commit b477fc13144be90ba7bbec3386f2e48fa3e7d604
parent d84b21b427d080a031ca1f9c75beb1dc8024b900
Author: Sergey M․ <dstftw@gmail.com>
Date:   Tue, 16 Jun 2020 02:29:04 +0700

[youtube] Fix thumbnails extraction and remove uploader id extraction warning (closes #25676)

Diffstat:
Myoutube_dl/extractor/youtube.py | 40+++++++++++++++++++++++++++-------------
1 file changed, 27 insertions(+), 13 deletions(-)

diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py @@ -2225,8 +2225,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor): if mobj is not None: video_uploader_id = mobj.group('uploader_id') video_uploader_url = mobj.group('uploader_url') - else: - self._downloader.report_warning('unable to extract uploader nickname') channel_id = ( str_or_none(video_details.get('channelId')) @@ -2237,17 +2235,33 @@ class YoutubeIE(YoutubeBaseInfoExtractor): video_webpage, 'channel id', default=None, group='id')) channel_url = 'http://www.youtube.com/channel/%s' % channel_id if channel_id else None - # thumbnail image - # We try first to get a high quality image: - m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">', - video_webpage, re.DOTALL) - if m_thumb is not None: - video_thumbnail = m_thumb.group(1) - elif 'thumbnail_url' not in video_info: - self._downloader.report_warning('unable to extract video thumbnail') + thumbnails = [] + thumbnails_list = try_get( + video_details, lambda x: x['thumbnail']['thumbnails'], list) or [] + for t in thumbnails_list: + if not isinstance(t, dict): + continue + thumbnail_url = url_or_none(t.get('url')) + if not thumbnail_url: + continue + thumbnails.append({ + 'url': thumbnail_url, + 'width': int_or_none(t.get('width')), + 'height': int_or_none(t.get('height')), + }) + + if not thumbnails: video_thumbnail = None - else: # don't panic if we can't find it - video_thumbnail = compat_urllib_parse_unquote_plus(video_info['thumbnail_url'][0]) + # We try first to get a high quality image: + m_thumb = re.search(r'<span itemprop="thumbnail".*?href="(.*?)">', + video_webpage, re.DOTALL) + if m_thumb is not None: + video_thumbnail = m_thumb.group(1) + thumbnail_url = try_get(video_info, lambda x: x['thumbnail_url'][0], compat_str) + if thumbnail_url: + video_thumbnail = compat_urllib_parse_unquote_plus(thumbnail_url) + if video_thumbnail: + thumbnails.append({'url': video_thumbnail}) # upload date upload_date = self._html_search_meta( @@ -2480,7 +2494,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): 'creator': video_creator or artist, 'title': video_title, 'alt_title': video_alt_title or track, - 'thumbnail': video_thumbnail, + 'thumbnails': thumbnails, 'description': video_description, 'categories': video_categories, 'tags': video_tags,