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