yt_dlp/extractor/brightcove.py

   1 import base64
   2 import re
   3 import struct
   4 import xml.etree.ElementTree
   5
   6 from .adobepass import AdobePassIE
   7 from .common import InfoExtractor
   8 from ..compat import (
   9     compat_etree_fromstring,
  10     compat_parse_qs,
  11     compat_urlparse,
  12 )
  13 from ..networking.exceptions import HTTPError
  14 from ..utils import (
  15     clean_html,
  16     dict_get,
  17     extract_attributes,
  18     ExtractorError,
  19     find_xpath_attr,
  20     fix_xml_ampersands,
  21     float_or_none,
  22     int_or_none,
  23     js_to_json,
  24     mimetype2ext,
  25     parse_iso8601,
  26     parse_qs,
  27     smuggle_url,
  28     str_or_none,
  29     try_get,
  30     unescapeHTML,
  31     unsmuggle_url,
  32     UnsupportedError,
  33     update_url_query,
  34     url_or_none,
  35 )
  36
  37
  38 class BrightcoveLegacyIE(InfoExtractor):
  39     IE_NAME = 'brightcove:legacy'
  40     _VALID_URL = r'(?:https?://.*brightcove\.com/(services|viewer).*?\?|brightcove:)(?P<query>.*)'
  41
  42     _TESTS = [
  43         {
  44             # From http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/
  45             'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1654948606001&flashID=myExperience&%40videoPlayer=2371591881001',
  46             'md5': '5423e113865d26e40624dce2e4b45d95',
  47             'note': 'Test Brightcove downloads and detection in GenericIE',
  48             'info_dict': {
  49                 'id': '2371591881001',
  50                 'ext': 'mp4',
  51                 'title': 'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”',
  52                 'uploader': '8TV',
  53                 'description': 'md5:a950cc4285c43e44d763d036710cd9cd',
  54                 'timestamp': 1368213670,
  55                 'upload_date': '20130510',
  56                 'uploader_id': '1589608506001',
  57             },
  58             'skip': 'The player has been deactivated by the content owner',
  59         },
  60         {
  61             # From http://medianetwork.oracle.com/video/player/1785452137001
  62             'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=1217746023001&flashID=myPlayer&%40videoPlayer=1785452137001',
  63             'info_dict': {
  64                 'id': '1785452137001',
  65                 'ext': 'flv',
  66                 'title': 'JVMLS 2012: Arrays 2.0 - Opportunities and Challenges',
  67                 'description': 'John Rose speaks at the JVM Language Summit, August 1, 2012.',
  68                 'uploader': 'Oracle',
  69                 'timestamp': 1344975024,
  70                 'upload_date': '20120814',
  71                 'uploader_id': '1460825906',
  72             },
  73             'skip': 'video not playable',
  74         },
  75         {
  76             # From http://mashable.com/2013/10/26/thermoelectric-bracelet-lets-you-control-your-body-temperature/
  77             'url': 'http://c.brightcove.com/services/viewer/federated_f9?&playerID=1265504713001&publisherID=AQ%7E%7E%2CAAABBzUwv1E%7E%2CxP-xFHVUstiMFlNYfvF4G9yFnNaqCw_9&videoID=2750934548001',
  78             'info_dict': {
  79                 'id': '2750934548001',
  80                 'ext': 'mp4',
  81                 'title': 'This Bracelet Acts as a Personal Thermostat',
  82                 'description': 'md5:547b78c64f4112766ccf4e151c20b6a0',
  83                 # 'uploader': 'Mashable',
  84                 'timestamp': 1382041798,
  85                 'upload_date': '20131017',
  86                 'uploader_id': '1130468786001',
  87             },
  88         },
  89         {
  90             # test that the default referer works
  91             # from http://national.ballet.ca/interact/video/Lost_in_Motion_II/
  92             'url': 'http://link.brightcove.com/services/player/bcpid756015033001?bckey=AQ~~,AAAApYJi_Ck~,GxhXCegT1Dp39ilhXuxMJxasUhVNZiil&bctid=2878862109001',
  93             'info_dict': {
  94                 'id': '2878862109001',
  95                 'ext': 'mp4',
  96                 'title': 'Lost in Motion II',
  97                 'description': 'md5:363109c02998fee92ec02211bd8000df',
  98                 'uploader': 'National Ballet of Canada',
  99             },
 100             'skip': 'Video gone',
 101         },
 102         {
 103             # test flv videos served by akamaihd.net
 104             # From http://www.redbull.com/en/bike/stories/1331655643987/replay-uci-dh-world-cup-2014-from-fort-william
 105             'url': 'http://c.brightcove.com/services/viewer/htmlFederated?%40videoPlayer=ref%3Aevent-stream-356&linkBaseURL=http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fvideos%2F1331655630249%2Freplay-uci-fort-william-2014-dh&playerKey=AQ%7E%7E%2CAAAApYJ7UqE%7E%2Cxqr_zXk0I-zzNndy8NlHogrCb5QdyZRf&playerID=1398061561001#__youtubedl_smuggle=%7B%22Referer%22%3A+%22http%3A%2F%2Fwww.redbull.com%2Fen%2Fbike%2Fstories%2F1331655643987%2Freplay-uci-dh-world-cup-2014-from-fort-william%22%7D',
 106             # The md5 checksum changes on each download
 107             'info_dict': {
 108                 'id': '3750436379001',
 109                 'ext': 'flv',
 110                 'title': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals',
 111                 'uploader': 'RBTV Old (do not use)',
 112                 'description': 'UCI MTB World Cup 2014: Fort William, UK - Downhill Finals',
 113                 'timestamp': 1409122195,
 114                 'upload_date': '20140827',
