[yahoo] Generalize, support arbitrary subdomains, support iframe videos, capture error message (Closes #2470)

This commit is contained in:
Sergey M․ 2014-10-10 22:11:30 +07:00
parent 762958d5af
commit 3c06d3715e
2 changed files with 81 additions and 37 deletions

View File

@ -463,7 +463,7 @@ from .xvideos import XVideosIE
from .xtube import XTubeUserIE, XTubeIE from .xtube import XTubeUserIE, XTubeIE
from .yahoo import ( from .yahoo import (
YahooIE, YahooIE,
YahooNewsIE, #YahooNewsIE,
YahooSearchIE, YahooSearchIE,
) )
from .ynet import YnetIE from .ynet import YnetIE

View File

@ -7,6 +7,7 @@ import re
from .common import InfoExtractor, SearchInfoExtractor from .common import InfoExtractor, SearchInfoExtractor
from ..utils import ( from ..utils import (
ExtractorError,
compat_urllib_parse, compat_urllib_parse,
compat_urlparse, compat_urlparse,
clean_html, clean_html,
@ -16,7 +17,7 @@ from ..utils import (
class YahooIE(InfoExtractor): class YahooIE(InfoExtractor):
IE_DESC = 'Yahoo screen and movies' IE_DESC = 'Yahoo screen and movies'
_VALID_URL = r'(?P<url>https?://(?:.+?\.)?(?:screen|movies)\.yahoo\.com/.*?-(?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',
@ -73,13 +74,70 @@ class YahooIE(InfoExtractor):
'duration': 97, 'duration': 97,
} }
}, },
{
'url': 'https://ca.sports.yahoo.com/video/program-makes-hockey-more-affordable-013127711.html',
'md5': '57e06440778b1828a6079d2f744212c4',
'info_dict': {
'id': 'c9fa2a36-0d4d-3937-b8f6-cc0fb1881e73',
'ext': 'mp4',
'title': 'Program that makes hockey more affordable not offered in Manitoba',
'description': 'md5:c54a609f4c078d92b74ffb9bf1f496f4',
'duration': 121,
}
}, {
'url': 'https://ca.finance.yahoo.com/news/20-most-valuable-brands-world-112600775.html',
'md5': '3e401e4eed6325aa29d9b96125fd5b4f',
'info_dict': {
'id': 'c1b4c09c-8ed8-3b65-8b05-169c55358a83',
'ext': 'mp4',
'title': "Apple Is The World's Most Valuable Brand",
'description': 'md5:73eabc1a11c6f59752593b2ceefa1262',
'duration': 21,
}
}, {
'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
'md5': '67010fdf3a08d290e060a4dd96baa07b',
'info_dict': {
'id': 'f885cf7f-43d4-3450-9fac-46ac30ece521',
'ext': 'mp4',
'title': 'China Moses Is Crazy About the Blues',
'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
'duration': 128,
}
}, {
'url': 'https://in.lifestyle.yahoo.com/video/connect-dots-dark-side-virgo-090247395.html',
'md5': 'd9a083ccf1379127bf25699d67e4791b',
'info_dict': {
'id': '52aeeaa3-b3d1-30d8-9ef8-5d0cf05efb7c',
'ext': 'mp4',
'title': 'Connect the Dots: Dark Side of Virgo',
'description': 'md5:1428185051cfd1949807ad4ff6d3686a',
'duration': 201,
}
}, {
'url': 'https://gma.yahoo.com/pizza-delivery-man-surprised-huge-tip-college-kids-195200785.html',
'only_matching': True,
}
] ]
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id') display_id = mobj.group('display_id')
url = mobj.group('url') url = mobj.group('url')
webpage = self._download_webpage(url, video_id) host = mobj.group('host')
webpage = self._download_webpage(url, display_id)
# Look for iframed media first
iframe_m = re.search(r'<iframe[^>]+src="(/video/.+?-\d+\.html\?format=embed.*?)"', webpage)
if iframe_m:
iframepage = self._download_webpage(
host + iframe_m.group(1), display_id, 'Downloading iframe webpage')
items_json = self._search_regex(
r'mediaItems: (\[.+?\])$', iframepage, 'items', flags=re.MULTILINE, default=None)
if items_json:
items = json.loads(items_json)
video_id = items[0]['id']
return self._get_info(video_id, display_id, webpage)
items_json = self._search_regex( items_json = self._search_regex(
r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE, r'mediaItems: ({.*?})$', webpage, 'items', flags=re.MULTILINE,
