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