_ 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         'note': 'Only available via webpage',
  89         'url': 'http://www.vevo.com/watch/GBUV71600656',
  90         'md5': '67e79210613865b66a47c33baa5e37fe',
  91         'info_dict': {
  92             'id': 'GBUV71600656',
  93             'ext': 'mp4',
  94             'title': 'Viva Love',
  95             'upload_date': '20160428',
  96             'age_limit': 0,
  97             'uploader': 'ABC',
  98             'timestamp': 1461830400,
  99         },
 100         'expected_warnings': ['Failed to download video versions info'],
 101     }]
 102     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com'
 103     _SOURCE_TYPES = {
 104         0: 'youtube',
 105         1: 'brightcove',
 106         2: 'http',
 107         3: 'hls_ios',
 108         4: 'hls',
 109         5: 'smil',  # http
 110         7: 'f4m_cc',
 111         8: 'f4m_ak',
 112         9: 'f4m_l3',
 113         10: 'ism',
 114         13: 'smil',  # rtmp
 115         18: 'dash',
 116     }
 117     _VERSIONS = {
 118         0: 'youtube',  # only in AuthenticateVideo videoVersions
 119         1: 'level3',
 120         2: 'akamai',
 121         3: 'level3',
 122         4: 'amazon',
 123     }
 124
 125     def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
 126         formats = []
 127         els = smil.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
 128         for el in els:
 129             src = el.attrib['src']
 130             m = re.match(r'''(?xi)
 131                 (?P<ext>[a-z0-9]+):
 132                 (?P<path>
 133                     [/a-z0-9]+     # The directory and main part of the URL
 134                     _(?P<tbr>[0-9]+)k
 135                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 136                     _(?P<vcodec>[a-z0-9]+)
 137                     _(?P<vbr>[0-9]+)
 138                     _(?P<acodec>[a-z0-9]+)
 139                     _(?P<abr>[0-9]+)
 140                     \.[a-z0-9]+  # File extension
 141                 )''', src)
 142             if not m:
 143                 continue
 144
 145             format_url = self._SMIL_BASE_URL + m.group('path')
 146             formats.append({
 147                 'url': format_url,
 148                 'format_id': 'smil_' + m.group('tbr'),
 149                 'vcodec': m.group('vcodec'),
 150                 'acodec': m.group('acodec'),
 151                 'tbr': int(m.group('tbr')),
 152                 'vbr': int(m.group('vbr')),
 153                 'abr': int(m.group('abr')),
 154                 'ext': m.group('ext'),
 155                 'width': int(m.group('width')),
 156                 'height': int(m.group('height')),
 157             })
 158         return formats
 159
 160     def _initialize_api(self, video_id):
 161         req = sanitized_Request(
 162             'http://www.vevo.com/auth', data=b'')
 163         webpage = self._download_webpage(
 164             req, None,
 165             note='Retrieving oauth token',
 166             errnote='Unable to retrieve oauth token')
 167
 168         if 'THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION' in webpage:
 169             raise ExtractorError(
 170                 '%s said: This page is currently unavailable in your region.' % self.IE_NAME, expected=True)
 171
 172         auth_info = self._parse_json(webpage, video_id)
 173         self._api_url_template = self.http_scheme() + '//apiv2.vevo.com/%s?token=' + auth_info['access_token']
 174
 175     def _call_api(self, path, *args, **kwargs):
 176         return self._download_json(self._api_url_template % path, *args, **kwargs)
 177
 178     def _real_extract(self, url):
 179         video_id = self._match_id(url)
 180
 181         json_url = 'http://api.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 182         response = self._download_json(
 183             json_url, video_id, 'Downloading video info', 'Unable to download info')
 184         video_info = response.get('video') or {}
 185         video_versions = video_info.get('videoVersions')
 186         uploader = None
 187         timestamp = None
 188         view_count = None
 189         formats = []
 190
 191         if not video_info:
 192             if response.get('statusCode') != 909:
 193                 ytid = response.get('errorInfo', {}).get('ytid')
 194                 if ytid:
 195                     self.report_warning(
 196                         'Video is geoblocked, trying with the YouTube video %s' % ytid)
 197                     return self.url_result(ytid, 'Youtube', ytid)
 198
 199                 if 'statusMessage' in response:
 200                     raise ExtractorError('%s said: %s' % (
 201                         self.IE_NAME, response['statusMessage']), expected=True)
 202                 raise ExtractorError('Unable to extract videos')
 203
 204             self._initialize_api(video_id)
 205             video_info = self._call_api(
 206                 'video/%s' % video_id, video_id, 'Downloading api video info',
 207                 'Failed to download video info')
 208
 209             video_versions = self._call_api(
 210                 'video/%s/streams' % video_id, video_id,
 211                 'Downloading video versions info',
 212                 'Failed to download video versions info',
 213                 fatal=False)
 214
 215             # Some videos are only available via webpage (e.g.
