeyValueOptions { + delimiter: KeyValueOptionsDelimiter::Comma, + options: vec![ + KeyValueOption { + option_name: "PASSWORD".to_string(), + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "secret".to_string() + )), + }, + KeyValueOption { + option_name: "MUST_CHANGE_PASSWORD".to_string(), + option_value: KeyValueOptionKind::Single(Value::Boolean(true)), + }, + KeyValueOption { + option_name: "MINS_TO_UNLOCK".to_string(), + option_value: KeyValueOptionKind::Single(number("10")), + }, + ] + } + ); + } + _ => unreachable!(), + } + + let stmt = verified_stmt("ALTER USER u1 UNSET PASSWORD"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!(alter.unset_props, vec!["PASSWORD".to_string()]); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 UNSET PASSWORD, MUST_CHANGE_PASSWORD, MINS_TO_UNLOCK"); + + let stmt = verified_stmt("ALTER USER u1 SET DEFAULT_SECONDARY_ROLES=('ALL')"); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!( + alter.set_props.options, + vec![KeyValueOption { + option_name: "DEFAULT_SECONDARY_ROLES".to_string(), + option_value: KeyValueOptionKind::Multi(vec![Value::SingleQuotedString( + "ALL".to_string() + )]) + }] + ); + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 SET DEFAULT_SECONDARY_ROLES=()"); + verified_stmt("ALTER USER u1 SET DEFAULT_SECONDARY_ROLES=('R1', 'R2', 'R3')"); + verified_stmt("ALTER USER u1 SET PASSWORD='secret', DEFAULT_SECONDARY_ROLES=('ALL')"); + verified_stmt("ALTER USER u1 SET DEFAULT_SECONDARY_ROLES=('ALL'), PASSWORD='secret'"); + let stmt = verified_stmt( + "ALTER USER u1 SET WORKLOAD_IDENTITY=(TYPE=AWS, ARN='arn:aws:iam::123456789:r1/')", + ); + match stmt { + Statement::AlterUser(alter) => { + assert_eq!( + alter.set_props.options, + vec![KeyValueOption { + option_name: "WORKLOAD_IDENTITY".to_string(), + option_value: KeyValueOptionKind::KeyValueOptions(Box::new(KeyValueOptions { + delimiter: KeyValueOptionsDelimiter::Comma, + options: vec![ + KeyValueOption { + option_name: "TYPE".to_string(), + option_value: KeyValueOptionKind::Single(Value::Placeholder( + "AWS".to_string() + )), + }, + KeyValueOption { + option_name: "ARN".to_string(), + option_value: KeyValueOptionKind::Single( + Value::SingleQuotedString( + "arn:aws:iam::123456789:r1/".to_string() + ) + ), + }, + ] + })) + }] + ) + } + _ => unreachable!(), + } + verified_stmt("ALTER USER u1 SET DEFAULT_SECONDARY_ROLES=('ALL'), PASSWORD='secret', WORKLOAD_IDENTITY=(TYPE=AWS, ARN='arn:aws:iam::123456789:r1/')"); +} diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 7c9e5261b..e04bfaf5d 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -19,7 +19,7 @@ //! Test SQL syntax specific to Snowflake. The parser based on the //! generic dialect is also tested (on the inputs it can handle). -use sqlparser::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionType}; +use sqlparser::ast::helpers::key_value_options::{KeyValueOption, KeyValueOptionKind}; use sqlparser::ast::helpers::stmt_data_loading::{StageLoadSelectItem, StageLoadSelectItemKind}; use sqlparser::ast::*; use sqlparser::dialect::{Dialect, GenericDialect, SnowflakeDialect}; @@ -2116,23 +2116,27 @@ fn test_create_stage_with_stage_params() { ); assert!(stage_params.credentials.options.contains(&KeyValueOption { option_name: "AWS_KEY_ID".to_string(), - option_type: KeyValueOptionType::STRING, - value: "1a2b3c".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "1a2b3c".to_string() + )), })); assert!(stage_params.credentials.options.contains(&KeyValueOption { option_name: "AWS_SECRET_KEY".to_string(), - option_type: KeyValueOptionType::STRING, - value: "4x5y6z".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "4x5y6z".to_string() + )), })); assert!(stage_params.encryption.options.contains(&KeyValueOption { option_name: "MASTER_KEY".to_string(), - option_type: KeyValueOptionType::STRING, - value: "key".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "key".to_string() + )), })); assert!(stage_params.encryption.options.contains(&KeyValueOption { option_name: "TYPE".to_string(), - option_type: KeyValueOptionType::STRING, - value: "AWS_SSE_KMS".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "AWS_SSE_KMS".to_string() + )), })); } _ => unreachable!(), @@ -2146,7 +2150,7 @@ fn test_create_stage_with_directory_table_params() { let sql = concat!( "CREATE OR REPLACE STAGE my_ext_stage ", "URL='s3://load/files/' ", - "DIRECTORY=(ENABLE=TRUE REFRESH_ON_CREATE=FALSE NOTIFICATION_INTEGRATION='some-string')" + "DIRECTORY=(ENABLE=true REFRESH_ON_CREATE=false NOTIFICATION_INTEGRATION='some-string')" ); match snowflake().verified_stmt(sql) { @@ -2156,18 +2160,17 @@ fn test_create_stage_with_directory_table_params() { } => { assert!(directory_table_params.options.contains(&KeyValueOption { option_name: "ENABLE".to_string(), - option_type: KeyValueOptionType::BOOLEAN, - value: "TRUE".to_string() + option_value: KeyValueOptionKind::Single(Value::Boolean(true)), })); assert!(directory_table_params.options.contains(&KeyValueOption { option_name: "REFRESH_ON_CREATE".to_string(), - option_type: KeyValueOptionType::BOOLEAN, - value: "FALSE".to_string() + option_value: KeyValueOptionKind::Single(Value::Boolean(false)), })); assert!(directory_table_params.options.contains(&KeyValueOption { option_name: "NOTIFICATION_INTEGRATION".to_string(), - option_type: KeyValueOptionType::STRING, - value: "some-string".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "some-string".to_string() + )), })); } _ => unreachable!(), @@ -2187,18 +2190,17 @@ fn test_create_stage_with_file_format() { Statement::CreateStage { file_format, .. } => { assert!(file_format.options.contains(&KeyValueOption { option_name: "COMPRESSION".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "AUTO".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder("AUTO".to_string())), })); assert!(file_format.options.contains(&KeyValueOption { option_name: "BINARY_FORMAT".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "HEX".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder("HEX".to_string())), })); assert!(file_format.options.contains(&KeyValueOption { option_name: "ESCAPE".to_string(), - option_type: KeyValueOptionType::STRING, - value: r#"\\"#.to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + r#"\\"#.to_string() + )), })); } _ => unreachable!(), @@ -2214,19 +2216,19 @@ fn test_create_stage_with_copy_options() { let sql = concat!( "CREATE OR REPLACE STAGE my_ext_stage ", "URL='s3://load/files/' ", - "COPY_OPTIONS=(ON_ERROR=CONTINUE FORCE=TRUE)" + "COPY_OPTIONS=(ON_ERROR=CONTINUE FORCE=true)" ); match snowflake().verified_stmt(sql) { Statement::CreateStage { copy_options, .. } => { assert!(copy_options.options.contains(&KeyValueOption { option_name: "ON_ERROR".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "CONTINUE".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder( + "CONTINUE".to_string() + )), })); assert!(copy_options.options.contains(&KeyValueOption { option_name: "FORCE".to_string(), - option_type: KeyValueOptionType::BOOLEAN, - value: "TRUE".to_string() + option_value: KeyValueOptionKind::Single(Value::Boolean(true)), })); } _ => unreachable!(), @@ -2357,23 +2359,27 @@ fn test_copy_into_with_stage_params() { ); assert!(stage_params.credentials.options.contains(&KeyValueOption { option_name: "AWS_KEY_ID".to_string(), - option_type: KeyValueOptionType::STRING, - value: "1a2b3c".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "1a2b3c".to_string() + )), })); assert!(stage_params.credentials.options.contains(&KeyValueOption { option_name: "AWS_SECRET_KEY".to_string(), - option_type: KeyValueOptionType::STRING, - value: "4x5y6z".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "4x5y6z".to_string() + )), })); assert!(stage_params.encryption.options.contains(&KeyValueOption { option_name: "MASTER_KEY".to_string(), - option_type: KeyValueOptionType::STRING, - value: "key".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "key".to_string() + )), })); assert!(stage_params.encryption.options.contains(&KeyValueOption { option_name: "TYPE".to_string(), - option_type: KeyValueOptionType::STRING, - value: "AWS_SSE_KMS".to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + "AWS_SSE_KMS".to_string() + )), })); } _ => unreachable!(), @@ -2524,18 +2530,17 @@ fn test_copy_into_file_format() { Statement::CopyIntoSnowflake { file_format, .. } => { assert!(file_format.options.contains(&KeyValueOption { option_name: "COMPRESSION".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "AUTO".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder("AUTO".to_string())), })); assert!(file_format.options.contains(&KeyValueOption { option_name: "BINARY_FORMAT".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "HEX".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder("HEX".to_string())), })); assert!(file_format.options.contains(&KeyValueOption { option_name: "ESCAPE".to_string(), - option_type: KeyValueOptionType::STRING, - value: r#"\\"#.to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + r#"\\"#.to_string() + )), })); } _ => unreachable!(), @@ -2563,18 +2568,17 @@ fn test_copy_into_file_format() { Statement::CopyIntoSnowflake { file_format, .. } => { assert!(file_format.options.contains(&KeyValueOption { option_name: "COMPRESSION".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "AUTO".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder("AUTO".to_string())), })); assert!(file_format.options.contains(&KeyValueOption { option_name: "BINARY_FORMAT".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "HEX".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder("HEX".to_string())), })); assert!(file_format.options.contains(&KeyValueOption { option_name: "ESCAPE".to_string(), - option_type: KeyValueOptionType::STRING, - value: r#"\\"#.to_string() + option_value: KeyValueOptionKind::Single(Value::SingleQuotedString( + r#"\\"#.to_string() + )), })); } _ => unreachable!(), @@ -2588,20 +2592,20 @@ fn test_copy_into_copy_options() { "FROM 'gcs://mybucket/./../a.csv' ", "FILES = ('file1.json', 'file2.json') ", "PATTERN = '.*employees0[1-5].csv.gz' ", - "COPY_OPTIONS=(ON_ERROR=CONTINUE FORCE=TRUE)" + "COPY_OPTIONS=(ON_ERROR=CONTINUE FORCE=true)" ); match snowflake().verified_stmt(sql) { Statement::CopyIntoSnowflake { copy_options, .. } => { assert!(copy_options.options.contains(&KeyValueOption { option_name: "ON_ERROR".to_string(), - option_type: KeyValueOptionType::ENUM, - value: "CONTINUE".to_string() + option_value: KeyValueOptionKind::Single(Value::Placeholder( + "CONTINUE".to_string() + )), })); assert!(copy_options.options.contains(&KeyValueOption { option_name: "FORCE".to_string(), - option_type: KeyValueOptionType::BOOLEAN, - value: "TRUE".to_string() + option_value: KeyValueOptionKind::Single(Value::Boolean(true)), })); } _ => unreachable!(), @@ -3863,17 +3867,20 @@ fn test_alter_session() { "sql parser error: expected at least one option" ); - snowflake().verified_stmt("ALTER SESSION SET AUTOCOMMIT=TRUE"); - snowflake().verified_stmt("ALTER SESSION SET AUTOCOMMIT=FALSE QUERY_TAG='tag'"); + snowflake().one_statement_parses_to( + "ALTER SESSION SET AUTOCOMMIT=TRUE", + "ALTER SESSION SET AUTOCOMMIT=true", + ); + snowflake().verified_stmt("ALTER SESSION SET AUTOCOMMIT=false QUERY_TAG='tag'"); snowflake().verified_stmt("ALTER SESSION UNSET AUTOCOMMIT"); snowflake().verified_stmt("ALTER SESSION UNSET AUTOCOMMIT, QUERY_TAG"); snowflake().one_statement_parses_to( "ALTER SESSION SET A=false, B='tag';", - "ALTER SESSION SET A=FALSE B='tag'", + "ALTER SESSION SET A=false B='tag'", ); snowflake().one_statement_parses_to( "ALTER SESSION SET A=true \nB='tag'", - "ALTER SESSION SET A=TRUE B='tag'", + "ALTER SESSION SET A=true B='tag'", ); snowflake().one_statement_parses_to("ALTER SESSION UNSET a\nB", "ALTER SESSION UNSET a, B"); } From c0998832a2f018915ad4464db6aefa5b1e2f4447 Mon Sep 17 00:00:00 2001 From: Joey Hain Date: Fri, 26 Sep 2025 01:05:23 -0700 Subject: [PATCH 267/286] Correctly tokenize nested comments in Databricks, Clickhouse, and ANSI (#2044) --- src/dialect/ansi.rs | 5 ++ src/dialect/clickhouse.rs | 6 ++ src/dialect/databricks.rs | 5 ++ src/tokenizer.rs | 141 ++++++++++++++++++-------------------- 4 files changed, 81 insertions(+), 76 deletions(-) diff --git a/src/dialect/ansi.rs b/src/dialect/ansi.rs index 32ba7b32a..ec3c095be 100644 --- a/src/dialect/ansi.rs +++ b/src/dialect/ansi.rs @@ -33,4 +33,9 @@ impl Dialect for AnsiDialect { fn require_interval_qualifier(&self) -> bool { true } + + /// The SQL standard explicitly states that block comments nest. + fn supports_nested_comments(&self) -> bool { + true + } } diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs index f5e70c309..bdac1f57b 100644 --- a/src/dialect/clickhouse.rs +++ b/src/dialect/clickhouse.rs @@ -94,4 +94,10 @@ impl Dialect for ClickHouseDialect { fn supports_group_by_with_modifier(&self) -> bool { true } + + /// Supported since 2020. + /// See + fn supports_nested_comments(&self) -> bool { + true + } } diff --git a/src/dialect/databricks.rs b/src/dialect/databricks.rs index a3476b1b8..4bb8c8d51 100644 --- a/src/dialect/databricks.rs +++ b/src/dialect/databricks.rs @@ -64,4 +64,9 @@ impl Dialect for DatabricksDialect { fn supports_struct_literal(&self) -> bool { true } + + /// See + fn supports_nested_comments(&self) -> bool { + true + } } diff --git a/src/tokenizer.rs b/src/tokenizer.rs index 8382a5344..54a158c1f 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -2419,7 +2419,7 @@ mod tests { use crate::dialect::{ BigQueryDialect, ClickHouseDialect, HiveDialect, MsSqlDialect, MySqlDialect, SQLiteDialect, }; - use crate::test_utils::all_dialects_where; + use crate::test_utils::{all_dialects_except, all_dialects_where}; use core::fmt::Debug; #[test] @@ -3169,90 +3169,79 @@ mod tests { #[test] fn tokenize_nested_multiline_comment() { - let dialect = GenericDialect {}; - let test_cases = vec![ - ( - "0/*multi-line\n* \n/* comment \n /*comment*/*/ */ /comment*/1", - vec![ - Token::Number("0".to_string(), false), - Token::Whitespace(Whitespace::MultiLineComment( - "multi-line\n* \n/* comment \n /*comment*/*/ ".into(), - )), - Token::Whitespace(Whitespace::Space), - Token::Div, - Token::Word(Word { - value: "comment".to_string(), - quote_style: None, - keyword: Keyword::COMMENT, - }), - Token::Mul, - Token::Div, - Token::Number("1".to_string(), false), - ], - ), - ( - "0/*multi-line\n* \n/* comment \n /*comment/**/ */ /comment*/*/1", - vec![ - Token::Number("0".to_string(), false), - Token::Whitespace(Whitespace::MultiLineComment( - "multi-line\n* \n/* comment \n /*comment/**/ */ /comment*/".into(), - )), - Token::Number("1".to_string(), false), - ], - ), - ( - "SELECT 1/* a /* b */ c */0", - vec![ - Token::make_keyword("SELECT"), - Token::Whitespace(Whitespace::Space), - Token::Number("1".to_string(), false), - Token::Whitespace(Whitespace::MultiLineComment(" a /* b */ c ".to_string())), - Token::Number("0".to_string(), false), - ], - ), - ]; + all_dialects_where(|d| d.supports_nested_comments()).tokenizes_to( + "0/*multi-line\n* \n/* comment \n /*comment*/*/ */ /comment*/1", + vec![ + Token::Number("0".to_string(), false), + Token::Whitespace(Whitespace::MultiLineComment( + "multi-line\n* \n/* comment \n /*comment*/*/ ".into(), + )), + Token::Whitespace(Whitespace::Space), + Token::Div, + Token::Word(Word { + value: "comment".to_string(), + quote_style: None, + keyword: Keyword::COMMENT, + }), + Token::Mul, + Token::Div, + Token::Number("1".to_string(), false), + ], + ); - for (sql, expected) in test_cases { - let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); - compare(expected, tokens); - } + all_dialects_where(|d| d.supports_nested_comments()).tokenizes_to( + "0/*multi-line\n* \n/* comment \n /*comment/**/ */ /comment*/*/1", + vec![ + Token::Number("0".to_string(), false), + Token::Whitespace(Whitespace::MultiLineComment( + "multi-line\n* \n/* comment \n /*comment/**/ */ /comment*/".into(), + )), + Token::Number("1".to_string(), false), + ], + ); + + all_dialects_where(|d| d.supports_nested_comments()).tokenizes_to( + "SELECT 1/* a /* b */ c */0", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Number("1".to_string(), false), + Token::Whitespace(Whitespace::MultiLineComment(" a /* b */ c ".to_string())), + Token::Number("0".to_string(), false), + ], + ); } #[test] fn tokenize_nested_multiline_comment_empty() { - let sql = "select 1/*/**/*/0"; - - let dialect = GenericDialect {}; - let tokens = Tokenizer::new(&dialect, sql).tokenize().unwrap(); - let expected = vec![ - Token::make_keyword("select"), - Token::Whitespace(Whitespace::Space), - Token::Number("1".to_string(), false), - Token::Whitespace(Whitespace::MultiLineComment("/**/".to_string())), - Token::Number("0".to_string(), false), - ]; - - compare(expected, tokens); + all_dialects_where(|d| d.supports_nested_comments()).tokenizes_to( + "select 1/*/**/*/0", + vec![ + Token::make_keyword("select"), + Token::Whitespace(Whitespace::Space), + Token::Number("1".to_string(), false), + Token::Whitespace(Whitespace::MultiLineComment("/**/".to_string())), + Token::Number("0".to_string(), false), + ], + ); } #[test] fn tokenize_nested_comments_if_not_supported() { - let dialect = SQLiteDialect {}; - let sql = "SELECT 1/*/* nested comment */*/0"; - let tokens = Tokenizer::new(&dialect, sql).tokenize(); - let expected = vec![ - Token::make_keyword("SELECT"), - Token::Whitespace(Whitespace::Space), - Token::Number("1".to_string(), false), - Token::Whitespace(Whitespace::MultiLineComment( - "/* nested comment ".to_string(), - )), - Token::Mul, - Token::Div, - Token::Number("0".to_string(), false), - ]; - - compare(expected, tokens.unwrap()); + all_dialects_except(|d| d.supports_nested_comments()).tokenizes_to( + "SELECT 1/*/* nested comment */*/0", + vec![ + Token::make_keyword("SELECT"), + Token::Whitespace(Whitespace::Space), + Token::Number("1".to_string(), false), + Token::Whitespace(Whitespace::MultiLineComment( + "/* nested comment ".to_string(), + )), + Token::Mul, + Token::Div, + Token::Number("0".to_string(), false), + ], + ); } #[test] From 7461d8bd02701de55960851c531c1b5326a737c9 Mon Sep 17 00:00:00 2001 From: Mohamed Abdeen <83442793+MohamedAbdeen21@users.noreply.github.com> Date: Fri, 26 Sep 2025 11:46:18 +0100 Subject: [PATCH 268/286] MySQL: `CREATE INDEX`: allow `USING` clause before `ON` (#2029) --- src/ast/ddl.rs | 2 ++ src/parser/mod.rs | 15 +++++++++----- tests/sqlparser_common.rs | 43 +++++++++++++++++++++++++++++++++++++++ 3 files changed, 55 insertions(+), 5 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 4c145f91e..c4f769675 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2361,6 +2361,8 @@ pub struct CreateIndex { pub name: Option, #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] pub table_name: ObjectName, + /// Index type used in the statement. Can also be found inside [`CreateIndex::index_options`] + /// depending on the position of the option within the statement. pub using: Option, pub columns: Vec, pub unique: bool, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 66089be7c..d97fa1bd0 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7063,19 +7063,24 @@ impl<'a> Parser<'a> { pub fn parse_create_index(&mut self, unique: bool) -> Result { let concurrently = self.parse_keyword(Keyword::CONCURRENTLY); let if_not_exists = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + + let mut using = None; + let index_name = if if_not_exists || !self.parse_keyword(Keyword::ON) { let index_name = self.parse_object_name(false)?; + // MySQL allows `USING index_type` either before or after `ON table_name` + using = self.parse_optional_using_then_index_type()?; self.expect_keyword_is(Keyword::ON)?; Some(index_name) } else { None }; + let table_name = self.parse_object_name(false)?; - let using = if self.parse_keyword(Keyword::USING) { - Some(self.parse_index_type()?) - } else { - None - }; + + // MySQL allows having two `USING` clauses. + // In that case, the second clause overwrites the first. + using = self.parse_optional_using_then_index_type()?.or(using); let columns = self.parse_parenthesized_index_column_list()?; diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index b9434581a..365d5469e 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -17252,6 +17252,49 @@ fn parse_invisible_column() { } } +#[test] +fn parse_create_index_different_using_positions() { + let sql = "CREATE INDEX idx_name USING BTREE ON table_name (col1)"; + let expected = "CREATE INDEX idx_name ON table_name USING BTREE (col1)"; + match all_dialects().one_statement_parses_to(sql, expected) { + Statement::CreateIndex(CreateIndex { + name, + table_name, + using, + columns, + unique, + .. + }) => { + assert_eq!(name.unwrap().to_string(), "idx_name"); + assert_eq!(table_name.to_string(), "table_name"); + assert_eq!(using, Some(IndexType::BTree)); + assert_eq!(columns.len(), 1); + assert!(!unique); + } + _ => unreachable!(), + } + + let sql = "CREATE INDEX idx_name USING BTREE ON table_name (col1) USING HASH"; + let expected = "CREATE INDEX idx_name ON table_name USING BTREE (col1) USING HASH"; + match all_dialects().one_statement_parses_to(sql, expected) { + Statement::CreateIndex(CreateIndex { + name, + table_name, + columns, + index_options, + .. + }) => { + assert_eq!(name.unwrap().to_string(), "idx_name"); + assert_eq!(table_name.to_string(), "table_name"); + assert_eq!(columns.len(), 1); + assert!(index_options + .iter() + .any(|o| o == &IndexOption::Using(IndexType::Hash))); + } + _ => unreachable!(), + } +} + #[test] fn test_parse_alter_user() { verified_stmt("ALTER USER u1"); From ade40826563451cc14c130af9d689f4050dbdb15 Mon Sep 17 00:00:00 2001 From: nick young Date: Fri, 3 Oct 2025 23:35:09 -0700 Subject: [PATCH 269/286] [databricks] update dialect to support grouping by with modifier (#2047) Co-authored-by: Ifeanyi Ubah --- src/dialect/databricks.rs | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/dialect/databricks.rs b/src/dialect/databricks.rs index 4bb8c8d51..c5d5f9740 100644 --- a/src/dialect/databricks.rs +++ b/src/dialect/databricks.rs @@ -69,4 +69,9 @@ impl Dialect for DatabricksDialect { fn supports_nested_comments(&self) -> bool { true } + + /// See + fn supports_group_by_with_modifier(&self) -> bool { + true + } } From 8c82fc0a190aa5c337a033d88fd0bbca3e61d642 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Wed, 8 Oct 2025 13:03:26 +0200 Subject: [PATCH 270/286] Moved constraint variant outside of `TableConstraint` enum (#2054) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 309 +-------------------- src/ast/mod.rs | 14 +- src/ast/spans.rs | 83 +----- src/ast/table_constraints.rs | 516 +++++++++++++++++++++++++++++++++++ src/parser/mod.rs | 132 +++++---- src/test_utils.rs | 35 ++- tests/sqlparser_common.rs | 40 +-- tests/sqlparser_mysql.rs | 10 +- tests/sqlparser_postgres.rs | 33 ++- 9 files changed, 688 insertions(+), 484 deletions(-) create mode 100644 src/ast/table_constraints.rs diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index c4f769675..b7e020f5b 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -30,15 +30,15 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, ArgMode, CommentDef, ConditionalStatements, - CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, DataType, - Expr, FileFormat, FunctionBehavior, FunctionCalledOnNull, FunctionDeterminismSpecifier, - FunctionParallel, HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident, - InitializeKind, MySQLColumnPosition, ObjectName, OnCommit, OneOrManyWithParens, - OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RefreshModeKind, RowAccessPolicy, - SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, TableVersion, Tag, - TriggerEvent, TriggerExecBody, TriggerObject, TriggerPeriod, TriggerReferencing, Value, - ValueWithSpan, WrappedCollection, + display_comma_separated, display_separated, table_constraints::TableConstraint, ArgMode, + CommentDef, ConditionalStatements, CreateFunctionBody, CreateFunctionUsing, + CreateTableLikeKind, CreateTableOptions, DataType, Expr, FileFormat, FunctionBehavior, + FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, HiveDistributionStyle, + HiveFormat, HiveIOFormat, HiveRowFormat, Ident, InitializeKind, MySQLColumnPosition, + ObjectName, OnCommit, OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, + Query, RefreshModeKind, RowAccessPolicy, SequenceOptions, Spanned, SqlOption, + StorageSerializationPolicy, TableVersion, Tag, TriggerEvent, TriggerExecBody, TriggerObject, + TriggerPeriod, TriggerReferencing, Value, ValueWithSpan, WrappedCollection, }; use crate::display_utils::{DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; use crate::keywords::Keyword; @@ -1029,291 +1029,6 @@ impl fmt::Display for AlterColumnOperation { } } -/// A table-level constraint, specified in a `CREATE TABLE` or an -/// `ALTER TABLE ADD ` statement. -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] -#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] -#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] -pub enum TableConstraint { - /// MySQL [definition][1] for `UNIQUE` constraints statements:\ - /// * `[CONSTRAINT []] UNIQUE [] [index_type] () ` - /// - /// where: - /// * [index_type][2] is `USING {BTREE | HASH}` - /// * [index_options][3] is `{index_type | COMMENT 'string' | ... %currently unsupported stmts% } ...` - /// * [index_type_display][4] is `[INDEX | KEY]` - /// - /// [1]: https://dev.mysql.com/doc/refman/8.3/en/create-table.html - /// [2]: IndexType - /// [3]: IndexOption - /// [4]: KeyOrIndexDisplay - Unique { - /// Constraint name. - /// - /// Can be not the same as `index_name` - name: Option, - /// Index name - index_name: Option, - /// Whether the type is followed by the keyword `KEY`, `INDEX`, or no keyword at all. - index_type_display: KeyOrIndexDisplay, - /// Optional `USING` of [index type][1] statement before columns. - /// - /// [1]: IndexType - index_type: Option, - /// Identifiers of the columns that are unique. - columns: Vec, - index_options: Vec, - characteristics: Option, - /// Optional Postgres nulls handling: `[ NULLS [ NOT ] DISTINCT ]` - nulls_distinct: NullsDistinctOption, - }, - /// MySQL [definition][1] for `PRIMARY KEY` constraints statements:\ - /// * `[CONSTRAINT []] PRIMARY KEY [index_name] [index_type] () ` - /// - /// Actually the specification have no `[index_name]` but the next query will complete successfully: - /// ```sql - /// CREATE TABLE unspec_table ( - /// xid INT NOT NULL, - /// CONSTRAINT p_name PRIMARY KEY index_name USING BTREE (xid) - /// ); - /// ``` - /// - /// where: - /// * [index_type][2] is `USING {BTREE | HASH}` - /// * [index_options][3] is `{index_type | COMMENT 'string' | ... %currently unsupported stmts% } ...` - /// - /// [1]: https://dev.mysql.com/doc/refman/8.3/en/create-table.html - /// [2]: IndexType - /// [3]: IndexOption - PrimaryKey { - /// Constraint name. - /// - /// Can be not the same as `index_name` - name: Option, - /// Index name - index_name: Option, - /// Optional `USING` of [index type][1] statement before columns. - /// - /// [1]: IndexType - index_type: Option, - /// Identifiers of the columns that form the primary key. - columns: Vec, - index_options: Vec, - characteristics: Option, - }, - /// A referential integrity constraint (`[ CONSTRAINT ] FOREIGN KEY () - /// REFERENCES () - /// { [ON DELETE ] [ON UPDATE ] | - /// [ON UPDATE ] [ON DELETE ] - /// }`). - ForeignKey { - name: Option, - /// MySQL-specific field - /// - index_name: Option, - columns: Vec, - foreign_table: ObjectName, - referred_columns: Vec, - on_delete: Option, - on_update: Option, - characteristics: Option, - }, - /// `[ CONSTRAINT ] CHECK () [[NOT] ENFORCED]` - Check { - name: Option, - expr: Box, - /// MySQL-specific syntax - /// - enforced: Option, - }, - /// MySQLs [index definition][1] for index creation. Not present on ANSI so, for now, the usage - /// is restricted to MySQL, as no other dialects that support this syntax were found. - /// - /// `{INDEX | KEY} [index_name] [index_type] (key_part,...) [index_option]...` - /// - /// [1]: https://dev.mysql.com/doc/refman/8.0/en/create-table.html - Index { - /// Whether this index starts with KEY (true) or INDEX (false), to maintain the same syntax. - display_as_key: bool, - /// Index name. - name: Option, - /// Optional [index type][1]. - /// - /// [1]: IndexType - index_type: Option, - /// Referred column identifier list. - columns: Vec, - /// Optional index options such as `USING`; see [`IndexOption`]. - index_options: Vec, - }, - /// MySQLs [fulltext][1] definition. Since the [`SPATIAL`][2] definition is exactly the same, - /// and MySQL displays both the same way, it is part of this definition as well. - /// - /// Supported syntax: - /// - /// ```markdown - /// {FULLTEXT | SPATIAL} [INDEX | KEY] [index_name] (key_part,...) - /// - /// key_part: col_name - /// ``` - /// - /// [1]: https://dev.mysql.com/doc/refman/8.0/en/fulltext-natural-language.html - /// [2]: https://dev.mysql.com/doc/refman/8.0/en/spatial-types.html - FulltextOrSpatial { - /// Whether this is a `FULLTEXT` (true) or `SPATIAL` (false) definition. - fulltext: bool, - /// Whether the type is followed by the keyword `KEY`, `INDEX`, or no keyword at all. - index_type_display: KeyOrIndexDisplay, - /// Optional index name. - opt_index_name: Option, - /// Referred column identifier list. - columns: Vec, - }, -} - -impl fmt::Display for TableConstraint { - fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - match self { - TableConstraint::Unique { - name, - index_name, - index_type_display, - index_type, - columns, - index_options, - characteristics, - nulls_distinct, - } => { - write!( - f, - "{}UNIQUE{nulls_distinct}{index_type_display:>}{}{} ({})", - display_constraint_name(name), - display_option_spaced(index_name), - display_option(" USING ", "", index_type), - display_comma_separated(columns), - )?; - - if !index_options.is_empty() { - write!(f, " {}", display_separated(index_options, " "))?; - } - - write!(f, "{}", display_option_spaced(characteristics))?; - Ok(()) - } - TableConstraint::PrimaryKey { - name, - index_name, - index_type, - columns, - index_options, - characteristics, - } => { - write!( - f, - "{}PRIMARY KEY{}{} ({})", - display_constraint_name(name), - display_option_spaced(index_name), - display_option(" USING ", "", index_type), - display_comma_separated(columns), - )?; - - if !index_options.is_empty() { - write!(f, " {}", display_separated(index_options, " "))?; - } - - write!(f, "{}", display_option_spaced(characteristics))?; - Ok(()) - } - TableConstraint::ForeignKey { - name, - index_name, - columns, - foreign_table, - referred_columns, - on_delete, - on_update, - characteristics, - } => { - write!( - f, - "{}FOREIGN KEY{} ({}) REFERENCES {}", - display_constraint_name(name), - display_option_spaced(index_name), - display_comma_separated(columns), - foreign_table, - )?; - if !referred_columns.is_empty() { - write!(f, "({})", display_comma_separated(referred_columns))?; - } - if let Some(action) = on_delete { - write!(f, " ON DELETE {action}")?; - } - if let Some(action) = on_update { - write!(f, " ON UPDATE {action}")?; - } - if let Some(characteristics) = characteristics { - write!(f, " {characteristics}")?; - } - Ok(()) - } - TableConstraint::Check { - name, - expr, - enforced, - } => { - write!(f, "{}CHECK ({})", display_constraint_name(name), expr)?; - if let Some(b) = enforced { - write!(f, " {}", if *b { "ENFORCED" } else { "NOT ENFORCED" }) - } else { - Ok(()) - } - } - TableConstraint::Index { - display_as_key, - name, - index_type, - columns, - index_options, - } => { - write!(f, "{}", if *display_as_key { "KEY" } else { "INDEX" })?; - if let Some(name) = name { - write!(f, " {name}")?; - } - if let Some(index_type) = index_type { - write!(f, " USING {index_type}")?; - } - write!(f, " ({})", display_comma_separated(columns))?; - if !index_options.is_empty() { - write!(f, " {}", display_comma_separated(index_options))?; - } - Ok(()) - } - Self::FulltextOrSpatial { - fulltext, - index_type_display, - opt_index_name, - columns, - } => { - if *fulltext { - write!(f, "FULLTEXT")?; - } else { - write!(f, "SPATIAL")?; - } - - write!(f, "{index_type_display:>}")?; - - if let Some(name) = opt_index_name { - write!