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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     unified_strdate,
   7     xpath_text,
   8 )
   9
  10
  11 class CinchcastIE(InfoExtractor):
  12     _VALID_URL = r'https?://player\.cinchcast\.com/.*?assetId=(?P<id>[0-9]+)'
  13     _TEST = {
  14         # Actual test is run in generic, look for undergroundwellness
  15         'url': 'http://player.cinchcast.com/?platformId=1&#038;assetType=single&#038;assetId=7141703',
  16         'only_matching': True,
  17     }
  18
  19     def _real_extract(self, url):
  20         video_id = self._match_id(url)
  21         doc = self._download_xml(
  22             'http://www.blogtalkradio.com/playerasset/mrss?assetType=single&assetId=%s' % video_id,
  23             video_id)
  24
  25         item = doc.find('.//item')
  26         title = xpath_text(item, './title', fatal=True)
  27         date_str = xpath_text(
  28             item, './{http://developer.longtailvideo.com/trac/}date')
  29         upload_date = unified_strdate(date_str, day_first=False)
  30         # duration is present but wrong
  31         formats = [{
  32             'format_id': 'main',
  33             'url': item.find('./{http://search.yahoo.com/mrss/}content').attrib['url'],
  34         }]
  35         backup_url = xpath_text(
  36             item, './{http://developer.longtailvideo.com/trac/}backupContent')
  37         if backup_url:
  38             formats.append({
  39                 'preference': 2,  # seems to be more reliable
  40                 'format_id': 'backup',
  41                 'url': backup_url,
  42             })
  43         self._sort_formats(formats)
  44
  45         return {
  46             'id': video_id,
  47             'title': title,
  48             'upload_date': upload_date,
  49             'formats': formats,
  50         }