yt_dlp/extractor/yandexmusic.py

   1 import hashlib
   2 import itertools
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     ExtractorError,
   7     float_or_none,
   8     int_or_none,
   9     try_get,
  10 )
  11
  12
  13 class YandexMusicBaseIE(InfoExtractor):
  14     _VALID_URL_BASE = r'https?://music\.yandex\.(?P<tld>ru|kz|ua|by|com)'
  15
  16     @staticmethod
  17     def _handle_error(response):
  18         if isinstance(response, dict):
  19             error = response.get('error')
  20             if error:
  21                 raise ExtractorError(error, expected=True)
  22             if response.get('type') == 'captcha' or 'captcha' in response:
  23                 YandexMusicBaseIE._raise_captcha()
  24
  25     @staticmethod
  26     def _raise_captcha():
  27         raise ExtractorError(
  28             'YandexMusic has considered yt-dlp requests automated and '
  29             'asks you to solve a CAPTCHA. You can either wait for some '
  30             'time until unblocked and optionally use --sleep-interval '
  31             'in future or alternatively you can go to https://music.yandex.ru/ '
  32             'solve CAPTCHA, then export cookies and pass cookie file to '
  33             'yt-dlp with --cookies',
  34             expected=True)
  35
  36     def _download_webpage_handle(self, *args, **kwargs):
  37         webpage = super()._download_webpage_handle(*args, **kwargs)
  38         if 'Нам очень жаль, но&nbsp;запросы, поступившие с&nbsp;вашего IP-адреса, похожи на&nbsp;автоматические.' in webpage:
  39             self._raise_captcha()
  40         return webpage
  41
  42     def _download_json(self, *args, **kwargs):
  43         response = super()._download_json(*args, **kwargs)
  44         self._handle_error(response)
  45         return response
  46
  47     def _call_api(self, ep, tld, url, item_id, note, query):
  48         return self._download_json(
  49             f'https://music.yandex.{tld}/handlers/{ep}.jsx',
  50             item_id, note,
  51             fatal=False,
  52             headers={
  53                 'Referer': url,
  54                 'X-Requested-With': 'XMLHttpRequest',
  55                 'X-Retpath-Y': url,
  56             },
  57             query=query)
  58
  59
  60 class YandexMusicTrackIE(YandexMusicBaseIE):
  61     IE_NAME = 'yandexmusic:track'
  62     IE_DESC = 'Яндекс.Музыка - Трек'
  63     _VALID_URL = rf'{YandexMusicBaseIE._VALID_URL_BASE}/album/(?P<album_id>\d+)/track/(?P<id>\d+)'
  64
  65     _TESTS = [{
  66         'url': 'http://music.yandex.ru/album/540508/track/4878838',
  67         'md5': 'dec8b661f12027ceaba33318787fff76',
  68         'info_dict': {
  69             'id': '4878838',
  70             'ext': 'mp3',
  71             'title': 'md5:c63e19341fdbe84e43425a30bc777856',
  72             'filesize': int,
  73             'duration': 193.04,
  74             'track': 'md5:210508c6ffdfd67a493a6c378f22c3ff',
  75             'album': 'md5:cd04fb13c4efeafdfa0a6a6aca36d01a',
  76             'album_artist': 'md5:5f54c35462c07952df33d97cfb5fc200',
  77             'artist': 'md5:e6fd86621825f14dc0b25db3acd68160',
  78             'release_year': 2009,
  79         },
  80         # 'skip': 'Travis CI servers blocked by YandexMusic',
  81     }, {
  82         # multiple disks
  83         'url': 'http://music.yandex.ru/album/3840501/track/705105',
  84         'md5': '82a54e9e787301dd45aba093cf6e58c0',
  85         'info_dict': {
  86             'id': '705105',
  87             'ext': 'mp3',
  88             'title': 'md5:f86d4a9188279860a83000277024c1a6',
  89             'filesize': int,
  90             'duration': 239.27,
  91             'track': 'md5:40f887f0666ba1aa10b835aca44807d1',
  92             'album': 'md5:624f5224b14f5c88a8e812fd7fbf1873',
  93             'album_artist': 'md5:dd35f2af4e8927100cbe6f5e62e1fb12',
  94             'artist': 'md5:dd35f2af4e8927100cbe6f5e62e1fb12',
  95             'release_year': 2016,
  96             'genre': 'pop',
  97             'disc_number': 2,
  98             'track_number': 9,
  99         },
 100         # 'skip': 'Travis CI servers blocked by YandexMusic',
 101     }, {
 102         'url': 'http://music.yandex.com/album/540508/track/4878838',
 103         'only_matching': True,
 104     }]
 105
 106     def _real_extract(self, url):
 107         mobj = self._match_valid_url(url)
