_ Git - youtube-dl/blob - youtube_dl/extractor/bbc.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     clean_html,
  10     dict_get,
  11     ExtractorError,
  12     float_or_none,
  13     get_element_by_class,
  14     int_or_none,
  15     js_to_json,
  16     parse_duration,
  17     parse_iso8601,
  18     try_get,
  19     unescapeHTML,
  20     url_or_none,
  21     urlencode_postdata,
  22     urljoin,
  23 )
  24 from ..compat import (
  25     compat_etree_Element,
  26     compat_HTTPError,
  27     compat_urlparse,
  28 )
  29
  30
  31 class BBCCoUkIE(InfoExtractor):
  32     IE_NAME = 'bbc.co.uk'
  33     IE_DESC = 'BBC iPlayer'
  34     _ID_REGEX = r'(?:[pbm][\da-z]{7}|w[\da-z]{7,14})'
  35     _VALID_URL = r'''(?x)
  36                     https?://
  37                         (?:www\.)?bbc\.co\.uk/
  38                         (?:
  39                             programmes/(?!articles/)|
  40                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
  41                             music/(?:clips|audiovideo/popular)[/#]|
  42                             radio/player/|
  43                             events/[^/]+/play/[^/]+/
  44                         )
  45                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
  46                     ''' % _ID_REGEX
  47
  48     _LOGIN_URL = 'https://account.bbc.com/signin'
  49     _NETRC_MACHINE = 'bbc'
  50
  51     _MEDIASELECTOR_URLS = [
  52         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  53         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  54         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  55         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  56         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  57     ]
  58
  59     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  60     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  61
  62     _NAMESPACES = (
  63         _MEDIASELECTION_NS,
  64         _EMP_PLAYLIST_NS,
  65     )
  66
  67     _TESTS = [
  68         {
  69             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  70             'info_dict': {
  71                 'id': 'b039d07m',
  72                 'ext': 'flv',
  73                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  74                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  75             },
  76             'params': {
  77                 # rtmp download
  78                 'skip_download': True,
  79             }
  80         },
  81         {
  82             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  83             'info_dict': {
  84                 'id': 'b00yng1d',
  85                 'ext': 'flv',
  86                 'title': 'The Man in Black: Series 3: The Printed Name',
  87                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  88                 'duration': 1800,
  89             },
  90             'params': {
  91                 # rtmp download
  92                 'skip_download': True,
  93             },
  94             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  95         },
  96         {
  97             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  98             'info_dict': {
  99                 'id': 'b00yng1d',
 100                 'ext': 'flv',
 101                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
 102                 '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.',
 103                 'duration': 5100,
 104             },
 105             'params': {
 106                 # rtmp download
 107                 'skip_download': True,
 108             },
 109             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 110         },
 111         {
 112             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
 113             'info_dict': {
 114                 'id': 'b03k3pb7',
 115                 'ext': 'flv',
 116                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 117                 'description': '2. Invasion',
 118                 'duration': 3600,
 119             },
 120             'params': {
 121                 # rtmp download
 122                 'skip_download': True,
 123             },
 124             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 125         }, {
 126             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 127             'info_dict': {
 128                 'id': 'b04v209v',
 129                 'ext': 'flv',
 130                 'title': 'Pete Tong, The Essential New Tune Special',
 131                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 132                 'duration': 10800,
 133             },
 134             'params': {
 135                 # rtmp download
 136                 'skip_download': True,
 137             },
 138             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 139         }, {
 140             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
 141             'note': 'Audio',
 142             'info_dict': {
 143                 'id': 'p022h44j',
 144                 'ext': 'flv',
 145                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 146                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 147                 'duration': 227,
 148             },
 149             'params': {
 150                 # rtmp download
 151                 'skip_download': True,
 152             }
 153         }, {
 154             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 155             'note': 'Video',
 156             'info_dict': {
 157                 'id': 'p025c103',
 158                 'ext': 'flv',
 159                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 160                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 161                 'duration': 226,
 162             },
 163             'params': {
 164                 # rtmp download
 165                 'skip_download': True,
 166             }
 167         }, {
 168             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 169             'info_dict': {
 170                 'id': 'p02n76xf',
 171                 'ext': 'flv',
 172                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 173                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 174                 'duration': 3540,
 175             },
 176             'params': {
 177                 # rtmp download
 178                 'skip_download': True,
 179             },
