yt_dlp/extractor/bbc.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import functools
   5 import itertools
   6 import json
   7 import re
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_etree_Element,
  12     compat_HTTPError,
  13     compat_str,
  14     compat_urlparse,
  15 )
  16 from ..utils import (
  17     ExtractorError,
  18     OnDemandPagedList,
  19     clean_html,
  20     dict_get,
  21     float_or_none,
  22     get_element_by_class,
  23     int_or_none,
  24     js_to_json,
  25     parse_duration,
  26     parse_iso8601,
  27     parse_qs,
  28     strip_or_none,
  29     try_get,
  30     unescapeHTML,
  31     unified_timestamp,
  32     url_or_none,
  33     urlencode_postdata,
  34     urljoin,
  35 )
  36
  37
  38 class BBCCoUkIE(InfoExtractor):
  39     IE_NAME = 'bbc.co.uk'
  40     IE_DESC = 'BBC iPlayer'
  41     _ID_REGEX = r'(?:[pbm][\da-z]{7}|w[\da-z]{7,14})'
  42     _VALID_URL = r'''(?x)
  43                     https?://
  44                         (?:www\.)?bbc\.co\.uk/
  45                         (?:
  46                             programmes/(?!articles/)|
  47                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
  48                             music/(?:clips|audiovideo/popular)[/#]|
  49                             radio/player/|
  50                             sounds/play/|
  51                             events/[^/]+/play/[^/]+/
  52                         )
  53                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
  54                     ''' % _ID_REGEX
  55
  56     _LOGIN_URL = 'https://account.bbc.com/signin'
  57     _NETRC_MACHINE = 'bbc'
  58
  59     _MEDIA_SELECTOR_URL_TEMPL = 'https://open.live.bbc.co.uk/mediaselector/6/select/version/2.0/mediaset/%s/vpid/%s'
  60     _MEDIA_SETS = [
  61         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  62         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  63         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  64         'iptv-all',
  65         'pc',
  66     ]
  67
  68     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  69
  70     _TESTS = [
  71         {
  72             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  73             'info_dict': {
  74                 'id': 'b039d07m',
  75                 'ext': 'flv',
  76                 'title': 'Kaleidoscope, Leonard Cohen',
  77                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  78             },
  79             'params': {
  80                 # rtmp download
  81                 'skip_download': True,
  82             }
  83         },
  84         {
  85             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  86             'info_dict': {
  87                 'id': 'b00yng1d',
  88                 'ext': 'flv',
  89                 'title': 'The Man in Black: Series 3: The Printed Name',
  90                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  91                 'duration': 1800,
  92             },
  93             'params': {
  94                 # rtmp download
  95                 'skip_download': True,
  96             },
  97             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  98         },
  99         {
 100             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
 101             'info_dict': {
 102                 'id': 'b00yng1d',
 103                 'ext': 'flv',
 104                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
 105                 'description': 'Emma Willis and Marvin Humes present the fifth set of blind auditions in the singing competition, as the coaches continue to build their teams based on voice alone.',
 106                 'duration': 5100,
 107             },
 108             'params': {
 109                 # rtmp download
 110                 'skip_download': True,
 111             },
 112             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 113         },
 114         {
 115             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
 116             'info_dict': {
 117                 'id': 'b03k3pb7',
 118                 'ext': 'flv',
 119                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 120                 'description': '2. Invasion',
 121                 'duration': 3600,
 122             },
 123             'params': {
 124                 # rtmp download
 125                 'skip_download': True,
 126             },
 127             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 128         }, {
 129             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 130             'info_dict': {
 131                 'id': 'b04v209v',
 132                 'ext': 'flv',
 133                 'title': 'Pete Tong, The Essential New Tune Special',
 134                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 135                 'duration': 10800,
 136             },
 137             'params': {
 138                 # rtmp download
 139                 'skip_download': True,
 140             },
 141             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 142         }, {
 143             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
 144             'note': 'Audio',
 145             'info_dict': {
 146                 'id': 'p022h44j',
 147                 'ext': 'flv',
 148                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 149                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 150                 'duration': 227,
 151             },
 152             'params': {
 153                 # rtmp download
 154                 'skip_download': True,
 155             }
 156         }, {
 157             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 158             'note': 'Video',
 159             'info_dict': {
 160                 'id': 'p025c103',
 161                 'ext': 'flv',
 162                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 163                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 164                 'duration': 226,
 165             },
 166             'params': {
 167                 # rtmp download
 168                 'skip_download': True,
 169             }
 170         }, {
 171             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 172             'info_dict': {
 173                 'id': 'p02n76xf',
 174                 'ext': 'flv',
 175                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 176                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 177                 'duration': 3540,
 178             },
 179             'params': {
 180                 # rtmp download
 181                 'skip_download': True,
 182             },
 183             'skip': 'geolocation',
 184         }, {
 185             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 186             'info_dict': {
 187                 'id': 'b05zmgw1',
 188                 'ext': 'flv',
 189                 'description': 'Kirsty Wark and Morgan Quaintance visit the Royal Academy as it prepares for its annual artistic extravaganza, meeting people who have come together to make the show unique.',
 190                 'title': 'Royal Academy Summer Exhibition',
 191                 'duration': 3540,
 192             },
 193             'params': {
 194                 # rtmp download
 195                 'skip_download': True,
 196             },
 197             'skip': 'geolocation',
 198         }, {
 199             # iptv-all mediaset fails with geolocation however there is no geo restriction
 200             # for this programme at all
 201             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
 202             'info_dict': {
 203                 'id': 'b06rkms3',
 204                 'ext': 'flv',
 205                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 206                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 207             },
 208             'params': {
 209                 # rtmp download
 210                 'skip_download': True,
 211             },
 212             'skip': 'Now it\'s really geo-restricted',
 213         }, {
 214             # compact player (https://github.com/ytdl-org/youtube-dl/issues/8147)
 215             'url': 'http://www.bbc.co.uk/programmes/p028bfkf/player',
 216             'info_dict': {
 217                 'id': 'p028bfkj',
 218                 'ext': 'flv',
 219                 'title': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 220                 'description': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 221             },
 222             'params': {
 223                 # rtmp download
 224                 'skip_download': True,
 225             },
 226         }, {
 227             'url': 'https://www.bbc.co.uk/sounds/play/m0007jzb',
 228             'note': 'Audio',
 229             'info_dict': {
 230                 'id': 'm0007jz9',
 231                 'ext': 'mp4',
 232                 'title': 'BBC Proms, 2019, Prom 34: West–Eastern Divan Orchestra',
 233                 'description': "Live BBC Proms. West–Eastern Divan Orchestra with Daniel Barenboim and Martha Argerich.",
 234                 'duration': 9840,
 235             },
 236             'params': {
 237                 # rtmp download
 238                 'skip_download': True,
 239             }
 240         }, {
 241             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 242             'only_matching': True,
 243         }, {
 244             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 245             'only_matching': True,
 246         }, {
 247             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 248             'only_matching': True,
 249         }, {
 250             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 251             'only_matching': True,
