_ 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'''(?x)
  27                     https?://
  28                         (?:www\.)?bbc\.co\.uk/
  29                         (?:
  30                             programmes/(?!articles/)|
  31                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
  32                             music/clips[/#]|
  33                             radio/player/
  34                         )
  35                         (?P<id>%s)
  36                     ''' % _ID_REGEX
  37
  38     _MEDIASELECTOR_URLS = [
  39         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  40         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  41         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  42         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  43         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  44     ]
  45
  46     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  47     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  48
  49     _NAMESPACES = (
  50         _MEDIASELECTION_NS,
  51         _EMP_PLAYLIST_NS,
  52     )
  53
  54     _TESTS = [
  55         {
  56             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  57             'info_dict': {
  58                 'id': 'b039d07m',
  59                 'ext': 'flv',
  60                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  61                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  62             },
  63             'params': {
  64                 # rtmp download
  65                 'skip_download': True,
  66             }
  67         },
  68         {
  69             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  70             'info_dict': {
  71                 'id': 'b00yng1d',
  72                 'ext': 'flv',
  73                 'title': 'The Man in Black: Series 3: The Printed Name',
  74                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  75                 'duration': 1800,
  76             },
  77             'params': {
  78                 # rtmp download
  79                 'skip_download': True,
  80             },
  81             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  82         },
  83         {
  84             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  85             'info_dict': {
  86                 'id': 'b00yng1d',
  87                 'ext': 'flv',
  88                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  89                 '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.",
  90                 'duration': 5100,
  91             },
  92             'params': {
  93                 # rtmp download
  94                 'skip_download': True,
  95             },
  96             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
  97         },
  98         {
  99             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
 100             'info_dict': {
 101                 'id': 'b03k3pb7',
 102                 'ext': 'flv',
 103                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 104                 'description': '2. Invasion',
 105                 'duration': 3600,
 106             },
 107             'params': {
 108                 # rtmp download
 109                 'skip_download': True,
 110             },
 111             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 112         }, {
 113             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 114             'info_dict': {
 115                 'id': 'b04v209v',
 116                 'ext': 'flv',
 117                 'title': 'Pete Tong, The Essential New Tune Special',
 118                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 119                 'duration': 10800,
 120             },
 121             'params': {
 122                 # rtmp download
 123                 'skip_download': True,
 124             },
 125             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 126         }, {
 127             'url': 'http://www.bbc.co.uk/music/clips/p02frcc3',
 128             'note': 'Audio',
 129             'info_dict': {
 130                 'id': 'p02frcch',
 131                 'ext': 'flv',
 132                 'title': 'Pete Tong, Past, Present and Future Special, Madeon - After Hours mix',
 133                 'description': 'French house superstar Madeon takes us out of the club and onto the after party.',
 134                 'duration': 3507,
 135             },
 136             'params': {
 137                 # rtmp download
 138                 'skip_download': True,
 139             }
 140         }, {
 141             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 142             'note': 'Video',
 143             'info_dict': {
 144                 'id': 'p025c103',
 145                 'ext': 'flv',
 146                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 147                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 148                 'duration': 226,
 149             },
 150             'params': {
 151                 # rtmp download
 152                 'skip_download': True,
 153             }
 154         }, {
 155             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 156             'info_dict': {
 157                 'id': 'p02n76xf',
 158                 'ext': 'flv',
 159                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 160                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 161                 'duration': 3540,
 162             },
 163             'params': {