 180             'skip': 'geolocation',
 181         }, {
 182             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 183             'info_dict': {
 184                 'id': 'b05zmgw1',
 185                 'ext': 'flv',
 186                 '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.',
 187                 'title': 'Royal Academy Summer Exhibition',
 188                 'duration': 3540,
 189             },
 190             'params': {
 191                 # rtmp download
 192                 'skip_download': True,
 193             },
 194             'skip': 'geolocation',
 195         }, {
 196             # iptv-all mediaset fails with geolocation however there is no geo restriction
 197             # for this programme at all
 198             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
 199             'info_dict': {
 200                 'id': 'b06rkms3',
 201                 'ext': 'flv',
 202                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 203                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 204             },
 205             'params': {
 206                 # rtmp download
 207                 'skip_download': True,
 208             },
 209             'skip': 'Now it\'s really geo-restricted',
 210         }, {
 211             # compact player (https://github.com/ytdl-org/youtube-dl/issues/8147)
 212             'url': 'http://www.bbc.co.uk/programmes/p028bfkf/player',
 213             'info_dict': {
 214                 'id': 'p028bfkj',
 215                 'ext': 'flv',
 216                 'title': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 217                 'description': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 218             },
 219             'params': {
 220                 # rtmp download
 221                 'skip_download': True,
 222             },
 223         }, {
 224             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 225             'only_matching': True,
 226         }, {
 227             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 228             'only_matching': True,
 229         }, {
 230             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 231             'only_matching': True,
 232         }, {
 233             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 234             'only_matching': True,
 235         }, {
 236             'url': 'https://www.bbc.co.uk/music/audiovideo/popular#p055bc55',
 237             'only_matching': True,
 238         }, {
 239             'url': 'http://www.bbc.co.uk/programmes/w3csv1y9',
 240             'only_matching': True,
 241         }, {
 242             'url': 'https://www.bbc.co.uk/programmes/m00005xn',
 243             'only_matching': True,
 244         }, {
 245             'url': 'https://www.bbc.co.uk/programmes/w172w4dww1jqt5s',
 246             'only_matching': True,
 247         }]
 248
 249     _USP_RE = r'/([^/]+?)\.ism(?:\.hlsv2\.ism)?/[^/]+\.m3u8'
 250
 251     def _login(self):
 252         username, password = self._get_login_info()
 253         if username is None:
 254             return
 255
 256         login_page = self._download_webpage(
 257             self._LOGIN_URL, None, 'Downloading signin page')
 258
 259         login_form = self._hidden_inputs(login_page)
 260
 261         login_form.update({
 262             'username': username,
 263             'password': password,
 264         })
 265
 266         post_url = urljoin(self._LOGIN_URL, self._search_regex(
 267             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 268             'post url', default=self._LOGIN_URL, group='url'))
 269
 270         response, urlh = self._download_webpage_handle(
 271             post_url, None, 'Logging in', data=urlencode_postdata(login_form),
 272             headers={'Referer': self._LOGIN_URL})
 273
 274         if self._LOGIN_URL in urlh.geturl():
 275             error = clean_html(get_element_by_class('form-message', response))
 276             if error:
 277                 raise ExtractorError(
 278                     'Unable to login: %s' % error, expected=True)
 279             raise ExtractorError('Unable to log in')
 280
 281     def _real_initialize(self):
 282         self._login()
 283
 284     class MediaSelectionError(Exception):
 285         def __init__(self, id):
 286             self.id = id
 287
 288     def _extract_asx_playlist(self, connection, programme_id):
 289         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 290         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 291
 292     def _extract_items(self, playlist):
 293         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 294
 295     def _findall_ns(self, element, xpath):
 296         elements = []
 297         for ns in self._NAMESPACES:
 298             elements.extend(element.findall(xpath % ns))
 299         return elements
 300
 301     def _extract_medias(self, media_selection):
 302         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 303         if error is None:
 304             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 305         if error is not None:
 306             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 307         return self._findall_ns(media_selection, './{%s}media')
 308
 309     def _extract_connections(self, media):
 310         return self._findall_ns(media, './{%s}connection')
 311
 312     def _get_subtitles(self, media, programme_id):
 313         subtitles = {}
 314         for connection in self._extract_connections(media):
 315             cc_url = url_or_none(connection.get('href'))
 316             if not cc_url:
 317                 continue
 318             captions = self._download_xml(
 319                 cc_url, programme_id, 'Downloading captions', fatal=False)
 320             if not isinstance(captions, compat_etree_Element):
 321                 continue
 322             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 323             subtitles[lang] = [
 324                 {
 325                     'url': connection.get('href'),
 326                     'ext': 'ttml',
 327                 },
 328             ]
 329         return subtitles
 330
 331     def _raise_extractor_error(self, media_selection_error):
 332         raise ExtractorError(
 333             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 334             expected=True)
 335
 336     def _download_media_selector(self, programme_id):
 337         last_exception = None
 338         for mediaselector_url in self._MEDIASELECTOR_URLS:
