_ 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/p022h44b',
 128             'note': 'Audio',
 129             'info_dict': {
 130                 'id': 'p022h44j',
 131                 'ext': 'flv',
 132                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 133                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 134                 'duration': 227,
 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/b06rkn85',
 186             'info_dict': {
 187                 'id': 'b06rkms3',
 188                 'ext': 'flv',
 189                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 190                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 191             },
 192             'params': {
 193                 # rtmp download
 194                 'skip_download': True,
 195             },
 196         }, {
 197             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 198             'only_matching': True,
 199         }, {
 200             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 201             'only_matching': True,
 202         }, {
 203             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 204             'only_matching': True,
 205         }, {
 206             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 207             'only_matching': True,
 208         }
 209     ]
 210
 211     class MediaSelectionError(Exception):
 212         def __init__(self, id):
 213             self.id = id
 214
 215     def _extract_asx_playlist(self, connection, programme_id):
 216         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 217         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 218
 219     def _extract_connection(self, connection, programme_id):
 220         formats = []
 221         kind = connection.get('kind')
 222         protocol = connection.get('protocol')
 223         supplier = connection.get('supplier')
 224         if protocol == 'http':
 225             href = connection.get('href')
 226             transfer_format = connection.get('transferFormat')
 227             # ASX playlist
 228             if supplier == 'asx':
 229                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 230                     formats.append({
 231                         'url': ref,
 232                         'format_id': 'ref%s_%s' % (i, supplier),
 233                     })
 234             # Skip DASH until supported
 235             elif transfer_format == 'dash':
 236                 pass
 237             elif transfer_format == 'hls':
 238                 formats.extend(self._extract_m3u8_formats(
 239                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 240                     m3u8_id=supplier, fatal=False))
 241             # Direct link
 242             else:
 243                 formats.append({
 244                     'url': href,
 245                     'format_id': supplier or kind or protocol,
 246                 })
 247         elif protocol == 'rtmp':
 248             application = connection.get('application', 'ondemand')
 249             auth_string = connection.get('authString')
 250             identifier = connection.get('identifier')
 251             server = connection.get('server')
 252             formats.append({
 253                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 254                 'play_path': identifier,
 255                 'app': '%s?%s' % (application, auth_string),
 256                 'page_url': 'http://www.bbc.co.uk',
 257                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 258                 'rtmp_live': False,
 259                 'ext': 'flv',
 260                 'format_id': supplier,
 261             })
 262         return formats
 263
 264     def _extract_items(self, playlist):
 265         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 266
 267     def _findall_ns(self, element, xpath):
 268         elements = []
 269         for ns in self._NAMESPACES:
 270             elements.extend(element.findall(xpath % ns))
 271         return elements
 272
 273     def _extract_medias(self, media_selection):
 274         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 275         if error is None:
 276             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 277         if error is not None:
 278             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 279         return self._findall_ns(media_selection, './{%s}media')
 280
 281     def _extract_connections(self, media):
 282         return self._findall_ns(media, './{%s}connection')
 283
 284     def _extract_video(self, media, programme_id):
 285         formats = []
 286         vbr = int_or_none(media.get('bitrate'))
 287         vcodec = media.get('encoding')
 288         service = media.get('service')
 289         width = int_or_none(media.get('width'))
 290         height = int_or_none(media.get('height'))
 291         file_size = int_or_none(media.get('media_file_size'))
 292         for connection in self._extract_connections(media):
 293             conn_formats = self._extract_connection(connection, programme_id)
 294             for format in conn_formats:
 295                 format.update({
 296                     'width': width,
 297                     'height': height,
 298                     'vbr': vbr,
 299                     'vcodec': vcodec,
 300                     'filesize': file_size,
 301                 })
 302                 if service:
 303                     format['format_id'] = '%s_%s' % (service, format['format_id'])
 304             formats.extend(conn_formats)
 305         return formats
 306
 307     def _extract_audio(self, media, programme_id):
 308         formats = []
