git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/bbc.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import itertools
   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     urlencode_postdata,
  21     urljoin,
  22 )
  23 from ..compat import (
  24     compat_HTTPError,
  25     compat_urlparse,
  26 )
  27
  28
  29 class BBCCoUkIE(InfoExtractor):
  30     IE_NAME = 'bbc.co.uk'
  31     IE_DESC = 'BBC iPlayer'
  32     _ID_REGEX = r'(?:[pbm][\da-z]{7}|w[\da-z]{7,14})'
  33     _VALID_URL = r'''(?x)
  34                     https?://
  35                         (?:www\.)?bbc\.co\.uk/
  36                         (?:
  37                             programmes/(?!articles/)|
  38                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
  39                             music/(?:clips|audiovideo/popular)[/#]|
  40                             radio/player/|
  41                             events/[^/]+/play/[^/]+/
  42                         )
  43                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
  44                     ''' % _ID_REGEX
  45
  46     _LOGIN_URL = 'https://account.bbc.com/signin'
  47     _NETRC_MACHINE = 'bbc'
  48
  49     _MEDIASELECTOR_URLS = [
  50         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  51         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  52         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  53         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  54         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  55     ]
  56
  57     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  58     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  59
  60     _NAMESPACES = (
  61         _MEDIASELECTION_NS,
  62         _EMP_PLAYLIST_NS,
  63     )
  64
  65     _TESTS = [
  66         {
  67             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  68             'info_dict': {
  69                 'id': 'b039d07m',
  70                 'ext': 'flv',
  71                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  72                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  73             },
  74             'params': {
  75                 # rtmp download
  76                 'skip_download': True,
  77             }
  78         },
  79         {
  80             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  81             'info_dict': {
  82                 'id': 'b00yng1d',
  83                 'ext': 'flv',
  84                 'title': 'The Man in Black: Series 3: The Printed Name',
  85                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  86                 'duration': 1800,
  87             },
  88             'params': {
  89                 # rtmp download
  90                 'skip_download': True,
  91             },
  92             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  93         },
  94         {
  95             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  96             'info_dict': {
  97                 'id': 'b00yng1d',
  98                 'ext': 'flv',
  99                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
 100                 '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.',
 101                 'duration': 5100,
 102             },
 103             'params': {
 104                 # rtmp download
 105                 'skip_download': True,
 106             },
 107             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 108         },
 109         {
 110             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
 111             'info_dict': {
 112                 'id': 'b03k3pb7',
 113                 'ext': 'flv',
 114                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 115                 'description': '2. Invasion',
 116                 'duration': 3600,
 117             },
 118             'params': {
 119                 # rtmp download
 120                 'skip_download': True,
 121             },
 122             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 123         }, {
 124             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 125             'info_dict': {
 126                 'id': 'b04v209v',
 127                 'ext': 'flv',
 128                 'title': 'Pete Tong, The Essential New Tune Special',
 129                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 130                 'duration': 10800,
 131             },
 132             'params': {
 133                 # rtmp download
 134                 'skip_download': True,
 135             },
 136             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 137         }, {
 138             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
 139             'note': 'Audio',
 140             'info_dict': {
 141                 'id': 'p022h44j',
 142                 'ext': 'flv',
 143                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 144                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 145                 'duration': 227,
 146             },
 147             'params': {
 148                 # rtmp download
 149                 'skip_download': True,
 150             }
 151         }, {
 152             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 153             'note': 'Video',
 154             'info_dict': {
 155                 'id': 'p025c103',
 156                 'ext': 'flv',
 157                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 158                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 159                 'duration': 226,
 160             },
 161             'params': {
 162                 # rtmp download
 163                 'skip_download': True,
 164             }
 165         }, {
 166             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 167             'info_dict': {
 168                 'id': 'p02n76xf',
 169                 'ext': 'flv',
 170                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 171                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 172                 'duration': 3540,
 173             },
 174             'params': {
 175                 # rtmp download
 176                 'skip_download': True,
 177             },
 178             'skip': 'geolocation',
 179         }, {
 180             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 181             'info_dict': {
 182                 'id': 'b05zmgw1',
 183                 'ext': 'flv',
 184                 '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.',
 185                 'title': 'Royal Academy Summer Exhibition',
 186                 'duration': 3540,
 187             },
 188             'params': {
 189                 # rtmp download
 190                 'skip_download': True,
 191             },
 192             'skip': 'geolocation',
 193         }, {
 194             # iptv-all mediaset fails with geolocation however there is no geo restriction
 195             # for this programme at all
 196             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
 197             'info_dict': {
 198                 'id': 'b06rkms3',
 199                 'ext': 'flv',
 200                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 201                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 202             },
 203             'params': {
 204                 # rtmp download
 205                 'skip_download': True,
 206             },
 207             'skip': 'Now it\'s really geo-restricted',
 208         }, {
 209             # compact player (https://github.com/rg3/youtube-dl/issues/8147)
 210             'url': 'http://www.bbc.co.uk/programmes/p028bfkf/player',
 211             'info_dict': {
 212                 'id': 'p028bfkj',
 213                 'ext': 'flv',
 214                 'title': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 215                 'description': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 216             },
 217             'params': {
 218                 # rtmp download
 219                 'skip_download': True,
 220             },
 221         }, {
 222             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 223             'only_matching': True,
 224         }, {
 225             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 226             'only_matching': True,
 227         }, {
 228             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 229             'only_matching': True,