 339             try:
 340                 return self._download_media_selector_url(
 341                     mediaselector_url % programme_id, programme_id)
 342             except BBCCoUkIE.MediaSelectionError as e:
 343                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 344                     last_exception = e
 345                     continue
 346                 self._raise_extractor_error(e)
 347         self._raise_extractor_error(last_exception)
 348
 349     def _download_media_selector_url(self, url, programme_id=None):
 350         media_selection = self._download_xml(
 351             url, programme_id, 'Downloading media selection XML',
 352             expected_status=(403, 404))
 353         return self._process_media_selector(media_selection, programme_id)
 354
 355     def _process_media_selector(self, media_selection, programme_id):
 356         formats = []
 357         subtitles = None
 358         urls = []
 359
 360         for media in self._extract_medias(media_selection):
 361             kind = media.get('kind')
 362             if kind in ('video', 'audio'):
 363                 bitrate = int_or_none(media.get('bitrate'))
 364                 encoding = media.get('encoding')
 365                 service = media.get('service')
 366                 width = int_or_none(media.get('width'))
 367                 height = int_or_none(media.get('height'))
 368                 file_size = int_or_none(media.get('media_file_size'))
 369                 for connection in self._extract_connections(media):
 370                     href = connection.get('href')
 371                     if href in urls:
 372                         continue
 373                     if href:
 374                         urls.append(href)
 375                     conn_kind = connection.get('kind')
 376                     protocol = connection.get('protocol')
 377                     supplier = connection.get('supplier')
 378                     transfer_format = connection.get('transferFormat')
 379                     format_id = supplier or conn_kind or protocol
 380                     if service:
 381                         format_id = '%s_%s' % (service, format_id)
 382                     # ASX playlist
 383                     if supplier == 'asx':
 384                         for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 385                             formats.append({
 386                                 'url': ref,
 387                                 'format_id': 'ref%s_%s' % (i, format_id),
 388                             })
 389                     elif transfer_format == 'dash':
 390                         formats.extend(self._extract_mpd_formats(
 391                             href, programme_id, mpd_id=format_id, fatal=False))
 392                     elif transfer_format == 'hls':
 393                         formats.extend(self._extract_m3u8_formats(
 394                             href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 395                             m3u8_id=format_id, fatal=False))
 396                         if re.search(self._USP_RE, href):
 397                             usp_formats = self._extract_m3u8_formats(
 398                                 re.sub(self._USP_RE, r'/\1.ism/\1.m3u8', href),
 399                                 programme_id, ext='mp4', entry_protocol='m3u8_native',
 400                                 m3u8_id=format_id, fatal=False)
 401                             for f in usp_formats:
 402                                 if f.get('height') and f['height'] > 720:
 403                                     continue
 404                                 formats.append(f)
 405                     elif transfer_format == 'hds':
 406                         formats.extend(self._extract_f4m_formats(
 407                             href, programme_id, f4m_id=format_id, fatal=False))
 408                     else:
 409                         if not service and not supplier and bitrate:
 410                             format_id += '-%d' % bitrate
 411                         fmt = {
 412                             'format_id': format_id,
 413                             'filesize': file_size,
 414                         }
 415                         if kind == 'video':
 416                             fmt.update({
 417                                 'width': width,
 418                                 'height': height,
 419                                 'tbr': bitrate,
 420                                 'vcodec': encoding,
 421                             })
 422                         else:
 423                             fmt.update({
 424                                 'abr': bitrate,
 425                                 'acodec': encoding,
 426                                 'vcodec': 'none',
 427                             })
 428                         if protocol in ('http', 'https'):
 429                             # Direct link
 430                             fmt.update({
 431                                 'url': href,
 432                             })
 433                         elif protocol == 'rtmp':
 434                             application = connection.get('application', 'ondemand')
 435                             auth_string = connection.get('authString')
 436                             identifier = connection.get('identifier')
 437                             server = connection.get('server')
 438                             fmt.update({
 439                                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 440                                 'play_path': identifier,
 441                                 'app': '%s?%s' % (application, auth_string),
 442                                 'page_url': 'http://www.bbc.co.uk',
 443                                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 444                                 'rtmp_live': False,
 445                                 'ext': 'flv',
 446                             })
 447                         else:
 448                             continue
 449                         formats.append(fmt)
 450             elif kind == 'captions':
 451                 subtitles = self.extract_subtitles(media, programme_id)
 452         return formats, subtitles
 453
 454     def _download_playlist(self, playlist_id):
 455         try:
 456             playlist = self._download_json(
 457                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 458                 playlist_id, 'Downloading playlist JSON')
