_ Git - youtube-dl/blob - youtube_dl/extractor/mediasite.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10     compat_urlparse,
  11 )
  12 from ..utils import (
  13     ExtractorError,
  14     float_or_none,
  15     mimetype2ext,
  16     unescapeHTML,
  17     unsmuggle_url,
  18     url_or_none,
  19     urljoin,
  20 )
  21
  22
  23 class MediasiteIE(InfoExtractor):
  24     _VALID_URL = r'(?xi)https?://[^/]+/Mediasite/(?:Play|Showcase/livebroadcast/Presentation)/(?P<id>[0-9a-f]{32,34})(?P<query>\?[^#]+|)'
  25     _TESTS = [
  26         {
  27             'url': 'https://hitsmediaweb.h-its.org/mediasite/Play/2db6c271681e4f199af3c60d1f82869b1d',
  28             'info_dict': {
  29                 'id': '2db6c271681e4f199af3c60d1f82869b1d',
  30                 'ext': 'mp4',
  31                 'title': 'Lecture: Tuesday, September 20, 2016 - Sir Andrew Wiles',
  32                 'description': 'Sir Andrew Wiles: “Equations in arithmetic”\\n\\nI will describe some of the interactions between modern number theory and the problem of solving equations in rational numbers or integers\\u0027.',
  33                 'timestamp': 1474268400.0,
  34                 'upload_date': '20160919',
  35             },
  36         },
  37         {
  38             'url': 'http://mediasite.uib.no/Mediasite/Play/90bb363295d945d6b548c867d01181361d?catalog=a452b7df-9ae1-46b7-a3ba-aceeb285f3eb',
  39             'info_dict': {
  40                 'id': '90bb363295d945d6b548c867d01181361d',
  41                 'ext': 'mp4',
  42                 'upload_date': '20150429',
  43                 'title': '5) IT-forum 2015-Dag 1  - Dungbeetle -  How and why Rain created a tiny bug tracker for Unity',
  44                 'timestamp': 1430311380.0,
  45             },
  46         },
  47         {
  48             'url': 'https://collegerama.tudelft.nl/Mediasite/Play/585a43626e544bdd97aeb71a0ec907a01d',
  49             'md5': '481fda1c11f67588c0d9d8fbdced4e39',
  50             'info_dict': {
  51                 'id': '585a43626e544bdd97aeb71a0ec907a01d',
  52                 'ext': 'mp4',
  53                 'title': 'Een nieuwe wereld: waarden, bewustzijn en techniek van de mensheid 2.0.',
  54                 'description': '',
  55                 'thumbnail': r're:^https?://.*\.jpg(?:\?.*)?$',
  56                 'duration': 7713.088,
  57                 'timestamp': 1413309600,
  58                 'upload_date': '20141014',
  59             },
  60         },
  61         {
  62             'url': 'https://collegerama.tudelft.nl/Mediasite/Play/86a9ea9f53e149079fbdb4202b521ed21d?catalog=fd32fd35-6c99-466c-89d4-cd3c431bc8a4',
  63             'md5': 'ef1fdded95bdf19b12c5999949419c92',
  64             'info_dict': {
  65                 'id': '86a9ea9f53e149079fbdb4202b521ed21d',
  66                 'ext': 'wmv',
  67                 'title': '64ste Vakantiecursus: Afvalwater',
  68                 'description': 'md5:7fd774865cc69d972f542b157c328305',
  69                 'thumbnail': r're:^https?://.*\.jpg(?:\?.*?)?$',
  70                 'duration': 10853,
  71                 'timestamp': 1326446400,
  72                 'upload_date': '20120113',
  73             },
  74         },
  75         {
  76             'url': 'http://digitalops.sandia.gov/Mediasite/Play/24aace4429fc450fb5b38cdbf424a66e1d',
  77             'md5': '9422edc9b9a60151727e4b6d8bef393d',
  78             'info_dict': {
  79                 'id': '24aace4429fc450fb5b38cdbf424a66e1d',
  80                 'ext': 'mp4',
  81                 'title': 'Xyce Software Training - Section 1',
  82                 'description': r're:(?s)SAND Number: SAND 2013-7800.{200,}',
  83                 'upload_date': '20120409',
  84                 'timestamp': 1333983600,
  85                 'duration': 7794,
  86             }
  87         },
  88         {
  89             'url': 'https://collegerama.tudelft.nl/Mediasite/Showcase/livebroadcast/Presentation/ada7020854f743c49fbb45c9ec7dbb351d',
  90             'only_matching': True,
  91         },
  92     ]
  93
  94     # look in Mediasite.Core.js (Mediasite.ContentStreamType[*])
  95     _STREAM_TYPES = {
  96         0: 'video1',  # the main video
  97         2: 'slide',
  98         3: 'presentation',
  99         4: 'video2',  # screencast?