 309         abr = int_or_none(media.get('bitrate'))
 310         acodec = media.get('encoding')
 311         service = media.get('service')
 312         for connection in self._extract_connections(media):
 313             conn_formats = self._extract_connection(connection, programme_id)
 314             for format in conn_formats:
 315                 format.update({
 316                     'format_id': '%s_%s' % (service, format['format_id']),
 317                     'abr': abr,
 318                     'acodec': acodec,
 319                 })
 320             formats.extend(conn_formats)
 321         return formats
 322
 323     def _get_subtitles(self, media, programme_id):
 324         subtitles = {}
 325         for connection in self._extract_connections(media):
 326             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 327             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 328             subtitles[lang] = [
 329                 {
 330                     'url': connection.get('href'),
 331                     'ext': 'ttml',
 332                 },
 333             ]
 334         return subtitles
 335
 336     def _raise_extractor_error(self, media_selection_error):
 337         raise ExtractorError(
 338             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 339             expected=True)
 340
 341     def _download_media_selector(self, programme_id):
 342         last_exception = None
 343         for mediaselector_url in self._MEDIASELECTOR_URLS:
 344             try:
 345                 return self._download_media_selector_url(
 346                     mediaselector_url % programme_id, programme_id)
 347             except BBCCoUkIE.MediaSelectionError as e:
 348                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 349                     last_exception = e
 350                     continue
 351                 self._raise_extractor_error(e)
 352         self._raise_extractor_error(last_exception)
 353
 354     def _download_media_selector_url(self, url, programme_id=None):
 355         try:
 356             media_selection = self._download_xml(
 357                 url, programme_id, 'Downloading media selection XML')
 358         except ExtractorError as ee:
 359             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 360                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 361             else:
 362                 raise
 363         return self._process_media_selector(media_selection, programme_id)
 364
 365     def _process_media_selector(self, media_selection, programme_id):
 366         formats = []
 367         subtitles = None
 368
 369         for media in self._extract_medias(media_selection):
 370             kind = media.get('kind')
 371             if kind == 'audio':
 372                 formats.extend(self._extract_audio(media, programme_id))
 373             elif kind == 'video':
 374                 formats.extend(self._extract_video(media, programme_id))
 375             elif kind == 'captions':
 376                 subtitles = self.extract_subtitles(media, programme_id)
 377         return formats, subtitles
 378
 379     def _download_playlist(self, playlist_id):
 380         try:
 381             playlist = self._download_json(
 382                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 383                 playlist_id, 'Downloading playlist JSON')
 384
 385             version = playlist.get('defaultAvailableVersion')
 386             if version:
 387                 smp_config = version['smpConfig']
 388                 title = smp_config['title']
 389                 description = smp_config['summary']
 390                 for item in smp_config['items']:
 391                     kind = item['kind']
 392                     if kind != 'programme' and kind != 'radioProgramme':
 393                         continue
 394                     programme_id = item.get('vpid')
 395                     duration = int_or_none(item.get('duration'))
 396                     formats, subtitles = self._download_media_selector(programme_id)
 397                 return programme_id, title, description, duration, formats, subtitles
 398         except ExtractorError as ee:
 399             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 400                 raise
 401
 402         # fallback to legacy playlist
 403         return self._process_legacy_playlist(playlist_id)
 404
 405     def _process_legacy_playlist_url(self, url, display_id):
 406         playlist = self._download_legacy_playlist_url(url, display_id)
 407         return self._extract_from_legacy_playlist(playlist, display_id)
 408
 409     def _process_legacy_playlist(self, playlist_id):
 410         return self._process_legacy_playlist_url(
 411             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 412
 413     def _download_legacy_playlist_url(self, url, playlist_id=None):
 414         return self._download_xml(
 415             url, playlist_id, 'Downloading legacy playlist XML')
 416
 417     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 418         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 419         if no_items is not None:
 420             reason = no_items.get('reason')
 421             if reason == 'preAvailability':
 422                 msg = 'Episode %s is not yet available' % playlist_id
 423             elif reason == 'postAvailability':
 424                 msg = 'Episode %s is no longer available' % playlist_id
 425             elif reason == 'noMedia':
 426                 msg = 'Episode %s is not currently available' % playlist_id
 427             else:
 428                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 429             raise ExtractorError(msg, expected=True)
 430
