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