plpgsql: pure parser and reentrant scanner

The plpgsql scanner is a wrapper around the core scanner, which
already uses the flex %option reentrant.  This patch only pushes up a
few levels the place where the scanner handle is allocated.  Before,
it was allocated in pl_scanner.c in a global variable, so to the
outside the scanner was not reentrant.  Now, it is allocated in
pl_comp.c and is passed as an argument to yyparse(), similar to how it
is handled in other reentrant scanners.

Also use flex yyextra to handle context information, instead of global
variables.  Again, this uses the existing yyextra support in the core
scanner.  This complements the other changes to make the scanner
reentrant.

The bison option %pure-parser is used to make the generated parser
pure.  This happens in the usual way, since plpgsql has its own bison
parser definition.

Reviewed-by: Heikki Linnakangas <hlinnaka@iki.fi>
Discussion: https://www.postgresql.org/message-id/flat/eb6faeac-2a8a-4b69-9189-c33c520e5b7b@eisentraut.org

Author: petere

src/pl/plpgsql/src/nls.mk

@@ -6,5 +6,5 @@ GETTEXT_FILES    = pl_comp.c \
                    pl_funcs.c \
                    pl_handler.c \
                    pl_scanner.c
-GETTEXT_TRIGGERS = $(BACKEND_COMMON_GETTEXT_TRIGGERS) yyerror plpgsql_yyerror
+GETTEXT_TRIGGERS = $(BACKEND_COMMON_GETTEXT_TRIGGERS) yyerror:3 plpgsql_yyerror:3
 GETTEXT_FLAGS    = $(BACKEND_COMMON_GETTEXT_FLAGS)

src/pl/plpgsql/src/pl_comp.c

@@ -237,6 +237,12 @@ plpgsql_compile(FunctionCallInfo fcinfo, bool forValidator)
 	return function;
 }
 
+struct compile_error_callback_arg
+{
+	const char *proc_source;
+	yyscan_t	yyscanner;
+};
+
 /*
  * This is the slow part of plpgsql_compile().
  *
@@ -269,13 +275,15 @@ do_compile(FunctionCallInfo fcinfo,
 	Form_pg_proc procStruct = (Form_pg_proc) GETSTRUCT(procTup);
 	bool		is_dml_trigger = CALLED_AS_TRIGGER(fcinfo);
 	bool		is_event_trigger = CALLED_AS_EVENT_TRIGGER(fcinfo);
+	yyscan_t	scanner;
 	Datum		prosrcdatum;
 	char	   *proc_source;
 	HeapTuple	typeTup;
 	Form_pg_type typeStruct;
 	PLpgSQL_variable *var;
 	PLpgSQL_rec *rec;
 	int			i;
+	struct compile_error_callback_arg cbarg;
 	ErrorContextCallback plerrcontext;
 	int			parse_rc;
 	Oid			rettypeid;
@@ -290,21 +298,21 @@ do_compile(FunctionCallInfo fcinfo,
 	MemoryContext func_cxt;
 
 	/*
-	 * Setup the scanner input and error info.  We assume that this function
-	 * cannot be invoked recursively, so there's no need to save and restore
-	 * the static variables used here.
+	 * Setup the scanner input and error info.
 	 */
 	prosrcdatum = SysCacheGetAttrNotNull(PROCOID, procTup, Anum_pg_proc_prosrc);
 	proc_source = TextDatumGetCString(prosrcdatum);
-	plpgsql_scanner_init(proc_source);
+	scanner = plpgsql_scanner_init(proc_source);
 
 	plpgsql_error_funcname = pstrdup(NameStr(procStruct->proname));
 
 	/*
 	 * Setup error traceback support for ereport()
 	 */
+	cbarg.proc_source = forValidator ? proc_source : NULL;
+	cbarg.yyscanner = scanner;
 	plerrcontext.callback = plpgsql_compile_error_callback;
-	plerrcontext.arg = forValidator ? proc_source : NULL;
+	plerrcontext.arg = &cbarg;
 	plerrcontext.previous = error_context_stack;
 	error_context_stack = &plerrcontext;
 
@@ -779,12 +787,12 @@ do_compile(FunctionCallInfo fcinfo,
 	/*
 	 * Now parse the function's text
 	 */
-	parse_rc = plpgsql_yyparse();
+	parse_rc = plpgsql_yyparse(scanner);
 	if (parse_rc != 0)
 		elog(ERROR, "plpgsql parser returned %d", parse_rc);
 	function->action = plpgsql_parse_result;
 
-	plpgsql_scanner_finish();
+	plpgsql_scanner_finish(scanner);
 	pfree(proc_source);
 
 	/*
@@ -841,27 +849,29 @@ do_compile(FunctionCallInfo fcinfo,
 PLpgSQL_function *
 plpgsql_compile_inline(char *proc_source)
 {
+	yyscan_t	scanner;
 	char	   *func_name = "inline_code_block";
 	PLpgSQL_function *function;
+	struct compile_error_callback_arg cbarg;
 	ErrorContextCallback plerrcontext;
 	PLpgSQL_variable *var;
 	int			parse_rc;
 	MemoryContext func_cxt;
 
 	/*
-	 * Setup the scanner input and error info.  We assume that this function
-	 * cannot be invoked recursively, so there's no need to save and restore
-	 * the static variables used here.
+	 * Setup the scanner input and error info.
 	 */
-	plpgsql_scanner_init(proc_source);
+	scanner = plpgsql_scanner_init(proc_source);
 
 	plpgsql_error_funcname = func_name;
 
 	/*
 	 * Setup error traceback support for ereport()
 	 */
+	cbarg.proc_source = proc_source;
+	cbarg.yyscanner = scanner;
 	plerrcontext.callback = plpgsql_compile_error_callback;
-	plerrcontext.arg = proc_source;
+	plerrcontext.arg = &cbarg;
 	plerrcontext.previous = error_context_stack;
 	error_context_stack = &plerrcontext;
 
@@ -935,12 +945,12 @@ plpgsql_compile_inline(char *proc_source)
 	/*
 	 * Now parse the function's text
 	 */
-	parse_rc = plpgsql_yyparse();
+	parse_rc = plpgsql_yyparse(scanner);
 	if (parse_rc != 0)
 		elog(ERROR, "plpgsql parser returned %d", parse_rc);
 	function->action = plpgsql_parse_result;
 
-	plpgsql_scanner_finish();
+	plpgsql_scanner_finish(scanner);
 
 	/*
 	 * If it returns VOID (always true at the moment), we allow control to
@@ -978,13 +988,16 @@ plpgsql_compile_inline(char *proc_source)
 static void
 plpgsql_compile_error_callback(void *arg)
 {
-	if (arg)
+	struct compile_error_callback_arg *cbarg = (struct compile_error_callback_arg *) arg;
+	yyscan_t	yyscanner = cbarg->yyscanner;
+
+	if (cbarg->proc_source)
 	{
 		/*
 		 * Try to convert syntax error position to reference text of original
 		 * CREATE FUNCTION or DO command.
 		 */
-		if (function_parse_error_transpose((const char *) arg))
+		if (function_parse_error_transpose(cbarg->proc_source))
 			return;
 
 		/*
@@ -995,7 +1008,7 @@ plpgsql_compile_error_callback(void *arg)
 
 	if (plpgsql_error_funcname)
 		errcontext("compilation of PL/pgSQL function \"%s\" near line %d",
-				   plpgsql_error_funcname, plpgsql_latest_lineno());
+				   plpgsql_error_funcname, plpgsql_latest_lineno(yyscanner));
 }