 100         5: 'video3',
 101     }
 102
 103     @staticmethod
 104     def _extract_urls(webpage):
 105         return [
 106             unescapeHTML(mobj.group('url'))
 107             for mobj in re.finditer(
 108                 r'(?xi)<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:(?:https?:)?//[^/]+)?/Mediasite/Play/[0-9a-f]{32,34}(?:\?.*?)?)\1',
 109                 webpage)]
 110
 111     def _real_extract(self, url):
 112         url, data = unsmuggle_url(url, {})
 113         mobj = re.match(self._VALID_URL, url)
 114         resource_id = mobj.group('id')
 115         query = mobj.group('query')
 116
 117         webpage, urlh = self._download_webpage_handle(url, resource_id)  # XXX: add UrlReferrer?
 118         redirect_url = compat_str(urlh.geturl())
 119
 120         # XXX: might have also extracted UrlReferrer and QueryString from the html
 121         service_path = compat_urlparse.urljoin(redirect_url, self._html_search_regex(
 122             r'<div[^>]+\bid=["\']ServicePath[^>]+>(.+?)</div>', webpage, resource_id,
 123             default='/Mediasite/PlayerService/PlayerService.svc/json'))
 124
 125         player_options = self._download_json(
 126             '%s/GetPlayerOptions' % service_path, resource_id,
 127             headers={
 128                 'Content-type': 'application/json; charset=utf-8',
 129                 'X-Requested-With': 'XMLHttpRequest',
 130             },
 131             data=json.dumps({
 132                 'getPlayerOptionsRequest': {
 133                     'ResourceId': resource_id,
 134                     'QueryString': query,
 135                     'UrlReferrer': data.get('UrlReferrer', ''),
 136                     'UseScreenReader': False,
 137                 }
 138             }).encode('utf-8'))['d']
 139
 140         presentation = player_options['Presentation']
 141         title = presentation['Title']
 142
 143         if presentation is None:
 144             raise ExtractorError(
 145                 'Mediasite says: %s' % player_options['PlayerPresentationStatusMessage'],
 146                 expected=True)
 147
 148         thumbnails = []
 149         formats = []
 150         for snum, Stream in enumerate(presentation['Streams']):
 151             stream_type = Stream.get('StreamType')
 152             if stream_type is None:
 153                 continue
 154
 155             video_urls = Stream.get('VideoUrls')
 156             if not isinstance(video_urls, list):
 157                 video_urls = []
 158
 159             stream_id = self._STREAM_TYPES.get(
 160                 stream_type, 'type%u' % stream_type)
 161
 162             stream_formats = []
 163             for unum, VideoUrl in enumerate(video_urls):
 164                 video_url = url_or_none(VideoUrl.get('Location'))
 165                 if not video_url:
 166                     continue
 167                 # XXX: if Stream.get('CanChangeScheme', False), switch scheme to HTTP/HTTPS
 168
 169                 media_type = VideoUrl.get('MediaType')
 170                 if media_type == 'SS':
 171                     stream_formats.extend(self._extract_ism_formats(
 172                         video_url, resource_id,
 173                         ism_id='%s-%u.%u' % (stream_id, snum, unum),
 174                         fatal=False))
 175                 elif media_type == 'Dash':
 176                     stream_formats.extend(self._extract_mpd_formats(
 177                         video_url, resource_id,
 178                         mpd_id='%s-%u.%u' % (stream_id, snum, unum),
 179                         fatal=False))
 180                 else:
 181                     stream_formats.append({
 182                         'format_id': '%s-%u.%u' % (stream_id, snum, unum),
 183                         'url': video_url,
 184                         'ext': mimetype2ext(VideoUrl.get('MimeType')),
 185                     })
 186
 187             # TODO: if Stream['HasSlideContent']:
 188             # synthesise an MJPEG video stream '%s-%u.slides' % (stream_type, snum)
 189             # from Stream['Slides']
 190             # this will require writing a custom downloader...
 191
 192             # disprefer 'secondary' streams
 193             if stream_type != 0:
 194                 for fmt in stream_formats:
 195                     fmt['preference'] = -1
 196
 197             thumbnail_url = Stream.get('ThumbnailUrl')
 198             if thumbnail_url:
 199                 thumbnails.append({
 200                     'id': '%s-%u' % (stream_id, snum),
 201                     'url': urljoin(redirect_url, thumbnail_url),
 202                     'preference': -1 if stream_type != 0 else 0,
 203                 })
 204             formats.extend(stream_formats)
 205
 206         self._sort_formats(formats)
 207
 208         # XXX: Presentation['Presenters']
 209         # XXX: Presentation['Transcript']
 210
 211         return {
 212             'id': resource_id,
 213             'title': title,
 214             'description': presentation.get('Description'),
 215             'duration': float_or_none(presentation.get('Duration'), 1000),
 216             'timestamp': float_or_none(presentation.get('UnixTime'), 1000),
 217             'formats': formats,
 218             'thumbnails': thumbnails,
 219         }