 459
 460             version = playlist.get('defaultAvailableVersion')
 461             if version:
 462                 smp_config = version['smpConfig']
 463                 title = smp_config['title']
 464                 description = smp_config['summary']
 465                 for item in smp_config['items']:
 466                     kind = item['kind']
 467                     if kind not in ('programme', 'radioProgramme'):
 468                         continue
 469                     programme_id = item.get('vpid')
 470                     duration = int_or_none(item.get('duration'))
 471                     formats, subtitles = self._download_media_selector(programme_id)
 472                 return programme_id, title, description, duration, formats, subtitles
 473         except ExtractorError as ee:
 474             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 475                 raise
 476
 477         # fallback to legacy playlist
 478         return self._process_legacy_playlist(playlist_id)
 479
 480     def _process_legacy_playlist_url(self, url, display_id):
 481         playlist = self._download_legacy_playlist_url(url, display_id)
 482         return self._extract_from_legacy_playlist(playlist, display_id)
 483
 484     def _process_legacy_playlist(self, playlist_id):
 485         return self._process_legacy_playlist_url(
 486             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 487
 488     def _download_legacy_playlist_url(self, url, playlist_id=None):
 489         return self._download_xml(
 490             url, playlist_id, 'Downloading legacy playlist XML')
 491
 492     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 493         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 494         if no_items is not None:
 495             reason = no_items.get('reason')
 496             if reason == 'preAvailability':
 497                 msg = 'Episode %s is not yet available' % playlist_id
 498             elif reason == 'postAvailability':
 499                 msg = 'Episode %s is no longer available' % playlist_id
 500             elif reason == 'noMedia':
 501                 msg = 'Episode %s is not currently available' % playlist_id
 502             else:
 503                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 504             raise ExtractorError(msg, expected=True)
 505
 506         for item in self._extract_items(playlist):
 507             kind = item.get('kind')
 508             if kind not in ('programme', 'radioProgramme'):
 509                 continue
 510             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 511             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 512             description = description_el.text if description_el is not None else None
 513
 514             def get_programme_id(item):
 515                 def get_from_attributes(item):
 516                     for p in('identifier', 'group'):
 517                         value = item.get(p)
 518                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 519                             return value
 520                 get_from_attributes(item)
 521                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 522                 if mediator is not None:
 523                     return get_from_attributes(mediator)
 524
 525             programme_id = get_programme_id(item)
 526             duration = int_or_none(item.get('duration'))
 527
 528             if programme_id:
 529                 formats, subtitles = self._download_media_selector(programme_id)
 530             else:
 531                 formats, subtitles = self._process_media_selector(item, playlist_id)
 532                 programme_id = playlist_id
 533
 534         return programme_id, title, description, duration, formats, subtitles
 535
 536     def _real_extract(self, url):
 537         group_id = self._match_id(url)
 538
 539         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 540
 541         error = self._search_regex(
 542             r'<div\b[^>]+\bclass=["\']smp__message delta["\'][^>]*>([^<]+)<',
 543             webpage, 'error', default=None)
 544         if error:
 545             raise ExtractorError(error, expected=True)
 546
 547         programme_id = None
 548         duration = None
 549
 550         tviplayer = self._search_regex(
 551             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 552             webpage, 'player', default=None)
 553
 554         if tviplayer:
 555             player = self._parse_json(tviplayer, group_id).get('player', {})
 556             duration = int_or_none(player.get('duration'))
 557             programme_id = player.get('vpid')
 558
 559         if not programme_id:
 560             programme_id = self._search_regex(
 561                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 562
 563         if programme_id:
 564             formats, subtitles = self._download_media_selector(programme_id)
 565             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 566                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 567                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
 568             description = self._search_regex(
 569                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 570                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
 571                 webpage, 'description', default=None)
 572             if not description:
 573                 description = self._html_search_meta('description', webpage)
 574         else:
 575             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 576
 577         self._sort_formats(formats)
 578
 579         return {
 580             'id': programme_id,
 581             'title': title,
 582             'description': description,
 583             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 584             'duration': duration,
 585             'formats': formats,
 586             'subtitles': subtitles,
 587         }
 588
 589
 590 class BBCIE(BBCCoUkIE):
 591     IE_NAME = 'bbc'
 592     IE_DESC = 'BBC'
 593     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 594
 595     _MEDIASELECTOR_URLS = [
 596         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 597         # even not geo restricted at all
 598         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 599         # Provides more formats, namely direct mp4 links, but fails on some videos with
 600         # notukerror for non UK (?) users (e.g.