 230         }, {
 231             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 232             'only_matching': True,
 233         }, {
 234             'url': 'https://www.bbc.co.uk/music/audiovideo/popular#p055bc55',
 235             'only_matching': True,
 236         }, {
 237             'url': 'http://www.bbc.co.uk/programmes/w3csv1y9',
 238             'only_matching': True,
 239         }, {
 240             'url': 'https://www.bbc.co.uk/programmes/m00005xn',
 241             'only_matching': True,
 242         }, {
 243             'url': 'https://www.bbc.co.uk/programmes/w172w4dww1jqt5s',
 244             'only_matching': True,
 245         }]
 246
 247     _USP_RE = r'/([^/]+?)\.ism(?:\.hlsv2\.ism)?/[^/]+\.m3u8'
 248
 249     def _login(self):
 250         username, password = self._get_login_info()
 251         if username is None:
 252             return
 253
 254         login_page = self._download_webpage(
 255             self._LOGIN_URL, None, 'Downloading signin page')
 256
 257         login_form = self._hidden_inputs(login_page)
 258
 259         login_form.update({
 260             'username': username,
 261             'password': password,
 262         })
 263
 264         post_url = urljoin(self._LOGIN_URL, self._search_regex(
 265             r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
 266             'post url', default=self._LOGIN_URL, group='url'))
 267
 268         response, urlh = self._download_webpage_handle(
 269             post_url, None, 'Logging in', data=urlencode_postdata(login_form),
 270             headers={'Referer': self._LOGIN_URL})
 271
 272         if self._LOGIN_URL in urlh.geturl():
 273             error = clean_html(get_element_by_class('form-message', response))
 274             if error:
 275                 raise ExtractorError(
 276                     'Unable to login: %s' % error, expected=True)
 277             raise ExtractorError('Unable to log in')
 278
 279     def _real_initialize(self):
 280         self._login()
 281
 282     class MediaSelectionError(Exception):
 283         def __init__(self, id):
 284             self.id = id
 285
 286     def _extract_asx_playlist(self, connection, programme_id):
 287         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 288         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 289
 290     def _extract_items(self, playlist):
 291         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 292
 293     def _findall_ns(self, element, xpath):
 294         elements = []
 295         for ns in self._NAMESPACES:
 296             elements.extend(element.findall(xpath % ns))
 297         return elements
 298
 299     def _extract_medias(self, media_selection):
 300         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 301         if error is None:
 302             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 303         if error is not None:
 304             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 305         return self._findall_ns(media_selection, './{%s}media')
 306
 307     def _extract_connections(self, media):
 308         return self._findall_ns(media, './{%s}connection')
 309
 310     def _get_subtitles(self, media, programme_id):
 311         subtitles = {}
 312         for connection in self._extract_connections(media):
 313             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 314             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 315             subtitles[lang] = [
 316                 {
 317                     'url': connection.get('href'),
 318                     'ext': 'ttml',
 319                 },
 320             ]
 321         return subtitles
 322
 323     def _raise_extractor_error(self, media_selection_error):
 324         raise ExtractorError(
 325             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 326             expected=True)
 327
 328     def _download_media_selector(self, programme_id):
 329         last_exception = None
 330         for mediaselector_url in self._MEDIASELECTOR_URLS:
 331             try:
 332                 return self._download_media_selector_url(
 333                     mediaselector_url % programme_id, programme_id)
 334             except BBCCoUkIE.MediaSelectionError as e:
 335                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 336                     last_exception = e
 337                     continue
 338                 self._raise_extractor_error(e)
 339         self._raise_extractor_error(last_exception)
 340
 341     def _download_media_selector_url(self, url, programme_id=None):
 342         media_selection = self._download_xml(
 343             url, programme_id, 'Downloading media selection XML',
 344             expected_status=(403, 404))
 345         return self._process_media_selector(media_selection, programme_id)
 346
 347     def _process_media_selector(self, media_selection, programme_id):
 348         formats = []
 349         subtitles = None
 350         urls = []
 351
 352         for media in self._extract_medias(media_selection):
 353             kind = media.get('kind')
 354             if kind in ('video', 'audio'):
 355                 bitrate = int_or_none(media.get('bitrate'))
 356                 encoding = media.get('encoding')
 357                 service = media.get('service')
 358                 width = int_or_none(media.get('width'))
 359                 height = int_or_none(media.get('height'))
 360                 file_size = int_or_none(media.get('media_file_size'))
 361                 for connection in self._extract_connections(media):
 362                     href = connection.get('href')
 363                     if href in urls:
 364                         continue
 365                     if href:
 366                         urls.append(href)
 367                     conn_kind = connection.get('kind')
 368                     protocol = connection.get('protocol')
 369                     supplier = connection.get('supplier')
 370                     transfer_format = connection.get('transferFormat')
 371                     format_id = supplier or conn_kind or protocol
 372                     if service:
 373                         format_id = '%s_%s' % (service, format_id)
 374                     # ASX playlist
 375                     if supplier == 'asx':
 376                         for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 377                             formats.append({
 378                                 'url': ref,
 379                                 'format_id': 'ref%s_%s' % (i, format_id),
 380                             })
 381                     elif transfer_format == 'dash':
 382                         formats.extend(self._extract_mpd_formats(
 383                             href, programme_id, mpd_id=format_id, fatal=False))
 384                     elif transfer_format == 'hls':
 385                         formats.extend(self._extract_m3u8_formats(
 386                             href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 387                             m3u8_id=format_id, fatal=False))
 388                         if re.search(self._USP_RE, href):
 389                             usp_formats = self._extract_m3u8_formats(
 390                                 re.sub(self._USP_RE, r'/\1.ism/\1.m3u8', href),
 391                                 programme_id, ext='mp4', entry_protocol='m3u8_native',
 392                                 m3u8_id=format_id, fatal=False)
 393                             for f in usp_formats:
 394                                 if f.get('height') and f['height'] > 720:
 395                                     continue
 396                                 formats.append(f)
 397                     elif transfer_format == 'hds':
 398                         formats.extend(self._extract_f4m_formats(
 399                             href, programme_id, f4m_id=format_id, fatal=False))
 400                     else:
 401                         if not service and not supplier and bitrate:
 402                             format_id += '-%d' % bitrate
 403                         fmt = {
 404                             'format_id': format_id,
 405                             'filesize': file_size,
 406                         }
 407                         if kind == 'video':
 408                             fmt.update({
 409                                 'width': width,
 410                                 'height': height,
 411                                 'tbr': bitrate,
 412                                 'vcodec': encoding,
