youtube_dl/extractor/vevo.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..compat import (
   7     compat_etree_fromstring,
   8     compat_str,
   9     compat_urlparse,
  10 )
  11 from ..utils import (
  12     ExtractorError,
  13     int_or_none,
  14     sanitized_Request,
  15     parse_iso8601,
  16 )
  17
  18
  19 class VevoBaseIE(InfoExtractor):
  20     def _extract_json(self, webpage, video_id, item):
  21         return self._parse_json(
  22             self._search_regex(
  23                 r'window\.__INITIAL_STORE__\s*=\s*({.+?});\s*</script>',
  24                 webpage, 'initial store'),
  25             video_id)['default'][item]
  26
  27
  28 class VevoIE(VevoBaseIE):
  29     '''
  30     Accepts urls from vevo.com or in the format 'vevo:{id}'
  31     (currently used by MTVIE and MySpaceIE)
  32     '''
  33     _VALID_URL = r'''(?x)
  34         (?:https?://www\.vevo\.com/watch/(?!playlist|genre)(?:[^/]+/(?:[^/]+/)?)?|
  35            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
  36            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
  37            vevo:)
  38         (?P<id>[^&?#]+)'''
  39
  40     _TESTS = [{
  41         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  42         'md5': '95ee28ee45e70130e3ab02b0f579ae23',
  43         'info_dict': {
  44             'id': 'GB1101300280',
  45             'ext': 'mp4',
  46             'title': 'Hurts - Somebody to Die For',
  47             'timestamp': 1372057200,
  48             'upload_date': '20130624',
  49             'uploader': 'Hurts',
  50             'track': 'Somebody to Die For',
  51             'artist': 'Hurts',
  52             'genre': 'Pop',
  53         },
  54         'expected_warnings': ['Unable to download SMIL file'],
  55     }, {
  56         'note': 'v3 SMIL format',
  57         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
  58         'md5': 'f6ab09b034f8c22969020b042e5ac7fc',
  59         'info_dict': {
  60             'id': 'USUV71302923',
  61             'ext': 'mp4',
  62             'title': 'Cassadee Pope - I Wish I Could Break Your Heart',
  63             'timestamp': 1392796919,
  64             'upload_date': '20140219',
  65             'uploader': 'Cassadee Pope',
  66             'track': 'I Wish I Could Break Your Heart',
  67             'artist': 'Cassadee Pope',
  68             'genre': 'Country',
  69         },
  70         'expected_warnings': ['Unable to download SMIL file'],
  71     }, {
  72         'note': 'Age-limited video',
  73         'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
  74         'info_dict': {
  75             'id': 'USRV81300282',
  76             'ext': 'mp4',
  77             'title': 'Justin Timberlake - Tunnel Vision (Explicit)',
  78             'age_limit': 18,
  79             'timestamp': 1372888800,
  80             'upload_date': '20130703',
  81             'uploader': 'Justin Timberlake',
  82             'track': 'Tunnel Vision (Explicit)',
  83             'artist': 'Justin Timberlake',
  84             'genre': 'Pop',
  85         },
  86         'expected_warnings': ['Unable to download SMIL file'],
  87     }, {
  88         'note': 'No video_info',
  89         'url': 'http://www.vevo.com/watch/k-camp-1/Till-I-Die/USUV71503000',
  90         'md5': '8b83cc492d72fc9cf74a02acee7dc1b0',
  91         'info_dict': {
  92             'id': 'USUV71503000',
  93             'ext': 'mp4',
  94             'title': 'K Camp - Till I Die',
  95             'age_limit': 18,
  96             'timestamp': 1449468000,
  97             'upload_date': '20151207',
  98             'uploader': 'K Camp',
  99             'track': 'Till I Die',
 100             'artist': 'K Camp',
 101             'genre': 'Rap/Hip-Hop',
 102         },
 103     }, {
 104         'note': 'Only available via webpage',
 105         'url': 'http://www.vevo.com/watch/GBUV71600656',
 106         'md5': '67e79210613865b66a47c33baa5e37fe',
 107         'info_dict': {
 108             'id': 'GBUV71600656',
 109             'ext': 'mp4',
 110             'title': 'ABC - Viva Love',
 111             'age_limit': 0,
 112             'timestamp': 1461830400,
 113             'upload_date': '20160428',
 114             'uploader': 'ABC',
 115             'track': 'Viva Love',
 116             'artist': 'ABC',
 117             'genre': 'Pop',
 118         },
 119         'expected_warnings': ['Failed to download video versions info'],
 120     }, {
 121         # no genres available
 122         'url': 'http://www.vevo.com/watch/INS171400764',
 123         'only_matching': True,
 124     }]
 125     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com'
 126     _SOURCE_TYPES = {
 127         0: 'youtube',