(f, " {name}")?; - } - - write!(f, " ({})", display_comma_separated(columns))?; - - Ok(()) - } - } - } -} - /// Representation whether a definition can can contains the KEY or INDEX keywords with the same /// meaning. /// @@ -2065,7 +1780,7 @@ pub enum GeneratedExpressionMode { } #[must_use] -fn display_constraint_name(name: &'_ Option) -> impl fmt::Display + '_ { +pub(crate) fn display_constraint_name(name: &'_ Option) -> impl fmt::Display + '_ { struct ConstraintName<'a>(&'a Option); impl fmt::Display for ConstraintName<'_> { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { @@ -2082,7 +1797,7 @@ fn display_constraint_name(name: &'_ Option) -> impl fmt::Display + '_ { /// * `Some(inner)` => create display struct for `"{prefix}{inner}{postfix}"` /// * `_` => do nothing #[must_use] -fn display_option<'a, T: fmt::Display>( +pub(crate) fn display_option<'a, T: fmt::Display>( prefix: &'a str, postfix: &'a str, option: &'a Option, @@ -2104,7 +1819,7 @@ fn display_option<'a, T: fmt::Display>( /// * `Some(inner)` => create display struct for `" {inner}"` /// * `_` => do nothing #[must_use] -fn display_option_spaced(option: &Option) -> impl fmt::Display + '_ { +pub(crate) fn display_option_spaced(option: &Option) -> impl fmt::Display + '_ { display_option(" ", "", option) } diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 4c1743feb..52968d7db 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -68,9 +68,8 @@ pub use self::ddl::{ DropBehavior, DropTrigger, GeneratedAs, GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, - ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, TableConstraint, - TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, - ViewColumnDef, + ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, TagsColumnOption, + UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{Delete, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; @@ -118,6 +117,11 @@ mod dcl; mod ddl; mod dml; pub mod helpers; +pub mod table_constraints; +pub use table_constraints::{ + CheckConstraint, ForeignKeyConstraint, FullTextOrSpatialConstraint, IndexConstraint, + PrimaryKeyConstraint, TableConstraint, UniqueConstraint, +}; mod operator; mod query; mod spans; @@ -152,14 +156,14 @@ where } } -pub fn display_separated<'a, T>(slice: &'a [T], sep: &'static str) -> DisplaySeparated<'a, T> +pub(crate) fn display_separated<'a, T>(slice: &'a [T], sep: &'static str) -> DisplaySeparated<'a, T> where T: fmt::Display, { DisplaySeparated { slice, sep } } -pub fn display_comma_separated(slice: &[T]) -> DisplaySeparated<'_, T> +pub(crate) fn display_comma_separated(slice: &[T]) -> DisplaySeparated<'_, T> where T: fmt::Display, { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 4c53e55ce..0a303fcfd 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -670,83 +670,12 @@ impl Spanned for ColumnOptionDef { impl Spanned for TableConstraint { fn span(&self) -> Span { match self { - TableConstraint::Unique { - name, - index_name, - index_type_display: _, - index_type: _, - columns, - index_options: _, - characteristics, - nulls_distinct: _, - } => union_spans( - name.iter() - .map(|i| i.span) - .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span())) - .chain(characteristics.iter().map(|i| i.span())), - ), - TableConstraint::PrimaryKey { - name, - index_name, - index_type: _, - columns, - index_options: _, - characteristics, - } => union_spans( - name.iter() - .map(|i| i.span) - .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span())) - .chain(characteristics.iter().map(|i| i.span())), - ), - TableConstraint::ForeignKey { - name, - columns, - index_name, - foreign_table, - referred_columns, - on_delete, - on_update, - characteristics, - } => union_spans( - name.iter() - .map(|i| i.span) - .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span)) - .chain(core::iter::once(foreign_table.span())) - .chain(referred_columns.iter().map(|i| i.span)) - .chain(on_delete.iter().map(|i| i.span())) - .chain(on_update.iter().map(|i| i.span())) - .chain(characteristics.iter().map(|i| i.span())), - ), - TableConstraint::Check { - name, - expr, - enforced: _, - } => expr.span().union_opt(&name.as_ref().map(|i| i.span)), - TableConstraint::Index { - display_as_key: _, - name, - index_type: _, - columns, - index_options: _, - } => union_spans( - name.iter() - .map(|i| i.span) - .chain(columns.iter().map(|i| i.span())), - ), - TableConstraint::FulltextOrSpatial { - fulltext: _, - index_type_display: _, - opt_index_name, - columns, - } => union_spans( - opt_index_name - .iter() - .map(|i| i.span) - .chain(columns.iter().map(|i| i.span())), - ), + TableConstraint::Unique(constraint) => constraint.span(), + TableConstraint::PrimaryKey(constraint) => constraint.span(), + TableConstraint::ForeignKey(constraint) => constraint.span(), + TableConstraint::Check(constraint) => constraint.span(), + TableConstraint::Index(constraint) => constraint.span(), + TableConstraint::FulltextOrSpatial(constraint) => constraint.span(), } } } diff --git a/src/ast/table_constraints.rs b/src/ast/table_constraints.rs new file mode 100644 index 000000000..afcf62959 --- /dev/null +++ b/src/ast/table_constraints.rs @@ -0,0 +1,516 @@ +// Licensed to the Apache Software Foundation (ASF) under one +// or more contributor license agreements. See the NOTICE file +// distributed with this work for additional information +// regarding copyright ownership. The ASF licenses this file +// to you under the Apache License, Version 2.0 (the +// "License"); you may not use this file except in compliance +// with the License. You may obtain a copy of the License at +// +// http://www.apache.org/licenses/LICENSE-2.0 +// +// Unless required by applicable law or agreed to in writing, +// software distributed under the License is distributed on an +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY +// KIND, either express or implied. See the License for the +// specific language governing permissions and limitations +// under the License. + +//! SQL Abstract Syntax Tree (AST) types for table constraints + +use crate::ast::{ + display_comma_separated, display_separated, ConstraintCharacteristics, Expr, Ident, + IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, ObjectName, + ReferentialAction, +}; +use crate::tokenizer::Span; +use core::fmt; + +#[cfg(not(feature = "std"))] +use alloc::{boxed::Box, vec::Vec}; + +#[cfg(feature = "serde")] +use serde::{Deserialize, Serialize}; + +#[cfg(feature = "visitor")] +use sqlparser_derive::{Visit, VisitMut}; + +/// A table-level constraint, specified in a `CREATE TABLE` or an +/// `ALTER TABLE ADD ` statement. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum TableConstraint { + /// MySQL [definition][1] for `UNIQUE` constraints statements:\ + /// * `[CONSTRAINT []] UNIQUE [] [index_type] () ` + /// + /// where: + /// * [index_type][2] is `USING {BTREE | HASH}` + /// * [index_options][3] is `{index_type | COMMENT 'string' | ... %currently unsupported stmts% } ...` + /// * [index_type_display][4] is `[INDEX | KEY]` + /// + /// [1]: https://dev.mysql.com/doc/refman/8.3/en/create-table.html + /// [2]: IndexType + /// [3]: IndexOption + /// [4]: KeyOrIndexDisplay + Unique(UniqueConstraint), + /// MySQL [definition][1] for `PRIMARY KEY` constraints statements:\ + /// * `[CONSTRAINT []] PRIMARY KEY [index_name] [index_type] () ` + /// + /// Actually the specification have no `[index_name]` but the next query will complete successfully: + /// ```sql + /// CREATE TABLE unspec_table ( + /// xid INT NOT NULL, + /// CONSTRAINT p_name PRIMARY KEY index_name USING BTREE (xid) + /// ); + /// ``` + /// + /// where: + /// * [index_type][2] is `USING {BTREE | HASH}` + /// * [index_options][3] is `{index_type | COMMENT 'string' | ... %currently unsupported stmts% } ...` + /// + /// [1]: https://dev.mysql.com/doc/refman/8.3/en/create-table.html + /// [2]: IndexType + /// [3]: IndexOption + PrimaryKey(PrimaryKeyConstraint), + /// A referential integrity constraint (`[ CONSTRAINT ] FOREIGN KEY () + /// REFERENCES () + /// { [ON DELETE ] [ON UPDATE ] | + /// [ON UPDATE ] [ON DELETE ] + /// }`). + ForeignKey(ForeignKeyConstraint), + /// `[ CONSTRAINT ] CHECK () [[NOT] ENFORCED]` + Check(CheckConstraint), + /// MySQLs [index definition][1] for index creation. Not present on ANSI so, for now, the usage + /// is restricted to MySQL, as no other dialects that support this syntax were found. + /// + /// `{INDEX | KEY} [index_name] [index_type] (key_part,...) [index_option]...` + /// + /// [1]: https://dev.mysql.com/doc/refman/8.0/en/create-table.html + Index(IndexConstraint), + /// MySQLs [fulltext][1] definition. Since the [`SPATIAL`][2] definition is exactly the same, + /// and MySQL displays both the same way, it is part of this definition as well. + /// + /// Supported syntax: + /// + /// ```markdown + /// {FULLTEXT | SPATIAL} [INDEX | KEY] [index_name] (key_part,...) + /// + /// key_part: col_name + /// ``` + /// + /// [1]: https://dev.mysql.com/doc/refman/8.0/en/fulltext-natural-language.html + /// [2]: https://dev.mysql.com/doc/refman/8.0/en/spatial-types.html + FulltextOrSpatial(FullTextOrSpatialConstraint), +} + +impl From for TableConstraint { + fn from(constraint: UniqueConstraint) -> Self { + TableConstraint::Unique(constraint) + } +} + +impl From for TableConstraint { + fn from(constraint: PrimaryKeyConstraint) -> Self { + TableConstraint::PrimaryKey(constraint) + } +} + +impl From for TableConstraint { + fn from(constraint: ForeignKeyConstraint) -> Self { + TableConstraint::ForeignKey(constraint) + } +} + +impl From for TableConstraint { + fn from(constraint: CheckConstraint) -> Self { + TableConstraint::Check(constraint) + } +} + +impl From for TableConstraint { + fn from(constraint: IndexConstraint) -> Self { + TableConstraint::Index(constraint) + } +} + +impl From for TableConstraint { + fn from(constraint: FullTextOrSpatialConstraint) -> Self { + TableConstraint::FulltextOrSpatial(constraint) + } +} + +impl fmt::Display for TableConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + TableConstraint::Unique(constraint) => constraint.fmt(f), + TableConstraint::PrimaryKey(constraint) => constraint.fmt(f), + TableConstraint::ForeignKey(constraint) => constraint.fmt(f), + TableConstraint::Check(constraint) => constraint.fmt(f), + TableConstraint::Index(constraint) => constraint.fmt(f), + TableConstraint::FulltextOrSpatial(constraint) => constraint.fmt(f), + } + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CheckConstraint { + pub name: Option, + pub expr: Box, + /// MySQL-specific syntax + /// + pub enforced: Option, +} + +impl fmt::Display for CheckConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + use crate::ast::ddl::display_constraint_name; + write!( + f, + "{}CHECK ({})", + display_constraint_name(&self.name), + self.expr + )?; + if let Some(b) = self.enforced { + write!(f, " {}", if b { "ENFORCED" } else { "NOT ENFORCED" }) + } else { + Ok(()) + } + } +} + +impl crate::ast::Spanned for CheckConstraint { + fn span(&self) -> Span { + self.expr + .span() + .union_opt(&self.name.as_ref().map(|i| i.span)) + } +} + +/// A referential integrity constraint (`[ CONSTRAINT ] FOREIGN KEY () +/// REFERENCES () +/// { [ON DELETE ] [ON UPDATE ] | +/// [ON UPDATE ] [ON DELETE ] +/// }`). +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ForeignKeyConstraint { + pub name: Option, + /// MySQL-specific field + /// + pub index_name: Option, + pub columns: Vec, + pub foreign_table: ObjectName, + pub referred_columns: Vec, + pub on_delete: Option, + pub on_update: Option, + pub characteristics: Option, +} + +impl fmt::Display for ForeignKeyConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + use crate::ast::ddl::{display_constraint_name, display_option_spaced}; + write!( + f, + "{}FOREIGN KEY{} ({}) REFERENCES {}", + display_constraint_name(&self.name), + display_option_spaced(&self.index_name), + display_comma_separated(&self.columns), + self.foreign_table, + )?; + if !self.referred_columns.is_empty() { + write!(f, "({})", display_comma_separated(&self.referred_columns))?; + } + if let Some(action) = &self.on_delete { + write!(f, " ON DELETE {action}")?; + } + if let Some(action) = &self.on_update { + write!(f, " ON UPDATE {action}")?; + } + if let Some(characteristics) = &self.characteristics { + write!(f, " {characteristics}")?; + } + Ok(()) + } +} + +impl crate::ast::Spanned for ForeignKeyConstraint { + fn span(&self) -> Span { + fn union_spans>(iter: I) -> Span { + Span::union_iter(iter) + } + + union_spans( + self.name + .iter() + .map(|i| i.span) + .chain(self.index_name.iter().map(|i| i.span)) + .chain(self.columns.iter().map(|i| i.span)) + .chain(core::iter::once(self.foreign_table.span())) + .chain(self.referred_columns.iter().map(|i| i.span)) + .chain(self.on_delete.iter().map(|i| i.span())) + .chain(self.on_update.iter().map(|i| i.span())) + .chain(self.characteristics.iter().map(|i| i.span())), + ) + } +} + +/// MySQLs [fulltext][1] definition. Since the [`SPATIAL`][2] definition is exactly the same, +/// and MySQL displays both the same way, it is part of this definition as well. +/// +/// Supported syntax: +/// +/// ```markdown +/// {FULLTEXT | SPATIAL} [INDEX | KEY] [index_name] (key_part,...) +/// +/// key_part: col_name +/// ``` +/// +/// [1]: https://dev.mysql.com/doc/refman/8.0/en/fulltext-natural-language.html +/// [2]: https://dev.mysql.com/doc/refman/8.0/en/spatial-types.html +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct FullTextOrSpatialConstraint { + /// Whether this is a `FULLTEXT` (true) or `SPATIAL` (false) definition. + pub fulltext: bool, + /// Whether the type is followed by the keyword `KEY`, `INDEX`, or no keyword at all. + pub index_type_display: KeyOrIndexDisplay, + /// Optional index name. + pub opt_index_name: Option, + /// Referred column identifier list. + pub columns: Vec, +} + +impl fmt::Display for FullTextOrSpatialConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + if self.fulltext { + write!(f, "FULLTEXT")?; + } else { + write!(f, "SPATIAL")?; + } + + write!(f, "{:>}", self.index_type_display)?; + + if let Some(name) = &self.opt_index_name { + write!(f, " {name}")?; + } + + write!(f, " ({})", display_comma_separated(&self.columns))?; + + Ok(()) + } +} + +impl crate::ast::Spanned for FullTextOrSpatialConstraint { + fn span(&self) -> Span { + fn union_spans>(iter: I) -> Span { + Span::union_iter(iter) + } + + union_spans( + self.opt_index_name + .iter() + .map(|i| i.span) + .chain(self.columns.iter().map(|i| i.span())), + ) + } +} + +/// MySQLs [index definition][1] for index creation. Not present on ANSI so, for now, the usage +/// is restricted to MySQL, as no other dialects that support this syntax were found. +/// +/// `{INDEX | KEY} [index_name] [index_type] (key_part,...) [index_option]...` +/// +/// [1]: https://dev.mysql.com/doc/refman/8.0/en/create-table.html +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct IndexConstraint { + /// Whether this index starts with KEY (true) or INDEX (false), to maintain the same syntax. + pub display_as_key: bool, + /// Index name. + pub name: Option, + /// Optional [index type][1]. + /// + /// [1]: IndexType + pub index_type: Option, + /// Referred column identifier list. + pub columns: Vec, + /// Optional index options such as `USING`; see [`IndexOption`]. + pub index_options: Vec, +} + +impl fmt::Display for IndexConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}", if self.display_as_key { "KEY" } else { "INDEX" })?; + if let Some(name) = &self.name { + write!(f, " {name}")?; + } + if let Some(index_type) = &self.index_type { + write!(f, " USING {index_type}")?; + } + write!(f, " ({})", display_comma_separated(&self.columns))?; + if !self.index_options.is_empty() { + write!(f, " {}", display_comma_separated(&self.index_options))?; + } + Ok(()) + } +} + +impl crate::ast::Spanned for IndexConstraint { + fn span(&self) -> Span { + fn union_spans>(iter: I) -> Span { + Span::union_iter(iter) + } + + union_spans( + self.name + .iter() + .map(|i| i.span) + .chain(self.columns.iter().map(|i| i.span())), + ) + } +} + +/// MySQL [definition][1] for `PRIMARY KEY` constraints statements: +/// * `[CONSTRAINT []] PRIMARY KEY [index_name] [index_type] () ` +/// +/// Actually the specification have no `[index_name]` but the next query will complete successfully: +/// ```sql +/// CREATE TABLE unspec_table ( +/// xid INT NOT NULL, +/// CONSTRAINT p_name PRIMARY KEY index_name USING BTREE (xid) +/// ); +/// ``` +/// +/// where: +/// * [index_type][2] is `USING {BTREE | HASH}` +/// * [index_options][3] is `{index_type | COMMENT 'string' | ... %currently unsupported stmts% } ...` +/// +/// [1]: https://dev.mysql.com/doc/refman/8.3/en/create-table.html +/// [2]: IndexType +/// [3]: IndexOption +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct PrimaryKeyConstraint { + /// Constraint name. + /// + /// Can be not the same as `index_name` + pub name: Option, + /// Index name + pub index_name: Option, + /// Optional `USING` of [index type][1] statement before columns. + /// + /// [1]: IndexType + pub index_type: Option, + /// Identifiers of the columns that form the primary key. + pub columns: Vec, + pub index_options: Vec, + pub characteristics: Option, +} + +impl fmt::Display for PrimaryKeyConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + use crate::ast::ddl::{display_constraint_name, display_option, display_option_spaced}; + write!( + f, + "{}PRIMARY KEY{}{} ({})", + display_constraint_name(&self.name), + display_option_spaced(&self.index_name), + display_option(" USING ", "", &self.index_type), + display_comma_separated(&self.columns), + )?; + + if !self.index_options.is_empty() { + write!(f, " {}", display_separated(&self.index_options, " "))?; + } + + write!(f, "{}", display_option_spaced(&self.characteristics))?; + Ok(()) + } +} + +impl crate::ast::Spanned for PrimaryKeyConstraint { + fn span(&self) -> Span { + fn union_spans>(iter: I) -> Span { + Span::union_iter(iter) + } + + union_spans( + self.name + .iter() + .map(|i| i.span) + .chain(self.index_name.iter().map(|i| i.span)) + .chain(self.columns.iter().map(|i| i.span())) + .chain(self.characteristics.iter().map(|i| i.span())), + ) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct UniqueConstraint { + /// Constraint name. + /// + /// Can be not the same as `index_name` + pub name: Option, + /// Index name + pub index_name: Option, + /// Whether the type is followed by the keyword `KEY`, `INDEX`, or no keyword at all. + pub index_type_display: KeyOrIndexDisplay, + /// Optional `USING` of [index type][1] statement before columns. + /// + /// [1]: IndexType + pub index_type: Option, + /// Identifiers of the columns that are unique. + pub columns: Vec, + pub index_options: Vec, + pub characteristics: Option, + /// Optional Postgres nulls handling: `[ NULLS [ NOT ] DISTINCT ]` + pub nulls_distinct: NullsDistinctOption, +} + +impl fmt::Display for UniqueConstraint { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + use crate::ast::ddl::{display_constraint_name, display_option, display_option_spaced}; + write!( + f, + "{}UNIQUE{}{:>}{}{} ({})", + display_constraint_name(&self.name), + self.nulls_distinct, + self.index_type_display, + display_option_spaced(&self.index_name), + display_option(" USING ", "", &self.index_type), + display_comma_separated(&self.columns), + )?; + + if !self.index_options.is_empty() { + write!(f, " {}", display_separated(&self.index_options, " "))?; + } + + write!(f, "{}", display_option_spaced(&self.characteristics))?; + Ok(()) + } +} + +impl crate::ast::Spanned for UniqueConstraint { + fn span(&self) -> Span { + fn union_spans>(iter: I) -> Span { + Span::union_iter(iter) + } + + union_spans( + self.name + .iter() + .map(|i| i.span) + .chain(self.index_name.iter().map(|i| i.span)) + .chain(self.columns.iter().map(|i| i.span())) + .chain(self.characteristics.iter().map(|i| i.span())), + ) + } +} diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d97fa1bd0..deca3b577 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8398,16 +8398,19 @@ impl<'a> Parser<'a> { let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; let characteristics = self.parse_constraint_characteristics()?; - Ok(Some(TableConstraint::Unique { - name, - index_name, - index_type_display, - index_type, - columns, - index_options, - characteristics, - nulls_distinct, - })) + Ok(Some( + UniqueConstraint { + name, + index_name, + index_type_display, + index_type, + columns, + index_options, + characteristics, + nulls_distinct, + } + .into(), + )) } Token::Word(w) if w.keyword == Keyword::PRIMARY => { // after `PRIMARY` always stay `KEY` @@ -8420,14 +8423,17 @@ impl<'a> Parser<'a> { let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; let characteristics = self.parse_constraint_characteristics()?; - Ok(Some(TableConstraint::PrimaryKey { - name, - index_name, - index_type, - columns, - index_options, - characteristics, - })) + Ok(Some( + PrimaryKeyConstraint { + name, + index_name, + index_type, + columns, + index_options, + characteristics, + } + .into(), + )) } Token::Word(w) if w.keyword == Keyword::FOREIGN => { self.expect_keyword_is(Keyword::KEY)?; @@ -8452,16 +8458,19 @@ impl<'a> Parser<'a> { let characteristics = self.parse_constraint_characteristics()?; - Ok(Some(TableConstraint::ForeignKey { - name, - index_name, - columns, - foreign_table, - referred_columns, - on_delete, - on_update, - characteristics, - })) + Ok(Some( + ForeignKeyConstraint { + name, + index_name, + columns, + foreign_table, + referred_columns, + on_delete, + on_update, + characteristics, + } + .into(), + )) } Token::Word(w) if w.keyword == Keyword::CHECK => { self.expect_token(&Token::LParen)?; @@ -8476,11 +8485,14 @@ impl<'a> Parser<'a> { None }; - Ok(Some(TableConstraint::Check { - name, - expr, - enforced, - })) + Ok(Some( + CheckConstraint { + name, + expr, + enforced, + } + .into(), + )) } Token::Word(w) if (w.keyword == Keyword::INDEX || w.keyword == Keyword::KEY) @@ -8498,13 +8510,16 @@ impl<'a> Parser<'a> { let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; - Ok(Some(TableConstraint::Index { - display_as_key, - name, - index_type, - columns, - index_options, - })) + Ok(Some( + IndexConstraint { + display_as_key, + name, + index_type, + columns, + index_options, + } + .into(), + )) } Token::Word(w) if (w.keyword == Keyword::FULLTEXT || w.keyword == Keyword::SPATIAL) @@ -8528,12 +8543,15 @@ impl<'a> Parser<'a> { let columns = self.parse_parenthesized_index_column_list()?; - Ok(Some(TableConstraint::FulltextOrSpatial { - fulltext, - index_type_display, - opt_index_name, - columns, - })) + Ok(Some( + FullTextOrSpatialConstraint { + fulltext, + index_type_display, + opt_index_name, + columns, + } + .into(), + )) } _ => { if name.is_some() { @@ -18136,85 +18154,91 @@ mod tests { test_parse_table_constraint!( dialect, "INDEX (c1)", - TableConstraint::Index { + IndexConstraint { display_as_key: false, name: None, index_type: None, columns: vec![mk_expected_col("c1")], index_options: vec![], } + .into() ); test_parse_table_constraint!( dialect, "KEY (c1)", - TableConstraint::Index { + IndexConstraint { display_as_key: true, name: None, index_type: None, columns: vec![mk_expected_col("c1")], index_options: vec![], } + .into() ); test_parse_table_constraint!( dialect, "INDEX 'index' (c1, c2)", - TableConstraint::Index { + TableConstraint::Index(IndexConstraint { display_as_key: false, name: Some(Ident::with_quote('\'', "index")), index_type: None, columns: vec![mk_expected_col("c1"), mk_expected_col("c2")], index_options: vec![], - } + }) ); test_parse_table_constraint!( dialect, "INDEX USING BTREE (c1)", - TableConstraint::Index { + IndexConstraint { display_as_key: false, name: None, index_type: Some(IndexType::BTree), columns: vec![mk_expected_col("c1")], index_options: vec![], } + .into() ); test_parse_table_constraint!( dialect, "INDEX USING HASH (c1)", - TableConstraint::Index { + IndexConstraint { display_as_key: false, name: None, index_type: Some(IndexType::Hash), columns: vec![mk_expected_col("c1")], index_options: vec![], } + .into() ); test_parse_table_constraint!( dialect, "INDEX idx_name USING BTREE (c1)", - TableConstraint::Index { + IndexConstraint { display_as_key: false, name: Some(Ident::new("idx_name")), index_type: Some(IndexType::BTree), columns: vec![mk_expected_col("c1")], index_options: vec![], } + .into() ); test_parse_table_constraint!( dialect, "INDEX idx_name USING HASH (c1)", - TableConstraint::Index { + IndexConstraint { display_as_key: false, name: Some(Ident::new("idx_name")), index_type: Some(IndexType::Hash), columns: vec![mk_expected_col("c1")], index_options: vec![], } + .into() ); } diff --git a/src/test_utils.rs b/src/test_utils.rs index ab2cf89b2..03fdee8ad 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -469,17 +469,17 @@ pub fn index_column(stmt: Statement) -> Expr { } Statement::CreateTable(CreateTable { constraints, .. }) => { match constraints.first().unwrap() { - TableConstraint::Index { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::Index(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } - TableConstraint::Unique { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::Unique(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } - TableConstraint::PrimaryKey { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::PrimaryKey(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } - TableConstraint::FulltextOrSpatial { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::FulltextOrSpatial(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), } @@ -487,19 +487,18 @@ pub fn index_column(stmt: Statement) -> Expr { Statement::AlterTable { operations, .. } => match operations.first().unwrap() { AlterTableOperation::AddConstraint { constraint, .. } => { match constraint { - TableConstraint::Index { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::Index(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } - TableConstraint::Unique { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::Unique(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } - TableConstraint::PrimaryKey { columns, .. } => { - columns.first().unwrap().column.expr.clone() + TableConstraint::PrimaryKey(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() + } + TableConstraint::FulltextOrSpatial(constraint) => { + constraint.columns.first().unwrap().column.expr.clone() } - TableConstraint::FulltextOrSpatial { - columns, - .. - } => columns.first().unwrap().column.expr.clone(), _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), } } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 365d5469e..343ea4955 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3818,7 +3818,7 @@ fn parse_create_table() { assert_eq!( constraints, vec![ - TableConstraint::ForeignKey { + ForeignKeyConstraint { name: Some("fkey".into()), index_name: None, columns: vec!["lat".into()], @@ -3827,8 +3827,9 @@ fn parse_create_table() { on_delete: Some(ReferentialAction::Restrict), on_update: None, characteristics: None, - }, - TableConstraint::ForeignKey { + } + .into(), + ForeignKeyConstraint { name: Some("fkey2".into()), index_name: None, columns: vec!["lat".into()], @@ -3837,8 +3838,9 @@ fn parse_create_table() { on_delete: Some(ReferentialAction::NoAction), on_update: Some(ReferentialAction::Restrict), characteristics: None, - }, - TableConstraint::ForeignKey { + } + .into(), + ForeignKeyConstraint { name: None, index_name: None, columns: vec!["lat".into()], @@ -3847,8 +3849,9 @@ fn parse_create_table() { on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::SetDefault), characteristics: None, - }, - TableConstraint::ForeignKey { + } + .into(), + ForeignKeyConstraint { name: None, index_name: None, columns: vec!["lng".into()], @@ -3857,7 +3860,8 @@ fn parse_create_table() { on_delete: None, on_update: Some(ReferentialAction::SetNull), characteristics: None, - }, + } + .into(), ] ); assert_eq!(table_options, CreateTableOptions::None); @@ -3945,7 +3949,7 @@ fn parse_create_table_with_constraint_characteristics() { assert_eq!( constraints, vec![ - TableConstraint::ForeignKey { + ForeignKeyConstraint { name: Some("fkey".into()), index_name: None, columns: vec!["lat".into()], @@ -3958,8 +3962,9 @@ fn parse_create_table_with_constraint_characteristics() { initially: Some(DeferrableInitial::Deferred), enforced: None }), - }, - TableConstraint::ForeignKey { + } + .into(), + ForeignKeyConstraint { name: Some("fkey2".into()), index_name: None, columns: vec!["lat".into()], @@ -3972,8 +3977,9 @@ fn parse_create_table_with_constraint_characteristics() { initially: Some(DeferrableInitial::Immediate), enforced: None, }), - }, - TableConstraint::ForeignKey { + } + .into(), + ForeignKeyConstraint { name: None, index_name: None, columns: vec!["lat".into()], @@ -3986,8 +3992,9 @@ fn parse_create_table_with_constraint_characteristics() { initially: Some(DeferrableInitial::Deferred), enforced: Some(false), }), - }, - TableConstraint::ForeignKey { + } + .into(), + ForeignKeyConstraint { name: None, index_name: None, columns: vec!["lng".into()], @@ -4000,7 +4007,8 @@ fn parse_create_table_with_constraint_characteristics() { initially: Some(DeferrableInitial::Immediate), enforced: Some(true), }), - }, + } + .into(), ] ); assert_eq!(table_options, CreateTableOptions::None); diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 5d75aa508..35bac548c 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -684,7 +684,7 @@ fn table_constraint_unique_primary_ctor( }) .collect(); match unique_index_type_display { - Some(index_type_display) => TableConstraint::Unique { + Some(index_type_display) => UniqueConstraint { name, index_name, index_type_display, @@ -693,15 +693,17 @@ fn table_constraint_unique_primary_ctor( index_options, characteristics, nulls_distinct: NullsDistinctOption::None, - }, - None => TableConstraint::PrimaryKey { + } + .into(), + None => PrimaryKeyConstraint { name, index_name, index_type, columns, index_options, characteristics, - }, + } + .into(), } } diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 196a82f54..45ae32dac 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -607,9 +607,10 @@ fn parse_alter_table_constraints_unique_nulls_distinct() { { Statement::AlterTable { operations, .. } => match &operations[0] { AlterTableOperation::AddConstraint { - constraint: TableConstraint::Unique { nulls_distinct, .. }, + constraint: TableConstraint::Unique(constraint), .. } => { + let nulls_distinct = &constraint.nulls_distinct; assert_eq!(nulls_distinct, &NullsDistinctOption::NotDistinct) } _ => unreachable!(), @@ -5578,7 +5579,7 @@ fn parse_create_domain() { data_type: DataType::Integer(None), collation: None, default: None, - constraints: vec![TableConstraint::Check { + constraints: vec![CheckConstraint { name: None, expr: Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("VALUE"))), @@ -5586,7 +5587,8 @@ fn parse_create_domain() { right: Box::new(Expr::Value(test_utils::number("0").into())), }), enforced: None, - }], + } + .into()], }); assert_eq!(pg().verified_stmt(sql1), expected); @@ -5597,7 +5599,7 @@ fn parse_create_domain() { data_type: DataType::Integer(None), collation: Some(Ident::with_quote('"', "en_US")), default: None, - constraints: vec![TableConstraint::Check { + constraints: vec![CheckConstraint { name: None, expr: Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("VALUE"))), @@ -5605,7 +5607,8 @@ fn parse_create_domain() { right: Box::new(Expr::Value(test_utils::number("0").into())), }), enforced: None, - }], + } + .into()], }); assert_eq!(pg().verified_stmt(sql2), expected); @@ -5616,7 +5619,7 @@ fn parse_create_domain() { data_type: DataType::Integer(None), collation: None, default: Some(Expr::Value(test_utils::number("1").into())), - constraints: vec![TableConstraint::Check { + constraints: vec![CheckConstraint { name: None, expr: Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("VALUE"))), @@ -5624,7 +5627,8 @@ fn parse_create_domain() { right: Box::new(Expr::Value(test_utils::number("0").into())), }), enforced: None, - }], + } + .into()], }); assert_eq!