yt_dlp/extractor/yandexmusic.py

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