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 )
  15
  16
  17 class VevoIE(InfoExtractor):
  18     """
  19     Accepts urls from vevo.com or in the format 'vevo:{id}'
  20     (currently used by MTVIE and MySpaceIE)
  21     """
  22     _VALID_URL = r'''(?x)
  23         (?:https?://www\.vevo\.com/watch/(?:[^/]+/(?:[^/]+/)?)?|
  24            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
  25            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
  26            vevo:)
  27         (?P<id>[^&?#]+)'''
  28
  29     _TESTS = [{
  30         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  31         "md5": "95ee28ee45e70130e3ab02b0f579ae23",
  32         'info_dict': {
  33             'id': 'GB1101300280',
  34             'ext': 'mp4',
  35             "upload_date": "20130624",
  36             "uploader": "Hurts",
  37             "title": "Somebody to Die For",
  38             "duration": 230.12,
  39             "width": 1920,
  40             "height": 1080,
  41             # timestamp and upload_date are often incorrect; seem to change randomly
  42             'timestamp': int,
  43         }
  44     }, {
  45         'note': 'v3 SMIL format',
  46         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
  47         'md5': 'f6ab09b034f8c22969020b042e5ac7fc',
  48         'info_dict': {
  49             'id': 'USUV71302923',
  50             'ext': 'mp4',
  51             'upload_date': '20140219',
  52             'uploader': 'Cassadee Pope',
  53             'title': 'I Wish I Could Break Your Heart',
  54             'duration': 226.101,
  55             'age_limit': 0,
  56             'timestamp': int,
  57         }
  58     }, {
  59         'note': 'Age-limited video',
  60         'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
  61         'info_dict': {
  62             'id': 'USRV81300282',
  63             'ext': 'mp4',
  64             'age_limit': 18,
  65             'title': 'Tunnel Vision (Explicit)',
  66             'uploader': 'Justin Timberlake',
  67             'upload_date': 're:2013070[34]',
  68             'timestamp': int,
  69         },
  70         'params': {
  71             'skip_download': 'true',
  72         }
  73     }, {
  74         'note': 'No video_info',
  75         'url': 'http://www.vevo.com/watch/k-camp-1/Till-I-Die/USUV71503000',
  76         'md5': '8b83cc492d72fc9cf74a02acee7dc1b0',
  77         'info_dict': {
  78             'id': 'USUV71503000',
  79             'ext': 'mp4',
  80             'title': 'Till I Die - K Camp ft. T.I.',
  81             'duration': 193,
  82         },
  83         'expected_warnings': ['Unable to download SMIL file'],
  84     }]
  85     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  86
  87     def _real_initialize(self):
  88         req = sanitized_Request(
  89             'http://www.vevo.com/auth', data=b'')
  90         webpage = self._download_webpage(
  91             req, None,
  92             note='Retrieving oauth token',
  93             errnote='Unable to retrieve oauth token',
  94             fatal=False)
  95         if webpage is False:
  96             self._oauth_token = None
  97         else:
  98             if 'THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION' in webpage:
  99                 raise ExtractorError('%s said: This page is currently unavailable in your region.' % self.IE_NAME, expected=True)
 100
 101             self._oauth_token = self._search_regex(
 102                 r'access_token":\s*"([^"]+)"',
 103                 webpage, 'access token', fatal=False)
 104
 105     def _formats_from_json(self, video_info):
 106         if not video_info:
 107             return []
 108
 109         last_version = {'version': -1}
 110         for version in video_info['videoVersions']:
 111             # These are the HTTP downloads, other types are for different manifests
 112             if version['sourceType'] == 2:
 113                 if version['version'] > last_version['version']:
 114                     last_version = version
 115         if last_version['version'] == -1:
 116             raise ExtractorError('Unable to extract last version of the video')
 117
 118         renditions = compat_etree_fromstring(last_version['data'])
 119         formats = []
 120         # Already sorted from worst to best quality
 121         for rend in renditions.findall('rendition'):
 122             attr = rend.attrib
 123             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
 124             formats.append({
 125                 'url': attr['url'],
 126                 'format_id': attr['name'],
 127                 'format_note': format_note,
 128                 'height': int(attr['frameheight']),
 129                 'width': int(attr['frameWidth']),
 130             })
 131         return formats
 132
 133     def _formats_from_smil(self, smil_doc):
 134         formats = []
 135         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
 136         for el in els:
 137             src = el.attrib['src']
 138             m = re.match(r'''(?xi)
