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