yt_dlp/extractor/ustream.py

   1 from __future__ import unicode_literals
   2
   3 import random
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_str,
   9     compat_urlparse,
  10 )
  11 from ..utils import (
  12     encode_data_uri,
  13     ExtractorError,
  14     int_or_none,
  15     float_or_none,
  16     join_nonempty,
  17     mimetype2ext,
  18     str_or_none,
  19 )
  20
  21
  22 class UstreamIE(InfoExtractor):
  23     _VALID_URL = r'https?://(?:www\.)?(?:ustream\.tv|video\.ibm\.com)/(?P<type>recorded|embed|embed/recorded)/(?P<id>\d+)'
  24     IE_NAME = 'ustream'
  25     _TESTS = [{
  26         'url': 'http://www.ustream.tv/recorded/20274954',
  27         'md5': '088f151799e8f572f84eb62f17d73e5c',
  28         'info_dict': {
  29             'id': '20274954',
  30             'ext': 'flv',
  31             'title': 'Young Americans for Liberty February 7, 2012 2:28 AM',
  32             'description': 'Young Americans for Liberty February 7, 2012 2:28 AM',
  33             'timestamp': 1328577035,
  34             'upload_date': '20120207',
  35             'uploader': 'yaliberty',
  36             'uploader_id': '6780869',
  37         },
  38     }, {
  39         # From http://sportscanada.tv/canadagames/index.php/week2/figure-skating/444
  40         # Title and uploader available only from params JSON
  41         'url': 'http://www.ustream.tv/embed/recorded/59307601?ub=ff0000&lc=ff0000&oc=ffffff&uc=ffffff&v=3&wmode=direct',
  42         'md5': '5a2abf40babeac9812ed20ae12d34e10',
  43         'info_dict': {
  44             'id': '59307601',
  45             'ext': 'flv',
  46             'title': '-CG11- Canada Games Figure Skating',
  47             'uploader': 'sportscanadatv',
  48         },
  49         'skip': 'This Pro Broadcaster has chosen to remove this video from the ustream.tv site.',
  50     }, {
  51         'url': 'http://www.ustream.tv/embed/10299409',
  52         'info_dict': {
  53             'id': '10299409',
  54         },
  55         'playlist_count': 3,
  56     }, {
  57         'url': 'http://www.ustream.tv/recorded/91343263',
  58         'info_dict': {
  59             'id': '91343263',
  60             'ext': 'mp4',
  61             'title': 'GitHub Universe - General Session - Day 1',
  62             'upload_date': '20160914',
  63             'description': 'GitHub Universe - General Session - Day 1',
  64             'timestamp': 1473872730,
  65             'uploader': 'wa0dnskeqkr',
  66             'uploader_id': '38977840',
  67         },
  68         'params': {
  69             'skip_download': True,  # m3u8 download
  70         },
  71     }, {
  72         'url': 'https://video.ibm.com/embed/recorded/128240221?&autoplay=true&controls=true&volume=100',
  73         'only_matching': True,
  74     }]
  75
  76     @staticmethod
  77     def _extract_url(webpage):
  78         mobj = re.search(
  79             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?(?:ustream\.tv|video\.ibm\.com)/embed/.+?)\1', webpage)
  80         if mobj is not None:
  81             return mobj.group('url')
  82
  83     def _get_stream_info(self, url, video_id, app_id_ver, extra_note=None):
  84         def num_to_hex(n):
  85             return hex(n)[2:]
  86
  87         rnd = random.randrange
  88
  89         if not extra_note:
  90             extra_note = ''
  91
  92         conn_info = self._download_json(
  93             'http://r%d-1-%s-recorded-lp-live.ums.ustream.tv/1/ustream' % (rnd(1e8), video_id),
  94             video_id, note='Downloading connection info' + extra_note,
  95             query={
  96                 'type': 'viewer',
  97                 'appId': app_id_ver[0],
  98                 'appVersion': app_id_ver[1],
  99                 'rsid': '%s:%s' % (num_to_hex(rnd(1e8)), num_to_hex(rnd(1e8))),
 100                 'rpin': '_rpin.%d' % rnd(1e15),
 101                 'referrer': url,
 102                 'media': video_id,
 103                 'application': 'recorded',
