bpo-47212: Improve error messages for un-parenthesized generator expressions (GH-32302)

Grammar/python.gram

@@ -1073,12 +1073,12 @@ func_type_comment[Token*]:
 invalid_arguments:
     | a=args ',' '*' { RAISE_SYNTAX_ERROR_KNOWN_LOCATION(a, "iterable argument unpacking follows keyword argument unpacking") }
     | a=expression b=for_if_clauses ',' [args | expression for_if_clauses] {
-        RAISE_SYNTAX_ERROR_KNOWN_RANGE(a, PyPegen_last_item(b, comprehension_ty)->target, "Generator expression must be parenthesized") }
+        RAISE_SYNTAX_ERROR_KNOWN_RANGE(a, _PyPegen_get_last_comprehension_item(PyPegen_last_item(b, comprehension_ty)), "Generator expression must be parenthesized") }
     | a=NAME b='=' expression for_if_clauses {
         RAISE_SYNTAX_ERROR_KNOWN_RANGE(a, b, "invalid syntax. Maybe you meant '==' or ':=' instead of '='?")}
     | a=args b=for_if_clauses { _PyPegen_nonparen_genexp_in_call(p, a, b) }
     | args ',' a=expression b=for_if_clauses {
-        RAISE_SYNTAX_ERROR_KNOWN_RANGE(a, asdl_seq_GET(b, b->size-1)->target, "Generator expression must be parenthesized") }
+        RAISE_SYNTAX_ERROR_KNOWN_RANGE(a, _PyPegen_get_last_comprehension_item(PyPegen_last_item(b, comprehension_ty)), "Generator expression must be parenthesized") }
     | a=args ',' args { _PyPegen_arguments_parsing_error(p, a) }
 invalid_kwarg:
     | a[Token*]=('True'|'False'|'None') b='=' {
@@ -1257,7 +1257,7 @@ invalid_finally_stmt:
 invalid_except_stmt_indent:
     | a='except' expression ['as' NAME ] ':' NEWLINE !INDENT {
         RAISE_INDENTATION_ERROR("expected an indented block after 'except' statement on line %d", a->lineno) }
-    | a='except' ':' NEWLINE !INDENT { RAISE_SYNTAX_ERROR("expected an indented block after except statement on line %d", a->lineno) }
+    | a='except' ':' NEWLINE !INDENT { RAISE_INDENTATION_ERROR("expected an indented block after 'except' statement on line %d", a->lineno) }
 invalid_except_star_stmt_indent:
     | a='except' '*' expression ['as' NAME ] ':' NEWLINE !INDENT {
         RAISE_INDENTATION_ERROR("expected an indented block after 'except*' statement on line %d", a->lineno) }

Lib/test/test_exceptions.py

@@ -198,12 +198,17 @@ class ExceptionTests(unittest.TestCase):
         s = '''if True:\n        print()\n\texec "mixed tabs and spaces"'''
         ckmsg(s, "inconsistent use of tabs and spaces in indentation", TabError)
 
-    def check(self, src, lineno, offset, encoding='utf-8'):
+    def check(self, src, lineno, offset, end_lineno=None, end_offset=None, encoding='utf-8'):
         with self.subTest(source=src, lineno=lineno, offset=offset):
             with self.assertRaises(SyntaxError) as cm:
                 compile(src, '<fragment>', 'exec')
             self.assertEqual(cm.exception.lineno, lineno)
             self.assertEqual(cm.exception.offset, offset)
+            if end_lineno is not None:
+                self.assertEqual(cm.exception.end_lineno, end_lineno)
+            if end_offset is not None:
+                self.assertEqual(cm.exception.end_offset, end_offset)
+
             if cm.exception.text is not None:
                 if not isinstance(src, str):
                     src = src.decode(encoding, 'replace')
@@ -235,6 +240,10 @@ class ExceptionTests(unittest.TestCase):
         check('match ...:\n    case {**rest, "key": value}:\n        ...', 2, 19)
         check("[a b c d e f]", 1, 2)
         check("for x yfff:", 1, 7)
+        check("f(a for a in b, c)", 1, 3, 1, 15)
+        check("f(a for a in b if a, c)", 1, 3, 1, 20)
+        check("f(a, b for b in c)", 1, 6, 1, 18)
+        check("f(a, b for b in c, d)", 1, 6, 1, 18)
 
