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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     ExtractorError,
   9     float_or_none,
  10     int_or_none,
  11     parse_duration,
  12     parse_iso8601,
  13     remove_end,
  14     unescapeHTML,
  15 )
  16 from ..compat import (
  17     compat_etree_fromstring,
  18     compat_HTTPError,
  19 )
  20
  21
  22 class BBCCoUkIE(InfoExtractor):
  23     IE_NAME = 'bbc.co.uk'
  24     IE_DESC = 'BBC iPlayer'
  25     _ID_REGEX = r'[pb][\da-z]{7}'
  26     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:(?:programmes/(?!articles/)|iplayer(?:/[^/]+)?/(?:episode/|playlist/))|music/clips[/#])(?P<id>%s)' % _ID_REGEX
  27
  28     _MEDIASELECTOR_URLS = [
  29         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  30         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  31         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  32         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  33         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  34     ]
  35
  36     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  37     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  38
  39     _NAMESPACES = (
  40         _MEDIASELECTION_NS,
  41         _EMP_PLAYLIST_NS,
  42     )
  43
  44     _TESTS = [
  45         {
  46             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  47             'info_dict': {
  48                 'id': 'b039d07m',
  49                 'ext': 'flv',
  50                 'title': 'Kaleidoscope, Leonard Cohen',
  51                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  52                 'duration': 1740,
  53             },
  54             'params': {
  55                 # rtmp download
  56                 'skip_download': True,
  57             }
  58         },
  59         {
  60             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  61             'info_dict': {
  62                 'id': 'b00yng1d',
  63                 'ext': 'flv',
  64                 'title': 'The Man in Black: Series 3: The Printed Name',
  65                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  66                 'duration': 1800,
  67             },
  68             'params': {
  69                 # rtmp download
  70                 'skip_download': True,
  71             },
  72             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  73         },
  74         {
  75             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  76             'info_dict': {
  77                 'id': 'b00yng1d',
  78                 'ext': 'flv',
  79                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  80                 '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.",
  81                 'duration': 5100,
  82             },
  83             'params': {
  84                 # rtmp download
  85                 'skip_download': True,
  86             },
  87             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
  88         },
  89         {
  90             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
  91             'info_dict': {
  92                 'id': 'b03k3pb7',
  93                 'ext': 'flv',
  94                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
  95                 'description': '2. Invasion',
  96                 'duration': 3600,
  97             },
  98             'params': {
  99                 # rtmp download
 100                 'skip_download': True,
 101             },
 102             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 103         }, {
 104             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 105             'info_dict': {
 106                 'id': 'b04v209v',
 107                 'ext': 'flv',
 108                 'title': 'Pete Tong, The Essential New Tune Special',
 109                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 110                 'duration': 10800,
 111             },
 112             'params': {
 113                 # rtmp download
 114                 'skip_download': True,
 115             }
 116         }, {
 117             'url': 'http://www.bbc.co.uk/music/clips/p02frcc3',
 118             'note': 'Audio',
 119             'info_dict': {
 120                 'id': 'p02frcch',
 121                 'ext': 'flv',
 122                 'title': 'Pete Tong, Past, Present and Future Special, Madeon - After Hours mix',
 123                 'description': 'French house superstar Madeon takes us out of the club and onto the after party.',
 124                 'duration': 3507,
 125             },
 126             'params': {
 127                 # rtmp download
 128                 'skip_download': True,
 129             }
 130         }, {
 131             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 132             'note': 'Video',
 133             'info_dict': {
 134                 'id': 'p025c103',
 135                 'ext': 'flv',
 136                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 137                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 138                 'duration': 226,
 139             },
 140             'params': {
 141                 # rtmp download
 142                 'skip_download': True,
 143             }
