youtube_dl/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/(?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
  72     @staticmethod
  73     def _extract_url(webpage):
  74         mobj = re.search(
  75             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
  76         if mobj is not None:
  77             return mobj.group('url')
  78
  79     def _get_stream_info(self, url, video_id, app_id_ver, extra_note=None):
  80         def num_to_hex(n):
  81             return hex(n)[2:]
  82
  83         rnd = random.randrange
  84
  85         if not extra_note:
  86             extra_note = ''
  87
  88         conn_info = self._download_json(
  89             'http://r%d-1-%s-recorded-lp-live.ums.ustream.tv/1/ustream' % (rnd(1e8), video_id),
  90             video_id, note='Downloading connection info' + extra_note,
  91             query={
  92                 'type': 'viewer',
  93                 'appId': app_id_ver[0],
  94                 'appVersion': app_id_ver[1],
  95                 'rsid': '%s:%s' % (num_to_hex(rnd(1e8)), num_to_hex(rnd(1e8))),
  96                 'rpin': '_rpin.%d' % rnd(1e15),
  97                 'referrer': url,
  98                 'media': video_id,
  99                 'application': 'recorded',
 100             })
 101         host = conn_info[0]['args'][0]['host']
 102         connection_id = conn_info[0]['args'][0]['connectionId']
 103
 104         return self._download_json(
 105             'http://%s/1/ustream?connectionId=%s' % (host, connection_id),
 106             video_id, note='Downloading stream info' + extra_note)
 107
 108     def _get_streams(self, url, video_id, app_id_ver):
 109         # Sometimes the return dict does not have 'stream'
 110         for trial_count in range(3):
 111             stream_info = self._get_stream_info(
 112                 url, video_id, app_id_ver,
 113                 extra_note=' (try %d)' % (trial_count + 1) if trial_count > 0 else '')
 114             if 'stream' in stream_info[0]['args'][0]:
 115                 return stream_info[0]['args'][0]['stream']
 116         return []
 117
 118     def _parse_segmented_mp4(self, dash_stream_info):
 119         def resolve_dash_template(template, idx, chunk_hash):
 120             return template.replace('%', compat_str(idx), 1).replace('%', chunk_hash)
 121
 122         formats = []
 123         for stream in dash_stream_info['streams']:
 124             # Use only one provider to avoid too many formats
 125             provider = dash_stream_info['providers'][0]
 126             fragments = [{
 127                 'url': resolve_dash_template(
 128                     provider['url'] + stream['initUrl'], 0, dash_stream_info['hashes']['0'])
 129             }]
 130             for idx in range(dash_stream_info['videoLength'] // dash_stream_info['chunkTime']):
 131                 fragments.append({
 132                     'url': resolve_dash_template(
 133                         provider['url'] + stream['segmentUrl'], idx,
 134                         dash_stream_info['hashes'][compat_str(idx // 10 * 10)])
 135                 })
 136             content_type = stream['contentType']
 137             kind = content_type.split('/')[0]
 138             f = {
 139                 'format_id': '-'.join(filter(None, [
 140                     'dash', kind, str_or_none(stream.get('bitrate'))])),
 141                 'protocol': 'http_dash_segments',
 142                 # TODO: generate a MPD doc for external players?
