w/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

[yahoo] Extend _VALID_URL and fix extraction

Browse source 
Closes #9271
This commit is contained in:
Yen Chi Hsuan 2016-04-24 17:01:18 +08:00
parent 949b6497cc
commit 2a7c38831c
No known key found for this signature in database
GPG key ID: 3FDDD575826C5C30
1 changed files with 13 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
youtube_dl/extractor/yahoo.py
View file

@ -24,7 +24,7 @@ from .nbc import NBCSportsVPlayerIE
class YahooIE(InfoExtractor): class YahooIE(InfoExtractor):
IE_DESC = 'Yahoo screen and movies' IE_DESC = 'Yahoo screen and movies'
_VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+)?-(?P<id>[0-9]+)(?:-[a-z]+)?\.html)' _VALID_URL = r'(?P<url>(?P<host>https?://(?:[a-zA-Z]{2}\.)?[\da-zA-Z_-]+\.yahoo\.com)/(?:[^/]+/)*(?P<display_id>.+)?-(?P<id>[0-9]+)(?:-[a-z]+)?(?:\.html)?)'
_TESTS = [ _TESTS = [
{ {
'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html', 'url': 'http://screen.yahoo.com/julian-smith-travis-legg-watch-214727115.html',
@ -166,6 +166,17 @@ class YahooIE(InfoExtractor):
'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.', 'description': 'While they play feuding fathers in \'Daddy\'s Home,\' star Will Ferrell & Mark Wahlberg share their true feelings on parenthood.',
}, },
}, },
{
# config['models']['applet_model']['data']['sapi'] has no query
'url': 'https://www.yahoo.com/music/livenation/event/galactic-2016',
'md5': 'dac0c72d502bc5facda80c9e6d5c98db',
'info_dict': {
'id': 'a6015640-e9e5-3efb-bb60-05589a183919',
'ext': 'mp4',
'description': 'Galactic',
'title': 'Dolla Diva (feat. Maggie Koerner)',
},
},
] ]
def _real_extract(self, url): def _real_extract(self, url):
@ -202,7 +213,7 @@ class YahooIE(InfoExtractor):
config = self._parse_json(config_json, display_id, fatal=False) config = self._parse_json(config_json, display_id, fatal=False)
if config: if config:
sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi') sapi = config.get('models', {}).get('applet_model', {}).get('data', {}).get('sapi')
if sapi: if sapi and 'query' in sapi:
return self._extract_info(display_id, sapi, webpage) return self._extract_info(display_id, sapi, webpage)
items_json = self._search_regex( items_json = self._search_regex(