 144         }, {
 145             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 146             'info_dict': {
 147                 'id': 'p02n76xf',
 148                 'ext': 'flv',
 149                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 150                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 151                 'duration': 3540,
 152             },
 153             'params': {
 154                 # rtmp download
 155                 'skip_download': True,
 156             },
 157             'skip': 'geolocation',
 158         }, {
 159             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 160             'info_dict': {
 161                 'id': 'b05zmgw1',
 162                 'ext': 'flv',
 163                 '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.',
 164                 'title': 'Royal Academy Summer Exhibition',
 165                 'duration': 3540,
 166             },
 167             'params': {
 168                 # rtmp download
 169                 'skip_download': True,
 170             },
 171             'skip': 'geolocation',
 172         }, {
 173             # iptv-all mediaset fails with geolocation however there is no geo restriction
 174             # for this programme at all
 175             'url': 'http://www.bbc.co.uk/programmes/b06bp7lf',
 176             'info_dict': {
 177                 'id': 'b06bp7kf',
 178                 'ext': 'flv',
 179                 'title': "Annie Mac's Friday Night, B.Traits sits in for Annie",
 180                 'description': 'B.Traits sits in for Annie Mac with a Mini-Mix from Disclosure.',
 181                 'duration': 10800,
 182             },
 183             'params': {
 184                 # rtmp download
 185                 'skip_download': True,
 186             },
 187         }, {
 188             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 189             'only_matching': True,
 190         }, {
 191             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 192             'only_matching': True,
 193         }, {
 194             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 195             'only_matching': True,
 196         }
 197     ]
 198
 199     class MediaSelectionError(Exception):
 200         def __init__(self, id):
 201             self.id = id
 202
 203     def _extract_asx_playlist(self, connection, programme_id):
 204         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 205         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 206
 207     def _extract_connection(self, connection, programme_id):
 208         formats = []
 209         kind = connection.get('kind')
 210         protocol = connection.get('protocol')
 211         supplier = connection.get('supplier')
 212         if protocol == 'http':
 213             href = connection.get('href')
 214             transfer_format = connection.get('transferFormat')
 215             # ASX playlist
 216             if supplier == 'asx':
 217                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 218                     formats.append({
 219                         'url': ref,
 220                         'format_id': 'ref%s_%s' % (i, supplier),
 221                     })
 222             # Skip DASH until supported
 223             elif transfer_format == 'dash':
 224                 pass
 225             elif transfer_format == 'hls':
 226                 m3u8_formats = self._extract_m3u8_formats(
 227                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 228                     m3u8_id=supplier, fatal=False)
 229                 if m3u8_formats:
 230                     formats.extend(m3u8_formats)
 231             # Direct link
 232             else:
 233                 formats.append({
 234                     'url': href,
 235                     'format_id': supplier or kind or protocol,
 236                 })
 237         elif protocol == 'rtmp':
 238             application = connection.get('application', 'ondemand')
 239             auth_string = connection.get('authString')
 240             identifier = connection.get('identifier')
 241             server = connection.get('server')
 242             formats.append({
 243                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 244                 'play_path': identifier,
 245                 'app': '%s?%s' % (application, auth_string),
 246                 'page_url': 'http://www.bbc.co.uk',
 247                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 248                 'rtmp_live': False,
 249                 'ext': 'flv',
 250                 'format_id': supplier,
 251             })
 252         return formats
 253
 254     def _extract_items(self, playlist):
 255         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 256
 257     def _findall_ns(self, element, xpath):
 258         elements = []
 259         for ns in self._NAMESPACES:
 260             elements.extend(element.findall(xpath % ns))
 261         return elements
 262
 263     def _extract_medias(self, media_selection):
 264         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 265         if error is None:
 266             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 267         if error is not None:
 268             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 269         return self._findall_ns(media_selection, './{%s}media')
 270
 271     def _extract_connections(self, media):
 272         return self._findall_ns(media, './{%s}connection')
 273
 274     def _extract_video(self, media, programme_id):
 275         formats = []
 276         vbr = int_or_none(media.get('bitrate'))
 277         vcodec = media.get('encoding')
 278         service = media.get('service')
 279         width = int_or_none(media.get('width'))
 280         height = int_or_none(media.get('height'))
 281         file_size = int_or_none(media.get('media_file_size'))
 282         for connection in self._extract_connections(media):
 283             conn_formats = self._extract_connection(connection, programme_id)
 284             for format in conn_formats:
 285                 format.update({
 286                     'width': width,
 287                     'height': height,
 288                     'vbr': vbr,
 289                     'vcodec': vcodec,
 290                     'filesize': file_size,
 291                 })
 292                 if service:
 293                     format['format_id'] = '%s_%s' % (service, format['format_id'])
 294             formats.extend(conn_formats)
 295         return formats
 296
 297     def _extract_audio(self, media, programme_id):
 298         formats = []
 299         abr = int_or_none(media.get('bitrate'))
 300         acodec = media.get('encoding')
 301         service = media.get('service')
 302         for connection in self._extract_connections(media):
 303             conn_formats = self._extract_connection(connection, programme_id)
 304             for format in conn_formats:
 305                 format.update({
 306                     'format_id': '%s_%s' % (service, format['format_id']),
 307                     'abr': abr,
 308                     'acodec': acodec,
 309                 })
 310             formats.extend(conn_formats)
 311         return formats
 312
 313     def _get_subtitles(self, media, programme_id):
 314         subtitles = {}
 315         for connection in self._extract_connections(media):
 316             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 317             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 318             subtitles[lang] = [
 319                 {
 320                     'url': connection.get('href'),
 321                     'ext': 'ttml',
 322                 },
 323             ]
 324         return subtitles
 325
 326     def _raise_extractor_error(self, media_selection_error):
 327         raise ExtractorError(
 328             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 329             expected=True)
 330
 331     def _download_media_selector(self, programme_id):
 332         last_exception = None
 333         for mediaselector_url in self._MEDIASELECTOR_URLS:
 334             try:
 335                 return self._download_media_selector_url(
 336                     mediaselector_url % programme_id, programme_id)
 337             except BBCCoUkIE.MediaSelectionError as e:
 338                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 339                     last_exception = e
 340                     continue
 341                 self._raise_extractor_error(e)
 342         self._raise_extractor_error(last_exception)
 343
 344     def _download_media_selector_url(self, url, programme_id=None):
 345         try:
 346             media_selection = self._download_xml(
 347                 url, programme_id, 'Downloading media selection XML')
 348         except ExtractorError as ee:
 349             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 350                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 351             else:
 352                 raise
 353         return self._process_media_selector(media_selection, programme_id)
 354
 355     def _process_media_selector(self, media_selection, programme_id):
 356         formats = []
 357         subtitles = None
 358
 359         for media in self._extract_medias(media_selection):
 360             kind = media.get('kind')
 361             if kind == 'audio':
 362                 formats.extend(self._extract_audio(media, programme_id))
 363             elif kind == 'video':
 364                 formats.extend(self._extract_video(media, programme_id))
 365             elif kind == 'captions':
 366                 subtitles = self.extract_subtitles(media, programme_id)
 367         return formats, subtitles
 368
 369     def _download_playlist(self, playlist_id):
 370         try:
 371             playlist = self._download_json(
 372                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 373                 playlist_id, 'Downloading playlist JSON')
 374
 375             version = playlist.get('defaultAvailableVersion')
 376             if version:
 377                 smp_config = version['smpConfig']
 378                 title = smp_config['title']
 379                 description = smp_config['summary']
 380                 for item in smp_config['items']:
 381                     kind = item['kind']
 382                     if kind != 'programme' and kind != 'radioProgramme':
 383                         continue
 384                     programme_id = item.get('vpid')
 385                     duration = int_or_none(item.get('duration'))
