return
try:
stat = os.stat(fullname)
- except OSError:
+ except (OSError, ValueError):
cache.pop(filename, None)
continue
if size != stat.st_size or mtime != stat.st_mtime:
try:
stat = os.stat(fullname)
break
- except OSError:
+ except (OSError, ValueError):
pass
else:
return []
+ except ValueError: # may be raised by os.stat()
+ return []
try:
with tokenize.open(fullname) as fp:
lines = fp.readlines()
self.assertEqual(linecache.getlines(filename, module_globals),
['source for x.y.z\n'])
+ def test_invalid_names(self):
+ for name, desc in [
+ ('\x00', 'NUL bytes filename'),
+ (__file__ + '\x00', 'filename with embedded NUL bytes'),
+ # A filename with surrogate codes. A UnicodeEncodeError is raised
+ # by os.stat() upon querying, which is a subclass of ValueError.
+ ("\uD834\uDD1E.py", 'surrogate codes (MUSICAL SYMBOL G CLEF)'),
+ # For POSIX platforms, an OSError will be raised but for Windows
+ # platforms, a ValueError is raised due to the path_t converter.
+ # See: https://github.com/python/cpython/issues/122170
+ ('a' * 1_000_000, 'very long filename'),
+ ]:
+ with self.subTest(f'updatecache: {desc}'):
+ linecache.clearcache()
+ lines = linecache.updatecache(name)
+ self.assertListEqual(lines, [])
+ self.assertNotIn(name, linecache.cache)
+
+ # hack into the cache (it shouldn't be allowed
+ # but we never know what people do...)
+ for key, fullname in [(name, 'ok'), ('key', name), (name, name)]:
+ with self.subTest(f'checkcache: {desc}',
+ key=key, fullname=fullname):
+ linecache.clearcache()
+ linecache.cache[key] = (0, 1234, [], fullname)
+ linecache.checkcache(key)
+ self.assertNotIn(key, linecache.cache)
+
+ # just to be sure that we did not mess with cache
+ linecache.clearcache()
+
class LineCacheInvalidationTests(unittest.TestCase):
def setUp(self):