 252         }, {
 253             'url': 'https://www.bbc.co.uk/music/audiovideo/popular#p055bc55',
 254             'only_matching': True,
 255         }, {
 256             'url': 'http://www.bbc.co.uk/programmes/w3csv1y9',
 257             'only_matching': True,
 258         }, {
 259             'url': 'https://www.bbc.co.uk/programmes/m00005xn',
 260             'only_matching': True,
 261         }, {
 262             'url': 'https://www.bbc.co.uk/programmes/w172w4dww1jqt5s',
 263             'only_matching': True,
 264         }]
 265
 266     def _login(self):
 267         username, password = self._get_login_info()
 268         if username is None:
 269             return
 270
 271         login_page = self._download_webpage(
 272             self._LOGIN_URL, None, 'Downloading signin page')
 273
 274         login_form = self._hidden_inputs(login_page)
 275
 276         login_form.update({
 277             'username': username,
 278             'password': password,
 279         })
 280
 281         post_url = urljoin(self._LOGIN_URL, self._search_regex(
 282             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 283             'post url', default=self._LOGIN_URL, group='url'))
 284
 285         response, urlh = self._download_webpage_handle(
 286             post_url, None, 'Logging in', data=urlencode_postdata(login_form),
 287             headers={'Referer': self._LOGIN_URL})
 288
 289         if self._LOGIN_URL in urlh.geturl():
 290             error = clean_html(get_element_by_class('form-message', response))
 291             if error:
 292                 raise ExtractorError(
 293                     'Unable to login: %s' % error, expected=True)
 294             raise ExtractorError('Unable to log in')
 295
 296     def _real_initialize(self):
 297         self._login()
 298
 299     class MediaSelectionError(Exception):
 300         def __init__(self, id):
 301             self.id = id
 302
 303     def _extract_asx_playlist(self, connection, programme_id):
 304         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 305         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 306
 307     def _extract_items(self, playlist):
 308         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 309
 310     def _extract_medias(self, media_selection):
 311         error = media_selection.get('result')
 312         if error:
 313             raise BBCCoUkIE.MediaSelectionError(error)
 314         return media_selection.get('media') or []
 315
 316     def _extract_connections(self, media):
 317         return media.get('connection') or []
 318
 319     def _get_subtitles(self, media, programme_id):
 320         subtitles = {}
 321         for connection in self._extract_connections(media):
 322             cc_url = url_or_none(connection.get('href'))
 323             if not cc_url:
 324                 continue
 325             captions = self._download_xml(
 326                 cc_url, programme_id, 'Downloading captions', fatal=False)
 327             if not isinstance(captions, compat_etree_Element):
 328                 continue
 329             subtitles['en'] = [
 330                 {
 331                     'url': connection.get('href'),
 332                     'ext': 'ttml',
 333                 },
 334             ]
 335             break
 336         return subtitles
 337
 338     def _raise_extractor_error(self, media_selection_error):
 339         raise ExtractorError(
 340             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 341             expected=True)
 342
 343     def _download_media_selector(self, programme_id):
 344         last_exception = None
 345         for media_set in self._MEDIA_SETS:
 346             try:
 347                 return self._download_media_selector_url(
 348                     self._MEDIA_SELECTOR_URL_TEMPL % (media_set, programme_id), programme_id)
 349             except BBCCoUkIE.MediaSelectionError as e:
 350                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 351                     last_exception = e
 352                     continue
 353                 self._raise_extractor_error(e)
 354         self._raise_extractor_error(last_exception)
 355
 356     def _download_media_selector_url(self, url, programme_id=None):
 357         media_selection = self._download_json(
 358             url, programme_id, 'Downloading media selection JSON',
 359             expected_status=(403, 404))
 360         return self._process_media_selector(media_selection, programme_id)
 361
 362     def _process_media_selector(self, media_selection, programme_id):
 363         formats = []
 364         subtitles = None
 365         urls = []
 366
 367         for media in self._extract_medias(media_selection):
 368             kind = media.get('kind')
 369             if kind in ('video', 'audio'):
 370                 bitrate = int_or_none(media.get('bitrate'))
 371                 encoding = media.get('encoding')
 372                 width = int_or_none(media.get('width'))
 373                 height = int_or_none(media.get('height'))
 374                 file_size = int_or_none(media.get('media_file_size'))
 375                 for connection in self._extract_connections(media):
 376                     href = connection.get('href')
 377                     if href in urls:
 378                         continue
 379                     if href:
 380                         urls.append(href)
 381                     conn_kind = connection.get('kind')
 382                     protocol = connection.get('protocol')
 383                     supplier = connection.get('supplier')
 384                     transfer_format = connection.get('transferFormat')
 385                     format_id = supplier or conn_kind or protocol
 386                     # ASX playlist
 387                     if supplier == 'asx':
 388                         for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 389                             formats.append({
 390                                 'url': ref,
 391                                 'format_id': 'ref%s_%s' % (i, format_id),
 392                             })
 393                     elif transfer_format == 'dash':
 394                         formats.extend(self._extract_mpd_formats(
 395                             href, programme_id, mpd_id=format_id, fatal=False))
 396                     elif transfer_format == 'hls':
 397                         formats.extend(self._extract_m3u8_formats(
 398                             href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 399                             m3u8_id=format_id, fatal=False))
 400                     elif transfer_format == 'hds':
 401                         formats.extend(self._extract_f4m_formats(
 402                             href, programme_id, f4m_id=format_id, fatal=False))
 403                     else:
 404                         if not supplier and bitrate:
 405                             format_id += '-%d' % bitrate
 406                         fmt = {
 407                             'format_id': format_id,
 408                             'filesize': file_size,
 409                         }
 410                         if kind == 'video':
 411                             fmt.update({
 412                                 'width': width,
 413                                 'height': height,
 414                                 'tbr': bitrate,
 415                                 'vcodec': encoding,
 416                             })
 417                         else:
 418                             fmt.update({
 419                                 'abr': bitrate,
 420                                 'acodec': encoding,
 421                                 'vcodec': 'none',
 422                             })
 423                         if protocol in ('http', 'https'):
 424                             # Direct link
 425                             fmt.update({
 426                                 'url': href,
 427                             })
 428                         elif protocol == 'rtmp':
 429                             application = connection.get('application', 'ondemand')
 430                             auth_string = connection.get('authString')
 431                             identifier = connection.get('identifier')
 432                             server = connection.get('server')
 433                             fmt.update({
 434                                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 435                                 'play_path': identifier,
 436                                 'app': '%s?%s' % (application, auth_string),
 437                                 'page_url': 'http://www.bbc.co.uk',
 438                                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 439                                 'rtmp_live': False,
 440                                 'ext': 'flv',
 441                             })
 442                         else:
 443                             continue
 444                         formats.append(fmt)
 445             elif kind == 'captions':
 446                 subtitles = self.extract_subtitles(media, programme_id)
 447         return formats, subtitles
 448
 449     def _download_playlist(self, playlist_id):
 450         try:
 451             playlist = self._download_json(
 452                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 453                 playlist_id, 'Downloading playlist JSON')
 454             formats = []
 455             subtitles = {}
 456
 457             for version in playlist.get('allAvailableVersions', []):
 458                 smp_config = version['smpConfig']
 459                 title = smp_config['title']
 460                 description = smp_config['summary']