@ -90,22 +148,22 @@ class YahooIE(InfoExtractor):
r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"', r'root\.App\.Cache\.context\.videoCache\.curVideo = \{"([^"]+)"',
r'"first_videoid"\s*:\s*"([^"]+)"', r'"first_videoid"\s*:\s*"([^"]+)"',
] ]
long_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID') video_id = self._search_regex(CONTENT_ID_REGEXES, webpage, 'content ID')
video_id = long_id
else: else:
items = json.loads(items_json) items = json.loads(items_json)
info = items['mediaItems']['query']['results']['mediaObj'][0] info = items['mediaItems']['query']['results']['mediaObj'][0]
# The 'meta' field is not always in the video webpage, we request it # The 'meta' field is not always in the video webpage, we request it
# from another page # from another page
long_id = info['id'] video_id = info['id']
return self._get_info(long_id, video_id, webpage) return self._get_info(video_id, display_id, webpage)
def _get_info(self, long_id, video_id, webpage): def _get_info(self, video_id, display_id, webpage):
region = self._search_regex( region = self._search_regex(
r'"region"\s*:\s*"([^"]+)"', webpage, 'region', fatal=False, default='US') r'\\?"region\\?"\s*:\s*\\?"([^"]+?)\\?"',
webpage, 'region', fatal=False, default='US')
query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"' query = ('SELECT * FROM yahoo.media.video.streams WHERE id="%s"'
' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="%s"' ' AND plrs="86Gj0vCaSzV_Iuf6hNylf2" AND region="%s"'
' AND protocol="http"' % (long_id, region)) ' AND protocol="http"' % (video_id, region))
data = compat_urllib_parse.urlencode({ data = compat_urllib_parse.urlencode({
'q': query, 'q': query,
'env': 'prod', 'env': 'prod',
@ -113,9 +171,17 @@ class YahooIE(InfoExtractor):
}) })
query_result = self._download_json( query_result = self._download_json(
'http://video.query.yahoo.com/v1/public/yql?' + data, 'http://video.query.yahoo.com/v1/public/yql?' + data,
video_id, 'Downloading video info') display_id, 'Downloading video info')
info = query_result['query']['results']['mediaObj'][0] info = query_result['query']['results']['mediaObj'][0]
meta = info['meta'] meta = info.get('meta')
if not meta:
msg = info['status'].get('msg')
if msg:
raise ExtractorError(
'%s returned error: %s' % (self.IE_NAME, msg), expected=True)
raise ExtractorError('Unable to extract media object meta')
formats = [] formats = []
for s in info['streams']: for s in info['streams']:
@ -142,6 +208,7 @@ class YahooIE(InfoExtractor):
return { return {
'id': video_id, 'id': video_id,
'display_id': display_id,
'title': meta['title'], 'title': meta['title'],
'formats': formats, 'formats': formats,
'description': clean_html(meta['description']), 'description': clean_html(meta['description']),
@ -150,29 +217,6 @@ class YahooIE(InfoExtractor):
} }
class YahooNewsIE(YahooIE):
IE_NAME = 'yahoo:news'
_VALID_URL = r'http://news\.yahoo\.com/video/.*?-(?P<id>\d*?)\.html'
_TESTS = [{
'url': 'http://news.yahoo.com/video/china-moses-crazy-blues-104538833.html',
'md5': '67010fdf3a08d290e060a4dd96baa07b',
'info_dict': {
'id': '104538833',
'ext': 'mp4',
'title': 'China Moses Is Crazy About the Blues',
'description': 'md5:9900ab8cd5808175c7b3fe55b979bed0',
},
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
video_id = mobj.group('id')
webpage = self._download_webpage(url, video_id)
long_id = self._search_regex(r'contentId: \'(.+?)\',', webpage, 'long id')
return self._get_info(long_id, video_id, webpage)
class YahooSearchIE(SearchInfoExtractor): class YahooSearchIE(SearchInfoExtractor):
IE_DESC = 'Yahoo screen search' IE_DESC = 'Yahoo screen search'
_MAX_RESULTS = 1000 _MAX_RESULTS = 1000