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