 108         tld, album_id, track_id = mobj.group('tld'), mobj.group('album_id'), mobj.group('id')
 109
 110         track = self._call_api(
 111             'track', tld, url, track_id, 'Downloading track JSON',
 112             {'track': f'{track_id}:{album_id}'})['track']
 113         track_title = track['title']
 114
 115         download_data = self._download_json(
 116             f'https://music.yandex.ru/api/v2.1/handlers/track/{track_id}:{album_id}/web-album_track-track-track-main/download/m',
 117             track_id, 'Downloading track location url JSON', query={'hq': 1}, headers={'X-Retpath-Y': url})
 118
 119         fd_data = self._download_json(
 120             download_data['src'], track_id,
 121             'Downloading track location JSON',
 122             query={'format': 'json'})
 123         key = hashlib.md5(('XGRlBW9FXlekgbPrRHuSiA' + fd_data['path'][1:] + fd_data['s']).encode()).hexdigest()
 124         f_url = 'http://{}/get-mp3/{}/{}?track-id={} '.format(fd_data['host'], key, fd_data['ts'] + fd_data['path'], track['id'])
 125
 126         thumbnail = None
 127         cover_uri = track.get('albums', [{}])[0].get('coverUri')
 128         if cover_uri:
 129             thumbnail = cover_uri.replace('%%', 'orig')
 130             if not thumbnail.startswith('http'):
 131                 thumbnail = 'http://' + thumbnail
 132
 133         track_info = {
 134             'id': track_id,
 135             'ext': 'mp3',
 136             'url': f_url,
 137             'filesize': int_or_none(track.get('fileSize')),
 138             'duration': float_or_none(track.get('durationMs'), 1000),
 139             'thumbnail': thumbnail,
 140             'track': track_title,
 141             'acodec': download_data.get('codec'),
 142             'abr': int_or_none(download_data.get('bitrate')),
 143         }
 144
 145         def extract_artist_name(artist):
 146             decomposed = artist.get('decomposed')
 147             if not isinstance(decomposed, list):
 148                 return artist['name']
 149             parts = [artist['name']]
 150             for element in decomposed:
 151                 if isinstance(element, dict) and element.get('name'):
 152                     parts.append(element['name'])
 153                 elif isinstance(element, str):
 154                     parts.append(element)
 155             return ''.join(parts)
 156
 157         def extract_artist(artist_list):
 158             if artist_list and isinstance(artist_list, list):
 159                 artists_names = [extract_artist_name(a) for a in artist_list if a.get('name')]
 160                 if artists_names:
 161                     return ', '.join(artists_names)
 162
 163         albums = track.get('albums')
 164         if albums and isinstance(albums, list):
 165             album = albums[0]
 166             if isinstance(album, dict):
 167                 year = album.get('year')
 168                 disc_number = int_or_none(try_get(
 169                     album, lambda x: x['trackPosition']['volume']))
 170                 track_number = int_or_none(try_get(
 171                     album, lambda x: x['trackPosition']['index']))
 172                 track_info.update({
 173                     'album': album.get('title'),
 174                     'album_artist': extract_artist(album.get('artists')),
 175                     'release_year': int_or_none(year),
 176                     'genre': album.get('genre'),
 177                     'disc_number': disc_number,
 178                     'track_number': track_number,
 179                 })
 180
 181         track_artist = extract_artist(track.get('artists'))
 182         if track_artist:
 183             track_info.update({
 184                 'artist': track_artist,
 185                 'title': f'{track_artist} - {track_title}',
 186             })
 187         else:
 188             track_info['title'] = track_title
 189
 190         return track_info
 191
 192
 193 class YandexMusicPlaylistBaseIE(YandexMusicBaseIE):
 194     def _extract_tracks(self, source, item_id, url, tld):
 195         tracks = source['tracks']
 196         track_ids = [str(track_id) for track_id in source['trackIds']]
 197
 198         # tracks dictionary shipped with playlist.jsx API is limited to 150 tracks,
 199         # missing tracks should be retrieved manually.
