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': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  51                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  52             },
  53             'params': {
  54                 # rtmp download
  55                 'skip_download': True,
  56             }
  57         },
  58         {
  59             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  60             'info_dict': {
  61                 'id': 'b00yng1d',
  62                 'ext': 'flv',
  63                 'title': 'The Man in Black: Series 3: The Printed Name',
  64                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  65                 'duration': 1800,
  66             },
  67             'params': {
  68                 # rtmp download
  69                 'skip_download': True,
  70             },
  71             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  72         },
  73         {
  74             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  75             'info_dict': {
  76                 'id': 'b00yng1d',
  77                 'ext': 'flv',
  78                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  79                 '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.",
  80                 'duration': 5100,
  81             },
  82             'params': {
  83                 # rtmp download
  84                 'skip_download': True,
  85             },
  86             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
  87         },
  88         {
  89             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
  90             'info_dict': {
  91                 'id': 'b03k3pb7',
  92                 'ext': 'flv',
  93                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
  94                 'description': '2. Invasion',
  95                 'duration': 3600,
  96             },
  97             'params': {
  98                 # rtmp download
  99                 'skip_download': True,
 100             },
 101             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 102         }, {
 103             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 104             'info_dict': {
 105                 'id': 'b04v209v',
 106                 'ext': 'flv',
 107                 'title': 'Pete Tong, The Essential New Tune Special',
 108                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 109                 'duration': 10800,
 110             },
 111             'params': {
 112                 # rtmp download
 113                 'skip_download': True,
 114             },
 115             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 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         duration = None
 458
 459         tviplayer = self._search_regex(
 460             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 461             webpage, 'player', default=None)
 462
 463         if tviplayer:
 464             player = self._parse_json(tviplayer, group_id).get('player', {})
 465             duration = int_or_none(player.get('duration'))
 466             programme_id = player.get('vpid')
 467
 468         if not programme_id:
 469             programme_id = self._search_regex(
 470                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 471
 472         if programme_id:
 473             formats, subtitles = self._download_media_selector(programme_id)
 474             title = self._og_search_title(webpage)
 475             description = self._search_regex(
 476                 r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 477                 webpage, 'description', default=None)
 478             if not description:
 479                 description = self._html_search_meta('description', webpage)
 480         else:
 481             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 482
 483         self._sort_formats(formats)
 484
 485         return {
 486             'id': programme_id,
 487             'title': title,
 488             'description': description,
 489             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 490             'duration': duration,
 491             'formats': formats,
 492             'subtitles': subtitles,
 493         }
 494
 495
 496 class BBCIE(BBCCoUkIE):
 497     IE_NAME = 'bbc'
 498     IE_DESC = 'BBC'
 499     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 500
 501     _MEDIASELECTOR_URLS = [
 502         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 503         # even not geo restricted at all
 504         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 505         # Provides more formats, namely direct mp4 links, but fails on some videos with
 506         # notukerror for non UK (?) users (e.g.
