w/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

[youtube] improve title and description extraction(closes #21934)

Browse source
This commit is contained in:
Remita Amine 2019-07-30 00:13:33 +01:00
parent 90634acfcf
commit 8dbf751aa2
1 changed files with 3 additions and 12 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
youtube_dl/extractor/youtube.py
View file

@ -1820,16 +1820,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
video_details = try_get( video_details = try_get(
player_response, lambda x: x['videoDetails'], dict) or {} player_response, lambda x: x['videoDetails'], dict) or {}
# title video_title = video_info.get('title', [None])[0] or video_details.get('title')
if 'title' in video_info: if not video_title:
video_title = video_info['title'][0]
elif 'title' in player_response:
video_title = video_details['title']
else:
self._downloader.report_warning('Unable to extract video title') self._downloader.report_warning('Unable to extract video title')
video_title = '_' video_title = '_'
# description
description_original = video_description = get_element_by_id("eow-description", video_webpage) description_original = video_description = get_element_by_id("eow-description", video_webpage)
if video_description: if video_description:
@ -1854,11 +1849,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
''', replace_url, video_description) ''', replace_url, video_description)
video_description = clean_html(video_description) video_description = clean_html(video_description)
else: else:
fd_mobj = re.search(r'<meta name="description" content="([^"]+)"', video_webpage) video_description = self._html_search_meta('description', video_webpage) or video_details.get('shortDescription')
if fd_mobj:
video_description = unescapeHTML(fd_mobj.group(1))
else:
video_description = ''
if not smuggled_data.get('force_singlefeed', False): if not smuggled_data.get('force_singlefeed', False):
if not self._downloader.params.get('noplaylist'): if not self._downloader.params.get('noplaylist'):