 386                     formats, subtitles = self._download_media_selector(programme_id)
 387                 return programme_id, title, description, duration, formats, subtitles
 388         except ExtractorError as ee:
 389             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 390                 raise
 391
 392         # fallback to legacy playlist
 393         return self._process_legacy_playlist(playlist_id)
 394
 395     def _process_legacy_playlist_url(self, url, display_id):
 396         playlist = self._download_legacy_playlist_url(url, display_id)
 397         return self._extract_from_legacy_playlist(playlist, display_id)
 398
 399     def _process_legacy_playlist(self, playlist_id):
 400         return self._process_legacy_playlist_url(
 401             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 402
 403     def _download_legacy_playlist_url(self, url, playlist_id=None):
 404         return self._download_xml(
 405             url, playlist_id, 'Downloading legacy playlist XML')
 406
 407     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 408         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 409         if no_items is not None:
 410             reason = no_items.get('reason')
 411             if reason == 'preAvailability':
 412                 msg = 'Episode %s is not yet available' % playlist_id
 413             elif reason == 'postAvailability':
 414                 msg = 'Episode %s is no longer available' % playlist_id
 415             elif reason == 'noMedia':
 416                 msg = 'Episode %s is not currently available' % playlist_id
 417             else:
 418                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 419             raise ExtractorError(msg, expected=True)
 420
 421         for item in self._extract_items(playlist):
 422             kind = item.get('kind')
 423             if kind != 'programme' and kind != 'radioProgramme':
 424                 continue
 425             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 426             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 427             description = description_el.text if description_el is not None else None
 428
 429             def get_programme_id(item):
 430                 def get_from_attributes(item):
 431                     for p in('identifier', 'group'):
 432                         value = item.get(p)
 433                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 434                             return value
 435                 get_from_attributes(item)
 436                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 437                 if mediator is not None:
 438                     return get_from_attributes(mediator)
 439
 440             programme_id = get_programme_id(item)
 441             duration = int_or_none(item.get('duration'))
 442
 443             if programme_id:
 444                 formats, subtitles = self._download_media_selector(programme_id)
 445             else:
 446                 formats, subtitles = self._process_media_selector(item, playlist_id)
 447                 programme_id = playlist_id
 448
 449         return programme_id, title, description, duration, formats, subtitles
 450
 451     def _real_extract(self, url):
 452         group_id = self._match_id(url)
 453
 454         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 455
 456         programme_id = None
 457
 458         tviplayer = self._search_regex(
 459             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 460             webpage, 'player', default=None)
 461
 462         if tviplayer:
 463             player = self._parse_json(tviplayer, group_id).get('player', {})
 464             duration = int_or_none(player.get('duration'))
 465             programme_id = player.get('vpid')
 466
 467         if not programme_id:
 468             programme_id = self._search_regex(
 469                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 470
 471         if programme_id:
 472             formats, subtitles = self._download_media_selector(programme_id)
 473             title = self._og_search_title(webpage)
 474             description = self._search_regex(
 475                 r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 476                 webpage, 'description', fatal=False)
 477         else:
 478             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 479
 480         self._sort_formats(formats)
 481
 482         return {
 483             'id': programme_id,
 484             'title': title,
 485             'description': description,
 486             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 487             'duration': duration,
 488             'formats': formats,
 489             'subtitles': subtitles,
 490         }
 491
 492
 493 class BBCIE(BBCCoUkIE):
 494     IE_NAME = 'bbc'
 495     IE_DESC = 'BBC'
 496     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 497
 498     _MEDIASELECTOR_URLS = [
 499         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 500         # even not geo restricted at all
 501         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 502         # Provides more formats, namely direct mp4 links, but fails on some videos with
 503         # notukerror for non UK (?) users (e.g.
