yt_dlp/extractor/livestream.py

   1 import re
   2 import itertools
   3
   4 from .common import InfoExtractor
   5 from ..compat import (
   6     compat_str,
   7     compat_urlparse,
   8 )
   9 from ..utils import (
  10     find_xpath_attr,
  11     xpath_attr,
  12     xpath_with_ns,
  13     xpath_text,
  14     orderedSet,
  15     update_url_query,
  16     int_or_none,
  17     float_or_none,
  18     parse_iso8601,
  19     determine_ext,
  20 )
  21
  22
  23 class LivestreamIE(InfoExtractor):
  24     IE_NAME = 'livestream'
  25     _VALID_URL = r'https?://(?:new\.)?livestream\.com/(?:accounts/(?P<account_id>\d+)|(?P<account_name>[^/]+))/(?:events/(?P<event_id>\d+)|(?P<event_name>[^/]+))(?:/videos/(?P<id>\d+))?'
  26     _EMBED_REGEX = [r'<iframe[^>]+src="(?P<url>https?://(?:new\.)?livestream\.com/[^"]+/player[^"]+)"']
  27
  28     _TESTS = [{
  29         'url': 'http://new.livestream.com/CoheedandCambria/WebsterHall/videos/4719370',
  30         'md5': '53274c76ba7754fb0e8d072716f2292b',
  31         'info_dict': {
  32             'id': '4719370',
  33             'ext': 'mp4',
  34             'title': 'Live from Webster Hall NYC',
  35             'timestamp': 1350008072,
  36             'upload_date': '20121012',
  37             'duration': 5968.0,
  38             'like_count': int,
  39             'view_count': int,
  40             'thumbnail': r're:^http://.*\.jpg$'
  41         }
  42     }, {
  43         'url': 'http://new.livestream.com/tedx/cityenglish',
  44         'info_dict': {
  45             'title': 'TEDCity2.0 (English)',
  46             'id': '2245590',
  47         },
  48         'playlist_mincount': 4,
  49     }, {
  50         'url': 'http://new.livestream.com/chess24/tatasteelchess',
  51         'info_dict': {
  52             'title': 'Tata Steel Chess',
  53             'id': '3705884',
  54         },
  55         'playlist_mincount': 60,
  56     }, {
  57         'url': 'https://new.livestream.com/accounts/362/events/3557232/videos/67864563/player?autoPlay=false&height=360&mute=false&width=640',
  58         'only_matching': True,
  59     }, {
  60         'url': 'http://livestream.com/bsww/concacafbeachsoccercampeonato2015',
  61         'only_matching': True,
  62     }]
  63     _API_URL_TEMPLATE = 'http://livestream.com/api/accounts/%s/events/%s'
  64
  65     def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None):
  66         base_ele = find_xpath_attr(
  67             smil, self._xpath_ns('.//meta', namespace), 'name', 'httpBase')
  68         base = base_ele.get('content') if base_ele is not None else 'http://livestreamvod-f.akamaihd.net/'
  69
  70         formats = []
  71         video_nodes = smil.findall(self._xpath_ns('.//video', namespace))
  72
  73         for vn in video_nodes:
  74             tbr = int_or_none(vn.attrib.get('system-bitrate'), 1000)
  75             furl = (
  76                 update_url_query(compat_urlparse.urljoin(base, vn.attrib['src']), {
  77                     'v': '3.0.3',
  78                     'fp': 'WIN% 14,0,0,145',
  79                 }))
  80             if 'clipBegin' in vn.attrib:
  81                 furl += '&ssek=' + vn.attrib['clipBegin']
  82             formats.append({
  83                 'url': furl,
  84                 'format_id': 'smil_%d' % tbr,
  85                 'ext': 'flv',
  86                 'tbr': tbr,
  87                 'preference': -1000,  # Strictly inferior than all other formats?
