Improve JsonLexContext's freeability

Previously, the JSON code didn't have to worry too much about freeing
JsonLexContext, because it was never too long-lived.  With new features
being added for SQL/JSON this is no longer the case.  Add a routine
that knows how to free this struct and apply that to a few places, to
prevent this from becoming problematic.

At the same time, we change the API of makeJsonLexContextCstringLen to
make it receive a pointer to JsonLexContext for callers that want it to
be stack-allocated; it can also be passed as NULL to get the original
behavior of a palloc'ed one.

This also causes an ABI break due to the addition of flags to
JsonLexContext, so we can't easily backpatch it.  AFAICS that's not much
of a problem; apparently some leaks might exist in JSON usage of
text-search, for example via json_to_tsvector, but I haven't seen any
complaints about that.

Per Coverity complaint about datum_to_jsonb_internal().

Discussion: https://postgr.es/m/20230808174110.oq3iymllsv6amkih@alvherre.pgsql

Author: alvherre

src/backend/utils/adt/json.c

@@ -106,11 +106,11 @@ json_in(PG_FUNCTION_ARGS)
 {
 	char	   *json = PG_GETARG_CSTRING(0);
 	text	   *result = cstring_to_text(json);
-	JsonLexContext *lex;
+	JsonLexContext lex;
 
 	/* validate it */
-	lex = makeJsonLexContext(result, false);
-	if (!pg_parse_json_or_errsave(lex, &nullSemAction, fcinfo->context))
+	makeJsonLexContext(&lex, result, false);
+	if (!pg_parse_json_or_errsave(&lex, &nullSemAction, fcinfo->context))
 		PG_RETURN_NULL();
 
 	/* Internal representation is the same as text */
@@ -152,13 +152,14 @@ json_recv(PG_FUNCTION_ARGS)
 	StringInfo	buf = (StringInfo) PG_GETARG_POINTER(0);
 	char	   *str;
 	int			nbytes;
-	JsonLexContext *lex;
+	JsonLexContext lex;
 
 	str = pq_getmsgtext(buf, buf->len - buf->cursor, &nbytes);
 
 	/* Validate it. */
-	lex = makeJsonLexContextCstringLen(str, nbytes, GetDatabaseEncoding(), false);
-	pg_parse_json_or_ereport(lex, &nullSemAction);
+	makeJsonLexContextCstringLen(&lex, str, nbytes, GetDatabaseEncoding(),
+								 false);
+	pg_parse_json_or_ereport(&lex, &nullSemAction);
 
 	PG_RETURN_TEXT_P(cstring_to_text_with_len(str, nbytes));
 }
@@ -1625,14 +1626,16 @@ json_unique_object_field_start(void *_state, char *field, bool isnull)
 bool
 json_validate(text *json, bool check_unique_keys, bool throw_error)
 {
-	JsonLexContext *lex = makeJsonLexContext(json, check_unique_keys);
+	JsonLexContext lex;
 	JsonSemAction uniqueSemAction = {0};
 	JsonUniqueParsingState state;
 	JsonParseErrorType result;
 
+	makeJsonLexContext(&lex, json, check_unique_keys);
+
 	if (check_unique_keys)
 	{
-		state.lex = lex;
+		state.lex = &lex;
 		state.stack = NULL;
 		state.id_counter = 0;
 		state.unique = true;
@@ -1644,12 +1647,12 @@ json_validate(text *json, bool check_unique_keys, bool throw_error)
 		uniqueSemAction.object_end = json_unique_object_end;
 	}
 
-	result = pg_parse_json(lex, check_unique_keys ? &uniqueSemAction : &nullSemAction);
+	result = pg_parse_json(&lex, check_unique_keys ? &uniqueSemAction : &nullSemAction);
 
 	if (result != JSON_SUCCESS)
 	{
 		if (throw_error)
-			json_errsave_error(result, lex, NULL);
+			json_errsave_error(result, &lex, NULL);
 
 		return false;			/* invalid json */
 	}
@@ -1664,6 +1667,9 @@ json_validate(text *json, bool check_unique_keys, bool throw_error)
 		return false;			/* not unique keys */
 	}
 
