commit: 20c50c65566be31607e59d7ca2467c664a29c843
parent f9f9699f2fa87ec54d59b1834cc56adb6147fec6
Author: Sergey M․ <dstftw@gmail.com>
Date: Sat, 28 Nov 2020 15:02:31 +0700
[youtube] Improve yt initial player response extraction (closes #27216)
Diffstat:
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py
@@ -283,6 +283,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
}
_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):
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',
'info_dict': {
'id': 'CHqg6qOn4no',
@@ -1686,7 +1690,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
if not video_info and not player_response:
player_response = extract_player_response(
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='{}'),
video_id)