 461                 for item in smp_config['items']:
 462                     kind = item['kind']
 463                     if kind not in ('programme', 'radioProgramme'):
 464                         continue
 465                     programme_id = item.get('vpid')
 466                     duration = int_or_none(item.get('duration'))
 467                     version_formats, version_subtitles = self._download_media_selector(programme_id)
 468                     types = version['types']
 469                     for f in version_formats:
 470                         f['format_note'] = ', '.join(types)
 471                         if any('AudioDescribed' in x for x in types):
 472                             f['language_preference'] = -10
 473                     formats += version_formats
 474                     for tag, subformats in (version_subtitles or {}).items():
 475                         subtitles.setdefault(tag, []).extend(subformats)
 476
 477             return programme_id, title, description, duration, formats, subtitles
 478         except ExtractorError as ee:
 479             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 480                 raise
 481
 482         # fallback to legacy playlist
 483         return self._process_legacy_playlist(playlist_id)
 484
 485     def _process_legacy_playlist_url(self, url, display_id):
 486         playlist = self._download_legacy_playlist_url(url, display_id)
 487         return self._extract_from_legacy_playlist(playlist, display_id)
 488
 489     def _process_legacy_playlist(self, playlist_id):
 490         return self._process_legacy_playlist_url(
 491             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 492
 493     def _download_legacy_playlist_url(self, url, playlist_id=None):
 494         return self._download_xml(
 495             url, playlist_id, 'Downloading legacy playlist XML')
 496
 497     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 498         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 499         if no_items is not None:
 500             reason = no_items.get('reason')
 501             if reason == 'preAvailability':
 502                 msg = 'Episode %s is not yet available' % playlist_id
 503             elif reason == 'postAvailability':
 504                 msg = 'Episode %s is no longer available' % playlist_id
 505             elif reason == 'noMedia':
 506                 msg = 'Episode %s is not currently available' % playlist_id
 507             else:
 508                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 509             raise ExtractorError(msg, expected=True)
 510
 511         for item in self._extract_items(playlist):
 512             kind = item.get('kind')
 513             if kind not in ('programme', 'radioProgramme'):
 514                 continue
 515             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 516             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 517             description = description_el.text if description_el is not None else None
 518
 519             def get_programme_id(item):
 520                 def get_from_attributes(item):
 521                     for p in ('identifier', 'group'):
 522                         value = item.get(p)
 523                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 524                             return value
 525                 get_from_attributes(item)
 526                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 527                 if mediator is not None:
 528                     return get_from_attributes(mediator)
 529
 530             programme_id = get_programme_id(item)
 531             duration = int_or_none(item.get('duration'))
 532
 533             if programme_id:
 534                 formats, subtitles = self._download_media_selector(programme_id)
 535             else:
 536                 formats, subtitles = self._process_media_selector(item, playlist_id)
 537                 programme_id = playlist_id
 538
 539         return programme_id, title, description, duration, formats, subtitles
 540
 541     def _real_extract(self, url):
 542         group_id = self._match_id(url)
 543
 544         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 545
 546         error = self._search_regex(
 547             r'<div\b[^>]+\bclass=["\'](?:smp|playout)__message delta["\'][^>]*>\s*([^<]+?)\s*<',
 548             webpage, 'error', default=None)
 549         if error:
 550             raise ExtractorError(error, expected=True)
 551
 552         programme_id = None
 553         duration = None
 554
 555         tviplayer = self._search_regex(
 556             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 557             webpage, 'player', default=None)
 558
 559         if tviplayer:
 560             player = self._parse_json(tviplayer, group_id).get('player', {})
 561             duration = int_or_none(player.get('duration'))
 562             programme_id = player.get('vpid')
 563
 564         if not programme_id:
 565             programme_id = self._search_regex(
 566                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 567
 568         if programme_id:
 569             formats, subtitles = self._download_media_selector(programme_id)
 570             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 571                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 572                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
 573             description = self._search_regex(
 574                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 575                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
 576                 webpage, 'description', default=None)
 577             if not description:
 578                 description = self._html_search_meta('description', webpage)
 579         else:
 580             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 581
 582         self._sort_formats(formats)
 583
 584         return {
 585             'id': programme_id,
 586             'title': title,
 587             'description': description,
 588             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 589             'duration': duration,
 590             'formats': formats,
 591             'subtitles': subtitles,
 592         }
 593
 594
 595 class BBCIE(BBCCoUkIE):
 596     IE_NAME = 'bbc'
 597     IE_DESC = 'BBC'
 598     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 599
 600     _MEDIA_SETS = [
 601         'pc',
 602         'mobile-tablet-main',
 603     ]
 604
 605     _TESTS = [{
 606         # article with multiple videos embedded with data-playable containing vpids
 607         'url': 'http://www.bbc.com/news/world-europe-32668511',
 608         'info_dict': {
 609             'id': 'world-europe-32668511',
 610             'title': 'Russia stages massive WW2 parade',
 611             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 612         },
 613         'playlist_count': 2,
 614     }, {
 615         # article with multiple videos embedded with data-playable (more videos)
 616         'url': 'http://www.bbc.com/news/business-28299555',
 617         'info_dict': {
 618             'id': 'business-28299555',
 619             'title': 'Farnborough Airshow: Video highlights',
 620             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 621         },
 622         'playlist_count': 9,
 623         'skip': 'Save time',
 624     }, {
 625         # article with multiple videos embedded with `new SMP()`
 626         # broken
 627         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 628         'info_dict': {
 629             'id': '3662a707-0af9-3149-963f-47bea720b460',
 630             'title': 'BUGGER',
 631         },
 632         'playlist_count': 18,
 633     }, {
 634         # single video embedded with data-playable containing vpid
 635         'url': 'http://www.bbc.com/news/world-europe-32041533',
 636         'info_dict': {
 637             'id': 'p02mprgb',
 638             'ext': 'mp4',
 639             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 640             'description': 'md5:2868290467291b37feda7863f7a83f54',
 641             'duration': 47,
 642             'timestamp': 1427219242,
 643             'upload_date': '20150324',
 644         },
 645         'params': {
 646             # rtmp download
 647             'skip_download': True,
 648         }
 649     }, {
 650         # article with single video embedded with data-playable containing XML playlist
 651         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 652         # and playlist with f4m and m3u8 as streamingUrl
 653         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 654         'info_dict': {
 655             'id': '150615_telabyad_kentin_cogu',
 656             'ext': 'mp4',
 657             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 658             'description': 'md5:33a4805a855c9baf7115fcbde57e7025',
 659             'timestamp': 1434397334,
 660             'upload_date': '20150615',
 661         },
 662         'params': {
 663             'skip_download': True,
 664         }
 665     }, {
 666         # single video embedded with data-playable containing XML playlists (regional section)
 667         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 668         'info_dict': {
 669             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 670             'ext': 'mp4',
 671             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 672             'description': 'md5:1525f17448c4ee262b64b8f0c9ce66c8',
 673             'timestamp': 1434713142,
 674             'upload_date': '20150619',
 675         },
 676         'params': {
 677             'skip_download': True,
 678         }
 679     }, {
 680         # single video from video playlist embedded with vxp-playlist-data JSON