 507         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 508         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 509         # Provides fewer formats, but works everywhere for everybody (hopefully)
 510         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 511     ]
 512
 513     _TESTS = [{
 514         # article with multiple videos embedded with data-playable containing vpids
 515         'url': 'http://www.bbc.com/news/world-europe-32668511',
 516         'info_dict': {
 517             'id': 'world-europe-32668511',
 518             'title': 'Russia stages massive WW2 parade despite Western boycott',
 519             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 520         },
 521         'playlist_count': 2,
 522     }, {
 523         # article with multiple videos embedded with data-playable (more videos)
 524         'url': 'http://www.bbc.com/news/business-28299555',
 525         'info_dict': {
 526             'id': 'business-28299555',
 527             'title': 'Farnborough Airshow: Video highlights',
 528             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 529         },
 530         'playlist_count': 9,
 531         'skip': 'Save time',
 532     }, {
 533         # article with multiple videos embedded with `new SMP()`
 534         # broken
 535         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 536         'info_dict': {
 537             'id': '3662a707-0af9-3149-963f-47bea720b460',
 538             'title': 'BBC Blogs - Adam Curtis - BUGGER',
 539         },
 540         'playlist_count': 18,
 541     }, {
 542         # single video embedded with data-playable containing vpid
 543         'url': 'http://www.bbc.com/news/world-europe-32041533',
 544         'info_dict': {
 545             'id': 'p02mprgb',
 546             'ext': 'mp4',
 547             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 548             'description': 'md5:2868290467291b37feda7863f7a83f54',
 549             'duration': 47,
 550             'timestamp': 1427219242,
 551             'upload_date': '20150324',
 552         },
 553         'params': {
 554             # rtmp download
 555             'skip_download': True,
 556         }
 557     }, {
 558         # article with single video embedded with data-playable containing XML playlist
 559         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 560         # and playlist with f4m and m3u8 as streamingUrl
 561         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 562         'info_dict': {
 563             'id': '150615_telabyad_kentin_cogu',
 564             'ext': 'mp4',
 565             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 566             'timestamp': 1434397334,
 567             'upload_date': '20150615',
 568         },
 569         'params': {
 570             'skip_download': True,
 571         }
 572     }, {
 573         # single video embedded with data-playable containing XML playlists (regional section)
 574         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 575         'info_dict': {
 576             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 577             'ext': 'mp4',
 578             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 579             'timestamp': 1434713142,
 580             'upload_date': '20150619',
 581         },
 582         'params': {
 583             'skip_download': True,
 584         }
 585     }, {
 586         # single video from video playlist embedded with vxp-playlist-data JSON
 587         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 588         'info_dict': {
 589             'id': 'p02w6qjc',
 590             'ext': 'mp4',
 591             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 592             'duration': 56,
 593             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 594         },
 595         'params': {
 596             'skip_download': True,
 597         }
 598     }, {
 599         # single video story with digitalData
 600         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 601         'info_dict': {
 602             'id': 'p02q6gc4',
 603             'ext': 'flv',
 604             'title': 'Sri Lanka’s spicy secret',
 605             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 606             'timestamp': 1437674293,
 607             'upload_date': '20150723',
 608         },
 609         'params': {
 610             # rtmp download
 611             'skip_download': True,
 612         }
 613     }, {
 614         # single video story without digitalData
 615         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 616         'info_dict': {
 617             'id': 'p018zqqg',
 618             'ext': 'mp4',
 619             'title': 'Hyundai Santa Fe Sport: Rock star',
 620             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 621             'timestamp': 1415867444,
 622             'upload_date': '20141113',
 623         },
 624         'params': {
 625             # rtmp download
 626             'skip_download': True,
 627         }
 628     }, {
 629         # single video with playlist.sxml URL in playlist param
 630         'url': 'http://www.bbc.com/sport/0/football/33653409',
 631         'info_dict': {
 632             'id': 'p02xycnp',
 633             'ext': 'mp4',
 634             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 635             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 636             'duration': 140,
 637         },
 638         'params': {
 639             # rtmp download
 640             'skip_download': True,
 641         }
 642     }, {
 643         # article with multiple videos embedded with playlist.sxml in playlist param
 644         'url': 'http://www.bbc.com/sport/0/football/34475836',
 645         'info_dict': {
 646             'id': '34475836',
 647             'title': 'What Liverpool can expect from Klopp',
 648         },
 649         'playlist_count': 3,
 650     }, {
 651         # single video with playlist URL from weather section
 652         'url': 'http://www.bbc.com/weather/features/33601775',
 653         'only_matching': True,
 654     }, {
 655         # custom redirection to www.bbc.com
 656         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 657         'only_matching': True,
 658     }]
 659
 660     @classmethod
 661     def suitable(cls, url):
 662         return False if BBCCoUkIE.suitable(url) or BBCCoUkArticleIE.suitable(url) else super(BBCIE, cls).suitable(url)
 663
 664     def _extract_from_media_meta(self, media_meta, video_id):
 665         # Direct links to media in media metadata (e.g.