 115                 'uploader_id': '710858724001',
 116             },
 117             'skip': 'Video gone',
 118         },
 119         {
 120             # playlist with 'videoList'
 121             # from http://support.brightcove.com/en/video-cloud/docs/playlist-support-single-video-players
 122             'url': 'http://c.brightcove.com/services/viewer/htmlFederated?playerID=3550052898001&playerKey=AQ%7E%7E%2CAAABmA9XpXk%7E%2C-Kp7jNgisre1fG5OdqpAFUTcs0lP_ZoL',
 123             'info_dict': {
 124                 'title': 'Sealife',
 125                 'id': '3550319591001',
 126             },
 127             'playlist_mincount': 7,
 128             'skip': 'Unsupported URL',
 129         },
 130         {
 131             # playlist with 'playlistTab' (https://github.com/ytdl-org/youtube-dl/issues/9965)
 132             'url': 'http://c.brightcove.com/services/json/experience/runtime/?command=get_programming_for_experience&playerKey=AQ%7E%7E,AAABXlLMdok%7E,NJ4EoMlZ4rZdx9eU1rkMVd8EaYPBBUlg',
 133             'info_dict': {
 134                 'id': '1522758701001',
 135                 'title': 'Lesson 08',
 136             },
 137             'playlist_mincount': 10,
 138             'skip': 'Unsupported URL',
 139         },
 140         {
 141             # playerID inferred from bcpid
 142             # from http://www.un.org/chinese/News/story.asp?NewsID=27724
 143             'url': 'https://link.brightcove.com/services/player/bcpid1722935254001/?bctid=5360463607001&autoStart=false&secureConnections=true&width=650&height=350',
 144             'only_matching': True,  # Tested in GenericIE
 145         }
 146     ]
 147
 148     _WEBPAGE_TESTS = [{
 149         # embedded brightcove video
 150         # it also tests brightcove videos that need to set the 'Referer'
 151         # in the http requests
 152         'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
 153         'info_dict': {
 154             'id': '2765128793001',
 155             'ext': 'mp4',
 156             'title': 'Le cours de bourse : l’analyse technique',
 157             'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
 158             'uploader': 'BFM BUSINESS',
 159         },
 160         'params': {
 161             'skip_download': True,
 162         },
 163         'skip': '404 Not Found',
 164     }, {
 165         # embedded with itemprop embedURL and video id spelled as `idVideo`
 166         'url': 'http://bfmbusiness.bfmtv.com/mediaplayer/chroniques/olivier-delamarche/',
 167         'info_dict': {
 168             'id': '5255628253001',
 169             'ext': 'mp4',
 170             'title': 'md5:37c519b1128915607601e75a87995fc0',
 171             'description': 'md5:37f7f888b434bb8f8cc8dbd4f7a4cf26',
 172             'uploader': 'BFM BUSINESS',
 173             'uploader_id': '876450612001',
 174             'timestamp': 1482255315,
 175             'upload_date': '20161220',
 176         },
 177         'params': {
 178             'skip_download': True,
 179         },
 180         'skip': 'Redirects, page gone',
 181     }, {
 182         # https://github.com/ytdl-org/youtube-dl/issues/2253
 183         'url': 'http://bcove.me/i6nfkrc3',
 184         'md5': '0ba9446db037002366bab3b3eb30c88c',
 185         'info_dict': {
 186             'id': '3101154703001',
 187             'ext': 'mp4',
 188             'title': 'Still no power',
 189             'uploader': 'thestar.com',
 190             'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
 191         },
 192         'skip': 'video gone',
 193     }, {
 194         # https://github.com/ytdl-org/youtube-dl/issues/3541
 195         'url': 'http://www.kijk.nl/sbs6/leermijvrouwenkennen/videos/jqMiXKAYan2S/aflevering-1',
 196         'info_dict': {
 197             'id': '3866516442001',
 198             'ext': 'mp4',
 199             'title': 'Leer mij vrouwen kennen: Aflevering 1',
 200             'description': 'Leer mij vrouwen kennen: Aflevering 1',
 201             'uploader': 'SBS Broadcasting',
 202         },
 203         'skip': 'Restricted to Netherlands, 404 Not Found',
 204         'params': {
 205             'skip_download': True,  # m3u8 download
 206         },
 207     }, {
 208         # Brightcove video in <iframe>
 209         'url': 'http://www.un.org/chinese/News/story.asp?NewsID=27724',
 210         'md5': '36d74ef5e37c8b4a2ce92880d208b968',
 211         'info_dict': {
 212             'id': '5360463607001',
 213             'ext': 'mp4',
 214             'title': '叙利亚失明儿童在废墟上演唱《心跳》  呼吁获得正常童年生活',
 215             'description': '联合国儿童基金会中东和北非区域大使、作曲家扎德·迪拉尼（Zade Dirani）在3月15日叙利亚冲突爆发7周年纪念日之际发布了为叙利亚谱写的歌曲《心跳》（HEARTBEAT），为受到六年冲突影响的叙利亚儿童发出强烈呐喊，呼吁世界做出共同努力，使叙利亚儿童重新获得享有正常童年生活的权利。',
 216             'uploader': 'United Nations',
 217             'uploader_id': '1362235914001',
 218             'timestamp': 1489593889,
 219             'upload_date': '20170315',
 220         },
 221         'skip': '404 Not Found',
 222     }, {
 223         # Brightcove with UUID in videoPlayer
 224         'url': 'http://www8.hp.com/cn/zh/home.html',
 225         'info_dict': {
 226             'id': '5255815316001',
 227             'ext': 'mp4',
 228             'title': 'Sprocket Video - China',
 229             'description': 'Sprocket Video - China',
 230             'uploader': 'HP-Video Gallery',
 231             'timestamp': 1482263210,
 232             'upload_date': '20161220',
 233             'uploader_id': '1107601872001',
 234         },
 235         'params': {