 216             # https://github.com/rg3/youtube-dl/issues/9366)
 217             if not video_versions:
 218                 webpage = self._download_webpage(url, video_id)
 219                 video_versions = self._extract_json(webpage, video_id, 'streams')[video_id][0]
 220
 221             timestamp = parse_iso8601(video_info.get('releaseDate'))
 222             artists = video_info.get('artists')
 223             if artists:
 224                 uploader = artists[0]['name']
 225             view_count = int_or_none(video_info.get('views', {}).get('total'))
 226
 227             for video_version in video_versions:
 228                 version = self._VERSIONS.get(video_version['version'])
 229                 version_url = video_version.get('url')
 230                 if not version_url:
 231                     continue
 232
 233                 if '.ism' in version_url:
 234                     continue
 235                 elif '.mpd' in version_url:
 236                     formats.extend(self._extract_mpd_formats(
 237                         version_url, video_id, mpd_id='dash-%s' % version,
 238                         note='Downloading %s MPD information' % version,
 239                         errnote='Failed to download %s MPD information' % version,
 240                         fatal=False))
 241                 elif '.m3u8' in version_url:
 242                     formats.extend(self._extract_m3u8_formats(
 243                         version_url, video_id, 'mp4', 'm3u8_native',
 244                         m3u8_id='hls-%s' % version,
 245                         note='Downloading %s m3u8 information' % version,
 246                         errnote='Failed to download %s m3u8 information' % version,
 247                         fatal=False))
 248                 else:
 249                     m = re.search(r'''(?xi)
 250                         _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 251                         _(?P<vcodec>[a-z0-9]+)
 252                         _(?P<vbr>[0-9]+)
 253                         _(?P<acodec>[a-z0-9]+)
 254                         _(?P<abr>[0-9]+)
 255                         \.(?P<ext>[a-z0-9]+)''', version_url)
 256                     if not m:
 257                         continue
 258
 259                     formats.append({
 260                         'url': version_url,
 261                         'format_id': 'http-%s-%s' % (version, video_version['quality']),
 262                         'vcodec': m.group('vcodec'),
 263                         'acodec': m.group('acodec'),
 264                         'vbr': int(m.group('vbr')),
 265                         'abr': int(m.group('abr')),
 266                         'ext': m.group('ext'),
 267                         'width': int(m.group('width')),
 268                         'height': int(m.group('height')),
 269                     })
 270         else:
 271             timestamp = int_or_none(self._search_regex(
 272                 r'/Date\((\d+)\)/',
 273                 video_info['releaseDate'], 'release date', fatal=False),
 274                 scale=1000)
 275             artists = video_info.get('mainArtists')
 276             if artists:
 277                 uploader = artists[0]['artistName']
 278
 279             smil_parsed = False
 280             for video_version in video_info['videoVersions']:
 281                 version = self._VERSIONS.get(video_version['version'])
 282                 if version == 'youtube':
 283                     continue
 284                 else:
 285                     source_type = self._SOURCE_TYPES.get(video_version['sourceType'])
 286                     renditions = compat_etree_fromstring(video_version['data'])
 287                     if source_type == 'http':
 288                         for rend in renditions.findall('rendition'):
 289                             attr = rend.attrib
 290                             formats.append({
 291                                 'url': attr['url'],
 292                                 'format_id': 'http-%s-%s' % (version, attr['name']),
 293                                 'height': int_or_none(attr.get('frameheight')),
 294                                 'width': int_or_none(attr.get('frameWidth')),
 295                                 'tbr': int_or_none(attr.get('totalBitrate')),
 296                                 'vbr': int_or_none(attr.get('videoBitrate')),
