_ 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 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         'note': 'No video_info',
  72         'url': 'http://www.vevo.com/watch/k-camp-1/Till-I-Die/USUV71503000',
  73         'md5': '8b83cc492d72fc9cf74a02acee7dc1b0',
  74         'info_dict': {
  75             'id': 'USUV71503000',
  76             'ext': 'mp4',
  77             'title': 'Till I Die - K Camp ft. T.I.',
  78             'duration': 193,
  79         },
  80         'expected_warnings': ['HTTP Error 404'],
  81     }]
  82     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  83
  84     def _real_initialize(self):
  85         req = sanitized_Request(
  86             'http://www.vevo.com/auth', data=b'')
  87         webpage = self._download_webpage(
  88             req, None,
  89             note='Retrieving oauth token',
  90             errnote='Unable to retrieve oauth token',
  91             fatal=False)
  92         if webpage is False:
  93             self._oauth_token = None
  94         else:
  95             if 'THIS PAGE IS CURRENTLY UNAVAILABLE IN YOUR REGION' in webpage:
  96                 raise ExtractorError('%s said: This page is currently unavailable in your region.' % self.IE_NAME, expected=True)
  97
  98             self._oauth_token = self._search_regex(
  99                 r'access_token":\s*"([^"]+)"',
 100                 webpage, 'access token', fatal=False)
 101
 102     def _formats_from_json(self, video_info):
 103         if not video_info:
 104             return []
 105
 106         last_version = {'version': -1}
 107         for version in video_info['videoVersions']:
 108             # These are the HTTP downloads, other types are for different manifests
 109             if version['sourceType'] == 2:
 110                 if version['version'] > last_version['version']:
 111                     last_version = version
 112         if last_version['version'] == -1:
 113             raise ExtractorError('Unable to extract last version of the video')
 114
 115         renditions = compat_etree_fromstring(last_version['data'])
 116         formats = []
 117         # Already sorted from worst to best quality
 118         for rend in renditions.findall('rendition'):
 119             attr = rend.attrib
 120             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
 121             formats.append({
 122                 'url': attr['url'],
 123                 'format_id': attr['name'],
 124                 'format_note': format_note,
 125                 'height': int(attr['frameheight']),
 126                 'width': int(attr['frameWidth']),
 127             })
 128         return formats
 129
 130     def _formats_from_smil(self, smil_xml):
 131         formats = []
 132         smil_doc = compat_etree_fromstring(smil_xml.encode('utf-8'))
 133         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
 134         for el in els:
 135             src = el.attrib['src']
 136             m = re.match(r'''(?xi)
 137                 (?P<ext>[a-z0-9]+):
 138                 (?P<path>
 139                     [/a-z0-9]+     # The directory and main part of the URL
 140                     _(?P<cbr>[0-9]+)k
 141                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 142                     _(?P<vcodec>[a-z0-9]+)
 143                     _(?P<vbr>[0-9]+)
 144                     _(?P<acodec>[a-z0-9]+)
 145                     _(?P<abr>[0-9]+)
 146                     \.[a-z0-9]+  # File extension
 147                 )''', src)
 148             if not m:
 149                 continue
 150
 151             format_url = self._SMIL_BASE_URL + m.group('path')
 152             formats.append({
 153                 'url': format_url,
 154                 'format_id': 'SMIL_' + m.group('cbr'),
 155                 'vcodec': m.group('vcodec'),
 156                 'acodec': m.group('acodec'),
 157                 'vbr': int(m.group('vbr')),
 158                 'abr': int(m.group('abr')),
 159                 'ext': m.group('ext'),
 160                 'width': int(m.group('width')),
 161                 'height': int(m.group('height')),
 162             })
 163         return formats
 164
 165     def _download_api_formats(self, video_id):
 166         if not self._oauth_token:
 167             self._downloader.report_warning(
 168                 'No oauth token available, skipping API HLS download')
 169             return []
 170
 171         api_url = 'https://apiv2.vevo.com/video/%s/streams/hls?token=%s' % (
 172             video_id, self._oauth_token)
 173         api_data = self._download_json(
 174             api_url, video_id,
 175             note='Downloading HLS formats',
 176             errnote='Failed to download HLS format list', fatal=False)
 177         if api_data is None:
 178             return []
 179
 180         m3u8_url = api_data[0]['url']
 181         return self._extract_m3u8_formats(
 182             m3u8_url, video_id, entry_protocol='m3u8_native', ext='mp4',
 183             preference=0)
 184
 185     def _real_extract(self, url):
 186         video_id = self._match_id(url)
 187
 188         webpage = None
 189
 190         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 191         response = self._download_json(json_url, video_id)
 192         video_info = response['video'] or {}
 193
 194         if not video_info and response.get('statusCode') != 909:
 195             if 'statusMessage' in response:
 196                 raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusMessage']), expected=True)
 197             raise ExtractorError('Unable to extract videos')
 198
 199         if not video_info:
 200             if url.startswith('vevo:'):
 201                 raise ExtractorError('Please specify full Vevo URL for downloading', expected=True)
 202             webpage = self._download_webpage(url, video_id)
 203
 204         title = video_info.get('title') or self._og_search_title(webpage)
 205
 206         formats = self._formats_from_json(video_info)
 207
 208         is_explicit = video_info.get('isExplicit')
 209         if is_explicit is True:
 210             age_limit = 18
 211         elif is_explicit is False:
 212             age_limit = 0
 213         else:
 214             age_limit = None
 215
 216         # Download via HLS API
 217         formats.extend(self._download_api_formats(video_id))
 218
 219         # Download SMIL
 220         smil_blocks = sorted((
 221             f for f in video_info.get('videoVersions', [])
 222             if f['sourceType'] == 13),
 223             key=lambda f: f['version'])
 224         smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 225             self._SMIL_BASE_URL, video_id, video_id.lower())
 226         if smil_blocks:
 227             smil_url_m = self._search_regex(
 228                 r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
 229                 default=None)
 230             if smil_url_m is not None:
 231                 smil_url = smil_url_m
 232         if smil_url:
 233             smil_xml = self._download_webpage(
 234                 smil_url, video_id, 'Downloading SMIL info', fatal=False)
 235             if smil_xml:
 236                 formats.extend(self._formats_from_smil(smil_xml))
 237
 238         self._sort_formats(formats)
 239         timestamp = int_or_none(self._search_regex(
 240             r'/Date\((\d+)\)/',
 241             video_info['launchDate'], 'launch date', fatal=False),
 242             scale=1000) if video_info else None
 243
 244         duration = video_info.get('duration') or int_or_none(
 245             self._html_search_meta('video:duration', webpage))
 246
 247         return {
 248             'id': video_id,
 249             'title': title,
 250             'formats': formats,
 251             'thumbnail': video_info.get('imageUrl'),
 252             'timestamp': timestamp,
 253             'uploader': video_info['mainArtists'][0]['artistName'] if video_info else None,
 254             'duration': duration,
 255             'age_limit': age_limit,
 256         }