 504         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 505         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 506         # Provides fewer formats, but works everywhere for everybody (hopefully)
 507         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 508     ]
 509
 510     _TESTS = [{
 511         # article with multiple videos embedded with data-playable containing vpids
 512         'url': 'http://www.bbc.com/news/world-europe-32668511',
 513         'info_dict': {
 514             'id': 'world-europe-32668511',
 515             'title': 'Russia stages massive WW2 parade despite Western boycott',
 516             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 517         },
 518         'playlist_count': 2,
 519     }, {
 520         # article with multiple videos embedded with data-playable (more videos)
 521         'url': 'http://www.bbc.com/news/business-28299555',
 522         'info_dict': {
 523             'id': 'business-28299555',
 524             'title': 'Farnborough Airshow: Video highlights',
 525             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 526         },
 527         'playlist_count': 9,
 528         'skip': 'Save time',
 529     }, {
 530         # article with multiple videos embedded with `new SMP()`
 531         # broken
 532         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 533         'info_dict': {
 534             'id': '3662a707-0af9-3149-963f-47bea720b460',
 535             'title': 'BBC Blogs - Adam Curtis - BUGGER',
 536         },
 537         'playlist_count': 18,
 538     }, {
 539         # single video embedded with data-playable containing vpid
 540         'url': 'http://www.bbc.com/news/world-europe-32041533',
 541         'info_dict': {
 542             'id': 'p02mprgb',
 543             'ext': 'mp4',
 544             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 545             'description': 'md5:2868290467291b37feda7863f7a83f54',
 546             'duration': 47,
 547             'timestamp': 1427219242,
 548             'upload_date': '20150324',
 549         },
 550         'params': {
 551             # rtmp download
 552             'skip_download': True,
 553         }
 554     }, {
 555         # article with single video embedded with data-playable containing XML playlist
 556         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 557         # and playlist with f4m and m3u8 as streamingUrl
 558         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 559         'info_dict': {
 560             'id': '150615_telabyad_kentin_cogu',
 561             'ext': 'mp4',
 562             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 563             'timestamp': 1434397334,
 564             'upload_date': '20150615',
 565         },
 566         'params': {
 567             'skip_download': True,
 568         }
 569     }, {
 570         # single video embedded with data-playable containing XML playlists (regional section)
 571         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 572         'info_dict': {
 573             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 574             'ext': 'mp4',
 575             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 576             'timestamp': 1434713142,
 577             'upload_date': '20150619',
 578         },
 579         'params': {
 580             'skip_download': True,
 581         }
 582     }, {
 583         # single video from video playlist embedded with vxp-playlist-data JSON
 584         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 585         'info_dict': {
 586             'id': 'p02w6qjc',
 587             'ext': 'mp4',
 588             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 589             'duration': 56,
 590         },
 591         'params': {
 592             'skip_download': True,
 593         }
 594     }, {
 595         # single video story with digitalData
 596         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 597         'info_dict': {
 598             'id': 'p02q6gc4',
 599             'ext': 'flv',
 600             'title': 'Sri Lanka’s spicy secret',
 601             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 602             'timestamp': 1437674293,
 603             'upload_date': '20150723',
 604         },
 605         'params': {
 606             # rtmp download
 607             'skip_download': True,
 608         }
 609     }, {
 610         # single video story without digitalData
 611         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 612         'info_dict': {
 613             'id': 'p018zqqg',
 614             'ext': 'mp4',
 615             'title': 'Hyundai Santa Fe Sport: Rock star',
 616             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 617             'timestamp': 1415867444,
 618             'upload_date': '20141113',
 619         },
 620         'params': {
 621             # rtmp download
 622             'skip_download': True,
 623         }
 624     }, {
 625         # single video with playlist.sxml URL in playlist param
 626         'url': 'http://www.bbc.com/sport/0/football/33653409',
 627         'info_dict': {
 628             'id': 'p02xycnp',
 629             'ext': 'mp4',
 630             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 631             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 632             'duration': 140,
 633         },
 634         'params': {
 635             # rtmp download
 636             'skip_download': True,
 637         }
 638     }, {
 639         # article with multiple videos embedded with playlist.sxml in playlist param
 640         'url': 'http://www.bbc.com/sport/0/football/34475836',
 641         'info_dict': {
 642             'id': '34475836',
 643             'title': 'What Liverpool can expect from Klopp',
 644         },
 645         'playlist_count': 3,
 646     }, {
 647         # single video with playlist URL from weather section
 648         'url': 'http://www.bbc.com/weather/features/33601775',
 649         'only_matching': True,
 650     }, {
 651         # custom redirection to www.bbc.com
 652         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 653         'only_matching': True,
 654     }]
 655
 656     @classmethod
 657     def suitable(cls, url):
 658         return False if BBCCoUkIE.suitable(url) or BBCCoUkArticleIE.suitable(url) else super(BBCIE, cls).suitable(url)
 659
 660     def _extract_from_media_meta(self, media_meta, video_id):
 661         # Direct links to media in media metadata (e.g.