 128         1: 'brightcove',
 129         2: 'http',
 130         3: 'hls_ios',
 131         4: 'hls',
 132         5: 'smil',  # http
 133         7: 'f4m_cc',
 134         8: 'f4m_ak',
 135         9: 'f4m_l3',
 136         10: 'ism',
 137         13: 'smil',  # rtmp
 138         18: 'dash',
 139     }
 140     _VERSIONS = {
 141         0: 'youtube',  # only in AuthenticateVideo videoVersions
 142         1: 'level3',
 143         2: 'akamai',
 144         3: 'level3',
 145         4: 'amazon',
 146     }
 147
 148     def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
 149         formats = []
 150         els = smil.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
 151         for el in els:
 152             src = el.attrib['src']
 153             m = re.match(r'''(?xi)
 154                 (?P<ext>[a-z0-9]+):
 155                 (?P<path>
 156                     [/a-z0-9]+     # The directory and main part of the URL
 157                     _(?P<tbr>[0-9]+)k
 158                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 159                     _(?P<vcodec>[a-z0-9]+)
 160                     _(?P<vbr>[0-9]+)
 161                     _(?P<acodec>[a-z0-9]+)
 162                     _(?P<abr>[0-9]+)
 163                     \.[a-z0-9]+  # File extension
 164                 )''', src)
 165             if not m:
 166                 continue
 167
 168             format_url = self._SMIL_BASE_URL + m.group('path')
 169             formats.append({
 170                 'url': format_url,
 171                 'format_id': 'smil_' + m.group('tbr'),
 172                 'vcodec': m.group('vcodec'),
 173                 'acodec': m.group('acodec'),
 174                 'tbr': int(m.group('tbr')),
 175                 'vbr': int(m.group('vbr')),
 176                 'abr': int(m.group('abr')),
 177                 'ext': m.group('ext'),
 178                 'width': int(m.group('width')),
 179                 'height': int(m.group('height')),
 180             })
 181         return formats
 182
 183     def _initialize_api(self, video_id):
 184         req = sanitized_Request(
 185             'http://www.vevo.com/auth', data=b'')
 186         webpage = self._download_webpage(
 187             req, None,
 188             note='Retrieving oauth token',
 189             errnote='Unable to retrieve oauth token')
 190
 191         if 'THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION' in webpage:
 192             self.raise_geo_restricted(
 193                 '%s said: This page is currently unavailable in your region' % self.IE_NAME)
 194
 195         auth_info = self._parse_json(webpage, video_id)
 196         self._api_url_template = self.http_scheme() + '//apiv2.vevo.com/%s?token=' + auth_info['access_token']
 197
 198     def _call_api(self, path, *args, **kwargs):
 199         return self._download_json(self._api_url_template % path, *args, **kwargs)
 200
 201     def _real_extract(self, url):
 202         video_id = self._match_id(url)
 203
 204         json_url = 'http://api.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 205         response = self._download_json(
 206             json_url, video_id, 'Downloading video info',
 207             'Unable to download info', fatal=False) or {}
 208         video_info = response.get('video') or {}
 209         artist = None
 210         featured_artist = None
 211         uploader = None
 212         view_count = None
 213         formats = []
 214
 215         if not video_info:
 216             if response and response.get('statusCode') != 909:
 217                 ytid = response.get('errorInfo', {}).get('ytid')
 218                 if ytid:
 219                     self.report_warning(
 220                         'Video is geoblocked, trying with the YouTube video %s' % ytid)
 221                     return self.url_result(ytid, 'Youtube', ytid)
 222
 223                 if 'statusMessage' in response:
 224                     raise ExtractorError('%s said: %s' % (
 225                         self.IE_NAME, response['statusMessage']), expected=True)
 226                 raise ExtractorError('Unable to extract videos')
 227
 228             self._initialize_api(video_id)
 229             video_info = self._call_api(
 230                 'video/%s' % video_id, video_id, 'Downloading api video info',
 231                 'Failed to download video info')
 232
 233             video_versions = self._call_api(
 234                 'video/%s/streams' % video_id, video_id,
 235                 'Downloading video versions info',
 236                 'Failed to download video versions info',
 237                 fatal=False)
 238
 239             # Some videos are only available via webpage (e.g.