 143                 'url': encode_data_uri(b'<MPD/>', 'text/xml'),
 144                 'ext': mimetype2ext(content_type),
 145                 'height': stream.get('height'),
 146                 'width': stream.get('width'),
 147                 'fragments': fragments,
 148             }
 149             if kind == 'video':
 150                 f.update({
 151                     'vcodec': stream.get('codec'),
 152                     'acodec': 'none',
 153                     'vbr': stream.get('bitrate'),
 154                 })
 155             else:
 156                 f.update({
 157                     'vcodec': 'none',
 158                     'acodec': stream.get('codec'),
 159                     'abr': stream.get('bitrate'),
 160                 })
 161             formats.append(f)
 162         return formats
 163
 164     def _real_extract(self, url):
 165         m = re.match(self._VALID_URL, url)
 166         video_id = m.group('id')
 167
 168         # some sites use this embed format (see: https://github.com/ytdl-org/youtube-dl/issues/2990)
 169         if m.group('type') == 'embed/recorded':
 170             video_id = m.group('id')
 171             desktop_url = 'http://www.ustream.tv/recorded/' + video_id
 172             return self.url_result(desktop_url, 'Ustream')
 173         if m.group('type') == 'embed':
 174             video_id = m.group('id')
 175             webpage = self._download_webpage(url, video_id)
 176             content_video_ids = self._parse_json(self._search_regex(
 177                 r'ustream\.vars\.offAirContentVideoIds=([^;]+);', webpage,
 178                 'content video IDs'), video_id)
 179             return self.playlist_result(
 180                 map(lambda u: self.url_result('http://www.ustream.tv/recorded/' + u, 'Ustream'), content_video_ids),
 181                 video_id)
 182
 183         params = self._download_json(
 184             'https://api.ustream.tv/videos/%s.json' % video_id, video_id)
 185
 186         error = params.get('error')
 187         if error:
 188             raise ExtractorError(
 189                 '%s returned error: %s' % (self.IE_NAME, error), expected=True)
 190
 191         video = params['video']
 192
 193         title = video['title']
 194         filesize = float_or_none(video.get('file_size'))
 195
 196         formats = [{
 197             'id': video_id,
 198             'url': video_url,
 199             'ext': format_id,
 200             'filesize': filesize,
 201         } for format_id, video_url in video['media_urls'].items() if video_url]
 202
 203         if not formats:
 204             hls_streams = self._get_streams(url, video_id, app_id_ver=(11, 2))
 205             if hls_streams:
 206                 # m3u8_native leads to intermittent ContentTooShortError
 207                 formats.extend(self._extract_m3u8_formats(
 208                     hls_streams[0]['url'], video_id, ext='mp4', m3u8_id='hls'))
 209
 210             '''
 211             # DASH streams handling is incomplete as 'url' is missing
 212             dash_streams = self._get_streams(url, video_id, app_id_ver=(3, 1))
 213             if dash_streams:
 214                 formats.extend(self._parse_segmented_mp4(dash_streams))
 215             '''
 216
 217         self._sort_formats(formats)
 218
 219         description = video.get('description')
 220         timestamp = int_or_none(video.get('created_at'))
 221         duration = float_or_none(video.get('length'))
 222         view_count = int_or_none(video.get('views'))
 223
 224         uploader = video.get('owner', {}).get('username')
 225         uploader_id = video.get('owner', {}).get('id')
 226
 227         thumbnails = [{
 228             'id': thumbnail_id,
 229             'url': thumbnail_url,
 230         } for thumbnail_id, thumbnail_url in video.get('thumbnail', {}).items()]
 231
 232         return {
 233             'id': video_id,
 234             'title': title,
 235             'description': description,
 236             'thumbnails': thumbnails,
 237             'timestamp': timestamp,
 238             'duration': duration,
 239             'view_count': view_count,
 240             'uploader': uploader,
 241             'uploader_id': uploader_id,
 242             'formats': formats,
 243         }
 244
 245
 246 class UstreamChannelIE(InfoExtractor):
 247     _VALID_URL = r'https?://(?:www\.)?ustream\.tv/channel/(?P<slug>.+)'
 248     IE_NAME = 'ustream:channel'
 249     _TEST = {
 250         'url': 'http://www.ustream.tv/channel/channeljapan',
 251         'info_dict': {
 252             'id': '10874166',
 253         },
 254         'playlist_mincount': 17,
 255     }
 256
 257     def _real_extract(self, url):
 258         m = re.match(self._VALID_URL, url)
 259         display_id = m.group('slug')
 260         webpage = self._download_webpage(url, display_id)
 261         channel_id = self._html_search_meta('ustream:channel_id', webpage)
 262
 263         BASE = 'http://www.ustream.tv'
 264         next_url = '/ajax/socialstream/videos/%s/1.json' % channel_id
 265         video_ids = []
 266         while next_url:
 267             reply = self._download_json(
 268                 compat_urlparse.urljoin(BASE, next_url), display_id,
 269                 note='Downloading video information (next: %d)' % (len(video_ids) + 1))
 270             video_ids.extend(re.findall(r'data-content-id="(\d.*)"', reply['data']))
 271             next_url = reply['nextUrl']
 272
 273         entries = [
 274             self.url_result('http://www.ustream.tv/recorded/' + vid, 'Ustream')
 275             for vid in video_ids]
 276         return {
 277             '_type': 'playlist',
 278             'id': channel_id,
 279             'display_id': display_id,
 280             'entries': entries,
 281         }