diff options
author | Sergey M․ <dstftw@gmail.com> | 2015-05-15 21:42:34 +0600 |
---|---|---|
committer | Sergey M․ <dstftw@gmail.com> | 2015-05-15 21:42:34 +0600 |
commit | 62c95fd5fcb8dbea2faeb4edac4c5177cbac5912 (patch) | |
tree | 7b4761274ee89593399cd25998e754b6c262f4bb /youtube_dl/extractor | |
parent | 25f14e9f93295a787e0cb436a5f6179d6174733d (diff) |
[youtube:feed] Check each 'load more' portion for unique video ids
Diffstat (limited to 'youtube_dl/extractor')
-rw-r--r-- | youtube_dl/extractor/youtube.py | 10 |
1 files changed, 8 insertions, 2 deletions
diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 9096a2975..1f9940cf5 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -1621,10 +1621,16 @@ class YoutubeFeedsInfoExtractor(YoutubeBaseInfoExtractor): # for the video ids doesn't contain an index ids = [] more_widget_html = content_html = page - for page_num in itertools.count(1): matches = re.findall(r'href="\s*/watch\?v=([0-9A-Za-z_-]{11})', content_html) - new_ids = orderedSet(matches) + + # 'recommended' feed has infinite 'load more' and each new portion spins + # the same videos in (sometimes) slightly different order, so we'll check + # for unicity and break when portion has no new videos + new_ids = filter(lambda video_id: video_id not in ids, orderedSet(matches)) + if not new_ids: + break + ids.extend(new_ids) mobj = re.search(r'data-uix-load-more-href="/?(?P<more>[^"]+)"', more_widget_html) |