 413                             })
 414                         else:
 415                             fmt.update({
 416                                 'abr': bitrate,
 417                                 'acodec': encoding,
 418                                 'vcodec': 'none',
 419                             })
 420                         if protocol in ('http', 'https'):
 421                             # Direct link
 422                             fmt.update({
 423                                 'url': href,
 424                             })
 425                         elif protocol == 'rtmp':
 426                             application = connection.get('application', 'ondemand')
 427                             auth_string = connection.get('authString')
 428                             identifier = connection.get('identifier')
 429                             server = connection.get('server')
 430                             fmt.update({
 431                                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 432                                 'play_path': identifier,
 433                                 'app': '%s?%s' % (application, auth_string),
 434                                 'page_url': 'http://www.bbc.co.uk',
 435                                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 436                                 'rtmp_live': False,
 437                                 'ext': 'flv',
 438                             })
 439                         else:
 440                             continue
 441                         formats.append(fmt)
 442             elif kind == 'captions':
 443                 subtitles = self.extract_subtitles(media, programme_id)
 444         return formats, subtitles
 445
 446     def _download_playlist(self, playlist_id):
 447         try:
 448             playlist = self._download_json(
 449                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 450                 playlist_id, 'Downloading playlist JSON')
 451
 452             version = playlist.get('defaultAvailableVersion')
 453             if version:
 454                 smp_config = version['smpConfig']
 455                 title = smp_config['title']
 456                 description = smp_config['summary']
 457                 for item in smp_config['items']:
 458                     kind = item['kind']
 459                     if kind not in ('programme', 'radioProgramme'):
 460                         continue
 461                     programme_id = item.get('vpid')
 462                     duration = int_or_none(item.get('duration'))
 463                     formats, subtitles = self._download_media_selector(programme_id)
 464                 return programme_id, title, description, duration, formats, subtitles
 465         except ExtractorError as ee:
 466             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 467                 raise
 468
 469         # fallback to legacy playlist
 470         return self._process_legacy_playlist(playlist_id)
 471
 472     def _process_legacy_playlist_url(self, url, display_id):
 473         playlist = self._download_legacy_playlist_url(url, display_id)
 474         return self._extract_from_legacy_playlist(playlist, display_id)
 475
 476     def _process_legacy_playlist(self, playlist_id):
 477         return self._process_legacy_playlist_url(
 478             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 479
 480     def _download_legacy_playlist_url(self, url, playlist_id=None):
 481         return self._download_xml(
 482             url, playlist_id, 'Downloading legacy playlist XML')
 483
 484     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 485         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 486         if no_items is not None:
 487             reason = no_items.get('reason')
 488             if reason == 'preAvailability':
 489                 msg = 'Episode %s is not yet available' % playlist_id
 490             elif reason == 'postAvailability':
 491                 msg = 'Episode %s is no longer available' % playlist_id
 492             elif reason == 'noMedia':
 493                 msg = 'Episode %s is not currently available' % playlist_id
 494             else:
 495                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 496             raise ExtractorError(msg, expected=True)
 497
 498         for item in self._extract_items(playlist):
 499             kind = item.get('kind')
 500             if kind not in ('programme', 'radioProgramme'):
 501                 continue
 502             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 503             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 504             description = description_el.text if description_el is not None else None
 505
 506             def get_programme_id(item):
 507                 def get_from_attributes(item):
 508                     for p in('identifier', 'group'):
 509                         value = item.get(p)
 510                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 511                             return value
 512                 get_from_attributes(item)
 513                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 514                 if mediator is not None:
 515                     return get_from_attributes(mediator)
 516
 517             programme_id = get_programme_id(item)
 518             duration = int_or_none(item.get('duration'))
 519
 520             if programme_id:
 521                 formats, subtitles = self._download_media_selector(programme_id)
 522             else:
 523                 formats, subtitles = self._process_media_selector(item, playlist_id)
 524                 programme_id = playlist_id
 525
 526         return programme_id, title, description, duration, formats, subtitles
 527
 528     def _real_extract(self, url):
 529         group_id = self._match_id(url)
 530
 531         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 532
 533         error = self._search_regex(
 534             r'<div\b[^>]+\bclass=["\']smp__message delta["\'][^>]*>([^<]+)<',
 535             webpage, 'error', default=None)
 536         if error:
 537             raise ExtractorError(error, expected=True)
 538
 539         programme_id = None
 540         duration = None
 541
 542         tviplayer = self._search_regex(
 543             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 544             webpage, 'player', default=None)
 545
 546         if tviplayer:
 547             player = self._parse_json(tviplayer, group_id).get('player', {})
 548             duration = int_or_none(player.get('duration'))
 549             programme_id = player.get('vpid')
 550
 551         if not programme_id:
 552             programme_id = self._search_regex(
 553                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 554
 555         if programme_id:
 556             formats, subtitles = self._download_media_selector(programme_id)
 557             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 558                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 559                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
 560             description = self._search_regex(
 561                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 562                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
 563                 webpage, 'description', default=None)
 564             if not description:
 565                 description = self._html_search_meta('description', webpage)
 566         else:
 567             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 568
 569         self._sort_formats(formats)
 570
 571         return {
 572             'id': programme_id,
 573             'title': title,
 574             'description': description,
 575             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 576             'duration': duration,
 577             'formats': formats,
 578             'subtitles': subtitles,
 579         }
 580
 581
 582 class BBCIE(BBCCoUkIE):
 583     IE_NAME = 'bbc'
 584     IE_DESC = 'BBC'
 585     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 586
 587     _MEDIASELECTOR_URLS = [
 588         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 589         # even not geo restricted at all
 590         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 591         # Provides more formats, namely direct mp4 links, but fails on some videos with
 592         # notukerror for non UK (?) users (e.g.