 236             'skip_download': True,  # m3u8 download
 237         },
 238         'skip': 'video rotates...weekly?',
 239     }, {
 240         # Multiple brightcove videos
 241         # https://github.com/ytdl-org/youtube-dl/issues/2283
 242         'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 243         'info_dict': {
 244             'id': 'always-never',
 245             'title': 'Always / Never - The New Yorker',
 246         },
 247         'playlist_count': 3,
 248         'params': {
 249             'extract_flat': False,
 250             'skip_download': True,
 251         },
 252         'skip': 'Redirects, page gone',
 253     }, {
 254         # BrightcoveInPageEmbed embed
 255         'url': 'http://www.geekandsundry.com/tabletop-bonus-wils-final-thoughts-on-dread/',
 256         'info_dict': {
 257             'id': '4238694884001',
 258             'ext': 'flv',
 259             'title': 'Tabletop: Dread, Last Thoughts',
 260             'description': 'Tabletop: Dread, Last Thoughts',
 261             'duration': 51690,
 262         },
 263         'skip': 'Redirects, page gone',
 264     }, {
 265         # Brightcove embed, with no valid 'renditions' but valid 'IOSRenditions'
 266         # This video can't be played in browsers if Flash disabled and UA set to iPhone, which is actually a false alarm
 267         'url': 'https://dl.dropboxusercontent.com/u/29092637/interview.html',
 268         'info_dict': {
 269             'id': '4785848093001',
 270             'ext': 'mp4',
 271             'title': 'The Cardinal Pell Interview',
 272             'description': 'Sky News Contributor Andrew Bolt interviews George Pell in Rome, following the Cardinal\'s evidence before the Royal Commission into Child Abuse. ',
 273             'uploader': 'GlobeCast Australia - GlobeStream',
 274             'uploader_id': '2733773828001',
 275             'upload_date': '20160304',
 276             'timestamp': 1457083087,
 277         },
 278         'params': {
 279             # m3u8 downloads
 280             'skip_download': True,
 281         },
 282         'skip': '404 Not Found',
 283     }, {
 284         # Brightcove embed with whitespace around attribute names
 285         'url': 'http://www.stack.com/video/3167554373001/learn-to-hit-open-three-pointers-with-damian-lillard-s-baseline-drift-drill',
 286         'info_dict': {
 287             'id': '3167554373001',
 288             'ext': 'mp4',
 289             'title': "Learn to Hit Open Three-Pointers With Damian Lillard's Baseline Drift Drill",
 290             'description': 'md5:57bacb0e0f29349de4972bfda3191713',
 291             'uploader_id': '1079349493',
 292             'upload_date': '20140207',
 293             'timestamp': 1391810548,
 294         },
 295         'params': {
 296             'skip_download': True,
 297         },
 298         'skip': '410 Gone',
 299     }]
 300
 301     @classmethod
 302     def _build_brightcove_url(cls, object_str):
 303         """
 304         Build a Brightcove url from a xml string containing
 305         <object class="BrightcoveExperience">{params}</object>
 306         """
 307
 308         # Fix up some stupid HTML, see https://github.com/ytdl-org/youtube-dl/issues/1553
 309         object_str = re.sub(r'(<param(?:\s+[a-zA-Z0-9_]+="[^"]*")*)>',
 310                             lambda m: m.group(1) + '/>', object_str)
 311         # Fix up some stupid XML, see https://github.com/ytdl-org/youtube-dl/issues/1608
 312         object_str = object_str.replace('<--', '<!--')
 313         # remove namespace to simplify extraction
 314         object_str = re.sub(r'(<object[^>]*)(xmlns=".*?")', r'\1', object_str)
 315         object_str = fix_xml_ampersands(object_str)
 316
 317         try:
 318             object_doc = compat_etree_fromstring(object_str.encode('utf-8'))
 319         except xml.etree.ElementTree.ParseError:
 320             return
 321
 322         fv_el = find_xpath_attr(object_doc, './param', 'name', 'flashVars')
 323         if fv_el is not None:
 324             flashvars = dict(
 325                 (k, v[0])
 326                 for k, v in compat_parse_qs(fv_el.attrib['value']).items())
 327         else:
 328             flashvars = {}
 329
 330         data_url = object_doc.attrib.get('data', '')
 331         data_url_params = parse_qs(data_url)
 332
 333         def find_param(name):
 334             if name in flashvars:
 335                 return flashvars[name]
 336             node = find_xpath_attr(object_doc, './param', 'name', name)
 337             if node is not None:
 338                 return node.attrib['value']
 339             return data_url_params.get(name)
 340
 341         params = {}
 342
 343         playerID = find_param('playerID') or find_param('playerId')
 344         if playerID is None:
 345             raise ExtractorError('Cannot find player ID')
 346         params['playerID'] = playerID
 347
 348         playerKey = find_param('playerKey')
 349         # Not all pages define this value
 350         if playerKey is not None:
 351             params['playerKey'] = playerKey
 352         # These fields hold the id of the video
 353         videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID') or find_param('@videoList')
 354         if videoPlayer is not None:
 355             if isinstance(videoPlayer, list):
 356                 videoPlayer = videoPlayer[0]
 357             videoPlayer = videoPlayer.strip()
 358             # UUID is also possible for videoPlayer (e.g.
