yt_dlp/extractor/livestream.py

   1 import itertools
   2 import re
   3
   4 from .common import InfoExtractor
   5 from ..compat import compat_str, compat_urlparse
   6 from ..utils import (
   7     determine_ext,
   8     find_xpath_attr,
   9     float_or_none,
  10     int_or_none,
  11     orderedSet,
  12     parse_iso8601,
  13     traverse_obj,
  14     update_url_query,
  15     xpath_attr,
  16     xpath_text,
  17     xpath_with_ns,
  18 )
  19
  20
  21 class LivestreamIE(InfoExtractor):
  22     IE_NAME = 'livestream'
  23     _VALID_URL = r'''(?x)
  24         https?://(?:new\.)?livestream\.com/
  25         (?:accounts/(?P<account_id>\d+)|(?P<account_name>[^/]+))
  26         (?:/events/(?P<event_id>\d+)|/(?P<event_name>[^/]+))?
  27         (?:/videos/(?P<id>\d+))?
  28     '''
  29     _EMBED_REGEX = [r'<iframe[^>]+src="(?P<url>https?://(?:new\.)?livestream\.com/[^"]+/player[^"]+)"']
  30
  31     _TESTS = [{
  32         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  33         'md5': '7876c5f5dc3e711b6b73acce4aac1527',
  34         'info_dict': {
  35             'id': '4719370',
  36             'ext': 'mp4',
  37             'title': 'Live from Webster Hall NYC',
  38             'timestamp': 1350008072,
  39             'upload_date': '20121012',
  40             'duration': 5968.0,
  41             'like_count': int,
  42             'view_count': int,
  43             'comment_count': int,
  44             'thumbnail': r're:^http://.*\.jpg$'
  45         }
  46     }, {
  47         'url': 'https://livestream.com/coheedandcambria/websterhall',
  48         'info_dict': {
  49             'id': '1585861',
  50             'title': 'Live From Webster Hall'
  51         },
  52         'playlist_mincount': 1,
  53     }, {
  54         'url': 'https://livestream.com/dayananda/events/7954027',
  55         'info_dict': {
  56             'title': 'Live from Mevo',
  57             'id': '7954027',
  58         },
  59         'playlist_mincount': 4,
  60     }, {
  61         'url': 'https://livestream.com/accounts/82',
  62         'info_dict': {
  63             'id': '253978',
  64             'view_count': int,
  65             'title': 'trsr',
  66             'comment_count': int,
  67             'like_count': int,
  68             'upload_date': '20120306',
  69             'timestamp': 1331042383,
  70             'thumbnail': 'http://img.new.livestream.com/videos/0000000000000372/cacbeed6-fb68-4b5e-ad9c-e148124e68a9_640x427.jpg',
  71             'duration': 15.332,
  72             'ext': 'mp4'
  73         }
  74     }, {
  75         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  76         'only_matching': True,
  77     }, {
  78         'url': 'http://livestream.com/bsww/concacafbeachsoccercampeonato2015',
  79         'only_matching': True,
  80     }]
  81     _API_URL_TEMPLATE = 'http://livestream.com/api/accounts/%s/events/%s'
  82
  83     def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
  84         base_ele = find_xpath_attr(
  85             smil, self._xpath_ns('.//meta', namespace), 'name', 'httpBase')
  86         base = base_ele.get('content') if base_ele is not None else 'http://livestreamvod-f.akamaihd.net/'
  87
  88         formats = []
  89         video_nodes = smil.findall(self._xpath_ns('.//video', namespace))
  90
  91         for vn in video_nodes:
  92             tbr = int_or_none(vn.attrib.get('system-bitrate'), 1000)
  93             furl = (
  94                 update_url_query(compat_urlparse.urljoin(base, vn.attrib['src']), {
  95                     'v': '3.0.3',
  96                     'fp': 'WIN% 14,0,0,145',
  97                 }))
  98             if 'clipBegin' in vn.attrib:
  99                 furl += '&ssek=' + vn.attrib['clipBegin']
 100             formats.append({
 101                 'url': furl,
 102                 'format_id': 'smil_%d' % tbr,
 103                 'ext': 'flv',
 104                 'tbr': tbr,
 105                 'preference': -1000,  # Strictly inferior than all other formats?
