[youtube] Improve yt initial data extraction (closes #27524)

This commit is contained in:
Sergey M․ 2020-12-29 02:29:34 +07:00
parent 71febd1c52
commit 1a95953867
No known key found for this signature in database
GPG key ID: 2C393E0F18A9236D

View file

@ -280,6 +280,7 @@ 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*;' _YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;'
_YT_INITIAL_BOUNDARY_RE = r'(?:var\s+meta|</script|\n)'
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()
@ -297,7 +298,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
def _extract_yt_initial_data(self, video_id, webpage): def _extract_yt_initial_data(self, video_id, webpage):
return self._parse_json( return self._parse_json(
self._search_regex( self._search_regex(
(r'%s\s*\n' % self._YT_INITIAL_DATA_RE, (r'%s\s*%s' % (self._YT_INITIAL_DATA_RE, self._YT_INITIAL_BOUNDARY_RE),
self._YT_INITIAL_DATA_RE), webpage, 'yt initial data'), self._YT_INITIAL_DATA_RE), webpage, 'yt initial data'),
video_id) video_id)
@ -1103,6 +1104,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
'skip_download': True, 'skip_download': True,
}, },
}, },
{
# another example of '};' in ytInitialData
'url': 'https://www.youtube.com/watch?v=gVfgbahppCY',
'only_matching': True,
},
{ {
'url': 'https://www.youtube.com/watch_popup?v=63RmMXCd_bQ', 'url': 'https://www.youtube.com/watch_popup?v=63RmMXCd_bQ',
'only_matching': True, 'only_matching': True,
@ -1706,7 +1712,7 @@ 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'%s\s*(?:var\s+meta|</script|\n)' % self._YT_INITIAL_PLAYER_RESPONSE_RE, (r'%s\s*%s' % (self._YT_INITIAL_PLAYER_RESPONSE_RE, self._YT_INITIAL_BOUNDARY_RE),
self._YT_INITIAL_PLAYER_RESPONSE_RE), video_webpage, self._YT_INITIAL_PLAYER_RESPONSE_RE), video_webpage,
'initial player response', default='{}'), 'initial player response', default='{}'),
video_id) video_id)