 681         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 682         'info_dict': {
 683             'id': 'p02w6qjc',
 684             'ext': 'mp4',
 685             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 686             'duration': 56,
 687             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 688         },
 689         'params': {
 690             'skip_download': True,
 691         }
 692     }, {
 693         # single video story with digitalData
 694         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 695         'info_dict': {
 696             'id': 'p02q6gc4',
 697             'ext': 'flv',
 698             'title': 'Sri Lanka’s spicy secret',
 699             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 700             'timestamp': 1437674293,
 701             'upload_date': '20150723',
 702         },
 703         'params': {
 704             # rtmp download
 705             'skip_download': True,
 706         }
 707     }, {
 708         # single video story without digitalData
 709         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 710         'info_dict': {
 711             'id': 'p018zqqg',
 712             'ext': 'mp4',
 713             'title': 'Hyundai Santa Fe Sport: Rock star',
 714             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 715             'timestamp': 1415867444,
 716             'upload_date': '20141113',
 717         },
 718         'params': {
 719             # rtmp download
 720             'skip_download': True,
 721         }
 722     }, {
 723         # single video embedded with Morph
 724         'url': 'http://www.bbc.co.uk/sport/live/olympics/36895975',
 725         'info_dict': {
 726             'id': 'p041vhd0',
 727             'ext': 'mp4',
 728             'title': "Nigeria v Japan - Men's First Round",
 729             'description': 'Live coverage of the first round from Group B at the Amazonia Arena.',
 730             'duration': 7980,
 731             'uploader': 'BBC Sport',
 732             'uploader_id': 'bbc_sport',
 733         },
 734         'params': {
 735             # m3u8 download
 736             'skip_download': True,
 737         },
 738         'skip': 'Georestricted to UK',
 739     }, {
 740         # single video with playlist.sxml URL in playlist param
 741         'url': 'http://www.bbc.com/sport/0/football/33653409',
 742         'info_dict': {
 743             'id': 'p02xycnp',
 744             'ext': 'mp4',
 745             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 746             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 747             'duration': 140,
 748         },
 749         'params': {
 750             # rtmp download
 751             'skip_download': True,
 752         }
 753     }, {
 754         # article with multiple videos embedded with playlist.sxml in playlist param
 755         'url': 'http://www.bbc.com/sport/0/football/34475836',
 756         'info_dict': {
 757             'id': '34475836',
 758             'title': 'Jurgen Klopp: Furious football from a witty and winning coach',
 759             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
 760         },
 761         'playlist_count': 3,
 762     }, {
 763         # school report article with single video
 764         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
 765         'info_dict': {
 766             'id': '35744779',
 767             'title': 'School which breaks down barriers in Jerusalem',
 768         },
 769         'playlist_count': 1,
 770     }, {
 771         # single video with playlist URL from weather section
 772         'url': 'http://www.bbc.com/weather/features/33601775',
 773         'only_matching': True,
 774     }, {
 775         # custom redirection to www.bbc.com
 776         # also, video with window.__INITIAL_DATA__
 777         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 778         'info_dict': {
 779             'id': 'p02xzws1',
 780             'ext': 'mp4',
 781             'title': "Pluto may have 'nitrogen glaciers'",
 782             'description': 'md5:6a95b593f528d7a5f2605221bc56912f',
 783             'thumbnail': r're:https?://.+/.+\.jpg',
 784             'timestamp': 1437785037,
 785             'upload_date': '20150725',
 786         },
 787     }, {
 788         # single video article embedded with data-media-vpid
 789         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
 790         'only_matching': True,
 791     }, {
 792         'url': 'https://www.bbc.co.uk/bbcthree/clip/73d0bbd0-abc3-4cea-b3c0-cdae21905eb1',
 793         'info_dict': {
 794             'id': 'p06556y7',
 795             'ext': 'mp4',
 796             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 797             'description': 'md5:4b7dfd063d5a789a1512e99662be3ddd',
 798         },
 799         'params': {
 800             'skip_download': True,
 801         }
 802     }, {
 803         # window.__PRELOADED_STATE__
 804         'url': 'https://www.bbc.co.uk/radio/play/b0b9z4yl',
 805         'info_dict': {
 806             'id': 'b0b9z4vz',
 807             'ext': 'mp4',
 808             'title': 'Prom 6: An American in Paris and Turangalila',
 809             'description': 'md5:51cf7d6f5c8553f197e58203bc78dff8',
 810             'uploader': 'Radio 3',
 811             'uploader_id': 'bbc_radio_three',
 812         },
 813     }, {
 814         'url': 'http://www.bbc.co.uk/learningenglish/chinese/features/lingohack/ep-181227',
 815         'info_dict': {
 816             'id': 'p06w9tws',
 817             'ext': 'mp4',
 818             'title': 'md5:2fabf12a726603193a2879a055f72514',
 819             'description': 'Learn English words and phrases from this story',
 820         },
 821         'add_ie': [BBCCoUkIE.ie_key()],
 822     }, {
 823         # BBC Reel
 824         'url': 'https://www.bbc.com/reel/video/p07c6sb6/how-positive-thinking-is-harming-your-happiness',
 825         'info_dict': {
 826             'id': 'p07c6sb9',
 827             'ext': 'mp4',
 828             'title': 'How positive thinking is harming your happiness',
 829             'alt_title': 'The downsides of positive thinking',
 830             'description': 'md5:fad74b31da60d83b8265954ee42d85b4',
 831             'duration': 235,
 832             'thumbnail': r're:https?://.+/p07c9dsr.jpg',
 833             'upload_date': '20190604',
 834             'categories': ['Psychology'],
 835         },
 836     }]
 837
 838     @classmethod
 839     def suitable(cls, url):
 840         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerEpisodesIE, BBCCoUkIPlayerGroupIE, BBCCoUkPlaylistIE)
 841         return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
 842                 else super(BBCIE, cls).suitable(url))
 843
 844     def _extract_from_media_meta(self, media_meta, video_id):
 845         # Direct links to media in media metadata (e.g.
 846         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 847         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 848         source_files = media_meta.get('sourceFiles')
 849         if source_files:
 850             return [{
 851                 'url': f['url'],
 852                 'format_id': format_id,
 853                 'ext': f.get('encoding'),
 854                 'tbr': float_or_none(f.get('bitrate'), 1000),
 855                 'filesize': int_or_none(f.get('filesize')),
 856             } for format_id, f in source_files.items() if f.get('url')], []
 857
 858         programme_id = media_meta.get('externalId')
 859         if programme_id:
 860             return self._download_media_selector(programme_id)
 861
 862         # Process playlist.sxml as legacy playlist
 863         href = media_meta.get('href')
 864         if href:
 865             playlist = self._download_legacy_playlist_url(href)
 866             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 867             return formats, subtitles
 868
 869         return [], []
 870
 871     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 872         programme_id, title, description, duration, formats, subtitles = \
 873             self._process_legacy_playlist_url(url, playlist_id)
 874         self._sort_formats(formats)
 875         return {
 876             'id': programme_id,
 877             'title': title,
 878             'description': description,
 879             'duration': duration,
 880             'timestamp': timestamp,
 881             'formats': formats,
 882             'subtitles': subtitles,
 883         }
 884
 885     def _real_extract(self, url):
 886         playlist_id = self._match_id(url)
 887
 888         webpage = self._download_webpage(url, playlist_id)
 889
 890         json_ld_info = self._search_json_ld(webpage, playlist_id, default={})
 891         timestamp = json_ld_info.get('timestamp')
 892
 893         playlist_title = json_ld_info.get('title')
 894         if not playlist_title:
 895             playlist_title = self._og_search_title(
 896                 webpage, default=None) or self._html_search_regex(
 897                 r'<title>(.+?)</title>', webpage, 'playlist title', default=None)
 898             if playlist_title:
 899                 playlist_title = re.sub(r'(.+)\s*-\s*BBC.*?$', r'\1', playlist_title).strip()
 900
 901         playlist_description = json_ld_info.get(
 902             'description') or self._og_search_description(webpage, default=None)
 903
 904         if not timestamp:
 905             timestamp = parse_iso8601(self._search_regex(
 906                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 907                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 908                  r'"datePublished":\s*"([^"]+)'],
 909                 webpage, 'date', default=None))
 910
 911         entries = []
 912
 913         # article with multiple videos embedded with playlist.sxml (e.g.
