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

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     int_or_none,
   8     parse_duration,
   9     remove_end,
  10     xpath_element,
  11     xpath_text,
  12 )
  13
  14
  15 class DigitallySpeakingIE(InfoExtractor):
  16     _VALID_URL = r'https?://(?:s?evt\.dispeak|events\.digitallyspeaking)\.com/(?:[^/]+/)+xml/(?P<id>[^.]+)\.xml'
  17
  18     _TESTS = [{
  19         # From http://gdcvault.com/play/1023460/Tenacious-Design-and-The-Interface
  20         'url': 'http://evt.dispeak.com/ubm/gdc/sf16/xml/840376_BQRC.xml',
  21         'md5': 'a8efb6c31ed06ca8739294960b2dbabd',
  22         'info_dict': {
  23             'id': '840376_BQRC',
  24             'ext': 'mp4',
  25             'title': 'Tenacious Design and The Interface of \'Destiny\'',
  26         },
  27     }, {
  28         # From http://www.gdcvault.com/play/1014631/Classic-Game-Postmortem-PAC
  29         'url': 'http://events.digitallyspeaking.com/gdc/sf11/xml/12396_1299111843500GMPX.xml',
  30         'only_matching': True,
  31     }, {
  32         # From http://www.gdcvault.com/play/1013700/Advanced-Material
  33         'url': 'http://sevt.dispeak.com/ubm/gdc/eur10/xml/11256_1282118587281VNIT.xml',
  34         'only_matching': True,
  35     }]
  36
  37     def _parse_mp4(self, metadata):
  38         video_formats = []
  39         video_root = None
  40
  41         mp4_video = xpath_text(metadata, './mp4video', default=None)
  42         if mp4_video is not None:
  43             mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video)
  44             video_root = mobj.group('root')
  45         if video_root is None:
  46             http_host = xpath_text(metadata, 'httpHost', default=None)
  47             if http_host:
  48                 video_root = 'http://%s/' % http_host
  49         if video_root is None:
  50             # Hard-coded in http://evt.dispeak.com/ubm/gdc/sf16/custom/player2.js
  51             # Works for GPUTechConf, too
  52             video_root = 'http://s3-2u.digitallyspeaking.com/'
  53
  54         formats = metadata.findall('./MBRVideos/MBRVideo')
  55         if not formats:
  56             return None
  57         for a_format in formats:
  58             stream_name = xpath_text(a_format, 'streamName', fatal=True)
  59             video_path = re.match(r'mp4\:(?P<path>.*)', stream_name).group('path')
  60             url = video_root + video_path
  61             vbr = xpath_text(a_format, 'bitrate')
  62             video_formats.append({
  63                 'url': url,
  64                 'vbr': int_or_none(vbr),
  65             })
  66         return video_formats
  67
  68     def _parse_flv(self, metadata):
  69         formats = []
  70         akamai_url = xpath_text(metadata, './akamaiHost', fatal=True)
  71         audios = metadata.findall('./audios/audio')
  72         for audio in audios:
  73             formats.append({
  74                 'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
  75                 'play_path': remove_end(audio.get('url'), '.flv'),
  76                 'ext': 'flv',
  77                 'vcodec': 'none',
  78                 'format_id': audio.get('code'),
  79             })
  80         slide_video_path = xpath_text(metadata, './slideVideo', fatal=True)
  81         formats.append({
  82             'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
  83             'play_path': remove_end(slide_video_path, '.flv'),
  84             'ext': 'flv',
  85             'format_note': 'slide deck video',
  86             'quality': -2,
  87             'preference': -2,
  88             'format_id': 'slides',
  89         })
  90         speaker_video_path = xpath_text(metadata, './speakerVideo', fatal=True)
  91         formats.append({
  92             'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
  93             'play_path': remove_end(speaker_video_path, '.flv'),
  94             'ext': 'flv',
  95             'format_note': 'speaker video',
  96             'quality': -1,
  97             'preference': -1,
  98             'format_id': 'speaker',
  99         })
 100         return formats
 101
 102     def _real_extract(self, url):
 103         video_id = self._match_id(url)
 104
 105         xml_description = self._download_xml(url, video_id)
 106         metadata = xpath_element(xml_description, 'metadata')
 107
 108         video_formats = self._parse_mp4(metadata)
 109         if video_formats is None:
 110             video_formats = self._parse_flv(metadata)
 111
 112         return {
 113             'id': video_id,
 114             'formats': video_formats,
 115             'title': xpath_text(metadata, 'title', fatal=True),
 116             'duration': parse_duration(xpath_text(metadata, 'endTime')),
 117             'creator': xpath_text(metadata, 'speaker'),
 118         }