 164                 # rtmp download
 165                 'skip_download': True,
 166             },
 167             'skip': 'geolocation',
 168         }, {
 169             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 170             'info_dict': {
 171                 'id': 'b05zmgw1',
 172                 'ext': 'flv',
 173                 '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.',
 174                 'title': 'Royal Academy Summer Exhibition',
 175                 'duration': 3540,
 176             },
 177             'params': {
 178                 # rtmp download
 179                 'skip_download': True,
 180             },
 181             'skip': 'geolocation',
 182         }, {
 183             # iptv-all mediaset fails with geolocation however there is no geo restriction
 184             # for this programme at all
 185             'url': 'http://www.bbc.co.uk/programmes/b06bp7lf',
 186             'info_dict': {
 187                 'id': 'b06bp7kf',
 188                 'ext': 'flv',
 189                 'title': "Annie Mac's Friday Night, B.Traits sits in for Annie",
 190                 'description': 'B.Traits sits in for Annie Mac with a Mini-Mix from Disclosure.',
 191                 'duration': 10800,
 192             },
 193             'params': {
 194                 # rtmp download
 195                 'skip_download': True,
 196             },
 197         }, {
 198             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 199             'only_matching': True,
 200         }, {
 201             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 202             'only_matching': True,
 203         }, {
 204             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 205             'only_matching': True,
 206         }, {
 207             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 208             'only_matching': True,
 209         }
 210     ]
 211
 212     class MediaSelectionError(Exception):
 213         def __init__(self, id):
 214             self.id = id
 215
 216     def _extract_asx_playlist(self, connection, programme_id):
 217         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 218         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 219
 220     def _extract_connection(self, connection, programme_id):
 221         formats = []
 222         kind = connection.get('kind')
 223         protocol = connection.get('protocol')
 224         supplier = connection.get('supplier')
 225         if protocol == 'http':
 226             href = connection.get('href')
 227             transfer_format = connection.get('transferFormat')
 228             # ASX playlist
 229             if supplier == 'asx':
 230                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 231                     formats.append({
 232                         'url': ref,
 233                         'format_id': 'ref%s_%s' % (i, supplier),
 234                     })
 235             # Skip DASH until supported
 236             elif transfer_format == 'dash':
 237                 pass
 238             elif transfer_format == 'hls':
 239                 formats.extend(self._extract_m3u8_formats(
 240                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 241                     m3u8_id=supplier, fatal=False))
 242             # Direct link
 243             else:
 244                 formats.append({
 245                     'url': href,
 246                     'format_id': supplier or kind or protocol,
 247                 })
 248         elif protocol == 'rtmp':
 249             application = connection.get('application', 'ondemand')
 250             auth_string = connection.get('authString')
 251             identifier = connection.get('identifier')
 252             server = connection.get('server')
 253             formats.append({
 254                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 255                 'play_path': identifier,
 256                 'app': '%s?%s' % (application, auth_string),
 257                 'page_url': 'http://www.bbc.co.uk',
 258                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 259                 'rtmp_live': False,
 260                 'ext': 'flv',
 261                 'format_id': supplier,
 262             })
 263         return formats
 264
 265     def _extract_items(self, playlist):
 266         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 267
 268     def _findall_ns(self, element, xpath):
 269         elements = []
 270         for ns in self._NAMESPACES:
 271             elements.extend(element.findall(xpath % ns))
 272         return elements
 273
 274     def _extract_medias(self, media_selection):
 275         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 276         if error is None:
 277             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 278         if error is not None:
 279             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 280         return self._findall_ns(media_selection, './{%s}media')
 281
 282     def _extract_connections(self, media):
 283         return self._findall_ns(media, './{%s}connection')
 284
 285     def _extract_video(self, media, programme_id):
 286         formats = []
 287         vbr = int_or_none(media.get('bitrate'))
 288         vcodec = media.get('encoding')
 289         service = media.get('service')
 290         width = int_or_none(media.get('width'))
 291         height = int_or_none(media.get('height'))
 292         file_size = int_or_none(media.get('media_file_size'))
 293         for connection in self._extract_connections(media):
