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

   1 import re
   2 import json
   3 import xml.etree.ElementTree
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     compat_urllib_parse,
   8     find_xpath_attr,
   9 )
  10
  11 class BrightcoveIE(InfoExtractor):
  12     _VALID_URL = r'https?://.*brightcove\.com/(services|viewer).*\?(?P<query>.*)'
  13     _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
  14     _PLAYLIST_URL_TEMPLATE = 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=%s'
  15
  16     # There is a test for Brigtcove in GenericIE, that way we test both the download
  17     # and the detection of videos, and we don't have to find an URL that is always valid
  18
  19     @classmethod
  20     def _build_brighcove_url(cls, object_str):
  21         """
  22         Build a Brightcove url from a xml string containing
  23         <object class="BrightcoveExperience">{params}</object>
  24         """
  25         object_doc = xml.etree.ElementTree.fromstring(object_str)
  26         assert u'BrightcoveExperience' in object_doc.attrib['class']
  27         params = {'flashID': object_doc.attrib['id'],
  28                   'playerID': find_xpath_attr(object_doc, './param', 'name', 'playerID').attrib['value'],
  29                   }
  30         playerKey = find_xpath_attr(object_doc, './param', 'name', 'playerKey')
  31         # Not all pages define this value
  32         if playerKey is not None:
  33             params['playerKey'] = playerKey.attrib['value']
  34         videoPlayer = find_xpath_attr(object_doc, './param', 'name', '@videoPlayer')
  35         if videoPlayer is not None:
  36             params['@videoPlayer'] = videoPlayer.attrib['value']
  37         data = compat_urllib_parse.urlencode(params)
  38         return cls._FEDERATED_URL_TEMPLATE % data
  39
  40     def _real_extract(self, url):
  41         mobj = re.match(self._VALID_URL, url)
  42         query = mobj.group('query')
  43
  44         m_video_id = re.search(r'videoPlayer=(\d+)', query)
  45         if m_video_id is not None:
  46             video_id = m_video_id.group(1)
  47             return self._get_video_info(video_id, query)
  48         else:
  49             player_key = self._search_regex(r'playerKey=(.+?)(&|$)', query, 'playlist_id')
  50             return self._get_playlist_info(player_key)
  51
  52     def _get_video_info(self, video_id, query):
  53         request_url = self._FEDERATED_URL_TEMPLATE % query
  54         webpage = self._download_webpage(request_url, video_id)
  55
  56         self.report_extraction(video_id)
  57         info = self._search_regex(r'var experienceJSON = ({.*?});', webpage, 'json')
  58         info = json.loads(info)['data']
  59         video_info = info['programmedContent']['videoPlayer']['mediaDTO']
  60
  61         return self._extract_video_info(video_info)
  62
  63     def _get_playlist_info(self, player_key):
  64         playlist_info = self._download_webpage(self._PLAYLIST_URL_TEMPLATE % player_key,
  65                                                player_key, u'Downloading playlist information')
  66
  67         playlist_info = json.loads(playlist_info)['videoList']
  68         videos = [self._extract_video_info(video_info) for video_info in playlist_info['mediaCollectionDTO']['videoDTOs']]
  69
  70         return self.playlist_result(videos, playlist_id=playlist_info['id'],
  71                                     playlist_title=playlist_info['mediaCollectionDTO']['displayName'])
  72
  73     def _extract_video_info(self, video_info):
  74         renditions = video_info['renditions']
  75         renditions = sorted(renditions, key=lambda r: r['size'])
  76         best_format = renditions[-1]
  77
  78         return {'id': video_info['id'],
  79                 'title': video_info['displayName'],
  80                 'url': best_format['defaultURL'],
  81                 'ext': 'mp4',
  82                 'description': video_info.get('shortDescription'),
  83                 'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
  84                 'uploader': video_info.get('publisherName'),
  85                 }