 593         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 594         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 595         # Provides fewer formats, but works everywhere for everybody (hopefully)
 596         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 597     ]
 598
 599     _TESTS = [{
 600         # article with multiple videos embedded with data-playable containing vpids
 601         'url': 'http://www.bbc.com/news/world-europe-32668511',
 602         'info_dict': {
 603             'id': 'world-europe-32668511',
 604             'title': 'Russia stages massive WW2 parade despite Western boycott',
 605             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 606         },
 607         'playlist_count': 2,
 608     }, {
 609         # article with multiple videos embedded with data-playable (more videos)
 610         'url': 'http://www.bbc.com/news/business-28299555',
 611         'info_dict': {
 612             'id': 'business-28299555',
 613             'title': 'Farnborough Airshow: Video highlights',
 614             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 615         },
 616         'playlist_count': 9,
 617         'skip': 'Save time',
 618     }, {
 619         # article with multiple videos embedded with `new SMP()`
 620         # broken
 621         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 622         'info_dict': {
 623             'id': '3662a707-0af9-3149-963f-47bea720b460',
 624             'title': 'BUGGER',
 625         },
 626         'playlist_count': 18,
 627     }, {
 628         # single video embedded with data-playable containing vpid
 629         'url': 'http://www.bbc.com/news/world-europe-32041533',
 630         'info_dict': {
 631             'id': 'p02mprgb',
 632             'ext': 'mp4',
 633             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 634             'description': 'md5:2868290467291b37feda7863f7a83f54',
 635             'duration': 47,
 636             'timestamp': 1427219242,
 637             'upload_date': '20150324',
 638         },
 639         'params': {
 640             # rtmp download
 641             'skip_download': True,
 642         }
 643     }, {
 644         # article with single video embedded with data-playable containing XML playlist
 645         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 646         # and playlist with f4m and m3u8 as streamingUrl
 647         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 648         'info_dict': {
 649             'id': '150615_telabyad_kentin_cogu',
 650             'ext': 'mp4',
 651             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 652             'description': 'md5:33a4805a855c9baf7115fcbde57e7025',
 653             'timestamp': 1434397334,
 654             'upload_date': '20150615',
 655         },
 656         'params': {
 657             'skip_download': True,
 658         }
 659     }, {
 660         # single video embedded with data-playable containing XML playlists (regional section)
 661         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 662         'info_dict': {
 663             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 664             'ext': 'mp4',
 665             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 666             'description': 'md5:1525f17448c4ee262b64b8f0c9ce66c8',
 667             'timestamp': 1434713142,
 668             'upload_date': '20150619',
 669         },
 670         'params': {
 671             'skip_download': True,
 672         }
 673     }, {
 674         # single video from video playlist embedded with vxp-playlist-data JSON
 675         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 676         'info_dict': {
 677             'id': 'p02w6qjc',
 678             'ext': 'mp4',
 679             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 680             'duration': 56,
 681             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 682         },
 683         'params': {
 684             'skip_download': True,
 685         }
 686     }, {
 687         # single video story with digitalData
 688         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 689         'info_dict': {
 690             'id': 'p02q6gc4',
 691             'ext': 'flv',
 692             'title': 'Sri Lanka’s spicy secret',
 693             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 694             'timestamp': 1437674293,
 695             'upload_date': '20150723',
 696         },
 697         'params': {
 698             # rtmp download
 699             'skip_download': True,
 700         }
 701     }, {
 702         # single video story without digitalData
 703         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 704         'info_dict': {
 705             'id': 'p018zqqg',
 706             'ext': 'mp4',
 707             'title': 'Hyundai Santa Fe Sport: Rock star',
 708             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 709             'timestamp': 1415867444,
 710             'upload_date': '20141113',
 711         },
 712         'params': {
 713             # rtmp download
 714             'skip_download': True,
 715         }
 716     }, {
 717         # single video embedded with Morph
 718         'url': 'http://www.bbc.co.uk/sport/live/olympics/36895975',
 719         'info_dict': {
 720             'id': 'p041vhd0',
 721             'ext': 'mp4',
 722             'title': "Nigeria v Japan - Men's First Round",
 723             'description': 'Live coverage of the first round from Group B at the Amazonia Arena.',
 724             'duration': 7980,
 725             'uploader': 'BBC Sport',
 726             'uploader_id': 'bbc_sport',
 727         },
 728         'params': {
 729             # m3u8 download
 730             'skip_download': True,
 731         },
 732         'skip': 'Georestricted to UK',
 733     }, {
 734         # single video with playlist.sxml URL in playlist param
 735         'url': 'http://www.bbc.com/sport/0/football/33653409',