 294             conn_formats = self._extract_connection(connection, programme_id)
 295             for format in conn_formats:
 296                 format.update({
 297                     'width': width,
 298                     'height': height,
 299                     'vbr': vbr,
 300                     'vcodec': vcodec,
 301                     'filesize': file_size,
 302                 })
 303                 if service:
 304                     format['format_id'] = '%s_%s' % (service, format['format_id'])
 305             formats.extend(conn_formats)
 306         return formats
 307
 308     def _extract_audio(self, media, programme_id):
 309         formats = []
 310         abr = int_or_none(media.get('bitrate'))
 311         acodec = media.get('encoding')
 312         service = media.get('service')
 313         for connection in self._extract_connections(media):
 314             conn_formats = self._extract_connection(connection, programme_id)
 315             for format in conn_formats:
 316                 format.update({
 317                     'format_id': '%s_%s' % (service, format['format_id']),
 318                     'abr': abr,
 319                     'acodec': acodec,
 320                 })
 321             formats.extend(conn_formats)
 322         return formats
 323
 324     def _get_subtitles(self, media, programme_id):
 325         subtitles = {}
 326         for connection in self._extract_connections(media):
 327             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 328             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 329             subtitles[lang] = [
 330                 {
 331                     'url': connection.get('href'),
 332                     'ext': 'ttml',
 333                 },
 334             ]
 335         return subtitles
 336
 337     def _raise_extractor_error(self, media_selection_error):
 338         raise ExtractorError(
 339             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 340             expected=True)
 341
 342     def _download_media_selector(self, programme_id):
 343         last_exception = None
 344         for mediaselector_url in self._MEDIASELECTOR_URLS:
 345             try:
 346                 return self._download_media_selector_url(
 347                     mediaselector_url % programme_id, programme_id)
 348             except BBCCoUkIE.MediaSelectionError as e:
 349                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 350                     last_exception = e
 351                     continue
 352                 self._raise_extractor_error(e)
 353         self._raise_extractor_error(last_exception)
 354
 355     def _download_media_selector_url(self, url, programme_id=None):
 356         try:
 357             media_selection = self._download_xml(
 358                 url, programme_id, 'Downloading media selection XML')
 359         except ExtractorError as ee:
 360             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 361                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 362             else:
 363                 raise
 364         return self._process_media_selector(media_selection, programme_id)
 365
 366     def _process_media_selector(self, media_selection, programme_id):
 367         formats = []
 368         subtitles = None
 369
 370         for media in self._extract_medias(media_selection):
 371             kind = media.get('kind')
 372             if kind == 'audio':
 373                 formats.extend(self._extract_audio(media, programme_id))
 374             elif kind == 'video':
 375                 formats.extend(self._extract_video(media, programme_id))
 376             elif kind == 'captions':
 377                 subtitles = self.extract_subtitles(media, programme_id)
 378         return formats, subtitles
 379
 380     def _download_playlist(self, playlist_id):
 381         try:
 382             playlist = self._download_json(
 383                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 384                 playlist_id, 'Downloading playlist JSON')
 385
 386             version = playlist.get('defaultAvailableVersion')
 387             if version:
 388                 smp_config = version['smpConfig']
 389                 title = smp_config['title']
 390                 description = smp_config['summary']
 391                 for item in smp_config['items']:
 392                     kind = item['kind']
 393                     if kind != 'programme' and kind != 'radioProgramme':
 394                         continue
 395                     programme_id = item.get('vpid')
 396                     duration = int_or_none(item.get('duration'))
 397                     formats, subtitles = self._download_media_selector(programme_id)
 398                 return programme_id, title, description, duration, formats, subtitles
 399         except ExtractorError as ee:
 400             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 401                 raise
 402
 403         # fallback to legacy playlist
 404         return self._process_legacy_playlist(playlist_id)
 405
 406     def _process_legacy_playlist_url(self, url, display_id):
 407         playlist = self._download_legacy_playlist_url(url, display_id)
 408         return self._extract_from_legacy_playlist(playlist, display_id)
 409
 410     def _process_legacy_playlist(self, playlist_id):
 411         return self._process_legacy_playlist_url(
 412             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 413
 414     def _download_legacy_playlist_url(self, url, playlist_id=None):
 415         return self._download_xml(
