youtube_dl/extractor/gputechconf.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5
   6
   7 class GPUTechConfIE(InfoExtractor):
   8     _VALID_URL = r'https?://on-demand\.gputechconf\.com/gtc/2015/video/S(?P<id>\d+)\.html'
   9     _TEST = {
  10         'url': 'http://on-demand.gputechconf.com/gtc/2015/video/S5156.html',
  11         'md5': 'a8862a00a0fd65b8b43acc5b8e33f798',
  12         'info_dict': {
  13             'id': '5156',
  14             'ext': 'mp4',
  15             'title': 'Coordinating More Than 3 Million CUDA Threads for Social Network Analysis',
  16             'duration': 1219,
  17         }
  18     }
  19
  20     def _real_extract(self, url):
  21         video_id = self._match_id(url)
  22         webpage = self._download_webpage(url, video_id)
  23
  24         root_path = self._search_regex(
  25             r'var\s+rootPath\s*=\s*"([^"]+)', webpage, 'root path',
  26             default='http://evt.dispeak.com/nvidia/events/gtc15/')
  27         xml_file_id = self._search_regex(
  28             r'var\s+xmlFileId\s*=\s*"([^"]+)', webpage, 'xml file id')
  29
  30         return {
  31             '_type': 'url_transparent',
  32             'id': video_id,
  33             'url': '%sxml/%s.xml' % (root_path, xml_file_id),
  34             'ie_key': 'DigitallySpeaking',
  35         }