+	if (check_unique_keys)
+		freeJsonLexContext(&lex);
+
 	return true;				/* ok */
 }
 
@@ -1683,18 +1689,17 @@ Datum
 json_typeof(PG_FUNCTION_ARGS)
 {
 	text	   *json = PG_GETARG_TEXT_PP(0);
-	JsonLexContext *lex = makeJsonLexContext(json, false);
+	JsonLexContext lex;
 	char	   *type;
-	JsonTokenType tok;
 	JsonParseErrorType result;
 
 	/* Lex exactly one token from the input and check its type. */
-	result = json_lex(lex);
+	makeJsonLexContext(&lex, json, false);
+	result = json_lex(&lex);
 	if (result != JSON_SUCCESS)
-		json_errsave_error(result, lex, NULL);
-	tok = lex->token_type;
+		json_errsave_error(result, &lex, NULL);
 
-	switch (tok)
+	switch (lex.token_type)
 	{
 		case JSON_TOKEN_OBJECT_START:
 			type = "object";
@@ -1716,7 +1721,7 @@ json_typeof(PG_FUNCTION_ARGS)
 			type = "null";
 			break;
 		default:
-			elog(ERROR, "unexpected json token: %d", tok);
+			elog(ERROR, "unexpected json token: %d", lex.token_type);
 	}
 
 	PG_RETURN_TEXT_P(cstring_to_text(type));

src/backend/utils/adt/jsonb.c

@@ -252,13 +252,13 @@ jsonb_typeof(PG_FUNCTION_ARGS)
 static inline Datum
 jsonb_from_cstring(char *json, int len, bool unique_keys, Node *escontext)
 {
-	JsonLexContext *lex;
+	JsonLexContext lex;
 	JsonbInState state;
 	JsonSemAction sem;
 
 	memset(&state, 0, sizeof(state));
 	memset(&sem, 0, sizeof(sem));
-	lex = makeJsonLexContextCstringLen(json, len, GetDatabaseEncoding(), true);
+	makeJsonLexContextCstringLen(&lex, json, len, GetDatabaseEncoding(), true);
 
 	state.unique_keys = unique_keys;
 	state.escontext = escontext;
@@ -271,7 +271,7 @@ jsonb_from_cstring(char *json, int len, bool unique_keys, Node *escontext)
 	sem.scalar = jsonb_in_scalar;
 	sem.object_field_start = jsonb_in_object_field_start;
 
-	if (!pg_parse_json_or_errsave(lex, &sem, escontext))
+	if (!pg_parse_json_or_errsave(&lex, &sem, escontext))
 		return (Datum) 0;
 
 	/* after parsing, the item member has the composed jsonb structure */
@@ -755,11 +755,11 @@ datum_to_jsonb_internal(Datum val, bool is_null, JsonbInState *result,
 			case JSONTYPE_JSON:
 				{
 					/* parse the json right into the existing result object */
-					JsonLexContext *lex;
+					JsonLexContext lex;
 					JsonSemAction sem;
 					text	   *json = DatumGetTextPP(val);
 
-					lex = makeJsonLexContext(json, true);
+					makeJsonLexContext(&lex, json, true);
 
 					memset(&sem, 0, sizeof(sem));
 
@@ -772,7 +772,8 @@ datum_to_jsonb_internal(Datum val, bool is_null, JsonbInState *result,
 					sem.scalar = jsonb_in_scalar;
 					sem.object_field_start = jsonb_in_object_field_start;
 
-					pg_parse_json_or_ereport(lex, &sem);
+					pg_parse_json_or_ereport(&lex, &sem);
+					freeJsonLexContext(&lex);
 				}
 				break;
 			case JSONTYPE_JSONB: