yt_dlp/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         # From https://gdcvault.com/play/1016624, empty speakerVideo
  37         'url': 'https://sevt.dispeak.com/ubm/gdc/online12/xml/201210-822101_1349794556671DDDD.xml',
  38         'info_dict': {
  39             'id': '201210-822101_1349794556671DDDD',
  40             'ext': 'flv',
  41             'title': 'Pre-launch - Preparing to Take the Plunge',
  42         },
  43     }, {
  44         # From http://www.gdcvault.com/play/1014846/Conference-Keynote-Shigeru, empty slideVideo
  45         'url': 'http://events.digitallyspeaking.com/gdc/project25/xml/p25-miyamoto1999_1282467389849HSVB.xml',
  46         'only_matching': True,
  47     }]
  48
  49     def _parse_mp4(self, metadata):
  50         video_formats = []
  51         video_root = None
  52
  53         mp4_video = xpath_text(metadata, './mp4video', default=None)
  54         if mp4_video is not None:
  55             mobj = re.match(r'(?P<root>https?://.*?/).*', mp4_video)
  56             video_root = mobj.group('root')
  57         if video_root is None:
  58             http_host = xpath_text(metadata, 'httpHost', default=None)
  59             if http_host:
  60                 video_root = 'http://%s/' % http_host
  61         if video_root is None:
  62             # Hard-coded in http://evt.dispeak.com/ubm/gdc/sf16/custom/player2.js
  63             # Works for GPUTechConf, too
  64             video_root = 'http://s3-2u.digitallyspeaking.com/'
  65
  66         formats = metadata.findall('./MBRVideos/MBRVideo')
  67         if not formats:
  68             return None
  69         for a_format in formats:
  70             stream_name = xpath_text(a_format, 'streamName', fatal=True)
  71             video_path = re.match(r'mp4\:(?P<path>.*)', stream_name).group('path')
  72             url = video_root + video_path
  73             bitrate = xpath_text(a_format, 'bitrate')
  74             tbr = int_or_none(bitrate)
  75             vbr = int_or_none(self._search_regex(
  76                 r'-(\d+)\.mp4', video_path, 'vbr', default=None))
  77             video_formats.append({
  78                 'format_id': bitrate,
  79                 'url': url,
  80                 'tbr': tbr,
  81                 'vbr': vbr,
  82             })
  83         return video_formats
  84
  85     def _parse_flv(self, metadata):
  86         formats = []
  87         akamai_url = xpath_text(metadata, './akamaiHost', fatal=True)
  88         audios = metadata.findall('./audios/audio')
  89         for audio in audios:
  90             formats.append({
  91                 'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
  92                 'play_path': remove_end(audio.get('url'), '.flv'),
  93                 'ext': 'flv',
  94                 'vcodec': 'none',
  95                 'quality': 1,
  96                 'format_id': audio.get('code'),
  97             })
  98         for video_key, format_id, preference in (
  99                 ('slide', 'slides', -2), ('speaker', 'speaker', -1)):
 100             video_path = xpath_text(metadata, './%sVideo' % video_key)
 101             if not video_path:
 102                 continue
 103             formats.append({
 104                 'url': 'rtmp://%s/ondemand?ovpfv=1.1' % akamai_url,
 105                 'play_path': remove_end(video_path, '.flv'),
 106                 'ext': 'flv',
 107                 'format_note': '%s video' % video_key,
 108                 'quality': preference,
 109                 'format_id': format_id,
 110             })
 111         return formats
 112
 113     def _real_extract(self, url):
 114         video_id = self._match_id(url)
 115
 116         xml_description = self._download_xml(url, video_id)
 117         metadata = xpath_element(xml_description, 'metadata')
 118
 119         video_formats = self._parse_mp4(metadata)
 120         if video_formats is None:
 121             video_formats = self._parse_flv(metadata)
 122         self._sort_formats(video_formats)
 123
 124         return {
 125             'id': video_id,
 126             'formats': video_formats,
 127             'title': xpath_text(metadata, 'title', fatal=True),
 128             'duration': parse_duration(xpath_text(metadata, 'endTime')),
 129             'creator': xpath_text(metadata, 'speaker'),
 130         }