 104             })
 105         host = conn_info[0]['args'][0]['host']
 106         connection_id = conn_info[0]['args'][0]['connectionId']
 107
 108         return self._download_json(
 109             'http://%s/1/ustream?connectionId=%s' % (host, connection_id),
 110             video_id, note='Downloading stream info' + extra_note)
 111
 112     def _get_streams(self, url, video_id, app_id_ver):
 113         # Sometimes the return dict does not have 'stream'
 114         for trial_count in range(3):
 115             stream_info = self._get_stream_info(
 116                 url, video_id, app_id_ver,
 117                 extra_note=' (try %d)' % (trial_count + 1) if trial_count > 0 else '')
 118             if 'stream' in stream_info[0]['args'][0]:
 119                 return stream_info[0]['args'][0]['stream']
 120         return []
 121
 122     def _parse_segmented_mp4(self, dash_stream_info):
 123         def resolve_dash_template(template, idx, chunk_hash):
 124             return template.replace('%', compat_str(idx), 1).replace('%', chunk_hash)
 125
 126         formats = []
 127         for stream in dash_stream_info['streams']:
 128             # Use only one provider to avoid too many formats
 129             provider = dash_stream_info['providers'][0]
 130             fragments = [{
 131                 'url': resolve_dash_template(
 132                     provider['url'] + stream['initUrl'], 0, dash_stream_info['hashes']['0'])
 133             }]
 134             for idx in range(dash_stream_info['videoLength'] // dash_stream_info['chunkTime']):
 135                 fragments.append({
 136                     'url': resolve_dash_template(
 137                         provider['url'] + stream['segmentUrl'], idx,
 138                         dash_stream_info['hashes'][compat_str(idx // 10 * 10)])
 139                 })
 140             content_type = stream['contentType']
 141             kind = content_type.split('/')[0]
 142             f = {
 143                 'format_id': join_nonempty(
 144                     'dash', kind, str_or_none(stream.get('bitrate'))),
 145                 'protocol': 'http_dash_segments',
 146                 # TODO: generate a MPD doc for external players?
 147                 'url': encode_data_uri(b'<MPD/>', 'text/xml'),
 148                 'ext': mimetype2ext(content_type),
 149                 'height': stream.get('height'),
 150                 'width': stream.get('width'),
 151                 'fragments': fragments,
 152             }
 153             if kind == 'video':
 154                 f.update({
 155                     'vcodec': stream.get('codec'),
 156                     'acodec': 'none',
 157                     'vbr': stream.get('bitrate'),
 158                 })
 159             else:
 160                 f.update({
 161                     'vcodec': 'none',
 162                     'acodec': stream.get('codec'),
 163                     'abr': stream.get('bitrate'),
 164                 })
 165             formats.append(f)
 166         return formats
 167
 168     def _real_extract(self, url):
 169         m = self._match_valid_url(url)
 170         video_id = m.group('id')
 171
 172         # some sites use this embed format (see: https://github.com/ytdl-org/youtube-dl/issues/2990)
 173         if m.group('type') == 'embed/recorded':
 174             video_id = m.group('id')
 175             desktop_url = 'http://www.ustream.tv/recorded/' + video_id
 176             return self.url_result(desktop_url, 'Ustream')
 177         if m.group('type') == 'embed':
 178             video_id = m.group('id')
 179             webpage = self._download_webpage(url, video_id)
 180             content_video_ids = self._parse_json(self._search_regex(
 181                 r'ustream\.vars\.offAirContentVideoIds=([^;]+);', webpage,
 182                 'content video IDs'), video_id)
 183             return self.playlist_result(
 184                 map(lambda u: self.url_result('http://www.ustream.tv/recorded/' + u, 'Ustream'), content_video_ids),