  88             })
  89         return formats
  90
  91     def _extract_video_info(self, video_data):
  92         video_id = compat_str(video_data['id'])
  93
  94         FORMAT_KEYS = (
  95             ('sd', 'progressive_url'),
  96             ('hd', 'progressive_url_hd'),
  97         )
  98
  99         formats = []
 100         for format_id, key in FORMAT_KEYS:
 101             video_url = video_data.get(key)
 102             if video_url:
 103                 ext = determine_ext(video_url)
 104                 if ext == 'm3u8':
 105                     continue
 106                 bitrate = int_or_none(self._search_regex(
 107                     r'(\d+)\.%s' % ext, video_url, 'bitrate', default=None))
 108                 formats.append({
 109                     'url': video_url,
 110                     'format_id': format_id,
 111                     'tbr': bitrate,
 112                     'ext': ext,
 113                 })
 114
 115         smil_url = video_data.get('smil_url')
 116         if smil_url:
 117             formats.extend(self._extract_smil_formats(smil_url, video_id, fatal=False))
 118
 119         m3u8_url = video_data.get('m3u8_url')
 120         if m3u8_url:
 121             formats.extend(self._extract_m3u8_formats(
 122                 m3u8_url, video_id, 'mp4', 'm3u8_native',
 123                 m3u8_id='hls', fatal=False))
 124
 125         f4m_url = video_data.get('f4m_url')
 126         if f4m_url:
 127             formats.extend(self._extract_f4m_formats(
 128                 f4m_url, video_id, f4m_id='hds', fatal=False))
 129
 130         comments = [{
 131             'author_id': comment.get('author_id'),
 132             'author': comment.get('author', {}).get('full_name'),
 133             'id': comment.get('id'),
 134             'text': comment['text'],
 135             'timestamp': parse_iso8601(comment.get('created_at')),
 136         } for comment in video_data.get('comments', {}).get('data', [])]
 137
 138         return {
 139             'id': video_id,
 140             'formats': formats,
 141             'title': video_data['caption'],
 142             'description': video_data.get('description'),
 143             'thumbnail': video_data.get('thumbnail_url'),
 144             'duration': float_or_none(video_data.get('duration'), 1000),
 145             'timestamp': parse_iso8601(video_data.get('publish_at')),
 146             'like_count': video_data.get('likes', {}).get('total'),
 147             'comment_count': video_data.get('comments', {}).get('total'),
 148             'view_count': video_data.get('views'),
 149             'comments': comments,
 150         }
 151
 152     def _extract_stream_info(self, stream_info):
 153         broadcast_id = compat_str(stream_info['broadcast_id'])
 154         is_live = stream_info.get('is_live')
 155
 156         formats = []
 157         smil_url = stream_info.get('play_url')
 158         if smil_url:
 159             formats.extend(self._extract_smil_formats(smil_url, broadcast_id))
 160
 161         m3u8_url = stream_info.get('m3u8_url')
 162         if m3u8_url:
 163             formats.extend(self._extract_m3u8_formats(
 164                 m3u8_url, broadcast_id, 'mp4', 'm3u8_native',
 165                 m3u8_id='hls', fatal=False))
 166
 167         rtsp_url = stream_info.get('rtsp_url')
 168         if rtsp_url:
 169             formats.append({
 170                 'url': rtsp_url,
 171                 'format_id': 'rtsp',
 172             })
 173
 174         return {
 175             'id': broadcast_id,
 176             'formats': formats,
 177             'title': stream_info['stream_title'],
 178             'thumbnail': stream_info.get('thumbnail_url'),
 179             'is_live': is_live,
 180         }
 181
 182     def _extract_event(self, event_data):
 183         event_id = compat_str(event_data['id'])
 184         account_id = compat_str(event_data['owner_account_id'])
 185         feed_root_url = self._API_URL_TEMPLATE % (account_id, event_id) + '/feed.json'
 186
 187         stream_info = event_data.get('stream_info')
 188         if stream_info:
 189             return self._extract_stream_info(stream_info)
 190
 191         last_video = None
 192         entries = []
 193         for i in itertools.count(1):
 194             if last_video is None:
 195                 info_url = feed_root_url
 196             else:
 197                 info_url = '{root}?&id={id}&newer=-1&type=video'.format(
 198                     root=feed_root_url, id=last_video)
 199             videos_info = self._download_json(
 200                 info_url, event_id, 'Downloading page {0}'.format(i))['data']
 201             videos_info = [v['data'] for v in videos_info if v['type'] == 'video']
 202             if not videos_info:
 203                 break
 204             for v in videos_info:
 205                 v_id = compat_str(v['id'])
 206                 entries.append(self.url_result(
 207                     'http://livestream.com/accounts/%s/events/%s/videos/%s' % (account_id, event_id, v_id),
 208                     'Livestream', v_id, v.get('caption')))
 209             last_video = videos_info[-1]['id']
 210         return self.playlist_result(entries, event_id, event_data['full_name'])
 211
 212     def _real_extract(self, url):
 213         mobj = self._match_valid_url(url)
 214         video_id = mobj.group('id')
 215         event = mobj.group('event_id') or mobj.group('event_name')
 216         account = mobj.group('account_id') or mobj.group('account_name')
 217         api_url = self._API_URL_TEMPLATE % (account, event)
 218         if video_id:
 219             video_data = self._download_json(
 220                 api_url + '/videos/%s' % video_id, video_id)
 221             return self._extract_video_info(video_data)
 222         else:
 223             event_data = self._download_json(api_url, video_id)
 224             return self._extract_event(event_data)
 225
 226
 227 # The original version of Livestream uses a different system
 228 class LivestreamOriginalIE(InfoExtractor):
 229     IE_NAME = 'livestream:original'
 230     _VALID_URL = r'''(?x)https?://original\.livestream\.com/
 231         (?P<user>[^/\?#]+)(?:/(?P<type>video|folder)
 232         (?:(?:\?.*?Id=|/)(?P<id>.*?)(&|$))?)?