 601         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 602         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 603         # Provides fewer formats, but works everywhere for everybody (hopefully)
 604         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 605     ]
 606
 607     _TESTS = [{
 608         # article with multiple videos embedded with data-playable containing vpids
 609         'url': 'http://www.bbc.com/news/world-europe-32668511',
 610         'info_dict': {
 611             'id': 'world-europe-32668511',
 612             'title': 'Russia stages massive WW2 parade despite Western boycott',
 613             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 614         },
 615         'playlist_count': 2,
 616     }, {
 617         # article with multiple videos embedded with data-playable (more videos)
 618         'url': 'http://www.bbc.com/news/business-28299555',
 619         'info_dict': {
 620             'id': 'business-28299555',
 621             'title': 'Farnborough Airshow: Video highlights',
 622             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 623         },
 624         'playlist_count': 9,
 625         'skip': 'Save time',
 626     }, {
 627         # article with multiple videos embedded with `new SMP()`
 628         # broken
 629         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 630         'info_dict': {
 631             'id': '3662a707-0af9-3149-963f-47bea720b460',
 632             'title': 'BUGGER',
 633         },
 634         'playlist_count': 18,
 635     }, {
 636         # single video embedded with data-playable containing vpid
 637         'url': 'http://www.bbc.com/news/world-europe-32041533',
 638         'info_dict': {
 639             'id': 'p02mprgb',
 640             'ext': 'mp4',
 641             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 642             'description': 'md5:2868290467291b37feda7863f7a83f54',
 643             'duration': 47,
 644             'timestamp': 1427219242,
 645             'upload_date': '20150324',
 646         },
 647         'params': {
 648             # rtmp download
 649             'skip_download': True,
 650         }
 651     }, {
 652         # article with single video embedded with data-playable containing XML playlist
 653         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 654         # and playlist with f4m and m3u8 as streamingUrl
 655         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 656         'info_dict': {
 657             'id': '150615_telabyad_kentin_cogu',
 658             'ext': 'mp4',
 659             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 660             'description': 'md5:33a4805a855c9baf7115fcbde57e7025',
 661             'timestamp': 1434397334,
 662             'upload_date': '20150615',
 663         },
 664         'params': {
 665             'skip_download': True,
 666         }
 667     }, {
 668         # single video embedded with data-playable containing XML playlists (regional section)
 669         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 670         'info_dict': {
 671             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 672             'ext': 'mp4',
 673             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 674             'description': 'md5:1525f17448c4ee262b64b8f0c9ce66c8',
 675             'timestamp': 1434713142,
 676             'upload_date': '20150619',
 677         },
 678         'params': {
 679             'skip_download': True,
 680         }
 681     }, {
 682         # single video from video playlist embedded with vxp-playlist-data JSON
 683         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 684         'info_dict': {
 685             'id': 'p02w6qjc',
 686             'ext': 'mp4',
 687             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 688             'duration': 56,
 689             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 690         },
 691         'params': {
 692             'skip_download': True,
 693         }
 694     }, {
 695         # single video story with digitalData
 696         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 697         'info_dict': {
 698             'id': 'p02q6gc4',
 699             'ext': 'flv',
 700             'title': 'Sri Lanka’s spicy secret',
 701             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 702             'timestamp': 1437674293,
 703             'upload_date': '20150723',
 704         },
 705         'params': {
 706             # rtmp download
 707             'skip_download': True,
 708         }
 709     }, {
 710         # single video story without digitalData
 711         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 712         'info_dict': {
 713             'id': 'p018zqqg',
 714             'ext': 'mp4',
 715             'title': 'Hyundai Santa Fe Sport: Rock star',
 716             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 717             'timestamp': 1415867444,
 718             'upload_date': '20141113',
 719         },
 720         'params': {
 721             # rtmp download
 722             'skip_download': True,
 723         }
 724     }, {
 725         # single video embedded with Morph
 726         'url': 'http://www.bbc.co.uk/sport/live/olympics/36895975',
 727         'info_dict': {
 728             'id': 'p041vhd0',
 729             'ext': 'mp4',
 730             'title': "Nigeria v Japan - Men's First Round",
 731             'description': 'Live coverage of the first round from Group B at the Amazonia Arena.',
 732             'duration': 7980,
 733             'uploader': 'BBC Sport',
 734             'uploader_id': 'bbc_sport',
 735         },
 736         'params': {
 737             # m3u8 download
 738             'skip_download': True,
 739         },
 740         'skip': 'Georestricted to UK',
 741     }, {
 742         # single video with playlist.sxml URL in playlist param
 743         'url': 'http://www.bbc.com/sport/0/football/33653409',
 744         'info_dict': {
 745             'id': 'p02xycnp',
 746             'ext': 'mp4',
 747             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 748             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 749             'duration': 140,
 750         },
 751         'params': {
 752             # rtmp download
 753             'skip_download': True,