 914         # http://www.bbc.com/sport/0/football/34475836)
 915         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 916         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 917         if playlists:
 918             entries = [
 919                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 920                 for playlist_url in playlists]
 921
 922         # news article with multiple videos embedded with data-playable
 923         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 924         if data_playables:
 925             for _, data_playable_json in data_playables:
 926                 data_playable = self._parse_json(
 927                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 928                 if not data_playable:
 929                     continue
 930                 settings = data_playable.get('settings', {})
 931                 if settings:
 932                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 933                     # http://www.bbc.com/news/world-us-canada-34473351)
 934                     playlist_object = settings.get('playlistObject', {})
 935                     if playlist_object:
 936                         items = playlist_object.get('items')
 937                         if items and isinstance(items, list):
 938                             title = playlist_object['title']
 939                             description = playlist_object.get('summary')
 940                             duration = int_or_none(items[0].get('duration'))
 941                             programme_id = items[0].get('vpid')
 942                             formats, subtitles = self._download_media_selector(programme_id)
 943                             self._sort_formats(formats)
 944                             entries.append({
 945                                 'id': programme_id,
 946                                 'title': title,
 947                                 'description': description,
 948                                 'timestamp': timestamp,
 949                                 'duration': duration,
 950                                 'formats': formats,
 951                                 'subtitles': subtitles,
 952                             })
 953                     else:
 954                         # data-playable without vpid but with a playlist.sxml URLs
 955                         # in otherSettings.playlist (e.g.
 956                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 957                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 958                         if playlist:
 959                             entry = None
 960                             for key in ('streaming', 'progressiveDownload'):
 961                                 playlist_url = playlist.get('%sUrl' % key)
 962                                 if not playlist_url:
 963                                     continue
 964                                 try:
 965                                     info = self._extract_from_playlist_sxml(
 966                                         playlist_url, playlist_id, timestamp)
 967                                     if not entry:
 968                                         entry = info
 969                                     else:
 970                                         entry['title'] = info['title']
 971                                         entry['formats'].extend(info['formats'])
 972                                 except ExtractorError as e:
 973                                     # Some playlist URL may fail with 500, at the same time
 974                                     # the other one may work fine (e.g.
 975                                     # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 976                                     if isinstance(e.cause, compat_HTTPError) and e.cause.code == 500:
 977                                         continue
 978                                     raise
 979                             if entry:
 980                                 self._sort_formats(entry['formats'])
 981                                 entries.append(entry)
 982
 983         if entries:
 984             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 985
 986         # http://www.bbc.co.uk/learningenglish/chinese/features/lingohack/ep-181227
 987         group_id = self._search_regex(
 988             r'<div[^>]+\bclass=["\']video["\'][^>]+\bdata-pid=["\'](%s)' % self._ID_REGEX,
 989             webpage, 'group id', default=None)
 990         if group_id:
 991             return self.url_result(
 992                 'https://www.bbc.co.uk/programmes/%s' % group_id,
 993                 ie=BBCCoUkIE.ie_key())
 994
 995         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 996         programme_id = self._search_regex(
 997             [r'data-(?:video-player|media)-vpid="(%s)"' % self._ID_REGEX,
 998              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 999              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
1000             webpage, 'vpid', default=None)
1001
1002         if programme_id:
1003             formats, subtitles = self._download_media_selector(programme_id)
1004             self._sort_formats(formats)
1005             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
1006             digital_data = self._parse_json(
1007                 self._search_regex(
1008                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
1009                 programme_id, fatal=False)
1010             page_info = digital_data.get('page', {}).get('pageInfo', {})
1011             title = page_info.get('pageName') or self._og_search_title(webpage)
1012             description = page_info.get('description') or self._og_search_description(webpage)
1013             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
1014             return {
1015                 'id': programme_id,
1016                 'title': title,
1017                 'description': description,
1018                 'timestamp': timestamp,
1019                 'formats': formats,
1020                 'subtitles': subtitles,
1021             }
1022
1023         # bbc reel (e.g. https://www.bbc.com/reel/video/p07c6sb6/how-positive-thinking-is-harming-your-happiness)
1024         initial_data = self._parse_json(self._html_search_regex(
1025             r'<script[^>]+id=(["\'])initial-data\1[^>]+data-json=(["\'])(?P<json>(?:(?!\2).)+)',
1026             webpage, 'initial data', default='{}', group='json'), playlist_id, fatal=False)
1027         if initial_data:
1028             init_data = try_get(
1029                 initial_data, lambda x: x['initData']['items'][0], dict) or {}
1030             smp_data = init_data.get('smpData') or {}
1031             clip_data = try_get(smp_data, lambda x: x['items'][0], dict) or {}
1032             version_id = clip_data.get('versionID')
1033             if version_id:
1034                 title = smp_data['title']
1035                 formats, subtitles = self._download_media_selector(version_id)
1036                 self._sort_formats(formats)
1037                 image_url = smp_data.get('holdingImageURL')
1038                 display_date = init_data.get('displayDate')
1039                 topic_title = init_data.get('topicTitle')
1040
1041                 return {
1042                     'id': version_id,
1043                     'title': title,
1044                     'formats': formats,
1045                     'alt_title': init_data.get('shortTitle'),
1046                     'thumbnail': image_url.replace('$recipe', 'raw') if image_url else None,
1047                     'description': smp_data.get('summary') or init_data.get('shortSummary'),
1048                     'upload_date': display_date.replace('-', '') if display_date else None,
1049                     'subtitles': subtitles,
1050                     'duration': int_or_none(clip_data.get('duration')),
1051                     'categories': [topic_title] if topic_title else None,
1052                 }
1053
1054         # Morph based embed (e.g. http://www.bbc.co.uk/sport/live/olympics/36895975)
1055         # There are several setPayload calls may be present but the video
1056         # seems to be always related to the first one
1057         morph_payload = self._parse_json(
1058             self._search_regex(
1059                 r'Morph\.setPayload\([^,]+,\s*({.+?})\);',
1060                 webpage, 'morph payload', default='{}'),
1061             playlist_id, fatal=False)
1062         if morph_payload:
1063             components = try_get(morph_payload, lambda x: x['body']['components'], list) or []
1064             for component in components:
1065                 if not isinstance(component, dict):
1066                     continue
1067                 lead_media = try_get(component, lambda x: x['props']['leadMedia'], dict)