 359             # http://www.popcornflix.com/hoodies-vs-hooligans/7f2d2b87-bbf2-4623-acfb-ea942b4f01dd
 360             # or http://www8.hp.com/cn/zh/home.html)
 361             if not (re.match(
 362                     r'^(?:\d+|[\da-fA-F]{8}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{12})$',
 363                     videoPlayer) or videoPlayer.startswith('ref:')):
 364                 return None
 365             params['@videoPlayer'] = videoPlayer
 366         linkBase = find_param('linkBaseURL')
 367         if linkBase is not None:
 368             params['linkBaseURL'] = linkBase
 369         return cls._make_brightcove_url(params)
 370
 371     @classmethod
 372     def _build_brightcove_url_from_js(cls, object_js):
 373         # The layout of JS is as follows:
 374         # customBC.createVideo = function (width, height, playerID, playerKey, videoPlayer, VideoRandomID) {
 375         #   // build Brightcove <object /> XML
 376         # }
 377         m = re.search(
 378             r'''(?x)customBC\.createVideo\(
 379                 .*?                                                  # skipping width and height
 380                 ["\'](?P<playerID>\d+)["\']\s*,\s*                   # playerID
 381                 ["\'](?P<playerKey>AQ[^"\']{48})[^"\']*["\']\s*,\s*  # playerKey begins with AQ and is 50 characters
 382                                                                      # in length, however it's appended to itself
 383                                                                      # in places, so truncate
 384                 ["\'](?P<videoID>\d+)["\']                           # @videoPlayer
 385             ''', object_js)
 386         if m:
 387             return cls._make_brightcove_url(m.groupdict())
 388
 389     @classmethod
 390     def _make_brightcove_url(cls, params):
 391         return update_url_query(
 392             'http://c.brightcove.com/services/viewer/htmlFederated', params)
 393
 394     @classmethod
 395     def _extract_brightcove_url(cls, webpage):
 396         """Try to extract the brightcove url from the webpage, returns None
 397         if it can't be found
 398         """
 399         urls = cls._extract_brightcove_urls(webpage)
 400         return urls[0] if urls else None
 401
 402     @classmethod
 403     def _extract_brightcove_urls(cls, webpage):
 404         """Return a list of all Brightcove URLs from the webpage """
 405
 406         url_m = re.search(
 407             r'''(?x)
 408                 <meta\s+
 409                     (?:property|itemprop)=([\'"])(?:og:video|embedURL)\1[^>]+
 410                     content=([\'"])(?P<url>https?://(?:secure|c)\.brightcove.com/(?:(?!\2).)+)\2
 411             ''', webpage)
 412         if url_m:
 413             url = unescapeHTML(url_m.group('url'))
 414             # Some sites don't add it, we can't download with this url, for example:
 415             # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/
 416             if 'playerKey' in url or 'videoId' in url or 'idVideo' in url:
 417                 return [url]
 418
 419         matches = re.findall(
 420             r'''(?sx)<object
 421             (?:
 422                 [^>]+?class=[\'"][^>]*?BrightcoveExperience.*?[\'"] |
 423                 [^>]*?>\s*<param\s+name="movie"\s+value="https?://[^/]*brightcove\.com/
 424             ).+?>\s*</object>''',
 425             webpage)
 426         if matches:
 427             return list(filter(None, [cls._build_brightcove_url(m) for m in matches]))
 428
 429         matches = re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)
 430         if matches:
 431             return list(filter(None, [
 432                 cls._build_brightcove_url_from_js(custom_bc)
 433                 for custom_bc in matches]))
 434         return [src for _, src in re.findall(
 435             r'<iframe[^>]+src=([\'"])((?:https?:)?//link\.brightcove\.com/services/player/(?!\1).+)\1', webpage)]
 436
 437     def _extract_from_webpage(self, url, webpage):
 438         bc_urls = self._extract_brightcove_urls(webpage)
 439         for bc_url in bc_urls:
 440             yield self.url_result(smuggle_url(bc_url, {'Referer': url}), BrightcoveLegacyIE)
 441
 442     def _real_extract(self, url):
 443         url, smuggled_data = unsmuggle_url(url, {})
 444
 445         # Change the 'videoId' and others field to '@videoPlayer'
 446         url = re.sub(r'(?<=[?&])(videoI(d|D)|idVideo|bctid)', '%40videoPlayer', url)
 447         # Change bckey (used by bcove.me urls) to playerKey
 448         url = re.sub(r'(?<=[?&])bckey', 'playerKey', url)
 449         mobj = self._match_valid_url(url)
 450         query_str = mobj.group('query')
 451         query = compat_urlparse.parse_qs(query_str)
 452
 453         videoPlayer = query.get('@videoPlayer')
 454         if videoPlayer:
 455             # We set the original url as the default 'Referer' header
 456             referer = query.get('linkBaseURL', [None])[0] or smuggled_data.get('Referer', url)
 457             video_id = videoPlayer[0]
 458             if 'playerID' not in query:
 459                 mobj = re.search(r'/bcpid(\d+)', url)
 460                 if mobj is not None:
 461                     query['playerID'] = [mobj.group(1)]
 462             publisher_id = query.get('publisherId')
 463             if publisher_id and publisher_id[0].isdigit():
 464                 publisher_id = publisher_id[0]
 465             if not publisher_id:
 466                 player_key = query.get('playerKey')
 467                 if player_key and ',' in player_key[0]:
 468                     player_key = player_key[0]
 469                 else:
 470                     player_id = query.get('playerID')
 471                     if player_id and player_id[0].isdigit():
 472                         headers = {}
 473                         if referer:
 474                             headers['Referer'] = referer
 475                         player_page = self._download_webpage(
 476                             'http://link.brightcove.com/services/player/bcpid' + player_id[0],