 416             url, playlist_id, 'Downloading legacy playlist XML')
 417
 418     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 419         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 420         if no_items is not None:
 421             reason = no_items.get('reason')
 422             if reason == 'preAvailability':
 423                 msg = 'Episode %s is not yet available' % playlist_id
 424             elif reason == 'postAvailability':
 425                 msg = 'Episode %s is no longer available' % playlist_id
 426             elif reason == 'noMedia':
 427                 msg = 'Episode %s is not currently available' % playlist_id
 428             else:
 429                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 430             raise ExtractorError(msg, expected=True)
 431
 432         for item in self._extract_items(playlist):
 433             kind = item.get('kind')
 434             if kind != 'programme' and kind != 'radioProgramme':
 435                 continue
 436             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 437             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 438             description = description_el.text if description_el is not None else None
 439
 440             def get_programme_id(item):
 441                 def get_from_attributes(item):
 442                     for p in('identifier', 'group'):
 443                         value = item.get(p)
 444                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 445                             return value
 446                 get_from_attributes(item)
 447                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 448                 if mediator is not None:
 449                     return get_from_attributes(mediator)
 450
 451             programme_id = get_programme_id(item)
 452             duration = int_or_none(item.get('duration'))
 453
 454             if programme_id:
 455                 formats, subtitles = self._download_media_selector(programme_id)
 456             else:
 457                 formats, subtitles = self._process_media_selector(item, playlist_id)
 458                 programme_id = playlist_id
 459
 460         return programme_id, title, description, duration, formats, subtitles
 461
 462     def _real_extract(self, url):
 463         group_id = self._match_id(url)
 464
 465         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 466
 467         programme_id = None
 468         duration = None
 469
 470         tviplayer = self._search_regex(
 471             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 472             webpage, 'player', default=None)
 473
 474         if tviplayer:
 475             player = self._parse_json(tviplayer, group_id).get('player', {})
 476             duration = int_or_none(player.get('duration'))
 477             programme_id = player.get('vpid')
 478
 479         if not programme_id:
 480             programme_id = self._search_regex(
 481                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 482
 483         if programme_id:
 484             formats, subtitles = self._download_media_selector(programme_id)
 485             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 486                 r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>', webpage, 'title')
 487             description = self._search_regex(
 488                 r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 489                 webpage, 'description', default=None)
 490             if not description:
 491                 description = self._html_search_meta('description', webpage)
 492         else:
 493             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 494
 495         self._sort_formats(formats)
 496
 497         return {
 498             'id': programme_id,
 499             'title': title,
 500             'description': description,
 501             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 502             'duration': duration,
 503             'formats': formats,
 504             'subtitles': subtitles,
 505         }
 506
 507
 508 class BBCIE(BBCCoUkIE):
 509     IE_NAME = 'bbc'
 510     IE_DESC = 'BBC'
 511     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 512
 513     _MEDIASELECTOR_URLS = [
 514         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 515         # even not geo restricted at all
 516         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 517         # Provides more formats, namely direct mp4 links, but fails on some videos with
 518         # notukerror for non UK (?) users (e.g.
 519         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 520         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 521         # Provides fewer formats, but works everywhere for everybody (hopefully)
 522         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 523     ]
 524
 525     _TESTS = [{
 526         # article with multiple videos embedded with data-playable containing vpids
 527         'url': 'http://www.bbc.com/news/world-europe-32668511',
 528         'info_dict': {
 529             'id': 'world-europe-32668511',
 530             'title': 'Russia stages massive WW2 parade despite Western boycott',
 531             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 532         },
 533         'playlist_count': 2,
 534     }, {
 535         # article with multiple videos embedded with data-playable (more videos)
 536         'url': 'http://www.bbc.com/news/business-28299555',
 537         'info_dict': {
 538             'id': 'business-28299555',