 662         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 663         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 664         source_files = media_meta.get('sourceFiles')
 665         if source_files:
 666             return [{
 667                 'url': f['url'],
 668                 'format_id': format_id,
 669                 'ext': f.get('encoding'),
 670                 'tbr': float_or_none(f.get('bitrate'), 1000),
 671                 'filesize': int_or_none(f.get('filesize')),
 672             } for format_id, f in source_files.items() if f.get('url')], []
 673
 674         programme_id = media_meta.get('externalId')
 675         if programme_id:
 676             return self._download_media_selector(programme_id)
 677
 678         # Process playlist.sxml as legacy playlist
 679         href = media_meta.get('href')
 680         if href:
 681             playlist = self._download_legacy_playlist_url(href)
 682             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 683             return formats, subtitles
 684
 685         return [], []
 686
 687     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 688         programme_id, title, description, duration, formats, subtitles = \
 689             self._process_legacy_playlist_url(url, playlist_id)
 690         self._sort_formats(formats)
 691         return {
 692             'id': programme_id,
 693             'title': title,
 694             'description': description,
 695             'duration': duration,
 696             'timestamp': timestamp,
 697             'formats': formats,
 698             'subtitles': subtitles,
 699         }
 700
 701     def _real_extract(self, url):
 702         playlist_id = self._match_id(url)
 703
 704         webpage = self._download_webpage(url, playlist_id)
 705
 706         timestamp = None
 707         playlist_title = None
 708         playlist_description = None
 709
 710         ld = self._parse_json(
 711             self._search_regex(
 712                 r'(?s)<script type="application/ld\+json">(.+?)</script>',
 713                 webpage, 'ld json', default='{}'),
 714             playlist_id, fatal=False)
 715         if ld:
 716             timestamp = parse_iso8601(ld.get('datePublished'))
 717             playlist_title = ld.get('headline')
 718             playlist_description = ld.get('articleBody')
 719
 720         if not timestamp:
 721             timestamp = parse_iso8601(self._search_regex(
 722                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 723                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 724                  r'"datePublished":\s*"([^"]+)'],
 725                 webpage, 'date', default=None))
 726
 727         entries = []
 728
 729         # article with multiple videos embedded with playlist.sxml (e.g.
 730         # http://www.bbc.com/sport/0/football/34475836)
 731         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 732         if playlists:
 733             entries = [
 734                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 735                 for playlist_url in playlists]
 736
 737         # news article with multiple videos embedded with data-playable
 738         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 739         if data_playables:
 740             for _, data_playable_json in data_playables:
 741                 data_playable = self._parse_json(
 742                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 743                 if not data_playable:
 744                     continue
 745                 settings = data_playable.get('settings', {})
 746                 if settings:
 747                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 748                     # http://www.bbc.com/news/world-us-canada-34473351)
 749                     playlist_object = settings.get('playlistObject', {})
 750                     if playlist_object:
 751                         items = playlist_object.get('items')
 752                         if items and isinstance(items, list):
 753                             title = playlist_object['title']
 754                             description = playlist_object.get('summary')
 755                             duration = int_or_none(items[0].get('duration'))
 756                             programme_id = items[0].get('vpid')
 757                             formats, subtitles = self._download_media_selector(programme_id)
 758                             self._sort_formats(formats)
 759                             entries.append({
 760                                 'id': programme_id,
 761                                 'title': title,
 762                                 'description': description,
 763                                 'timestamp': timestamp,
 764                                 'duration': duration,
 765                                 'formats': formats,
 766                                 'subtitles': subtitles,
 767                             })
 768                     else:
 769                         # data-playable without vpid but with a playlist.sxml URLs
 770                         # in otherSettings.playlist (e.g.