 106             })
 107         return formats
 108
 109     def _extract_video_info(self, video_data):
 110         video_id = compat_str(video_data['id'])
 111
 112         FORMAT_KEYS = (
 113             ('sd', 'progressive_url'),
 114             ('hd', 'progressive_url_hd'),
 115         )
 116
 117         formats = []
 118         for format_id, key in FORMAT_KEYS:
 119             video_url = video_data.get(key)
 120             if video_url:
 121                 ext = determine_ext(video_url)
 122                 if ext == 'm3u8':
 123                     continue
 124                 bitrate = int_or_none(self._search_regex(
 125                     r'(\d+)\.%s' % ext, video_url, 'bitrate', default=None))
 126                 formats.append({
 127                     'url': video_url,
 128                     'format_id': format_id,
 129                     'tbr': bitrate,
 130                     'ext': ext,
 131                 })
 132
 133         smil_url = video_data.get('smil_url')
 134         if smil_url:
 135             formats.extend(self._extract_smil_formats(smil_url, video_id, fatal=False))
 136
 137         m3u8_url = video_data.get('m3u8_url')
 138         if m3u8_url:
 139             formats.extend(self._extract_m3u8_formats(
 140                 m3u8_url, video_id, 'mp4', 'm3u8_native',
 141                 m3u8_id='hls', fatal=False))
 142
 143         f4m_url = video_data.get('f4m_url')
 144         if f4m_url:
 145             formats.extend(self._extract_f4m_formats(
 146                 f4m_url, video_id, f4m_id='hds', fatal=False))
 147
 148         comments = [{
 149             'author_id': comment.get('author_id'),
 150             'author': comment.get('author', {}).get('full_name'),
 151             'id': comment.get('id'),
 152             'text': comment['text'],
 153             'timestamp': parse_iso8601(comment.get('created_at')),
 154         } for comment in video_data.get('comments', {}).get('data', [])]
 155
 156         return {
 157             'id': video_id,
 158             'formats': formats,
 159             'title': video_data['caption'],
 160             'description': video_data.get('description'),
 161             'thumbnail': video_data.get('thumbnail_url'),
 162             'duration': float_or_none(video_data.get('duration'), 1000),
 163             'timestamp': parse_iso8601(video_data.get('publish_at')),
 164             'like_count': video_data.get('likes', {}).get('total'),
 165             'comment_count': video_data.get('comments', {}).get('total'),
 166             'view_count': video_data.get('views'),
 167             'comments': comments,
 168         }
 169
 170     def _extract_stream_info(self, stream_info):
 171         broadcast_id = compat_str(stream_info['broadcast_id'])
 172         is_live = stream_info.get('is_live')
 173
 174         formats = []
 175         smil_url = stream_info.get('play_url')
 176         if smil_url:
 177             formats.extend(self._extract_smil_formats(smil_url, broadcast_id))
 178
 179         m3u8_url = stream_info.get('m3u8_url')
 180         if m3u8_url:
 181             formats.extend(self._extract_m3u8_formats(
 182                 m3u8_url, broadcast_id, 'mp4', 'm3u8_native',
 183                 m3u8_id='hls', fatal=False))
 184
 185         rtsp_url = stream_info.get('rtsp_url')
 186         if rtsp_url:
 187             formats.append({
 188                 'url': rtsp_url,
 189                 'format_id': 'rtsp',
 190             })
 191
 192         return {
 193             'id': broadcast_id,
 194             'formats': formats,
 195             'title': stream_info['stream_title'],
 196             'thumbnail': stream_info.get('thumbnail_url'),
 197             'is_live': is_live,
 198         }
 199
 200     def _generate_event_playlist(self, event_data):
 201         event_id = compat_str(event_data['id'])
 202         account_id = compat_str(event_data['owner_account_id'])
 203         feed_root_url = self._API_URL_TEMPLATE % (account_id, event_id) + '/feed.json'
 204
 205         stream_info = event_data.get('stream_info')
 206         if stream_info:
 207             return self._extract_stream_info(stream_info)
 208
 209         last_video = None
 210         for i in itertools.count(1):
 211             if last_video is None:
 212                 info_url = feed_root_url
 213             else:
 214                 info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 215                     root=feed_root_url, id=last_video)
 216             videos_info = self._download_json(
 217                 info_url, event_id, f'Downloading page {i}')['data']
 218             videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 219             if not videos_info:
 220                 break
 221             for v in videos_info:
 222                 v_id = compat_str(v['id'])
 223                 yield self.url_result(
 224                     f'http://livestream.com/accounts/{account_id}/events/{event_id}/videos/{v_id}',
 225                     LivestreamIE, v_id, v.get('caption'))
 226             last_video = videos_info[-1]['id']
 227
 228     def _real_extract(self, url):
 229         mobj = self._match_valid_url(url)
 230         video_id = mobj.group('id')
 231         event = mobj.group('event_id') or mobj.group('event_name')
 232         account = mobj.group('account_id') or mobj.group('account_name')
 233         api_url = f'http://livestream.com/api/accounts/{account}'
 234
 235         if video_id:
 236             video_data = self._download_json(
 237                 f'{api_url}/events/{event}/videos/{video_id}', video_id)
 238             return self._extract_video_info(video_data)
 239         elif event:
 240             event_data = self._download_json(f'{api_url}/events/{event}', None)
 241             return self.playlist_result(
 242                 self._generate_event_playlist(event_data), str(event_data['id']), event_data['full_name'])
 243
 244         account_data = self._download_json(api_url, None)
 245         items = traverse_obj(account_data, (('upcoming_events', 'past_events'), 'data', ...))