 233         '''
 234     _TESTS = [{
 235         'url': 'http://original.livestream.com/dealbook/video?clipId=pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 236         'info_dict': {
 237             'id': 'pla_8aa4a3f1-ba15-46a4-893b-902210e138fb',
 238             'ext': 'mp4',
 239             'title': 'Spark 1 (BitCoin) with Cameron Winklevoss & Tyler Winklevoss of Winklevoss Capital',
 240             'duration': 771.301,
 241             'view_count': int,
 242         },
 243     }, {
 244         'url': 'https://original.livestream.com/newplay/folder?dirId=a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 245         'info_dict': {
 246             'id': 'a07bf706-d0e4-4e75-a747-b021d84f2fd3',
 247         },
 248         'playlist_mincount': 4,
 249     }, {
 250         # live stream
 251         'url': 'http://original.livestream.com/znsbahamas',
 252         'only_matching': True,
 253     }]
 254
 255     def _extract_video_info(self, user, video_id):
 256         api_url = 'http://x%sx.api.channel.livestream.com/2.0/clipdetails?extendedInfo=true&id=%s' % (user, video_id)
 257         info = self._download_xml(api_url, video_id)
 258
 259         item = info.find('channel').find('item')
 260         title = xpath_text(item, 'title')
 261         media_ns = {'media': 'http://search.yahoo.com/mrss'}
 262         thumbnail_url = xpath_attr(
 263             item, xpath_with_ns('media:thumbnail', media_ns), 'url')
 264         duration = float_or_none(xpath_attr(
 265             item, xpath_with_ns('media:content', media_ns), 'duration'))
 266         ls_ns = {'ls': 'http://api.channel.livestream.com/2.0'}
 267         view_count = int_or_none(xpath_text(
 268             item, xpath_with_ns('ls:viewsCount', ls_ns)))
 269
 270         return {
 271             'id': video_id,
 272             'title': title,
 273             'thumbnail': thumbnail_url,
 274             'duration': duration,
 275             'view_count': view_count,
 276         }
 277
 278     def _extract_video_formats(self, video_data, video_id):
 279         formats = []
 280
 281         progressive_url = video_data.get('progressiveUrl')
 282         if progressive_url:
 283             formats.append({
 284                 'url': progressive_url,
 285                 'format_id': 'http',
 286             })
 287
 288         m3u8_url = video_data.get('httpUrl')
 289         if m3u8_url:
 290             formats.extend(self._extract_m3u8_formats(
 291                 m3u8_url, video_id, 'mp4', 'm3u8_native',
 292                 m3u8_id='hls', fatal=False))
 293
 294         rtsp_url = video_data.get('rtspUrl')
 295         if rtsp_url:
 296             formats.append({
 297                 'url': rtsp_url,
 298                 'format_id': 'rtsp',
 299             })
 300
 301         return formats
 302
 303     def _extract_folder(self, url, folder_id):
 304         webpage = self._download_webpage(url, folder_id)
 305         paths = orderedSet(re.findall(
 306             r'''(?x)(?:
 307                 <li\s+class="folder">\s*<a\s+href="|
 308                 <a\s+href="(?=https?://livestre\.am/)
 309             )([^"]+)"''', webpage))
 310
 311         entries = [{
 312             '_type': 'url',
 313             'url': compat_urlparse.urljoin(url, p),
 314         } for p in paths]
 315
 316         return self.playlist_result(entries, folder_id)
 317
 318     def _real_extract(self, url):
 319         mobj = self._match_valid_url(url)
 320         user = mobj.group('user')
 321         url_type = mobj.group('type')
 322         content_id = mobj.group('id')
 323         if url_type == 'folder':
 324             return self._extract_folder(url, content_id)
 325         else:
 326             # this url is used on mobile devices
 327             stream_url = 'http://x%sx.api.channel.livestream.com/3.0/getstream.json' % user
 328             info = {}
 329             if content_id:
 330                 stream_url += '?id=%s' % content_id
 331                 info = self._extract_video_info(user, content_id)
 332             else:
 333                 content_id = user
 334                 webpage = self._download_webpage(url, content_id)
 335                 info = {
 336                     'title': self._og_search_title(webpage),
 337                     'description': self._og_search_description(webpage),
 338                     'thumbnail': self._search_regex(r'channelLogo\.src\s*=\s*"([^"]+)"', webpage, 'thumbnail', None),
 339                 }
 340             video_data = self._download_json(stream_url, content_id)
 341             is_live = video_data.get('isLive')
 342             info.update({
 343                 'id': content_id,
 344                 'title': info['title'],
 345                 'formats': self._extract_video_formats(video_data, content_id),
 346                 'is_live': is_live,
 347             })
 348             return info
 349
 350
 351 # The server doesn't support HEAD request, the generic extractor can't detect
 352 # the redirection
 353 class LivestreamShortenerIE(InfoExtractor):
 354     IE_NAME = 'livestream:shortener'
 355     IE_DESC = False  # Do not list
 356     _VALID_URL = r'https?://livestre\.am/(?P<id>.+)'
 357
 358     def _real_extract(self, url):
 359         mobj = self._match_valid_url(url)
 360         id = mobj.group('id')
 361         webpage = self._download_webpage(url, id)
 362
 363         return self.url_result(self._og_search_url(webpage))