bpo-45408: Don't override previous tokenizer errors in the second parser pass (GH-28812)

This commit is contained in:
Pablo Galindo Salgado 2021-10-07 22:33:05 +01:00 committed by GitHub
parent 6811fdaec8
commit 0219017df7
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 15 additions and 2 deletions

View File

@ -1075,6 +1075,14 @@ Module(
with self.assertRaisesRegex(ValueError, msg): with self.assertRaisesRegex(ValueError, msg):
ast.literal_eval(node) ast.literal_eval(node)
def test_literal_eval_syntax_errors(self):
msg = "unexpected character after line continuation character"
with self.assertRaisesRegex(SyntaxError, msg):
ast.literal_eval(r'''
\
(\
\ ''')
def test_bad_integer(self): def test_bad_integer(self):
# issue13436: Bad error message with invalid numeric values # issue13436: Bad error message with invalid numeric values
body = [ast.ImportFrom(module='time', body = [ast.ImportFrom(module='time',

View File

@ -223,7 +223,7 @@ class ExceptionTests(unittest.TestCase):
check('x = "a', 1, 5) check('x = "a', 1, 5)
check('lambda x: x = 2', 1, 1) check('lambda x: x = 2', 1, 1)
check('f{a + b + c}', 1, 1) check('f{a + b + c}', 1, 1)
check('[file for str(file) in []\n])', 2, 2) check('[file for str(file) in []\n])', 1, 11)
check('a = « hello » « world »', 1, 5) check('a = « hello » « world »', 1, 5)
check('[\nfile\nfor str(file)\nin\n[]\n]', 3, 5) check('[\nfile\nfor str(file)\nin\n[]\n]', 3, 5)
check('[file for\n str(file) in []]', 2, 2) check('[file for\n str(file) in []]', 2, 2)

View File

@ -0,0 +1,2 @@
Fix a crash in the parser when reporting tokenizer errors that occur at the
same time unclosed parentheses are detected. Patch by Pablo Galindo.

View File

@ -1342,13 +1342,16 @@ _PyPegen_run_parser(Parser *p)
{ {
void *res = _PyPegen_parse(p); void *res = _PyPegen_parse(p);
if (res == NULL) { if (res == NULL) {
if (PyErr_Occurred() && !PyErr_ExceptionMatches(PyExc_SyntaxError)) {
return NULL;
}
Token *last_token = p->tokens[p->fill - 1]; Token *last_token = p->tokens[p->fill - 1];
reset_parser_state(p); reset_parser_state(p);
_PyPegen_parse(p); _PyPegen_parse(p);
if (PyErr_Occurred()) { if (PyErr_Occurred()) {
// Prioritize tokenizer errors to custom syntax errors raised // Prioritize tokenizer errors to custom syntax errors raised
// on the second phase only if the errors come from the parser. // on the second phase only if the errors come from the parser.
if (p->tok->done != E_ERROR && PyErr_ExceptionMatches(PyExc_SyntaxError)) { if (p->tok->done == E_DONE && PyErr_ExceptionMatches(PyExc_SyntaxError)) {
_PyPegen_check_tokenizer_errors(p); _PyPegen_check_tokenizer_errors(p);
} }
return NULL; return NULL;