 539             'title': 'Farnborough Airshow: Video highlights',
 540             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 541         },
 542         'playlist_count': 9,
 543         'skip': 'Save time',
 544     }, {
 545         # article with multiple videos embedded with `new SMP()`
 546         # broken
 547         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 548         'info_dict': {
 549             'id': '3662a707-0af9-3149-963f-47bea720b460',
 550             'title': 'BBC Blogs - Adam Curtis - BUGGER',
 551         },
 552         'playlist_count': 18,
 553     }, {
 554         # single video embedded with data-playable containing vpid
 555         'url': 'http://www.bbc.com/news/world-europe-32041533',
 556         'info_dict': {
 557             'id': 'p02mprgb',
 558             'ext': 'mp4',
 559             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 560             'description': 'md5:2868290467291b37feda7863f7a83f54',
 561             'duration': 47,
 562             'timestamp': 1427219242,
 563             'upload_date': '20150324',
 564         },
 565         'params': {
 566             # rtmp download
 567             'skip_download': True,
 568         }
 569     }, {
 570         # article with single video embedded with data-playable containing XML playlist
 571         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 572         # and playlist with f4m and m3u8 as streamingUrl
 573         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 574         'info_dict': {
 575             'id': '150615_telabyad_kentin_cogu',
 576             'ext': 'mp4',
 577             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 578             'timestamp': 1434397334,
 579             'upload_date': '20150615',
 580         },
 581         'params': {
 582             'skip_download': True,
 583         }
 584     }, {
 585         # single video embedded with data-playable containing XML playlists (regional section)
 586         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 587         'info_dict': {
 588             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 589             'ext': 'mp4',
 590             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 591             'timestamp': 1434713142,
 592             'upload_date': '20150619',
 593         },
 594         'params': {
 595             'skip_download': True,
 596         }
 597     }, {
 598         # single video from video playlist embedded with vxp-playlist-data JSON
 599         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 600         'info_dict': {
 601             'id': 'p02w6qjc',
 602             'ext': 'mp4',
 603             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 604             'duration': 56,
 605             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 606         },
 607         'params': {
 608             'skip_download': True,
 609         }
 610     }, {
 611         # single video story with digitalData
 612         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 613         'info_dict': {
 614             'id': 'p02q6gc4',
 615             'ext': 'flv',
 616             'title': 'Sri Lanka’s spicy secret',
 617             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 618             'timestamp': 1437674293,
 619             'upload_date': '20150723',
 620         },
 621         'params': {
 622             # rtmp download
 623             'skip_download': True,
 624         }
 625     }, {
 626         # single video story without digitalData
 627         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 628         'info_dict': {
 629             'id': 'p018zqqg',
 630             'ext': 'mp4',
 631             'title': 'Hyundai Santa Fe Sport: Rock star',
 632             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 633             'timestamp': 1415867444,
 634             'upload_date': '20141113',
 635         },
 636         'params': {
 637             # rtmp download
 638             'skip_download': True,
 639         }
 640     }, {
 641         # single video with playlist.sxml URL in playlist param
 642         'url': 'http://www.bbc.com/sport/0/football/33653409',
 643         'info_dict': {
 644             'id': 'p02xycnp',
 645             'ext': 'mp4',
 646             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 647             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 648             'duration': 140,
 649         },
 650         'params': {
 651             # rtmp download
 652             'skip_download': True,
 653         }
 654     }, {
 655         # article with multiple videos embedded with playlist.sxml in playlist param
 656         'url': 'http://www.bbc.com/sport/0/football/34475836',
 657         'info_dict': {
 658             'id': '34475836',
 659             'title': 'What Liverpool can expect from Klopp',
 660         },
 661         'playlist_count': 3,
 662     }, {
 663         # single video with playlist URL from weather section
 664         'url': 'http://www.bbc.com/weather/features/33601775',
 665         'only_matching': True,
 666     }, {
 667         # custom redirection to www.bbc.com
 668         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 669         'only_matching': True,
 670     }]
 671
 672     @classmethod
 673     def suitable(cls, url):
 674         return False if BBCCoUkIE.suitable(url) or BBCCoUkArticleIE.suitable(url) else super(BBCIE, cls).suitable(url)
 675
 676     def _extract_from_media_meta(self, media_meta, video_id):
 677         # Direct links to media in media metadata (e.g.