 200         if len(tracks) < len(track_ids):
 201             present_track_ids = {
 202                 str(track['id'])
 203                 for track in tracks if track.get('id')}
 204             missing_track_ids = [
 205                 track_id for track_id in track_ids
 206                 if track_id not in present_track_ids]
 207             # Request missing tracks in chunks to avoid exceeding max HTTP header size,
 208             # see https://github.com/ytdl-org/youtube-dl/issues/27355
 209             _TRACKS_PER_CHUNK = 250
 210             for chunk_num in itertools.count(0):
 211                 start = chunk_num * _TRACKS_PER_CHUNK
 212                 end = start + _TRACKS_PER_CHUNK
 213                 missing_track_ids_req = missing_track_ids[start:end]
 214                 assert missing_track_ids_req
 215                 missing_tracks = self._call_api(
 216                     'track-entries', tld, url, item_id,
 217                     f'Downloading missing tracks JSON chunk {chunk_num + 1}', {
 218                         'entries': ','.join(missing_track_ids_req),
 219                         'lang': tld,
 220                         'external-domain': f'music.yandex.{tld}',
 221                         'overembed': 'false',
 222                         'strict': 'true',
 223                     })
 224                 if missing_tracks:
 225                     tracks.extend(missing_tracks)
 226                 if end >= len(missing_track_ids):
 227                     break
 228
 229         return tracks
 230
 231     def _build_playlist(self, tracks):
 232         entries = []
 233         for track in tracks:
 234             track_id = track.get('id') or track.get('realId')
 235             if not track_id:
 236                 continue
 237             albums = track.get('albums')
 238             if not albums or not isinstance(albums, list):
 239                 continue
 240             album = albums[0]
 241             if not isinstance(album, dict):
 242                 continue
 243             album_id = album.get('id')
 244             if not album_id:
 245                 continue
 246             entries.append(self.url_result(
 247                 f'http://music.yandex.ru/album/{album_id}/track/{track_id}',
 248                 ie=YandexMusicTrackIE.ie_key(), video_id=track_id))
 249         return entries
 250
 251
 252 class YandexMusicAlbumIE(YandexMusicPlaylistBaseIE):
 253     IE_NAME = 'yandexmusic:album'
 254     IE_DESC = 'Яндекс.Музыка - Альбом'
 255     _VALID_URL = rf'{YandexMusicBaseIE._VALID_URL_BASE}/album/(?P<id>\d+)'
 256
 257     _TESTS = [{
 258         'url': 'http://music.yandex.ru/album/540508',
 259         'info_dict': {
 260             'id': '540508',
 261             'title': 'md5:7ed1c3567f28d14be9f61179116f5571',
 262         },
 263         'playlist_count': 50,
 264         # 'skip': 'Travis CI servers blocked by YandexMusic',
 265     }, {
 266         'url': 'https://music.yandex.ru/album/3840501',
 267         'info_dict': {
 268             'id': '3840501',
 269             'title': 'md5:36733472cdaa7dcb1fd9473f7da8e50f',
 270         },
 271         'playlist_count': 33,
 272         # 'skip': 'Travis CI servers blocked by YandexMusic',
 273     }, {
 274         # empty artists
 275         'url': 'https://music.yandex.ru/album/9091882',
 276         'info_dict': {
 277             'id': '9091882',
 278             'title': 'ТЕД на русском',
 279         },
 280         'playlist_count': 187,
