Add support for dumping raw parse tree with debug_print_raw_parse

This patch introduces a small change to log the raw parse tree in the
same way we currently log the parse tree, rewritten tree, and plan tree.
While tracing some queries, I found that being able to inspect the raw
parse tree is also helpful for understanding query transformation.

Although the raw parse tree can be inspected via a debugger, having it
logged simplifies the workflow for those interested in this stage of
query processing, without requiring a debugging session.

To avoid unnecessary log noise for users not interested in this detail,
a new GUC option, "debug_print_raw_parse", has been added.

When starting the PostgreSQL process with "-d N", and N is 3 or higher,
debug_print_raw_parse is enabled automatically, alongside
debug_print_parse.

Author: Chao Li <[email protected]>
Discussion: https://www.postgresql.org/message-id/CAEoWx2mcO0Gpo4vd8kPMAFWeJLSp0MeUUnaLdE1x0tSVd-VzUw%40mail.gmail.com

Author: evanchaoli

doc/src/sgml/config.sgml

@@ -7383,6 +7383,11 @@ local0.*    /var/log/postgresql
      </varlistentry>
 
      <varlistentry id="guc-debug-print-parse">
+      <term><varname>debug_print_raw_parse</varname> (<type>boolean</type>)
+      <indexterm>
+       <primary><varname>debug_print_raw_parse</varname> configuration parameter</primary>
+      </indexterm>
+      </term>
       <term><varname>debug_print_parse</varname> (<type>boolean</type>)
       <indexterm>
        <primary><varname>debug_print_parse</varname> configuration parameter</primary>
@@ -7401,8 +7406,8 @@ local0.*    /var/log/postgresql
       <listitem>
        <para>
         These parameters enable various debugging output to be emitted.
-        When set, they print the resulting parse tree, the query rewriter
-        output, or the execution plan for each executed query.
+        When set, they print the resulting raw parse tree, the parse tree, the query
+        rewriter output, or the execution plan for each executed query.
         These messages are emitted at <literal>LOG</literal> message level, so by
         default they will appear in the server log but will not be sent to the
         client.  You can change that by adjusting
@@ -7422,7 +7427,8 @@ local0.*    /var/log/postgresql
       <listitem>
        <para>
         When set, <varname>debug_pretty_print</varname> indents the messages
-        produced by <varname>debug_print_parse</varname>,
+        produced by <varname>debug_print_raw_parse</varname>,
+        <varname>debug_print_parse</varname>,
         <varname>debug_print_rewritten</varname>, or
         <varname>debug_print_plan</varname>.  This results in more readable
         but much longer output than the <quote>compact</quote> format used when

doc/src/sgml/rules.sgml

@@ -60,6 +60,7 @@
     <acronym>SQL</acronym> statement where the single parts that it is
     built from are stored separately. These query trees can be shown
     in the server log if you set the configuration parameters
+    <varname>debug_print_raw_parse</varname>,
     <varname>debug_print_parse</varname>,
     <varname>debug_print_rewritten</varname>, or
     <varname>debug_print_plan</varname>.  The rule actions are also

src/backend/tcop/postgres.c

@@ -649,6 +649,10 @@ pg_parse_query(const char *query_string)
 
 	TRACE_POSTGRESQL_QUERY_PARSE_DONE(query_string);
 
+	if (Debug_print_raw_parse)
+		elog_node_display(LOG, "raw parse tree", raw_parsetree_list,
+						  Debug_pretty_print);
+
 	return raw_parsetree_list;
 }
 
@@ -3697,7 +3701,10 @@ set_debug_options(int debug_flag, GucContext context, GucSource source)
 	if (debug_flag >= 2)
 		SetConfigOption("log_statement", "all", context, source);
 	if (debug_flag >= 3)
+	{
+		SetConfigOption("debug_print_raw_parse", "true", context, source);
 		SetConfigOption("debug_print_parse", "true", context, source);
+	}
 	if (debug_flag >= 4)
 		SetConfigOption("debug_print_plan", "true", context, source);
 	if (debug_flag >= 5)

src/backend/utils/misc/guc_parameters.dat

@@ -414,6 +414,12 @@
   ifdef => 'DEBUG_NODE_TESTS_ENABLED',
 },
 
+{ name => 'debug_print_raw_parse', type => 'bool', context => 'PGC_USERSET', group => 'LOGGING_WHAT',
+  short_desc => 'Logs each query\'s raw parse tree.',
+  variable => 'Debug_print_raw_parse',
+  boot_val => 'false',
+},
+
 { name => 'debug_print_parse', type => 'bool', context => 'PGC_USERSET', group => 'LOGGING_WHAT',
   short_desc => 'Logs each query\'s parse tree.',
   variable => 'Debug_print_parse',

src/backend/utils/misc/guc_tables.c

@@ -507,6 +507,7 @@ bool		AllowAlterSystem = true;
 bool		log_duration = false;
 bool		Debug_print_plan = false;
 bool		Debug_print_parse = false;
+bool		Debug_print_raw_parse = false;
 bool		Debug_print_rewritten = false;
 bool		Debug_pretty_print = true;
 

src/backend/utils/misc/postgresql.conf.sample

@@ -581,6 +581,7 @@
 
 # - What to Log -
 
+#debug_print_raw_parse = off
 #debug_print_parse = off
 #debug_print_rewritten = off
 #debug_print_plan = off

src/include/utils/guc.h

@@ -247,6 +247,7 @@ typedef enum
 /* GUC vars that are actually defined in guc_tables.c, rather than elsewhere */
 extern PGDLLIMPORT bool Debug_print_plan;
 extern PGDLLIMPORT bool Debug_print_parse;
+extern PGDLLIMPORT bool Debug_print_raw_parse;
 extern PGDLLIMPORT bool Debug_print_rewritten;
 extern PGDLLIMPORT bool Debug_pretty_print;