 678         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 679         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 680         source_files = media_meta.get('sourceFiles')
 681         if source_files:
 682             return [{
 683                 'url': f['url'],
 684                 'format_id': format_id,
 685                 'ext': f.get('encoding'),
 686                 'tbr': float_or_none(f.get('bitrate'), 1000),
 687                 'filesize': int_or_none(f.get('filesize')),
 688             } for format_id, f in source_files.items() if f.get('url')], []
 689
 690         programme_id = media_meta.get('externalId')
 691         if programme_id:
 692             return self._download_media_selector(programme_id)
 693
 694         # Process playlist.sxml as legacy playlist
 695         href = media_meta.get('href')
 696         if href:
 697             playlist = self._download_legacy_playlist_url(href)
 698             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 699             return formats, subtitles
 700
 701         return [], []
 702
 703     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 704         programme_id, title, description, duration, formats, subtitles = \
 705             self._process_legacy_playlist_url(url, playlist_id)
 706         self._sort_formats(formats)
 707         return {
 708             'id': programme_id,
 709             'title': title,
 710             'description': description,
 711             'duration': duration,
 712             'timestamp': timestamp,
 713             'formats': formats,
 714             'subtitles': subtitles,
 715         }
 716
 717     def _real_extract(self, url):
 718         playlist_id = self._match_id(url)
 719
 720         webpage = self._download_webpage(url, playlist_id)
 721
 722         timestamp = None
 723         playlist_title = None
 724         playlist_description = None
 725
 726         ld = self._parse_json(
 727             self._search_regex(
 728                 r'(?s)<script type="application/ld\+json">(.+?)</script>',
 729                 webpage, 'ld json', default='{}'),
 730             playlist_id, fatal=False)
 731         if ld:
 732             timestamp = parse_iso8601(ld.get('datePublished'))
 733             playlist_title = ld.get('headline')
 734             playlist_description = ld.get('articleBody')
 735
 736         if not timestamp:
 737             timestamp = parse_iso8601(self._search_regex(
 738                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 739                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 740                  r'"datePublished":\s*"([^"]+)'],
 741                 webpage, 'date', default=None))
 742
 743         entries = []
 744
 745         # article with multiple videos embedded with playlist.sxml (e.g.
 746         # http://www.bbc.com/sport/0/football/34475836)
 747         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 748         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 749         if playlists:
 750             entries = [
 751                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 752                 for playlist_url in playlists]
 753
 754         # news article with multiple videos embedded with data-playable
 755         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 756         if data_playables:
 757             for _, data_playable_json in data_playables:
 758                 data_playable = self._parse_json(
 759                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 760                 if not data_playable:
 761                     continue
 762                 settings = data_playable.get('settings', {})
 763                 if settings:
 764                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 765                     # http://www.bbc.com/news/world-us-canada-34473351)
 766                     playlist_object = settings.get('playlistObject', {})
 767                     if playlist_object:
 768                         items = playlist_object.get('items')
 769                         if items and isinstance(items, list):
 770                             title = playlist_object['title']
 771                             description = playlist_object.get('summary')
 772                             duration = int_or_none(items[0].get('duration'))
 773                             programme_id = items[0].get('vpid')
 774                             formats, subtitles = self._download_media_selector(programme_id)
 775                             self._sort_formats(formats)
 776                             entries.append({
 777                                 'id': programme_id,
 778                                 'title': title,
 779                                 'description': description,
 780                                 'timestamp': timestamp,
 781                                 'duration': duration,
 782                                 'formats': formats,
 783                                 'subtitles': subtitles,
 784                             })
 785                     else:
 786                         # data-playable without vpid but with a playlist.sxml URLs
 787                         # in otherSettings.playlist (e.g.
