s = _parser.State()
s.flags = flags
for phrase, action in lexicon:
+ sub_pattern = _parser.parse(phrase, flags)
+ if sub_pattern.state.groups != 1:
+ raise ValueError("Cannot use capturing groups in re.Scanner")
gid = s.opengroup()
p.append(_parser.SubPattern(s, [
- (SUBPATTERN, (gid, 0, 0, _parser.parse(phrase, flags))),
+ (SUBPATTERN, (gid, 0, 0, sub_pattern)),
]))
s.closegroup(gid, p[-1])
p = _parser.SubPattern(s, [(BRANCH, (None, p))])
(['sum', 'op=', 3, 'op*', 'foo', 'op+', 312.5,
'op+', 'bar'], ''))
+ def test_bug_gh140797(self):
+ # gh140797: Capturing groups are not allowed in re.Scanner
+
+ msg = r"Cannot use capturing groups in re\.Scanner"
+ # Capturing group throws an error
+ with self.assertRaisesRegex(ValueError, msg):
+ Scanner([("(a)b", None)])
+
+ # Named Group
+ with self.assertRaisesRegex(ValueError, msg):
+ Scanner([("(?P<name>a)", None)])
+
+ # Non-capturing groups should pass normally
+ s = Scanner([("(?:a)b", lambda scanner, token: token)])
+ result, rem = s.scan("ab")
+ self.assertEqual(result,['ab'])
+ self.assertEqual(rem,'')
+
def test_bug_448951(self):
# bug 448951 (similar to 429357, but with single char match)
# (Also test greedy matches.)