 477                             video_id, headers=headers, fatal=False)
 478                         if player_page:
 479                             player_key = self._search_regex(
 480                                 r'<param\s+name="playerKey"\s+value="([\w~,-]+)"',
 481                                 player_page, 'player key', fatal=False)
 482                 if player_key:
 483                     enc_pub_id = player_key.split(',')[1].replace('~', '=')
 484                     publisher_id = struct.unpack('>Q', base64.urlsafe_b64decode(enc_pub_id))[0]
 485             if publisher_id:
 486                 brightcove_new_url = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s' % (publisher_id, video_id)
 487                 if referer:
 488                     brightcove_new_url = smuggle_url(brightcove_new_url, {'referrer': referer})
 489                 return self.url_result(brightcove_new_url, BrightcoveNewIE.ie_key(), video_id)
 490         # TODO: figure out if it's possible to extract playlistId from playerKey
 491         # elif 'playerKey' in query:
 492         #     player_key = query['playerKey']
 493         #     return self._get_playlist_info(player_key[0])
 494         raise UnsupportedError(url)
 495
 496
 497 class BrightcoveNewBaseIE(AdobePassIE):
 498     def _parse_brightcove_metadata(self, json_data, video_id, headers={}):
 499         title = json_data['name'].strip()
 500
 501         formats, subtitles = [], {}
 502         sources = json_data.get('sources') or []
 503         for source in sources:
 504             container = source.get('container')
 505             ext = mimetype2ext(source.get('type'))
 506             src = source.get('src')
 507             if ext == 'm3u8' or container == 'M2TS':
 508                 if not src:
 509                     continue
 510                 fmts, subs = self._extract_m3u8_formats_and_subtitles(
 511                     src, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
 512                 subtitles = self._merge_subtitles(subtitles, subs)
 513             elif ext == 'mpd':
 514                 if not src:
 515                     continue
 516                 fmts, subs = self._extract_mpd_formats_and_subtitles(src, video_id, 'dash', fatal=False)
 517                 subtitles = self._merge_subtitles(subtitles, subs)
 518             else:
 519                 streaming_src = source.get('streaming_src')
 520                 stream_name, app_name = source.get('stream_name'), source.get('app_name')
 521                 if not src and not streaming_src and (not stream_name or not app_name):
 522                     continue
 523                 tbr = float_or_none(source.get('avg_bitrate'), 1000)
 524                 height = int_or_none(source.get('height'))
 525                 width = int_or_none(source.get('width'))
 526                 f = {
 527                     'tbr': tbr,
 528                     'filesize': int_or_none(source.get('size')),
 529                     'container': container,
 530                     'ext': ext or container.lower(),
 531                 }
 532                 if width == 0 and height == 0:
 533                     f.update({
 534                         'vcodec': 'none',
 535                     })
 536                 else:
 537                     f.update({
 538                         'width': width,
 539                         'height': height,
 540                         'vcodec': source.get('codec'),
 541                     })
 542
 543                 def build_format_id(kind):
 544                     format_id = kind
 545                     if tbr:
 546                         format_id += '-%dk' % int(tbr)
 547                     if height:
 548                         format_id += '-%dp' % height
 549                     return format_id
 550
 551                 if src or streaming_src:
 552                     f.update({
 553                         'url': src or streaming_src,
 554                         'format_id': build_format_id('http' if src else 'http-streaming'),
 555                         'source_preference': 0 if src else -1,
 556                     })
 557                 else:
 558                     f.update({
 559                         'url': app_name,
 560                         'play_path': stream_name,
 561                         'format_id': build_format_id('rtmp'),
 562                     })
 563                 fmts = [f]
 564
 565             # https://support.brightcove.com/playback-api-video-fields-reference#key_systems_object
 566             if container == 'WVM' or source.get('key_systems') or ext == 'ism':
 567                 for f in fmts:
 568                     f['has_drm'] = True
 569             formats.extend(fmts)
 570
 571         if not formats:
 572             errors = json_data.get('errors')
 573             if errors:
 574                 error = errors[0]
 575                 self.raise_no_formats(
 576                     error.get('message') or error.get('error_subcode') or error['error_code'], expected=True)
 577
 578         headers.pop('Authorization', None)  # or else http formats will give error 400
 579         for f in formats:
 580             f.setdefault('http_headers', {}).update(headers)
 581
 582         for text_track in json_data.get('text_tracks', []):
 583             if text_track.get('kind') != 'captions':
 584                 continue
 585             text_track_url = url_or_none(text_track.get('src'))
 586             if not text_track_url:
 587                 continue
 588             lang = (str_or_none(text_track.get('srclang'))
 589                     or str_or_none(text_track.get('label')) or 'en').lower()
 590             subtitles.setdefault(lang, []).append({
 591                 'url': text_track_url,
 592             })
 593
 594         is_live = False
 595         duration = float_or_none(json_data.get('duration'), 1000)
 596         if duration is not None and duration <= 0:
 597             is_live = True
 598
 599         common_res = [(160, 90), (320, 180), (480, 720), (640, 360), (768, 432), (1024, 576), (1280, 720), (1366, 768), (1920, 1080)]