1068                 if not lead_media:
1069                     continue
1070                 identifiers = lead_media.get('identifiers')
1071                 if not identifiers or not isinstance(identifiers, dict):
1072                     continue
1073                 programme_id = identifiers.get('vpid') or identifiers.get('playablePid')
1074                 if not programme_id:
1075                     continue
1076                 title = lead_media.get('title') or self._og_search_title(webpage)
1077                 formats, subtitles = self._download_media_selector(programme_id)
1078                 self._sort_formats(formats)
1079                 description = lead_media.get('summary')
1080                 uploader = lead_media.get('masterBrand')
1081                 uploader_id = lead_media.get('mid')
1082                 duration = None
1083                 duration_d = lead_media.get('duration')
1084                 if isinstance(duration_d, dict):
1085                     duration = parse_duration(dict_get(
1086                         duration_d, ('rawDuration', 'formattedDuration', 'spokenDuration')))
1087                 return {
1088                     'id': programme_id,
1089                     'title': title,
1090                     'description': description,
1091                     'duration': duration,
1092                     'uploader': uploader,
1093                     'uploader_id': uploader_id,
1094                     'formats': formats,
1095                     'subtitles': subtitles,
1096                 }
1097
1098         preload_state = self._parse_json(self._search_regex(
1099             r'window\.__PRELOADED_STATE__\s*=\s*({.+?});', webpage,
1100             'preload state', default='{}'), playlist_id, fatal=False)
1101         if preload_state:
1102             current_programme = preload_state.get('programmes', {}).get('current') or {}
1103             programme_id = current_programme.get('id')
1104             if current_programme and programme_id and current_programme.get('type') == 'playable_item':
1105                 title = current_programme.get('titles', {}).get('tertiary') or playlist_title
1106                 formats, subtitles = self._download_media_selector(programme_id)
1107                 self._sort_formats(formats)
1108                 synopses = current_programme.get('synopses') or {}
1109                 network = current_programme.get('network') or {}
1110                 duration = int_or_none(
1111                     current_programme.get('duration', {}).get('value'))
1112                 thumbnail = None
1113                 image_url = current_programme.get('image_url')
1114                 if image_url:
1115                     thumbnail = image_url.replace('{recipe}', 'raw')
1116                 return {
1117                     'id': programme_id,
1118                     'title': title,
1119                     'description': dict_get(synopses, ('long', 'medium', 'short')),
1120                     'thumbnail': thumbnail,
1121                     'duration': duration,
1122                     'uploader': network.get('short_title'),
1123                     'uploader_id': network.get('id'),
1124                     'formats': formats,
1125                     'subtitles': subtitles,
1126                 }
1127
1128         bbc3_config = self._parse_json(
1129             self._search_regex(
1130                 r'(?s)bbcthreeConfig\s*=\s*({.+?})\s*;\s*<', webpage,
1131                 'bbcthree config', default='{}'),
1132             playlist_id, transform_source=js_to_json, fatal=False) or {}
1133         payload = bbc3_config.get('payload') or {}
1134         if payload:
1135             clip = payload.get('currentClip') or {}
1136             clip_vpid = clip.get('vpid')
1137             clip_title = clip.get('title')
1138             if clip_vpid and clip_title:
1139                 formats, subtitles = self._download_media_selector(clip_vpid)
1140                 self._sort_formats(formats)
1141                 return {
1142                     'id': clip_vpid,
1143                     'title': clip_title,
1144                     'thumbnail': dict_get(clip, ('poster', 'imageUrl')),
1145                     'description': clip.get('description'),
1146                     'duration': parse_duration(clip.get('duration')),
1147                     'formats': formats,
1148                     'subtitles': subtitles,
1149                 }
1150             bbc3_playlist = try_get(
1151                 payload, lambda x: x['content']['bbcMedia']['playlist'],
1152                 dict)
1153             if bbc3_playlist:
1154                 playlist_title = bbc3_playlist.get('title') or playlist_title
1155                 thumbnail = bbc3_playlist.get('holdingImageURL')
1156                 entries = []
1157                 for bbc3_item in bbc3_playlist['items']:
1158                     programme_id = bbc3_item.get('versionID')
1159                     if not programme_id:
1160                         continue
1161                     formats, subtitles = self._download_media_selector(programme_id)
1162                     self._sort_formats(formats)
1163                     entries.append({
1164                         'id': programme_id,
1165                         'title': playlist_title,
1166                         'thumbnail': thumbnail,
1167                         'timestamp': timestamp,
1168                         'formats': formats,
1169                         'subtitles': subtitles,
1170                     })
1171                 return self.playlist_result(
1172                     entries, playlist_id, playlist_title, playlist_description)
1173
1174         initial_data = self._parse_json(self._parse_json(self._search_regex(
1175             r'window\.__INITIAL_DATA__\s*=\s*("{.+?}");', webpage,
1176             'preload state', default='"{}"'), playlist_id, fatal=False), playlist_id, fatal=False)
1177         if initial_data:
1178             def parse_media(media):
1179                 if not media:
1180                     return
1181                 for item in (try_get(media, lambda x: x['media']['items'], list) or []):
1182                     item_id = item.get('id')
1183                     item_title = item.get('title')
1184                     if not (item_id and item_title):
1185                         continue
1186                     formats, subtitles = self._download_media_selector(item_id)
1187                     self._sort_formats(formats)
1188                     item_desc = None
1189                     blocks = try_get(media, lambda x: x['summary']['blocks'], list)
1190                     if blocks:
1191                         summary = []
1192                         for block in blocks:
1193                             text = try_get(block, lambda x: x['model']['text'], compat_str)
1194                             if text:
1195                                 summary.append(text)
1196                         if summary:
1197                             item_desc = '\n\n'.join(summary)
1198                     item_time = None
1199                     for meta in try_get(media, lambda x: x['metadata']['items'], list) or []:
1200                         if try_get(meta, lambda x: x['label']) == 'Published':
1201                             item_time = unified_timestamp(meta.get('timestamp'))
1202                             break
1203                     entries.append({
1204                         'id': item_id,
1205                         'title': item_title,
1206                         'thumbnail': item.get('holdingImageUrl'),
1207                         'formats': formats,
1208                         'subtitles': subtitles,
1209                         'timestamp': item_time,
1210                         'description': strip_or_none(item_desc),
1211                     })
1212             for resp in (initial_data.get('data') or {}).values():
1213                 name = resp.get('name')
1214                 if name == 'media-experience':
1215                     parse_media(try_get(resp, lambda x: x['data']['initialItem']['mediaItem'], dict))
1216                 elif name == 'article':
1217                     for block in (try_get(resp, lambda x: x['data']['content']['model']['blocks'], list) or []):
1218                         if block.get('type') != 'media':
1219                             continue
1220                         parse_media(block.get('model'))
1221             return self.playlist_result(
1222                 entries, playlist_id, playlist_title, playlist_description)
1223
1224         def extract_all(pattern):
1225             return list(filter(None, map(
1226                 lambda s: self._parse_json(s, playlist_id, fatal=False),
1227                 re.findall(pattern, webpage))))
1228
1229         # Multiple video article (e.g.