 281     }]
 282
 283     @classmethod
 284     def suitable(cls, url):
 285         return False if YandexMusicTrackIE.suitable(url) else super().suitable(url)
 286
 287     def _real_extract(self, url):
 288         mobj = self._match_valid_url(url)
 289         tld = mobj.group('tld')
 290         album_id = mobj.group('id')
 291
 292         album = self._call_api(
 293             'album', tld, url, album_id, 'Downloading album JSON',
 294             {'album': album_id})
 295
 296         entries = self._build_playlist([track for volume in album['volumes'] for track in volume])
 297
 298         title = album['title']
 299         artist = try_get(album, lambda x: x['artists'][0]['name'], str)
 300         if artist:
 301             title = f'{artist} - {title}'
 302         year = album.get('year')
 303         if year:
 304             title += f' ({year})'
 305
 306         return self.playlist_result(entries, str(album['id']), title)
 307
 308
 309 class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
 310     IE_NAME = 'yandexmusic:playlist'
 311     IE_DESC = 'Яндекс.Музыка - Плейлист'
 312     _VALID_URL = rf'{YandexMusicBaseIE._VALID_URL_BASE}/users/(?P<user>[^/]+)/playlists/(?P<id>\d+)'
 313
 314     _TESTS = [{
 315         'url': 'http://music.yandex.ru/users/music.partners/playlists/1245',
 316         'info_dict': {
 317             'id': '1245',
 318             'title': 'md5:841559b3fe2b998eca88d0d2e22a3097',
 319             'description': 'md5:3b9f27b0efbe53f2ee1e844d07155cc9',
 320         },
 321         'playlist_count': 5,
 322         # 'skip': 'Travis CI servers blocked by YandexMusic',
 323     }, {
 324         'url': 'https://music.yandex.ru/users/ya.playlist/playlists/1036',
 325         'only_matching': True,
 326     }, {
 327         # playlist exceeding the limit of 150 tracks (see
 328         # https://github.com/ytdl-org/youtube-dl/issues/6666)
 329         'url': 'https://music.yandex.ru/users/mesiaz/playlists/1364',
 330         'info_dict': {
 331             'id': '1364',
 332             'title': 'md5:b3b400f997d3f878a13ae0699653f7db',
 333         },
 334         'playlist_mincount': 437,
 335         # 'skip': 'Travis CI servers blocked by YandexMusic',
 336     }]
 337
 338     def _real_extract(self, url):
 339         mobj = self._match_valid_url(url)
 340         tld = mobj.group('tld')
 341         user = mobj.group('user')
 342         playlist_id = mobj.group('id')
 343
 344         playlist = self._call_api(
 345             'playlist', tld, url, playlist_id, 'Downloading playlist JSON', {
 346                 'owner': user,
 347                 'kinds': playlist_id,
 348                 'light': 'true',
 349                 'lang': tld,
 350                 'external-domain': f'music.yandex.{tld}',
 351                 'overembed': 'false',
 352             })['playlist']
 353
 354         tracks = self._extract_tracks(playlist, playlist_id, url, tld)
 355
 356         return self.playlist_result(
 357             self._build_playlist(tracks),
 358             str(playlist_id),
 359             playlist.get('title'), playlist.get('description'))
 360
 361
 362 class YandexMusicArtistBaseIE(YandexMusicPlaylistBaseIE):
 363     def _call_artist(self, tld, url, artist_id):
 364         return self._call_api(
 365             'artist', tld, url, artist_id,
 366             f'Downloading artist {self._ARTIST_WHAT} JSON', {
 367                 'artist': artist_id,
 368                 'what': self._ARTIST_WHAT,