 600         thumb_base_url = dict_get(json_data, ('poster', 'thumbnail'))
 601         thumbnails = [{
 602             'url': re.sub(r'\d+x\d+', f'{w}x{h}', thumb_base_url),
 603             'width': w,
 604             'height': h,
 605         } for w, h in common_res] if thumb_base_url else None
 606
 607         return {
 608             'id': video_id,
 609             'title': title,
 610             'description': clean_html(json_data.get('description')),
 611             'thumbnails': thumbnails,
 612             'duration': duration,
 613             'timestamp': parse_iso8601(json_data.get('published_at')),
 614             'uploader_id': json_data.get('account_id'),
 615             'formats': formats,
 616             'subtitles': subtitles,
 617             'tags': json_data.get('tags', []),
 618             'is_live': is_live,
 619         }
 620
 621
 622 class BrightcoveNewIE(BrightcoveNewBaseIE):
 623     IE_NAME = 'brightcove:new'
 624     _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*(?P<content_type>video|playlist)Id=(?P<video_id>\d+|ref:[^&]+)'
 625     _TESTS = [{
 626         'url': 'http://players.brightcove.net/929656772001/e41d32dc-ec74-459e-a845-6c69f7b724ea_default/index.html?videoId=4463358922001',
 627         'md5': 'c8100925723840d4b0d243f7025703be',
 628         'info_dict': {
 629             'id': '4463358922001',
 630             'ext': 'mp4',
 631             'title': 'Meet the man behind Popcorn Time',
 632             'description': 'md5:eac376a4fe366edc70279bfb681aea16',
 633             'duration': 165.768,
 634             'timestamp': 1441391203,
 635             'upload_date': '20150904',
 636             'uploader_id': '929656772001',
 637             'formats': 'mincount:20',
 638         },
 639         'skip': '404 Not Found',
 640     }, {
 641         # with rtmp streams
 642         'url': 'http://players.brightcove.net/4036320279001/5d112ed9-283f-485f-a7f9-33f42e8bc042_default/index.html?videoId=4279049078001',
 643         'info_dict': {
 644             'id': '4279049078001',
 645             'ext': 'mp4',
 646             'title': 'Titansgrave: Chapter 0',
 647             'description': 'Titansgrave: Chapter 0',
 648             'duration': 1242.058,
 649             'timestamp': 1433556729,
 650             'upload_date': '20150606',
 651             'uploader_id': '4036320279001',
 652             'formats': 'mincount:39',
 653         },
 654         'params': {
 655             # m3u8 download
 656             'skip_download': True,
 657         }
 658     }, {
 659         # playlist stream
 660         'url': 'https://players.brightcove.net/1752604059001/S13cJdUBz_default/index.html?playlistId=5718313430001',
 661         'info_dict': {
 662             'id': '5718313430001',
 663             'title': 'No Audio Playlist',
 664         },
 665         'playlist_count': 7,
 666         'params': {
 667             # m3u8 download
 668             'skip_download': True,
 669         }
 670     }, {
 671         'url': 'http://players.brightcove.net/5690807595001/HyZNerRl7_default/index.html?playlistId=5743160747001',
 672         'only_matching': True,
 673     }, {
 674         # ref: prefixed video id
 675         'url': 'http://players.brightcove.net/3910869709001/21519b5c-4b3b-4363-accb-bdc8f358f823_default/index.html?videoId=ref:7069442',
 676         'only_matching': True,
 677     }, {
 678         # non numeric ref: prefixed video id
 679         'url': 'http://players.brightcove.net/710858724001/default_default/index.html?videoId=ref:event-stream-356',
 680         'only_matching': True,
 681     }, {
 682         # unavailable video without message but with error_code
 683         'url': 'http://players.brightcove.net/1305187701/c832abfb-641b-44eb-9da0-2fe76786505f_default/index.html?videoId=4377407326001',
 684         'only_matching': True,
 685     }]
 686
 687     _WEBPAGE_TESTS = [{
 688         # brightcove player url embed
 689         'url': 'https://nbc-2.com/weather/forecast/2022/11/16/forecast-warmest-day-of-the-week/',
 690         'md5': '2934d5372b354d27083ccf8575dbfee2',
 691         'info_dict': {
 692             'id': '6315650313112',
 693             'title': 'First Alert Forecast: November 15, 2022',
 694             'ext': 'mp4',
 695             'tags': ['nbc2', 'forecast'],
 696             'uploader_id': '6146886170001',
 697             'thumbnail': r're:^https?://.*\.jpg$',
 698             'timestamp': 1668574571,
 699             'duration': 233.375,
 700             'upload_date': '20221116',
 701         },
 702     }, {
 703         # embedded with video tag only
 704         'url': 'https://www.gooddishtv.com/tiktok-rapping-chef-mr-pyrex',
 705         'info_dict': {
 706             'id': 'tiktok-rapping-chef-mr-pyrex',
 707             'title': 'TikTok\'s Rapping Chef Makes Jambalaya for the Hosts',
 708             'thumbnail': r're:^https?://.*\.jpg$',
 709             'age_limit': 0,
 710             'description': 'Just in time for Mardi Gras',
 711         },
 712         'playlist': [{
 713             'info_dict': {
 714                 'id': '6299189544001',
 715                 'ext': 'mp4',
 716                 'title': 'TGD_01-032_5',
 717                 'thumbnail': r're:^https?://.*\.jpg$',
 718                 'tags': [],
 719                 'timestamp': 1646078943,
 720                 'uploader_id': '1569565978001',
 721                 'upload_date': '20220228',
 722                 'duration': 217.195,
 723             },
 724         }, {
 725             'info_dict': {
 726                 'id': '6305565995112',
 727                 'ext': 'mp4',
 728                 'title': 'TGD 01-087 (Airs 05.25.22)_Segment 5',
 729                 'thumbnail': r're:^https?://.*\.jpg$',
 730                 'tags': [],
 731                 'timestamp': 1651604591,
 732                 'uploader_id': '1569565978001',
 733                 'upload_date': '20220503',
 734                 'duration': 310.421,
 735             },
 736         }],
 737     }, {
 738         # Brightcove:new type [2].
