_ Git - youtube-dl/blob - youtube_dl/extractor/vevo.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4 import xml.etree.ElementTree
   5 import datetime
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     compat_HTTPError,
  10     ExtractorError,
  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)
  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     _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         }
  38     }, {
  39         'note': 'v3 SMIL format',
  40         'url': 'http://www.vevo.com/watch/cassadee-pope/i-wish-i-could-break-your-heart/USUV71302923',
  41         'md5': '893ec0e0d4426a1d96c01de8f2bdff58',
  42         'info_dict': {
  43             'id': 'USUV71302923',
  44             'ext': 'mp4',
  45             'upload_date': '20140219',
  46             'uploader': 'Cassadee Pope',
  47             'title': 'I Wish I Could Break Your Heart',
  48             'duration': 226.101,
  49         }
  50     }]
  51     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  52
  53     def _formats_from_json(self, video_info):
  54         last_version = {'version': -1}
  55         for version in video_info['videoVersions']:
  56             # These are the HTTP downloads, other types are for different manifests
  57             if version['sourceType'] == 2:
  58                 if version['version'] > last_version['version']:
  59                     last_version = version
  60         if last_version['version'] == -1:
  61             raise ExtractorError('Unable to extract last version of the video')
  62
  63         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
  64         formats = []
  65         # Already sorted from worst to best quality
  66         for rend in renditions.findall('rendition'):
  67             attr = rend.attrib
  68             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
  69             formats.append({
  70                 'url': attr['url'],
  71                 'format_id': attr['name'],
  72                 'format_note': format_note,
  73                 'height': int(attr['frameheight']),
  74                 'width': int(attr['frameWidth']),
  75             })
  76         return formats
  77
  78     def _formats_from_smil(self, smil_xml):
  79         formats = []
  80         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
  81         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
  82         for el in els:
  83             src = el.attrib['src']
  84             m = re.match(r'''(?xi)
  85                 (?P<ext>[a-z0-9]+):
  86                 (?P<path>
  87                     [/a-z0-9]+     # The directory and main part of the URL
  88                     _(?P<cbr>[0-9]+)k
  89                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
  90                     _(?P<vcodec>[a-z0-9]+)
  91                     _(?P<vbr>[0-9]+)
  92                     _(?P<acodec>[a-z0-9]+)
  93                     _(?P<abr>[0-9]+)
  94                     \.[a-z0-9]+  # File extension
  95                 )''', src)
  96             if not m:
  97                 continue
  98
  99             format_url = self._SMIL_BASE_URL + m.group('path')
 100             formats.append({
 101                 'url': format_url,
 102                 'format_id': 'SMIL_' + m.group('cbr'),
 103                 'vcodec': m.group('vcodec'),
 104                 'acodec': m.group('acodec'),
 105                 'vbr': int(m.group('vbr')),
 106                 'abr': int(m.group('abr')),
 107                 'ext': m.group('ext'),
 108                 'width': int(m.group('width')),
 109                 'height': int(m.group('height')),
 110             })
 111         return formats
 112
 113     def _real_extract(self, url):
 114         mobj = re.match(self._VALID_URL, url)
 115         video_id = mobj.group('id')
 116
 117         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 118         video_info = self._download_json(json_url, video_id)['video']
 119
 120         formats = self._formats_from_json(video_info)
 121
 122         # Download SMIL
 123         smil_blocks = sorted((
 124             f for f in video_info['videoVersions']
 125             if f['sourceType'] == 13),
 126             key=lambda f: f['version'])
 127
 128         smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 129             self._SMIL_BASE_URL, video_id, video_id.lower())
 130         if smil_blocks:
 131             smil_url_m = self._search_regex(
 132                 r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
 133                 fatal=False)
 134             if smil_url_m is not None:
 135                 smil_url = smil_url_m
 136
 137         try:
 138             smil_xml = self._download_webpage(smil_url, video_id,
 139                                               'Downloading SMIL info')
 140             formats.extend(self._formats_from_smil(smil_xml))
 141         except ExtractorError as ee:
 142             if not isinstance(ee.cause, compat_HTTPError):
 143                 raise
 144             self._downloader.report_warning(
 145                 'Cannot download SMIL information, falling back to JSON ..')
 146
 147         timestamp_ms = int(self._search_regex(
 148             r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
 149         upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
 150         return {
 151             'id': video_id,
 152             'title': video_info['title'],
 153             'formats': formats,
 154             'thumbnail': video_info['imageUrl'],
 155             'upload_date': upload_date.strftime('%Y%m%d'),
 156             'uploader': video_info['mainArtists'][0]['artistName'],
 157             'duration': video_info['duration'],
 158         }