 771                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 772                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 773                         if playlist:
 774                             entries.append(self._extract_from_playlist_sxml(
 775                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 776
 777         if entries:
 778             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
 779             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
 780             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 781
 782         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 783         programme_id = self._search_regex(
 784             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
 785              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 786              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 787             webpage, 'vpid', default=None)
 788
 789         if programme_id:
 790             formats, subtitles = self._download_media_selector(programme_id)
 791             self._sort_formats(formats)
 792             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 793             digital_data = self._parse_json(
 794                 self._search_regex(
 795                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 796                 programme_id, fatal=False)
 797             page_info = digital_data.get('page', {}).get('pageInfo', {})
 798             title = page_info.get('pageName') or self._og_search_title(webpage)
 799             description = page_info.get('description') or self._og_search_description(webpage)
 800             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 801             return {
 802                 'id': programme_id,
 803                 'title': title,
 804                 'description': description,
 805                 'timestamp': timestamp,
 806                 'formats': formats,
 807                 'subtitles': subtitles,
 808             }
 809
 810         playlist_title = self._html_search_regex(
 811             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
 812         playlist_description = self._og_search_description(webpage, default=None)
 813
 814         def extract_all(pattern):
 815             return list(filter(None, map(
 816                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 817                 re.findall(pattern, webpage))))
 818
 819         # Multiple video article (e.g.
 820         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 821         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 822         entries = []
 823         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 824             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 825             if embed_url and re.match(EMBED_URL, embed_url):
 826                 entries.append(embed_url)
 827         entries.extend(re.findall(
 828             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 829         if entries:
 830             return self.playlist_result(
 831                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 832                 playlist_id, playlist_title, playlist_description)
 833
 834         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 835         medias = extract_all(r"data-media-meta='({[^']+})'")
 836
 837         if not medias:
 838             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 839             media_asset = self._search_regex(
 840                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 841                 webpage, 'media asset', default=None)
 842             if media_asset:
 843                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 844                 medias = []
 845                 for video in media_asset_page.get('videos', {}).values():
 846                     medias.extend(video.values())
 847
 848         if not medias:
 849             # Multiple video playlist with single `now playing` entry (e.g.
 850             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 851             vxp_playlist = self._parse_json(
 852                 self._search_regex(
 853                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 854                     webpage, 'playlist data'),
 855                 playlist_id)
 856             playlist_medias = []
 857             for item in vxp_playlist:
 858                 media = item.get('media')
 859                 if not media:
 860                     continue
 861                 playlist_medias.append(media)
 862                 # Download single video if found media with asset id matching the video id from URL
 863                 if item.get('advert', {}).get('assetId') == playlist_id:
 864                     medias = [media]
 865                     break
 866             # Fallback to the whole playlist
 867             if not medias:
 868                 medias = playlist_medias
 869
 870         entries = []
 871         for num, media_meta in enumerate(medias, start=1):
 872             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 873             if not formats:
 874                 continue
 875             self._sort_formats(formats)
 876
 877             video_id = media_meta.get('externalId')
 878             if not video_id:
 879                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 880
 881             title = media_meta.get('caption')
 882             if not title:
 883                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 884
 885             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 886
 887             images = []
 888             for image in media_meta.get('images', {}).values():
 889                 images.extend(image.values())
 890             if 'image' in media_meta:
 891                 images.append(media_meta['image'])
 892
 893             thumbnails = [{
 894                 'url': image.get('href'),
 895                 'width': int_or_none(image.get('width')),
 896                 'height': int_or_none(image.get('height')),
 897             } for image in images]
 898
 899             entries.append({
 900                 'id': video_id,
 901                 'title': title,
 902                 'thumbnails': thumbnails,
 903                 'duration': duration,
 904                 'timestamp': timestamp,
 905                 'formats': formats,
 906                 'subtitles': subtitles,
 907             })
 908
 909         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 910
 911
 912 class BBCCoUkArticleIE(InfoExtractor):
 913     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 914     IE_NAME = 'bbc.co.uk:article'
 915     IE_DESC = 'BBC articles'
 916
 917     _TEST = {
 918         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 919         'info_dict': {
 920             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 921             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 922             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 923         },
 924         'playlist_count': 4,
 925         'add_ie': ['BBCCoUk'],
 926     }
 927
 928     def _real_extract(self, url):
 929         playlist_id = self._match_id(url)
 930
 931         webpage = self._download_webpage(url, playlist_id)
 932
 933         title = self._og_search_title(webpage)
 934         description = self._og_search_description(webpage).strip()
 935
 936         entries = [self.url_result(programme_url) for programme_url in re.findall(
 937             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 938
 939         return self.playlist_result(entries, playlist_id, title, description)