 754         }
 755     }, {
 756         # article with multiple videos embedded with playlist.sxml in playlist param
 757         'url': 'http://www.bbc.com/sport/0/football/34475836',
 758         'info_dict': {
 759             'id': '34475836',
 760             'title': 'Jurgen Klopp: Furious football from a witty and winning coach',
 761             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
 762         },
 763         'playlist_count': 3,
 764     }, {
 765         # school report article with single video
 766         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
 767         'info_dict': {
 768             'id': '35744779',
 769             'title': 'School which breaks down barriers in Jerusalem',
 770         },
 771         'playlist_count': 1,
 772     }, {
 773         # single video with playlist URL from weather section
 774         'url': 'http://www.bbc.com/weather/features/33601775',
 775         'only_matching': True,
 776     }, {
 777         # custom redirection to www.bbc.com
 778         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 779         'only_matching': True,
 780     }, {
 781         # single video article embedded with data-media-vpid
 782         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
 783         'only_matching': True,
 784     }, {
 785         'url': 'https://www.bbc.co.uk/bbcthree/clip/73d0bbd0-abc3-4cea-b3c0-cdae21905eb1',
 786         'info_dict': {
 787             'id': 'p06556y7',
 788             'ext': 'mp4',
 789             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 790             'description': 'md5:4b7dfd063d5a789a1512e99662be3ddd',
 791         },
 792         'params': {
 793             'skip_download': True,
 794         }
 795     }, {
 796         # window.__PRELOADED_STATE__
 797         'url': 'https://www.bbc.co.uk/radio/play/b0b9z4yl',
 798         'info_dict': {
 799             'id': 'b0b9z4vz',
 800             'ext': 'mp4',
 801             'title': 'Prom 6: An American in Paris and Turangalila',
 802             'description': 'md5:51cf7d6f5c8553f197e58203bc78dff8',
 803             'uploader': 'Radio 3',
 804             'uploader_id': 'bbc_radio_three',
 805         },
 806     }, {
 807         'url': 'http://www.bbc.co.uk/learningenglish/chinese/features/lingohack/ep-181227',
 808         'info_dict': {
 809             'id': 'p06w9tws',
 810             'ext': 'mp4',
 811             'title': 'md5:2fabf12a726603193a2879a055f72514',
 812             'description': 'Learn English words and phrases from this story',
 813         },
 814         'add_ie': [BBCCoUkIE.ie_key()],
 815     }]
 816
 817     @classmethod
 818     def suitable(cls, url):
 819         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerPlaylistIE, 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 Exception 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 playlist_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         # Morph based embed (e.g. http://www.bbc.co.uk/sport/live/olympics/36895975)
1003         # There are several setPayload calls may be present but the video
1004         # seems to be always related to the first one
1005         morph_payload = self._parse_json(
1006             self._search_regex(
1007                 r'Morph\.setPayload\([^,]+,\s*({.+?})\);',
1008                 webpage, 'morph payload', default='{}'),
1009             playlist_id, fatal=False)
1010         if morph_payload:
1011             components = try_get(morph_payload, lambda x: x['body']['components'], list) or []
1012             for component in components:
1013                 if not isinstance(component, dict):
1014                     continue
1015                 lead_media = try_get(component, lambda x: x['props']['leadMedia'], dict)
1016                 if not lead_media:
1017                     continue
1018                 identifiers = lead_media.get('identifiers')
1019                 if not identifiers or not isinstance(identifiers, dict):
1020                     continue
1021                 programme_id = identifiers.get('vpid') or identifiers.get('playablePid')
1022                 if not programme_id:
1023                     continue
1024                 title = lead_media.get('title') or self._og_search_title(webpage)
1025                 formats, subtitles = self._download_media_selector(programme_id)
1026                 self._sort_formats(formats)
1027                 description = lead_media.get('summary')
1028                 uploader = lead_media.get('masterBrand')
1029                 uploader_id = lead_media.get('mid')
1030                 duration = None
1031                 duration_d = lead_media.get('duration')
1032                 if isinstance(duration_d, dict):
1033                     duration = parse_duration(dict_get(
1034                         duration_d, ('rawDuration', 'formattedDuration', 'spokenDuration')))
1035                 return {
1036                     'id': programme_id,
1037                     'title': title,
1038                     'description': description,
1039                     'duration': duration,
1040                     'uploader': uploader,
1041                     'uploader_id': uploader_id,
1042                     'formats': formats,
1043                     'subtitles': subtitles,
1044                 }
1045
1046         preload_state = self._parse_json(self._search_regex(
1047             r'window\.__PRELOADED_STATE__\s*=\s*({.+?});', webpage,
1048             'preload state', default='{}'), playlist_id, fatal=False)
1049         if preload_state:
1050             current_programme = preload_state.get('programmes', {}).get('current') or {}
1051             programme_id = current_programme.get('id')
1052             if current_programme and programme_id and current_programme.get('type') == 'playable_item':
1053                 title = current_programme.get('titles', {}).get('tertiary') or playlist_title
1054                 formats, subtitles = self._download_media_selector(programme_id)
1055                 self._sort_formats(formats)
1056                 synopses = current_programme.get('synopses') or {}
1057                 network = current_programme.get('network') or {}
1058                 duration = int_or_none(
1059                     current_programme.get('duration', {}).get('value'))
1060                 thumbnail = None
