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