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 ..utils import (
   8     compat_HTTPError,
   9     ExtractorError,
  10 )
  11
  12
  13 class VevoIE(InfoExtractor):
  14     """
  15     Accepts urls from vevo.com or in the format 'vevo:{id}'
  16     (currently used by MTVIE)
  17     """
  18     _VALID_URL = r'''(?x)
  19         (?:https?://www\.vevo\.com/watch/(?:[^/]+/(?:[^/]+/)?)?|
  20            https?://cache\.vevo\.com/m/html/embed\.html\?video=|
  21            https?://videoplayer\.vevo\.com/embed/embedded\?videoId=|
  22            vevo:)
  23         (?P<id>[^&?#]+)'''
  24
  25     _TESTS = [{
  26         'url': 'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  27         "md5": "06bea460acb744eab74a9d7dcb4bfd61",
  28         'info_dict': {
  29             'id': 'GB1101300280',
  30             'ext': 'mp4',
  31             "upload_date": "20130624",
  32             "uploader": "Hurts",
  33             "title": "Somebody to Die For",
  34             "duration": 230.12,
  35             "width": 1920,
  36             "height": 1080,
  37             # timestamp and upload_date are often incorrect; seem to change randomly
  38             'timestamp': int,
  39         }
  40     }, {
  41         'note': 'v3 SMIL format',
  42         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
  43         'md5': '893ec0e0d4426a1d96c01de8f2bdff58',
  44         'info_dict': {
  45             'id': 'USUV71302923',
  46             'ext': 'mp4',
  47             'upload_date': '20140219',
  48             'uploader': 'Cassadee Pope',
  49             'title': 'I Wish I Could Break Your Heart',
  50             'duration': 226.101,
  51             'age_limit': 0,
  52             'timestamp': int,
  53         }
  54     }, {
  55         'note': 'Age-limited video',
  56         'url': 'https://www.vevo.com/watch/justin-timberlake/tunnel-vision-explicit/USRV81300282',
  57         'info_dict': {
  58             'id': 'USRV81300282',
  59             'ext': 'mp4',
  60             'age_limit': 18,
  61             'title': 'Tunnel Vision (Explicit)',
  62             'uploader': 'Justin Timberlake',
  63             'upload_date': 're:2013070[34]',
  64             'timestamp': int,
  65         },
  66         'params': {
  67             'skip_download': 'true',
  68         }
  69     }]
  70     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  71
  72     def _formats_from_json(self, video_info):
  73         last_version = {'version': -1}
  74         for version in video_info['videoVersions']:
  75             # These are the HTTP downloads, other types are for different manifests
  76             if version['sourceType'] == 2:
  77                 if version['version'] > last_version['version']:
  78                     last_version = version
  79         if last_version['version'] == -1:
  80             raise ExtractorError('Unable to extract last version of the video')
  81
  82         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
  83         formats = []
  84         # Already sorted from worst to best quality
  85         for rend in renditions.findall('rendition'):
  86             attr = rend.attrib
  87             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
  88             formats.append({
  89                 'url': attr['url'],
  90                 'format_id': attr['name'],
  91                 'format_note': format_note,
  92                 'height': int(attr['frameheight']),
  93                 'width': int(attr['frameWidth']),
  94             })
  95         return formats
  96
  97     def _formats_from_smil(self, smil_xml):
  98         formats = []
  99         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
 100         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
 101         for el in els:
 102             src = el.attrib['src']
 103             m = re.match(r'''(?xi)
 104                 (?P<ext>[a-z0-9]+):
 105                 (?P<path>
 106                     [/a-z0-9]+     # The directory and main part of the URL
 107                     _(?P<cbr>[0-9]+)k
 108                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
 109                     _(?P<vcodec>[a-z0-9]+)
 110                     _(?P<vbr>[0-9]+)
 111                     _(?P<acodec>[a-z0-9]+)
 112                     _(?P<abr>[0-9]+)
 113                     \.[a-z0-9]+  # File extension
 114                 )''', src)
 115             if not m:
 116                 continue
 117
 118             format_url = self._SMIL_BASE_URL + m.group('path')
 119             formats.append({
 120                 'url': format_url,
 121                 'format_id': 'SMIL_' + m.group('cbr'),
 122                 'vcodec': m.group('vcodec'),
 123                 'acodec': m.group('acodec'),
 124                 'vbr': int(m.group('vbr')),
 125                 'abr': int(m.group('abr')),
 126                 'ext': m.group('ext'),
 127                 'width': int(m.group('width')),
 128                 'height': int(m.group('height')),
 129             })
 130         return formats
 131
 132     def _real_extract(self, url):
 133         mobj = re.match(self._VALID_URL, url)
 134         video_id = mobj.group('id')
 135
 136         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 137         response = self._download_json(json_url, video_id)
 138         video_info = response['video']
 139
 140         if not video_info:
 141             if 'statusMessage' in response:
 142                 raise ExtractorError('%s said: %s' % (self.IE_NAME, response['statusMessage']), expected=True)
 143             raise ExtractorError('Unable to extract videos')
 144
 145         formats = self._formats_from_json(video_info)
 146
 147         is_explicit = video_info.get('isExplicit')
 148         if is_explicit is True:
 149             age_limit = 18
 150         elif is_explicit is False:
 151             age_limit = 0
 152         else:
 153             age_limit = None
 154
 155         # Download SMIL
 156         smil_blocks = sorted((
 157             f for f in video_info['videoVersions']
 158             if f['sourceType'] == 13),
 159             key=lambda f: f['version'])
 160
 161         smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 162             self._SMIL_BASE_URL, video_id, video_id.lower())
 163         if smil_blocks:
 164             smil_url_m = self._search_regex(
 165                 r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
 166                 fatal=False)
 167             if smil_url_m is not None:
 168                 smil_url = smil_url_m
 169
 170         try:
 171             smil_xml = self._download_webpage(smil_url, video_id,
 172                                               'Downloading SMIL info')
 173             formats.extend(self._formats_from_smil(smil_xml))
 174         except ExtractorError as ee:
 175             if not isinstance(ee.cause, compat_HTTPError):
 176                 raise
 177             self._downloader.report_warning(
 178                 'Cannot download SMIL information, falling back to JSON ..')
 179
 180         self._sort_formats(formats)
 181         timestamp_ms = int(self._search_regex(
 182             r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
 183
 184         return {
 185             'id': video_id,
 186             'title': video_info['title'],
 187             'formats': formats,
 188             'thumbnail': video_info['imageUrl'],
 189             'timestamp': timestamp_ms // 1000,
 190             'uploader': video_info['mainArtists'][0]['artistName'],
 191             'duration': video_info['duration'],
 192             'age_limit': age_limit,
 193         }