[youtube] Try to extract the video_info from the webpage before requesting the 'get_video_info' pages

The YouTube player doesn't seem to use them except for embedded videos, so we can skip a network request.
But they still provide better error mesagges (for removed videos for example).
This commit is contained in:
Jaime Marquínez Ferrándiz 2014-11-30 20:56:32 +01:00
parent 4472f84f0c
commit 4e62ebe250
1 changed files with 24 additions and 36 deletions

View File

@ -675,7 +675,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
player_url = None player_url = None
# Get video info # Get video info
self.report_video_info_webpage_download(video_id)
if re.search(r'player-age-gate-content">', video_webpage) is not None: if re.search(r'player-age-gate-content">', video_webpage) is not None:
age_gate = True age_gate = True
# We simulate the access to the video from www.youtube.com/v/{video_id} # We simulate the access to the video from www.youtube.com/v/{video_id}
@ -694,11 +693,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
video_info = compat_parse_qs(video_info_webpage) video_info = compat_parse_qs(video_info_webpage)
else: else:
age_gate = False age_gate = False
try:
# Try looking directly into the video webpage
mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
if not mobj:
raise ValueError('Could not find ytplayer.config') # caught below
json_code = uppercase_escape(mobj.group(1))
ytplayer_config = json.loads(json_code)
args = ytplayer_config['args']
# Convert to the same format returned by compat_parse_qs
video_info = dict((k, [v]) for k, v in args.items())
if 'url_encoded_fmt_stream_map' not in args:
raise ValueError('No stream_map present') # caught below
except ValueError:
# We fallback to the get_video_info pages (used by the embed page)
self.report_video_info_webpage_download(video_id)
for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']: for el_type in ['&el=embedded', '&el=detailpage', '&el=vevo', '']:
video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en' video_info_url = (proto + '://www.youtube.com/get_video_info?&video_id=%s%s&ps=default&eurl=&gl=US&hl=en'
% (video_id, el_type)) % (video_id, el_type))
video_info_webpage = self._download_webpage(video_info_url, video_id, video_info_webpage = self._download_webpage(video_info_url,
note=False, video_id, note=False,
errnote='unable to download video info webpage') errnote='unable to download video info webpage')
video_info = compat_parse_qs(video_info_webpage) video_info = compat_parse_qs(video_info_webpage)
if 'token' in video_info: if 'token' in video_info:
@ -827,32 +841,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
if self._downloader.params.get('writeannotations', False): if self._downloader.params.get('writeannotations', False):
video_annotations = self._extract_annotations(video_id) video_annotations = self._extract_annotations(video_id)
# Decide which formats to download
try:
mobj = re.search(r';ytplayer\.config\s*=\s*({.*?});', video_webpage)
if not mobj:
raise ValueError('Could not find vevo ID')
json_code = uppercase_escape(mobj.group(1))
ytplayer_config = json.loads(json_code)
args = ytplayer_config['args']
# Easy way to know if the 's' value is in url_encoded_fmt_stream_map
# this signatures are encrypted
if 'url_encoded_fmt_stream_map' not in args:
raise ValueError('No stream_map present') # caught below
re_signature = re.compile(r'[&,]s=')
m_s = re_signature.search(args['url_encoded_fmt_stream_map'])
if m_s is not None:
self.to_screen('%s: Encrypted signatures detected.' % video_id)
video_info['url_encoded_fmt_stream_map'] = [args['url_encoded_fmt_stream_map']]
m_s = re_signature.search(args.get('adaptive_fmts', ''))
if m_s is not None:
if 'adaptive_fmts' in video_info:
video_info['adaptive_fmts'][0] += ',' + args['adaptive_fmts']
else:
video_info['adaptive_fmts'] = [args['adaptive_fmts']]
except ValueError:
pass
def _map_to_format_list(urlmap): def _map_to_format_list(urlmap):
formats = [] formats = []
for itag, video_real_url in urlmap.items(): for itag, video_real_url in urlmap.items():