hoelee/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2025-08-10 10:49:45 +00:00
Code Issues Actions 6 Packages Projects Releases Wiki Activity

[ie/theplatform] Extract more metadata (#8635)

Browse Source 
Authored by: trainman261
This commit is contained in:
trainman261
2023-12-12 01:00:35 +01:00
committed by GitHub
parent e370f9ec36
commit 7e09c147fd
7 changed files with 73 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
yt_dlp/extractor/theplatform.py
View File

@@ -104,6 +104,10 @@ class ThePlatformBaseIE(OnceIE):
_add_chapter(chapter.get('startTime'), chapter.get('endTime'))
_add_chapter(tp_chapters[-1].get('startTime'), tp_chapters[-1].get('endTime') or duration)
def extract_site_specific_field(field):
# A number of sites have custom-prefixed keys, e.g. 'cbc$seasonNumber'
return traverse_obj(info, lambda k, v: v and k.endswith(f'${field}'), get_all=False)
return {
'title': info['title'],
'subtitles': subtitles,
@@ -113,6 +117,14 @@ class ThePlatformBaseIE(OnceIE):
'timestamp': int_or_none(info.get('pubDate'), 1000) or None,
'uploader': info.get('billingCode'),
'chapters': chapters,
'creator': traverse_obj(info, ('author', {str})) or None,
'categories': traverse_obj(info, (
'categories', lambda _, v: v.get('label') in ('category', None), 'name', {str})) or None,
'tags': traverse_obj(info, ('keywords', {lambda x: re.split(r'[;,]\s?', x) if x else None})),
'location': extract_site_specific_field('region'),
'series': extract_site_specific_field('show'),
'season_number': int_or_none(extract_site_specific_field('seasonNumber')),
'media_type': extract_site_specific_field('programmingType') or extract_site_specific_field('type'),
}
def _extract_theplatform_metadata(self, path, video_id):