 736         'info_dict': {
 737             'id': 'p02xycnp',
 738             'ext': 'mp4',
 739             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 740             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 741             'duration': 140,
 742         },
 743         'params': {
 744             # rtmp download
 745             'skip_download': True,
 746         }
 747     }, {
 748         # article with multiple videos embedded with playlist.sxml in playlist param
 749         'url': 'http://www.bbc.com/sport/0/football/34475836',
 750         'info_dict': {
 751             'id': '34475836',
 752             'title': 'Jurgen Klopp: Furious football from a witty and winning coach',
 753             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
 754         },
 755         'playlist_count': 3,
 756     }, {
 757         # school report article with single video
 758         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
 759         'info_dict': {
 760             'id': '35744779',
 761             'title': 'School which breaks down barriers in Jerusalem',
 762         },
 763         'playlist_count': 1,
 764     }, {
 765         # single video with playlist URL from weather section
 766         'url': 'http://www.bbc.com/weather/features/33601775',
 767         'only_matching': True,
 768     }, {
 769         # custom redirection to www.bbc.com
 770         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 771         'only_matching': True,
 772     }, {
 773         # single video article embedded with data-media-vpid
 774         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
 775         'only_matching': True,
 776     }, {
 777         'url': 'https://www.bbc.co.uk/bbcthree/clip/73d0bbd0-abc3-4cea-b3c0-cdae21905eb1',
 778         'info_dict': {
 779             'id': 'p06556y7',
 780             'ext': 'mp4',
 781             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 782             'description': 'md5:4b7dfd063d5a789a1512e99662be3ddd',
 783         },
 784         'params': {
 785             'skip_download': True,
 786         }
 787     }, {
 788         # window.__PRELOADED_STATE__
 789         'url': 'https://www.bbc.co.uk/radio/play/b0b9z4yl',
 790         'info_dict': {
 791             'id': 'b0b9z4vz',
 792             'ext': 'mp4',
 793             'title': 'Prom 6: An American in Paris and Turangalila',
 794             'description': 'md5:51cf7d6f5c8553f197e58203bc78dff8',
 795             'uploader': 'Radio 3',
 796             'uploader_id': 'bbc_radio_three',
 797         },
 798     }]
 799
 800     @classmethod
 801     def suitable(cls, url):
 802         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerPlaylistIE, BBCCoUkPlaylistIE)
 803         return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
 804                 else super(BBCIE, cls).suitable(url))
 805
 806     def _extract_from_media_meta(self, media_meta, video_id):
 807         # Direct links to media in media metadata (e.g.
 808         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 809         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 810         source_files = media_meta.get('sourceFiles')
 811         if source_files:
 812             return [{
 813                 'url': f['url'],
 814                 'format_id': format_id,
 815                 'ext': f.get('encoding'),
 816                 'tbr': float_or_none(f.get('bitrate'), 1000),
 817                 'filesize': int_or_none(f.get('filesize')),
 818             } for format_id, f in source_files.items() if f.get('url')], []
 819
 820         programme_id = media_meta.get('externalId')
 821         if programme_id:
 822             return self._download_media_selector(programme_id)
 823
 824         # Process playlist.sxml as legacy playlist
 825         href = media_meta.get('href')
 826         if href:
 827             playlist = self._download_legacy_playlist_url(href)
 828             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 829             return formats, subtitles
 830
 831         return [], []
 832
 833     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 834         programme_id, title, description, duration, formats, subtitles = \
 835             self._process_legacy_playlist_url(url, playlist_id)
 836         self._sort_formats(formats)
 837         return {
 838             'id': programme_id,
 839             'title': title,
 840             'description': description,
 841             'duration': duration,
 842             'timestamp': timestamp,
 843             'formats': formats,
 844             'subtitles': subtitles,
 845         }
 846
 847     def _real_extract(self, url):
 848         playlist_id = self._match_id(url)
 849
 850         webpage = self._download_webpage(url, playlist_id)
 851
 852         json_ld_info = self._search_json_ld(webpage, playlist_id, default={})
 853         timestamp = json_ld_info.get('timestamp')
 854
 855         playlist_title = json_ld_info.get('title')
 856         if not playlist_title:
 857             playlist_title = self._og_search_title(
 858                 webpage, default=None) or self._html_search_regex(
 859                 r'<title>(.+?)</title>', webpage, 'playlist title', default=None)
 860             if playlist_title:
 861                 playlist_title = re.sub(r'(.+)\s*-\s*BBC.*?$', r'\1', playlist_title).strip()
 862
 863         playlist_description = json_ld_info.get(
 864             'description') or self._og_search_description(webpage, default=None)
 865
 866         if not timestamp:
 867             timestamp = parse_iso8601(self._search_regex(
 868                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 869                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 870                  r'"datePublished":\s*"([^"]+)'],
 871                 webpage, 'date', default=None))
 872
 873         entries = []
 874
 875         # article with multiple videos embedded with playlist.sxml (e.g.