 739         'url': 'http://www.delawaresportszone.com/video-st-thomas-more-earns-first-trip-to-basketball-semis',
 740         'md5': '2b35148fcf48da41c9fb4591650784f3',
 741         'info_dict': {
 742             'id': '5348741021001',
 743             'ext': 'mp4',
 744             'upload_date': '20170306',
 745             'uploader_id': '4191638492001',
 746             'timestamp': 1488769918,
 747             'title': 'VIDEO:  St. Thomas More earns first trip to basketball semis',
 748         },
 749         'skip': '404 Not Found',
 750     }, {
 751         # Alternative brightcove <video> attributes
 752         'url': 'http://www.programme-tv.net/videos/extraits/81095-guillaume-canet-evoque-les-rumeurs-d-infidelite-de-marion-cotillard-avec-brad-pitt-dans-vivement-dimanche/',
 753         'info_dict': {
 754             'id': '81095-guillaume-canet-evoque-les-rumeurs-d-infidelite-de-marion-cotillard-avec-brad-pitt-dans-vivement-dimanche',
 755             'title': "Guillaume Canet évoque les rumeurs d'infidélité de Marion Cotillard avec Brad Pitt dans Vivement Dimanche, Extraits : toutes les vidéos avec Télé-Loisirs",
 756         },
 757         'playlist': [{
 758             'md5': '732d22ba3d33f2f3fc253c39f8f36523',
 759             'info_dict': {
 760                 'id': '5311302538001',
 761                 'ext': 'mp4',
 762                 'title': "Guillaume Canet évoque les rumeurs d'infidélité de Marion Cotillard avec Brad Pitt dans Vivement Dimanche",
 763                 'description': "Guillaume Canet évoque les rumeurs d'infidélité de Marion Cotillard avec Brad Pitt dans Vivement Dimanche (France 2, 5 février 2017)",
 764                 'timestamp': 1486321708,
 765                 'upload_date': '20170205',
 766                 'uploader_id': '800000640001',
 767             },
 768             'only_matching': True,
 769         }],
 770         'skip': '404 Not Found',
 771     }, {
 772         # Brightcove URL in single quotes
 773         'url': 'http://www.sportsnet.ca/baseball/mlb/sn-presents-russell-martin-world-citizen/',
 774         'md5': '4ae374f1f8b91c889c4b9203c8c752af',
 775         'info_dict': {
 776             'id': '4255764656001',
 777             'ext': 'mp4',
 778             'title': 'SN Presents: Russell Martin, World Citizen',
 779             'description': 'To understand why he was the Toronto Blue Jays’ top off-season priority is to appreciate his background and upbringing in Montreal, where he first developed his baseball skills. Written and narrated by Stephen Brunt.',
 780             'uploader': 'Rogers Sportsnet',
 781             'uploader_id': '1704050871',
 782             'upload_date': '20150525',
 783             'timestamp': 1432570283,
 784         },
 785         'skip': 'Page no longer has URL, now has javascript',
 786     }]
 787
 788     @staticmethod
 789     def _extract_url(ie, webpage):
 790         urls = BrightcoveNewIE._extract_brightcove_urls(ie, webpage)
 791         return urls[0] if urls else None
 792
 793     @staticmethod
 794     def _extract_brightcove_urls(ie, webpage):
 795         # Reference:
 796         # 1. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideoiniframe
 797         # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#tag
 798         # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
 799         # 4. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/in-page-embed-player-implementation.html
 800         # 5. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
 801
 802         entries = []
 803
 804         # Look for iframe embeds [1]
 805         for _, url in re.findall(
 806                 r'<iframe[^>]+src=(["\'])((?:https?:)?//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage):
 807             entries.append(url if url.startswith('http') else 'http:' + url)
 808
 809         # Look for <video> tags [2] and embed_in_page embeds [3]
 810         # [2] looks like:
 811         for video, script_tag, account_id, player_id, embed in re.findall(
 812                 r'''(?isx)
 813                     (<video(?:-js)?\s+[^>]*\bdata-video-id\s*=\s*['"]?[^>]+>)
 814                     (?:.*?
 815                         (<script[^>]+
 816                             src=["\'](?:https?:)?//players\.brightcove\.net/
 817                             (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
 818                         )
 819                     )?