(pg().verified_stmt(sql3), expected); @@ -5635,7 +5639,7 @@ fn parse_create_domain() { data_type: DataType::Integer(None), collation: Some(Ident::with_quote('"', "en_US")), default: Some(Expr::Value(test_utils::number("1").into())), - constraints: vec![TableConstraint::Check { + constraints: vec![CheckConstraint { name: None, expr: Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("VALUE"))), @@ -5643,7 +5647,8 @@ fn parse_create_domain() { right: Box::new(Expr::Value(test_utils::number("0").into())), }), enforced: None, - }], + } + .into()], }); assert_eq!(pg().verified_stmt(sql4), expected); @@ -5654,7 +5659,7 @@ fn parse_create_domain() { data_type: DataType::Integer(None), collation: None, default: None, - constraints: vec![TableConstraint::Check { + constraints: vec![CheckConstraint { name: Some(Ident::new("my_constraint")), expr: Box::new(Expr::BinaryOp { left: Box::new(Expr::Identifier(Ident::new("VALUE"))), @@ -5662,7 +5667,8 @@ fn parse_create_domain() { right: Box::new(Expr::Value(test_utils::number("0").into())), }), enforced: None, - }], + } + .into()], }); assert_eq!(pg().verified_stmt(sql5), expected); @@ -6467,7 +6473,7 @@ fn parse_alter_table_constraint_not_valid() { assert_eq!( operations, vec![AlterTableOperation::AddConstraint { - constraint: TableConstraint::ForeignKey { + constraint: ForeignKeyConstraint { name: Some("bar".into()), index_name: None, columns: vec!["baz".into()], @@ -6476,7 +6482,8 @@ fn parse_alter_table_constraint_not_valid() { on_delete: None, on_update: None, characteristics: None, - }, + } + .into(), not_valid: true, }] ); From 0fb3b6b11c0976731e95b00f1edfd5c57e658bde Mon Sep 17 00:00:00 2001 From: Andrew Harper Date: Wed, 8 Oct 2025 07:04:36 -0400 Subject: [PATCH 271/286] Add support for procedure parameter default values (#2041) --- src/ast/ddl.rs | 9 +++++- src/parser/mod.rs | 7 +++++ tests/sqlparser_common.rs | 58 ++++++++++++++++++++++++++++++++++++--- tests/sqlparser_mssql.rs | 6 ++++ 4 files changed, 75 insertions(+), 5 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index b7e020f5b..c6ab7ad10 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -1173,12 +1173,19 @@ pub struct ProcedureParam { pub name: Ident, pub data_type: DataType, pub mode: Option, + pub default: Option, } impl fmt::Display for ProcedureParam { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { if let Some(mode) = &self.mode { - write!(f, "{mode} {} {}", self.name, self.data_type) + if let Some(default) = &self.default { + write!(f, "{mode} {} {} = {}", self.name, self.data_type, default) + } else { + write!(f, "{mode} {} {}", self.name, self.data_type) + } + } else if let Some(default) = &self.default { + write!(f, "{} {} = {}", self.name, self.data_type, default) } else { write!(f, "{} {}", self.name, self.data_type) } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index deca3b577..d1e9b1e78 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7904,10 +7904,17 @@ impl<'a> Parser<'a> { }; let name = self.parse_identifier()?; let data_type = self.parse_data_type()?; + let default = if self.consume_token(&Token::Eq) { + Some(self.parse_expr()?) + } else { + None + }; + Ok(ProcedureParam { name, data_type, mode, + default, }) } diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 343ea4955..a1fd48d3e 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -16505,7 +16505,8 @@ fn parse_create_procedure_with_parameter_modes() { span: fake_span, }, data_type: DataType::Integer(None), - mode: Some(ArgMode::In) + mode: Some(ArgMode::In), + default: None, }, ProcedureParam { name: Ident { @@ -16514,7 +16515,8 @@ fn parse_create_procedure_with_parameter_modes() { span: fake_span, }, data_type: DataType::Text, - mode: Some(ArgMode::Out) + mode: Some(ArgMode::Out), + default: None, }, ProcedureParam { name: Ident { @@ -16523,7 +16525,8 @@ fn parse_create_procedure_with_parameter_modes() { span: fake_span, }, data_type: DataType::Timestamp(None, TimezoneInfo::None), - mode: Some(ArgMode::InOut) + mode: Some(ArgMode::InOut), + default: None, }, ProcedureParam { name: Ident { @@ -16532,13 +16535,60 @@ fn parse_create_procedure_with_parameter_modes() { span: fake_span, }, data_type: DataType::Bool, - mode: None + mode: None, + default: None, }, ]) ); } _ => unreachable!(), } + + // parameters with default values + let sql = r#"CREATE PROCEDURE test_proc (IN a INTEGER = 1, OUT b TEXT = '2', INOUT c TIMESTAMP = NULL, d BOOL = 0) AS BEGIN SELECT 1; END"#; + match verified_stmt(sql) { + Statement::CreateProcedure { + or_alter, + name, + params, + .. + } => { + assert_eq!(or_alter, false); + assert_eq!(name.to_string(), "test_proc"); + assert_eq!( + params, + Some(vec![ + ProcedureParam { + name: Ident::new("a"), + data_type: DataType::Integer(None), + mode: Some(ArgMode::In), + default: Some(Expr::Value((number("1")).with_empty_span())), + }, + ProcedureParam { + name: Ident::new("b"), + data_type: DataType::Text, + mode: Some(ArgMode::Out), + default: Some(Expr::Value( + Value::SingleQuotedString("2".into()).with_empty_span() + )), + }, + ProcedureParam { + name: Ident::new("c"), + data_type: DataType::Timestamp(None, TimezoneInfo::None), + mode: Some(ArgMode::InOut), + default: Some(Expr::Value(Value::Null.with_empty_span())), + }, + ProcedureParam { + name: Ident::new("d"), + data_type: DataType::Bool, + mode: None, + default: Some(Expr::Value((number("0")).with_empty_span())), + } + ]), + ); + } + _ => unreachable!(), + } } #[test] diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index b1ad422ec..181899170 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -156,6 +156,7 @@ fn parse_create_procedure() { }, data_type: DataType::Int(None), mode: None, + default: None, }, ProcedureParam { name: Ident { @@ -168,6 +169,7 @@ fn parse_create_procedure() { unit: None })), mode: None, + default: None, } ]), name: ObjectName::from(vec![Ident { @@ -196,6 +198,10 @@ fn parse_mssql_create_procedure() { let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN SELECT [foo], CASE WHEN [foo] IS NULL THEN 'empty' ELSE 'notempty' END AS [foo]; END"); // Multiple statements let _ = ms().verified_stmt("CREATE PROCEDURE [foo] AS BEGIN UPDATE bar SET col = 'test'; SELECT [foo] FROM BAR WHERE [FOO] > 10; END"); + + // parameters with default values + let sql = r#"CREATE PROCEDURE foo (IN @a INTEGER = 1, OUT @b TEXT = '2', INOUT @c DATETIME = NULL, @d BOOL = 0) AS BEGIN SELECT 1; END"#; + let _ = ms().verified_stmt(sql); } #[test] From 35220ee38c94eddd8a584520d8235d647edbc752 Mon Sep 17 00:00:00 2001 From: Dmitrii Blaginin Date: Fri, 10 Oct 2025 12:10:13 +0100 Subject: [PATCH 272/286] Require PR (#2052) --- .asf.yaml | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/.asf.yaml b/.asf.yaml index 0534705bc..08b31804d 100644 --- a/.asf.yaml +++ b/.asf.yaml @@ -50,3 +50,7 @@ github: - "test (beta)" - "test (nightly)" - "Release Audit Tool (RAT)" + pull_requests: + # enable updating head branches of pull requests + allow_update_branch: true + allow_auto_merge: true From e7d42f3d1a3fb6a542b70d8bef945db2c67bcf06 Mon Sep 17 00:00:00 2001 From: Joris Bayer Date: Fri, 10 Oct 2025 13:11:07 +0200 Subject: [PATCH 273/286] Increase version of sqlparser_derive from 0.3.0 to 0.4.0 (#2060) --- Cargo.toml | 2 +- derive/Cargo.toml | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/Cargo.toml b/Cargo.toml index 48db33b3f..ed94bbbdd 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -54,7 +54,7 @@ serde = { version = "1.0", default-features = false, features = ["derive", "allo # of dev-dependencies because of # https://github.com/rust-lang/cargo/issues/1596 serde_json = { version = "1.0", optional = true } -sqlparser_derive = { version = "0.3.0", path = "derive", optional = true } +sqlparser_derive = { version = "0.4.0", path = "derive", optional = true } [dev-dependencies] simple_logger = "5.0" diff --git a/derive/Cargo.toml b/derive/Cargo.toml index 7b6477300..549477041 100644 --- a/derive/Cargo.toml +++ b/derive/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser_derive" description = "Procedural (proc) macros for sqlparser" -version = "0.3.0" +version = "0.4.0" authors = ["sqlparser-rs authors"] homepage = "https://github.com/sqlparser-rs/sqlparser-rs" documentation = "https://docs.rs/sqlparser_derive/" From cc595cfd847f6e4b8edb6def67535e0a499b8845 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Fri, 10 Oct 2025 16:08:20 +0200 Subject: [PATCH 274/286] Moving several struct variants out of `Statement` enum to allow for trait impls for specific sub-variants (#2057) --- src/ast/dcl.rs | 113 +++++- src/ast/ddl.rs | 412 ++++++++++++++++++- src/ast/dml.rs | 65 ++- src/ast/mod.rs | 719 +++++++--------------------------- src/ast/spans.rs | 200 +++++----- src/parser/mod.rs | 48 ++- src/test_utils.rs | 23 +- tests/sqlparser_bigquery.rs | 12 +- tests/sqlparser_clickhouse.rs | 30 +- tests/sqlparser_common.rs | 111 +++--- tests/sqlparser_mssql.rs | 10 +- tests/sqlparser_mysql.rs | 38 +- tests/sqlparser_postgres.rs | 193 ++++----- tests/sqlparser_snowflake.rs | 10 +- tests/sqlparser_sqlite.rs | 10 +- 15 files changed, 1032 insertions(+), 962 deletions(-) diff --git a/src/ast/dcl.rs b/src/ast/dcl.rs index 079894075..d04875a73 100644 --- a/src/ast/dcl.rs +++ b/src/ast/dcl.rs @@ -28,8 +28,9 @@ use serde::{Deserialize, Serialize}; #[cfg(feature = "visitor")] use sqlparser_derive::{Visit, VisitMut}; -use super::{display_comma_separated, Expr, Ident, Password}; +use super::{display_comma_separated, Expr, Ident, Password, Spanned}; use crate::ast::{display_separated, ObjectName}; +use crate::tokenizer::Span; /// An option in `ROLE` statement. /// @@ -252,3 +253,113 @@ impl fmt::Display for SecondaryRoles { } } } + +/// CREATE ROLE statement +/// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-createrole.html) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateRole { + pub names: Vec, + pub if_not_exists: bool, + // Postgres + pub login: Option, + pub inherit: Option, + pub bypassrls: Option, + pub password: Option, + pub superuser: Option, + pub create_db: Option, + pub create_role: Option, + pub replication: Option, + pub connection_limit: Option, + pub valid_until: Option, + pub in_role: Vec, + pub in_group: Vec, + pub role: Vec, + pub user: Vec, + pub admin: Vec, + // MSSQL + pub authorization_owner: Option, +} + +impl fmt::Display for CreateRole { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "CREATE ROLE {if_not_exists}{names}{superuser}{create_db}{create_role}{inherit}{login}{replication}{bypassrls}", + if_not_exists = if self.if_not_exists { "IF NOT EXISTS " } else { "" }, + names = display_separated(&self.names, ", "), + superuser = match self.superuser { + Some(true) => " SUPERUSER", + Some(false) => " NOSUPERUSER", + None => "" + }, + create_db = match self.create_db { + Some(true) => " CREATEDB", + Some(false) => " NOCREATEDB", + None => "" + }, + create_role = match self.create_role { + Some(true) => " CREATEROLE", + Some(false) => " NOCREATEROLE", + None => "" + }, + inherit = match self.inherit { + Some(true) => " INHERIT", + Some(false) => " NOINHERIT", + None => "" + }, + login = match self.login { + Some(true) => " LOGIN", + Some(false) => " NOLOGIN", + None => "" + }, + replication = match self.replication { + Some(true) => " REPLICATION", + Some(false) => " NOREPLICATION", + None => "" + }, + bypassrls = match self.bypassrls { + Some(true) => " BYPASSRLS", + Some(false) => " NOBYPASSRLS", + None => "" + } + )?; + if let Some(limit) = &self.connection_limit { + write!(f, " CONNECTION LIMIT {limit}")?; + } + match &self.password { + Some(Password::Password(pass)) => write!(f, " PASSWORD {pass}")?, + Some(Password::NullPassword) => write!(f, " PASSWORD NULL")?, + None => {} + }; + if let Some(until) = &self.valid_until { + write!(f, " VALID UNTIL {until}")?; + } + if !self.in_role.is_empty() { + write!(f, " IN ROLE {}", display_comma_separated(&self.in_role))?; + } + if !self.in_group.is_empty() { + write!(f, " IN GROUP {}", display_comma_separated(&self.in_group))?; + } + if !self.role.is_empty() { + write!(f, " ROLE {}", display_comma_separated(&self.role))?; + } + if !self.user.is_empty() { + write!(f, " USER {}", display_comma_separated(&self.user))?; + } + if !self.admin.is_empty() { + write!(f, " ADMIN {}", display_comma_separated(&self.admin))?; + } + if let Some(owner) = &self.authorization_owner { + write!(f, " AUTHORIZATION {owner}")?; + } + Ok(()) + } +} + +impl Spanned for CreateRole { + fn span(&self) -> Span { + Span::empty() + } +} diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index c6ab7ad10..aafa8e644 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -19,7 +19,7 @@ //! (commonly referred to as Data Definition Language, or DDL) #[cfg(not(feature = "std"))] -use alloc::{boxed::Box, string::String, vec::Vec}; +use alloc::{boxed::Box, format, string::String, vec, vec::Vec}; use core::fmt::{self, Display, Write}; #[cfg(feature = "serde")] @@ -30,15 +30,16 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, table_constraints::TableConstraint, ArgMode, - CommentDef, ConditionalStatements, CreateFunctionBody, CreateFunctionUsing, - CreateTableLikeKind, CreateTableOptions, DataType, Expr, FileFormat, FunctionBehavior, - FunctionCalledOnNull, FunctionDeterminismSpecifier, FunctionParallel, HiveDistributionStyle, - HiveFormat, HiveIOFormat, HiveRowFormat, Ident, InitializeKind, MySQLColumnPosition, - ObjectName, OnCommit, OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, - Query, RefreshModeKind, RowAccessPolicy, SequenceOptions, Spanned, SqlOption, - StorageSerializationPolicy, TableVersion, Tag, TriggerEvent, TriggerExecBody, TriggerObject, - TriggerPeriod, TriggerReferencing, Value, ValueWithSpan, WrappedCollection, + display_comma_separated, display_separated, ArgMode, AttachedToken, CommentDef, + ConditionalStatements, CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, + CreateTableOptions, CreateViewParams, DataType, Expr, FileFormat, FunctionBehavior, + FunctionCalledOnNull, FunctionDesc, FunctionDeterminismSpecifier, FunctionParallel, + HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, HiveSetLocation, Ident, + InitializeKind, MySQLColumnPosition, ObjectName, OnCommit, OneOrManyWithParens, + OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RefreshModeKind, RowAccessPolicy, + SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, TableConstraint, TableVersion, + Tag, TriggerEvent, TriggerExecBody, TriggerObject, TriggerPeriod, TriggerReferencing, Value, + ValueWithSpan, WrappedCollection, }; use crate::display_utils::{DisplayCommaSeparated, Indent, NewLine, SpaceOrNewline}; use crate::keywords::Keyword; @@ -3138,3 +3139,394 @@ impl fmt::Display for DropTrigger { Ok(()) } } + +/// A `TRUNCATE` statement. +/// +/// ```sql +/// TRUNCATE TABLE table_names [PARTITION (partitions)] [RESTART IDENTITY | CONTINUE IDENTITY] [CASCADE | RESTRICT] [ON CLUSTER cluster_name] +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct Truncate { + /// Table names to truncate + pub table_names: Vec, + /// Optional partition specification + pub partitions: Option>, + /// TABLE - optional keyword + pub table: bool, + /// Postgres-specific option: [ RESTART IDENTITY | CONTINUE IDENTITY ] + pub identity: Option, + /// Postgres-specific option: [ CASCADE | RESTRICT ] + pub cascade: Option, + /// ClickHouse-specific option: [ ON CLUSTER cluster_name ] + /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/truncate/) + pub on_cluster: Option, +} + +impl fmt::Display for Truncate { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + let table = if self.table { "TABLE " } else { "" }; + + write!( + f, + "TRUNCATE {table}{table_names}", + table_names = display_comma_separated(&self.table_names) + )?; + + if let Some(identity) = &self.identity { + match identity { + super::TruncateIdentityOption::Restart => write!(f, " RESTART IDENTITY")?, + super::TruncateIdentityOption::Continue => write!(f, " CONTINUE IDENTITY")?, + } + } + if let Some(cascade) = &self.cascade { + match cascade { + super::CascadeOption::Cascade => write!(f, " CASCADE")?, + super::CascadeOption::Restrict => write!(f, " RESTRICT")?, + } + } + + if let Some(ref parts) = &self.partitions { + if !parts.is_empty() { + write!(f, " PARTITION ({})", display_comma_separated(parts))?; + } + } + if let Some(on_cluster) = &self.on_cluster { + write!(f, " ON CLUSTER {on_cluster}")?; + } + Ok(()) + } +} + +impl Spanned for Truncate { + fn span(&self) -> Span { + Span::union_iter( + self.table_names.iter().map(|i| i.name.span()).chain( + self.partitions + .iter() + .flat_map(|i| i.iter().map(|k| k.span())), + ), + ) + } +} + +/// An `MSCK` statement. +/// +/// ```sql +/// MSCK [REPAIR] TABLE table_name [ADD|DROP|SYNC PARTITIONS] +/// ``` +/// MSCK (Hive) - MetaStore Check command +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct Msck { + /// Table name to check + #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] + pub table_name: ObjectName, + /// Whether to repair the table + pub repair: bool, + /// Partition action (ADD, DROP, or SYNC) + pub partition_action: Option, +} + +impl fmt::Display for Msck { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "MSCK {repair}TABLE {table}", + repair = if self.repair { "REPAIR " } else { "" }, + table = self.table_name + )?; + if let Some(pa) = &self.partition_action { + write!(f, " {pa}")?; + } + Ok(()) + } +} + +impl Spanned for Msck { + fn span(&self) -> Span { + self.table_name.span() + } +} + +/// CREATE VIEW statement. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateView { + /// True if this is a `CREATE OR ALTER VIEW` statement + /// + /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-view-transact-sql) + pub or_alter: bool, + pub or_replace: bool, + pub materialized: bool, + /// Snowflake: SECURE view modifier + /// + pub secure: bool, + /// View name + pub name: ObjectName, + /// If `if_not_exists` is true, this flag is set to true if the view name comes before the `IF NOT EXISTS` clause. + /// Example: + /// ```sql + /// CREATE VIEW myview IF NOT EXISTS AS SELECT 1` + /// ``` + /// Otherwise, the flag is set to false if the view name comes after the clause + /// Example: + /// ```sql + /// CREATE VIEW IF NOT EXISTS myview AS SELECT 1` + /// ``` + pub name_before_not_exists: bool, + pub columns: Vec, + pub query: Box, + pub options: CreateTableOptions, + pub cluster_by: Vec, + /// Snowflake: Views can have comments in Snowflake. + /// + pub comment: Option, + /// if true, has RedShift [`WITH NO SCHEMA BINDING`] clause + pub with_no_schema_binding: bool, + /// if true, has SQLite `IF NOT EXISTS` clause + pub if_not_exists: bool, + /// if true, has SQLite `TEMP` or `TEMPORARY` clause + pub temporary: bool, + /// if not None, has Clickhouse `TO` clause, specify the table into which to insert results + /// + pub to: Option, + /// MySQL: Optional parameters for the view algorithm, definer, and security context + pub params: Option, +} + +impl fmt::Display for CreateView { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "CREATE {or_alter}{or_replace}", + or_alter = if self.or_alter { "OR ALTER " } else { "" }, + or_replace = if self.or_replace { "OR REPLACE " } else { "" }, + )?; + if let Some(ref params) = self.params { + params.fmt(f)?; + } + write!( + f, + "{secure}{materialized}{temporary}VIEW {if_not_and_name}{to}", + if_not_and_name = if self.if_not_exists { + if self.name_before_not_exists { + format!("{} IF NOT EXISTS", self.name) + } else { + format!("IF NOT EXISTS {}", self.name) + } + } else { + format!("{}", self.name) + }, + secure = if self.secure { "SECURE " } else { "" }, + materialized = if self.materialized { + "MATERIALIZED " + } else { + "" + }, + temporary = if self.temporary { "TEMPORARY " } else { "" }, + to = self + .to + .as_ref() + .map(|to| format!(" TO {to}")) + .unwrap_or_default() + )?; + if !self.columns.is_empty() { + write!(f, " ({})", display_comma_separated(&self.columns))?; + } + if matches!(self.options, CreateTableOptions::With(_)) { + write!(f, " {}", self.options)?; + } + if let Some(ref comment) = self.comment { + write!(f, " COMMENT = '{}'", escape_single_quote_string(comment))?; + } + if !self.cluster_by.is_empty() { + write!( + f, + " CLUSTER BY ({})", + display_comma_separated(&self.cluster_by) + )?; + } + if matches!(self.options, CreateTableOptions::Options(_)) { + write!(f, " {}", self.options)?; + } + f.write_str(" AS")?; + SpaceOrNewline.fmt(f)?; + self.query.fmt(f)?; + if self.with_no_schema_binding { + write!(f, " WITH NO SCHEMA BINDING")?; + } + Ok(()) + } +} + +/// CREATE EXTENSION statement +/// Note: this is a PostgreSQL-specific statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateExtension { + pub name: Ident, + pub if_not_exists: bool, + pub cascade: bool, + pub schema: Option, + pub version: Option, +} + +impl fmt::Display for CreateExtension { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "CREATE EXTENSION {if_not_exists}{name}", + if_not_exists = if self.if_not_exists { + "IF NOT EXISTS " + } else { + "" + }, + name = self.name + )?; + if self.cascade || self.schema.is_some() || self.version.is_some() { + write!(f, " WITH")?; + + if let Some(name) = &self.schema { + write!(f, " SCHEMA {name}")?; + } + if let Some(version) = &self.version { + write!(f, " VERSION {version}")?; + } + if self.cascade { + write!(f, " CASCADE")?; + } + } + + Ok(()) + } +} + +impl Spanned for CreateExtension { + fn span(&self) -> Span { + Span::empty() + } +} + +/// DROP EXTENSION statement +/// Note: this is a PostgreSQL-specific statement +/// +/// # References +/// +/// PostgreSQL Documentation: +/// +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct DropExtension { + pub names: Vec, + pub if_exists: bool, + /// `CASCADE` or `RESTRICT` + pub cascade_or_restrict: Option, +} + +impl fmt::Display for DropExtension { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "DROP EXTENSION")?; + if self.if_exists { + write!(f, " IF EXISTS")?; + } + write!(f, " {}", display_comma_separated(&self.names))?; + if let Some(cascade_or_restrict) = &self.cascade_or_restrict { + write!(f, " {cascade_or_restrict}")?; + } + Ok(()) + } +} + +impl Spanned for DropExtension { + fn span(&self) -> Span { + Span::empty() + } +} + +/// ALTER TABLE statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct AlterTable { + /// Table name + #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] + pub name: ObjectName, + pub if_exists: bool, + pub only: bool, + pub operations: Vec, + pub location: Option, + /// ClickHouse dialect supports `ON CLUSTER` clause for ALTER TABLE + /// For example: `ALTER TABLE table_name ON CLUSTER cluster_name ADD COLUMN c UInt32` + /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/alter/update) + pub on_cluster: Option, + /// Snowflake "ICEBERG" clause for Iceberg tables + /// + pub iceberg: bool, + /// Token that represents the end of the statement (semicolon or EOF) + pub end_token: AttachedToken, +} + +impl fmt::Display for AlterTable { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + if self.iceberg { + write!(f, "ALTER ICEBERG TABLE ")?; + } else { + write!(f, "ALTER TABLE ")?; + } + + if self.if_exists { + write!(f, "IF EXISTS ")?; + } + if self.only { + write!(f, "ONLY ")?; + } + write!(f, "{} ", &self.name)?; + if let Some(cluster) = &self.on_cluster { + write!(f, "ON CLUSTER {cluster} ")?; + } + write!(f, "{}", display_comma_separated(&self.operations))?; + if let Some(loc) = &self.location { + write!(f, " {loc}")? + } + Ok(()) + } +} + +/// DROP FUNCTION statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct DropFunction { + pub if_exists: bool, + /// One or more functions to drop + pub func_desc: Vec, + /// `CASCADE` or `RESTRICT` + pub drop_behavior: Option, +} + +impl fmt::Display for DropFunction { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "DROP FUNCTION{} {}", + if self.if_exists { " IF EXISTS" } else { "" }, + display_comma_separated(&self.func_desc), + )?; + if let Some(op) = &self.drop_behavior { + write!(f, " {op}")?; + } + Ok(()) + } +} + +impl Spanned for DropFunction { + fn span(&self) -> Span { + Span::empty() + } +} diff --git a/src/ast/dml.rs b/src/ast/dml.rs index e4d99bcfc..c0bfcb19f 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -29,7 +29,7 @@ use crate::display_utils::{indented_list, Indent, SpaceOrNewline}; use super::{ display_comma_separated, query::InputFormatClause, Assignment, Expr, FromTable, Ident, InsertAliases, MysqlInsertPriority, ObjectName, OnInsert, OrderByExpr, Query, SelectItem, - Setting, SqliteOnConflict, TableObject, TableWithJoins, + Setting, SqliteOnConflict, TableObject, TableWithJoins, UpdateTableFromKind, }; /// INSERT statement. @@ -240,3 +240,66 @@ impl Display for Delete { Ok(()) } } + +/// UPDATE statement. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct Update { + /// TABLE + pub table: TableWithJoins, + /// Column assignments + pub assignments: Vec, + /// Table which provide value to be set + pub from: Option, + /// WHERE + pub selection: Option, + /// RETURNING + pub returning: Option>, + /// SQLite-specific conflict resolution clause + pub or: Option, + /// LIMIT + pub limit: Option, +} + +impl Display for Update { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.write_str("UPDATE ")?; + if let Some(or) = &self.or { + or.fmt(f)?; + f.write_str(" ")?; + } + self.table.fmt(f)?; + if let Some(UpdateTableFromKind::BeforeSet(from)) = &self.from { + SpaceOrNewline.fmt(f)?; + f.write_str("FROM")?; + indented_list(f, from)?; + } + if !self.assignments.is_empty() { + SpaceOrNewline.fmt(f)?; + f.write_str("SET")?; + indented_list(f, &self.assignments)?; + } + if let Some(UpdateTableFromKind::AfterSet(from)) = &self.from { + SpaceOrNewline.fmt(f)?; + f.write_str("FROM")?; + indented_list(f, from)?; + } + if let Some(selection) = &self.selection { + SpaceOrNewline.fmt(f)?; + f.write_str("WHERE")?; + SpaceOrNewline.fmt(f)?; + Indent(selection).fmt(f)?; + } + if let Some(returning) = &self.returning { + SpaceOrNewline.fmt(f)?; + f.write_str("RETURNING")?; + indented_list(f, returning)?; + } + if let Some(limit) = &self.limit { + SpaceOrNewline.fmt(f)?; + write!(f, "LIMIT {limit}")?; + } + Ok(()) + } +} diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 52968d7db..0c83b3203 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -43,7 +43,7 @@ use serde::{Deserialize, Serialize}; use sqlparser_derive::{Visit, VisitMut}; use crate::{ - display_utils::{indented_list, SpaceOrNewline}, + display_utils::SpaceOrNewline, tokenizer::{Span, Token}, }; use crate::{ @@ -56,22 +56,24 @@ pub use self::data_type::{ ExactNumberInfo, IntervalFields, StructBracketKind, TimezoneInfo, }; pub use self::dcl::{ - AlterRoleOperation, ResetConfig, RoleOption, SecondaryRoles, SetConfigValue, Use, + AlterRoleOperation, CreateRole, ResetConfig, RoleOption, SecondaryRoles, SetConfigValue, Use, }; pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, - AlterSchema, AlterSchemaOperation, AlterTableAlgorithm, AlterTableLock, AlterTableOperation, - AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, - AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, - ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, - CreateFunction, CreateIndex, CreateTable, CreateTrigger, Deduplicate, DeferrableInitial, - DropBehavior, DropTrigger, GeneratedAs, GeneratedExpressionMode, IdentityParameters, - IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, - IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, - ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity, TagsColumnOption, - UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, + AlterSchema, AlterSchemaOperation, AlterTable, AlterTableAlgorithm, AlterTableLock, + AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, + AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, + ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, ColumnPolicyProperty, + ConstraintCharacteristics, CreateConnector, CreateDomain, CreateExtension, CreateFunction, + CreateIndex, CreateTable, CreateTrigger, CreateView, Deduplicate, DeferrableInitial, + DropBehavior, DropExtension, DropFunction, DropTrigger, GeneratedAs, GeneratedExpressionMode, + IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, + IdentityPropertyOrder, IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, Msck, + NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, RenameTableNameKind, + ReplicaIdentity, TagsColumnOption, Truncate, UserDefinedTypeCompositeAttributeDef, + UserDefinedTypeRepresentation, ViewColumnDef, }; -pub use self::dml::{Delete, Insert}; +pub use self::dml::{Delete, Insert, Update}; pub use self::operator::{BinaryOperator, UnaryOperator}; pub use self::query::{ AfterMatchSkip, ConnectBy, Cte, CteAsMaterialized, Distinct, EmptyMatchesMode, @@ -3064,6 +3066,59 @@ impl Display for ExceptionWhen { } } +/// ANALYZE TABLE statement (Hive-specific) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct Analyze { + #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] + pub table_name: ObjectName, + pub partitions: Option>, + pub for_columns: bool, + pub columns: Vec, + pub cache_metadata: bool, + pub noscan: bool, + pub compute_statistics: bool, + pub has_table_keyword: bool, +} + +impl fmt::Display for Analyze { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "ANALYZE{}{table_name}", + if self.has_table_keyword { + " TABLE " + } else { + " " + }, + table_name = self.table_name + )?; + if let Some(ref parts) = self.partitions { + if !parts.is_empty() { + write!(f, " PARTITION ({})", display_comma_separated(parts))?; + } + } + + if self.compute_statistics { + write!(f, " COMPUTE STATISTICS")?; + } + if self.noscan { + write!(f, " NOSCAN")?; + } + if self.cache_metadata { + write!(f, " CACHE METADATA")?; + } + if self.for_columns { + write!(f, " FOR COLUMNS")?; + if !self.columns.is_empty() { + write!