 431         for item in self._extract_items(playlist):
 432             kind = item.get('kind')
 433             if kind != 'programme' and kind != 'radioProgramme':
 434                 continue
 435             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 436             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 437             description = description_el.text if description_el is not None else None
 438
 439             def get_programme_id(item):
 440                 def get_from_attributes(item):
 441                     for p in('identifier', 'group'):
 442                         value = item.get(p)
 443                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 444                             return value
 445                 get_from_attributes(item)
 446                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 447                 if mediator is not None:
 448                     return get_from_attributes(mediator)
 449
 450             programme_id = get_programme_id(item)
 451             duration = int_or_none(item.get('duration'))
 452
 453             if programme_id:
 454                 formats, subtitles = self._download_media_selector(programme_id)
 455             else:
 456                 formats, subtitles = self._process_media_selector(item, playlist_id)
 457                 programme_id = playlist_id
 458
 459         return programme_id, title, description, duration, formats, subtitles
 460
 461     def _real_extract(self, url):
 462         group_id = self._match_id(url)
 463
 464         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 465
 466         programme_id = None
 467         duration = None
 468
 469         tviplayer = self._search_regex(
 470             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 471             webpage, 'player', default=None)
 472
 473         if tviplayer:
 474             player = self._parse_json(tviplayer, group_id).get('player', {})
 475             duration = int_or_none(player.get('duration'))
 476             programme_id = player.get('vpid')
 477
 478         if not programme_id:
 479             programme_id = self._search_regex(
 480                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 481
 482         if programme_id:
 483             formats, subtitles = self._download_media_selector(programme_id)
 484             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 485                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 486                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), 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         json_ld_info = self._search_json_ld(webpage, playlist_id, default=None)
 723         timestamp = json_ld_info.get('timestamp')
 724         playlist_title = json_ld_info.get('title')
 725         playlist_description = json_ld_info.get('description')
 726
 727         if not timestamp:
 728             timestamp = parse_iso8601(self._search_regex(
 729                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 730                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 731                  r'"datePublished":\s*"([^"]+)'],
 732                 webpage, 'date', default=None))
 733
 734         entries = []
 735
 736         # article with multiple videos embedded with playlist.sxml (e.g.
 737         # http://www.bbc.com/sport/0/football/34475836)
 738         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 739         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 740         if playlists:
 741             entries = [
 742                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 743                 for playlist_url in playlists]
 744
 745         # news article with multiple videos embedded with data-playable
 746         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 747         if data_playables:
 748             for _, data_playable_json in data_playables:
 749                 data_playable = self._parse_json(
 750                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 751                 if not data_playable:
 752                     continue
 753                 settings = data_playable.get('settings', {})
 754                 if settings:
 755                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 756                     # http://www.bbc.com/news/world-us-canada-34473351)
 757                     playlist_object = settings.get('playlistObject', {})
 758                     if playlist_object:
 759                         items = playlist_object.get('items')
 760                         if items and isinstance(items, list):
 761                             title = playlist_object['title']
 762                             description = playlist_object.get('summary')
 763                             duration = int_or_none(items[0].get('duration'))
 764                             programme_id = items[0].get('vpid')
 765                             formats, subtitles = self._download_media_selector(programme_id)
 766                             self._sort_formats(formats)
 767                             entries.append({
 768                                 'id': programme_id,
 769                                 'title': title,
 770                                 'description': description,
 771                                 'timestamp': timestamp,
 772                                 'duration': duration,
 773                                 'formats': formats,
 774                                 'subtitles': subtitles,
 775                             })
 776                     else:
 777                         # data-playable without vpid but with a playlist.sxml URLs
 778                         # in otherSettings.playlist (e.g.