 139                 (?P<ext>[a-z0-9]+):
 140                 (?P<path>
 141                     [/a-z0-9]+     # The directory and main part of the URL
 142                     _(?P<cbr>[0-9]+)k
 143                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 144                     _(?P<vcodec>[a-z0-9]+)
 145                     _(?P<vbr>[0-9]+)
 146                     _(?P<acodec>[a-z0-9]+)
 147                     _(?P<abr>[0-9]+)
 148                     \.[a-z0-9]+  # File extension
 149                 )''', src)
 150             if not m:
 151                 continue
 152
 153             format_url = self._SMIL_BASE_URL + m.group('path')
 154             formats.append({
 155                 'url': format_url,
 156                 'format_id': 'SMIL_' + m.group('cbr'),
 157                 'vcodec': m.group('vcodec'),
 158                 'acodec': m.group('acodec'),
 159                 'vbr': int(m.group('vbr')),
 160                 'abr': int(m.group('abr')),
 161                 'ext': m.group('ext'),
 162                 'width': int(m.group('width')),
 163                 'height': int(m.group('height')),
 164             })
 165         return formats
 166
 167     def _download_api_formats(self, video_id, video_url):
 168         if not self._oauth_token:
 169             self._downloader.report_warning(
 170                 'No oauth token available, skipping API HLS download')
 171             return []
 172
 173         api_url = compat_urlparse.urljoin(video_url, '//apiv2.vevo.com/video/%s/streams/hls?token=%s' % (
 174             video_id, self._oauth_token))
 175         api_data = self._download_json(
 176             api_url, video_id,
 177             note='Downloading HLS formats',
 178             errnote='Failed to download HLS format list', fatal=False)
 179         if api_data is None:
 180             return []
 181
 182         m3u8_url = api_data[0]['url']
 183         return self._extract_m3u8_formats(
 184             m3u8_url, video_id, entry_protocol='m3u8_native', ext='mp4',
 185             preference=0)
 186
 187     def _real_extract(self, url):
 188         video_id = self._match_id(url)
 189
 190         webpage = None
 191
 192         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 193         response = self._download_json(json_url, video_id)
 194         video_info = response['video'] or {}
 195
 196         if not video_info and response.get('statusCode') != 909:
 197             if 'statusMessage' in response:
 198                 raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusMessage']), expected=True)
 199             raise ExtractorError('Unable to extract videos')
 200
 201         if not video_info:
 202             if url.startswith('vevo:'):
 203                 raise ExtractorError('Please specify full Vevo URL for downloading', expected=True)
 204             webpage = self._download_webpage(url, video_id)
 205
 206         title = video_info.get('title') or self._og_search_title(webpage)
 207
 208         formats = self._formats_from_json(video_info)
 209
 210         is_explicit = video_info.get('isExplicit')
 211         if is_explicit is True:
 212             age_limit = 18
 213         elif is_explicit is False:
 214             age_limit = 0
 215         else:
 216             age_limit = None
 217
 218         # Download via HLS API
 219         formats.extend(self._download_api_formats(video_id, url))
 220
 221         # Download SMIL
 222         smil_blocks = sorted((
 223             f for f in video_info.get('videoVersions', [])
 224             if f['sourceType'] == 13),
 225             key=lambda f: f['version'])
 226         smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 227             self._SMIL_BASE_URL, video_id, video_id.lower())
 228         if smil_blocks:
 229             smil_url_m = self._search_regex(
 230                 r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
 231                 default=None)
 232             if smil_url_m is not None:
 233                 smil_url = smil_url_m
 234         if smil_url:
 235             smil_doc = self._download_smil(smil_url, video_id, fatal=False)
 236             if smil_doc:
 237                 formats.extend(self._formats_from_smil(smil_doc))
 238
 239         self._sort_formats(formats)
 240         timestamp = int_or_none(self._search_regex(
 241             r'/Date\((\d+)\)/',
 242             video_info['launchDate'], 'launch date', fatal=False),
 243             scale=1000) if video_info else None
 244
 245         duration = video_info.get('duration') or int_or_none(
 246             self._html_search_meta('video:duration', webpage))
 247
 248         return {
 249             'id': video_id,
 250             'title': title,
 251             'formats': formats,
 252             'thumbnail': video_info.get('imageUrl'),
 253             'timestamp': timestamp,
 254             'uploader': video_info['mainArtists'][0]['artistName'] if video_info else None,
 255             'duration': duration,
 256             'age_limit': age_limit,
 257         }