(f, " {}", display_comma_separated(&self.columns))?; + } + } + Ok(()) + } +} + /// A top-level statement (SELECT, INSERT, CREATE, etc.) #[allow(clippy::large_enum_variant)] #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -3078,49 +3133,18 @@ pub enum Statement { /// ANALYZE /// ``` /// Analyze (Hive) - Analyze { - #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] - table_name: ObjectName, - partitions: Option>, - for_columns: bool, - columns: Vec, - cache_metadata: bool, - noscan: bool, - compute_statistics: bool, - has_table_keyword: bool, - }, + Analyze(Analyze), Set(Set), /// ```sql /// TRUNCATE /// ``` /// Truncate (Hive) - Truncate { - table_names: Vec, - partitions: Option>, - /// TABLE - optional keyword; - table: bool, - /// Postgres-specific option - /// [ RESTART IDENTITY | CONTINUE IDENTITY ] - identity: Option, - /// Postgres-specific option - /// [ CASCADE | RESTRICT ] - cascade: Option, - /// ClickHouse-specific option - /// [ ON CLUSTER cluster_name ] - /// - /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/truncate/) - on_cluster: Option, - }, + Truncate(Truncate), /// ```sql /// MSCK /// ``` /// Msck (Hive) - Msck { - #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] - table_name: ObjectName, - repair: bool, - partition_action: Option, - }, + Msck(Msck), /// ```sql /// SELECT /// ``` @@ -3223,22 +3247,7 @@ pub enum Statement { /// ```sql /// UPDATE /// ``` - Update { - /// TABLE - table: TableWithJoins, - /// Column assignments - assignments: Vec, - /// Table which provide value to be set - from: Option, - /// WHERE - selection: Option, - /// RETURNING - returning: Option>, - /// SQLite-specific conflict resolution clause - or: Option, - /// LIMIT - limit: Option, - }, + Update(Update), /// ```sql /// DELETE /// ``` @@ -3246,48 +3255,7 @@ pub enum Statement { /// ```sql /// CREATE VIEW /// ``` - CreateView { - /// True if this is a `CREATE OR ALTER VIEW` statement - /// - /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-view-transact-sql) - or_alter: bool, - or_replace: bool, - materialized: bool, - /// Snowflake: SECURE view modifier - /// - secure: bool, - /// View name - name: ObjectName, - /// If `if_not_exists` is true, this flag is set to true if the view name comes before the `IF NOT EXISTS` clause. - /// Example: - /// ```sql - /// CREATE VIEW myview IF NOT EXISTS AS SELECT 1` - /// ``` - /// Otherwise, the flag is set to false if the view name comes after the clause - /// Example: - /// ```sql - /// CREATE VIEW IF NOT EXISTS myview AS SELECT 1` - /// ``` - name_before_not_exists: bool, - columns: Vec, - query: Box, - options: CreateTableOptions, - cluster_by: Vec, - /// Snowflake: Views can have comments in Snowflake. - /// - comment: Option, - /// if true, has RedShift [`WITH NO SCHEMA BINDING`] clause - with_no_schema_binding: bool, - /// if true, has SQLite `IF NOT EXISTS` clause - if_not_exists: bool, - /// if true, has SQLite `TEMP` or `TEMPORARY` clause - temporary: bool, - /// if not None, has Clickhouse `TO` clause, specify the table into which to insert results - /// - to: Option, - /// MySQL: Optional parameters for the view algorithm, definer, and security context - params: Option, - }, + CreateView(CreateView), /// ```sql /// CREATE TABLE /// ``` @@ -3311,28 +3279,7 @@ pub enum Statement { /// CREATE ROLE /// ``` /// See [PostgreSQL](https://www.postgresql.org/docs/current/sql-createrole.html) - CreateRole { - names: Vec, - if_not_exists: bool, - // Postgres - login: Option, - inherit: Option, - bypassrls: Option, - password: Option, - superuser: Option, - create_db: Option, - create_role: Option, - replication: Option, - connection_limit: Option, - valid_until: Option, - in_role: Vec, - in_group: Vec, - role: Vec, - user: Vec, - admin: Vec, - // MSSQL - authorization_owner: Option, - }, + CreateRole(CreateRole), /// ```sql /// CREATE SECRET /// ``` @@ -3370,24 +3317,7 @@ pub enum Statement { /// ```sql /// ALTER TABLE /// ``` - AlterTable { - /// Table name - #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] - name: ObjectName, - if_exists: bool, - only: bool, - operations: Vec, - location: Option, - /// ClickHouse dialect supports `ON CLUSTER` clause for ALTER TABLE - /// For example: `ALTER TABLE table_name ON CLUSTER cluster_name ADD COLUMN c UInt32` - /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/alter/update) - on_cluster: Option, - /// Snowflake "ICEBERG" clause for Iceberg tables - /// - iceberg: bool, - /// Token that represents the end of the statement (semicolon or EOF) - end_token: AttachedToken, - }, + AlterTable(AlterTable), /// ```sql /// ALTER SCHEMA /// ``` @@ -3523,13 +3453,7 @@ pub enum Statement { /// ```sql /// DROP FUNCTION /// ``` - DropFunction { - if_exists: bool, - /// One or more function to drop - func_desc: Vec, - /// `CASCADE` or `RESTRICT` - drop_behavior: Option, - }, + DropFunction(DropFunction), /// ```sql /// DROP DOMAIN /// ``` @@ -3593,25 +3517,13 @@ pub enum Statement { /// ``` /// /// Note: this is a PostgreSQL-specific statement, - CreateExtension { - name: Ident, - if_not_exists: bool, - cascade: bool, - schema: Option, - version: Option, - }, + CreateExtension(CreateExtension), /// ```sql /// DROP EXTENSION [ IF EXISTS ] name [, ...] [ CASCADE | RESTRICT ] - /// - /// Note: this is a PostgreSQL-specific statement. - /// https://www.postgresql.org/docs/current/sql-dropextension.html /// ``` - DropExtension { - names: Vec, - if_exists: bool, - /// `CASCADE` or `RESTRICT` - cascade_or_restrict: Option, - }, + /// Note: this is a PostgreSQL-specific statement. + /// + DropExtension(DropExtension), /// ```sql /// FETCH /// ``` @@ -4328,6 +4240,24 @@ pub enum Statement { Vacuum(VacuumStatement), } +impl From for Statement { + fn from(analyze: Analyze) -> Self { + Statement::Analyze(analyze) + } +} + +impl From for Statement { + fn from(truncate: ddl::Truncate) -> Self { + Statement::Truncate(truncate) + } +} + +impl From for Statement { + fn from(msck: ddl::Msck) -> Self { + Statement::Msck(msck) + } +} + /// ```sql /// {COPY | REVOKE} CURRENT GRANTS /// ``` @@ -4528,61 +4458,8 @@ impl fmt::Display for Statement { } write!(f, " {source}") } - Statement::Msck { - table_name, - repair, - partition_action, - } => { - write!( - f, - "MSCK {repair}TABLE {table}", - repair = if *repair { "REPAIR " } else { "" }, - table = table_name - )?; - if let Some(pa) = partition_action { - write!(f, " {pa}")?; - } - Ok(()) - } - Statement::Truncate { - table_names, - partitions, - table, - identity, - cascade, - on_cluster, - } => { - let table = if *table { "TABLE " } else { "" }; - - write!( - f, - "TRUNCATE {table}{table_names}", - table_names = display_comma_separated(table_names) - )?; - - if let Some(identity) = identity { - match identity { - TruncateIdentityOption::Restart => write!(f, " RESTART IDENTITY")?, - TruncateIdentityOption::Continue => write!(f, " CONTINUE IDENTITY")?, - } - } - if let Some(cascade) = cascade { - match cascade { - CascadeOption::Cascade => write!(f, " CASCADE")?, - CascadeOption::Restrict => write!(f, " RESTRICT")?, - } - } - - if let Some(ref parts) = partitions { - if !parts.is_empty() { - write!(f, " PARTITION ({})", display_comma_separated(parts))?; - } - } - if let Some(on_cluster) = on_cluster { - write!(f, " ON CLUSTER {on_cluster}")?; - } - Ok(()) - } + Statement::Msck(msck) => msck.fmt(f), + Statement::Truncate(truncate) => truncate.fmt(f), Statement::Case(stmt) => { write!(f, "{stmt}") } @@ -4637,44 +4514,7 @@ impl fmt::Display for Statement { )?; Ok(()) } - Statement::Analyze { - table_name, - partitions, - for_columns, - columns, - cache_metadata, - noscan, - compute_statistics, - has_table_keyword, - } => { - write!( - f, - "ANALYZE{}{table_name}", - if *has_table_keyword { " TABLE " } else { " " } - )?; - if let Some(ref parts) = partitions { - if !parts.is_empty() { - write!(f, " PARTITION ({})", display_comma_separated(parts))?; - } - } - - if *compute_statistics { - write!(f, " COMPUTE STATISTICS")?; - } - if *noscan { - write!(f, " NOSCAN")?; - } - if *cache_metadata { - write!(f, " CACHE METADATA")?; - } - if *for_columns { - write!(f, " FOR COLUMNS")?; - if !columns.is_empty() { - write!(f, " {}", display_comma_separated(columns))?; - } - } - Ok(()) - } + Statement::Analyze(analyze) => analyze.fmt(f), Statement::Insert(insert) => insert.fmt(f), Statement::Install { extension_name: name, @@ -4730,53 +4570,7 @@ impl fmt::Display for Statement { } Ok(()) } - Statement::Update { - table, - assignments, - from, - selection, - returning, - or, - limit, - } => { - f.write_str("UPDATE ")?; - if let Some(or) = or { - or.fmt(f)?; - f.write_str(" ")?; - } - table.fmt(f)?; - if let Some(UpdateTableFromKind::BeforeSet(from)) = from { - SpaceOrNewline.fmt(f)?; - f.write_str("FROM")?; - indented_list(f, from)?; - } - if !assignments.is_empty() { - SpaceOrNewline.fmt(f)?; - f.write_str("SET")?; - indented_list(f, assignments)?; - } - if let Some(UpdateTableFromKind::AfterSet(from)) = from { - SpaceOrNewline.fmt(f)?; - f.write_str("FROM")?; - indented_list(f, from)?; - } - if let Some(selection) = selection { - SpaceOrNewline.fmt(f)?; - f.write_str("WHERE")?; - SpaceOrNewline.fmt(f)?; - Indent(selection).fmt(f)?; - } - if let Some(returning) = returning { - SpaceOrNewline.fmt(f)?; - f.write_str("RETURNING")?; - indented_list(f, returning)?; - } - if let Some(limit) = limit { - SpaceOrNewline.fmt(f)?; - write!(f, "LIMIT {limit}")?; - } - Ok(()) - } + Statement::Update(update) => update.fmt(f), Statement::Delete(delete) => delete.fmt(f), Statement::Open(open) => open.fmt(f), Statement::Close { cursor } => { @@ -4932,80 +4726,7 @@ impl fmt::Display for Statement { } Ok(()) } - Statement::CreateView { - or_alter, - name, - or_replace, - columns, - query, - materialized, - secure, - options, - cluster_by, - comment, - with_no_schema_binding, - if_not_exists, - temporary, - to, - params, - name_before_not_exists, - } => { - write!( - f, - "CREATE {or_alter}{or_replace}", - or_alter = if *or_alter { "OR ALTER " } else { "" }, - or_replace = if *or_replace { "OR REPLACE " } else { "" }, - )?; - if let Some(params) = params { - params.fmt(f)?; - } - write!( - f, - "{secure}{materialized}{temporary}VIEW {if_not_and_name}{to}", - if_not_and_name = if *if_not_exists { - if *name_before_not_exists { - format!("{name} IF NOT EXISTS") - } else { - format!("IF NOT EXISTS {name}") - } - } else { - format!("{name}") - }, - secure = if *secure { "SECURE " } else { "" }, - materialized = if *materialized { "MATERIALIZED " } else { "" }, - temporary = if *temporary { "TEMPORARY " } else { "" }, - to = to - .as_ref() - .map(|to| format!(" TO {to}")) - .unwrap_or_default() - )?; - if !columns.is_empty() { - write!(f, " ({})", display_comma_separated(columns))?; - } - if matches!(options, CreateTableOptions::With(_)) { - write!(f, " {options}")?; - } - if let Some(comment) = comment { - write!( - f, - " COMMENT = '{}'", - value::escape_single_quote_string(comment) - )?; - } - if !cluster_by.is_empty() { - write!(f, " CLUSTER BY ({})", display_comma_separated(cluster_by))?; - } - if matches!(options, CreateTableOptions::Options(_)) { - write!(f, " {options}")?; - } - f.write_str(" AS")?; - SpaceOrNewline.fmt(f)?; - query.fmt(f)?; - if *with_no_schema_binding { - write!(f, " WITH NO SCHEMA BINDING")?; - } - Ok(()) - } + Statement::CreateView(create_view) => create_view.fmt(f), Statement::CreateTable(create_table) => create_table.fmt(f), Statement::LoadData { local, @@ -5056,141 +4777,9 @@ impl fmt::Display for Statement { Ok(()) } Statement::CreateIndex(create_index) => create_index.fmt(f), - Statement::CreateExtension { - name, - if_not_exists, - cascade, - schema, - version, - } => { - write!( - f, - "CREATE EXTENSION {if_not_exists}{name}", - if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" } - )?; - if *cascade || schema.is_some() || version.is_some() { - write!(f, " WITH")?; - - if let Some(name) = schema { - write!(f, " SCHEMA {name}")?; - } - if let Some(version) = version { - write!(f, " VERSION {version}")?; - } - if *cascade { - write!(f, " CASCADE")?; - } - } - - Ok(()) - } - Statement::DropExtension { - names, - if_exists, - cascade_or_restrict, - } => { - write!(f, "DROP EXTENSION")?; - if *if_exists { - write!(f, " IF EXISTS")?; - } - write!(f, " {}", display_comma_separated(names))?; - if let Some(cascade_or_restrict) = cascade_or_restrict { - write!(f, " {cascade_or_restrict}")?; - } - Ok(()) - } - Statement::CreateRole { - names, - if_not_exists, - inherit, - login, - bypassrls, - password, - create_db, - create_role, - superuser, - replication, - connection_limit, - valid_until, - in_role, - in_group, - role, - user, - admin, - authorization_owner, - } => { - write!( - f, - "CREATE ROLE {if_not_exists}{names}{superuser}{create_db}{create_role}{inherit}{login}{replication}{bypassrls}", - if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, - names = display_separated(names, ", "), - superuser = match *superuser { - Some(true) => " SUPERUSER", - Some(false) => " NOSUPERUSER", - None => "" - }, - create_db = match *create_db { - Some(true) => " CREATEDB", - Some(false) => " NOCREATEDB", - None => "" - }, - create_role = match *create_role { - Some(true) => " CREATEROLE", - Some(false) => " NOCREATEROLE", - None => "" - }, - inherit = match *inherit { - Some(true) => " INHERIT", - Some(false) => " NOINHERIT", - None => "" - }, - login = match *login { - Some(true) => " LOGIN", - Some(false) => " NOLOGIN", - None => "" - }, - replication = match *replication { - Some(true) => " REPLICATION", - Some(false) => " NOREPLICATION", - None => "" - }, - bypassrls = match *bypassrls { - Some(true) => " BYPASSRLS", - Some(false) => " NOBYPASSRLS", - None => "" - } - )?; - if let Some(limit) = connection_limit { - write!(f, " CONNECTION LIMIT {limit}")?; - } - match password { - Some(Password::Password(pass)) => write!(f, " PASSWORD {pass}"), - Some(Password::NullPassword) => write!(f, " PASSWORD NULL"), - None => Ok(()), - }?; - if let Some(until) = valid_until { - write!(f, " VALID UNTIL {until}")?; - } - if !in_role.is_empty() { - write!(f, " IN ROLE {}", display_comma_separated(in_role))?; - } - if !in_group.is_empty() { - write!(f, " IN GROUP {}", display_comma_separated(in_group))?; - } - if !role.is_empty() { - write!(f, " ROLE {}", display_comma_separated(role))?; - } - if !user.is_empty() { - write!(f, " USER {}", display_comma_separated(user))?; - } - if !admin.is_empty() { - write!(f, " ADMIN {}", display_comma_separated(admin))?; - } - if let Some(owner) = authorization_owner { - write!(f, " AUTHORIZATION {owner}")?; - } - Ok(()) - } + Statement::CreateExtension(create_extension) => write!(f, "{create_extension}"), + Statement::DropExtension(drop_extension) => write!(f, "{drop_extension}"), + Statement::CreateRole(create_role) => write!(f, "{create_role}"), Statement::CreateSecret { or_replace, temporary, @@ -5272,42 +4861,7 @@ impl fmt::Display for Statement { Ok(()) } Statement::CreateConnector(create_connector) => create_connector.fmt(f), - Statement::AlterTable { - name, - if_exists, - only, - operations, - location, - on_cluster, - iceberg, - end_token: _, - } => { - if *iceberg { - write!(f, "ALTER ICEBERG TABLE ")?; - } else { - write!(f, "ALTER TABLE ")?; - } - - if *if_exists { - write!(f, "IF EXISTS ")?; - } - if *only { - write!(f, "ONLY ")?; - } - write!(f, "{name} ")?; - if let Some(cluster) = on_cluster { - write!(f, "ON CLUSTER {cluster} ")?; - } - write!( - f, - "{operations}", - operations = display_comma_separated(operations) - )?; - if let Some(loc) = location { - write!(f, " {loc}")? - } - Ok(()) - } + Statement::AlterTable(alter_table) => write!(f, "{alter_table}"), Statement::AlterIndex { name, operation } => { write!(f, "ALTER INDEX {name} {operation}") } @@ -5410,22 +4964,7 @@ impl fmt::Display for Statement { }; Ok(()) } - Statement::DropFunction { - if_exists, - func_desc, - drop_behavior, - } => { - write!( - f, - "DROP FUNCTION{} {}", - if *if_exists { " IF EXISTS" } else { "" }, - display_comma_separated(func_desc), - )?; - if let Some(op) = drop_behavior { - write!(f, " {op}")?; - } - Ok(()) - } + Statement::DropFunction(drop_function) => write!(f, "{drop_function}"), Statement::DropDomain(DropDomain { if_exists, name, @@ -10945,6 +10484,48 @@ impl From for Statement { } } +impl From for Statement { + fn from(u: Update) -> Self { + Self::Update(u) + } +} + +impl From for Statement { + fn from(cv: CreateView) -> Self { + Self::CreateView(cv) + } +} + +impl From for Statement { + fn from(cr: CreateRole) -> Self { + Self::CreateRole(cr) + } +} + +impl From for Statement { + fn from(at: AlterTable) -> Self { + Self::AlterTable(at) + } +} + +impl From for Statement { + fn from(df: DropFunction) -> Self { + Self::DropFunction(df) + } +} + +impl From for Statement { + fn from(ce: CreateExtension) -> Self { + Self::CreateExtension(ce) + } +} + +impl From for Statement { + fn from(de: DropExtension) -> Self { + Self::DropExtension(de) + } +} + impl From for Statement { fn from(c: CaseStatement) -> Self { Self::Case(c) diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 0a303fcfd..de3439cfc 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -16,8 +16,8 @@ // under the License. use crate::ast::{ - ddl::AlterSchema, query::SelectItemQualifiedWildcardKind, AlterSchemaOperation, ColumnOptions, - ExportData, Owner, TypedString, + ddl::AlterSchema, query::SelectItemQualifiedWildcardKind, AlterSchemaOperation, AlterTable, + ColumnOptions, CreateView, ExportData, Owner, TypedString, }; use core::iter; @@ -25,23 +25,23 @@ use crate::tokenizer::Span; use super::{ dcl::SecondaryRoles, value::ValueWithSpan, AccessExpr, AlterColumnOperation, - AlterIndexOperation, AlterTableOperation, Array, Assignment, AssignmentTarget, AttachedToken, - BeginEndStatements, CaseStatement, CloseCursor, ClusteredIndex, ColumnDef, ColumnOption, - ColumnOptionDef, ConditionalStatementBlock, ConditionalStatements, ConflictTarget, ConnectBy, - ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, - Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, - Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, - FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, IndexColumn, Insert, - Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, - LateralView, LimitClause, MatchRecognizePattern, Measure, NamedParenthesizedList, - NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, - OnInsert, OpenStatement, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, - ProjectionSelect, Query, RaiseStatement, RaiseStatementValue, ReferentialAction, - RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, - SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, - TableConstraint, TableFactor, TableObject, TableOptionsClustered, TableWithJoins, - UpdateTableFromKind, Use, Value, Values, ViewColumnDef, WhileStatement, - WildcardAdditionalOptions, With, WithFill, + AlterIndexOperation, AlterTableOperation, Analyze, Array, Assignment, AssignmentTarget, + AttachedToken, BeginEndStatements, CaseStatement, CloseCursor, ClusteredIndex, ColumnDef, + ColumnOption, ColumnOptionDef, ConditionalStatementBlock, ConditionalStatements, + ConflictTarget, ConnectBy, ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, + CreateTableOptions, Cte, Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, + ExprWithAlias, Fetch, FromTable, Function, FunctionArg, FunctionArgExpr, + FunctionArgumentClause, FunctionArgumentList, FunctionArguments, GroupByExpr, HavingBound, + IfStatement, IlikeSelectItem, IndexColumn, Insert, Interpolate, InterpolateExpr, Join, + JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, LimitClause, + MatchRecognizePattern, Measure, NamedParenthesizedList, NamedWindowDefinition, ObjectName, + ObjectNamePart, Offset, OnConflict, OnConflictAction, OnInsert, OpenStatement, OrderBy, + OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, RaiseStatement, + RaiseStatementValue, ReferentialAction, RenameSelectItem, ReplaceSelectElement, + ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, + SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, + TableOptionsClustered, TableWithJoins, Update, UpdateTableFromKind, Use, Value, Values, + ViewColumnDef, WhileStatement, WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -298,38 +298,9 @@ impl Spanned for Values { impl Spanned for Statement { fn span(&self) -> Span { match self { - Statement::Analyze { - table_name, - partitions, - for_columns: _, - columns, - cache_metadata: _, - noscan: _, - compute_statistics: _, - has_table_keyword: _, - } => union_spans( - core::iter::once(table_name.span()) - .chain(partitions.iter().flat_map(|i| i.iter().map(|k| k.span()))) - .chain(columns.iter().map(|i| i.span)), - ), - Statement::Truncate { - table_names, - partitions, - table: _, - identity: _, - cascade: _, - on_cluster: _, - } => union_spans( - table_names - .iter() - .map(|i| i.name.span()) - .chain(partitions.iter().flat_map(|i| i.iter().map(|k| k.span()))), - ), - Statement::Msck { - table_name, - repair: _, - partition_action: _, - } => table_name.span(), + Statement::Analyze(analyze) => analyze.span(), + Statement::Truncate(truncate) => truncate.span(), + Statement::Msck(msck) => msck.span(), Statement::Query(query) => query.span(), Statement::Insert(insert) => insert.span(), Statement::Install { extension_name } => extension_name.span, @@ -375,47 +346,9 @@ impl Spanned for Statement { CloseCursor::All => Span::empty(), CloseCursor::Specific { name } => name.span, }, - Statement::Update { - table, - assignments, - from, - selection, - returning, - or: _, - limit: _, - } => union_spans( - core::iter::once(table.span()) - .chain(assignments.iter().map(|i| i.span())) - .chain(from.iter().map(|i| i.span())) - .chain(selection.iter().map(|i| i.span())) - .chain(returning.iter().flat_map(|i| i.iter().map(|k| k.span()))), - ), + Statement::Update(update) => update.span(), Statement::Delete(delete) => delete.span(), - Statement::CreateView { - or_alter: _, - or_replace: _, - materialized: _, - secure: _, - name, - columns, - query, - options, - cluster_by, - comment: _, - with_no_schema_binding: _, - if_not_exists: _, - temporary: _, - to, - name_before_not_exists: _, - params: _, - } => union_spans( - core::iter::once(name.span()) - .chain(columns.iter().map(|i| i.span())) - .chain(core::iter::once(query.span())) - .chain(core::iter::once(options.span())) - .chain(cluster_by.iter().map(|i| i.span)) - .chain(to.iter().map(|i| i.span())), - ), + Statement::CreateView(create_view) => create_view.span(), Statement::CreateTable(create_table) => create_table.span(), Statement::CreateVirtualTable { name, @@ -428,25 +361,13 @@ impl Spanned for Statement { .chain(module_args.iter().map(|i| i.span)), ), Statement::CreateIndex(create_index) => create_index.span(), - Statement::CreateRole { .. } => Span::empty(), + Statement::CreateRole(create_role) => create_role.span(), + Statement::CreateExtension(create_extension) => create_extension.span(), + Statement::DropExtension(drop_extension) => drop_extension.span(), Statement::CreateSecret { .. } => Span::empty(), Statement::CreateServer { .. } => Span::empty(), Statement::CreateConnector { .. } => Span::empty(), - Statement::AlterTable { - name, - if_exists: _, - only: _, - operations, - location: _, - on_cluster, - iceberg: _, - end_token, - } => union_spans( - core::iter::once(name.span()) - .chain(operations.iter().map(|i| i.span())) - .chain(on_cluster.iter().map(|i| i.span)) - .chain(core::iter::once(end_token.0.span)), - ), + Statement::AlterTable(alter_table) => alter_table.span(), Statement::AlterIndex { name, operation } => name.span().union(&operation.span()), Statement::AlterView { name, @@ -467,13 +388,11 @@ impl Spanned for Statement { Statement::AttachDuckDBDatabase { .. } => Span::empty(), Statement::DetachDuckDBDatabase { .. } => Span::empty(), Statement::Drop { .. } => Span::empty(), - Statement::DropFunction { .. } => Span::empty(), + Statement::DropFunction(drop_function) => drop_function.span(), Statement::DropDomain { .. } => Span::empty(), Statement::DropProcedure { .. } => Span::empty(), Statement::DropSecret { .. } => Span::empty(), Statement::Declare { .. } => Span::empty(), - Statement::CreateExtension { .. } => Span::empty(), - Statement::DropExtension { .. } => Span::empty(), Statement::Fetch { .. } => Span::empty(), Statement::Flush { .. } => Span::empty(), Statement::Discard { .. } => Span::empty(), @@ -873,6 +792,20 @@ impl Spanned for ConstraintCharacteristics { } } +impl Spanned for Analyze { + fn span(&self) -> Span { + union_spans( + core::iter::once(self.table_name.span()) + .chain( + self.partitions + .iter() + .flat_map(|i| i.iter().map(|k| k.span())), + ) + .chain(self.columns.iter().map(|i| i.span)), + ) + } +} + /// # partial span /// /// Missing spans: @@ -941,6 +874,29 @@ impl Spanned for Delete { } } +impl Spanned for Update { + fn span(&self) -> Span { + let Update { + table, + assignments, + from, + selection, + returning, + or: _, + limit, + } = self; + + union_spans( + core::iter::once(table.span()) + .chain(assignments.iter().map(|i| i.span())) + .chain(from.iter().map(|i| i.span())) + .chain(selection.iter().map(|i| i.span())) + .chain(returning.iter().flat_map(|i| i.iter().map(|k| k.span()))) + .chain(limit.iter().map(|i| i.span())), + ) + } +} + impl Spanned for FromTable { fn span(&self) -> Span { match self { @@ -2375,6 +2331,30 @@ impl Spanned for AlterSchema { } } +impl Spanned for CreateView { + fn span(&self) -> Span { + union_spans( + core::iter::once(self.name.span()) + .chain(self.columns.iter().map(|i| i.span())) + .chain(core::iter::once(self.query.span())) + .chain(core::iter::once(self.options.span())) + .chain(self.cluster_by.iter().map(|i| i.span)) + .chain(self.to.iter().map(|i| i.span())), + ) + } +} + +impl Spanned for AlterTable { + fn span(&self) -> Span { + union_spans( + core::iter::once(self.name.span()) + .chain(self.operations.iter().map(|i| i.span())) + .chain(self.on_cluster.iter().map(|i| i.span)) + .chain(core::iter::once(self.end_token.0.span)), + ) + } +} + #[cfg(test)] pub mod tests { use crate::dialect::{Dialect, GenericDialect, SnowflakeDialect}; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index d1e9b1e78..2b365e297 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -982,11 +982,12 @@ impl<'a> Parser<'a> { Ok(pa) })? .unwrap_or_default(); - Ok(Statement::Msck { + Ok(Msck { repair, table_name, partition_action, - }) + } + .into()) } pub fn parse_truncate(&mut self) -> Result { @@ -1024,14 +1025,15 @@ impl<'a> Parser<'a> { let on_cluster = self.parse_optional_on_cluster()?; - Ok(Statement::Truncate { + Ok(Truncate { table_names, partitions, table, identity, cascade, on_cluster, - }) + } + .into()) } fn parse_cascade_option(&mut self) -> Option { @@ -1167,7 +1169,7 @@ impl<'a> Parser<'a> { } } - Ok(Statement::Analyze { + Ok(Analyze { has_table_keyword, table_name, for_columns, @@ -1176,7 +1178,8 @@ impl<'a> Parser<'a> { cache_metadata, noscan, compute_statistics, - }) + } + .into()) } /// Parse a new expression including wildcard & qualified wildcard. @@ -5926,7 +5929,7 @@ impl<'a> Parser<'a> { Keyword::BINDING, ]); - Ok(Statement::CreateView { + Ok(CreateView { or_alter, name, columns, @@ -5943,7 +5946,8 @@ impl<'a> Parser<'a> { to, params: create_view_params, name_before_not_exists, - }) + } + .into()) } /// Parse optional parameters for the `CREATE VIEW` statement supported by [MySQL]. @@ -6206,7 +6210,7 @@ impl<'a> Parser<'a> { }? } - Ok(Statement::CreateRole { + Ok(CreateRole { names, if_not_exists, login, @@ -6225,7 +6229,8 @@ impl<'a> Parser<'a> { user, admin, authorization_owner, - }) + } + .into()) } pub fn parse_owner(&mut self) -> Result { @@ -6503,11 +6508,11 @@ impl<'a> Parser<'a> { let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); let func_desc = self.parse_comma_separated(Parser::parse_function_desc)?; let drop_behavior = self.parse_optional_drop_behavior(); - Ok(Statement::DropFunction { + Ok(Statement::DropFunction(DropFunction { if_exists, func_desc, drop_behavior, - }) + })) } /// ```sql @@ -7175,13 +7180,14 @@ impl<'a> Parser<'a> { (None, None, false) }; - Ok(Statement::CreateExtension { + Ok(CreateExtension { name, if_not_exists, schema, version, cascade, - }) + } + .into()) } /// Parse a PostgreSQL-specific [Statement::DropExtension] statement. @@ -7190,7 +7196,7 @@ impl<'a> Parser<'a> { let names = self.parse_comma_separated(|p| p.parse_identifier())?; let cascade_or_restrict = self.parse_one_of_keywords(&[Keyword::CASCADE, Keyword::RESTRICT]); - Ok(Statement::DropExtension { + Ok(Statement::DropExtension(DropExtension { names, if_exists, cascade_or_restrict: cascade_or_restrict @@ -7200,7 +7206,7 @@ impl<'a> Parser<'a> { _ => self.expected("CASCADE or RESTRICT", self.peek_token()), }) .transpose()?, - }) + })) } //TODO: Implement parsing for Skewed @@ -9378,7 +9384,7 @@ impl<'a> Parser<'a> { self.get_current_token().clone() }; - Ok(Statement::AlterTable { + Ok(AlterTable { name: table_name, if_exists, only, @@ -9387,7 +9393,8 @@ impl<'a> Parser<'a> { on_cluster, iceberg, end_token: AttachedToken(end_token), - }) + } + .into()) } pub fn parse_alter_view(&mut self) -> Result { @@ -15662,7 +15669,7 @@ impl<'a> Parser<'a> { } else { None }; - Ok(Statement::Update { + Ok(Update { table, assignments, from, @@ -15670,7 +15677,8 @@ impl<'a> Parser<'a> { returning, or, limit, - }) + } + .into()) } /// Parse a `var = expr` assignment, used in an UPDATE statement diff --git a/src/test_utils.rs b/src/test_utils.rs index 03fdee8ad..a8c8afd59 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -343,21 +343,12 @@ pub fn expr_from_projection(item: &SelectItem) -> &Expr { pub fn alter_table_op_with_name(stmt: Statement, expected_name: &str) -> AlterTableOperation { match stmt { - Statement::AlterTable { - name, - if_exists, - only: is_only, - operations, - on_cluster: _, - location: _, - iceberg, - end_token: _, - } => { - assert_eq!(name.to_string(), expected_name); - assert!(!if_exists); - assert!(!is_only); - assert!(!iceberg); - only(operations) + Statement::AlterTable(alter_table) => { + assert_eq!(alter_table.name.to_string(), expected_name); + assert!(!alter_table.if_exists); + assert!(!alter_table.only); + assert!(!alter_table.iceberg); + only(alter_table.operations) } _ => panic!("Expected ALTER TABLE statement"), } @@ -484,7 +475,7 @@ pub fn index_column(stmt: Statement) -> Expr { _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), } } - Statement::AlterTable { operations, .. } => match operations.first().unwrap() { + Statement::AlterTable(alter_table) => match alter_table.operations.first().unwrap() { AlterTableOperation::AddConstraint { constraint, .. } => { match constraint { TableConstraint::Index(constraint) => { diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 4f0cfa3e8..03a0ac813 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -332,13 +332,13 @@ fn parse_create_view_with_options() { "AS SELECT column_1, column_2, column_3 FROM myproject.mydataset.mytable", ); match bigquery().verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { name, query, options, columns, .. - } => { + }) => { assert_eq!( name, ObjectName::from(vec![ @@ -401,7 +401,7 @@ fn parse_create_view_with_options() { fn parse_create_view_if_not_exists() { let sql = "CREATE VIEW IF NOT EXISTS mydataset.newview AS SELECT foo FROM bar"; match bigquery().verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { name, columns, query, @@ -414,7 +414,7 @@ fn parse_create_view_if_not_exists() { if_not_exists, temporary, .. - } => { + }) => { assert_eq!("mydataset.newview", name.to_string()); assert_eq!(Vec::::new(), columns); assert_eq!("SELECT foo FROM bar", query.to_string()); @@ -435,12 +435,12 @@ fn parse_create_view_if_not_exists() { fn parse_create_view_with_unquoted_hyphen() { let sql = "CREATE VIEW IF NOT EXISTS my-pro-ject.mydataset.myview AS SELECT 1"; match bigquery().verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { name, query, if_not_exists, .. - } => { + }) => { assert_eq!("my-pro-ject.mydataset.myview", name.to_string()); assert_eq!("SELECT 1", query.to_string()); assert!(if_not_exists); diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index bc1431f9c..44bfcda42 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -243,12 +243,10 @@ fn parse_alter_table_attach_and_detach_partition() { match clickhouse_and_generic() .verified_stmt(format!("ALTER TABLE t0 {operation} PARTITION part").as_str()) { - Statement::AlterTable { - name, operations, .. - } => { - pretty_assertions::assert_eq!("t0", name.to_string()); + Statement::AlterTable(alter_table) => { + pretty_assertions::assert_eq!("t0", alter_table.name.to_string()); pretty_assertions::assert_eq!( - operations[0], + alter_table.operations[0], if operation == &"ATTACH" { AlterTableOperation::AttachPartition { partition: Partition::Expr(Identifier(Ident::new("part"))), @@ -266,9 +264,9 @@ fn parse_alter_table_attach_and_detach_partition() { match clickhouse_and_generic() .verified_stmt(format!("ALTER TABLE t1 {operation} PART part").as_str()) { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, operations, .. - } => { + }) => { pretty_assertions::assert_eq!("t1", name.to_string()); pretty_assertions::assert_eq!( operations[0], @@ -308,9 +306,9 @@ fn parse_alter_table_add_projection() { "ALTER TABLE t0 ADD PROJECTION IF NOT EXISTS my_name", " (SELECT a, b GROUP BY a ORDER BY b)", )) { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, operations, .. - } => { + }) => { assert_eq!(name, ObjectName::from(vec!["t0".into()])); assert_eq!(1, operations.len()); assert_eq!( @@ -380,9 +378,9 @@ fn parse_alter_table_add_projection() { fn parse_alter_table_drop_projection() { match clickhouse_and_generic().verified_stmt("ALTER TABLE t0 DROP PROJECTION IF EXISTS my_name") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, operations, .. - } => { + }) => { assert_eq!(name, ObjectName::from(vec!["t0".into()])); assert_eq!(1, operations.len()); assert_eq!( @@ -413,9 +411,9 @@ fn parse_alter_table_clear_and_materialize_projection() { format!("ALTER TABLE t0 {keyword} PROJECTION IF EXISTS my_name IN PARTITION p0",) .as_str(), ) { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, operations, .. - } => { + }) => { assert_eq!(name, ObjectName::from(vec!["t0".into()])); assert_eq!(1, operations.len()); assert_eq!( @@ -904,7 +902,7 @@ fn parse_create_table_with_variant_default_expressions() { #[test] fn parse_create_view_with_fields_data_types() { match clickhouse().verified_stmt(r#"CREATE VIEW v (i "int", f "String") AS SELECT * FROM t"#) { - Statement::CreateView { name, columns, .. } => { + Statement::CreateView(CreateView { name, columns, .. }) => { assert_eq!(name, ObjectName::from(vec!