 876         # http://www.bbc.com/sport/0/football/34475836)
 877         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 878         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 879         if playlists:
 880             entries = [
 881                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 882                 for playlist_url in playlists]
 883
 884         # news article with multiple videos embedded with data-playable
 885         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 886         if data_playables:
 887             for _, data_playable_json in data_playables:
 888                 data_playable = self._parse_json(
 889                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 890                 if not data_playable:
 891                     continue
 892                 settings = data_playable.get('settings', {})
 893                 if settings:
 894                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 895                     # http://www.bbc.com/news/world-us-canada-34473351)
 896                     playlist_object = settings.get('playlistObject', {})
 897                     if playlist_object:
 898                         items = playlist_object.get('items')
 899                         if items and isinstance(items, list):
 900                             title = playlist_object['title']
 901                             description = playlist_object.get('summary')
 902                             duration = int_or_none(items[0].get('duration'))
 903                             programme_id = items[0].get('vpid')
 904                             formats, subtitles = self._download_media_selector(programme_id)
 905                             self._sort_formats(formats)
 906                             entries.append({
 907                                 'id': programme_id,
 908                                 'title': title,
 909                                 'description': description,
 910                                 'timestamp': timestamp,
 911                                 'duration': duration,
 912                                 'formats': formats,
 913                                 'subtitles': subtitles,
 914                             })
 915                     else:
 916                         # data-playable without vpid but with a playlist.sxml URLs
 917                         # in otherSettings.playlist (e.g.
 918                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 919                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 920                         if playlist:
 921                             entry = None
 922                             for key in ('streaming', 'progressiveDownload'):
 923                                 playlist_url = playlist.get('%sUrl' % key)
 924                                 if not playlist_url:
 925                                     continue
 926                                 try:
 927                                     info = self._extract_from_playlist_sxml(
 928                                         playlist_url, playlist_id, timestamp)
 929                                     if not entry:
 930                                         entry = info
 931                                     else:
 932                                         entry['title'] = info['title']
 933                                         entry['formats'].extend(info['formats'])
 934                                 except Exception as e:
 935                                     # Some playlist URL may fail with 500, at the same time
 936                                     # the other one may work fine (e.g.
 937                                     # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 938                                     if isinstance(e.cause, compat_HTTPError) and e.cause.code == 500:
 939                                         continue
 940                                     raise
 941                             if entry:
 942                                 self._sort_formats(entry['formats'])
 943                                 entries.append(entry)
 944
 945         if entries:
 946             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 947
 948         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 949         programme_id = self._search_regex(
 950             [r'data-(?:video-player|media)-vpid="(%s)"' % self._ID_REGEX,
 951              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 952              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 953             webpage, 'vpid', default=None)
 954
 955         if programme_id:
 956             formats, subtitles = self._download_media_selector(programme_id)
 957             self._sort_formats(formats)
 958             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 959             digital_data = self._parse_json(
 960                 self._search_regex(
 961                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 962                 programme_id, fatal=False)
 963             page_info = digital_data.get('page', {}).get('pageInfo', {})
 964             title = page_info.get('pageName') or self._og_search_title(webpage)
 965             description = page_info.get('description') or self._og_search_description(webpage)
 966             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 967             return {
 968                 'id': programme_id,
 969                 'title': title,
 970                 'description': description,
 971                 'timestamp': timestamp,
 972                 'formats': formats,
 973                 'subtitles': subtitles,
 974             }
 975
 976         # Morph based embed (e.g. http://www.bbc.co.uk/sport/live/olympics/36895975)
 977         # There are several setPayload calls may be present but the video
 978         # seems to be always related to the first one
 979         morph_payload = self._parse_json(
 980             self._search_regex(
 981                 r'Morph\.setPayload\([^,]+,\s*({.+?})\);',
 982                 webpage, 'morph payload', default='{}'),
 983             playlist_id, fatal=False)
 984         if morph_payload:
 985             components = try_get(morph_payload, lambda x: x['body']['components'], list) or []
 986             for component in components:
 987                 if not isinstance(component, dict):
 988                     continue
 989                 lead_media = try_get(component, lambda x: x['props']['leadMedia'], dict)
 990                 if not lead_media:
 991                     continue
 992                 identifiers = lead_media.get('identifiers')
 993                 if not identifiers or not isinstance(identifiers, dict):
 994                     continue
 995                 programme_id = identifiers.get('vpid') or identifiers.get('playablePid')
 996                 if not programme_id:
 997                     continue
 998                 title = lead_media.get('title') or self._og_search_title(webpage)
 999                 formats, subtitles = self._download_media_selector(programme_id)
1000                 self._sort_formats(formats)
1001                 description = lead_media.get('summary')
1002                 uploader = lead_media.get('masterBrand')
