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