["v".into()])); assert_eq!( columns, @@ -1518,7 +1516,7 @@ fn parse_freeze_and_unfreeze_partition() { Value::SingleQuotedString("2024-08-14".to_string()).with_empty_span(), )); match clickhouse_and_generic().verified_stmt(&sql) { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!(operations.len(), 1); let expected_operation = if operation_name == &"FREEZE" { AlterTableOperation::FreezePartition { @@ -1542,7 +1540,7 @@ fn parse_freeze_and_unfreeze_partition() { let sql = format!("ALTER TABLE t {operation_name} PARTITION '2024-08-14' WITH NAME 'hello'"); match clickhouse_and_generic().verified_stmt(&sql) { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!(operations.len(), 1); let expected_partition = Partition::Expr(Expr::Value( Value::SingleQuotedString("2024-08-14".to_string()).with_empty_span(), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index a1fd48d3e..773937c5a 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -377,12 +377,12 @@ fn parse_insert_sqlite() { fn parse_update() { let sql = "UPDATE t SET a = 1, b = 2, c = 3 WHERE d"; match verified_stmt(sql) { - Statement::Update { + Statement::Update(Update { table, assignments, selection, .. - } => { + }) => { assert_eq!(table.to_string(), "t".to_string()); assert_eq!( assignments, @@ -439,7 +439,7 @@ fn parse_update_set_from() { let stmt = dialects.verified_stmt(sql); assert_eq!( stmt, - Statement::Update { + Statement::Update(Update { table: TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("t1")])), joins: vec![], @@ -516,7 +516,7 @@ fn parse_update_set_from() { returning: None, or: None, limit: None - } + }) ); let sql = "UPDATE T SET a = b FROM U, (SELECT foo FROM V) AS W WHERE 1 = 1"; @@ -527,7 +527,7 @@ fn parse_update_set_from() { fn parse_update_with_table_alias() { let sql = "UPDATE users AS u SET u.username = 'new_user' WHERE u.username = 'old_user'"; match verified_stmt(sql) { - Statement::Update { + Statement::Update(Update { table, assignments, from: _from, @@ -535,7 +535,7 @@ fn parse_update_with_table_alias() { returning, or: None, limit: None, - } => { + }) => { assert_eq!( TableWithJoins { relation: TableFactor::Table { @@ -591,7 +591,7 @@ fn parse_update_with_table_alias() { #[test] fn parse_update_or() { let expect_or_clause = |sql: &str, expected_action: SqliteOnConflict| match verified_stmt(sql) { - Statement::Update { or, .. } => assert_eq!(or, Some(expected_action)), + Statement::Update(Update { or, .. }) => assert_eq!(or, Some(expected_action)), other => unreachable!("Expected update with or, got {:?}", other), }; expect_or_clause( @@ -4846,9 +4846,9 @@ fn test_alter_table_with_on_cluster() { match all_dialects() .verified_stmt("ALTER TABLE t ON CLUSTER 'cluster' ADD CONSTRAINT bar PRIMARY KEY (baz)") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, on_cluster, .. - } => { + }) => { assert_eq!(name.to_string(), "t"); assert_eq!(on_cluster, Some(Ident::with_quote('\'', "cluster"))); } @@ -4858,9 +4858,9 @@ fn test_alter_table_with_on_cluster() { match all_dialects() .verified_stmt("ALTER TABLE t ON CLUSTER cluster_name ADD CONSTRAINT bar PRIMARY KEY (baz)") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, on_cluster, .. - } => { + }) => { assert_eq!(name.to_string(), "t"); assert_eq!(on_cluster, Some(Ident::new("cluster_name"))); } @@ -7615,7 +7615,7 @@ fn parse_ctes() { // CTE in a view let sql = &format!("CREATE VIEW v AS {with}"); match verified_stmt(sql) { - Statement::CreateView { query, .. } => assert_ctes_in_select(&cte_sqls, &query), + Statement::CreateView(create_view) => assert_ctes_in_select(&cte_sqls, &create_view.query), _ => panic!("Expected: CREATE VIEW"), } // CTE in a CTE... @@ -8103,7 +8103,7 @@ fn parse_drop_database_if_exists() { fn parse_create_view() { let sql = "CREATE VIEW myschema.myview AS SELECT foo FROM bar"; match verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { or_alter, name, columns, @@ -8120,7 +8120,7 @@ fn parse_create_view() { params, name_before_not_exists: _, secure: _, - } => { + }) => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); @@ -8146,7 +8146,7 @@ fn parse_create_view() { fn parse_create_view_with_options() { let sql = "CREATE VIEW v WITH (foo = 'bar', a = 123) AS SELECT 1"; match verified_stmt(sql) { - Statement::CreateView { options, .. } => { + Statement::CreateView(create_view) => { assert_eq!( CreateTableOptions::With(vec![ SqlOption::KeyValue { @@ -8160,7 +8160,7 @@ fn parse_create_view_with_options() { value: Expr::value(number("123")), }, ]), - options + create_view.options ); } _ => unreachable!(), @@ -8173,24 +8173,21 @@ fn parse_create_view_with_columns() { // TODO: why does this fail for ClickHouseDialect? (#1449) // match all_dialects().verified_stmt(sql) { match all_dialects_except(|d| d.is::()).verified_stmt(sql) { - Statement::CreateView { - or_alter, - name, - columns, - or_replace, - options, - query, - materialized, - cluster_by, - comment, - with_no_schema_binding: late_binding, - if_not_exists, - temporary, - to, - params, - name_before_not_exists: _, - secure: _, - } => { + Statement::CreateView(create_view) => { + let or_alter = create_view.or_alter; + let name = create_view.name; + let columns = create_view.columns; + let or_replace = create_view.or_replace; + let options = create_view.options; + let query = create_view.query; + let materialized = create_view.materialized; + let cluster_by = create_view.cluster_by; + let comment = create_view.comment; + let late_binding = create_view.with_no_schema_binding; + let if_not_exists = create_view.if_not_exists; + let temporary = create_view.temporary; + let to = create_view.to; + let params = create_view.params; assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); assert_eq!( @@ -8224,7 +8221,7 @@ fn parse_create_view_with_columns() { fn parse_create_view_temporary() { let sql = "CREATE TEMPORARY VIEW myschema.myview AS SELECT foo FROM bar"; match verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { or_alter, name, columns, @@ -8241,7 +8238,7 @@ fn parse_create_view_temporary() { params, name_before_not_exists: _, secure: _, - } => { + }) => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); @@ -8265,7 +8262,7 @@ fn parse_create_view_temporary() { fn parse_create_or_replace_view() { let sql = "CREATE OR REPLACE VIEW v AS SELECT 1"; match verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { or_alter, name, columns, @@ -8282,7 +8279,7 @@ fn parse_create_or_replace_view() { params, name_before_not_exists: _, secure: _, - } => { + }) => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); assert_eq!(columns, vec![]); @@ -8310,7 +8307,7 @@ fn parse_create_or_replace_materialized_view() { // https://docs.snowflake.com/en/sql-reference/sql/create-materialized-view.html let sql = "CREATE OR REPLACE MATERIALIZED VIEW v AS SELECT 1"; match verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { or_alter, name, columns, @@ -8327,7 +8324,7 @@ fn parse_create_or_replace_materialized_view() { params, name_before_not_exists: _, secure: _, - } => { + }) => { assert_eq!(or_alter, false); assert_eq!("v", name.to_string()); assert_eq!(columns, vec![]); @@ -8351,7 +8348,7 @@ fn parse_create_or_replace_materialized_view() { fn parse_create_materialized_view() { let sql = "CREATE MATERIALIZED VIEW myschema.myview AS SELECT foo FROM bar"; match verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { or_alter, name, or_replace, @@ -8368,7 +8365,7 @@ fn parse_create_materialized_view() { params, name_before_not_exists: _, secure: _, - } => { + }) => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); @@ -8392,7 +8389,7 @@ fn parse_create_materialized_view() { fn parse_create_materialized_view_with_cluster_by() { let sql = "CREATE MATERIALIZED VIEW myschema.myview CLUSTER BY (foo) AS SELECT foo FROM bar"; match verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { or_alter, name, or_replace, @@ -8409,7 +8406,7 @@ fn parse_create_materialized_view_with_cluster_by() { params, name_before_not_exists: _, secure: _, - } => { + }) => { assert_eq!(or_alter, false); assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); @@ -9417,21 +9414,17 @@ fn parse_drop_index() { fn parse_create_role() { let sql = "CREATE ROLE consultant"; match verified_stmt(sql) { - Statement::CreateRole { names, .. } => { - assert_eq_vec(&["consultant"], &names); + Statement::CreateRole(create_role) => { + assert_eq_vec(&["consultant"], &create_role.names); } _ => unreachable!(), } let sql = "CREATE ROLE IF NOT EXISTS mysql_a, mysql_b"; match verified_stmt(sql) { - Statement::CreateRole { - names, - if_not_exists, - .. - } => { - assert_eq_vec(&["mysql_a", "mysql_b"], &names); - assert!(if_not_exists); + Statement::CreateRole(create_role) => { + assert_eq_vec(&["mysql_a", "mysql_b"], &create_role.names); + assert!(create_role.if_not_exists); } _ => unreachable!(), } @@ -13351,8 +13344,8 @@ fn test_extract_seconds_single_quote_err() { fn test_truncate_table_with_on_cluster() { let sql = "TRUNCATE TABLE t ON CLUSTER cluster_name"; match all_dialects().verified_stmt(sql) { - Statement::Truncate { on_cluster, .. } => { - assert_eq!(on_cluster, Some(Ident::new("cluster_name"))); + Statement::Truncate(truncate) => { + assert_eq!(truncate.on_cluster, Some(Ident::new("cluster_name"))); } _ => panic!("Expected: TRUNCATE TABLE statement"), } @@ -16407,14 +16400,14 @@ fn parse_truncate_only() { ]; assert_eq!( - Statement::Truncate { + Statement::Truncate(Truncate { table_names, partitions: None, table: true, identity: None, cascade: None, on_cluster: None, - }, + }), truncate ); } @@ -17288,9 +17281,9 @@ fn parse_invisible_column() { let sql = r#"ALTER TABLE t ADD COLUMN bar INT INVISIBLE"#; let stmt = verified_stmt(sql); match stmt { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(alter_table) => { assert_eq!( - operations, + alter_table.operations, vec![AlterTableOperation::AddColumn { column_keyword: true, if_not_exists: false, diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 181899170..9dccce491 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -778,14 +778,10 @@ fn parse_mssql_bin_literal() { fn parse_mssql_create_role() { let sql = "CREATE ROLE mssql AUTHORIZATION helena"; match ms().verified_stmt(sql) { - Statement::CreateRole { - names, - authorization_owner, - .. - } => { - assert_eq_vec(&["mssql"], &names); + Statement::CreateRole(create_role) => { + assert_eq_vec(&["mssql"], &create_role.names); assert_eq!( - authorization_owner, + create_role.authorization_owner, Some(ObjectName::from(vec![Ident { value: "helena".into(), quote_style: None, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 35bac548c..9aaa35ba0 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2603,7 +2603,7 @@ fn parse_insert_with_numeric_prefix_column_name() { fn parse_update_with_joins() { let sql = "UPDATE orders AS o JOIN customers AS c ON o.customer_id = c.id SET o.completed = true WHERE c.firstname = 'Peter'"; match mysql().verified_stmt(sql) { - Statement::Update { + Statement::Update(Update { table, assignments, from: _from, @@ -2611,7 +2611,7 @@ fn parse_update_with_joins() { returning, or: None, limit: None, - } => { + }) => { assert_eq!( TableWithJoins { relation: TableFactor::Table { @@ -2729,7 +2729,7 @@ fn parse_delete_with_limit() { #[test] fn parse_alter_table_add_column() { match mysql().verified_stmt("ALTER TABLE tab ADD COLUMN b INT FIRST") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, if_exists, only, @@ -2738,7 +2738,7 @@ fn parse_alter_table_add_column() { location: _, on_cluster: _, end_token: _, - } => { + }) => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); assert!(!iceberg); @@ -2761,13 +2761,13 @@ fn parse_alter_table_add_column() { } match mysql().verified_stmt("ALTER TABLE tab ADD COLUMN b INT AFTER foo") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, if_exists, only, operations, .. - } => { + }) => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); assert!(!only); @@ -2798,13 +2798,13 @@ fn parse_alter_table_add_columns() { match mysql() .verified_stmt("ALTER TABLE tab ADD COLUMN a TEXT FIRST, ADD COLUMN b INT AFTER foo") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, if_exists, only, operations, .. - } => { + }) => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); assert!(!only); @@ -3026,7 +3026,7 @@ fn parse_alter_table_with_algorithm() { "ALTER TABLE users DROP COLUMN password_digest, ALGORITHM = COPY, RENAME COLUMN name TO username"; let stmt = mysql_and_generic().verified_stmt(sql); match stmt { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!( operations, vec![ @@ -3074,7 +3074,7 @@ fn parse_alter_table_with_lock() { "ALTER TABLE users DROP COLUMN password_digest, LOCK = EXCLUSIVE, RENAME COLUMN name TO username"; let stmt = mysql_and_generic().verified_stmt(sql); match stmt { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!( operations, vec![ @@ -3857,7 +3857,7 @@ fn parse_revoke() { fn parse_create_view_algorithm_param() { let sql = "CREATE ALGORITHM = MERGE VIEW foo AS SELECT 1"; let stmt = mysql().verified_stmt(sql); - if let Statement::CreateView { + if let Statement::CreateView(CreateView { params: Some(CreateViewParams { algorithm, @@ -3865,7 +3865,7 @@ fn parse_create_view_algorithm_param() { security, }), .. - } = stmt + }) = stmt { assert_eq!(algorithm, Some(CreateViewAlgorithm::Merge)); assert!(definer.is_none()); @@ -3881,7 +3881,7 @@ fn parse_create_view_algorithm_param() { fn parse_create_view_definer_param() { let sql = "CREATE DEFINER = 'jeffrey'@'localhost' VIEW foo AS SELECT 1"; let stmt = mysql().verified_stmt(sql); - if let Statement::CreateView { + if let Statement::CreateView(CreateView { params: Some(CreateViewParams { algorithm, @@ -3889,7 +3889,7 @@ fn parse_create_view_definer_param() { security, }), .. - } = stmt + }) = stmt { assert!(algorithm.is_none()); if let Some(GranteeName::UserHost { user, host }) = definer { @@ -3910,7 +3910,7 @@ fn parse_create_view_definer_param() { fn parse_create_view_security_param() { let sql = "CREATE SQL SECURITY DEFINER VIEW foo AS SELECT 1"; let stmt = mysql().verified_stmt(sql); - if let Statement::CreateView { + if let Statement::CreateView(CreateView { params: Some(CreateViewParams { algorithm, @@ -3918,7 +3918,7 @@ fn parse_create_view_security_param() { security, }), .. - } = stmt + }) = stmt { assert!(algorithm.is_none()); assert!(definer.is_none()); @@ -3933,7 +3933,7 @@ fn parse_create_view_security_param() { fn parse_create_view_multiple_params() { let sql = "CREATE ALGORITHM = UNDEFINED DEFINER = `root`@`%` SQL SECURITY INVOKER VIEW foo AS SELECT 1"; let stmt = mysql().verified_stmt(sql); - if let Statement::CreateView { + if let Statement::CreateView(CreateView { params: Some(CreateViewParams { algorithm, @@ -3941,7 +3941,7 @@ fn parse_create_view_multiple_params() { security, }), .. - } = stmt + }) = stmt { assert_eq!(algorithm, Some(CreateViewAlgorithm::Undefined)); if let Some(GranteeName::UserHost { user, host }) = definer { @@ -4179,7 +4179,7 @@ fn test_variable_assignment_using_colon_equal() { let stmt = mysql().verified_stmt(sql_update); match stmt { - Statement::Update { assignments, .. } => { + Statement::Update(Update { assignments, .. }) => { assert_eq!( assignments, vec![Assignment { diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 45ae32dac..3c2a98e1a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -605,7 +605,7 @@ fn parse_alter_table_constraints_unique_nulls_distinct() { match pg_and_generic() .verified_stmt("ALTER TABLE t ADD CONSTRAINT b UNIQUE NULLS NOT DISTINCT (c)") { - Statement::AlterTable { operations, .. } => match &operations[0] { + Statement::AlterTable(alter_table) => match &alter_table.operations[0] { AlterTableOperation::AddConstraint { constraint: TableConstraint::Unique(constraint), .. @@ -674,93 +674,93 @@ fn parse_create_extension() { fn parse_drop_extension() { assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION extension_name"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into()], if_exists: false, cascade_or_restrict: None, - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION extension_name CASCADE"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into()], if_exists: false, cascade_or_restrict: Some(ReferentialAction::Cascade), - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION extension_name RESTRICT"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into()], if_exists: false, cascade_or_restrict: Some(ReferentialAction::Restrict), - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION extension_name, extension_name2 CASCADE"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into(), "extension_name2".into()], if_exists: false, cascade_or_restrict: Some(ReferentialAction::Cascade), - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION extension_name, extension_name2 RESTRICT"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into(), "extension_name2".into()], if_exists: false, cascade_or_restrict: Some(ReferentialAction::Restrict), - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION IF EXISTS extension_name"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into()], if_exists: true, cascade_or_restrict: None, - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION IF EXISTS extension_name CASCADE"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into()], if_exists: true, cascade_or_restrict: Some(ReferentialAction::Cascade), - } + }) ); assert_eq!( pg_and_generic().verified_stmt("DROP EXTENSION IF EXISTS extension_name RESTRICT"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name".into()], if_exists: true, cascade_or_restrict: Some(ReferentialAction::Restrict), - } + }) ); assert_eq!( pg_and_generic() .verified_stmt("DROP EXTENSION IF EXISTS extension_name1, extension_name2 CASCADE"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name1".into(), "extension_name2".into()], if_exists: true, cascade_or_restrict: Some(ReferentialAction::Cascade), - } + }) ); assert_eq!( pg_and_generic() .verified_stmt("DROP EXTENSION IF EXISTS extension_name1, extension_name2 RESTRICT"), - Statement::DropExtension { + Statement::DropExtension(DropExtension { names: vec!["extension_name1".into(), "extension_name2".into()], if_exists: true, cascade_or_restrict: Some(ReferentialAction::Restrict), - } + }) ); } @@ -829,13 +829,13 @@ fn parse_alter_table_alter_column_add_generated() { #[test] fn parse_alter_table_add_columns() { match pg().verified_stmt("ALTER TABLE IF EXISTS ONLY tab ADD COLUMN a TEXT, ADD COLUMN b INT") { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, if_exists, only, operations, .. - } => { + }) => { assert_eq!(name.to_string(), "tab"); assert!(if_exists); assert!(only); @@ -909,13 +909,13 @@ fn parse_alter_table_owner_to() { for case in test_cases { match pg_and_generic().verified_stmt(case.sql) { - Statement::AlterTable { + Statement::AlterTable(AlterTable { name, if_exists: _, only: _, operations, .. - } => { + }) => { assert_eq!(name.to_string(), "tab"); assert_eq!( operations, @@ -2009,7 +2009,7 @@ fn parse_pg_returning() { RETURNING temp_lo AS lo, temp_hi AS hi, prcp", ); match stmt { - Statement::Update { returning, .. } => { + Statement::Update(Update { returning, .. }) => { assert_eq!( Some(vec![ SelectItem::ExprWithAlias { @@ -3833,47 +3833,29 @@ fn parse_custom_operator() { fn parse_create_role() { let sql = "CREATE ROLE IF NOT EXISTS mysql_a, mysql_b"; match pg().verified_stmt(sql) { - Statement::CreateRole { - names, - if_not_exists, - .. - } => { - assert_eq_vec(&["mysql_a", "mysql_b"], &names); - assert!(if_not_exists); + Statement::CreateRole(create_role) => { + assert_eq_vec(&["mysql_a", "mysql_b"], &create_role.names); + assert!(create_role.if_not_exists); } _ => unreachable!(), } let sql = "CREATE ROLE abc LOGIN PASSWORD NULL"; match pg().parse_sql_statements(sql).as_deref() { - Ok( - [Statement::CreateRole { - names, - login, - password, - .. - }], - ) => { - assert_eq_vec(&["abc"], names); - assert_eq!(*login, Some(true)); - assert_eq!(*password, Some(Password::NullPassword)); + Ok([Statement::CreateRole(create_role)]) => { + assert_eq_vec(&["abc"], &create_role.names); + assert_eq!(create_role.login, Some(true)); + assert_eq!(create_role.password, Some(Password::NullPassword)); } err => panic!("Failed to parse CREATE ROLE test case: {err:?}"), } let sql = "CREATE ROLE abc WITH LOGIN PASSWORD NULL"; match pg().parse_sql_statements(sql).as_deref() { - Ok( - [Statement::CreateRole { - names, - login, - password, - .. - }], - ) => { - assert_eq_vec(&["abc"], names); - assert_eq!(*login, Some(true)); - assert_eq!(*password, Some(Password::NullPassword)); + Ok([Statement::CreateRole(create_role)]) => { + assert_eq_vec(&["abc"], &create_role.names); + assert_eq!(create_role.login, Some(true)); + assert_eq!(create_role.password, Some(Password::NullPassword)); } err => panic!("Failed to parse CREATE ROLE test case: {err:?}"), } @@ -3881,69 +3863,44 @@ fn parse_create_role() { let sql = "CREATE ROLE magician WITH SUPERUSER CREATEROLE NOCREATEDB BYPASSRLS INHERIT PASSWORD 'abcdef' LOGIN VALID UNTIL '2025-01-01' IN ROLE role1, role2 ROLE role3 ADMIN role4, role5 REPLICATION"; // Roundtrip order of optional parameters is not preserved match pg().parse_sql_statements(sql).as_deref() { - Ok( - [Statement::CreateRole { - names, - if_not_exists, - bypassrls, - login, - inherit, - password, - superuser, - create_db, - create_role, - replication, - connection_limit, - valid_until, - in_role, - in_group, - role, - user: _, - admin, - authorization_owner, - }], - ) => { - assert_eq_vec(&["magician"], names); - assert!(!*if_not_exists); - assert_eq!(*login, Some(true)); - assert_eq!(*inherit, Some(true)); - assert_eq!(*bypassrls, Some(true)); + Ok([Statement::CreateRole(create_role)]) => { + assert_eq_vec(&["magician"], &create_role.names); + assert!(!create_role.if_not_exists); + assert_eq!(create_role.login, Some(true)); + assert_eq!(create_role.inherit, Some(true)); + assert_eq!(create_role.bypassrls, Some(true)); assert_eq!( - *password, + create_role.password, Some(Password::Password(Expr::Value( (Value::SingleQuotedString("abcdef".into())).with_empty_span() ))) ); - assert_eq!(*superuser, Some(true)); - assert_eq!(*create_db, Some(false)); - assert_eq!(*create_role, Some(true)); - assert_eq!(*replication, Some(true)); - assert_eq!(*connection_limit, None); + assert_eq!(create_role.superuser, Some(true)); + assert_eq!(create_role.create_db, Some(false)); + assert_eq!(create_role.create_role, Some(true)); + assert_eq!(create_role.replication, Some(true)); + assert_eq!(create_role.connection_limit, None); assert_eq!( - *valid_until, + create_role.valid_until, Some(Expr::Value( (Value::SingleQuotedString("2025-01-01".into())).with_empty_span() )) ); - assert_eq_vec(&["role1", "role2"], in_role); - assert!(in_group.is_empty()); - assert_eq_vec(&["role3"], role); - assert_eq_vec(&["role4", "role5"], admin); - assert_eq!(*authorization_owner, None); + assert_eq_vec(&["role1", "role2"], &create_role.in_role); + assert!(create_role.in_group.is_empty()); + assert_eq_vec(&["role3"], &create_role.role); + assert_eq_vec(&["role4", "role5"], &create_role.admin); + assert_eq!(create_role.authorization_owner, None); } err => panic!("Failed to parse CREATE ROLE test case: {err:?}"), } let sql = "CREATE ROLE abc WITH USER foo, bar ROLE baz "; match pg().parse_sql_statements(sql).as_deref() { - Ok( - [Statement::CreateRole { - names, user, role, .. - }], - ) => { - assert_eq_vec(&["abc"], names); - assert_eq_vec(&["foo", "bar"], user); - assert_eq_vec(&["baz"], role); + Ok([Statement::CreateRole(create_role)]) => { + assert_eq_vec(&["abc"], &create_role.names); + assert_eq_vec(&["foo", "bar"], &create_role.user); + assert_eq_vec(&["baz"], &create_role.role); } err => panic!("Failed to parse CREATE ROLE test case: {err:?}"), } @@ -4532,7 +4489,7 @@ fn parse_drop_function() { let sql = "DROP FUNCTION IF EXISTS test_func"; assert_eq!( pg().verified_stmt(sql), - Statement::DropFunction { + Statement::DropFunction(DropFunction { if_exists: true, func_desc: vec![FunctionDesc { name: ObjectName::from(vec![Ident { @@ -4543,13 +4500,13 @@ fn parse_drop_function() { args: None }], drop_behavior: None - } + }) ); let sql = "DROP FUNCTION IF EXISTS test_func(a INTEGER, IN b INTEGER = 1)"; assert_eq!( pg().verified_stmt(sql), - Statement::DropFunction { + Statement::DropFunction(DropFunction { if_exists: true, func_desc: vec![FunctionDesc { name: ObjectName::from(vec![Ident { @@ -4570,13 +4527,13 @@ fn parse_drop_function() { ]), }], drop_behavior: None - } + }) ); let sql = "DROP FUNCTION IF EXISTS test_func1(a INTEGER, IN b INTEGER = 1), test_func2(a VARCHAR, IN b INTEGER = 1)"; assert_eq!( pg().verified_stmt(sql), - Statement::DropFunction { + Statement::DropFunction(DropFunction { if_exists: true, func_desc: vec![ FunctionDesc { @@ -4617,7 +4574,7 @@ fn parse_drop_function() { } ], drop_behavior: None - } + }) ); } @@ -4957,14 +4914,14 @@ fn parse_truncate() { only: false, }]; assert_eq!( - Statement::Truncate { + Statement::Truncate(Truncate { table_names, partitions: None, table: false, identity: None, cascade: None, on_cluster: None, - }, + }), truncate ); } @@ -4981,14 +4938,14 @@ fn parse_truncate_with_options() { }]; assert_eq!( - Statement::Truncate { + Statement::Truncate(Truncate { table_names, partitions: None, table: true, identity: Some(TruncateIdentityOption::Restart), cascade: Some(CascadeOption::Cascade), on_cluster: None, - }, + }), truncate ); } @@ -5014,14 +4971,14 @@ fn parse_truncate_with_table_list() { ]; assert_eq!( - Statement::Truncate { + Statement::Truncate(Truncate { table_names, partitions: None, table: true, identity: Some(TruncateIdentityOption::Restart), cascade: Some(CascadeOption::Cascade), on_cluster: None, - }, + }), truncate ); } @@ -6409,7 +6366,7 @@ fn parse_varbit_datatype() { #[test] fn parse_alter_table_replica_identity() { match pg_and_generic().verified_stmt("ALTER TABLE foo REPLICA IDENTITY FULL") { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!( operations, vec![AlterTableOperation::ReplicaIdentity { @@ -6421,7 +6378,7 @@ fn parse_alter_table_replica_identity() { } match pg_and_generic().verified_stmt("ALTER TABLE foo REPLICA IDENTITY USING INDEX foo_idx") { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!( operations, vec![AlterTableOperation::ReplicaIdentity { @@ -6469,7 +6426,7 @@ fn parse_alter_table_constraint_not_valid() { match pg_and_generic().verified_stmt( "ALTER TABLE foo ADD CONSTRAINT bar FOREIGN KEY (baz) REFERENCES other(ref) NOT VALID", ) { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!( operations, vec![AlterTableOperation::AddConstraint { @@ -6495,7 +6452,7 @@ fn parse_alter_table_constraint_not_valid() { #[test] fn parse_alter_table_validate_constraint() { match pg_and_generic().verified_stmt("ALTER TABLE foo VALIDATE CONSTRAINT bar") { - Statement::AlterTable { operations, .. } => { + Statement::AlterTable(AlterTable { operations, .. }) => { assert_eq!( operations, vec![AlterTableOperation::ValidateConstraint { name: "bar".into() }] diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index e04bfaf5d..e7a128343 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -53,11 +53,11 @@ fn parse_sf_create_secure_view_and_materialized_view() { "CREATE OR REPLACE SECURE MATERIALIZED VIEW v AS SELECT 1", ] { match snowflake().verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { secure, materialized, .. - } => { + }) => { assert!(secure); if sql.contains("MATERIALIZED") { assert!(materialized); @@ -1047,7 +1047,7 @@ fn parse_sf_create_or_replace_with_comment_for_snowflake() { test_utils::TestedDialects::new(vec![Box::new(SnowflakeDialect {}) as Box]); match dialect.verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { name, columns, or_replace, @@ -1060,7 +1060,7 @@ fn parse_sf_create_or_replace_with_comment_for_snowflake() { if_not_exists, temporary, .. - } => { + }) => { assert_eq!("v", name.to_string()); assert_eq!(columns, vec![]); assert_eq!(options, CreateTableOptions::None); @@ -3281,7 +3281,7 @@ fn parse_view_column_descriptions() { let sql = "CREATE OR REPLACE VIEW v (a COMMENT 'Comment', b) AS SELECT a, b FROM table1"; match snowflake().verified_stmt(sql) { - Statement::CreateView { name, columns, .. } => { + Statement::CreateView(CreateView { name, columns, .. }) => { assert_eq!(name.to_string(), "v"); assert_eq!( columns, diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 114aca03a..5083ecd08 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -166,7 +166,7 @@ fn parse_create_virtual_table() { fn parse_create_view_temporary_if_not_exists() { let sql = "CREATE TEMPORARY VIEW IF NOT EXISTS myschema.myview AS SELECT foo FROM bar"; match sqlite_and_generic().verified_stmt(sql) { - Statement::CreateView { + Statement::CreateView(CreateView { name, columns, query, @@ -179,7 +179,7 @@ fn parse_create_view_temporary_if_not_exists() { if_not_exists, temporary, .. - } => { + }) => { assert_eq!("myschema.myview", name.to_string()); assert_eq!(Vec::::new(), columns); assert_eq!("SELECT foo FROM bar", query.to_string()); @@ -467,7 +467,7 @@ fn parse_update_tuple_row_values() { // See https://github.com/sqlparser-rs/sqlparser-rs/issues/1311 assert_eq!( sqlite().verified_stmt("UPDATE x SET (a, b) = (1, 2)"), - Statement::Update { + Statement::Update(Update { or: None, assignments: vec![Assignment { target: AssignmentTarget::Tuple(vec![ @@ -487,7 +487,7 @@ fn parse_update_tuple_row_values() { from: None, returning: None, limit: None - } + }) ); } @@ -596,7 +596,7 @@ fn test_regexp_operator() { #[test] fn test_update_delete_limit() { match sqlite().verified_stmt("UPDATE foo SET bar = 1 LIMIT 99") { - Statement::Update { limit, .. } => { + Statement::Update(Update { limit, .. }) => { assert_eq!(limit, Some(Expr::value(number("99")))); } _ => unreachable!(), From 4490c8c55cbe1e6b96a29fd9ba448e06320af50a Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Sat, 11 Oct 2025 13:16:56 +0200 Subject: [PATCH 275/286] Added support for SQLite triggers (#2037) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 56 ++++++-- src/ast/mod.rs | 4 +- src/dialect/mssql.rs | 6 +- src/parser/mod.rs | 45 ++++-- tests/sqlparser_mssql.rs | 4 +- tests/sqlparser_mysql.rs | 4 +- tests/sqlparser_postgres.rs | 26 ++-- tests/sqlparser_sqlite.rs | 275 ++++++++++++++++++++++++++++++++++++ 8 files changed, 374 insertions(+), 46 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index aafa8e644..3294a7a81 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -2922,6 +2922,26 @@ impl Spanned for RenameTableNameKind { } } +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +/// Whether the syntax used for the trigger object (ROW or STATEMENT) is `FOR` or `FOR EACH`. +pub enum TriggerObjectKind { + /// The `FOR` syntax is used. + For(TriggerObject), + /// The `FOR EACH` syntax is used. + ForEach(TriggerObject), +} + +impl Display for TriggerObjectKind { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + TriggerObjectKind::For(obj) => write!(f, "FOR {obj}"), + TriggerObjectKind::ForEach(obj) => write!(f, "FOR EACH {obj}"), + } + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -2943,6 +2963,23 @@ pub struct CreateTrigger { /// /// [MsSql](https://learn.microsoft.com/en-us/sql/t-sql/statements/create-trigger-transact-sql?view=sql-server-ver16#arguments) pub or_alter: bool, + /// True if this is a temporary trigger. + /// + /// Examples: + /// + /// ```sql + /// CREATE TEMP TRIGGER trigger_name + /// ``` + /// + /// or + /// + /// ```sql + /// CREATE TEMPORARY TRIGGER trigger_name; + /// CREATE TEMP TRIGGER trigger_name; + /// ``` + /// + /// [SQLite](https://sqlite.org/lang_createtrigger.html#temp_triggers_on_non_temp_tables) + pub temporary: bool, /// The `OR REPLACE` clause is used to re-create the trigger if it already exists. /// /// Example: @@ -2987,6 +3024,8 @@ pub struct CreateTrigger { /// ``` pub period: TriggerPeriod, /// Whether the trigger period was specified before the target table name. + /// This does not refer to whether the period is BEFORE, AFTER, or INSTEAD OF, + /// but rather the position of the period clause in relation to the table name. /// /// ```sql /// -- period_before_table == true: Postgres, MySQL, and standard SQL @@ -3006,9 +3045,9 @@ pub struct CreateTrigger { pub referencing: Vec, /// This specifies whether the trigger function should be fired once for /// every row affected by the trigger event, or just once per SQL statement. - pub trigger_object: TriggerObject, - /// Whether to include the `EACH` term of the `FOR EACH`, as it is optional syntax. - pub include_each: bool, + /// This is optional in some SQL dialects, such as SQLite, and if not specified, in + /// those cases, the implied default is `FOR EACH ROW`. + pub trigger_object: Option, /// Triggering conditions pub condition: Option, /// Execute logic block @@ -3025,6 +3064,7 @@ impl Display for CreateTrigger { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { let CreateTrigger { or_alter, + temporary, or_replace, is_constraint, name, @@ -3036,7 +3076,6 @@ impl Display for CreateTrigger { referencing, trigger_object, condition, - include_each, exec_body, statements_as, statements, @@ -3044,7 +3083,8 @@ impl Display for CreateTrigger { } = self; write!