1003                 uploader_id = lead_media.get('mid')
1004                 duration = None
1005                 duration_d = lead_media.get('duration')
1006                 if isinstance(duration_d, dict):
1007                     duration = parse_duration(dict_get(
1008                         duration_d, ('rawDuration', 'formattedDuration', 'spokenDuration')))
1009                 return {
1010                     'id': programme_id,
1011                     'title': title,
1012                     'description': description,
1013                     'duration': duration,
1014                     'uploader': uploader,
1015                     'uploader_id': uploader_id,
1016                     'formats': formats,
1017                     'subtitles': subtitles,
1018                 }
1019
1020         preload_state = self._parse_json(self._search_regex(
1021             r'window\.__PRELOADED_STATE__\s*=\s*({.+?});', webpage,
1022             'preload state', default='{}'), playlist_id, fatal=False)
1023         if preload_state:
1024             current_programme = preload_state.get('programmes', {}).get('current') or {}
1025             programme_id = current_programme.get('id')
1026             if current_programme and programme_id and current_programme.get('type') == 'playable_item':
1027                 title = current_programme.get('titles', {}).get('tertiary') or playlist_title
1028                 formats, subtitles = self._download_media_selector(programme_id)
1029                 self._sort_formats(formats)
1030                 synopses = current_programme.get('synopses') or {}
1031                 network = current_programme.get('network') or {}
1032                 duration = int_or_none(
1033                     current_programme.get('duration', {}).get('value'))
1034                 thumbnail = None
1035                 image_url = current_programme.get('image_url')
1036                 if image_url:
1037                     thumbnail = image_url.replace('{recipe}', '1920x1920')
1038                 return {
1039                     'id': programme_id,
1040                     'title': title,
1041                     'description': dict_get(synopses, ('long', 'medium', 'short')),
1042                     'thumbnail': thumbnail,
1043                     'duration': duration,
1044                     'uploader': network.get('short_title'),
1045                     'uploader_id': network.get('id'),
1046                     'formats': formats,
1047                     'subtitles': subtitles,
1048                 }
1049
1050         bbc3_config = self._parse_json(
1051             self._search_regex(
1052                 r'(?s)bbcthreeConfig\s*=\s*({.+?})\s*;\s*<', webpage,
1053                 'bbcthree config', default='{}'),
1054             playlist_id, transform_source=js_to_json, fatal=False)
1055         if bbc3_config:
1056             bbc3_playlist = try_get(
1057                 bbc3_config, lambda x: x['payload']['content']['bbcMedia']['playlist'],
1058                 dict)
1059             if bbc3_playlist:
1060                 playlist_title = bbc3_playlist.get('title') or playlist_title
1061                 thumbnail = bbc3_playlist.get('holdingImageURL')
1062                 entries = []
1063                 for bbc3_item in bbc3_playlist['items']:
1064                     programme_id = bbc3_item.get('versionID')
1065                     if not programme_id:
1066                         continue
1067                     formats, subtitles = self._download_media_selector(programme_id)
1068                     self._sort_formats(formats)
1069                     entries.append({
1070                         'id': programme_id,
1071                         'title': playlist_title,
1072                         'thumbnail': thumbnail,
1073                         'timestamp': timestamp,
1074                         'formats': formats,
1075                         'subtitles': subtitles,
1076                     })
1077                 return self.playlist_result(
1078                     entries, playlist_id, playlist_title, playlist_description)
1079
1080         def extract_all(pattern):
1081             return list(filter(None, map(
1082                 lambda s: self._parse_json(s, playlist_id, fatal=False),
1083                 re.findall(pattern, webpage))))
1084
1085         # Multiple video article (e.g.
1086         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
1087         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
1088         entries = []
1089         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
1090             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
1091             if embed_url and re.match(EMBED_URL, embed_url):
1092                 entries.append(embed_url)
1093         entries.extend(re.findall(
1094             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
1095         if entries:
1096             return self.playlist_result(
1097                 [self.url_result(entry_, 'BBCCoUk') for entry_ in entries],
1098                 playlist_id, playlist_title, playlist_description)
1099
1100         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
1101         medias = extract_all(r"data-media-meta='({[^']+})'")
1102
1103         if not medias:
1104             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
1105             media_asset = self._search_regex(
1106                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
1107                 webpage, 'media asset', default=None)
1108             if media_asset:
1109                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
1110                 medias = []
1111                 for video in media_asset_page.get('videos', {}).values():
1112                     medias.extend(video.values())
1113
1114         if not medias:
1115             # Multiple video playlist with single `now playing` entry (e.g.
1116             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
1117             vxp_playlist = self._parse_json(
1118                 self._search_regex(
1119                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
1120                     webpage, 'playlist data'),
1121                 playlist_id)
1122             playlist_medias = []
1123             for item in vxp_playlist:
1124                 media = item.get('media')
1125                 if not media:
1126                     continue
1127                 playlist_medias.append(media)
1128                 # Download single video if found media with asset id matching the video id from URL
1129                 if item.get('advert', {}).get('assetId') == playlist_id:
1130                     medias = [media]
1131                     break
1132             # Fallback to the whole playlist
1133             if not medias:
1134                 medias = playlist_medias
1135
1136         entries = []
1137         for num, media_meta in enumerate(medias, start=1):
1138             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
1139             if not formats:
1140                 continue
1141             self._sort_formats(formats)
1142
1143             video_id = media_meta.get('externalId')
1144             if not video_id:
1145                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
1146
