yt_dlp/extractor/ustream.py

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