yt_dlp/extractor/bbc.py

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