1147             title = media_meta.get('caption')
1148             if not title:
1149                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
1150
1151             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
1152
1153             images = []
1154             for image in media_meta.get('images', {}).values():
1155                 images.extend(image.values())
1156             if 'image' in media_meta:
1157                 images.append(media_meta['image'])
1158
1159             thumbnails = [{
1160                 'url': image.get('href'),
1161                 'width': int_or_none(image.get('width')),
1162                 'height': int_or_none(image.get('height')),
1163             } for image in images]
1164
1165             entries.append({
1166                 'id': video_id,
1167                 'title': title,
1168                 'thumbnails': thumbnails,
1169                 'duration': duration,
1170                 'timestamp': timestamp,
1171                 'formats': formats,
1172                 'subtitles': subtitles,
1173             })
1174
1175         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
1176
1177
1178 class BBCCoUkArticleIE(InfoExtractor):
1179     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
1180     IE_NAME = 'bbc.co.uk:article'
1181     IE_DESC = 'BBC articles'
1182
1183     _TEST = {
1184         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
1185         'info_dict': {
1186             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
1187             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
1188             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
1189         },
1190         'playlist_count': 4,
1191         'add_ie': ['BBCCoUk'],
1192     }
1193
1194     def _real_extract(self, url):
1195         playlist_id = self._match_id(url)
1196
1197         webpage = self._download_webpage(url, playlist_id)
1198
1199         title = self._og_search_title(webpage)
1200         description = self._og_search_description(webpage).strip()
1201
1202         entries = [self.url_result(programme_url) for programme_url in re.findall(
1203             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
1204
1205         return self.playlist_result(entries, playlist_id, title, description)
1206
1207
1208 class BBCCoUkPlaylistBaseIE(InfoExtractor):
1209     def _entries(self, webpage, url, playlist_id):
1210         single_page = 'page' in compat_urlparse.parse_qs(
1211             compat_urlparse.urlparse(url).query)
1212         for page_num in itertools.count(2):
1213             for video_id in re.findall(
1214                     self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage):
1215                 yield self.url_result(
1216                     self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
1217             if single_page:
1218                 return
1219             next_page = self._search_regex(
1220                 r'<li[^>]+class=(["\'])pagination_+next\1[^>]*><a[^>]+href=(["\'])(?P<url>(?:(?!\2).)+)\2',
1221                 webpage, 'next page url', default=None, group='url')
1222             if not next_page:
1223                 break
1224             webpage = self._download_webpage(
1225                 compat_urlparse.urljoin(url, next_page), playlist_id,
1226                 'Downloading page %d' % page_num, page_num)
1227
1228     def _real_extract(self, url):
1229         playlist_id = self._match_id(url)
1230
1231         webpage = self._download_webpage(url, playlist_id)
1232
1233         title, description = self._extract_title_and_description(webpage)
1234
1235         return self.playlist_result(
1236             self._entries(webpage, url, playlist_id),
1237             playlist_id, title, description)
1238
1239
1240 class BBCCoUkIPlayerPlaylistIE(BBCCoUkPlaylistBaseIE):
1241     IE_NAME = 'bbc.co.uk:iplayer:playlist'
1242     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/(?:episodes|group)/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
1243     _URL_TEMPLATE = 'http://www.bbc.co.uk/iplayer/episode/%s'
1244     _VIDEO_ID_TEMPLATE = r'data-ip-id=["\'](%s)'
1245     _TESTS = [{
1246         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1247         'info_dict': {
1248             'id': 'b05rcz9v',
1249             'title': 'The Disappearance',
1250             'description': 'French thriller serial about a missing teenager.',
1251         },
1252         'playlist_mincount': 6,
1253         'skip': 'This programme is not currently available on BBC iPlayer',
1254     }, {
1255         # Available for over a year unlike 30 days for most other programmes
1256         'url': 'http://www.bbc.co.uk/iplayer/group/p02tcc32',
1257         'info_dict': {
1258             'id': 'p02tcc32',
1259             'title': 'Bohemian Icons',
1260             'description': 'md5:683e901041b2fe9ba596f2ab04c4dbe7',
1261         },
1262         'playlist_mincount': 10,
1263     }]
1264
1265     def _extract_title_and_description(self, webpage):
1266         title = self._search_regex(r'<h1>([^<]+)</h1>', webpage, 'title', fatal=False)
1267         description = self._search_regex(
1268             r'<p[^>]+class=(["\'])subtitle\1[^>]*>(?P<value>[^<]+)</p>',
1269             webpage, 'description', fatal=False, group='value')
1270         return title, description
1271
1272
1273 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1274     IE_NAME = 'bbc.co.uk:playlist'
1275     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1276     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1277     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1278     _TESTS = [{
1279         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1280         'info_dict': {
1281             'id': 'b05rcz9v',
1282             'title': 'The Disappearance - Clips - BBC Four',
1283             'description': 'French thriller serial about a missing teenager.',
1284         },
1285         'playlist_mincount': 7,
1286     }, {
1287         # multipage playlist, explicit page
1288         'url': 'http://www.bbc.co.uk/programmes/b00mfl7n/clips?page=1',
1289         'info_dict': {
1290             'id': 'b00mfl7n',
1291             'title': 'Frozen Planet - Clips - BBC One',
1292             'description': 'md5:65dcbf591ae628dafe32aa6c4a4a0d8c',
1293         },
1294         'playlist_mincount': 24,
1295     }, {
1296         # multipage playlist, all pages
1297         'url': 'http://www.bbc.co.uk/programmes/b00mfl7n/clips',
1298         'info_dict': {
1299             'id': 'b00mfl7n',
1300             'title': 'Frozen Planet - Clips - BBC One',
1301             'description': 'md5:65dcbf591ae628dafe32aa6c4a4a0d8c',
1302         },
1303         'playlist_mincount': 142,
1304     }, {
1305         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1306         'only_matching': True,
1307     }, {
1308         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1309         'only_matching': True,
1310     }, {
1311         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1312         'only_matching': True,
1313     }]
1314
1315     def _extract_title_and_description(self, webpage):
1316         title = self._og_search_title(webpage, fatal=False)
1317         description = self._og_search_description(webpage)
1318         return title, description