Add "**" as a new token in scanners

l-wang · l-wang · commit 21680e948af9 · 2025-03-12T15:13:50.000-05:00
DOUBLE_ASTERISK is now recognized but not yet implemented.

This change improves error message readability, as "**" is supported
in jsonpath_scan.l as ANY_P.  However, the SQL standard for simplified
JSON accessors does not specify "**", so it is not supported in that
context.
diff --git a/src/backend/parser/gram.y b/src/backend/parser/gram.y
@@ -687,7 +687,7 @@ static Node *makeRecursiveViewSelect(char *relname, List *aliases, Node *query);
  */
 %token <str>	IDENT UIDENT FCONST SCONST USCONST BCONST XCONST Op
 %token <ival>	ICONST PARAM
-%token			TYPECAST DOT_DOT COLON_EQUALS EQUALS_GREATER
+%token			TYPECAST DOT_DOT COLON_EQUALS EQUALS_GREATER DOUBLE_ASTERISK
 %token			LESS_EQUALS GREATER_EQUALS NOT_EQUALS
 
 /*
@@ -16941,6 +16941,13 @@ indirection_el:
 				{
 					$$ = (Node *) makeNode(A_Star);
 				}
+			| '.' DOUBLE_ASTERISK
+				{
+			ereport(ERROR,
+					(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+					errmsg("arbitrary depth wild card in simple json accessor not supported"),
+					parser_errposition(@2)));
+				}
 			| '[' a_expr ']'
 				{
 					A_Indices *ai = makeNode(A_Indices);
diff --git a/src/backend/parser/scan.l b/src/backend/parser/scan.l
@@ -338,6 +338,7 @@ identifier		{ident_start}{ident_cont}*
 typecast		"::"
 dot_dot			\.\.
 colon_equals	":="
+double_asterisk	"**"
 
 /*
  * These operator-like tokens (unlike the above ones) also match the {operator}
@@ -851,6 +852,11 @@ other			.
 					return COLON_EQUALS;
 				}
 
+{double_asterisk}	{
+						SET_YYLLOC();
+						return DOUBLE_ASTERISK;
+					}
+
 {equals_greater} {
 					SET_YYLLOC();
 					return EQUALS_GREATER;
diff --git a/src/include/parser/scanner.h b/src/include/parser/scanner.h
@@ -50,7 +50,7 @@ typedef union core_YYSTYPE
  * the ASCII characters plus these:
  *	%token <str>	IDENT UIDENT FCONST SCONST USCONST BCONST XCONST Op
  *	%token <ival>	ICONST PARAM
- *	%token			TYPECAST DOT_DOT COLON_EQUALS EQUALS_GREATER
+ *	%token			TYPECAST DOT_DOT COLON_EQUALS EQUALS_GREATER DOUBLE_ASTERISK
  *	%token			LESS_EQUALS GREATER_EQUALS NOT_EQUALS
  * The above token definitions *must* be the first ones declared in any
  * bison parser built atop this scanner, so that they will have consistent
diff --git a/src/interfaces/ecpg/preproc/pgc.l b/src/interfaces/ecpg/preproc/pgc.l
@@ -321,6 +321,7 @@ array			({ident_cont}|{whitespace}|[\[\]\+\-\*\%\/\(\)\>\.])*
 typecast		"::"
 dot_dot			\.\.
 colon_equals	":="
+double_asterisk	"**"
 
 /*
  * These operator-like tokens (unlike the above ones) also match the {operator}
@@ -832,6 +833,10 @@ cppline			{space}*#([^i][A-Za-z]*|{if}|{ifdef}|{ifndef}|{import})((\/\*[^*/]*\*+
 					return COLON_EQUALS;
 				}
 
+{double_asterisk}	{
+						return DOUBLE_ASTERISK;
+					}
+
 {equals_greater} {
 					return EQUALS_GREATER;
 				}
diff --git a/src/pl/plpgsql/src/pl_gram.y b/src/pl/plpgsql/src/pl_gram.y
@@ -245,7 +245,7 @@ static	void			check_raise_parameters(PLpgSQL_stmt_raise *stmt);
  */
 %token <str>	IDENT UIDENT FCONST SCONST USCONST BCONST XCONST Op
 %token <ival>	ICONST PARAM
-%token			TYPECAST DOT_DOT COLON_EQUALS EQUALS_GREATER
+%token			TYPECAST DOT_DOT COLON_EQUALS EQUALS_GREATER DOUBLE_ASTERISK
 %token			LESS_EQUALS GREATER_EQUALS NOT_EQUALS
 
 /*
diff --git a/src/test/regress/expected/jsonb.out b/src/test/regress/expected/jsonb.out
@@ -6131,7 +6131,7 @@ SELECT (jb).a.b.c.* FROM test_jsonb_dot_notation;
 (1 row)
 
 SELECT (jb).a.**.x FROM test_jsonb_dot_notation; -- not supported
-ERROR:  syntax error at or near "**"
+ERROR:  arbitrary depth wild card in simple json accessor not supported yet
 LINE 1: SELECT (jb).a.**.x FROM test_jsonb_dot_notation;
                       ^
 EXPLAIN (VERBOSE, COSTS OFF) SELECT (jb).* FROM test_jsonb_dot_notation;

Original file line number	Diff line number	Diff line change
`@@ -6131,7 +6131,7 @@ SELECT (jb).a.b.c.* FROM test_jsonb_dot_notation;`
`6131`	`6131`	`(1 row)`
`6132`	`6132`
`6133`	`6133`	`SELECT (jb).a.**.x FROM test_jsonb_dot_notation; -- not supported`
`6134`		`-ERROR: syntax error at or near "**"`
	`6134`	`+ERROR: arbitrary depth wild card in simple json accessor not supported yet`
`6135`	`6135`	`LINE 1: SELECT (jb).a.**.x FROM test_jsonb_dot_notation;`
`6136`	`6136`	`^`
`6137`	`6137`	`EXPLAIN (VERBOSE, COSTS OFF) SELECT (jb).* FROM test_jsonb_dot_notation;`