 297                                 'abr': int_or_none(attr.get('audioBitrate')),
 298                                 'vcodec': attr.get('videoCodec'),
 299                                 'acodec': attr.get('audioCodec'),
 300                             })
 301                     elif source_type == 'hls':
 302                         formats.extend(self._extract_m3u8_formats(
 303                             renditions.find('rendition').attrib['url'], video_id,
 304                             'mp4', 'm3u8_native', m3u8_id='hls-%s' % version,
 305                             note='Downloading %s m3u8 information' % version,
 306                             errnote='Failed to download %s m3u8 information' % version,
 307                             fatal=False))
 308                     elif source_type == 'smil' and version == 'level3' and not smil_parsed:
 309                         formats.extend(self._extract_smil_formats(
 310                             renditions.find('rendition').attrib['url'], video_id, False))
 311                         smil_parsed = True
 312         self._sort_formats(formats)
 313
 314         title = video_info['title']
 315
 316         is_explicit = video_info.get('isExplicit')
 317         if is_explicit is True:
 318             age_limit = 18
 319         elif is_explicit is False:
 320             age_limit = 0
 321         else:
 322             age_limit = None
 323
 324         duration = video_info.get('duration')
 325
 326         return {
 327             'id': video_id,
 328             'title': title,
 329             'formats': formats,
 330             'thumbnail': video_info.get('imageUrl') or video_info.get('thumbnailUrl'),
 331             'timestamp': timestamp,
 332             'uploader': uploader,
 333             'duration': duration,
 334             'view_count': view_count,
 335             'age_limit': age_limit,
 336         }
 337
 338
 339 class VevoPlaylistIE(VevoBaseIE):
 340     _VALID_URL = r'https?://www\.vevo\.com/watch/(?P<kind>playlist|genre)/(?P<id>[^/?#&]+)'
 341
 342     _TESTS = [{
 343         'url': 'http://www.vevo.com/watch/playlist/dadbf4e7-b99f-4184-9670-6f0e547b6a29',
 344         'info_dict': {
 345             'id': 'dadbf4e7-b99f-4184-9670-6f0e547b6a29',
 346             'title': 'Best-Of: Birdman',
 347         },
 348         'playlist_count': 10,
 349     }, {
 350         'url': 'http://www.vevo.com/watch/genre/rock',
 351         'info_dict': {
 352             'id': 'rock',
 353             'title': 'Rock',
 354         },
 355         'playlist_count': 20,
 356     }, {
 357         'url': 'http://www.vevo.com/watch/playlist/dadbf4e7-b99f-4184-9670-6f0e547b6a29?index=0',
 358         'md5': '32dcdfddddf9ec6917fc88ca26d36282',
 359         'info_dict': {
 360             'id': 'USCMV1100073',
 361             'ext': 'mp4',
 362             'title': 'Y.U. MAD',
 363             'timestamp': 1323417600,
 364             'upload_date': '20111209',
 365             'uploader': 'Birdman',
 366         },
 367         'expected_warnings': ['Unable to download SMIL file'],
 368     }, {
 369         'url': 'http://www.vevo.com/watch/genre/rock?index=0',
 370         'only_matching': True,
 371     }]
 372
 373     def _real_extract(self, url):
 374         mobj = re.match(self._VALID_URL, url)
 375         playlist_id = mobj.group('id')
 376         playlist_kind = mobj.group('kind')
 377
 378         webpage = self._download_webpage(url, playlist_id)
 379
 380         qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 381         index = qs.get('index', [None])[0]
 382
 383         if index:
 384             video_id = self._search_regex(
 385                 r'<meta[^>]+content=(["\'])vevo://video/(?P<id>.+?)\1[^>]*>',
 386                 webpage, 'video id', default=None, group='id')
 387             if video_id:
 388                 return self.url_result('vevo:%s' % video_id, VevoIE.ie_key())
 389
 390         playlists = self._extract_json(webpage, playlist_id, '%ss' % playlist_kind)
 391
 392         playlist = (list(playlists.values())[0]
 393                     if playlist_kind == 'playlist' else playlists[playlist_id])
 394
 395         entries = [
 396             self.url_result('vevo:%s' % src, VevoIE.ie_key())
 397             for src in playlist['isrcs']]
 398
 399         return self.playlist_result(
 400             entries, playlist.get('playlistId'),
 401             playlist.get('name'), playlist.get('description'))