youtube_dl/extractor/yandexmusic.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import hashlib
   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 )
  14
  15
  16 class YandexMusicBaseIE(InfoExtractor):
  17     @staticmethod
  18     def _handle_error(response):
  19         if isinstance(response, dict):
  20             error = response.get('error')
  21             if error:
  22                 raise ExtractorError(error, expected=True)
  23             if response.get('type') == 'captcha' or 'captcha' in response:
  24                 YandexMusicBaseIE._raise_captcha()
  25
  26     @staticmethod
  27     def _raise_captcha():
  28         raise ExtractorError(
  29             'YandexMusic has considered youtube-dl requests automated and '
  30             'asks you to solve a CAPTCHA. You can either wait for some '
  31             'time until unblocked and optionally use --sleep-interval '
  32             'in future or alternatively you can go to https://music.yandex.ru/ '
  33             'solve CAPTCHA, then export cookies and pass cookie file to '
  34             'youtube-dl with --cookies',
  35             expected=True)
  36
  37     def _download_webpage_handle(self, *args, **kwargs):
  38         webpage = super(YandexMusicBaseIE, self)._download_webpage_handle(*args, **kwargs)
  39         if 'Нам очень жаль, но&nbsp;запросы, поступившие с&nbsp;вашего IP-адреса, похожи на&nbsp;автоматические.' in webpage:
  40             self._raise_captcha()
  41         return webpage
  42
  43     def _download_json(self, *args, **kwargs):
  44         response = super(YandexMusicBaseIE, self)._download_json(*args, **kwargs)
  45         self._handle_error(response)
  46         return response
  47
  48
  49 class YandexMusicTrackIE(YandexMusicBaseIE):
  50     IE_NAME = 'yandexmusic:track'
  51     IE_DESC = 'Яндекс.Музыка - Трек'
  52     _VALID_URL = r'https?://music\.yandex\.(?:ru|kz|ua|by)/album/(?P<album_id>\d+)/track/(?P<id>\d+)'
  53
  54     _TEST = {
  55         'url': 'http://music.yandex.ru/album/540508/track/4878838',
  56         'md5': 'f496818aa2f60b6c0062980d2e00dc20',
  57         'info_dict': {
  58             'id': '4878838',
  59             'ext': 'mp3',
  60             'title': 'Carlo Ambrosio, Carlo Ambrosio & Fabio Di Bari - Gypsy Eyes 1',
  61             'filesize': 4628061,
  62             'duration': 193.04,
  63             'track': 'Gypsy Eyes 1',
  64             'album': 'Gypsy Soul',
  65             'album_artist': 'Carlo Ambrosio',
  66             'artist': 'Carlo Ambrosio, Carlo Ambrosio & Fabio Di Bari',
  67             'release_year': 2009,
  68         },
  69         'skip': 'Travis CI servers blocked by YandexMusic',
  70     }
  71
  72     def _get_track_url(self, storage_dir, track_id):
  73         data = self._download_json(
  74             'http://music.yandex.ru/api/v1.5/handlers/api-jsonp.jsx?action=getTrackSrc&p=download-info/%s'
  75             % storage_dir,
  76             track_id, 'Downloading track location JSON')
  77
  78         # Each string is now wrapped in a list, this is probably only temporarily thus
  79         # supporting both scenarios (see https://github.com/rg3/youtube-dl/issues/10193)
  80         for k, v in data.items():
  81             if v and isinstance(v, list):
  82                 data[k] = v[0]
  83
  84         key = hashlib.md5(('XGRlBW9FXlekgbPrRHuSiA' + data['path'][1:] + data['s']).encode('utf-8')).hexdigest()
  85         storage = storage_dir.split('.')
  86
  87         return ('http://%s/get-mp3/%s/%s?track-id=%s&from=service-10-track&similarities-experiment=default'
  88                 % (data['host'], key, data['ts'] + data['path'], storage[1]))
  89
  90     def _get_track_info(self, track):
  91         thumbnail = None
  92         cover_uri = track.get('albums', [{}])[0].get('coverUri')
  93         if cover_uri:
  94             thumbnail = cover_uri.replace('%%', 'orig')
  95             if not thumbnail.startswith('http'):
  96                 thumbnail = 'http://' + thumbnail
  97
  98         track_title = track['title']
  99         track_info = {
 100             'id': track['id'],
 101             'ext': 'mp3',
 102             'url': self._get_track_url(track['storageDir'], track['id']),
 103             'filesize': int_or_none(track.get('fileSize')),
 104             'duration': float_or_none(track.get('durationMs'), 1000),
 105             'thumbnail': thumbnail,
 106             'track': track_title,
 107         }
 108
 109         def extract_artist(artist_list):
 110             if artist_list and isinstance(artist_list, list):
 111                 artists_names = [a['name'] for a in artist_list if a.get('name')]
 112                 if artists_names:
 113                     return ', '.join(artists_names)
 114
 115         albums = track.get('albums')
 116         if albums and isinstance(albums, list):
 117             album = albums[0]
 118             if isinstance(album, dict):
 119                 year = album.get('year')
 120                 track_info.update({
 121                     'album': album.get('title'),
 122                     'album_artist': extract_artist(album.get('artists')),
 123                     'release_year': int_or_none(year),
 124                 })
 125
 126         track_artist = extract_artist(track.get('artists'))
 127         if track_artist:
 128             track_info.update({
 129                 'artist': track_artist,
 130                 'title': '%s - %s' % (track_artist, track_title),
 131             })
 132         else:
 133             track_info['title'] = track_title
 134         return track_info
 135
