aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--youtube_dl/extractor/yandexmusic.py40
1 files changed, 37 insertions, 3 deletions
diff --git a/youtube_dl/extractor/yandexmusic.py b/youtube_dl/extractor/yandexmusic.py
index 025716958..7a90cc60c 100644
--- a/youtube_dl/extractor/yandexmusic.py
+++ b/youtube_dl/extractor/yandexmusic.py
@@ -39,9 +39,14 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
'info_dict': {
'id': '4878838',
'ext': 'mp3',
- 'title': 'Carlo Ambrosio - Gypsy Eyes 1',
+ 'title': 'Carlo Ambrosio & Fabio Di Bari, Carlo Ambrosio - Gypsy Eyes 1',
'filesize': 4628061,
'duration': 193.04,
+ 'track': 'Gypsy Eyes 1',
+ 'album': 'Gypsy Soul',
+ 'album_artist': 'Carlo Ambrosio',
+ 'artist': 'Carlo Ambrosio & Fabio Di Bari, Carlo Ambrosio',
+ 'release_year': '2009',
}
}
@@ -64,16 +69,45 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
thumbnail = cover_uri.replace('%%', 'orig')
if not thumbnail.startswith('http'):
thumbnail = 'http://' + thumbnail
- return {
+
+ track_title = track['title']
+ track_info = {
'id': track['id'],
'ext': 'mp3',
'url': self._get_track_url(track['storageDir'], track['id']),
- 'title': '%s - %s' % (track['artists'][0]['name'], track['title']),
'filesize': int_or_none(track.get('fileSize')),
'duration': float_or_none(track.get('durationMs'), 1000),
'thumbnail': thumbnail,
+ 'track': track_title,
}
+ def extract_artist(artist_list):
+ if artist_list and isinstance(artist_list, list):
+ artists_names = [a['name'] for a in artist_list if a.get('name')]
+ if artists_names:
+ return ', '.join(artists_names)
+
+ albums = track.get('albums')
+ if albums and isinstance(albums, list):
+ album = albums[0]
+ if isinstance(album, dict):
+ year = album.get('year')
+ track_info.update({
+ 'album': album.get('title'),
+ 'album_artist': extract_artist(album.get('artists')),
+ 'release_year': compat_str(year) if year else None,
+ })
+
+ track_artist = extract_artist(track.get('artists'))
+ if track_artist:
+ track_info.update({
+ 'artist': track_artist,
+ 'title': '%s - %s' % (track_artist, track_title),
+ })
+ else:
+ track_info['title'] = track_title
+ return track_info
+
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
album_id, track_id = mobj.group('album_id'), mobj.group('id')