 240             # https://github.com/rg3/youtube-dl/issues/9366)
 241             if not video_versions:
 242                 webpage = self._download_webpage(url, video_id)
 243                 video_versions = self._extract_json(webpage, video_id, 'streams')[video_id][0]
 244
 245             timestamp = parse_iso8601(video_info.get('releaseDate'))
 246             artists = video_info.get('artists')
 247             if artists:
 248                 artist = uploader = artists[0]['name']
 249             view_count = int_or_none(video_info.get('views', {}).get('total'))
 250
 251             for video_version in video_versions:
 252                 version = self._VERSIONS.get(video_version['version'])
 253                 version_url = video_version.get('url')
 254                 if not version_url:
 255                     continue
 256
 257                 if '.ism' in version_url:
 258                     continue
 259                 elif '.mpd' in version_url:
 260                     formats.extend(self._extract_mpd_formats(
 261                         version_url, video_id, mpd_id='dash-%s' % version,
 262                         note='Downloading %s MPD information' % version,
 263                         errnote='Failed to download %s MPD information' % version,
 264                         fatal=False))
 265                 elif '.m3u8' in version_url:
 266                     formats.extend(self._extract_m3u8_formats(
 267                         version_url, video_id, 'mp4', 'm3u8_native',
 268                         m3u8_id='hls-%s' % version,
 269                         note='Downloading %s m3u8 information' % version,
 270                         errnote='Failed to download %s m3u8 information' % version,
 271                         fatal=False))
 272                 else:
 273                     m = re.search(r'''(?xi)
 274                         _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 275                         _(?P<vcodec>[a-z0-9]+)
 276                         _(?P<vbr>[0-9]+)
 277                         _(?P<acodec>[a-z0-9]+)
 278                         _(?P<abr>[0-9]+)
 279                         \.(?P<ext>[a-z0-9]+)''', version_url)
 280                     if not m:
 281                         continue
 282
 283                     formats.append({
 284                         'url': version_url,
 285                         'format_id': 'http-%s-%s' % (version, video_version['quality']),
 286                         'vcodec': m.group('vcodec'),
 287                         'acodec': m.group('acodec'),
 288                         'vbr': int(m.group('vbr')),
 289                         'abr': int(m.group('abr')),
 290                         'ext': m.group('ext'),
 291                         'width': int(m.group('width')),
 292                         'height': int(m.group('height')),
 293                     })
 294         else:
 295             timestamp = int_or_none(self._search_regex(
 296                 r'/Date\((\d+)\)/',
 297                 video_info['releaseDate'], 'release date', fatal=False),
 298                 scale=1000)
 299             artists = video_info.get('mainArtists')
 300             if artists:
 301                 artist = uploader = artists[0]['artistName']
 302
 303             featured_artists = video_info.get('featuredArtists')
 304             if featured_artists:
 305                 featured_artist = featured_artists[0]['artistName']
 306
 307             smil_parsed = False
 308             for video_version in video_info['videoVersions']:
 309                 version = self._VERSIONS.get(video_version['version'])
 310                 if version == 'youtube':
 311                     continue
 312                 else:
 313                     source_type = self._SOURCE_TYPES.get(video_version['sourceType'])
 314                     renditions = compat_etree_fromstring(video_version['data'])
 315                     if source_type == 'http':
 316                         for rend in renditions.findall('rendition'):
 317                             attr = rend.attrib
 318                             formats.append({
 319                                 'url': attr['url'],
 320                                 'format_id': 'http-%s-%s' % (version, attr['name']),
 321                                 'height': int_or_none(attr.get('frameheight')),
 322                                 'width': int_or_none(attr.get('frameWidth')),
 323                                 'tbr': int_or_none(attr.get('totalBitrate')),
 324                                 'vbr': int_or_none(attr.get('videoBitrate')),
 325                                 'abr': int_or_none(attr.get('audioBitrate')),
 326                                 'vcodec': attr.get('videoCodec'),
 327                                 'acodec': attr.get('audioCodec'),
