From deef31955bb6106939606cd2a8b677db700055e3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sergey=20M=E2=80=A4?= Date: Sun, 30 Apr 2017 21:07:30 +0700 Subject: [PATCH] [utils] Improve unified_timestamp Seen at http://zaq1.pl/video/xev0e --- test/test_utils.py | 1 + youtube_dl/utils.py | 5 +++++ 2 files changed, 6 insertions(+) diff --git a/test/test_utils.py b/test/test_utils.py index 9e1808e6b..05fdc0e95 100644 --- a/test/test_utils.py +++ b/test/test_utils.py @@ -338,6 +338,7 @@ def test_unified_timestamps(self): self.assertEqual(unified_timestamp('UNKNOWN DATE FORMAT'), None) self.assertEqual(unified_timestamp('May 16, 2016 11:15 PM'), 1463440500) self.assertEqual(unified_timestamp('Feb 7, 2016 at 6:35 pm'), 1454870100) + self.assertEqual(unified_timestamp('2017-03-30T17:52:41Q'), 1490896361) def test_determine_ext(self): self.assertEqual(determine_ext('http://example.com/foo/bar.mp4/?download'), 'mp4') diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 9d7798ee8..86fc5ccac 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -1194,6 +1194,11 @@ def unified_timestamp(date_str, day_first=True): # Remove AM/PM + timezone date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str) + # Remove unrecognized timezones from ISO 8601 alike timestamps + m = re.search(r'\d{1,2}:\d{1,2}(?:\.\d+)?(?P\s*[A-Z]+)$', date_str) + if m: + date_str = date_str[:-len(m.group('tz'))] + for expression in date_formats(day_first): try: dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)