 788                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 789                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 790                         if playlist:
 791                             entries.append(self._extract_from_playlist_sxml(
 792                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 793
 794         if entries:
 795             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
 796             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
 797             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 798
 799         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 800         programme_id = self._search_regex(
 801             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
 802              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 803              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 804             webpage, 'vpid', default=None)
 805
 806         if programme_id:
 807             formats, subtitles = self._download_media_selector(programme_id)
 808             self._sort_formats(formats)
 809             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 810             digital_data = self._parse_json(
 811                 self._search_regex(
 812                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 813                 programme_id, fatal=False)
 814             page_info = digital_data.get('page', {}).get('pageInfo', {})
 815             title = page_info.get('pageName') or self._og_search_title(webpage)
 816             description = page_info.get('description') or self._og_search_description(webpage)
 817             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 818             return {
 819                 'id': programme_id,
 820                 'title': title,
 821                 'description': description,
 822                 'timestamp': timestamp,
 823                 'formats': formats,
 824                 'subtitles': subtitles,
 825             }
 826
 827         playlist_title = self._html_search_regex(
 828             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
 829         playlist_description = self._og_search_description(webpage, default=None)
 830
 831         def extract_all(pattern):
 832             return list(filter(None, map(
 833                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 834                 re.findall(pattern, webpage))))
 835
 836         # Multiple video article (e.g.
 837         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 838         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 839         entries = []
 840         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 841             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 842             if embed_url and re.match(EMBED_URL, embed_url):
 843                 entries.append(embed_url)
 844         entries.extend(re.findall(
 845             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 846         if entries:
 847             return self.playlist_result(
 848                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 849                 playlist_id, playlist_title, playlist_description)
 850
 851         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 852         medias = extract_all(r"data-media-meta='({[^']+})'")
 853
 854         if not medias:
 855             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 856             media_asset = self._search_regex(
 857                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 858                 webpage, 'media asset', default=None)
 859             if media_asset:
 860                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 861                 medias = []
 862                 for video in media_asset_page.get('videos', {}).values():
 863                     medias.extend(video.values())
 864
 865         if not medias:
 866             # Multiple video playlist with single `now playing` entry (e.g.
 867             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 868             vxp_playlist = self._parse_json(
 869                 self._search_regex(
 870                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 871                     webpage, 'playlist data'),
 872                 playlist_id)
 873             playlist_medias = []
 874             for item in vxp_playlist:
 875                 media = item.get('media')
 876                 if not media:
 877                     continue
 878                 playlist_medias.append(media)
 879                 # Download single video if found media with asset id matching the video id from URL
 880                 if item.get('advert', {}).get('assetId') == playlist_id:
 881                     medias = [media]
 882                     break
 883             # Fallback to the whole playlist
 884             if not medias:
 885                 medias = playlist_medias
 886
 887         entries = []
 888         for num, media_meta in enumerate(medias, start=1):
 889             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 890             if not formats:
 891                 continue
 892             self._sort_formats(formats)
 893
 894             video_id = media_meta.get('externalId')
 895             if not video_id:
 896                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 897
 898             title = media_meta.get('caption')
 899             if not title:
 900                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 901
 902             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 903
 904             images = []
 905             for image in media_meta.get('images', {}).values():
 906                 images.extend(image.values())
 907             if 'image' in media_meta:
 908                 images.append(media_meta['image'])
 909
 910             thumbnails = [{
 911                 'url': image.get('href'),
 912                 'width': int_or_none(image.get('width')),
 913                 'height': int_or_none(image.get('height')),
 914             } for image in images]
 915
 916             entries.append({
 917                 'id': video_id,
 918                 'title': title,
 919                 'thumbnails': thumbnails,
 920                 'duration': duration,
 921                 'timestamp': timestamp,
 922                 'formats': formats,
 923                 'subtitles': subtitles,
 924             })
 925
 926         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 927
 928
 929 class BBCCoUkArticleIE(InfoExtractor):
 930     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 931     IE_NAME = 'bbc.co.uk:article'
 932     IE_DESC = 'BBC articles'
 933
 934     _TEST = {
 935         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 936         'info_dict': {
 937             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 938             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 939             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 940         },
 941         'playlist_count': 4,
 942         'add_ie': ['BBCCoUk'],
 943     }
 944
 945     def _real_extract(self, url):
 946         playlist_id = self._match_id(url)
 947
 948         webpage = self._download_webpage(url, playlist_id)
 949
 950         title = self._og_search_title(webpage)
 951         description = self._og_search_description(webpage).strip()
 952
 953         entries = [self.url_result(programme_url) for programme_url in re.findall(
 954             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 955
 956         return self.playlist_result(entries, playlist_id, title, description)