1061                 image_url = current_programme.get('image_url')
1062                 if image_url:
1063                     thumbnail = image_url.replace('{recipe}', '1920x1920')
1064                 return {
1065                     'id': programme_id,
1066                     'title': title,
1067                     'description': dict_get(synopses, ('long', 'medium', 'short')),
1068                     'thumbnail': thumbnail,
1069                     'duration': duration,
1070                     'uploader': network.get('short_title'),
1071                     'uploader_id': network.get('id'),
1072                     'formats': formats,
1073                     'subtitles': subtitles,
1074                 }
1075
1076         bbc3_config = self._parse_json(
1077             self._search_regex(
1078                 r'(?s)bbcthreeConfig\s*=\s*({.+?})\s*;\s*<', webpage,
1079                 'bbcthree config', default='{}'),
1080             playlist_id, transform_source=js_to_json, fatal=False)
1081         if bbc3_config:
1082             bbc3_playlist = try_get(
1083                 bbc3_config, lambda x: x['payload']['content']['bbcMedia']['playlist'],
1084                 dict)
1085             if bbc3_playlist:
1086                 playlist_title = bbc3_playlist.get('title') or playlist_title
1087                 thumbnail = bbc3_playlist.get('holdingImageURL')
1088                 entries = []
1089                 for bbc3_item in bbc3_playlist['items']:
1090                     programme_id = bbc3_item.get('versionID')
1091                     if not programme_id:
1092                         continue
1093                     formats, subtitles = self._download_media_selector(programme_id)
1094                     self._sort_formats(formats)
1095                     entries.append({
1096                         'id': programme_id,
1097                         'title': playlist_title,
1098                         'thumbnail': thumbnail,
1099                         'timestamp': timestamp,
1100                         'formats': formats,
1101                         'subtitles': subtitles,
1102                     })
1103                 return self.playlist_result(
1104                     entries, playlist_id, playlist_title, playlist_description)
1105
1106         def extract_all(pattern):
1107             return list(filter(None, map(
1108                 lambda s: self._parse_json(s, playlist_id, fatal=False),
1109                 re.findall(pattern, webpage))))
1110
1111         # Multiple video article (e.g.
1112         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
1113         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
1114         entries = []
1115         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
1116             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
1117             if embed_url and re.match(EMBED_URL, embed_url):
1118                 entries.append(embed_url)
1119         entries.extend(re.findall(
1120             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
1121         if entries:
1122             return self.playlist_result(
1123                 [self.url_result(entry_, 'BBCCoUk') for entry_ in entries],
1124                 playlist_id, playlist_title, playlist_description)
1125
1126         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
1127         medias = extract_all(r"data-media-meta='({[^']+})'")
1128
1129         if not medias:
1130             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
1131             media_asset = self._search_regex(
1132                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
1133                 webpage, 'media asset', default=None)
1134             if media_asset:
1135                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
1136                 medias = []
1137                 for video in media_asset_page.get('videos', {}).values():
1138                     medias.extend(video.values())
1139
1140         if not medias:
1141             # Multiple video playlist with single `now playing` entry (e.g.
1142             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
1143             vxp_playlist = self._parse_json(
1144                 self._search_regex(
1145                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
1146                     webpage, 'playlist data'),
1147                 playlist_id)
1148             playlist_medias = []
1149             for item in vxp_playlist:
1150                 media = item.get('media')
1151                 if not media:
1152                     continue
1153                 playlist_medias.append(media)
1154                 # Download single video if found media with asset id matching the video id from URL
1155                 if item.get('advert', {}).get('assetId') == playlist_id:
1156                     medias = [media]
1157                     break
1158             # Fallback to the whole playlist
1159             if not medias:
1160                 medias = playlist_medias
1161
1162         entries = []
1163         for num, media_meta in enumerate(medias, start=1):
1164             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
1165             if not formats:
1166                 continue
1167             self._sort_formats(formats)
1168
1169             video_id = media_meta.get('externalId')
1170             if not video_id:
1171                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
1172
1173             title = media_meta.get('caption')
1174             if not title:
1175                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
1176
1177             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
1178
1179             images = []
1180             for image in media_meta.get('images', {}).values():
1181                 images.extend(image.values())
1182             if 'image' in media_meta:
1183                 images.append(media_meta['image'])
1184
1185             thumbnails = [{
1186                 'url': image.get('href'),
1187                 'width': int_or_none(image.get('width')),
1188                 'height': int_or_none(image.get('height')),
1189             } for image in images]
1190
1191             entries.append({
1192                 'id': video_id,
1193                 'title': title,
1194                 'thumbnails': thumbnails,
1195                 'duration': duration,