 779                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 780                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 781                         if playlist:
 782                             entries.append(self._extract_from_playlist_sxml(
 783                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 784
 785         if entries:
 786             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
 787             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
 788             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 789
 790         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 791         programme_id = self._search_regex(
 792             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
 793              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 794              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 795             webpage, 'vpid', default=None)
 796
 797         if programme_id:
 798             formats, subtitles = self._download_media_selector(programme_id)
 799             self._sort_formats(formats)
 800             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 801             digital_data = self._parse_json(
 802                 self._search_regex(
 803                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 804                 programme_id, fatal=False)
 805             page_info = digital_data.get('page', {}).get('pageInfo', {})
 806             title = page_info.get('pageName') or self._og_search_title(webpage)
 807             description = page_info.get('description') or self._og_search_description(webpage)
 808             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 809             return {
 810                 'id': programme_id,
 811                 'title': title,
 812                 'description': description,
 813                 'timestamp': timestamp,
 814                 'formats': formats,
 815                 'subtitles': subtitles,
 816             }
 817
 818         playlist_title = self._html_search_regex(
 819             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
 820         playlist_description = self._og_search_description(webpage, default=None)
 821
 822         def extract_all(pattern):
 823             return list(filter(None, map(
 824                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 825                 re.findall(pattern, webpage))))
 826
 827         # Multiple video article (e.g.
 828         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 829         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 830         entries = []
 831         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 832             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 833             if embed_url and re.match(EMBED_URL, embed_url):
 834                 entries.append(embed_url)
 835         entries.extend(re.findall(
 836             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 837         if entries:
 838             return self.playlist_result(
 839                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 840                 playlist_id, playlist_title, playlist_description)
 841
 842         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 843         medias = extract_all(r"data-media-meta='({[^']+})'")
 844
 845         if not medias:
 846             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 847             media_asset = self._search_regex(
 848                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 849                 webpage, 'media asset', default=None)
 850             if media_asset:
 851                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 852                 medias = []
 853                 for video in media_asset_page.get('videos', {}).values():
 854                     medias.extend(video.values())
 855
 856         if not medias:
 857             # Multiple video playlist with single `now playing` entry (e.g.
 858             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 859             vxp_playlist = self._parse_json(
 860                 self._search_regex(
 861                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 862                     webpage, 'playlist data'),
 863                 playlist_id)
 864             playlist_medias = []
 865             for item in vxp_playlist:
 866                 media = item.get('media')
 867                 if not media:
 868                     continue
 869                 playlist_medias.append(media)
 870                 # Download single video if found media with asset id matching the video id from URL
 871                 if item.get('advert', {}).get('assetId') == playlist_id:
 872                     medias = [media]
 873                     break
 874             # Fallback to the whole playlist
 875             if not medias:
 876                 medias = playlist_medias
 877
 878         entries = []
 879         for num, media_meta in enumerate(medias, start=1):
 880             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 881             if not formats:
 882                 continue
 883             self._sort_formats(formats)
 884
 885             video_id = media_meta.get('externalId')
 886             if not video_id:
 887                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 888
 889             title = media_meta.get('caption')
 890             if not title:
 891                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 892
 893             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 894
 895             images = []
 896             for image in media_meta.get('images', {}).values():
 897                 images.extend(image.values())
 898             if 'image' in media_meta:
 899                 images.append(media_meta['image'])
 900
 901             thumbnails = [{
 902                 'url': image.get('href'),
 903                 'width': int_or_none(image.get('width')),
 904                 'height': int_or_none(image.get('height')),
 905             } for image in images]
 906
 907             entries.append({
 908                 'id': video_id,
 909                 'title': title,
 910                 'thumbnails': thumbnails,
 911                 'duration': duration,
 912                 'timestamp': timestamp,
 913                 'formats': formats,
 914                 'subtitles': subtitles,
 915             })
 916
 917         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 918
 919
 920 class BBCCoUkArticleIE(InfoExtractor):
 921     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 922     IE_NAME = 'bbc.co.uk:article'
 923     IE_DESC = 'BBC articles'
 924
 925     _TEST = {
 926         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 927         'info_dict': {
 928             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 929             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 930             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 931         },
 932         'playlist_count': 4,
 933         'add_ie': ['BBCCoUk'],
 934     }
 935
 936     def _real_extract(self, url):
 937         playlist_id = self._match_id(url)
 938
 939         webpage = self._download_webpage(url, playlist_id)
 940
 941         title = self._og_search_title(webpage)
 942         description = self._og_search_description(webpage).strip()
 943
 944         entries = [self.url_result(programme_url) for programme_url in re.findall(
 945             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 946
 947         return self.playlist_result(entries, playlist_id, title, description)