 185                 video_id)
 186
 187         params = self._download_json(
 188             'https://api.ustream.tv/videos/%s.json' % video_id, video_id)
 189
 190         error = params.get('error')
 191         if error:
 192             raise ExtractorError(
 193                 '%s returned error: %s' % (self.IE_NAME, error), expected=True)
 194
 195         video = params['video']
 196
 197         title = video['title']
 198         filesize = float_or_none(video.get('file_size'))
 199
 200         formats = [{
 201             'id': video_id,
 202             'url': video_url,
 203             'ext': format_id,
 204             'filesize': filesize,
 205         } for format_id, video_url in video['media_urls'].items() if video_url]
 206
 207         if not formats:
 208             hls_streams = self._get_streams(url, video_id, app_id_ver=(11, 2))
 209             if hls_streams:
 210                 # m3u8_native leads to intermittent ContentTooShortError
 211                 formats.extend(self._extract_m3u8_formats(
 212                     hls_streams[0]['url'], video_id, ext='mp4', m3u8_id='hls'))
 213
 214             '''
 215             # DASH streams handling is incomplete as 'url' is missing
 216             dash_streams = self._get_streams(url, video_id, app_id_ver=(3, 1))
 217             if dash_streams:
 218                 formats.extend(self._parse_segmented_mp4(dash_streams))
 219             '''
 220
 221         self._sort_formats(formats)
 222
 223         description = video.get('description')
 224         timestamp = int_or_none(video.get('created_at'))
 225         duration = float_or_none(video.get('length'))
 226         view_count = int_or_none(video.get('views'))
 227
 228         uploader = video.get('owner', {}).get('username')
 229         uploader_id = video.get('owner', {}).get('id')
 230
 231         thumbnails = [{
 232             'id': thumbnail_id,
 233             'url': thumbnail_url,
 234         } for thumbnail_id, thumbnail_url in video.get('thumbnail', {}).items()]
 235
 236         return {
 237             'id': video_id,
 238             'title': title,
 239             'description': description,
 240             'thumbnails': thumbnails,
 241             'timestamp': timestamp,
 242             'duration': duration,
 243             'view_count': view_count,
 244             'uploader': uploader,
 245             'uploader_id': uploader_id,
 246             'formats': formats,
 247         }
 248
 249
 250 class UstreamChannelIE(InfoExtractor):
 251     _VALID_URL = r'https?://(?:www\.)?ustream\.tv/channel/(?P<slug>.+)'
 252     IE_NAME = 'ustream:channel'
 253     _TEST = {
 254         'url': 'http://www.ustream.tv/channel/channeljapan',
 255         'info_dict': {
 256             'id': '10874166',
 257         },
 258         'playlist_mincount': 17,
 259     }
 260
 261     def _real_extract(self, url):
 262         m = self._match_valid_url(url)
 263         display_id = m.group('slug')
 264         webpage = self._download_webpage(url, display_id)
 265         channel_id = self._html_search_meta('ustream:channel_id', webpage)
 266
 267         BASE = 'http://www.ustream.tv'
 268         next_url = '/ajax/socialstream/videos/%s/1.json' % channel_id
 269         video_ids = []
 270         while next_url:
 271             reply = self._download_json(
 272                 compat_urlparse.urljoin(BASE, next_url), display_id,
 273                 note='Downloading video information (next: %d)' % (len(video_ids) + 1))
 274             video_ids.extend(re.findall(r'data-content-id="(\d.*)"', reply['data']))
 275             next_url = reply['nextUrl']
 276
 277         entries = [
 278             self.url_result('http://www.ustream.tv/recorded/' + vid, 'Ustream')
 279             for vid in video_ids]
 280         return {
 281             '_type': 'playlist',
 282             'id': channel_id,
 283             'display_id': display_id,
 284             'entries': entries,
 285         }