Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/vevo.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import xml.etree.ElementTree
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_urllib_request,
   9 )
  10 from ..utils import (
  11     ExtractorError,
  12     int_or_none,
  13 )
  14
  15
  16 class VevoIE(InfoExtractor):
  17     """
  18     Accepts urls from vevo.com or in the format 'vevo:{id}'
  19     (currently used by MTVIE and MySpaceIE)
  20     """
  21     _VALID_URL = r'''(?x)
  22         (?:https?://www\.vevo\.com/watch/(?:[^/]+/(?:[^/]+/)?)?|
  23            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
  24            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
  25            vevo:)
  26         (?P<id>[^&?#]+)'''
  27
  28     _TESTS = [{
  29         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  30         "md5": "95ee28ee45e70130e3ab02b0f579ae23",
  31         'info_dict': {
  32             'id': 'GB1101300280',
  33             'ext': 'mp4',
  34             "upload_date": "20130624",
  35             "uploader": "Hurts",
  36             "title": "Somebody to Die For",
  37             "duration": 230.12,
  38             "width": 1920,
  39             "height": 1080,
  40             # timestamp and upload_date are often incorrect; seem to change randomly
  41             'timestamp': int,
  42         }
  43     }, {
  44         'note': 'v3 SMIL format',
  45         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
  46         'md5': 'f6ab09b034f8c22969020b042e5ac7fc',
  47         'info_dict': {
  48             'id': 'USUV71302923',
  49             'ext': 'mp4',
  50             'upload_date': '20140219',
  51             'uploader': 'Cassadee Pope',
  52             'title': 'I Wish I Could Break Your Heart',
  53             'duration': 226.101,
  54             'age_limit': 0,
  55             'timestamp': int,
  56         }
  57     }, {
  58         'note': 'Age-limited video',
  59         'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
  60         'info_dict': {
  61             'id': 'USRV81300282',
  62             'ext': 'mp4',
  63             'age_limit': 18,
  64             'title': 'Tunnel Vision (Explicit)',
  65             'uploader': 'Justin Timberlake',
  66             'upload_date': 're:2013070[34]',
  67             'timestamp': int,
  68         },
  69         'params': {
  70             'skip_download': 'true',
  71         }
  72     }]
  73     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  74
  75     def _real_initialize(self):
  76         req = compat_urllib_request.Request(
  77             'http://www.vevo.com/auth', data=b'')
  78         webpage = self._download_webpage(
  79             req, None,
  80             note='Retrieving oauth token',
  81             errnote='Unable to retrieve oauth token',
  82             fatal=False)
  83         if webpage is False:
  84             self._oauth_token = None
  85         else:
  86             self._oauth_token = self._search_regex(
  87                 r'access_token":\s*"([^"]+)"',
  88                 webpage, 'access token', fatal=False)
  89
  90     def _formats_from_json(self, video_info):
  91         last_version = {'version': -1}
  92         for version in video_info['videoVersions']:
  93             # These are the HTTP downloads, other types are for different manifests
  94             if version['sourceType'] == 2:
  95                 if version['version'] > last_version['version']:
  96                     last_version = version
  97         if last_version['version'] == -1:
  98             raise ExtractorError('Unable to extract last version of the video')
  99
 100         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
 101         formats = []
 102         # Already sorted from worst to best quality
 103         for rend in renditions.findall('rendition'):
 104             attr = rend.attrib
 105             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
 106             formats.append({
 107                 'url': attr['url'],
 108                 'format_id': attr['name'],
 109                 'format_note': format_note,
 110                 'height': int(attr['frameheight']),
 111                 'width': int(attr['frameWidth']),
 112             })
 113         return formats
 114
 115     def _formats_from_smil(self, smil_xml):
 116         formats = []
 117         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
 118         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
 119         for el in els:
 120             src = el.attrib['src']
 121             m = re.match(r'''(?xi)
 122                 (?P<ext>[a-z0-9]+):
 123                 (?P<path>
 124                     [/a-z0-9]+     # The directory and main part of the URL
 125                     _(?P<cbr>[0-9]+)k
 126                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 127                     _(?P<vcodec>[a-z0-9]+)
 128                     _(?P<vbr>[0-9]+)
 129                     _(?P<acodec>[a-z0-9]+)
 130                     _(?P<abr>[0-9]+)
 131                     \.[a-z0-9]+  # File extension
 132                 )''', src)
 133             if not m:
 134                 continue
 135
 136             format_url = self._SMIL_BASE_URL + m.group('path')
 137             formats.append({
 138                 'url': format_url,
 139                 'format_id': 'SMIL_' + m.group('cbr'),
 140                 'vcodec': m.group('vcodec'),
 141                 'acodec': m.group('acodec'),
 142                 'vbr': int(m.group('vbr')),
 143                 'abr': int(m.group('abr')),
 144                 'ext': m.group('ext'),
 145                 'width': int(m.group('width')),
 146                 'height': int(m.group('height')),
 147             })
 148         return formats
 149
 150     def _download_api_formats(self, video_id):
 151         if not self._oauth_token:
 152             self._downloader.report_warning(
 153                 'No oauth token available, skipping API HLS download')
 154             return []
 155
 156         api_url = 'https://apiv2.vevo.com/video/%s/streams/hls?token=%s' % (
 157             video_id, self._oauth_token)
 158         api_data = self._download_json(
 159             api_url, video_id,
 160             note='Downloading HLS formats',
 161             errnote='Failed to download HLS format list', fatal=False)
 162         if api_data is None:
 163             return []
 164
 165         m3u8_url = api_data[0]['url']
 166         return self._extract_m3u8_formats(
 167             m3u8_url, video_id, entry_protocol='m3u8_native', ext='mp4',
 168             preference=0)
 169
 170     def _real_extract(self, url):
 171         mobj = re.match(self._VALID_URL, url)
 172         video_id = mobj.group('id')
 173
 174         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 175         response = self._download_json(json_url, video_id)
 176         video_info = response['video']
 177
 178         if not video_info:
 179             if 'statusMessage' in response:
 180                 raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusMessage']), expected=True)
 181             raise ExtractorError('Unable to extract videos')
 182
 183         formats = self._formats_from_json(video_info)
 184
 185         is_explicit = video_info.get('isExplicit')
 186         if is_explicit is True:
 187             age_limit = 18
 188         elif is_explicit is False:
 189             age_limit = 0
 190         else:
 191             age_limit = None
 192
 193         # Download via HLS API
 194         formats.extend(self._download_api_formats(video_id))
 195
 196         # Download SMIL
 197         smil_blocks = sorted((
 198             f for f in video_info['videoVersions']
 199             if f['sourceType'] == 13),
 200             key=lambda f: f['version'])
 201         smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 202             self._SMIL_BASE_URL, video_id, video_id.lower())
 203         if smil_blocks:
 204             smil_url_m = self._search_regex(
 205                 r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
 206                 default=None)
 207             if smil_url_m is not None:
 208                 smil_url = smil_url_m
 209         if smil_url:
 210             smil_xml = self._download_webpage(
 211                 smil_url, video_id, 'Downloading SMIL info', fatal=False)
 212             if smil_xml:
 213                 formats.extend(self._formats_from_smil(smil_xml))
 214
 215         self._sort_formats(formats)
 216         timestamp_ms = int_or_none(self._search_regex(
 217             r'/Date\((\d+)\)/',
 218             video_info['launchDate'], 'launch date', fatal=False))
 219
 220         return {
 221             'id': video_id,
 222             'title': video_info['title'],
 223             'formats': formats,
 224             'thumbnail': video_info['imageUrl'],
 225             'timestamp': timestamp_ms // 1000,
 226             'uploader': video_info['mainArtists'][0]['artistName'],
 227             'duration': video_info['duration'],
 228             'age_limit': age_limit,
 229         }