 328                             })
 329                     elif source_type == 'hls':
 330                         formats.extend(self._extract_m3u8_formats(
 331                             renditions.find('rendition').attrib['url'], video_id,
 332                             'mp4', 'm3u8_native', m3u8_id='hls-%s' % version,
 333                             note='Downloading %s m3u8 information' % version,
 334                             errnote='Failed to download %s m3u8 information' % version,
 335                             fatal=False))
 336                     elif source_type == 'smil' and version == 'level3' and not smil_parsed:
 337                         formats.extend(self._extract_smil_formats(
 338                             renditions.find('rendition').attrib['url'], video_id, False))
 339                         smil_parsed = True
 340         self._sort_formats(formats)
 341
 342         track = video_info['title']
 343         if featured_artist:
 344             artist = '%s ft. %s' % (artist, featured_artist)
 345         title = '%s - %s' % (artist, track) if artist else track
 346
 347         genres = video_info.get('genres')
 348         genre = (
 349             genres[0] if genres and isinstance(genres, list) and
 350             isinstance(genres[0], compat_str) else None)
 351
 352         is_explicit = video_info.get('isExplicit')
 353         if is_explicit is True:
 354             age_limit = 18
 355         elif is_explicit is False:
 356             age_limit = 0
 357         else:
 358             age_limit = None
 359
 360         duration = video_info.get('duration')
 361
 362         return {
 363             'id': video_id,
 364             'title': title,
 365             'formats': formats,
 366             'thumbnail': video_info.get('imageUrl') or video_info.get('thumbnailUrl'),
 367             'timestamp': timestamp,
 368             'uploader': uploader,
 369             'duration': duration,
 370             'view_count': view_count,
 371             'age_limit': age_limit,
 372             'track': track,
 373             'artist': uploader,
 374             'genre': genre,
 375         }
 376
 377
 378 class VevoPlaylistIE(VevoBaseIE):
 379     _VALID_URL = r'https?://www\.vevo\.com/watch/(?P<kind>playlist|genre)/(?P<id>[^/?#&]+)'
 380
 381     _TESTS = [{
 382         'url': 'http://www.vevo.com/watch/playlist/dadbf4e7-b99f-4184-9670-6f0e547b6a29',
 383         'info_dict': {
 384             'id': 'dadbf4e7-b99f-4184-9670-6f0e547b6a29',
 385             'title': 'Best-Of: Birdman',
 386         },
 387         'playlist_count': 10,
 388     }, {
 389         'url': 'http://www.vevo.com/watch/genre/rock',
 390         'info_dict': {
 391             'id': 'rock',
 392             'title': 'Rock',
 393         },
 394         'playlist_count': 20,
 395     }, {
 396         'url': 'http://www.vevo.com/watch/playlist/dadbf4e7-b99f-4184-9670-6f0e547b6a29?index=0',
 397         'md5': '32dcdfddddf9ec6917fc88ca26d36282',
 398         'info_dict': {
 399             'id': 'USCMV1100073',
 400             'ext': 'mp4',
 401             'title': 'Birdman - Y.U. MAD',
 402             'timestamp': 1323417600,
 403             'upload_date': '20111209',
 404             'uploader': 'Birdman',
 405             'track': 'Y.U. MAD',
 406             'artist': 'Birdman',
 407             'genre': 'Rap/Hip-Hop',
 408         },
 409         'expected_warnings': ['Unable to download SMIL file'],
 410     }, {
 411         'url': 'http://www.vevo.com/watch/genre/rock?index=0',
 412         'only_matching': True,
 413     }]
 414
 415     def _real_extract(self, url):
 416         mobj = re.match(self._VALID_URL, url)
 417         playlist_id = mobj.group('id')
 418         playlist_kind = mobj.group('kind')
 419
 420         webpage = self._download_webpage(url, playlist_id)
 421
 422         qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 423         index = qs.get('index', [None])[0]
 424
 425         if index:
 426             video_id = self._search_regex(
 427                 r'<meta[^>]+content=(["\'])vevo://video/(?P<id>.+?)\1[^>]*>',
 428                 webpage, 'video id', default=None, group='id')
 429             if video_id:
 430                 return self.url_result('vevo:%s' % video_id, VevoIE.ie_key())
 431
 432         playlists = self._extract_json(webpage, playlist_id, '%ss' % playlist_kind)
 433
 434         playlist = (list(playlists.values())[0]
 435                     if playlist_kind == 'playlist' else playlists[playlist_id])
 436
 437         entries = [
 438             self.url_result('vevo:%s' % src, VevoIE.ie_key())
 439             for src in playlist['isrcs']]
 440
 441         return self.playlist_result(
 442             entries, playlist.get('playlistId') or playlist_id,
 443             playlist.get('name'), playlist.get('description'))