 666         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 667         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 668         source_files = media_meta.get('sourceFiles')
 669         if source_files:
 670             return [{
 671                 'url': f['url'],
 672                 'format_id': format_id,
 673                 'ext': f.get('encoding'),
 674                 'tbr': float_or_none(f.get('bitrate'), 1000),
 675                 'filesize': int_or_none(f.get('filesize')),
 676             } for format_id, f in source_files.items() if f.get('url')], []
 677
 678         programme_id = media_meta.get('externalId')
 679         if programme_id:
 680             return self._download_media_selector(programme_id)
 681
 682         # Process playlist.sxml as legacy playlist
 683         href = media_meta.get('href')
 684         if href:
 685             playlist = self._download_legacy_playlist_url(href)
 686             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 687             return formats, subtitles
 688
 689         return [], []
 690
 691     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 692         programme_id, title, description, duration, formats, subtitles = \
 693             self._process_legacy_playlist_url(url, playlist_id)
 694         self._sort_formats(formats)
 695         return {
 696             'id': programme_id,
 697             'title': title,
 698             'description': description,
 699             'duration': duration,
 700             'timestamp': timestamp,
 701             'formats': formats,
 702             'subtitles': subtitles,
 703         }
 704
 705     def _real_extract(self, url):
 706         playlist_id = self._match_id(url)
 707
 708         webpage = self._download_webpage(url, playlist_id)
 709
 710         timestamp = None
 711         playlist_title = None
 712         playlist_description = None
 713
 714         ld = self._parse_json(
 715             self._search_regex(
 716                 r'(?s)<script type="application/ld\+json">(.+?)</script>',
 717                 webpage, 'ld json', default='{}'),
 718             playlist_id, fatal=False)
 719         if ld:
 720             timestamp = parse_iso8601(ld.get('datePublished'))
 721             playlist_title = ld.get('headline')
 722             playlist_description = ld.get('articleBody')
 723
 724         if not timestamp:
 725             timestamp = parse_iso8601(self._search_regex(
 726                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 727                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 728                  r'"datePublished":\s*"([^"]+)'],
 729                 webpage, 'date', default=None))
 730
 731         entries = []
 732
 733         # article with multiple videos embedded with playlist.sxml (e.g.
 734         # http://www.bbc.com/sport/0/football/34475836)
 735         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 736         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 737         if playlists:
 738             entries = [
 739                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 740                 for playlist_url in playlists]
 741
 742         # news article with multiple videos embedded with data-playable
 743         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 744         if data_playables:
 745             for _, data_playable_json in data_playables:
 746                 data_playable = self._parse_json(
 747                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 748                 if not data_playable:
 749                     continue
 750                 settings = data_playable.get('settings', {})
 751                 if settings:
 752                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 753                     # http://www.bbc.com/news/world-us-canada-34473351)
 754                     playlist_object = settings.get('playlistObject', {})
 755                     if playlist_object:
 756                         items = playlist_object.get('items')
 757                         if items and isinstance(items, list):
 758                             title = playlist_object['title']
 759                             description = playlist_object.get('summary')
 760                             duration = int_or_none(items[0].get('duration'))
 761                             programme_id = items[0].get('vpid')
 762                             formats, subtitles = self._download_media_selector(programme_id)
 763                             self._sort_formats(formats)
 764                             entries.append({
 765                                 'id': programme_id,
 766                                 'title': title,
 767                                 'description': description,
 768                                 'timestamp': timestamp,
 769                                 'duration': duration,
 770                                 'formats': formats,
 771                                 'subtitles': subtitles,
 772                             })
 773                     else:
 774                         # data-playable without vpid but with a playlist.sxml URLs
 775                         # in otherSettings.playlist (e.g.