( f, - "CREATE {or_alter}{or_replace}{is_constraint}TRIGGER {name} ", + "CREATE {temporary}{or_alter}{or_replace}{is_constraint}TRIGGER {name} ", + temporary = if *temporary { "TEMPORARY " } else { "" }, or_alter = if *or_alter { "OR ALTER " } else { "" }, or_replace = if *or_replace { "OR REPLACE " } else { "" }, is_constraint = if *is_constraint { "CONSTRAINT " } else { "" }, @@ -3076,10 +3116,8 @@ impl Display for CreateTrigger { write!(f, " REFERENCING {}", display_separated(referencing, " "))?; } - if *include_each { - write!(f, " FOR EACH {trigger_object}")?; - } else if exec_body.is_some() { - write!(f, " FOR {trigger_object}")?; + if let Some(trigger_object) = trigger_object { + write!(f, " {trigger_object}")?; } if let Some(condition) = condition { write!(f, " WHEN {condition}")?; diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 0c83b3203..fef8943ef 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -70,8 +70,8 @@ pub use self::ddl::{ IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, Msck, NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction, RenameTableNameKind, - ReplicaIdentity, TagsColumnOption, Truncate, UserDefinedTypeCompositeAttributeDef, - UserDefinedTypeRepresentation, ViewColumnDef, + ReplicaIdentity, TagsColumnOption, TriggerObjectKind, Truncate, + UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, }; pub use self::dml::{Delete, Insert, Update}; pub use self::operator::{BinaryOperator, UnaryOperator}; diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index 4fcc0e4b6..f1d54cd67 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -18,7 +18,7 @@ use crate::ast::helpers::attached_token::AttachedToken; use crate::ast::{ BeginEndStatements, ConditionalStatementBlock, ConditionalStatements, CreateTrigger, - GranteesType, IfStatement, Statement, TriggerObject, + GranteesType, IfStatement, Statement, }; use crate::dialect::Dialect; use crate::keywords::{self, Keyword}; @@ -254,6 +254,7 @@ impl MsSqlDialect { Ok(CreateTrigger { or_alter, + temporary: false, or_replace: false, is_constraint: false, name, @@ -263,8 +264,7 @@ impl MsSqlDialect { table_name, referenced_table_name: None, referencing: Vec::new(), - trigger_object: TriggerObject::Statement, - include_each: false, + trigger_object: None, condition: None, exec_body: None, statements_as: true, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 2b365e297..70f4d8568 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -4753,9 +4753,9 @@ impl<'a> Parser<'a> { } else if self.parse_keyword(Keyword::DOMAIN) { self.parse_create_domain() } else if self.parse_keyword(Keyword::TRIGGER) { - self.parse_create_trigger(or_alter, or_replace, false) + self.parse_create_trigger(temporary, or_alter, or_replace, false) } else if self.parse_keywords(&[Keyword::CONSTRAINT, Keyword::TRIGGER]) { - self.parse_create_trigger(or_alter, or_replace, true) + self.parse_create_trigger(temporary, or_alter, or_replace, true) } else if self.parse_keyword(Keyword::MACRO) { self.parse_create_macro(or_replace, temporary) } else if self.parse_keyword(Keyword::SECRET) { @@ -5551,7 +5551,8 @@ impl<'a> Parser<'a> { /// DROP TRIGGER [ IF EXISTS ] name ON table_name [ CASCADE | RESTRICT ] /// ``` pub fn parse_drop_trigger(&mut self) -> Result { - if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect | MsSqlDialect) { + if !dialect_of!(self is PostgreSqlDialect | SQLiteDialect | GenericDialect | MySqlDialect | MsSqlDialect) + { self.prev_token(); return self.expected("an object type after DROP", self.peek_token()); } @@ -5579,11 +5580,13 @@ impl<'a> Parser<'a> { pub fn parse_create_trigger( &mut self, + temporary: bool, or_alter: bool, or_replace: bool, is_constraint: bool, ) -> Result { - if !dialect_of!(self is PostgreSqlDialect | GenericDialect | MySqlDialect | MsSqlDialect) { + if !dialect_of!(self is PostgreSqlDialect | SQLiteDialect | GenericDialect | MySqlDialect | MsSqlDialect) + { self.prev_token(); return self.expected("an object type after CREATE", self.peek_token()); } @@ -5610,14 +5613,25 @@ impl<'a> Parser<'a> { } } - self.expect_keyword_is(Keyword::FOR)?; - let include_each = self.parse_keyword(Keyword::EACH); - let trigger_object = - match self.expect_one_of_keywords(&[Keyword::ROW, Keyword::STATEMENT])? { - Keyword::ROW => TriggerObject::Row, - Keyword::STATEMENT => TriggerObject::Statement, - _ => unreachable!(), - }; + let trigger_object = if self.parse_keyword(Keyword::FOR) { + let include_each = self.parse_keyword(Keyword::EACH); + let trigger_object = + match self.expect_one_of_keywords(&[Keyword::ROW, Keyword::STATEMENT])? { + Keyword::ROW => TriggerObject::Row, + Keyword::STATEMENT => TriggerObject::Statement, + _ => unreachable!(), + }; + + Some(if include_each { + TriggerObjectKind::ForEach(trigger_object) + } else { + TriggerObjectKind::For(trigger_object) + }) + } else { + let _ = self.parse_keyword(Keyword::FOR); + + None + }; let condition = self .parse_keyword(Keyword::WHEN) @@ -5632,8 +5646,9 @@ impl<'a> Parser<'a> { statements = Some(self.parse_conditional_statements(&[Keyword::END])?); } - Ok(Statement::CreateTrigger(CreateTrigger { + Ok(CreateTrigger { or_alter, + temporary, or_replace, is_constraint, name, @@ -5644,13 +5659,13 @@ impl<'a> Parser<'a> { referenced_table_name, referencing, trigger_object, - include_each, condition, exec_body, statements_as: false, statements, characteristics, - })) + } + .into()) } pub fn parse_trigger_period(&mut self) -> Result { diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 9dccce491..e11c79f01 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2388,6 +2388,7 @@ fn parse_create_trigger() { create_stmt, Statement::CreateTrigger(CreateTrigger { or_alter: true, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("reminder1")]), @@ -2397,8 +2398,7 @@ fn parse_create_trigger() { table_name: ObjectName::from(vec![Ident::new("Sales"), Ident::new("Customer")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Statement, - include_each: false, + trigger_object: None, condition: None, exec_body: None, statements_as: true, diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 9aaa35ba0..e0ddecf32 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -4018,6 +4018,7 @@ fn parse_create_trigger() { create_stmt, Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), @@ -4027,8 +4028,7 @@ fn parse_create_trigger() { table_name: ObjectName::from(vec![Ident::new("emp")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: None, exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 3c2a98e1a..e18bf662a 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5636,6 +5636,7 @@ fn parse_create_simple_before_insert_trigger() { let sql = "CREATE TRIGGER check_insert BEFORE INSERT ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_insert"; let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_insert")]), @@ -5645,8 +5646,7 @@ fn parse_create_simple_before_insert_trigger() { table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: None, exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, @@ -5668,6 +5668,7 @@ fn parse_create_after_update_trigger_with_condition() { let sql = "CREATE TRIGGER check_update AFTER UPDATE ON accounts FOR EACH ROW WHEN (NEW.balance > 10000) EXECUTE FUNCTION check_account_update"; let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_update")]), @@ -5677,8 +5678,7 @@ fn parse_create_after_update_trigger_with_condition() { table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: Some(Expr::Nested(Box::new(Expr::BinaryOp { left: Box::new(Expr::CompoundIdentifier(vec![ Ident::new("NEW"), @@ -5707,6 +5707,7 @@ fn parse_create_instead_of_delete_trigger() { let sql = "CREATE TRIGGER check_delete INSTEAD OF DELETE ON accounts FOR EACH ROW EXECUTE FUNCTION check_account_deletes"; let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_delete")]), @@ -5716,8 +5717,7 @@ fn parse_create_instead_of_delete_trigger() { table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: None, exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, @@ -5739,6 +5739,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { let sql = "CREATE CONSTRAINT TRIGGER check_multiple_events BEFORE INSERT OR UPDATE OR DELETE ON accounts DEFERRABLE INITIALLY DEFERRED FOR EACH ROW EXECUTE FUNCTION check_account_changes"; let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: true, name: ObjectName::from(vec![Ident::new("check_multiple_events")]), @@ -5752,8 +5753,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { table_name: ObjectName::from(vec![Ident::new("accounts")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: None, exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, @@ -5779,6 +5779,7 @@ fn parse_create_trigger_with_referencing() { let sql = "CREATE TRIGGER check_referencing BEFORE INSERT ON accounts REFERENCING NEW TABLE AS new_accounts OLD TABLE AS old_accounts FOR EACH ROW EXECUTE FUNCTION check_account_referencing"; let expected = Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_referencing")]), @@ -5799,8 +5800,7 @@ fn parse_create_trigger_with_referencing() { transition_relation_name: ObjectName::from(vec![Ident::new("old_accounts")]), }, ], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: None, exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, @@ -5826,7 +5826,7 @@ fn parse_create_trigger_invalid_cases() { let invalid_cases = vec![ ( "CREATE TRIGGER check_update BEFORE UPDATE ON accounts FUNCTION check_account_update", - "Expected: FOR, found: FUNCTION" + "Expected: an SQL statement, found: FUNCTION" ), ( "CREATE TRIGGER check_update TOMORROW UPDATE ON accounts EXECUTE FUNCTION check_account_update", @@ -6095,6 +6095,7 @@ fn parse_trigger_related_functions() { create_trigger, Statement::CreateTrigger(CreateTrigger { or_alter: false, + temporary: false, or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), @@ -6104,8 +6105,7 @@ fn parse_trigger_related_functions() { table_name: ObjectName::from(vec![Ident::new("emp")]), referenced_table_name: None, referencing: vec![], - trigger_object: TriggerObject::Row, - include_each: true, + trigger_object: Some(TriggerObjectKind::ForEach(TriggerObject::Row)), condition: None, exec_body: Some(TriggerExecBody { exec_type: TriggerExecBodyType::Function, diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 5083ecd08..f0d6d9b72 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -610,6 +610,281 @@ fn test_update_delete_limit() { } } +#[test] +fn test_create_trigger() { + let statement1 = "CREATE TRIGGER trg_inherit_asset_models AFTER INSERT ON assets FOR EACH ROW BEGIN INSERT INTO users (name) SELECT pam.name FROM users AS pam; END"; + + match sqlite().verified_stmt(statement1) { + Statement::CreateTrigger(CreateTrigger { + or_alter, + temporary, + or_replace, + is_constraint, + name, + period, + period_before_table, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + exec_body: _, + statements_as, + statements: _, + characteristics, + }) => { + assert!(!or_alter); + assert!(!temporary); + assert!(!or_replace); + assert!(!is_constraint); + assert_eq!(name.to_string(), "trg_inherit_asset_models"); + assert_eq!(period, TriggerPeriod::After); + assert!(period_before_table); + assert_eq!(events, vec![TriggerEvent::Insert]); + assert_eq!(table_name.to_string(), "assets"); + assert!(referenced_table_name.is_none()); + assert!(referencing.is_empty()); + assert_eq!( + trigger_object, + Some(TriggerObjectKind::ForEach(TriggerObject::Row)) + ); + assert!(condition.is_none()); + assert!(!statements_as); + assert!(characteristics.is_none()); + } + _ => unreachable!("Expected CREATE TRIGGER statement"), + } + + // Here we check that the variant of CREATE TRIGGER that omits the `FOR EACH ROW` clause, + // which in SQLite may be implicitly assumed, is parsed correctly. + let statement2 = "CREATE TRIGGER log_new_user AFTER INSERT ON users BEGIN INSERT INTO user_log (user_id, action, timestamp) VALUES (NEW.id, 'created', datetime('now')); END"; + + match sqlite().verified_stmt(statement2) { + Statement::CreateTrigger(CreateTrigger { + or_alter, + temporary, + or_replace, + is_constraint, + name, + period, + period_before_table, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + exec_body: _, + statements_as, + statements: _, + characteristics, + }) => { + assert!(!or_alter); + assert!(!temporary); + assert!(!or_replace); + assert!(!is_constraint); + assert_eq!(name.to_string(), "log_new_user"); + assert_eq!(period, TriggerPeriod::After); + assert!(period_before_table); + assert_eq!(events, vec![TriggerEvent::Insert]); + assert_eq!(table_name.to_string(), "users"); + assert!(referenced_table_name.is_none()); + assert!(referencing.is_empty()); + assert!(trigger_object.is_none()); + assert!(condition.is_none()); + assert!(!statements_as); + assert!(characteristics.is_none()); + } + _ => unreachable!("Expected CREATE TRIGGER statement"), + } + + let statement3 = "CREATE TRIGGER cleanup_orders AFTER DELETE ON customers BEGIN DELETE FROM orders WHERE customer_id = OLD.id; DELETE FROM invoices WHERE customer_id = OLD.id; END"; + match sqlite().verified_stmt(statement3) { + Statement::CreateTrigger(CreateTrigger { + or_alter, + temporary, + or_replace, + is_constraint, + name, + period, + period_before_table, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + exec_body: _, + statements_as, + statements: _, + characteristics, + }) => { + assert!(!or_alter); + assert!(!temporary); + assert!(!or_replace); + assert!(!is_constraint); + assert_eq!(name.to_string(), "cleanup_orders"); + assert_eq!(period, TriggerPeriod::After); + assert!(period_before_table); + assert_eq!(events, vec![TriggerEvent::Delete]); + assert_eq!(table_name.to_string(), "customers"); + assert!(referenced_table_name.is_none()); + assert!(referencing.is_empty()); + assert!(trigger_object.is_none()); + assert!(condition.is_none()); + assert!(!statements_as); + assert!(characteristics.is_none()); + } + _ => unreachable!("Expected CREATE TRIGGER statement"), + } + + let statement4 = "CREATE TRIGGER trg_before_update BEFORE UPDATE ON products FOR EACH ROW WHEN NEW.price < 0 BEGIN SELECT RAISE(ABORT, 'Price cannot be negative'); END"; + match sqlite().verified_stmt(statement4) { + Statement::CreateTrigger(CreateTrigger { + or_alter, + temporary, + or_replace, + is_constraint, + name, + period, + period_before_table, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + exec_body: _, + statements_as, + statements: _, + characteristics, + }) => { + assert!(!or_alter); + assert!(!temporary); + assert!(!or_replace); + assert!(!is_constraint); + assert_eq!(name.to_string(), "trg_before_update"); + assert_eq!(period, TriggerPeriod::Before); + assert!(period_before_table); + assert_eq!(events, vec![TriggerEvent::Update(Vec::new())]); + assert_eq!(table_name.to_string(), "products"); + assert!(referenced_table_name.is_none()); + assert!(referencing.is_empty()); + assert_eq!( + trigger_object, + Some(TriggerObjectKind::ForEach(TriggerObject::Row)) + ); + assert!(condition.is_some()); + assert!(!statements_as); + assert!(characteristics.is_none()); + } + _ => unreachable!("Expected CREATE TRIGGER statement"), + } + + // We test a INSTEAD OF trigger on a view + let statement5 = "CREATE TRIGGER trg_instead_of_insert INSTEAD OF INSERT ON my_view BEGIN INSERT INTO my_table (col1, col2) VALUES (NEW.col1, NEW.col2); END"; + match sqlite().verified_stmt(statement5) { + Statement::CreateTrigger(CreateTrigger { + or_alter, + temporary, + or_replace, + is_constraint, + name, + period, + period_before_table, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + exec_body: _, + statements_as, + statements: _, + characteristics, + }) => { + assert!(!or_alter); + assert!(!temporary); + assert!(!or_replace); + assert!(!is_constraint); + assert_eq!(name.to_string(), "trg_instead_of_insert"); + assert_eq!(period, TriggerPeriod::InsteadOf); + assert!(period_before_table); + assert_eq!(events, vec![TriggerEvent::Insert]); + assert_eq!(table_name.to_string(), "my_view"); + assert!(referenced_table_name.is_none()); + assert!(referencing.is_empty()); + assert!(trigger_object.is_none()); + assert!(condition.is_none()); + assert!(!statements_as); + assert!(characteristics.is_none()); + } + _ => unreachable!("Expected CREATE TRIGGER statement"), + } + + // We test a temporary trigger + let statement6 = "CREATE TEMPORARY TRIGGER temp_trigger AFTER INSERT ON temp_table BEGIN UPDATE log_table SET count = count + 1; END"; + match sqlite().verified_stmt(statement6) { + Statement::CreateTrigger(CreateTrigger { + or_alter, + temporary, + or_replace, + is_constraint, + name, + period, + period_before_table, + events, + table_name, + referenced_table_name, + referencing, + trigger_object, + condition, + exec_body: _, + statements_as, + statements: _, + characteristics, + }) => { + assert!(!or_alter); + assert!(temporary); + assert!(!or_replace); + assert!(!is_constraint); + assert_eq!(name.to_string(), "temp_trigger"); + assert_eq!(period, TriggerPeriod::After); + assert!(period_before_table); + assert_eq!(events, vec![TriggerEvent::Insert]); + assert_eq!(table_name.to_string(), "temp_table"); + assert!(referenced_table_name.is_none()); + assert!(referencing.is_empty()); + assert!(trigger_object.is_none()); + assert!(condition.is_none()); + assert!(!statements_as); + assert!(characteristics.is_none()); + } + _ => unreachable!("Expected CREATE TRIGGER statement"), + } +} + +#[test] +fn test_drop_trigger() { + let statement = "DROP TRIGGER IF EXISTS trg_inherit_asset_models"; + + match sqlite().verified_stmt(statement) { + Statement::DropTrigger(DropTrigger { + if_exists, + trigger_name, + table_name, + option, + }) => { + assert!(if_exists); + assert_eq!(trigger_name.to_string(), "trg_inherit_asset_models"); + assert!(table_name.is_none()); + assert!(option.is_none()); + } + _ => unreachable!("Expected DROP TRIGGER statement"), + } +} + fn sqlite() -> TestedDialects { TestedDialects::new(vec![Box::new(SQLiteDialect {})]) } From c8531d41a1985951538dae3a78d73ea69436d4f2 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Wed, 15 Oct 2025 13:15:55 +0200 Subject: [PATCH 276/286] Added support for MATCH syntax and unified column option ForeignKey (#2062) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 64 +++++++++++++++++++----------------- src/ast/mod.rs | 25 ++++++++++++++ src/ast/spans.rs | 14 +------- src/ast/table_constraints.rs | 12 ++++--- src/keywords.rs | 2 ++ src/parser/mod.rs | 52 ++++++++++++++++++++++------- tests/sqlparser_common.rs | 24 +++++++++++--- tests/sqlparser_postgres.rs | 49 +++++++++++++++++++++++++++ 8 files changed, 179 insertions(+), 63 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 3294a7a81..4a8678e44 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -30,14 +30,15 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, ArgMode, AttachedToken, CommentDef, - ConditionalStatements, CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, - CreateTableOptions, CreateViewParams, DataType, Expr, FileFormat, FunctionBehavior, - FunctionCalledOnNull, FunctionDesc, FunctionDeterminismSpecifier, FunctionParallel, - HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, HiveSetLocation, Ident, - InitializeKind, MySQLColumnPosition, ObjectName, OnCommit, OneOrManyWithParens, - OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RefreshModeKind, RowAccessPolicy, - SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, TableConstraint, TableVersion, + display_comma_separated, display_separated, + table_constraints::{ForeignKeyConstraint, TableConstraint}, + ArgMode, AttachedToken, CommentDef, ConditionalStatements, CreateFunctionBody, + CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, CreateViewParams, DataType, Expr, + FileFormat, FunctionBehavior, FunctionCalledOnNull, FunctionDesc, FunctionDeterminismSpecifier, + FunctionParallel, HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, + HiveSetLocation, Ident, InitializeKind, MySQLColumnPosition, ObjectName, OnCommit, + OneOrManyWithParens, OperateFunctionArg, OrderByExpr, ProjectionSelect, Query, RefreshModeKind, + RowAccessPolicy, SequenceOptions, Spanned, SqlOption, StorageSerializationPolicy, TableVersion, Tag, TriggerEvent, TriggerExecBody, TriggerObject, TriggerPeriod, TriggerReferencing, Value, ValueWithSpan, WrappedCollection, }; @@ -1559,20 +1560,14 @@ pub enum ColumnOption { is_primary: bool, characteristics: Option, }, - /// A referential integrity constraint (`[FOREIGN KEY REFERENCES - /// () + /// A referential integrity constraint (`REFERENCES () + /// [ MATCH { FULL | PARTIAL | SIMPLE } ] /// { [ON DELETE ] [ON UPDATE ] | /// [ON UPDATE ] [ON DELETE ] - /// } + /// } /// [] /// `). - ForeignKey { - foreign_table: ObjectName, - referred_columns: Vec, - on_delete: Option, - on_update: Option, - characteristics: Option, - }, + ForeignKey(ForeignKeyConstraint), /// `CHECK ()` Check(Expr), /// Dialect-specific options, such as: @@ -1643,6 +1638,12 @@ pub enum ColumnOption { Invisible, } +impl From for ColumnOption { + fn from(fk: ForeignKeyConstraint) -> Self { + ColumnOption::ForeignKey(fk) + } +} + impl fmt::Display for ColumnOption { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { use ColumnOption::*; @@ -1669,24 +1670,25 @@ impl fmt::Display for ColumnOption { } Ok(()) } - ForeignKey { - foreign_table, - referred_columns, - on_delete, - on_update, - characteristics, - } => { - write!(f, "REFERENCES {foreign_table}")?; - if !referred_columns.is_empty() { - write!(f, " ({})", display_comma_separated(referred_columns))?; + ForeignKey(constraint) => { + write!(f, "REFERENCES {}", constraint.foreign_table)?; + if !constraint.referred_columns.is_empty() { + write!( + f, + " ({})", + display_comma_separated(&constraint.referred_columns) + )?; } - if let Some(action) = on_delete { + if let Some(match_kind) = &constraint.match_kind { + write!(f, " {match_kind}")?; + } + if let Some(action) = &constraint.on_delete { write!(f, " ON DELETE {action}")?; } - if let Some(action) = on_update { + if let Some(action) = &constraint.on_update { write!(f, " ON UPDATE {action}")?; } - if let Some(characteristics) = characteristics { + if let Some(characteristics) = &constraint.characteristics { write!(f, " {characteristics}")?; } Ok(()) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index fef8943ef..f4e2825dd 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -657,6 +657,31 @@ pub enum CastKind { DoubleColon, } +/// `MATCH` type for constraint references +/// +/// See: +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ConstraintReferenceMatchKind { + /// `MATCH FULL` + Full, + /// `MATCH PARTIAL` + Partial, + /// `MATCH SIMPLE` + Simple, +} + +impl fmt::Display for ConstraintReferenceMatchKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + Self::Full => write!(f, "MATCH FULL"), + Self::Partial => write!(f, "MATCH PARTIAL"), + Self::Simple => write!(f, "MATCH SIMPLE"), + } + } +} + /// `EXTRACT` syntax variants. /// /// In Snowflake dialect, the `EXTRACT` expression can support either the `from` syntax diff --git a/src/ast/spans.rs b/src/ast/spans.rs index de3439cfc..1e5a96bc4 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -741,19 +741,7 @@ impl Spanned for ColumnOption { ColumnOption::Ephemeral(expr) => expr.as_ref().map_or(Span::empty(), |e| e.span()), ColumnOption::Alias(expr) => expr.span(), ColumnOption::Unique { .. } => Span::empty(), - ColumnOption::ForeignKey { - foreign_table, - referred_columns, - on_delete, - on_update, - characteristics, - } => union_spans( - core::iter::once(foreign_table.span()) - .chain(referred_columns.iter().map(|i| i.span)) - .chain(on_delete.iter().map(|i| i.span())) - .chain(on_update.iter().map(|i| i.span())) - .chain(characteristics.iter().map(|i| i.span())), - ), + ColumnOption::ForeignKey(constraint) => constraint.span(), ColumnOption::Check(expr) => expr.span(), ColumnOption::DialectSpecific(_) => Span::empty(), ColumnOption::CharacterSet(object_name) => object_name.span(), diff --git a/src/ast/table_constraints.rs b/src/ast/table_constraints.rs index afcf62959..ddf0c1253 100644 --- a/src/ast/table_constraints.rs +++ b/src/ast/table_constraints.rs @@ -18,9 +18,9 @@ //! SQL Abstract Syntax Tree (AST) types for table constraints use crate::ast::{ - display_comma_separated, display_separated, ConstraintCharacteristics, Expr, Ident, - IndexColumn, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, ObjectName, - ReferentialAction, + display_comma_separated, display_separated, ConstraintCharacteristics, + ConstraintReferenceMatchKind, Expr, Ident, IndexColumn, IndexOption, IndexType, + KeyOrIndexDisplay, NullsDistinctOption, ObjectName, ReferentialAction, }; use crate::tokenizer::Span; use core::fmt; @@ -189,7 +189,7 @@ impl crate::ast::Spanned for CheckConstraint { } /// A referential integrity constraint (`[ CONSTRAINT ] FOREIGN KEY () -/// REFERENCES () +/// REFERENCES () [ MATCH { FULL | PARTIAL | SIMPLE } ] /// { [ON DELETE ] [ON UPDATE ] | /// [ON UPDATE ] [ON DELETE ] /// }`). @@ -206,6 +206,7 @@ pub struct ForeignKeyConstraint { pub referred_columns: Vec, pub on_delete: Option, pub on_update: Option, + pub match_kind: Option, pub characteristics: Option, } @@ -223,6 +224,9 @@ impl fmt::Display for ForeignKeyConstraint { if !self.referred_columns.is_empty() { write!(f, "({})", display_comma_separated(&self.referred_columns))?; } + if let Some(match_kind) = &self.match_kind { + write!(f, " {match_kind}")?; + } if let Some(action) = &self.on_delete { write!(f, " ON DELETE {action}")?; } diff --git a/src/keywords.rs b/src/keywords.rs index 3c9855222..35bf616d9 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -713,6 +713,7 @@ define_keywords!( PARAMETER, PARQUET, PART, + PARTIAL, PARTITION, PARTITIONED, PARTITIONS, @@ -885,6 +886,7 @@ define_keywords!( SHOW, SIGNED, SIMILAR, + SIMPLE, SKIP, SLOW, SMALLINT, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 70f4d8568..ef583dd37 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -7940,7 +7940,7 @@ impl<'a> Parser<'a> { } pub fn parse_column_def(&mut self) -> Result { - let name = self.parse_identifier()?; + let col_name = self.parse_identifier()?; let data_type = if self.is_column_type_sqlite_unspecified() { DataType::Unspecified } else { @@ -7965,7 +7965,7 @@ impl<'a> Parser<'a> { }; } Ok(ColumnDef { - name, + name: col_name, data_type, options, }) @@ -8065,10 +8065,15 @@ impl<'a> Parser<'a> { // PostgreSQL allows omitting the column list and // uses the primary key column of the foreign table by default let referred_columns = self.parse_parenthesized_column_list(Optional, false)?; + let mut match_kind = None; let mut on_delete = None; let mut on_update = None; loop { - if on_delete.is_none() && self.parse_keywords(&[Keyword::ON, Keyword::DELETE]) { + if match_kind.is_none() && self.parse_keyword(Keyword::MATCH) { + match_kind = Some(self.parse_match_kind()?); + } else if on_delete.is_none() + && self.parse_keywords(&[Keyword::ON, Keyword::DELETE]) + { on_delete = Some(self.parse_referential_action()?); } else if on_update.is_none() && self.parse_keywords(&[Keyword::ON, Keyword::UPDATE]) @@ -8080,13 +8085,20 @@ impl<'a> Parser<'a> { } let characteristics = self.parse_constraint_characteristics()?; - Ok(Some(ColumnOption::ForeignKey { - foreign_table, - referred_columns, - on_delete, - on_update, - characteristics, - })) + Ok(Some( + ForeignKeyConstraint { + name: None, // Column-level constraints don't have names + index_name: None, // Not applicable for column-level constraints + columns: vec![], // Not applicable for column-level constraints + foreign_table, + referred_columns, + on_delete, + on_update, + match_kind, + characteristics, + } + .into(), + )) } else if self.parse_keyword(Keyword::CHECK) { self.expect_token(&Token::LParen)?; // since `CHECK` requires parentheses, we can parse the inner expression in ParserState::Normal @@ -8360,6 +8372,18 @@ impl<'a> Parser<'a> { } } + pub fn parse_match_kind(&mut self) -> Result { + if self.parse_keyword(Keyword::FULL) { + Ok(ConstraintReferenceMatchKind::Full) + } else if self.parse_keyword(Keyword::PARTIAL) { + Ok(ConstraintReferenceMatchKind::Partial) + } else if self.parse_keyword(Keyword::SIMPLE) { + Ok(ConstraintReferenceMatchKind::Simple) + } else { + self.expected("one of FULL, PARTIAL or SIMPLE", self.peek_token()) + } + } + pub fn parse_constraint_characteristics( &mut self, ) -> Result, ParserError> { @@ -8470,10 +8494,15 @@ impl<'a> Parser<'a> { self.expect_keyword_is(Keyword::REFERENCES)?; let foreign_table = self.parse_object_name(false)?; let referred_columns = self.parse_parenthesized_column_list(Optional, false)?; + let mut match_kind = None; let mut on_delete = None; let mut on_update = None; loop { - if on_delete.is_none() && self.parse_keywords(&[Keyword::ON, Keyword::DELETE]) { + if match_kind.is_none() && self.parse_keyword(Keyword::MATCH) { + match_kind = Some(self.parse_match_kind()?); + } else if on_delete.is_none() + && self.parse_keywords(&[Keyword::ON, Keyword::DELETE]) + { on_delete = Some(self.parse_referential_action()?); } else if on_update.is_none() && self.parse_keywords(&[Keyword::ON, Keyword::UPDATE]) @@ -8495,6 +8524,7 @@ impl<'a> Parser<'a> { referred_columns, on_delete, on_update, + match_kind, characteristics, } .into(), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 773937c5a..52f38b10e 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3790,13 +3790,17 @@ fn parse_create_table() { data_type: DataType::Int(None), options: vec![ColumnOptionDef { name: None, - option: ColumnOption::ForeignKey { + option: ColumnOption::ForeignKey(ForeignKeyConstraint { + name: None, + index_name: None, + columns: vec![], foreign_table: ObjectName::from(vec!["othertable".into()]), referred_columns: vec!["a".into(), "b".into()], on_delete: None, on_update: None, + match_kind: None, characteristics: None, - }, + }), }], }, ColumnDef { @@ -3804,13 +3808,17 @@ fn parse_create_table() { data_type: DataType::Int(None), options: vec![ColumnOptionDef { name: None, - option: ColumnOption::ForeignKey { + option: ColumnOption::ForeignKey(ForeignKeyConstraint { + name: None, + index_name: None, + columns: vec![], foreign_table: ObjectName::from(vec!["othertable2".into()]), referred_columns: vec![], on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::NoAction), + match_kind: None, characteristics: None, - }, + }), },], }, ] @@ -3826,6 +3834,7 @@ fn parse_create_table() { referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Restrict), on_update: None, + match_kind: None, characteristics: None, } .into(), @@ -3837,6 +3846,7 @@ fn parse_create_table() { referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::NoAction), on_update: Some(ReferentialAction::Restrict), + match_kind: None, characteristics: None, } .into(), @@ -3848,6 +3858,7 @@ fn parse_create_table() { referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::SetDefault), + match_kind: None, characteristics: None, } .into(), @@ -3859,6 +3870,7 @@ fn parse_create_table() { referred_columns: vec!