minhngoc25a
/
yt-dlc
mirror of https://github.com/blackjack4494/yt-dlc
1
0
Fork 1
Code Issues Releases Wiki Activity

[anvato] Extract more metadata

This commit is contained in:
Sergey M․ 2016-11-04 21:17:56 +07:00
parent c4c9b8440c
commit f93ac1d175
No known key found for this signature in database
GPG Key ID: 2C393E0F18A9236D
1 changed files with 13 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
youtube_dl/extractor/anvato.py
View File

@ -157,14 +157,7 @@ class AnvatoIE(InfoExtractor):
video_data_url, video_id, transform_source=strip_jsonp, video_data_url, video_id, transform_source=strip_jsonp,
data=json.dumps(payload).encode('utf-8')) data=json.dumps(payload).encode('utf-8'))
def _extract_anvato_videos(self, webpage, video_id): def _get_anvato_videos(self, access_key, video_id):
anvplayer_data = self._parse_json(self._html_search_regex(
r'<script[^>]+data-anvp=\'([^\']+)\'', webpage,
'Anvato player data'), video_id)
video_id = anvplayer_data['video']
access_key = anvplayer_data['accessKey']
video_data = self._get_video_json(access_key, video_id) video_data = self._get_video_json(access_key, video_id)
formats = [] formats = []
@ -218,7 +211,19 @@ class AnvatoIE(InfoExtractor):
'formats': formats, 'formats': formats,
'title': video_data.get('def_title'), 'title': video_data.get('def_title'),
'description': video_data.get('def_description'), 'description': video_data.get('def_description'),
'tags': video_data.get('def_tags', '').split(','),
'categories': video_data.get('categories'), 'categories': video_data.get('categories'),
'thumbnail': video_data.get('thumbnail'), 'thumbnail': video_data.get('thumbnail'),
'timestamp': int_or_none(video_data.get(
'ts_published') or video_data.get('ts_added')),
'uploader': video_data.get('mcp_id'),
'duration': int_or_none(video_data.get('duration')),
'subtitles': subtitles, 'subtitles': subtitles,
} }
def _extract_anvato_videos(self, webpage, video_id):
anvplayer_data = self._parse_json(self._html_search_regex(
r'<script[^>]+data-anvp=\'([^\']+)\'', webpage,
'Anvato player data'), video_id)
return self._get_anvato_videos(
anvplayer_data['accessKey'], anvplayer_data['video'])