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

   1 # encoding: utf-8
   2
   3 import re
   4 import json
   5 import xml.etree.ElementTree
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     compat_urllib_parse,
  10     find_xpath_attr,
  11     compat_urlparse,
  12     compat_str,
  13     compat_urllib_request,
  14
  15     ExtractorError,
  16 )
  17
  18
  19 class BrightcoveIE(InfoExtractor):
  20     _VALID_URL = r'https?://.*brightcove\.com/(services|viewer).*\?(?P<query>.*)'
  21     _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
  22     _PLAYLIST_URL_TEMPLATE = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%s'
  23
  24     _TESTS = [
  25         {
  26             # From http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/
  27             u'url': u'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1654948606001&flashID=myExperience&%40videoPlayer=2371591881001',
  28             u'file': u'2371591881001.mp4',
  29             u'md5': u'8eccab865181d29ec2958f32a6a754f5',
  30             u'note': u'Test Brightcove downloads and detection in GenericIE',
  31             u'info_dict': {
  32                 u'title': u'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
  33                 u'uploader': u'8TV',
  34                 u'description': u'md5:a950cc4285c43e44d763d036710cd9cd',
  35             }
  36         },
  37         {
  38             # From http://medianetwork.oracle.com/video/player/1785452137001
  39             u'url': u'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1217746023001&flashID=myPlayer&%40videoPlayer=1785452137001',
  40             u'file': u'1785452137001.flv',
  41             u'info_dict': {
  42                 u'title': u'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges',
  43                 u'description': u'John Rose speaks at the JVM Language Summit, August 1, 2012.',
  44                 u'uploader': u'Oracle',
  45             },
  46         },
  47         {
  48             # From http://mashable.com/2013/10/26/thermoelectric-bracelet-lets-you-control-your-body-temperature/
  49             u'url': u'http://c.brightcove.com/services/viewer/federated_f9?&playerID=1265504713001&publisherID=AQ%7E%7E%2CAAABBzUwv1E%7E%2CxP-xFHVUstiMFlNYfvF4G9yFnNaqCw_9&videoID=2750934548001',
  50             u'info_dict': {
  51                 u'id': u'2750934548001',
  52                 u'ext': u'mp4',
  53                 u'title': u'This Bracelet Acts as a Personal Thermostat',
  54                 u'description': u'md5:547b78c64f4112766ccf4e151c20b6a0',
  55                 u'uploader': u'Mashable',
  56             },
  57         },
  58     ]
  59
  60     @classmethod
  61     def _build_brighcove_url(cls, object_str):
  62         """
  63         Build a Brightcove url from a xml string containing
  64         <object class="BrightcoveExperience">{params}</object>
  65         """
  66
  67         # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
  68         object_str = re.sub(r'(<param name="[^"]+" value="[^"]+")>',
  69                             lambda m: m.group(1) + '/>', object_str)
  70         # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608
  71         object_str = object_str.replace(u'<--', u'<!--')
  72
  73         object_doc = xml.etree.ElementTree.fromstring(object_str)
  74         assert u'BrightcoveExperience' in object_doc.attrib['class']
  75         params = {'flashID': object_doc.attrib['id'],
  76                   'playerID': find_xpath_attr(object_doc, './param', 'name', 'playerID').attrib['value'],
  77                   }
  78         def find_param(name):
  79             node = find_xpath_attr(object_doc, './param', 'name', name)
  80             if node is not None:
  81                 return node.attrib['value']
  82             return None
  83         playerKey = find_param('playerKey')
  84         # Not all pages define this value
  85         if playerKey is not None:
  86             params['playerKey'] = playerKey
  87         # The three fields hold the id of the video
  88         videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID')
  89         if videoPlayer is not None:
  90             params['@videoPlayer'] = videoPlayer
  91         linkBase = find_param('linkBaseURL')
  92         if linkBase is not None:
  93             params['linkBaseURL'] = linkBase
  94         data = compat_urllib_parse.urlencode(params)
  95         return cls._FEDERATED_URL_TEMPLATE % data
  96
  97     @classmethod
  98     def _extract_brightcove_url(cls, webpage):
  99         """Try to extract the brightcove url from the wepbage, returns None
 100         if it can't be found
 101         """
 102         m_brightcove = re.search(
 103             r'<object[^>]+?class=([\'"])[^>]*?BrightcoveExperience.*?\1.+?</object>',
 104             webpage, re.DOTALL)
 105         if m_brightcove is not None:
 106             return cls._build_brighcove_url(m_brightcove.group())
 107         else:
 108             return None
 109
 110     def _real_extract(self, url):
 111         # Change the 'videoId' and others field to '@videoPlayer'
 112         url = re.sub(r'(?<=[?&])(videoI(d|D)|bctid)', '%40videoPlayer', url)
 113         # Change bckey (used by bcove.me urls) to playerKey
 114         url = re.sub(r'(?<=[?&])bckey', 'playerKey', url)
 115         mobj = re.match(self._VALID_URL, url)
 116         query_str = mobj.group('query')
 117         query = compat_urlparse.parse_qs(query_str)
 118
 119         videoPlayer = query.get('@videoPlayer')
 120         if videoPlayer:
 121             return self._get_video_info(videoPlayer[0], query_str, query)
 122         else:
 123             player_key = query['playerKey']
 124             return self._get_playlist_info(player_key[0])
 125
 126     def _get_video_info(self, video_id, query_str, query):
 127         request_url = self._FEDERATED_URL_TEMPLATE % query_str
 128         req = compat_urllib_request.Request(request_url)
 129         linkBase = query.get('linkBaseURL')
 130         if linkBase is not None:
 131             req.add_header('Referer', linkBase[0])
 132         webpage = self._download_webpage(req, video_id)
 133
 134         self.report_extraction(video_id)
 135         info = self._search_regex(r'var experienceJSON = ({.*?});', webpage, 'json')
 136         info = json.loads(info)['data']
 137         video_info = info['programmedContent']['videoPlayer']['mediaDTO']
 138
 139         return self._extract_video_info(video_info)
 140
 141     def _get_playlist_info(self, player_key):
 142         playlist_info = self._download_webpage(self._PLAYLIST_URL_TEMPLATE % player_key,
 143                                                player_key, u'Downloading playlist information')
 144
 145         json_data = json.loads(playlist_info)
 146         if 'videoList' not in json_data:
 147             raise ExtractorError(u'Empty playlist')
 148         playlist_info = json_data['videoList']
 149         videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
 150
 151         return self.playlist_result(videos, playlist_id=playlist_info['id'],
 152                                     playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
 153
 154     def _extract_video_info(self, video_info):
 155         info = {
 156             'id': compat_str(video_info['id']),
 157             'title': video_info['displayName'],
 158             'description': video_info.get('shortDescription'),
 159             'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
 160             'uploader': video_info.get('publisherName'),
 161         }
 162
 163         renditions = video_info.get('renditions')
 164         if renditions:
 165             renditions = sorted(renditions, key=lambda r: r['size'])
 166             info['formats'] = [{
 167                 'url': rend['defaultURL'],
 168                 'height': rend.get('frameHeight'),
 169                 'width': rend.get('frameWidth'),
 170             } for rend in renditions]
 171         elif video_info.get('FLVFullLengthURL') is not None:
 172             info.update({
 173                 'url': video_info['FLVFullLengthURL'],
 174             })
 175         else:
 176             raise ExtractorError(u'Unable to extract video url for %s' % info['id'])
 177         return info