["longitude".into()], on_delete: None, on_update: Some(ReferentialAction::SetNull), + match_kind: None, characteristics: None, } .into(), @@ -3957,6 +3969,7 @@ fn parse_create_table_with_constraint_characteristics() { referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Restrict), on_update: None, + match_kind: None, characteristics: Some(ConstraintCharacteristics { deferrable: Some(true), initially: Some(DeferrableInitial::Deferred), @@ -3972,6 +3985,7 @@ fn parse_create_table_with_constraint_characteristics() { referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::NoAction), on_update: Some(ReferentialAction::Restrict), + match_kind: None, characteristics: Some(ConstraintCharacteristics { deferrable: Some(true), initially: Some(DeferrableInitial::Immediate), @@ -3987,6 +4001,7 @@ fn parse_create_table_with_constraint_characteristics() { referred_columns: vec!["lat".into()], on_delete: Some(ReferentialAction::Cascade), on_update: Some(ReferentialAction::SetDefault), + match_kind: None, characteristics: Some(ConstraintCharacteristics { deferrable: Some(false), initially: Some(DeferrableInitial::Deferred), @@ -4002,6 +4017,7 @@ fn parse_create_table_with_constraint_characteristics() { referred_columns: vec!["longitude".into()], on_delete: None, on_update: Some(ReferentialAction::SetNull), + match_kind: None, characteristics: Some(ConstraintCharacteristics { deferrable: Some(false), initially: Some(DeferrableInitial::Immediate), diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index e18bf662a..9d08540ad 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -6438,6 +6438,7 @@ fn parse_alter_table_constraint_not_valid() { referred_columns: vec!["ref".into()], on_delete: None, on_update: None, + match_kind: None, characteristics: None, } .into(), @@ -6603,3 +6604,51 @@ fn parse_alter_schema() { _ => unreachable!(), } } + +#[test] +fn parse_foreign_key_match() { + let test_cases = [ + ("MATCH FULL", ConstraintReferenceMatchKind::Full), + ("MATCH SIMPLE", ConstraintReferenceMatchKind::Simple), + ("MATCH PARTIAL", ConstraintReferenceMatchKind::Partial), + ]; + + for (match_clause, expected_kind) in test_cases { + // Test column-level foreign key + let sql = format!("CREATE TABLE t (id INT REFERENCES other_table (id) {match_clause})"); + let statement = pg_and_generic().verified_stmt(&sql); + match statement { + Statement::CreateTable(CreateTable { columns, .. }) => { + match &columns[0].options[0].option { + ColumnOption::ForeignKey(constraint) => { + assert_eq!(constraint.match_kind, Some(expected_kind)); + } + _ => panic!("Expected ColumnOption::ForeignKey"), + } + } + _ => unreachable!("{:?} should parse to Statement::CreateTable", sql), + } + + // Test table-level foreign key constraint + let sql = format!( + "CREATE TABLE t (id INT, FOREIGN KEY (id) REFERENCES other_table(id) {match_clause})" + ); + let statement = pg_and_generic().verified_stmt(&sql); + match statement { + Statement::CreateTable(CreateTable { constraints, .. }) => match &constraints[0] { + TableConstraint::ForeignKey(constraint) => { + assert_eq!(constraint.match_kind, Some(expected_kind)); + } + _ => panic!("Expected TableConstraint::ForeignKey"), + }, + _ => unreachable!("{:?} should parse to Statement::CreateTable", sql), + } + } +} + +#[test] +fn parse_foreign_key_match_with_actions() { + let sql = "CREATE TABLE orders (order_id INT REFERENCES another_table (id) MATCH FULL ON DELETE CASCADE ON UPDATE RESTRICT, customer_id INT, CONSTRAINT fk_customer FOREIGN KEY (customer_id) REFERENCES customers(customer_id) MATCH SIMPLE ON DELETE SET NULL ON UPDATE CASCADE)"; + + pg_and_generic().verified_stmt(sql); +} From 218f43cf2de1b436637430bbd936613b30e277a9 Mon Sep 17 00:00:00 2001 From: niebayes Date: Thu, 16 Oct 2025 16:37:01 +0800 Subject: [PATCH 277/286] chore: add stack overflow warning for Visitor and VisitorMut (#2068) --- src/ast/visitor.rs | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index 7840f0e14..328f925f7 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -182,6 +182,10 @@ visit_noop!(bigdecimal::BigDecimal); /// ``` pub trait Visitor { /// Type returned when the recursion returns early. + /// + /// Important note: The `Break` type should be kept as small as possible to prevent + /// stack overflow during recursion. If you need to return an error, consider + /// boxing it with `Box` to minimize stack usage. type Break; /// Invoked for any queries that appear in the AST before visiting children @@ -290,6 +294,10 @@ pub trait Visitor { /// ``` pub trait VisitorMut { /// Type returned when the recursion returns early. + /// + /// Important note: The `Break` type should be kept as small as possible to prevent + /// stack overflow during recursion. If you need to return an error, consider + /// boxing it with `Box` to minimize stack usage. type Break; /// Invoked for any queries that appear in the AST before visiting children From f861566c8dca813d504df74cc99383680c9dac51 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Thu, 16 Oct 2025 10:38:50 +0200 Subject: [PATCH 278/286] Reused `CheckConstraint` in `ColumnOption` (#2063) --- src/ast/ddl.rs | 11 ++++++++--- src/ast/spans.rs | 2 +- src/parser/mod.rs | 9 ++++++++- tests/sqlparser_common.rs | 6 +++++- 4 files changed, 22 insertions(+), 6 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 4a8678e44..84dc74d96 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -31,7 +31,7 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ display_comma_separated, display_separated, - table_constraints::{ForeignKeyConstraint, TableConstraint}, + table_constraints::{CheckConstraint, ForeignKeyConstraint, TableConstraint}, ArgMode, AttachedToken, CommentDef, ConditionalStatements, CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, CreateViewParams, DataType, Expr, FileFormat, FunctionBehavior, FunctionCalledOnNull, FunctionDesc, FunctionDeterminismSpecifier, @@ -1569,7 +1569,7 @@ pub enum ColumnOption { /// `). ForeignKey(ForeignKeyConstraint), /// `CHECK ()` - Check(Expr), + Check(CheckConstraint), /// Dialect-specific options, such as: /// - MySQL's `AUTO_INCREMENT` or SQLite's `AUTOINCREMENT` /// - ... @@ -1638,6 +1638,11 @@ pub enum ColumnOption { Invisible, } +impl From for ColumnOption { + fn from(c: CheckConstraint) -> Self { + ColumnOption::Check(c) + } +} impl From for ColumnOption { fn from(fk: ForeignKeyConstraint) -> Self { ColumnOption::ForeignKey(fk) @@ -1693,7 +1698,7 @@ impl fmt::Display for ColumnOption { } Ok(()) } - Check(expr) => write!(f, "CHECK ({expr})"), + Check(constraint) => write!(f, "{constraint}"), DialectSpecific(val) => write!(f, "{}", display_separated(val, " ")), CharacterSet(n) => write!(f, "CHARACTER SET {n}"), Collation(n) => write!(f, "COLLATE {n}"), diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 1e5a96bc4..5d82c7339 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -741,8 +741,8 @@ impl Spanned for ColumnOption { ColumnOption::Ephemeral(expr) => expr.as_ref().map_or(Span::empty(), |e| e.span()), ColumnOption::Alias(expr) => expr.span(), ColumnOption::Unique { .. } => Span::empty(), + ColumnOption::Check(constraint) => constraint.span(), ColumnOption::ForeignKey(constraint) => constraint.span(), - ColumnOption::Check(expr) => expr.span(), ColumnOption::DialectSpecific(_) => Span::empty(), ColumnOption::CharacterSet(object_name) => object_name.span(), ColumnOption::Collation(object_name) => object_name.span(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ef583dd37..ef31c41f2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8104,7 +8104,14 @@ impl<'a> Parser<'a> { // since `CHECK` requires parentheses, we can parse the inner expression in ParserState::Normal let expr: Expr = self.with_state(ParserState::Normal, |p| p.parse_expr())?; self.expect_token(&Token::RParen)?; - Ok(Some(ColumnOption::Check(expr))) + Ok(Some( + CheckConstraint { + name: None, // Column-level check constraints don't have names + expr: Box::new(expr), + enforced: None, // Could be extended later to support MySQL ENFORCED/NOT ENFORCED + } + .into(), + )) } else if self.parse_keyword(Keyword::AUTO_INCREMENT) && dialect_of!(self is MySqlDialect | GenericDialect) { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 52f38b10e..9f807ecfe 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3781,7 +3781,11 @@ fn parse_create_table() { }, ColumnOptionDef { name: None, - option: ColumnOption::Check(verified_expr("constrained > 0")), + option: ColumnOption::Check(CheckConstraint { + name: None, + expr: Box::new(verified_expr("constrained > 0")), + enforced: None, + }), }, ], }, From 9cc9f9ab38747926f09a9d5570a91adf9c2fce28 Mon Sep 17 00:00:00 2001 From: Luca Cappelletti Date: Tue, 21 Oct 2025 10:05:02 +0200 Subject: [PATCH 279/286] Refactored `ColumnOption::Unique` to reuse `UniqueConstraint` and `PrimaryKeyConstraint` (#2064) --- src/ast/ddl.rs | 58 +++++++++++++++++++++++++++++++-------- src/ast/query.rs | 12 +++++++- src/ast/spans.rs | 6 ++-- src/parser/mod.rs | 32 +++++++++++++++------ tests/sqlparser_common.rs | 40 +++++++++++++++++++-------- tests/sqlparser_mysql.rs | 36 ++++++++++++++++-------- tests/sqlparser_sqlite.rs | 22 ++++++++++----- 7 files changed, 152 insertions(+), 54 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 84dc74d96..f66165368 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -31,7 +31,10 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ display_comma_separated, display_separated, - table_constraints::{CheckConstraint, ForeignKeyConstraint, TableConstraint}, + table_constraints::{ + CheckConstraint, ForeignKeyConstraint, PrimaryKeyConstraint, TableConstraint, + UniqueConstraint, + }, ArgMode, AttachedToken, CommentDef, ConditionalStatements, CreateFunctionBody, CreateFunctionUsing, CreateTableLikeKind, CreateTableOptions, CreateViewParams, DataType, Expr, FileFormat, FunctionBehavior, FunctionCalledOnNull, FunctionDesc, FunctionDeterminismSpecifier, @@ -55,6 +58,22 @@ pub struct IndexColumn { pub operator_class: Option, } +impl From for IndexColumn { + fn from(c: Ident) -> Self { + Self { + column: OrderByExpr::from(c), + operator_class: None, + } + } +} + +impl<'a> From<&'a str> for IndexColumn { + fn from(c: &'a str) -> Self { + let ident = Ident::new(c); + ident.into() + } +} + impl fmt::Display for IndexColumn { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.column)?; @@ -1555,11 +1574,10 @@ pub enum ColumnOption { /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/create/table#default_values) Alias(Expr), - /// `{ PRIMARY KEY | UNIQUE } []` - Unique { - is_primary: bool, - characteristics: Option, - }, + /// `PRIMARY KEY []` + PrimaryKey(PrimaryKeyConstraint), + /// `UNIQUE []` + Unique(UniqueConstraint), /// A referential integrity constraint (`REFERENCES () /// [ MATCH { FULL | PARTIAL | SIMPLE } ] /// { [ON DELETE ] [ON UPDATE ] | @@ -1638,6 +1656,18 @@ pub enum ColumnOption { Invisible, } +impl From for ColumnOption { + fn from(c: UniqueConstraint) -> Self { + ColumnOption::Unique(c) + } +} + +impl From for ColumnOption { + fn from(c: PrimaryKeyConstraint) -> Self { + ColumnOption::PrimaryKey(c) + } +} + impl From for ColumnOption { fn from(c: CheckConstraint) -> Self { ColumnOption::Check(c) @@ -1665,12 +1695,16 @@ impl fmt::Display for ColumnOption { } } Alias(expr) => write!(f, "ALIAS {expr}"), - Unique { - is_primary, - characteristics, - } => { - write!(f, "{}", if *is_primary { "PRIMARY KEY" } else { "UNIQUE" })?; - if let Some(characteristics) = characteristics { + PrimaryKey(constraint) => { + write!(f, "PRIMARY KEY")?; + if let Some(characteristics) = &constraint.characteristics { + write!(f, " {characteristics}")?; + } + Ok(()) + } + Unique(constraint) => { + write!(f, "UNIQUE")?; + if let Some(characteristics) = &constraint.characteristics { write!(f, " {characteristics}")?; } Ok(()) diff --git a/src/ast/query.rs b/src/ast/query.rs index 3b414cc6e..599b013ab 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -2506,6 +2506,16 @@ pub struct OrderByExpr { pub with_fill: Option, } +impl From for OrderByExpr { + fn from(ident: Ident) -> Self { + OrderByExpr { + expr: Expr::Identifier(ident), + options: OrderByOptions::default(), + with_fill: None, + } + } +} + impl fmt::Display for OrderByExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}{}", self.expr, self.options)?; @@ -2574,7 +2584,7 @@ impl fmt::Display for InterpolateExpr { } } -#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[derive(Default, Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct OrderByOptions { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 5d82c7339..7d2a00095 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -728,7 +728,8 @@ impl Spanned for RaiseStatementValue { /// - [ColumnOption::Null] /// - [ColumnOption::NotNull] /// - [ColumnOption::Comment] -/// - [ColumnOption::Unique]¨ +/// - [ColumnOption::PrimaryKey] +/// - [ColumnOption::Unique] /// - [ColumnOption::DialectSpecific] /// - [ColumnOption::Generated] impl Spanned for ColumnOption { @@ -740,7 +741,8 @@ impl Spanned for ColumnOption { ColumnOption::Materialized(expr) => expr.span(), ColumnOption::Ephemeral(expr) => expr.as_ref().map_or(Span::empty(), |e| e.span()), ColumnOption::Alias(expr) => expr.span(), - ColumnOption::Unique { .. } => Span::empty(), + ColumnOption::PrimaryKey(constraint) => constraint.span(), + ColumnOption::Unique(constraint) => constraint.span(), ColumnOption::Check(constraint) => constraint.span(), ColumnOption::ForeignKey(constraint) => constraint.span(), ColumnOption::DialectSpecific(_) => Span::empty(), diff --git a/src/parser/mod.rs b/src/parser/mod.rs index ef31c41f2..e1dd8c0ec 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -8050,16 +8050,32 @@ impl<'a> Parser<'a> { } } else if self.parse_keywords(&[Keyword::PRIMARY, Keyword::KEY]) { let characteristics = self.parse_constraint_characteristics()?; - Ok(Some(ColumnOption::Unique { - is_primary: true, - characteristics, - })) + Ok(Some( + PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics, + } + .into(), + )) } else if self.parse_keyword(Keyword::UNIQUE) { let characteristics = self.parse_constraint_characteristics()?; - Ok(Some(ColumnOption::Unique { - is_primary: false, - characteristics, - })) + Ok(Some( + UniqueConstraint { + name: None, + index_name: None, + index_type_display: KeyOrIndexDisplay::None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics, + nulls_distinct: NullsDistinctOption::None, + } + .into(), + )) } else if self.parse_keyword(Keyword::REFERENCES) { let foreign_table = self.parse_object_name(false)?; // PostgreSQL allows omitting the column list and diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 9f807ecfe..659b37ca6 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -3763,10 +3763,14 @@ fn parse_create_table() { }, ColumnOptionDef { name: Some("pkey".into()), - option: ColumnOption::Unique { - is_primary: true, - characteristics: None - }, + option: ColumnOption::PrimaryKey(PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: None, + }), }, ColumnOptionDef { name: None, @@ -3774,10 +3778,16 @@ fn parse_create_table() { }, ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: false, - characteristics: None - }, + option: ColumnOption::Unique(UniqueConstraint { + name: None, + index_name: None, + index_type_display: KeyOrIndexDisplay::None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: None, + nulls_distinct: NullsDistinctOption::None, + }), }, ColumnOptionDef { name: None, @@ -4106,10 +4116,16 @@ fn parse_create_table_column_constraint_characteristics() { data_type: DataType::Int(None), options: vec![ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: false, - characteristics: expected_value - } + option: ColumnOption::Unique(UniqueConstraint { + name: None, + index_name: None, + index_type_display: KeyOrIndexDisplay::None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: expected_value, + nulls_distinct: NullsDistinctOption::None, + }) }] }], "{message}" diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index e0ddecf32..26ef58fd6 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -638,10 +638,14 @@ fn parse_create_table_auto_increment() { options: vec![ ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: true, - characteristics: None - }, + option: ColumnOption::PrimaryKey(PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: None, + }), }, ColumnOptionDef { name: None, @@ -743,10 +747,14 @@ fn parse_create_table_primary_and_unique_key() { options: vec![ ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: true, - characteristics: None - }, + option: ColumnOption::PrimaryKey(PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: None, + }), }, ColumnOptionDef { name: None, @@ -1382,10 +1390,14 @@ fn parse_quote_identifiers() { data_type: DataType::Int(None), options: vec![ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: true, - characteristics: None - }, + option: ColumnOption::PrimaryKey(PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: None, + }), }], }], columns diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index f0d6d9b72..943e7a230 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -217,10 +217,14 @@ fn parse_create_table_auto_increment() { options: vec![ ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: true, - characteristics: None - }, + option: ColumnOption::PrimaryKey(PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], + characteristics: None, + }), }, ColumnOptionDef { name: None, @@ -245,10 +249,14 @@ fn parse_create_table_primary_key_asc_desc() { options: vec![ ColumnOptionDef { name: None, - option: ColumnOption::Unique { - is_primary: true, + option: ColumnOption::PrimaryKey(PrimaryKeyConstraint { + name: None, + index_name: None, + index_type: None, + columns: vec![], + index_options: vec![], characteristics: None, - }, + }), }, ColumnOptionDef { name: None, From 6b352eaffdd6e8339d062ca8bc671c2c9b415c1e Mon Sep 17 00:00:00 2001 From: Yoav Cohen <59807311+yoavcloud@users.noreply.github.com> Date: Wed, 22 Oct 2025 11:11:19 +0200 Subject: [PATCH 280/286] Redshift: more copy options (#2072) --- src/ast/mod.rs | 34 ++++++++++++++++++++++++++++++++++ src/keywords.rs | 4 ++++ src/parser/mod.rs | 31 +++++++++++++++++++++++++++++++ tests/sqlparser_common.rs | 28 ++++++++++++++++++++++++++-- 4 files changed, 95 insertions(+), 2 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index f4e2825dd..176d36545 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -8210,6 +8210,8 @@ pub enum CopyLegacyOption { Bzip2, /// CLEANPATH CleanPath, + /// COMPUPDATE [ PRESET | { ON | TRUE } | { OFF | FALSE } ] + CompUpdate { preset: bool, enabled: Option }, /// CSV ... Csv(Vec), /// DATEFORMAT \[ AS \] {'dateformat_string' | 'auto' } @@ -8250,8 +8252,12 @@ pub enum CopyLegacyOption { PartitionBy(UnloadPartitionBy), /// REGION \[ AS \] 'aws-region' } Region(String), + /// REMOVEQUOTES + RemoveQuotes, /// ROWGROUPSIZE \[ AS \] size \[ MB | GB \] RowGroupSize(FileSize), + /// STATUPDATE [ { ON | TRUE } | { OFF | FALSE } ] + StatUpdate(Option), /// TIMEFORMAT \[ AS \] {'timeformat_string' | 'auto' | 'epochsecs' | 'epochmillisecs' } TimeFormat(Option), /// TRUNCATECOLUMNS @@ -8278,6 +8284,22 @@ impl fmt::Display for CopyLegacyOption { BlankAsNull => write!(f, "BLANKSASNULL"), Bzip2 => write!(f, "BZIP2"), CleanPath => write!(f, "CLEANPATH"), + CompUpdate { preset, enabled } => { + write!(f, "COMPUPDATE")?; + if *preset { + write!(f, " PRESET")?; + } else if let Some(enabled) = enabled { + write!( + f, + "{}", + match enabled { + true => " TRUE", + false => " FALSE", + } + )?; + } + Ok(()) + } Csv(opts) => { write!(f, "CSV")?; if !opts.is_empty() { @@ -8324,7 +8346,19 @@ impl fmt::Display for CopyLegacyOption { Parquet => write!(f, "PARQUET"), PartitionBy(p) => write!(f, "{p}"), Region(region) => write!(f, "REGION '{}'", value::escape_single_quote_string(region)), + RemoveQuotes => write!(f, "REMOVEQUOTES"), RowGroupSize(file_size) => write!(f, "ROWGROUPSIZE {file_size}"), + StatUpdate(enabled) => { + write!( + f, + "STATUPDATE{}", + match enabled { + Some(true) => " TRUE", + Some(false) => " FALSE", + _ => "", + } + ) + } TimeFormat(fmt) => { write!(f, "TIMEFORMAT")?; if let Some(fmt) = fmt { diff --git a/src/keywords.rs b/src/keywords.rs index 35bf616d9..319c57827 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -215,6 +215,7 @@ define_keywords!( COMMITTED, COMPATIBLE, COMPRESSION, + COMPUPDATE, COMPUTE, CONCURRENTLY, CONDITION, @@ -749,6 +750,7 @@ define_keywords!( PRECISION, PREPARE, PRESERVE, + PRESET, PREWHERE, PRIMARY, PRINT, @@ -801,6 +803,7 @@ define_keywords!( RELEASES, REMOTE, REMOVE, + REMOVEQUOTES, RENAME, REORG, REPAIR, @@ -915,6 +918,7 @@ define_keywords!( STATS_AUTO_RECALC, STATS_PERSISTENT, STATS_SAMPLE_PAGES, + STATUPDATE, STATUS, STDDEV_POP, STDDEV_SAMP, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index e1dd8c0ec..b7d69f308 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9741,6 +9741,7 @@ impl<'a> Parser<'a> { Keyword::BLANKSASNULL, Keyword::BZIP2, Keyword::CLEANPATH, + Keyword::COMPUPDATE, Keyword::CSV, Keyword::DATEFORMAT, Keyword::DELIMITER, @@ -9761,7 +9762,9 @@ impl<'a> Parser<'a> { Keyword::PARQUET, Keyword::PARTITION, Keyword::REGION, + Keyword::REMOVEQUOTES, Keyword::ROWGROUPSIZE, + Keyword::STATUPDATE, Keyword::TIMEFORMAT, Keyword::TRUNCATECOLUMNS, Keyword::ZSTD, @@ -9782,6 +9785,20 @@ impl<'a> Parser<'a> { Some(Keyword::BLANKSASNULL) => CopyLegacyOption::BlankAsNull, Some(Keyword::BZIP2) => CopyLegacyOption::Bzip2, Some(Keyword::CLEANPATH) => CopyLegacyOption::CleanPath, + Some(Keyword::COMPUPDATE) => { + let preset = self.parse_keyword(Keyword::PRESET); + let enabled = match self.parse_one_of_keywords(&[ + Keyword::TRUE, + Keyword::FALSE, + Keyword::ON, + Keyword::OFF, + ]) { + Some(Keyword::TRUE) | Some(Keyword::ON) => Some(true), + Some(Keyword::FALSE) | Some(Keyword::OFF) => Some(false), + _ => None, + }; + CopyLegacyOption::CompUpdate { preset, enabled } + } Some(Keyword::CSV) => CopyLegacyOption::Csv({ let mut opts = vec![]; while let Some(opt) = @@ -9870,11 +9887,25 @@ impl<'a> Parser<'a> { let region = self.parse_literal_string()?; CopyLegacyOption::Region(region) } + Some(Keyword::REMOVEQUOTES) => CopyLegacyOption::RemoveQuotes, Some(Keyword::ROWGROUPSIZE) => { let _ = self.parse_keyword(Keyword::AS); let file_size = self.parse_file_size()?; CopyLegacyOption::RowGroupSize(file_size) } + Some(Keyword::STATUPDATE) => { + let enabled = match self.parse_one_of_keywords(&[ + Keyword::TRUE, + Keyword::FALSE, + Keyword::ON, + Keyword::OFF, + ]) { + Some(Keyword::TRUE) | Some(Keyword::ON) => Some(true), + Some(Keyword::FALSE) | Some(Keyword::OFF) => Some(false), + _ => None, + }; + CopyLegacyOption::StatUpdate(enabled) + } Some(Keyword::TIMEFORMAT) => { let _ = self.parse_keyword(Keyword::AS); let fmt = if matches!(self.peek_token().token, Token::SingleQuotedString(_)) { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 659b37ca6..99b7ac3fa 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -17126,7 +17126,19 @@ fn parse_copy_options() { "IAM_ROLE DEFAULT ", "IGNOREHEADER AS 1 ", "TIMEFORMAT AS 'auto' ", - "TRUNCATECOLUMNS", + "TRUNCATECOLUMNS ", + "REMOVEQUOTES ", + "COMPUPDATE ", + "COMPUPDATE PRESET ", + "COMPUPDATE ON ", + "COMPUPDATE OFF ", + "COMPUPDATE TRUE ", + "COMPUPDATE FALSE ", + "STATUPDATE ", + "STATUPDATE ON ", + "STATUPDATE OFF ", + "STATUPDATE TRUE ", + "STATUPDATE FALSE", ), concat!( "COPY dst (c1, c2, c3) FROM 's3://redshift-downloads/tickit/category_pipe.txt' ", @@ -17139,7 +17151,19 @@ fn parse_copy_options() { "IAM_ROLE DEFAULT ", "IGNOREHEADER 1 ", "TIMEFORMAT 'auto' ", - "TRUNCATECOLUMNS", + "TRUNCATECOLUMNS ", + "REMOVEQUOTES ", + "COMPUPDATE ", + "COMPUPDATE PRESET ", + "COMPUPDATE TRUE ", + "COMPUPDATE FALSE ", + "COMPUPDATE TRUE ", + "COMPUPDATE FALSE ", + "STATUPDATE ", + "STATUPDATE TRUE ", + "STATUPDATE FALSE ", + "STATUPDATE TRUE ", + "STATUPDATE FALSE", ), ); one_statement_parses_to( From 67684c84d4c2589356c411ea4917dcf1defcd77c Mon Sep 17 00:00:00 2001 From: Thomas Kluyver Date: Wed, 22 Oct 2025 10:12:06 +0100 Subject: [PATCH 281/286] SQLite: make period optional for CREATE TRIGGER (#2071) Co-authored-by: Ifeanyi Ubah --- src/ast/ddl.rs | 16 ++++++++++------ src/dialect/mssql.rs | 2 +- src/parser/mod.rs | 2 +- tests/sqlparser_mssql.rs | 2 +- tests/sqlparser_mysql.rs | 2 +- tests/sqlparser_postgres.rs | 14 +++++++------- tests/sqlparser_sqlite.rs | 16 ++++++++++------ 7 files changed, 31 insertions(+), 23 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index f66165368..fd481213f 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -3063,7 +3063,7 @@ pub struct CreateTrigger { /// FOR EACH ROW /// EXECUTE FUNCTION trigger_function(); /// ``` - pub period: TriggerPeriod, + pub period: Option, /// Whether the trigger period was specified before the target table name. /// This does not refer to whether the period is BEFORE, AFTER, or INSTEAD OF, /// but rather the position of the period clause in relation to the table name. @@ -3132,14 +3132,18 @@ impl Display for CreateTrigger { )?; if *period_before_table { - write!(f, "{period}")?; + if let Some(p) = period { + write!(f, "{p} ")?; + } if !events.is_empty() { - write!(f, " {}", display_separated(events, " OR "))?; + write!(f, "{} ", display_separated(events, " OR "))?; } - write!(f, " ON {table_name}")?; - } else { write!(f, "ON {table_name}")?; - write!(f, " {period}")?; + } else { + write!(f, "ON {table_name} ")?; + if let Some(p) = period { + write!(f, "{p}")?; + } if !events.is_empty() { write!(f, " {}", display_separated(events, ", "))?; } diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs index f1d54cd67..e1902b389 100644 --- a/src/dialect/mssql.rs +++ b/src/dialect/mssql.rs @@ -258,7 +258,7 @@ impl MsSqlDialect { or_replace: false, is_constraint: false, name, - period, + period: Some(period), period_before_table: false, events, table_name, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b7d69f308..b44171c7d 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -5592,7 +5592,7 @@ impl<'a> Parser<'a> { } let name = self.parse_object_name(false)?; - let period = self.parse_trigger_period()?; + let period = self.maybe_parse(|parser| parser.parse_trigger_period())?; let events = self.parse_keyword_separated(Keyword::OR, Parser::parse_trigger_event)?; self.expect_keyword_is(Keyword::ON)?; diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index e11c79f01..a947db49b 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -2392,7 +2392,7 @@ fn parse_create_trigger() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("reminder1")]), - period: TriggerPeriod::After, + period: Some(TriggerPeriod::After), period_before_table: false, events: vec![TriggerEvent::Insert, TriggerEvent::Update(vec![]),], table_name: ObjectName::from(vec![Ident::new("Sales"), Ident::new("Customer")]), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 26ef58fd6..e43df87ab 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -4034,7 +4034,7 @@ fn parse_create_trigger() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), - period: TriggerPeriod::Before, + period: Some(TriggerPeriod::Before), period_before_table: true, events: vec![TriggerEvent::Insert], table_name: ObjectName::from(vec![Ident::new("emp")]), diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 9d08540ad..bcc154287 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5640,7 +5640,7 @@ fn parse_create_simple_before_insert_trigger() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_insert")]), - period: TriggerPeriod::Before, + period: Some(TriggerPeriod::Before), period_before_table: true, events: vec![TriggerEvent::Insert], table_name: ObjectName::from(vec![Ident::new("accounts")]), @@ -5672,7 +5672,7 @@ fn parse_create_after_update_trigger_with_condition() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_update")]), - period: TriggerPeriod::After, + period: Some(TriggerPeriod::After), period_before_table: true, events: vec![TriggerEvent::Update(vec![])], table_name: ObjectName::from(vec![Ident::new("accounts")]), @@ -5711,7 +5711,7 @@ fn parse_create_instead_of_delete_trigger() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_delete")]), - period: TriggerPeriod::InsteadOf, + period: Some(TriggerPeriod::InsteadOf), period_before_table: true, events: vec![TriggerEvent::Delete], table_name: ObjectName::from(vec![Ident::new("accounts")]), @@ -5743,7 +5743,7 @@ fn parse_create_trigger_with_multiple_events_and_deferrable() { or_replace: false, is_constraint: true, name: ObjectName::from(vec![Ident::new("check_multiple_events")]), - period: TriggerPeriod::Before, + period: Some(TriggerPeriod::Before), period_before_table: true, events: vec![ TriggerEvent::Insert, @@ -5783,7 +5783,7 @@ fn parse_create_trigger_with_referencing() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("check_referencing")]), - period: TriggerPeriod::Before, + period: Some(TriggerPeriod::Before), period_before_table: true, events: vec![TriggerEvent::Insert], table_name: ObjectName::from(vec![Ident::new("accounts")]), @@ -5830,7 +5830,7 @@ fn parse_create_trigger_invalid_cases() { ), ( "CREATE TRIGGER check_update TOMORROW UPDATE ON accounts EXECUTE FUNCTION check_account_update", - "Expected: one of FOR or BEFORE or AFTER or INSTEAD, found: TOMORROW" + "Expected: one of INSERT or UPDATE or DELETE or TRUNCATE, found: TOMORROW" ), ( "CREATE TRIGGER check_update BEFORE SAVE ON accounts EXECUTE FUNCTION check_account_update", @@ -6099,7 +6099,7 @@ fn parse_trigger_related_functions() { or_replace: false, is_constraint: false, name: ObjectName::from(vec![Ident::new("emp_stamp")]), - period: TriggerPeriod::Before, + period: Some(TriggerPeriod::Before), period_before_table: true, events: vec![TriggerEvent::Insert, TriggerEvent::Update(vec![])], table_name: ObjectName::from(vec![Ident::new("emp")]), diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index 943e7a230..f1f6cf49b 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -647,7 +647,7 @@ fn test_create_trigger() { assert!(!or_replace); assert!(!is_constraint); assert_eq!(name.to_string(), "trg_inherit_asset_models"); - assert_eq!(period, TriggerPeriod::After); + assert_eq!(period, Some(TriggerPeriod::After)); assert!(period_before_table); assert_eq!(events, vec![TriggerEvent::Insert]); assert_eq!(table_name.to_string(), "assets"); @@ -693,7 +693,7 @@ fn test_create_trigger() { assert!(!or_replace); assert!(!is_constraint); assert_eq!(name.to_string(), "log_new_user"); - assert_eq!(period, TriggerPeriod::After); + assert_eq!(period, Some(TriggerPeriod::After)); assert!(period_before_table); assert_eq!(events, vec![TriggerEvent::Insert]); assert_eq!(table_name.to_string(), "users"); @@ -733,7 +733,7 @@ fn test_create_trigger() { assert!(!or_replace); assert!(!is_constraint); assert_eq!(name.to_string(), "cleanup_orders"); - assert_eq!(period, TriggerPeriod::After); + assert_eq!(period, Some(TriggerPeriod::After)); assert!(period_before_table); assert_eq!(events, vec![TriggerEvent::Delete]); assert_eq!(table_name.to_string(), "customers"); @@ -773,7 +773,7 @@ fn test_create_trigger() { assert!(!or_replace); assert!(!is_constraint); assert_eq!(name.to_string(), "trg_before_update"); - assert_eq!(period, TriggerPeriod::Before); + assert_eq!(period, Some(TriggerPeriod::Before)); assert!(period_before_table); assert_eq!(events, vec![TriggerEvent::Update(Vec::new())]); assert_eq!(table_name.to_string(), "products"); @@ -817,7 +817,7 @@ fn test_create_trigger() { assert!(!or_replace); assert!(!is_constraint); assert_eq!(name.to_string(), "trg_instead_of_insert"); - assert_eq!(period, TriggerPeriod::InsteadOf); + assert_eq!(period, Some(TriggerPeriod::InsteadOf)); assert!(period_before_table); assert_eq!(events, vec![TriggerEvent::Insert]); assert_eq!(table_name.to_string(), "my_view"); @@ -858,7 +858,7 @@ fn test_create_trigger() { assert!(!or_replace); assert!(!is_constraint); assert_eq!(name.to_string(), "temp_trigger"); - assert_eq!(period, TriggerPeriod::After); + assert_eq!(period, Some(TriggerPeriod::After)); assert!(period_before_table); assert_eq!(events, vec![TriggerEvent::Insert]); assert_eq!(table_name.to_string(), "temp_table"); @@ -871,6 +871,10 @@ fn test_create_trigger() { } _ => unreachable!