         # Errors thrown by compile.c
         check('class foo:return 1', 1, 11)

Lib/test/test_syntax.py

@@ -1307,6 +1307,13 @@ Specialized indentation errors:
    Traceback (most recent call last):
    IndentationError: expected an indented block after 'try' statement on line 1
 
+   >>> try:
+   ...     something()
+   ... except:
+   ... pass
+   Traceback (most recent call last):
+   IndentationError: expected an indented block after 'except' statement on line 3
+
    >>> try:
    ...     something()
    ... except A:

Misc/NEWS.d/next/Core and Builtins/2022-04-05-11-29-21.bpo-47212.leF4pz.rst

@@ -0,0 +1,3 @@
+Raise :exc:`IndentationError` instead of :exc:`SyntaxError` for a bare
+``except`` with no following indent. Improve :exc:`SyntaxError` locations for
+an un-parenthesized generator used as arguments. Patch by Matthieu Dartiailh.

Parser/action_helpers.c

@@ -1145,7 +1145,7 @@ _PyPegen_get_expr_name(expr_ty e)
     }
 }
 
-static inline expr_ty
+expr_ty
 _PyPegen_get_last_comprehension_item(comprehension_ty comprehension) {
     if (comprehension->ifs == NULL || asdl_seq_LEN(comprehension->ifs) == 0) {
         return comprehension->iter;

Parser/parser.c

@@ -18968,7 +18968,7 @@ invalid_arguments_rule(Parser *p)
         )
         {
             D(fprintf(stderr, "%*c+ invalid_arguments[%d-%d]: %s succeeded!\n", p->level, ' ', _mark, p->mark, "expression for_if_clauses ',' [args | expression for_if_clauses]"));
-            _res = RAISE_SYNTAX_ERROR_KNOWN_RANGE ( a , PyPegen_last_item ( b , comprehension_ty ) -> target , "Generator expression must be parenthesized" );
+            _res = RAISE_SYNTAX_ERROR_KNOWN_RANGE ( a , _PyPegen_get_last_comprehension_item ( PyPegen_last_item ( b , comprehension_ty ) ) , "Generator expression must be parenthesized" );
             if (_res == NULL && PyErr_Occurred()) {
                 p->error_indicator = 1;
                 p->level--;
@@ -19061,7 +19061,7 @@ invalid_arguments_rule(Parser *p)
         )
         {
             D(fprintf(stderr, "%*c+ invalid_arguments[%d-%d]: %s succeeded!\n", p->level, ' ', _mark, p->mark, "args ',' expression for_if_clauses"));
-            _res = RAISE_SYNTAX_ERROR_KNOWN_RANGE ( a , asdl_seq_GET ( b , b -> size - 1 ) -> target , "Generator expression must be parenthesized" );
+            _res = RAISE_SYNTAX_ERROR_KNOWN_RANGE ( a , _PyPegen_get_last_comprehension_item ( PyPegen_last_item ( b , comprehension_ty ) ) , "Generator expression must be parenthesized" );
             if (_res == NULL && PyErr_Occurred()) {
                 p->error_indicator = 1;
                 p->level--;
@@ -22190,7 +22190,7 @@ invalid_except_stmt_indent_rule(Parser *p)
         )
         {
             D(fprintf(stderr, "%*c+ invalid_except_stmt_indent[%d-%d]: %s succeeded!\n", p->level, ' ', _mark, p->mark, "'except' ':' NEWLINE !INDENT"));
-            _res = RAISE_SYNTAX_ERROR ( "expected an indented block after except statement on line %d" , a -> lineno );
+            _res = RAISE_INDENTATION_ERROR ( "expected an indented block after 'except' statement on line %d" , a -> lineno );
             if (_res == NULL && PyErr_Occurred()) {
                 p->error_indicator = 1;
                 p->level--;

Parser/pegen.h

@@ -324,6 +324,7 @@ int _PyPegen_check_barry_as_flufl(Parser *, Token *);
 int _PyPegen_check_legacy_stmt(Parser *p, expr_ty t);
 mod_ty _PyPegen_make_module(Parser *, asdl_stmt_seq *);
 void *_PyPegen_arguments_parsing_error(Parser *, expr_ty);
+expr_ty _PyPegen_get_last_comprehension_item(comprehension_ty comprehension);
 void *_PyPegen_nonparen_genexp_in_call(Parser *p, expr_ty args, asdl_comprehension_seq *comprehensions);
 
 // Parser API