 369                 'sort': self._ARTIST_SORT or '',
 370                 'dir': '',
 371                 'period': '',
 372                 'lang': tld,
 373                 'external-domain': f'music.yandex.{tld}',
 374                 'overembed': 'false',
 375             })
 376
 377     def _real_extract(self, url):
 378         mobj = self._match_valid_url(url)
 379         tld = mobj.group('tld')
 380         artist_id = mobj.group('id')
 381         data = self._call_artist(tld, url, artist_id)
 382         tracks = self._extract_tracks(data, artist_id, url, tld)
 383         title = try_get(data, lambda x: x['artist']['name'], str)
 384         return self.playlist_result(
 385             self._build_playlist(tracks), artist_id, title)
 386
 387
 388 class YandexMusicArtistTracksIE(YandexMusicArtistBaseIE):
 389     IE_NAME = 'yandexmusic:artist:tracks'
 390     IE_DESC = 'Яндекс.Музыка - Артист - Треки'
 391     _VALID_URL = rf'{YandexMusicBaseIE._VALID_URL_BASE}/artist/(?P<id>\d+)/tracks'
 392
 393     _TESTS = [{
 394         'url': 'https://music.yandex.ru/artist/617526/tracks',
 395         'info_dict': {
 396             'id': '617526',
 397             'title': 'md5:131aef29d45fd5a965ca613e708c040b',
 398         },
 399         'playlist_count': 507,
 400         # 'skip': 'Travis CI servers blocked by YandexMusic',
 401     }]
 402
 403     _ARTIST_SORT = ''
 404     _ARTIST_WHAT = 'tracks'
 405
 406     def _real_extract(self, url):
 407         mobj = self._match_valid_url(url)
 408         tld = mobj.group('tld')
 409         artist_id = mobj.group('id')
 410         data = self._call_artist(tld, url, artist_id)
 411         tracks = self._extract_tracks(data, artist_id, url, tld)
 412         artist = try_get(data, lambda x: x['artist']['name'], str)
 413         title = '{} - {}'.format(artist or artist_id, 'Треки')
 414         return self.playlist_result(
 415             self._build_playlist(tracks), artist_id, title)
 416
 417
 418 class YandexMusicArtistAlbumsIE(YandexMusicArtistBaseIE):
 419     IE_NAME = 'yandexmusic:artist:albums'
 420     IE_DESC = 'Яндекс.Музыка - Артист - Альбомы'
 421     _VALID_URL = rf'{YandexMusicBaseIE._VALID_URL_BASE}/artist/(?P<id>\d+)/albums'
 422
 423     _TESTS = [{
 424         'url': 'https://music.yandex.ru/artist/617526/albums',
 425         'info_dict': {
 426             'id': '617526',
 427             'title': 'md5:55dc58d5c85699b7fb41ee926700236c',
 428         },
 429         'playlist_count': 8,
 430         # 'skip': 'Travis CI servers blocked by YandexMusic',
 431     }]
 432
 433     _ARTIST_SORT = 'year'
 434     _ARTIST_WHAT = 'albums'
 435
 436     def _real_extract(self, url):
 437         mobj = self._match_valid_url(url)
 438         tld = mobj.group('tld')
 439         artist_id = mobj.group('id')
 440         data = self._call_artist(tld, url, artist_id)
 441         entries = []
 442         for album in data['albums']:
 443             if not isinstance(album, dict):
 444                 continue
 445             album_id = album.get('id')
 446             if not album_id:
 447                 continue
 448             entries.append(self.url_result(
 449                 f'http://music.yandex.ru/album/{album_id}',
 450                 ie=YandexMusicAlbumIE.ie_key(), video_id=album_id))
 451         artist = try_get(data, lambda x: x['artist']['name'], str)
 452         title = '{} - {}'.format(artist or artist_id, 'Альбомы')
 453         return self.playlist_result(entries, artist_id, title)