 820                 ''', webpage):
 821             attrs = extract_attributes(video)
 822
 823             # According to examples from [4] it's unclear whether video id
 824             # may be optional and what to do when it is
 825             video_id = attrs.get('data-video-id')
 826             if not video_id:
 827                 continue
 828
 829             account_id = account_id or attrs.get('data-account')
 830             if not account_id:
 831                 continue
 832
 833             player_id = player_id or attrs.get('data-player') or 'default'
 834             embed = embed or attrs.get('data-embed') or 'default'
 835
 836             bc_url = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' % (
 837                 account_id, player_id, embed, video_id)
 838
 839             # Some brightcove videos may be embedded with video tag only and
 840             # without script tag or any mentioning of brightcove at all. Such
 841             # embeds are considered ambiguous since they are matched based only
 842             # on data-video-id and data-account attributes and in the wild may
 843             # not be brightcove embeds at all. Let's check reconstructed
 844             # brightcove URLs in case of such embeds and only process valid
 845             # ones. By this we ensure there is indeed a brightcove embed.
 846             if not script_tag and not ie._is_valid_url(
 847                     bc_url, video_id, 'possible brightcove video'):
 848                 continue
 849
 850             entries.append(bc_url)
 851
 852         return entries
 853
 854     def _extract_from_webpage(self, url, webpage):
 855         bc_urls = self._extract_brightcove_urls(self, webpage)
 856         for bc_url in bc_urls:
 857             yield self.url_result(smuggle_url(bc_url, {'referrer': url}), BrightcoveNewIE)
 858
 859     def _real_extract(self, url):
 860         url, smuggled_data = unsmuggle_url(url, {})
 861         self._initialize_geo_bypass({
 862             'countries': smuggled_data.get('geo_countries'),
 863             'ip_blocks': smuggled_data.get('geo_ip_blocks'),
 864         })
 865
 866         account_id, player_id, embed, content_type, video_id = self._match_valid_url(url).groups()
 867
 868         policy_key_id = '%s_%s' % (account_id, player_id)
 869         policy_key = self.cache.load('brightcove', policy_key_id)
 870         policy_key_extracted = False
 871         store_pk = lambda x: self.cache.store('brightcove', policy_key_id, x)
 872
 873         def extract_policy_key():
 874             base_url = 'http://players.brightcove.net/%s/%s_%s/' % (account_id, player_id, embed)
 875             config = self._download_json(
 876                 base_url + 'config.json', video_id, fatal=False) or {}
 877             policy_key = try_get(
 878                 config, lambda x: x['video_cloud']['policy_key'])
 879             if not policy_key:
 880                 webpage = self._download_webpage(
 881                     base_url + 'index.min.js', video_id)
 882
 883                 catalog = self._search_regex(
 884                     r'catalog\(({.+?})\);', webpage, 'catalog', default=None)
 885                 if catalog:
 886                     catalog = self._parse_json(
 887                         js_to_json(catalog), video_id, fatal=False)
 888                     if catalog:
 889                         policy_key = catalog.get('policyKey')
 890
 891                 if not policy_key:
 892                     policy_key = self._search_regex(
 893                         r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1',
 894                         webpage, 'policy key', group='pk')
 895
 896             store_pk(policy_key)
 897             return policy_key
 898
 899         token = smuggled_data.get('token')
 900         api_url = f'https://{"edge-auth" if token else "edge"}.api.brightcove.com/playback/v1/accounts/{account_id}/{content_type}s/{video_id}'
 901         headers = {'Authorization': f'Bearer {token}'} if token else {}
 902         referrer = smuggled_data.get('referrer')  # XXX: notice the spelling/case of the key
 903         if referrer:
 904             headers.update({
 905                 'Referer': referrer,
 906                 'Origin': re.search(r'https?://[^/]+', referrer).group(0),
 907             })
 908
 909         for _ in range(2):
 910             if not policy_key:
 911                 policy_key = extract_policy_key()
 912                 policy_key_extracted = True
 913             headers['Accept'] = 'application/json;pk=%s' % policy_key
 914             try:
 915                 json_data = self._download_json(api_url, video_id, headers=headers)
 916                 break
 917             except ExtractorError as e:
 918                 if isinstance(e.cause, HTTPError) and e.cause.status in (401, 403):
 919                     json_data = self._parse_json(e.cause.response.read().decode(), video_id)[0]
 920                     message = json_data.get('message') or json_data['error_code']
 921                     if json_data.get('error_subcode') == 'CLIENT_GEO':
 922                         self.raise_geo_restricted(msg=message)
 923                     elif json_data.get('error_code') == 'INVALID_POLICY_KEY' and not policy_key_extracted:
 924                         policy_key = None
 925                         store_pk(None)
 926                         continue
 927                     raise ExtractorError(message, expected=True)
 928                 raise
 929
 930         errors = json_data.get('errors')
 931         if errors and errors[0].get('error_subcode') == 'TVE_AUTH':
 932             custom_fields = json_data['custom_fields']
 933             tve_token = self._extract_mvpd_auth(
 934                 smuggled_data['source_url'], video_id,
 935                 custom_fields['bcadobepassrequestorid'],
 936                 custom_fields['bcadobepassresourceid'])
 937             json_data = self._download_json(
 938                 api_url, video_id, headers={
 939                     'Accept': 'application/json;pk=%s' % policy_key
 940                 }, query={
 941                     'tveToken': tve_token,
 942                 })
 943
 944         if content_type == 'playlist':
 945             return self.playlist_result(
 946                 [self._parse_brightcove_metadata(vid, vid.get('id'), headers)
 947                  for vid in json_data.get('videos', []) if vid.get('id')],
 948                 json_data.get('id'), json_data.get('name'),
 949                 json_data.get('description'))
 950
 951         return self._parse_brightcove_metadata(
 952             json_data, video_id, headers=headers)