_ 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 )
   9
  10 class BrightcoveIE(InfoExtractor):
  11     _VALID_URL = r'http://.*brightcove\.com/.*\?(?P<query>.*videoPlayer=(?P<id>\d*).*)'
  12     _FEDERATED_URL_TEMPLATE = 'http://c.brightcove.com/services/viewer/htmlFederated?%s'
  13
  14     # There is a test for Brigtcove in GenericIE, that way we test both the download
  15     # and the detection of videos, and we don't have to find an URL that is always valid
  16
  17     @classmethod
  18     def _build_brighcove_url(cls, object_str):
  19         """
  20         Build a Brightcove url from a xml string containing
  21         <object class="BrightcoveExperience">{params}</object>
  22         """
  23         object_doc = xml.etree.ElementTree.fromstring(object_str)
  24         assert object_doc.attrib['class'] == u'BrightcoveExperience'
  25         params = {'flashID': object_doc.attrib['id'],
  26                   'playerID': object_doc.find('./param[@name="playerID"]').attrib['value'],
  27                   '@videoPlayer': object_doc.find('./param[@name="@videoPlayer"]').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         data = compat_urllib_parse.urlencode(params)
  34         return cls._FEDERATED_URL_TEMPLATE % data
  35
  36     def _real_extract(self, url):
  37         mobj = re.match(self._VALID_URL, url)
  38         query = mobj.group('query')
  39         video_id = mobj.group('id')
  40
  41         request_url = self._FEDERATED_URL_TEMPLATE % query
  42         webpage = self._download_webpage(request_url, video_id)
  43
  44         self.report_extraction(video_id)
  45         info = self._search_regex(r'var experienceJSON = ({.*?});', webpage, 'json')
  46         info = json.loads(info)['data']
  47         video_info = info['programmedContent']['videoPlayer']['mediaDTO']
  48         renditions = video_info['renditions']
  49         renditions = sorted(renditions, key=lambda r: r['size'])
  50         best_format = renditions[-1]
  51
  52         return {'id': video_id,
  53                 'title': video_info['displayName'],
  54                 'url': best_format['defaultURL'],
  55                 'ext': 'mp4',
  56                 'description': video_info.get('shortDescription'),
  57                 'thumbnail': video_info.get('videoStillURL') or video_info.get('thumbnailURL'),
  58                 'uploader': video_info.get('publisherName'),
  59                 }