1230         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
1231         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
1232         entries = []
1233         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
1234             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
1235             if embed_url and re.match(EMBED_URL, embed_url):
1236                 entries.append(embed_url)
1237         entries.extend(re.findall(
1238             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
1239         if entries:
1240             return self.playlist_result(
1241                 [self.url_result(entry_, 'BBCCoUk') for entry_ in entries],
1242                 playlist_id, playlist_title, playlist_description)
1243
1244         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
1245         medias = extract_all(r"data-media-meta='({[^']+})'")
1246
1247         if not medias:
1248             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
1249             media_asset = self._search_regex(
1250                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
1251                 webpage, 'media asset', default=None)
1252             if media_asset:
1253                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
1254                 medias = []
1255                 for video in media_asset_page.get('videos', {}).values():
1256                     medias.extend(video.values())
1257
1258         if not medias:
1259             # Multiple video playlist with single `now playing` entry (e.g.
1260             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
1261             vxp_playlist = self._parse_json(
1262                 self._search_regex(
1263                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
1264                     webpage, 'playlist data'),
1265                 playlist_id)
1266             playlist_medias = []
1267             for item in vxp_playlist:
1268                 media = item.get('media')
1269                 if not media:
1270                     continue
1271                 playlist_medias.append(media)
1272                 # Download single video if found media with asset id matching the video id from URL
1273                 if item.get('advert', {}).get('assetId') == playlist_id:
1274                     medias = [media]
1275                     break
1276             # Fallback to the whole playlist
1277             if not medias:
1278                 medias = playlist_medias
1279
1280         entries = []
1281         for num, media_meta in enumerate(medias, start=1):
1282             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
1283             if not formats and not self.get_param('ignore_no_formats'):
1284                 continue
1285             self._sort_formats(formats)
1286
1287             video_id = media_meta.get('externalId')
1288             if not video_id:
1289                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
1290
1291             title = media_meta.get('caption')
1292             if not title:
1293                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
1294
1295             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
1296
1297             images = []
1298             for image in media_meta.get('images', {}).values():
1299                 images.extend(image.values())
1300             if 'image' in media_meta:
1301                 images.append(media_meta['image'])
1302
1303             thumbnails = [{
1304                 'url': image.get('href'),
1305                 'width': int_or_none(image.get('width')),
1306                 'height': int_or_none(image.get('height')),
1307             } for image in images]
1308
1309             entries.append({
1310                 'id': video_id,
1311                 'title': title,
1312                 'thumbnails': thumbnails,
1313                 'duration': duration,
1314                 'timestamp': timestamp,
1315                 'formats': formats,
1316                 'subtitles': subtitles,
1317             })
1318
1319         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
1320
1321
1322 class BBCCoUkArticleIE(InfoExtractor):
1323     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
1324     IE_NAME = 'bbc.co.uk:article'
1325     IE_DESC = 'BBC articles'
1326
1327     _TEST = {
1328         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
1329         'info_dict': {
1330             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
1331             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
1332             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
1333         },
1334         'playlist_count': 4,
1335         'add_ie': ['BBCCoUk'],
1336     }
1337
1338     def _real_extract(self, url):
1339         playlist_id = self._match_id(url)
1340
1341         webpage = self._download_webpage(url, playlist_id)
1342
1343         title = self._og_search_title(webpage)
1344         description = self._og_search_description(webpage).strip()
1345
1346         entries = [self.url_result(programme_url) for programme_url in re.findall(
1347             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
1348
1349         return self.playlist_result(entries, playlist_id, title, description)
1350
1351
1352 class BBCCoUkPlaylistBaseIE(InfoExtractor):
1353     def _entries(self, webpage, url, playlist_id):
1354         single_page = 'page' in compat_urlparse.parse_qs(
1355             compat_urlparse.urlparse(url).query)
1356         for page_num in itertools.count(2):
1357             for video_id in re.findall(
1358                     self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage):
1359                 yield self.url_result(
1360                     self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
1361             if single_page:
1362                 return
1363             next_page = self._search_regex(
1364                 r'<li[^>]+class=(["\'])pagination_+next\1[^>]*><a[^>]+href=(["\'])(?P<url>(?:(?!\2).)+)\2',
1365                 webpage, 'next page url', default=None, group='url')
1366             if not next_page:
1367                 break
1368             webpage = self._download_webpage(
1369                 compat_urlparse.urljoin(url, next_page), playlist_id,
1370                 'Downloading page %d' % page_num, page_num)
1371
1372     def _real_extract(self, url):
1373         playlist_id = self._match_id(url)
1374
1375         webpage = self._download_webpage(url, playlist_id)
1376
1377         title, description = self._extract_title_and_description(webpage)
1378
1379         return self.playlist_result(
1380             self._entries(webpage, url, playlist_id),
1381             playlist_id, title, description)
1382
1383
1384 class BBCCoUkIPlayerPlaylistBaseIE(InfoExtractor):
1385     _VALID_URL_TMPL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/%%s/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
1386
1387     @staticmethod
1388     def _get_default(episode, key, default_key='default'):
1389         return try_get(episode, lambda x: x[key][default_key])
1390
1391     def _get_description(self, data):
1392         synopsis = data.get(self._DESCRIPTION_KEY) or {}
1393         return dict_get(synopsis, ('large', 'medium', 'small'))
1394
1395     def _fetch_page(self, programme_id, per_page, series_id, page):
1396         elements = self._get_elements(self._call_api(
1397             programme_id, per_page, page + 1, series_id))
1398         for element in elements:
1399             episode = self._get_episode(element)
1400             episode_id = episode.get('id')
1401             if not episode_id:
1402                 continue
1403             thumbnail = None
1404             image = self._get_episode_image(episode)
1405             if image:
1406                 thumbnail = image.replace('{recipe}', 'raw')
1407             category = self._get_default(episode, 'labels', 'category')
1408             yield {
1409                 '_type': 'url',
1410                 'id': episode_id,
1411                 'title': self._get_episode_field(episode, 'subtitle'),
1412                 'url': 'https://www.bbc.co.uk/iplayer/episode/' + episode_id,
1413                 'thumbnail': thumbnail,
1414                 'description': self._get_description(episode),
1415                 'categories': [category] if category else None,
1416                 'series': self._get_episode_field(episode, 'title'),
