[youtube] Modify the regex to match ids of length 11 (fixes #1396)

In urls like http://www.youtube.com/watch?v=BaW_jenozKcsharePLED17F32AD9753930 you can't split the query string and ids always have that length.
This commit is contained in:
Jaime Marquínez Ferrándiz 2013-09-09 10:33:12 +02:00
parent 890f62e868
commit 8963d9c266
2 changed files with 8 additions and 5 deletions

View file

@ -72,10 +72,13 @@ class TestAllURLsMatching(unittest.TestCase):
self.assertTrue(JustinTVIE.suitable(u"http://www.twitch.tv/tsm_theoddone/c/2349361"))
def test_youtube_extract(self):
self.assertEqual(YoutubeIE()._extract_id('http://www.youtube.com/watch?&v=BaW_jenozKc'), 'BaW_jenozKc')
self.assertEqual(YoutubeIE()._extract_id('https://www.youtube.com/watch?&v=BaW_jenozKc'), 'BaW_jenozKc')
self.assertEqual(YoutubeIE()._extract_id('https://www.youtube.com/watch?feature=player_embedded&v=BaW_jenozKc'), 'BaW_jenozKc')
self.assertEqual(YoutubeIE()._extract_id('https://www.youtube.com/watch_popup?v=BaW_jenozKc'), 'BaW_jenozKc')
assertExtractId = lambda url, id: self.assertEqual(YoutubeIE()._extract_id(url), id)
assertExtractId('http://www.youtube.com/watch?&v=BaW_jenozKc', 'BaW_jenozKc')
assertExtractId('https://www.youtube.com/watch?&v=BaW_jenozKc', 'BaW_jenozKc')
assertExtractId('https://www.youtube.com/watch?feature=player_embedded&v=BaW_jenozKc', 'BaW_jenozKc')
assertExtractId('https://www.youtube.com/watch_popup?v=BaW_jenozKc', 'BaW_jenozKc')
assertExtractId('http://www.youtube.com/watch?v=BaW_jenozKcsharePLED17F32AD9753930', 'BaW_jenozKc')
assertExtractId('BaW_jenozKc', 'BaW_jenozKc')
def test_no_duplicates(self):
ies = gen_extractors()

View file

@ -150,7 +150,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
|youtu\.be/ # just youtu.be/xxxx
)
)? # all until now is optional -> you can pass the naked ID
([0-9A-Za-z_-]+) # here is it! the YouTube video ID
([0-9A-Za-z_-]{11}) # here is it! the YouTube video ID
(?(1).+)? # if we found the ID, everything can follow
$"""
_NEXT_URL_RE = r'[\?&]next_url=([^&]+)'