mirrors/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-13 20:03:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix #93 YoutubePlaylistsIE

Browse source
This commit is contained in:
Ali Sherief 2020-11-09 16:06:48 +00:00
parent 651bae3d23
commit 876f1c17ff
No known key found for this signature in database
GPG key ID: 1C3D1D19C8E8B6D8
1 changed files with 4 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
youtube_dlc/extractor/youtube.py
View file

@ -300,11 +300,12 @@ class YoutubeEntryListBaseInfoExtractor(YoutubeBaseInfoExtractor):
# Extract entries from page with "Load more" button
def _entries(self, page, playlist_id):
more_widget_html = content_html = page
mobj_reg = r'(?:(?:data-uix-load-more-href="[^"]+?;continuation=)|(?:"continuation":"))(?P<more>[^"]+)"'
for page_num in itertools.count(1):
for entry in self._process_page(content_html):
yield entry
mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html)
mobj = re.search(mobj_reg, more_widget_html)
if not mobj:
break
@ -315,7 +316,7 @@ def _entries(self, page, playlist_id):
# Downloading page may result in intermittent 5xx HTTP error
# that is usually worked around with a retry
more = self._download_json(
'https://www.youtube.com/%s' % mobj.group('more'), playlist_id,
'https://www.youtube.com/browse_ajax?ctoken=%s' % mobj.group('more'), playlist_id,
'Downloading page #%s%s'
% (page_num, ' (retry #%d)' % count if count else ''),
transform_source=uppercase_escape,
@ -372,7 +373,7 @@ def extract_videos_from_page(self, page):
class YoutubePlaylistsBaseInfoExtractor(YoutubeEntryListBaseInfoExtractor):
def _process_page(self, content):
for playlist_id in orderedSet(re.findall(
r'<h3[^>]+class="[^"]*yt-lockup-title[^"]*"[^>]*><a[^>]+href="/?playlist\?list=([0-9A-Za-z-_]{10,})"',
r'"/?playlist\?list=([0-9A-Za-z-_]{10,})"',
content)):
yield self.url_result(
'https://www.youtube.com/playlist?list=%s' % playlist_id, 'YoutubePlaylist')