 246         return self.playlist_result(
 247             itertools.chain.from_iterable(map(self._generate_event_playlist, items)),
 248             account_data.get('id'), account_data.get('full_name'))
 249
 250
 251 # The original version of Livestream uses a different system
 252 class LivestreamOriginalIE(InfoExtractor):
 253     IE_NAME = 'livestream:original'
 254     _VALID_URL = r'''(?x)https?://original\.livestream\.com/
 255         (?P<user>[^/\?#]+)(?:/(?P<type>video|folder)
 256         (?:(?:\?.*?Id=|/)(?P<id>.*?)(&|$))?)?
 257         '''
 258     _TESTS = [{
 259         'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 260         'info_dict': {
 261             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 262             'ext': 'mp4',
 263             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 264             'duration': 771.301,
 265             'view_count': int,
 266         },
 267     }, {
 268         'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 269         'info_dict': {
 270             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 271         },
 272         'playlist_mincount': 4,
 273     }, {
 274         # live stream
 275         'url': 'http://original.livestream.com/znsbahamas',
 276         'only_matching': True,
 277     }]
 278
 279     def _extract_video_info(self, user, video_id):
 280         api_url = 'http://x%sx.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id=%s' % (user, video_id)
 281         info = self._download_xml(api_url, video_id)
 282
 283         item = info.find('channel').find('item')
 284         title = xpath_text(item, 'title')
 285         media_ns = {'media': 'http://search.yahoo.com/mrss'}
 286         thumbnail_url = xpath_attr(
 287             item, xpath_with_ns('media:thumbnail', media_ns), 'url')
 288         duration = float_or_none(xpath_attr(
 289             item, xpath_with_ns('media:content', media_ns), 'duration'))
 290         ls_ns = {'ls': 'http://api.channel.livestream.com/2.0'}
 291         view_count = int_or_none(xpath_text(
 292             item, xpath_with_ns('ls:viewsCount', ls_ns)))
 293
 294         return {
 295             'id': video_id,
 296             'title': title,
 297             'thumbnail': thumbnail_url,
 298             'duration': duration,
 299             'view_count': view_count,
 300         }
 301
 302     def _extract_video_formats(self, video_data, video_id):
 303         formats = []
 304
 305         progressive_url = video_data.get('progressiveUrl')
 306         if progressive_url:
 307             formats.append({
 308                 'url': progressive_url,
 309                 'format_id': 'http',
 310             })
 311
 312         m3u8_url = video_data.get('httpUrl')
 313         if m3u8_url:
 314             formats.extend(self._extract_m3u8_formats(
 315                 m3u8_url, video_id, 'mp4', 'm3u8_native',
 316                 m3u8_id='hls', fatal=False))
 317
 318         rtsp_url = video_data.get('rtspUrl')
 319         if rtsp_url:
 320             formats.append({
 321                 'url': rtsp_url,
 322                 'format_id': 'rtsp',
 323             })
 324
 325         return formats
 326
 327     def _extract_folder(self, url, folder_id):
 328         webpage = self._download_webpage(url, folder_id)
 329         paths = orderedSet(re.findall(
 330             r'''(?x)(?:
 331                 <li\s+class="folder">\s*<a\s+href="|
 332                 <a\s+href="(?=https?://livestre\.am/)
 333             )([^"]+)"''', webpage))
 334
 335         entries = [{
 336             '_type': 'url',
 337             'url': compat_urlparse.urljoin(url, p),
 338         } for p in paths]
 339
 340         return self.playlist_result(entries, folder_id)
 341
 342     def _real_extract(self, url):
 343         mobj = self._match_valid_url(url)
 344         user = mobj.group('user')
 345         url_type = mobj.group('type')
 346         content_id = mobj.group('id')
 347         if url_type == 'folder':
 348             return self._extract_folder(url, content_id)
 349         else:
 350             # this url is used on mobile devices
 351             stream_url = 'http://x%sx.api.channel.livestream.com/3.0/getstream.json' % user
 352             info = {}
 353             if content_id:
 354                 stream_url += '?id=%s' % content_id
 355                 info = self._extract_video_info(user, content_id)
 356             else:
 357                 content_id = user
 358                 webpage = self._download_webpage(url, content_id)
 359                 info = {
 360                     'title': self._og_search_title(webpage),
 361                     'description': self._og_search_description(webpage),
 362                     'thumbnail': self._search_regex(r'channelLogo\.src\s*=\s*"([^"]+)"', webpage, 'thumbnail', None),
 363                 }
 364             video_data = self._download_json(stream_url, content_id)
 365             is_live = video_data.get('isLive')
 366             info.update({
 367                 'id': content_id,
 368                 'title': info['title'],
 369                 'formats': self._extract_video_formats(video_data, content_id),
 370                 'is_live': is_live,
 371             })
 372             return info
 373
 374
 375 # The server doesn't support HEAD request, the generic extractor can't detect
 376 # the redirection
 377 class LivestreamShortenerIE(InfoExtractor):
 378     IE_NAME = 'livestream:shortener'
 379     IE_DESC = False  # Do not list
 380     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 381
 382     def _real_extract(self, url):
 383         mobj = self._match_valid_url(url)
 384         id = mobj.group('id')
 385         webpage = self._download_webpage(url, id)
 386
 387         return self.url_result(self._og_search_url(webpage))