1417                 'ie_key': BBCCoUkIE.ie_key(),
1418             }
1419
1420     def _real_extract(self, url):
1421         pid = self._match_id(url)
1422         qs = parse_qs(url)
1423         series_id = qs.get('seriesId', [None])[0]
1424         page = qs.get('page', [None])[0]
1425         per_page = 36 if page else self._PAGE_SIZE
1426         fetch_page = functools.partial(self._fetch_page, pid, per_page, series_id)
1427         entries = fetch_page(int(page) - 1) if page else OnDemandPagedList(fetch_page, self._PAGE_SIZE)
1428         playlist_data = self._get_playlist_data(self._call_api(pid, 1))
1429         return self.playlist_result(
1430             entries, pid, self._get_playlist_title(playlist_data),
1431             self._get_description(playlist_data))
1432
1433
1434 class BBCCoUkIPlayerEpisodesIE(BBCCoUkIPlayerPlaylistBaseIE):
1435     IE_NAME = 'bbc.co.uk:iplayer:episodes'
1436     _VALID_URL = BBCCoUkIPlayerPlaylistBaseIE._VALID_URL_TMPL % 'episodes'
1437     _TESTS = [{
1438         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1439         'info_dict': {
1440             'id': 'b05rcz9v',
1441             'title': 'The Disappearance',
1442             'description': 'md5:58eb101aee3116bad4da05f91179c0cb',
1443         },
1444         'playlist_mincount': 8,
1445     }, {
1446         # all seasons
1447         'url': 'https://www.bbc.co.uk/iplayer/episodes/b094m5t9/doctor-foster',
1448         'info_dict': {
1449             'id': 'b094m5t9',
1450             'title': 'Doctor Foster',
1451             'description': 'md5:5aa9195fad900e8e14b52acd765a9fd6',
1452         },
1453         'playlist_mincount': 10,
1454     }, {
1455         # explicit season
1456         'url': 'https://www.bbc.co.uk/iplayer/episodes/b094m5t9/doctor-foster?seriesId=b094m6nv',
1457         'info_dict': {
1458             'id': 'b094m5t9',
1459             'title': 'Doctor Foster',
1460             'description': 'md5:5aa9195fad900e8e14b52acd765a9fd6',
1461         },
1462         'playlist_mincount': 5,
1463     }, {
1464         # all pages
1465         'url': 'https://www.bbc.co.uk/iplayer/episodes/m0004c4v/beechgrove',
1466         'info_dict': {
1467             'id': 'm0004c4v',
1468             'title': 'Beechgrove',
1469             'description': 'Gardening show that celebrates Scottish horticulture and growing conditions.',
1470         },
1471         'playlist_mincount': 37,
1472     }, {
1473         # explicit page
1474         'url': 'https://www.bbc.co.uk/iplayer/episodes/m0004c4v/beechgrove?page=2',
1475         'info_dict': {
1476             'id': 'm0004c4v',
1477             'title': 'Beechgrove',
1478             'description': 'Gardening show that celebrates Scottish horticulture and growing conditions.',
1479         },
1480         'playlist_mincount': 1,
1481     }]
1482     _PAGE_SIZE = 100
1483     _DESCRIPTION_KEY = 'synopsis'
1484
1485     def _get_episode_image(self, episode):
1486         return self._get_default(episode, 'image')
1487
1488     def _get_episode_field(self, episode, field):
1489         return self._get_default(episode, field)
1490
1491     @staticmethod
1492     def _get_elements(data):
1493         return data['entities']['results']
1494
1495     @staticmethod
1496     def _get_episode(element):
1497         return element.get('episode') or {}
1498
1499     def _call_api(self, pid, per_page, page=1, series_id=None):
1500         variables = {
1501             'id': pid,
1502             'page': page,
1503             'perPage': per_page,
1504         }
1505         if series_id:
1506             variables['sliceId'] = series_id
1507         return self._download_json(
1508             'https://graph.ibl.api.bbc.co.uk/', pid, headers={
1509                 'Content-Type': 'application/json'
1510             }, data=json.dumps({
1511                 'id': '5692d93d5aac8d796a0305e895e61551',
1512                 'variables': variables,
1513             }).encode('utf-8'))['data']['programme']
1514
1515     @staticmethod
1516     def _get_playlist_data(data):
1517         return data
1518
1519     def _get_playlist_title(self, data):
1520         return self._get_default(data, 'title')
1521
1522
1523 class BBCCoUkIPlayerGroupIE(BBCCoUkIPlayerPlaylistBaseIE):
1524     IE_NAME = 'bbc.co.uk:iplayer:group'
1525     _VALID_URL = BBCCoUkIPlayerPlaylistBaseIE._VALID_URL_TMPL % 'group'
1526     _TESTS = [{
1527         # Available for over a year unlike 30 days for most other programmes
1528         'url': 'http://www.bbc.co.uk/iplayer/group/p02tcc32',
1529         'info_dict': {
1530             'id': 'p02tcc32',
1531             'title': 'Bohemian Icons',
1532             'description': 'md5:683e901041b2fe9ba596f2ab04c4dbe7',
1533         },
1534         'playlist_mincount': 10,
1535     }, {
1536         # all pages
1537         'url': 'https://www.bbc.co.uk/iplayer/group/p081d7j7',
1538         'info_dict': {
1539             'id': 'p081d7j7',
1540             'title': 'Music in Scotland',
1541             'description': 'Perfomances in Scotland and programmes featuring Scottish acts.',
1542         },
1543         'playlist_mincount': 47,
1544     }, {
1545         # explicit page
1546         'url': 'https://www.bbc.co.uk/iplayer/group/p081d7j7?page=2',
1547         'info_dict': {
1548             'id': 'p081d7j7',
1549             'title': 'Music in Scotland',
1550             'description': 'Perfomances in Scotland and programmes featuring Scottish acts.',
1551         },
1552         'playlist_mincount': 11,
1553     }]
1554     _PAGE_SIZE = 200
1555     _DESCRIPTION_KEY = 'synopses'
1556
1557     def _get_episode_image(self, episode):
1558         return self._get_default(episode, 'images', 'standard')
1559
1560     def _get_episode_field(self, episode, field):
1561         return episode.get(field)
1562
1563     @staticmethod
1564     def _get_elements(data):
1565         return data['elements']
1566
1567     @staticmethod
1568     def _get_episode(element):
1569         return element
1570
1571     def _call_api(self, pid, per_page, page=1, series_id=None):
1572         return self._download_json(
1573             'http://ibl.api.bbc.co.uk/ibl/v1/groups/%s/episodes' % pid,
1574             pid, query={
1575                 'page': page,
1576                 'per_page': per_page,
1577             })['group_episodes']
1578
1579     @staticmethod
1580     def _get_playlist_data(data):
1581         return data['group']
1582
1583     def _get_playlist_title(self, data):
1584         return data.get('title')
1585
1586
1587 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1588     IE_NAME = 'bbc.co.uk:playlist'
1589     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1590     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1591     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1592     _TESTS = [{
1593         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1594         'info_dict': {
1595             'id': 'b05rcz9v',
1596             'title': 'The Disappearance - Clips - BBC Four',
1597             'description': 'French thriller serial about a missing teenager.',
1598         },
1599         'playlist_mincount': 7,
1600     }, {
1601         # multipage playlist, explicit page
1602         'url': 'http://www.bbc.co.uk/programmes/b00mfl7n/clips?page=1',
1603         'info_dict': {
1604             'id': 'b00mfl7n',
1605             'title': 'Frozen Planet - Clips - BBC One',
1606             'description': 'md5:65dcbf591ae628dafe32aa6c4a4a0d8c',
1607         },
1608         'playlist_mincount': 24,
1609     }, {
1610         # multipage playlist, all pages
1611         'url': 'http://www.bbc.co.uk/programmes/b00mfl7n/clips',
1612         'info_dict': {
1613             'id': 'b00mfl7n',
1614             'title': 'Frozen Planet - Clips - BBC One',
1615             'description': 'md5:65dcbf591ae628dafe32aa6c4a4a0d8c',
1616         },
1617         'playlist_mincount': 142,
1618     }, {
1619         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1620         'only_matching': True,
1621     }, {
1622         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1623         'only_matching': True,
1624     }, {
1625         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1626         'only_matching': True,
1627     }]
1628
1629     def _extract_title_and_description(self, webpage):
1630         title = self._og_search_title(webpage, fatal=False)
1631         description = self._og_search_description(webpage)
1632         return title, description