1196                 'timestamp': timestamp,
1197                 'formats': formats,
1198                 'subtitles': subtitles,
1199             })
1200
1201         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
1202
1203
1204 class BBCCoUkArticleIE(InfoExtractor):
1205     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
1206     IE_NAME = 'bbc.co.uk:article'
1207     IE_DESC = 'BBC articles'
1208
1209     _TEST = {
1210         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
1211         'info_dict': {
1212             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
1213             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
1214             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
1215         },
1216         'playlist_count': 4,
1217         'add_ie': ['BBCCoUk'],
1218     }
1219
1220     def _real_extract(self, url):
1221         playlist_id = self._match_id(url)
1222
1223         webpage = self._download_webpage(url, playlist_id)
1224
1225         title = self._og_search_title(webpage)
1226         description = self._og_search_description(webpage).strip()
1227
1228         entries = [self.url_result(programme_url) for programme_url in re.findall(
1229             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
1230
1231         return self.playlist_result(entries, playlist_id, title, description)
1232
1233
1234 class BBCCoUkPlaylistBaseIE(InfoExtractor):
1235     def _entries(self, webpage, url, playlist_id):
1236         single_page = 'page' in compat_urlparse.parse_qs(
1237             compat_urlparse.urlparse(url).query)
1238         for page_num in itertools.count(2):
1239             for video_id in re.findall(
1240                     self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage):
1241                 yield self.url_result(
1242                     self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
1243             if single_page:
1244                 return
1245             next_page = self._search_regex(
1246                 r'<li[^>]+class=(["\'])pagination_+next\1[^>]*><a[^>]+href=(["\'])(?P<url>(?:(?!\2).)+)\2',
1247                 webpage, 'next page url', default=None, group='url')
1248             if not next_page:
1249                 break
1250             webpage = self._download_webpage(
1251                 compat_urlparse.urljoin(url, next_page), playlist_id,
1252                 'Downloading page %d' % page_num, page_num)
1253
1254     def _real_extract(self, url):
1255         playlist_id = self._match_id(url)
1256
1257         webpage = self._download_webpage(url, playlist_id)
1258
1259         title, description = self._extract_title_and_description(webpage)
1260
1261         return self.playlist_result(
1262             self._entries(webpage, url, playlist_id),
1263             playlist_id, title, description)
1264
1265
1266 class BBCCoUkIPlayerPlaylistIE(BBCCoUkPlaylistBaseIE):
1267     IE_NAME = 'bbc.co.uk:iplayer:playlist'
1268     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/(?:episodes|group)/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
1269     _URL_TEMPLATE = 'http://www.bbc.co.uk/iplayer/episode/%s'
1270     _VIDEO_ID_TEMPLATE = r'data-ip-id=["\'](%s)'
1271     _TESTS = [{
1272         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1273         'info_dict': {
1274             'id': 'b05rcz9v',
1275             'title': 'The Disappearance',
1276             'description': 'French thriller serial about a missing teenager.',
1277         },
1278         'playlist_mincount': 6,
1279         'skip': 'This programme is not currently available on BBC iPlayer',
1280     }, {
1281         # Available for over a year unlike 30 days for most other programmes
1282         'url': 'http://www.bbc.co.uk/iplayer/group/p02tcc32',
1283         'info_dict': {
1284             'id': 'p02tcc32',
1285             'title': 'Bohemian Icons',
1286             'description': 'md5:683e901041b2fe9ba596f2ab04c4dbe7',
1287         },
1288         'playlist_mincount': 10,
1289     }]
1290
1291     def _extract_title_and_description(self, webpage):
1292         title = self._search_regex(r'<h1>([^<]+)</h1>', webpage, 'title', fatal=False)
1293         description = self._search_regex(
1294             r'<p[^>]+class=(["\'])subtitle\1[^>]*>(?P<value>[^<]+)</p>',
1295             webpage, 'description', fatal=False, group='value')
1296         return title, description
1297
1298
1299 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1300     IE_NAME = 'bbc.co.uk:playlist'
1301     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1302     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1303     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1304     _TESTS = [{
1305         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1306         'info_dict': {
1307             'id': 'b05rcz9v',
1308             'title': 'The Disappearance - Clips - BBC Four',
1309             'description': 'French thriller serial about a missing teenager.',
1310         },
1311         'playlist_mincount': 7,
1312     }, {
1313         # multipage playlist, explicit page
1314         'url': 'http://www.bbc.co.uk/programmes/b00mfl7n/clips?page=1',
1315         'info_dict': {
1316             'id': 'b00mfl7n',
1317             'title': 'Frozen Planet - Clips - BBC One',
1318             'description': 'md5:65dcbf591ae628dafe32aa6c4a4a0d8c',
1319         },
1320         'playlist_mincount': 24,
1321     }, {
1322         # multipage playlist, all pages
1323         'url': 'http://www.bbc.co.uk/programmes/b00mfl7n/clips',
1324         'info_dict': {
1325             'id': 'b00mfl7n',
1326             'title': 'Frozen Planet - Clips - BBC One',
1327             'description': 'md5:65dcbf591ae628dafe32aa6c4a4a0d8c',
1328         },
1329         'playlist_mincount': 142,
1330     }, {
1331         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1332         'only_matching': True,
1333     }, {
1334         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1335         'only_matching': True,
1336     }, {
1337         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1338         'only_matching': True,
1339     }]
1340
1341     def _extract_title_and_description(self, webpage):
1342         title = self._og_search_title(webpage, fatal=False)
1343         description = self._og_search_description(webpage)
1344         return title, description