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