 776                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 777                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 778                         if playlist:
 779                             entries.append(self._extract_from_playlist_sxml(
 780                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 781
 782         if entries:
 783             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
 784             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
 785             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 786
 787         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 788         programme_id = self._search_regex(
 789             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
 790              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 791              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 792             webpage, 'vpid', default=None)
 793
 794         if programme_id:
 795             formats, subtitles = self._download_media_selector(programme_id)
 796             self._sort_formats(formats)
 797             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 798             digital_data = self._parse_json(
 799                 self._search_regex(
 800                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 801                 programme_id, fatal=False)
 802             page_info = digital_data.get('page', {}).get('pageInfo', {})
 803             title = page_info.get('pageName') or self._og_search_title(webpage)
 804             description = page_info.get('description') or self._og_search_description(webpage)
 805             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 806             return {
 807                 'id': programme_id,
 808                 'title': title,
 809                 'description': description,
 810                 'timestamp': timestamp,
 811                 'formats': formats,
 812                 'subtitles': subtitles,
 813             }
 814
 815         playlist_title = self._html_search_regex(
 816             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
 817         playlist_description = self._og_search_description(webpage, default=None)
 818
 819         def extract_all(pattern):
 820             return list(filter(None, map(
 821                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 822                 re.findall(pattern, webpage))))
 823
 824         # Multiple video article (e.g.
 825         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 826         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 827         entries = []
 828         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 829             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 830             if embed_url and re.match(EMBED_URL, embed_url):
 831                 entries.append(embed_url)
 832         entries.extend(re.findall(
 833             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 834         if entries:
 835             return self.playlist_result(
 836                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 837                 playlist_id, playlist_title, playlist_description)
 838
 839         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 840         medias = extract_all(r"data-media-meta='({[^']+})'")
 841
 842         if not medias:
 843             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 844             media_asset = self._search_regex(
 845                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 846                 webpage, 'media asset', default=None)
 847             if media_asset:
 848                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 849                 medias = []
 850                 for video in media_asset_page.get('videos', {}).values():
 851                     medias.extend(video.values())
 852
 853         if not medias:
 854             # Multiple video playlist with single `now playing` entry (e.g.
 855             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 856             vxp_playlist = self._parse_json(
 857                 self._search_regex(
 858                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 859                     webpage, 'playlist data'),
 860                 playlist_id)
 861             playlist_medias = []
 862             for item in vxp_playlist:
 863                 media = item.get('media')
 864                 if not media:
 865                     continue
 866                 playlist_medias.append(media)
 867                 # Download single video if found media with asset id matching the video id from URL
 868                 if item.get('advert', {}).get('assetId') == playlist_id:
 869                     medias = [media]
 870                     break
 871             # Fallback to the whole playlist
 872             if not medias:
 873                 medias = playlist_medias
 874
 875         entries = []
 876         for num, media_meta in enumerate(medias, start=1):
 877             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 878             if not formats:
 879                 continue
 880             self._sort_formats(formats)
 881
 882             video_id = media_meta.get('externalId')
 883             if not video_id:
 884                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 885
 886             title = media_meta.get('caption')
 887             if not title:
 888                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 889
 890             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 891
 892             images = []
 893             for image in media_meta.get('images', {}).values():
 894                 images.extend(image.values())
 895             if 'image' in media_meta:
 896                 images.append(media_meta['image'])
 897
 898             thumbnails = [{
 899                 'url': image.get('href'),
 900                 'width': int_or_none(image.get('width')),
 901                 'height': int_or_none(image.get('height')),
 902             } for image in images]
 903
 904             entries.append({
 905                 'id': video_id,
 906                 'title': title,
 907                 'thumbnails': thumbnails,
 908                 'duration': duration,
 909                 'timestamp': timestamp,
 910                 'formats': formats,
 911                 'subtitles': subtitles,
 912             })
 913
 914         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 915
 916
 917 class BBCCoUkArticleIE(InfoExtractor):
 918     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 919     IE_NAME = 'bbc.co.uk:article'
 920     IE_DESC = 'BBC articles'
 921
 922     _TEST = {
 923         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 924         'info_dict': {
 925             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 926             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 927             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 928         },
 929         'playlist_count': 4,
 930         'add_ie': ['BBCCoUk'],
 931     }
 932
 933     def _real_extract(self, url):
 934         playlist_id = self._match_id(url)
 935
 936         webpage = self._download_webpage(url, playlist_id)
 937
 938         title = self._og_search_title(webpage)
 939         description = self._og_search_description(webpage).strip()
 940
 941         entries = [self.url_result(programme_url) for programme_url in re.findall(
 942             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 943
 944         return self.playlist_result(entries, playlist_id, title, description)