midgard/youtube-dl
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

[youtube] Improve yt initial player response extraction (closes #27216)

Browse source
This commit is contained in:
Sergey M․ 2020-11-28 15:02:31 +07:00
parent f9f9699f2f
commit 20c50c6556
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D
1 changed files with 7 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
youtube_dl/extractor/youtube.py
View file

@ -283,6 +283,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
} }
_YT_INITIAL_DATA_RE = r'(?:window\s*\[\s*["\']ytInitialData["\']\s*\]|ytInitialData)\s*=\s*({.+?})\s*;' _YT_INITIAL_DATA_RE = r'(?:window\s*\[\s*["\']ytInitialData["\']\s*\]|ytInitialData)\s*=\s*({.+?})\s*;'
_YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;'
def _call_api(self, ep, query, video_id): def _call_api(self, ep, query, video_id):
data = self._DEFAULT_API_DATA.copy() data = self._DEFAULT_API_DATA.copy()
@ -1068,7 +1069,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
}, },
}, },
{ {
# with '};' inside yt initial data (see https://github.com/ytdl-org/youtube-dl/issues/27093) # with '};' inside yt initial data (see [1])
# see [2] for an example with '};' inside ytInitialPlayerResponse
# 1. https://github.com/ytdl-org/youtube-dl/issues/27093
# 2. https://github.com/ytdl-org/youtube-dl/issues/27216
'url': 'https://www.youtube.com/watch?v=CHqg6qOn4no', 'url': 'https://www.youtube.com/watch?v=CHqg6qOn4no',
'info_dict': { 'info_dict': {
'id': 'CHqg6qOn4no', 'id': 'CHqg6qOn4no',
@ -1686,7 +1690,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
if not video_info and not player_response: if not video_info and not player_response:
player_response = extract_player_response( player_response = extract_player_response(
self._search_regex( self._search_regex(
r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;', video_webpage, (r'%s\s*(?:var\s+meta|</script|\n)' % self._YT_INITIAL_PLAYER_RESPONSE_RE,
self._YT_INITIAL_PLAYER_RESPONSE_RE), video_webpage,
'initial player response', default='{}'), 'initial player response', default='{}'),
video_id) video_id)