 136     def _real_extract(self, url):
 137         mobj = re.match(self._VALID_URL, url)
 138         album_id, track_id = mobj.group('album_id'), mobj.group('id')
 139
 140         track = self._download_json(
 141             'http://music.yandex.ru/handlers/track.jsx?track=%s:%s' % (track_id, album_id),
 142             track_id, 'Downloading track JSON')['track']
 143
 144         return self._get_track_info(track)
 145
 146
 147 class YandexMusicPlaylistBaseIE(YandexMusicBaseIE):
 148     def _build_playlist(self, tracks):
 149         return [
 150             self.url_result(
 151                 'http://music.yandex.ru/album/%s/track/%s' % (track['albums'][0]['id'], track['id']))
 152             for track in tracks if track.get('albums') and isinstance(track.get('albums'), list)]
 153
 154
 155 class YandexMusicAlbumIE(YandexMusicPlaylistBaseIE):
 156     IE_NAME = 'yandexmusic:album'
 157     IE_DESC = 'Яндекс.Музыка - Альбом'
 158     _VALID_URL = r'https?://music\.yandex\.(?:ru|kz|ua|by)/album/(?P<id>\d+)/?(\?|$)'
 159
 160     _TEST = {
 161         'url': 'http://music.yandex.ru/album/540508',
 162         'info_dict': {
 163             'id': '540508',
 164             'title': 'Carlo Ambrosio - Gypsy Soul (2009)',
 165         },
 166         'playlist_count': 50,
 167         'skip': 'Travis CI servers blocked by YandexMusic',
 168     }
 169
 170     def _real_extract(self, url):
 171         album_id = self._match_id(url)
 172
 173         album = self._download_json(
 174             'http://music.yandex.ru/handlers/album.jsx?album=%s' % album_id,
 175             album_id, 'Downloading album JSON')
 176
 177         entries = self._build_playlist(album['volumes'][0])
 178
 179         title = '%s - %s' % (album['artists'][0]['name'], album['title'])
 180         year = album.get('year')
 181         if year:
 182             title += ' (%s)' % year
 183
 184         return self.playlist_result(entries, compat_str(album['id']), title)
 185
 186
 187 class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
 188     IE_NAME = 'yandexmusic:playlist'
 189     IE_DESC = 'Яндекс.Музыка - Плейлист'
 190     _VALID_URL = r'https?://music\.yandex\.(?P<tld>ru|kz|ua|by)/users/(?P<user>[^/]+)/playlists/(?P<id>\d+)'
 191
 192     _TESTS = [{
 193         'url': 'http://music.yandex.ru/users/music.partners/playlists/1245',
 194         'info_dict': {
 195             'id': '1245',
 196             'title': 'Что слушают Enter Shikari',
 197             'description': 'md5:3b9f27b0efbe53f2ee1e844d07155cc9',
 198         },
 199         'playlist_count': 6,
 200         'skip': 'Travis CI servers blocked by YandexMusic',
 201     }, {
 202         # playlist exceeding the limit of 150 tracks shipped with webpage (see
 203         # https://github.com/rg3/youtube-dl/issues/6666)
 204         'url': 'https://music.yandex.ru/users/ya.playlist/playlists/1036',
 205         'info_dict': {
 206             'id': '1036',
 207             'title': 'Музыка 90-х',
 208         },
 209         'playlist_mincount': 300,
 210         'skip': 'Travis CI servers blocked by YandexMusic',
 211     }]
 212
 213     def _real_extract(self, url):
 214         mobj = re.match(self._VALID_URL, url)
 215         tld = mobj.group('tld')
 216         user = mobj.group('user')
 217         playlist_id = mobj.group('id')
 218
 219         playlist = self._download_json(
 220             'https://music.yandex.%s/handlers/playlist.jsx' % tld,
 221             playlist_id, 'Downloading missing tracks JSON',
 222             fatal=False,
 223             headers={
 224                 'Referer': url,
 225                 'X-Requested-With': 'XMLHttpRequest',
 226                 'X-Retpath-Y': url,
 227             },
 228             query={
 229                 'owner': user,
 230                 'kinds': playlist_id,
 231                 'light': 'true',
 232                 'lang': tld,
 233                 'external-domain': 'music.yandex.%s' % tld,
 234                 'overembed': 'false',
 235             })['playlist']
 236
 237         tracks = playlist['tracks']
 238         track_ids = [compat_str(track_id) for track_id in playlist['trackIds']]
 239
 240         # tracks dictionary shipped with playlist.jsx API is limited to 150 tracks,
 241         # missing tracks should be retrieved manually.
 242         if len(tracks) < len(track_ids):
 243             present_track_ids = set([
 244                 compat_str(track['id'])
 245                 for track in tracks if track.get('id')])
 246             missing_track_ids = [
 247                 track_id for track_id in track_ids
 248                 if track_id not in present_track_ids]
 249             missing_tracks = self._download_json(
 250                 'https://music.yandex.%s/handlers/track-entries.jsx' % tld,
 251                 playlist_id, 'Downloading missing tracks JSON',
 252                 fatal=False,
 253                 headers={
 254                     'Referer': url,
 255                     'X-Requested-With': 'XMLHttpRequest',
 256                 },
 257                 query={
 258                     'entries': ','.join(missing_track_ids),
 259                     'lang': tld,
 260                     'external-domain': 'music.yandex.%s' % tld,
 261                     'overembed': 'false',
 262                     'strict': 'true',
 263                 })
 264             if missing_tracks:
 265                 tracks.extend(missing_tracks)
 266
 267         return self.playlist_result(
 268             self._build_playlist(tracks),
 269             compat_str(playlist_id),
 270             playlist.get('title'), playlist.get('description'))