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

   1 import re
   2 import json
   3 import xml.etree.ElementTree
   4 import datetime
   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     _TESTS = [{
  25         u'url': u'http://www.vevo.com/watch/hurts/somebody-to-die-for/GB1101300280',
  26         u'file': u'GB1101300280.mp4',
  27         u"md5": u"06bea460acb744eab74a9d7dcb4bfd61",
  28         u'info_dict': {
  29             u"upload_date": u"20130624",
  30             u"uploader": u"Hurts",
  31             u"title": u"Somebody to Die For",
  32             u"duration": 230.12,
  33             u"width": 1920,
  34             u"height": 1080,
  35         }
  36     }]
  37     _SMIL_BASE_URL = 'http://smil.lvl3.vevo.com/'
  38
  39     def _formats_from_json(self, video_info):
  40         last_version = {'version': -1}
  41         for version in video_info['videoVersions']:
  42             # These are the HTTP downloads, other types are for different manifests
  43             if version['sourceType'] == 2:
  44                 if version['version'] > last_version['version']:
  45                     last_version = version
  46         if last_version['version'] == -1:
  47             raise ExtractorError(u'Unable to extract last version of the video')
  48
  49         renditions = xml.etree.ElementTree.fromstring(last_version['data'])
  50         formats = []
  51         # Already sorted from worst to best quality
  52         for rend in renditions.findall('rendition'):
  53             attr = rend.attrib
  54             format_note = '%(videoCodec)s@%(videoBitrate)4sk, %(audioCodec)s@%(audioBitrate)3sk' % attr
  55             formats.append({
  56                 'url': attr['url'],
  57                 'format_id': attr['name'],
  58                 'format_note': format_note,
  59                 'height': int(attr['frameheight']),
  60                 'width': int(attr['frameWidth']),
  61             })
  62         return formats
  63
  64     def _formats_from_smil(self, smil_xml):
  65         formats = []
  66         smil_doc = xml.etree.ElementTree.fromstring(smil_xml.encode('utf-8'))
  67         els = smil_doc.findall('.//{http://www.w3.org/2001/SMIL20/Language}video')
  68         for el in els:
  69             src = el.attrib['src']
  70             m = re.match(r'''(?xi)
  71                 (?P<ext>[a-z0-9]+):
  72                 (?P<path>
  73                     [/a-z0-9]+     # The directory and main part of the URL
  74                     _(?P<cbr>[0-9]+)k
  75                     _(?P<width>[0-9]+)x(?P<height>[0-9]+)
  76                     _(?P<vcodec>[a-z0-9]+)
  77                     _(?P<vbr>[0-9]+)
  78                     _(?P<acodec>[a-z0-9]+)
  79                     _(?P<abr>[0-9]+)
  80                     \.[a-z0-9]+  # File extension
  81                 )''', src)
  82             if not m:
  83                 continue
  84
  85             format_url = self._SMIL_BASE_URL + m.group('path')
  86             formats.append({
  87                 'url': format_url,
  88                 'format_id': u'SMIL_' + m.group('cbr'),
  89                 'vcodec': m.group('vcodec'),
  90                 'acodec': m.group('acodec'),
  91                 'vbr': int(m.group('vbr')),
  92                 'abr': int(m.group('abr')),
  93                 'ext': m.group('ext'),
  94                 'width': int(m.group('width')),
  95                 'height': int(m.group('height')),
  96             })
  97         return formats
  98
  99     def _real_extract(self, url):
 100         mobj = re.match(self._VALID_URL, url)
 101         video_id = mobj.group('id')
 102
 103         json_url = 'http://videoplayer.vevo.com/VideoService/AuthenticateVideo?isrc=%s' % video_id
 104         info_json = self._download_webpage(json_url, video_id, u'Downloading json info')
 105         video_info = json.loads(info_json)['video']
 106
 107         formats = self._formats_from_json(video_info)
 108         try:
 109             smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
 110                 self._SMIL_BASE_URL, video_id, video_id.lower())
 111             smil_xml = self._download_webpage(smil_url, video_id,
 112                                               u'Downloading SMIL info')
 113             formats.extend(self._formats_from_smil(smil_xml))
 114         except ExtractorError as ee:
 115             if not isinstance(ee.cause, compat_HTTPError):
 116                 raise
 117             self._downloader.report_warning(
 118                 u'Cannot download SMIL information, falling back to JSON ..')
 119
 120         timestamp_ms = int(self._search_regex(
 121             r'/Date\((\d+)\)/', video_info['launchDate'], u'launch date'))
 122         upload_date = datetime.datetime.fromtimestamp(timestamp_ms // 1000)
 123         info = {
 124             'id': video_id,
 125             'title': video_info['title'],
 126             'formats': formats,
 127             'thumbnail': video_info['imageUrl'],
 128             'upload_date': upload_date.strftime('%Y%m%d'),
 129             'uploader': video_info['mainArtists'][0]['artistName'],
 130             'duration': video_info['duration'],
 131         }
 132
 133         return info