("Expected CREATE TRIGGER statement"), } + + // We test a trigger defined without a period (BEFORE/AFTER/INSTEAD OF) + let statement7 = "CREATE TRIGGER trg_inherit_asset_models INSERT ON assets FOR EACH ROW BEGIN INSERT INTO users (name) SELECT pam.name FROM users AS pam; END"; + sqlite().verified_stmt(statement7); } #[test] From bc6e1d6e1a960c0c220c712a16b5046f563a5f6e Mon Sep 17 00:00:00 2001 From: Andriy Romanov Date: Thu, 30 Oct 2025 02:20:41 -0700 Subject: [PATCH 282/286] Added TIMESTAMP_NTZ type support with precision to Snowflake dialect (#2080) --- src/ast/data_type.rs | 6 ++++-- src/parser/mod.rs | 4 +++- tests/sqlparser_databricks.rs | 6 +++--- tests/sqlparser_snowflake.rs | 15 +++++++++++++++ 4 files changed, 25 insertions(+), 6 deletions(-) diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index f8523e844..6da6a90d0 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -375,7 +375,7 @@ pub enum DataType { /// Databricks timestamp without time zone. See [1]. /// /// [1]: https://docs.databricks.com/aws/en/sql/language-manual/data-types/timestamp-ntz-type - TimestampNtz, + TimestampNtz(Option), /// Interval type. Interval { /// [PostgreSQL] fields specification like `INTERVAL YEAR TO MONTH`. @@ -676,7 +676,9 @@ impl fmt::Display for DataType { DataType::Timestamp(precision, timezone_info) => { format_datetime_precision_and_tz(f, "TIMESTAMP", precision, timezone_info) } - DataType::TimestampNtz => write!(f, "TIMESTAMP_NTZ"), + DataType::TimestampNtz(precision) => { + format_type_with_optional_length(f, "TIMESTAMP_NTZ", precision, false) + } DataType::Datetime64(precision, timezone) => { format_clickhouse_datetime_precision_and_timezone( f, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index b44171c7d..1248c918f 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -10542,7 +10542,9 @@ impl<'a> Parser<'a> { self.parse_optional_precision()?, TimezoneInfo::Tz, )), - Keyword::TIMESTAMP_NTZ => Ok(DataType::TimestampNtz), + Keyword::TIMESTAMP_NTZ => { + Ok(DataType::TimestampNtz(self.parse_optional_precision()?)) + } Keyword::TIME => { let precision = self.parse_optional_precision()?; let tz = if self.parse_keyword(Keyword::WITH) { diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index e01611b6f..92b635339 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -328,7 +328,7 @@ fn data_type_timestamp_ntz() { assert_eq!( databricks().verified_expr("TIMESTAMP_NTZ '2025-03-29T18:52:00'"), Expr::TypedString(TypedString { - data_type: DataType::TimestampNtz, + data_type: DataType::TimestampNtz(None), value: ValueWithSpan { value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()), span: Span::empty(), @@ -345,7 +345,7 @@ fn data_type_timestamp_ntz() { expr: Box::new(Expr::Nested(Box::new(Expr::Identifier( "created_at".into() )))), - data_type: DataType::TimestampNtz, + data_type: DataType::TimestampNtz(None), format: None } ); @@ -357,7 +357,7 @@ fn data_type_timestamp_ntz() { columns, vec![ColumnDef { name: "x".into(), - data_type: DataType::TimestampNtz, + data_type: DataType::TimestampNtz(None), options: vec![], }] ); diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index e7a128343..2be5eae8c 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4638,6 +4638,21 @@ fn test_create_database() { assert!(err.contains("Expected"), "Unexpected error: {err}"); } +#[test] +fn test_timestamp_ntz_with_precision() { + snowflake().verified_stmt("SELECT CAST('2024-01-01 01:00:00' AS TIMESTAMP_NTZ(1))"); + snowflake().verified_stmt("SELECT CAST('2024-01-01 01:00:00' AS TIMESTAMP_NTZ(9))"); + + let select = + snowflake().verified_only_select("SELECT CAST('2024-01-01 01:00:00' AS TIMESTAMP_NTZ(9))"); + match expr_from_projection(only(&select.projection)) { + Expr::Cast { data_type, .. } => { + assert_eq!(*data_type, DataType::TimestampNtz(Some(9))); + } + _ => unreachable!(), + } +} + #[test] fn test_drop_constraints() { snowflake().verified_stmt("ALTER TABLE tbl DROP PRIMARY KEY"); From 308a7231bcbc5c1c8ab71fe38f17b1a21632a6c6 Mon Sep 17 00:00:00 2001 From: Alexander Beedie Date: Thu, 30 Oct 2025 15:57:25 +0400 Subject: [PATCH 283/286] Make `BitwiseNot` ("~") available for all dialects, not just PostgreSQL (#2081) --- src/ast/operator.rs | 50 ++++++++++++++++++------------------- src/parser/mod.rs | 6 +++-- tests/sqlparser_common.rs | 19 ++++++++++++++ tests/sqlparser_postgres.rs | 2 -- 4 files changed, 48 insertions(+), 29 deletions(-) diff --git a/src/ast/operator.rs b/src/ast/operator.rs index d0bb05e3c..58c401f7d 100644 --- a/src/ast/operator.rs +++ b/src/ast/operator.rs @@ -33,35 +33,35 @@ use super::display_separated; #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum UnaryOperator { + /// `@-@` Length or circumference (PostgreSQL/Redshift geometric operator) + /// see + AtDashAt, + /// Unary logical not operator: e.g. `! false` (Hive-specific) + BangNot, + /// Bitwise Not, e.g. `~9` + BitwiseNot, + /// `@@` Center (PostgreSQL/Redshift geometric operator) + /// see + DoubleAt, + /// `#` Number of points in path or polygon (PostgreSQL/Redshift geometric operator) + /// see + Hash, /// Plus, e.g. `+9` Plus, /// Minus, e.g. `-9` Minus, /// Not, e.g. `NOT(true)` Not, - /// Bitwise Not, e.g. `~9` (PostgreSQL-specific) - PGBitwiseNot, - /// Square root, e.g. `|/9` (PostgreSQL-specific) - PGSquareRoot, + /// Absolute value, e.g. `@ -9` (PostgreSQL-specific) + PGAbs, /// Cube root, e.g. `||/27` (PostgreSQL-specific) PGCubeRoot, /// Factorial, e.g. `9!` (PostgreSQL-specific) PGPostfixFactorial, /// Factorial, e.g. `!!9` (PostgreSQL-specific) PGPrefixFactorial, - /// Absolute value, e.g. `@ -9` (PostgreSQL-specific) - PGAbs, - /// Unary logical not operator: e.g. `! false` (Hive-specific) - BangNot, - /// `#` Number of points in path or polygon (PostgreSQL/Redshift geometric operator) - /// see - Hash, - /// `@-@` Length or circumference (PostgreSQL/Redshift geometric operator) - /// see - AtDashAt, - /// `@@` Center (PostgreSQL/Redshift geometric operator) - /// see - DoubleAt, + /// Square root, e.g. `|/9` (PostgreSQL-specific) + PGSquareRoot, /// `?-` Is horizontal? (PostgreSQL/Redshift geometric operator) /// see QuestionDash, @@ -73,19 +73,19 @@ pub enum UnaryOperator { impl fmt::Display for UnaryOperator { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { f.write_str(match self { - UnaryOperator::Plus => "+", + UnaryOperator::AtDashAt => "@-@", + UnaryOperator::BangNot => "!", + UnaryOperator::BitwiseNot => "~", + UnaryOperator::DoubleAt => "@@", + UnaryOperator::Hash => "#", UnaryOperator::Minus => "-", UnaryOperator::Not => "NOT", - UnaryOperator::PGBitwiseNot => "~", - UnaryOperator::PGSquareRoot => "|/", + UnaryOperator::PGAbs => "@", UnaryOperator::PGCubeRoot => "||/", UnaryOperator::PGPostfixFactorial => "!", UnaryOperator::PGPrefixFactorial => "!!", - UnaryOperator::PGAbs => "@", - UnaryOperator::BangNot => "!", - UnaryOperator::Hash => "#", - UnaryOperator::AtDashAt => "@-@", - UnaryOperator::DoubleAt => "@@", + UnaryOperator::PGSquareRoot => "|/", + UnaryOperator::Plus => "+", UnaryOperator::QuestionDash => "?-", UnaryOperator::QuestionPipe => "?|", }) diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 1248c918f..9a01e510b 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -1633,7 +1633,6 @@ impl<'a> Parser<'a> { | tok @ Token::PGSquareRoot | tok @ Token::PGCubeRoot | tok @ Token::AtSign - | tok @ Token::Tilde if dialect_is!(dialect is PostgreSqlDialect) => { let op = match tok { @@ -1641,7 +1640,6 @@ impl<'a> Parser<'a> { Token::PGSquareRoot => UnaryOperator::PGSquareRoot, Token::PGCubeRoot => UnaryOperator::PGCubeRoot, Token::AtSign => UnaryOperator::PGAbs, - Token::Tilde => UnaryOperator::PGBitwiseNot, _ => unreachable!(), }; Ok(Expr::UnaryOp { @@ -1651,6 +1649,10 @@ impl<'a> Parser<'a> { ), }) } + Token::Tilde => Ok(Expr::UnaryOp { + op: UnaryOperator::BitwiseNot, + expr: Box::new(self.parse_subexpr(self.dialect.prec_value(Precedence::PlusMinus))?), + }), tok @ Token::Sharp | tok @ Token::AtDashAt | tok @ Token::AtAt diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 99b7ac3fa..f1ba5df04 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -17613,3 +17613,22 @@ fn test_parse_alter_user() { } verified_stmt("ALTER USER u1 SET DEFAULT_SECONDARY_ROLES=('ALL'), PASSWORD='secret', WORKLOAD_IDENTITY=(TYPE=AWS, ARN='arn:aws:iam::123456789:r1/')"); } + +#[test] +fn parse_generic_unary_ops() { + let unary_ops = &[ + ("~", UnaryOperator::BitwiseNot), + ("-", UnaryOperator::Minus), + ("+", UnaryOperator::Plus), + ]; + for (str_op, op) in unary_ops { + let select = verified_only_select(&format!("SELECT {}expr", &str_op)); + assert_eq!( + UnnamedExpr(UnaryOp { + op: *op, + expr: Box::new(Identifier(Ident::new("expr"))), + }), + select.projection[0] + ); + } +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index bcc154287..9ba0fb978 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -2144,13 +2144,11 @@ fn parse_ampersand_arobase() { #[test] fn parse_pg_unary_ops() { let pg_unary_ops = &[ - ("~", UnaryOperator::PGBitwiseNot), ("|/", UnaryOperator::PGSquareRoot), ("||/", UnaryOperator::PGCubeRoot), ("!!", UnaryOperator::PGPrefixFactorial), ("@", UnaryOperator::PGAbs), ]; - for (str_op, op) in pg_unary_ops { let select = pg().verified_only_select(&format!("SELECT {}a", &str_op)); assert_eq!( From eabde4b41e9607722c23b96699bbd83e0e9bd343 Mon Sep 17 00:00:00 2001 From: Christopher Watford Date: Tue, 11 Nov 2025 03:10:32 -0500 Subject: [PATCH 284/286] feat: Add RESET to the base dialect #2078 (#2079) --- src/ast/mod.rs | 56 ++++++++++++++++++++++++++++++++++----- src/ast/spans.rs | 1 + src/parser/mod.rs | 13 +++++++++ tests/sqlparser_common.rs | 23 ++++++++++++++++ 4 files changed, 86 insertions(+), 7 deletions(-) diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 176d36545..4636e4ba2 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -2787,10 +2787,11 @@ impl fmt::Display for Declare { } /// Sql options of a `CREATE TABLE` statement. -#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash, Default)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum CreateTableOptions { + #[default] None, /// Options specified using the `WITH` keyword. /// e.g. `WITH (description = "123")` @@ -2819,12 +2820,6 @@ pub enum CreateTableOptions { TableProperties(Vec), } -impl Default for CreateTableOptions { - fn default() -> Self { - Self::None - } -} - impl fmt::Display for CreateTableOptions { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { @@ -4263,6 +4258,14 @@ pub enum Statement { /// ``` /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_VACUUM_command.html) Vacuum(VacuumStatement), + /// Restore the value of a run-time parameter to the default value. + /// + /// ```sql + /// RESET configuration_parameter; + /// RESET ALL; + /// ``` + /// [PostgreSQL](https://www.postgresql.org/docs/current/sql-reset.html) + Reset(ResetStatement), } impl From for Statement { @@ -5757,6 +5760,7 @@ impl fmt::Display for Statement { Statement::AlterSchema(s) => write!(f, "{s}"), Statement::Vacuum(s) => write!(f, "{s}"), Statement::AlterUser(s) => write!(f, "{s}"), + Statement::Reset(s) => write!(f, "{s}"), } } } @@ -10519,6 +10523,38 @@ impl fmt::Display for VacuumStatement { } } +/// Variants of the RESET statement +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum Reset { + /// Resets all session parameters to their default values. + ALL, + + /// Resets a specific session parameter to its default value. + ConfigurationParameter(ObjectName), +} + +/// Resets a session parameter to its default value. +/// ```sql +/// RESET { ALL | } +/// ``` +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ResetStatement { + pub reset: Reset, +} + +impl fmt::Display for ResetStatement { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match &self.reset { + Reset::ALL => write!(f, "RESET ALL"), + Reset::ConfigurationParameter(param) => write!(f, "RESET {}", param), + } + } +} + impl From for Statement { fn from(s: Set) -> Self { Self::Set(s) @@ -10759,6 +10795,12 @@ impl From for Statement { } } +impl From for Statement { + fn from(r: ResetStatement) -> Self { + Self::Reset(r) + } +} + #[cfg(test)] mod tests { use crate::tokenizer::Location; diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 7d2a00095..34edabd97 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -475,6 +475,7 @@ impl Spanned for Statement { Statement::AlterSchema(s) => s.span(), Statement::Vacuum(..) => Span::empty(), Statement::AlterUser(..) => Span::empty(), + Statement::Reset(..) => Span::empty(), } } } diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 9a01e510b..f43329be6 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -656,6 +656,7 @@ impl<'a> Parser<'a> { self.prev_token(); self.parse_vacuum() } + Keyword::RESET => self.parse_reset(), _ => self.expected("an SQL statement", next_token), }, Token::LParen => { @@ -17727,6 +17728,18 @@ impl<'a> Parser<'a> { _ => self.expected("expected option value", self.peek_token()), } } + + /// Parses a RESET statement + fn parse_reset(&mut self) -> Result { + if self.parse_keyword(Keyword::ALL) { + return Ok(Statement::Reset(ResetStatement { reset: Reset::ALL })); + } + + let obj = self.parse_object_name(false)?; + Ok(Statement::Reset(ResetStatement { + reset: Reset::ConfigurationParameter(obj), + })) + } } fn maybe_prefixed_expr(expr: Expr, prefix: Option) -> Expr { diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index f1ba5df04..9ea91c642 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -17632,3 +17632,26 @@ fn parse_generic_unary_ops() { ); } } + +#[test] +fn parse_reset_statement() { + match verified_stmt("RESET some_parameter") { + Statement::Reset(ResetStatement { + reset: Reset::ConfigurationParameter(o), + }) => assert_eq!(o, ObjectName::from(vec!["some_parameter".into()])), + _ => unreachable!(), + } + match verified_stmt("RESET some_extension.some_parameter") { + Statement::Reset(ResetStatement { + reset: Reset::ConfigurationParameter(o), + }) => assert_eq!( + o, + ObjectName::from(vec!["some_extension".into(), "some_parameter".into()]) + ), + _ => unreachable!(), + } + match verified_stmt("RESET ALL") { + Statement::Reset(ResetStatement { reset }) => assert_eq!(reset, Reset::ALL), + _ => unreachable!(), + } +} From c439ee9419a8b3e91ae6fc406a2fa88bb2bf1bdd Mon Sep 17 00:00:00 2001 From: Andriy Romanov Date: Tue, 11 Nov 2025 00:10:37 -0800 Subject: [PATCH 285/286] Add snowflake dynamic table parsing (#2083) --- src/ast/ddl.rs | 48 ++++++++++++++++++++++++++----- src/ast/mod.rs | 2 +- src/ast/spans.rs | 3 ++ src/dialect/snowflake.rs | 55 ++++++++++++++++++++++++++++++++---- src/keywords.rs | 1 + src/parser/mod.rs | 6 +++- src/test_utils.rs | 2 +- tests/sqlparser_mysql.rs | 4 +-- tests/sqlparser_snowflake.rs | 8 ++++++ 9 files changed, 112 insertions(+), 17 deletions(-) diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index fd481213f..1ae24a7f4 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -365,6 +365,18 @@ pub enum AlterTableOperation { DropClusteringKey, SuspendRecluster, ResumeRecluster, + /// `REFRESH` + /// + /// Note: this is Snowflake specific for dynamic tables + Refresh, + /// `SUSPEND` + /// + /// Note: this is Snowflake specific for dynamic tables + Suspend, + /// `RESUME` + /// + /// Note: this is Snowflake specific for dynamic tables + Resume, /// `ALGORITHM [=] { DEFAULT | INSTANT | INPLACE | COPY }` /// /// [MySQL]-specific table alter algorithm. @@ -845,6 +857,15 @@ impl fmt::Display for AlterTableOperation { write!(f, "RESUME RECLUSTER")?; Ok(()) } + AlterTableOperation::Refresh => { + write!(f, "REFRESH") + } + AlterTableOperation::Suspend => { + write!(f, "SUSPEND") + } + AlterTableOperation::Resume => { + write!(f, "RESUME") + } AlterTableOperation::AutoIncrement { equals, value } => { write!( f, @@ -3532,6 +3553,20 @@ impl Spanned for DropExtension { } } +/// Table type for ALTER TABLE statements. +/// Used to distinguish between regular tables, Iceberg tables, and Dynamic tables. +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AlterTableType { + /// Iceberg table type + /// + Iceberg, + /// Dynamic table type + /// + Dynamic, +} + /// ALTER TABLE statement #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] @@ -3548,19 +3583,18 @@ pub struct AlterTable { /// For example: `ALTER TABLE table_name ON CLUSTER cluster_name ADD COLUMN c UInt32` /// [ClickHouse](https://clickhouse.com/docs/en/sql-reference/statements/alter/update) pub on_cluster: Option, - /// Snowflake "ICEBERG" clause for Iceberg tables - /// - pub iceberg: bool, + /// Table type: None for regular tables, Some(AlterTableType) for Iceberg or Dynamic tables + pub table_type: Option, /// Token that represents the end of the statement (semicolon or EOF) pub end_token: AttachedToken, } impl fmt::Display for AlterTable { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - if self.iceberg { - write!(f, "ALTER ICEBERG TABLE ")?; - } else { - write!(f, "ALTER TABLE ")?; + match &self.table_type { + Some(AlterTableType::Iceberg) => write!(f, "ALTER ICEBERG TABLE ")?, + Some(AlterTableType::Dynamic) => write!(f, "ALTER DYNAMIC TABLE ")?, + None => write!(f, "ALTER TABLE ")?, } if self.if_exists { diff --git a/src/ast/mod.rs b/src/ast/mod.rs index 4636e4ba2..b32697f60 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -61,7 +61,7 @@ pub use self::dcl::{ pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, AlterSchema, AlterSchemaOperation, AlterTable, AlterTableAlgorithm, AlterTableLock, - AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, + AlterTableOperation, AlterTableType, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, CreateExtension, CreateFunction, diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 34edabd97..80244e691 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -1108,6 +1108,9 @@ impl Spanned for AlterTableOperation { AlterTableOperation::DropClusteringKey => Span::empty(), AlterTableOperation::SuspendRecluster => Span::empty(), AlterTableOperation::ResumeRecluster => Span::empty(), + AlterTableOperation::Refresh => Span::empty(), + AlterTableOperation::Suspend => Span::empty(), + AlterTableOperation::Resume => Span::empty(), AlterTableOperation::Algorithm { .. } => Span::empty(), AlterTableOperation::AutoIncrement { value, .. } => value.span(), AlterTableOperation::Lock { .. } => Span::empty(), diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index 825fd45f0..bb0d4f16b 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -17,6 +17,7 @@ #[cfg(not(feature = "std"))] use crate::alloc::string::ToString; +use crate::ast::helpers::attached_token::AttachedToken; use crate::ast::helpers::key_value_options::{ KeyValueOption, KeyValueOptionKind, KeyValueOptions, KeyValueOptionsDelimiter, }; @@ -26,11 +27,12 @@ use crate::ast::helpers::stmt_data_loading::{ FileStagingCommand, StageLoadSelectItem, StageLoadSelectItemKind, StageParamsObject, }; use crate::ast::{ - CatalogSyncNamespaceMode, ColumnOption, ColumnPolicy, ColumnPolicyProperty, ContactEntry, - CopyIntoSnowflakeKind, CreateTableLikeKind, DollarQuotedString, Ident, IdentityParameters, - IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, - InitializeKind, ObjectName, ObjectNamePart, RefreshModeKind, RowAccessPolicy, ShowObjects, - SqlOption, Statement, StorageSerializationPolicy, TagsColumnOption, Value, WrappedCollection, + AlterTable, AlterTableOperation, AlterTableType, CatalogSyncNamespaceMode, ColumnOption, + ColumnPolicy, ColumnPolicyProperty, ContactEntry, CopyIntoSnowflakeKind, CreateTableLikeKind, + DollarQuotedString, Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, + IdentityPropertyKind, IdentityPropertyOrder, InitializeKind, ObjectName, ObjectNamePart, + RefreshModeKind, RowAccessPolicy, ShowObjects, SqlOption, Statement, + StorageSerializationPolicy, TagsColumnOption, Value, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -214,6 +216,11 @@ impl Dialect for SnowflakeDialect { return Some(parser.parse_begin_exception_end()); } + if parser.parse_keywords(&[Keyword::ALTER, Keyword::DYNAMIC, Keyword::TABLE]) { + // ALTER DYNAMIC TABLE + return Some(parse_alter_dynamic_table(parser)); + } + if parser.parse_keywords(&[Keyword::ALTER, Keyword::SESSION]) { // ALTER SESSION let set = match parser.parse_one_of_keywords(&[Keyword::SET, Keyword::UNSET]) { @@ -604,6 +611,44 @@ fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result +fn parse_alter_dynamic_table(parser: &mut Parser) -> Result { + // Use parse_object_name(true) to support IDENTIFIER() function + let table_name = parser.parse_object_name(true)?; + + // Parse the operation (REFRESH, SUSPEND, or RESUME) + let operation = if parser.parse_keyword(Keyword::REFRESH) { + AlterTableOperation::Refresh + } else if parser.parse_keyword(Keyword::SUSPEND) { + AlterTableOperation::Suspend + } else if parser.parse_keyword(Keyword::RESUME) { + AlterTableOperation::Resume + } else { + return parser.expected( + "REFRESH, SUSPEND, or RESUME after ALTER DYNAMIC TABLE", + parser.peek_token(), + ); + }; + + let end_token = if parser.peek_token_ref().token == Token::SemiColon { + parser.peek_token_ref().clone() + } else { + parser.get_current_token().clone() + }; + + Ok(Statement::AlterTable(AlterTable { + name: table_name, + if_exists: false, + only: false, + operations: vec![operation], + location: None, + on_cluster: None, + table_type: Some(AlterTableType::Dynamic), + end_token: AttachedToken(end_token), + })) +} + /// Parse snowflake alter session. /// fn parse_alter_session(parser: &mut Parser, set: bool) -> Result { diff --git a/src/keywords.rs b/src/keywords.rs index 319c57827..dc4ecd2f9 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -783,6 +783,7 @@ define_keywords!( REF, REFERENCES, REFERENCING, + REFRESH, REFRESH_MODE, REGCLASS, REGEXP, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index f43329be6..026f6249c 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -9462,7 +9462,11 @@ impl<'a> Parser<'a> { operations, location, on_cluster, - iceberg, + table_type: if iceberg { + Some(AlterTableType::Iceberg) + } else { + None + }, end_token: AttachedToken(end_token), } .into()) diff --git a/src/test_utils.rs b/src/test_utils.rs index a8c8afd59..b6100d498 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -347,7 +347,7 @@ pub fn alter_table_op_with_name(stmt: Statement, expected_name: &str) -> AlterTa assert_eq!(alter_table.name.to_string(), expected_name); assert!(!alter_table.if_exists); assert!(!alter_table.only); - assert!(!alter_table.iceberg); + assert_eq!(alter_table.table_type, None); only(alter_table.operations) } _ => panic!("Expected ALTER TABLE statement"), diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index e43df87ab..86c1013c3 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -2746,14 +2746,14 @@ fn parse_alter_table_add_column() { if_exists, only, operations, - iceberg, + table_type, location: _, on_cluster: _, end_token: _, }) => { assert_eq!(name.to_string(), "tab"); assert!(!if_exists); - assert!(!iceberg); + assert_eq!(table_type, None); assert!(!only); assert_eq!( operations, diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 2be5eae8c..f187af1bd 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -4662,3 +4662,11 @@ fn test_drop_constraints() { snowflake().verified_stmt("ALTER TABLE tbl DROP FOREIGN KEY k1 RESTRICT"); snowflake().verified_stmt("ALTER TABLE tbl DROP CONSTRAINT c1 CASCADE"); } + +#[test] +fn test_alter_dynamic_table() { + snowflake().verified_stmt("ALTER DYNAMIC TABLE MY_DYNAMIC_TABLE REFRESH"); + snowflake().verified_stmt("ALTER DYNAMIC TABLE my_database.my_schema.my_dynamic_table REFRESH"); + snowflake().verified_stmt("ALTER DYNAMIC TABLE my_dyn_table SUSPEND"); + snowflake().verified_stmt("ALTER DYNAMIC TABLE my_dyn_table RESUME"); +} From f69407b344a6ca80266289c827bb5c23b202adfc Mon Sep 17 00:00:00 2001 From: etgarperets Date: Tue, 11 Nov 2025 10:10:42 +0200 Subject: [PATCH 286/286] Add support for `INSERT INTO VALUE` (#2085) --- src/ast/query.rs | 8 +++++++- src/ast/spans.rs | 1 + src/parser/mod.rs | 21 ++++++++++++++++----- tests/sqlparser_bigquery.rs | 3 +++ tests/sqlparser_common.rs | 31 ++++++++++++++++++++++++------- tests/sqlparser_databricks.rs | 1 + tests/sqlparser_mysql.rs | 9 +++++++++ tests/sqlparser_postgres.rs | 3 +++ 8 files changed, 64 insertions(+), 13 deletions(-) diff --git a/src/ast/query.rs b/src/ast/query.rs index 599b013ab..33c92614f 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -3135,12 +3135,18 @@ pub struct Values { /// Was there an explicit ROWs keyword (MySQL)? /// pub explicit_row: bool, + // MySql supports both VALUES and VALUE keywords. + // + pub value_keyword: bool, pub rows: Vec>, } impl fmt::Display for Values { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - f.write_str("VALUES")?; + match self.value_keyword { + true => f.write_str("VALUE")?, + false => f.write_str("VALUES")?, + }; let prefix = if self.explicit_row { "ROW" } else { "" }; let mut delim = ""; for row in &self.rows { diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 80244e691..719e261cb 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -223,6 +223,7 @@ impl Spanned for Values { fn span(&self) -> Span { let Values { explicit_row: _, // bool, + value_keyword: _, rows, } = self; diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 026f6249c..9615343c2 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -12533,7 +12533,10 @@ impl<'a> Parser<'a> { SetExpr::Query(subquery) } else if self.parse_keyword(Keyword::VALUES) { let is_mysql = dialect_of!(self is MySqlDialect); - SetExpr::Values(self.parse_values(is_mysql)?) + SetExpr::Values(self.parse_values(is_mysql, false)?) + } else if self.parse_keyword(Keyword::VALUE) { + let is_mysql = dialect_of!(self is MySqlDialect); + SetExpr::Values(self.parse_values(is_mysql, true)?) } else if self.parse_keyword(Keyword::TABLE) { SetExpr::Table(Box::new(self.parse_as_table()?)) } else { @@ -13837,7 +13840,7 @@ impl<'a> Parser<'a> { // Snowflake and Databricks allow syntax like below: // SELECT * FROM VALUES (1, 'a'), (2, 'b') AS t (col1, col2) // where there are no parentheses around the VALUES clause. - let values = SetExpr::Values(self.parse_values(false)?); + let values = SetExpr::Values(self.parse_values(false, false)?); let alias = self.maybe_parse_table_alias()?; Ok(TableFactor::Derived { lateral: false, @@ -16504,7 +16507,11 @@ impl<'a> Parser<'a> { }) } - pub fn parse_values(&mut self, allow_empty: bool) -> Result { + pub fn parse_values( + &mut self, + allow_empty: bool, + value_keyword: bool, + ) -> Result { let mut explicit_row = false; let rows = self.parse_comma_separated(|parser| { @@ -16522,7 +16529,11 @@ impl<'a> Parser<'a> { Ok(exprs) } })?; - Ok(Values { explicit_row, rows }) + Ok(Values { + explicit_row, + rows, + value_keyword, + }) } pub fn parse_start_transaction(&mut self) -> Result { @@ -16937,7 +16948,7 @@ impl<'a> Parser<'a> { MergeInsertKind::Row } else { self.expect_keyword_is(Keyword::VALUES)?; - let values = self.parse_values(is_mysql)?; + let values = self.parse_values(is_mysql, false)?; MergeInsertKind::Values(values) }; MergeAction::Insert(MergeInsertExpr { columns, kind }) diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 03a0ac813..0ef1c4f0c 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -1807,6 +1807,7 @@ fn parse_merge() { let insert_action = MergeAction::Insert(MergeInsertExpr { columns: vec![Ident::new("product"), Ident::new("quantity")], kind: MergeInsertKind::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![Expr::value(number("1")), Expr::value(number("2"))]], }), @@ -1951,6 +1952,7 @@ fn parse_merge() { action: MergeAction::Insert(MergeInsertExpr { columns: vec![Ident::new("a"), Ident::new("b"),], kind: MergeInsertKind::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::value(number("1")), @@ -1965,6 +1967,7 @@ fn parse_merge() { action: MergeAction::Insert(MergeInsertExpr { columns: vec![], kind: MergeInsertKind::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::value(number("1")), diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index 9ea91c642..a235c392c 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -106,19 +106,19 @@ fn parse_insert_values() { let rows2 = vec![row.clone(), row]; let sql = "INSERT customer VALUES (1, 2, 3)"; - check_one(sql, "customer", &[], &rows1); + check_one(sql, "customer", &[], &rows1, false); let sql = "INSERT INTO customer VALUES (1, 2, 3)"; - check_one(sql, "customer", &[], &rows1); + check_one(sql, "customer", &[], &rows1, false); let sql = "INSERT INTO customer VALUES (1, 2, 3), (1, 2, 3)"; - check_one(sql, "customer", &[], &rows2); + check_one(sql, "customer", &[], &rows2, false); let sql = "INSERT INTO public.customer VALUES (1, 2, 3)"; - check_one(sql, "public.customer", &[], &rows1); + check_one(sql, "public.customer", &[], &rows1, false); let sql = "INSERT INTO db.public.customer VALUES (1, 2, 3)"; - check_one(sql, "db.public.customer", &[], &rows1); + check_one(sql, "db.public.customer", &[], &rows1, false); let sql = "INSERT INTO public.customer (id, name, active) VALUES (1, 2, 3)"; check_one( @@ -126,6 +126,16 @@ fn parse_insert_values() { "public.customer", &["id".to_string(), "name".to_string(), "active".to_string()], &rows1, + false, + ); + + let sql = r"INSERT INTO t (id, name, active) VALUE (1, 2, 3)"; + check_one( + sql, + "t", + &["id".to_string(), "name".to_string(), "active".to_string()], + &rows1, + true, ); fn check_one( @@ -133,6 +143,7 @@ fn parse_insert_values() { expected_table_name: &str, expected_columns: &[String], expected_rows: &[Vec], + expected_value_keyword: bool, ) { match verified_stmt(sql) { Statement::Insert(Insert { @@ -147,8 +158,13 @@ fn parse_insert_values() { assert_eq!(column, &Ident::new(expected_columns[index].clone())); } match *source.body { - SetExpr::Values(Values { rows, .. }) => { - assert_eq!(rows.as_slice(), expected_rows) + SetExpr::Values(Values { + rows, + value_keyword, + .. + }) => { + assert_eq!(rows.as_slice(), expected_rows); + assert!(value_keyword == expected_value_keyword); } _ => unreachable!(), } @@ -9908,6 +9924,7 @@ fn parse_merge() { action: MergeAction::Insert(MergeInsertExpr { columns: vec![Ident::new("A"), Ident::new("B"), Ident::new("C")], kind: MergeInsertKind::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::CompoundIdentifier(vec![ diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 92b635339..065e8f9e7 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -157,6 +157,7 @@ fn test_databricks_lambdas() { #[test] fn test_values_clause() { let values = Values { + value_keyword: false, explicit_row: false, rows: vec![ vec![ diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 86c1013c3..b31a5b7c0 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -1885,6 +1885,7 @@ fn parse_simple_insert() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![ vec![ @@ -1950,6 +1951,7 @@ fn parse_ignore_insert() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Value( @@ -1999,6 +2001,7 @@ fn parse_priority_insert() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Value( @@ -2045,6 +2048,7 @@ fn parse_priority_insert() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Value( @@ -2097,6 +2101,7 @@ fn parse_insert_as() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![Expr::Value( (Value::SingleQuotedString("2024-01-01".to_string())).with_empty_span() @@ -2156,6 +2161,7 @@ fn parse_insert_as() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::value(number("1")), @@ -2206,6 +2212,7 @@ fn parse_replace_insert() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Value( @@ -2253,6 +2260,7 @@ fn parse_empty_row_insert() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![], vec![]] })), @@ -2303,6 +2311,7 @@ fn parse_insert_with_on_duplicate_update() { Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Value( diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 9ba0fb978..87cb43edd 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -5169,6 +5169,7 @@ fn test_simple_postgres_insert_with_alias() { source: Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Identifier(Ident::new("DEFAULT")), @@ -5238,6 +5239,7 @@ fn test_simple_postgres_insert_with_alias() { source: Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Identifier(Ident::new("DEFAULT")), @@ -5309,6 +5311,7 @@ fn test_simple_insert_with_quoted_alias() { source: Some(Box::new(Query { with: None, body: Box::new(SetExpr::Values(Values { + value_keyword: false, explicit_row: false, rows: vec![vec![ Expr::Identifier(Ident::new("DEFAULT")),