ClickHouse CREATE TABLE Fixes: add ORDER BY and fix clause ordering (apache#824)

* Fix ClickHouse (add ORDER BY)

* Improve test case

Author: ankrgyl

src/ast/helpers/stmt_create_table.rs

@@ -8,7 +8,7 @@ use serde::{Deserialize, Serialize};
 use sqlparser_derive::{Visit, VisitMut};
 
 use crate::ast::{
-    ColumnDef, FileFormat, HiveDistributionStyle, HiveFormat, ObjectName, OnCommit, Query,
+    ColumnDef, FileFormat, HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, Query,
     SqlOption, Statement, TableConstraint,
 };
 use crate::parser::ParserError;
@@ -69,6 +69,7 @@ pub struct CreateTableBuilder {
     pub collation: Option<String>,
     pub on_commit: Option<OnCommit>,
     pub on_cluster: Option<String>,
+    pub order_by: Option<Vec<Ident>>,
 }
 
 impl CreateTableBuilder {
@@ -98,6 +99,7 @@ impl CreateTableBuilder {
             collation: None,
             on_commit: None,
             on_cluster: None,
+            order_by: None,
         }
     }
     pub fn or_replace(mut self, or_replace: bool) -> Self {
@@ -213,6 +215,11 @@ impl CreateTableBuilder {
         self
     }
 
+    pub fn order_by(mut self, order_by: Option<Vec<Ident>>) -> Self {
+        self.order_by = order_by;
+        self
+    }
+
     pub fn build(self) -> Statement {
         Statement::CreateTable {
             or_replace: self.or_replace,
@@ -239,6 +246,7 @@ impl CreateTableBuilder {
             collation: self.collation,
             on_commit: self.on_commit,
             on_cluster: self.on_cluster,
+            order_by: self.order_by,
         }
     }
 }
@@ -275,6 +283,7 @@ impl TryFrom<Statement> for CreateTableBuilder {
                 collation,
                 on_commit,
                 on_cluster,
+                order_by,
             } => Ok(Self {
                 or_replace,
                 temporary,
@@ -300,6 +309,7 @@ impl TryFrom<Statement> for CreateTableBuilder {
                 collation,
                 on_commit,
                 on_cluster,
+                order_by,
             }),
             _ => Err(ParserError::ParserError(format!(
                 "Expected create table statement, but received: {stmt}"

src/ast/mod.rs

@@ -1254,9 +1254,13 @@ pub enum Statement {
         default_charset: Option<String>,
         collation: Option<String>,
         on_commit: Option<OnCommit>,
-        /// Click house "ON CLUSTER" clause:
+        /// ClickHouse "ON CLUSTER" clause:
         /// <https://clickhouse.com/docs/en/sql-reference/distributed-ddl/>
         on_cluster: Option<String>,
+        /// ClickHouse "ORDER BY " clause. Note that omitted ORDER BY is different
+        /// than empty (represented as ()), the latter meaning "no sorting".
+        /// <https://clickhouse.com/docs/en/sql-reference/statements/create/table/>
+        order_by: Option<Vec<Ident>>,
     },
     /// SQLite's `CREATE VIRTUAL TABLE .. USING <module_name> (<module_args>)`
     CreateVirtualTable {
@@ -2053,6 +2057,7 @@ impl fmt::Display for Statement {
                 collation,
                 on_commit,
                 on_cluster,
+                order_by,
             } => {
                 // We want to allow the following options
                 // Empty column list, allowed by PostgreSQL:
@@ -2196,12 +2201,15 @@ impl fmt::Display for Statement {
                 if !with_options.is_empty() {
                     write!(f, " WITH ({})", display_comma_separated(with_options))?;
                 }
-                if let Some(query) = query {
-                    write!(f, " AS {query}")?;
-                }
                 if let Some(engine) = engine {
                     write!(f, " ENGINE={engine}")?;
                 }
+                if let Some(order_by) = order_by {
+                    write!(f, " ORDER BY ({})", display_comma_separated(order_by))?;
+                }
+                if let Some(query) = query {
+                    write!(f, " AS {query}")?;
+                }
                 if let Some(default_charset) = default_charset {
                     write!(f, " DEFAULT CHARSET={default_charset}")?;
                 }

src/parser.rs

@@ -3340,12 +3340,6 @@ impl<'a> Parser<'a> {
         // PostgreSQL supports `WITH ( options )`, before `AS`
         let with_options = self.parse_options(Keyword::WITH)?;
         let table_properties = self.parse_options(Keyword::TBLPROPERTIES)?;
-        // Parse optional `AS ( query )`
-        let query = if self.parse_keyword(Keyword::AS) {
-            Some(Box::new(self.parse_query()?))
-        } else {
-            None
-        };
 
         let engine = if self.parse_keyword(Keyword::ENGINE) {
             self.expect_token(&Token::Eq)?;
@@ -3358,6 +3352,29 @@ impl<'a> Parser<'a> {
             None
         };
 
+        let order_by = if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) {
+            if self.consume_token(&Token::LParen) {
+                let columns = if self.peek_token() != Token::RParen {
+                    self.parse_comma_separated(Parser::parse_identifier)?
+                } else {
+                    vec![]
+                };
+                self.expect_token(&Token::RParen)?;
+                Some(columns)
+            } else {
+                Some(vec![self.parse_identifier()?])
+            }
+        } else {
+            None
+        };
+
+        // Parse optional `AS ( query )`
+        let query = if self.parse_keyword(Keyword::AS) {
+            Some(Box::new(self.parse_query()?))
+        } else {
+            None
+        };
+
         let default_charset = if self.parse_keywords(&[Keyword::DEFAULT, Keyword::CHARSET]) {
             self.expect_token(&Token::Eq)?;
             let next_token = self.next_token();
@@ -3414,6 +3431,7 @@ impl<'a> Parser<'a> {
             .like(like)
             .clone_clause(clone)
             .engine(engine)
+            .order_by(order_by)
             .default_charset(default_charset)
             .collation(collation)
             .on_commit(on_commit)

tests/sqlparser_clickhouse.rs

@@ -317,6 +317,18 @@ fn parse_similar_to() {
     chk(true);
 }
 
+#[test]
+fn parse_create_table() {
+    clickhouse().verified_stmt(r#"CREATE TABLE "x" ("a" "int") ENGINE=MergeTree ORDER BY ("x")"#);
+    clickhouse().one_statement_parses_to(
+        r#"CREATE TABLE "x" ("a" "int") ENGINE=MergeTree ORDER BY "x""#,
+        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"#,
+    );
+}
+
 fn clickhouse() -> TestedDialects {
     TestedDialects {
         dialects: vec![Box::new(ClickHouseDialect {})],