youtube_dlc/extractor/yandexmusic.py

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