yt_dlp/extractor/ustream.py

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