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