_ 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)
 486             description = self._search_regex(
 487                 r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 488                 webpage, 'description', default=None)
 489             if not description:
 490                 description = self._html_search_meta('description', webpage)
 491         else:
 492             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 493
 494         self._sort_formats(formats)
 495
 496         return {
 497             'id': programme_id,
 498             'title': title,
 499             'description': description,
 500             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 501             'duration': duration,
 502             'formats': formats,
 503             'subtitles': subtitles,
 504         }
 505
 506
 507 class BBCIE(BBCCoUkIE):
 508     IE_NAME = 'bbc'
 509     IE_DESC = 'BBC'
 510     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 511
 512     _MEDIASELECTOR_URLS = [
 513         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 514         # even not geo restricted at all
 515         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 516         # Provides more formats, namely direct mp4 links, but fails on some videos with
 517         # notukerror for non UK (?) users (e.g.
 518         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 519         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 520         # Provides fewer formats, but works everywhere for everybody (hopefully)
 521         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 522     ]
 523
 524     _TESTS = [{
 525         # article with multiple videos embedded with data-playable containing vpids
 526         'url': 'http://www.bbc.com/news/world-europe-32668511',
 527         'info_dict': {
 528             'id': 'world-europe-32668511',
 529             'title': 'Russia stages massive WW2 parade despite Western boycott',
 530             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 531         },
 532         'playlist_count': 2,
 533     }, {
 534         # article with multiple videos embedded with data-playable (more videos)
 535         'url': 'http://www.bbc.com/news/business-28299555',
 536         'info_dict': {
 537             'id': 'business-28299555',
 538             'title': 'Farnborough Airshow: Video highlights',
 539             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 540         },
 541         'playlist_count': 9,
 542         'skip': 'Save time',
 543     }, {
 544         # article with multiple videos embedded with `new SMP()`
 545         # broken
 546         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 547         'info_dict': {
 548             'id': '3662a707-0af9-3149-963f-47bea720b460',
 549             'title': 'BBC Blogs - Adam Curtis - BUGGER',
 550         },
 551         'playlist_count': 18,
 552     }, {
 553         # single video embedded with data-playable containing vpid
 554         'url': 'http://www.bbc.com/news/world-europe-32041533',
 555         'info_dict': {
 556             'id': 'p02mprgb',
 557             'ext': 'mp4',
 558             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 559             'description': 'md5:2868290467291b37feda7863f7a83f54',
 560             'duration': 47,
 561             'timestamp': 1427219242,
 562             'upload_date': '20150324',
 563         },
 564         'params': {
 565             # rtmp download
 566             'skip_download': True,
 567         }
 568     }, {
 569         # article with single video embedded with data-playable containing XML playlist
 570         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 571         # and playlist with f4m and m3u8 as streamingUrl
 572         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 573         'info_dict': {
 574             'id': '150615_telabyad_kentin_cogu',
 575             'ext': 'mp4',
 576             'title': "YPG: Tel Abyad'ın tamamı kontrolümüzde",
 577             'timestamp': 1434397334,
 578             'upload_date': '20150615',
 579         },
 580         'params': {
 581             'skip_download': True,
 582         }
 583     }, {
 584         # single video embedded with data-playable containing XML playlists (regional section)
 585         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 586         'info_dict': {
 587             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 588             'ext': 'mp4',
 589             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 590             'timestamp': 1434713142,
 591             'upload_date': '20150619',
 592         },
 593         'params': {
 594             'skip_download': True,
 595         }
 596     }, {
 597         # single video from video playlist embedded with vxp-playlist-data JSON
 598         'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376',
 599         'info_dict': {
 600             'id': 'p02w6qjc',
 601             'ext': 'mp4',
 602             'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 603             'duration': 56,
 604             'description': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''',
 605         },
 606         'params': {
 607             'skip_download': True,
 608         }
 609     }, {
 610         # single video story with digitalData
 611         'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret',
 612         'info_dict': {
 613             'id': 'p02q6gc4',
 614             'ext': 'flv',
 615             'title': 'Sri Lanka’s spicy secret',
 616             'description': 'As a new train line to Jaffna opens up the country’s north, travellers can experience a truly distinct slice of Tamil culture.',
 617             'timestamp': 1437674293,
 618             'upload_date': '20150723',
 619         },
 620         'params': {
 621             # rtmp download
 622             'skip_download': True,
 623         }
 624     }, {
 625         # single video story without digitalData
 626         'url': 'http://www.bbc.com/autos/story/20130513-hyundais-rock-star',
 627         'info_dict': {
 628             'id': 'p018zqqg',
 629             'ext': 'mp4',
 630             'title': 'Hyundai Santa Fe Sport: Rock star',
 631             'description': 'md5:b042a26142c4154a6e472933cf20793d',
 632             'timestamp': 1415867444,
 633             'upload_date': '20141113',
 634         },
 635         'params': {
 636             # rtmp download
 637             'skip_download': True,
 638         }
 639     }, {
 640         # single video with playlist.sxml URL in playlist param
 641         'url': 'http://www.bbc.com/sport/0/football/33653409',
 642         'info_dict': {
 643             'id': 'p02xycnp',
 644             'ext': 'mp4',
 645             'title': 'Transfers: Cristiano Ronaldo to Man Utd, Arsenal to spend?',
 646             'description': 'BBC Sport\'s David Ornstein has the latest transfer gossip, including rumours of a Manchester United return for Cristiano Ronaldo.',
 647             'duration': 140,
 648         },
 649         'params': {
 650             # rtmp download
 651             'skip_download': True,
 652         }
 653     }, {
 654         # article with multiple videos embedded with playlist.sxml in playlist param
 655         'url': 'http://www.bbc.com/sport/0/football/34475836',
 656         'info_dict': {
 657             'id': '34475836',
 658             'title': 'What Liverpool can expect from Klopp',
 659         },
 660         'playlist_count': 3,
 661     }, {
 662         # single video with playlist URL from weather section
 663         'url': 'http://www.bbc.com/weather/features/33601775',
 664         'only_matching': True,
 665     }, {
 666         # custom redirection to www.bbc.com
 667         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 668         'only_matching': True,
 669     }]
 670
 671     @classmethod
 672     def suitable(cls, url):
 673         return False if BBCCoUkIE.suitable(url) or BBCCoUkArticleIE.suitable(url) else super(BBCIE, cls).suitable(url)
 674
 675     def _extract_from_media_meta(self, media_meta, video_id):
 676         # Direct links to media in media metadata (e.g.
 677         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 678         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 679         source_files = media_meta.get('sourceFiles')
 680         if source_files:
 681             return [{
 682                 'url': f['url'],
 683                 'format_id': format_id,
 684                 'ext': f.get('encoding'),
 685                 'tbr': float_or_none(f.get('bitrate'), 1000),
 686                 'filesize': int_or_none(f.get('filesize')),
 687             } for format_id, f in source_files.items() if f.get('url')], []
 688
 689         programme_id = media_meta.get('externalId')
 690         if programme_id:
 691             return self._download_media_selector(programme_id)
 692
 693         # Process playlist.sxml as legacy playlist
 694         href = media_meta.get('href')
 695         if href:
 696             playlist = self._download_legacy_playlist_url(href)
 697             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 698             return formats, subtitles
 699
 700         return [], []
 701
 702     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 703         programme_id, title, description, duration, formats, subtitles = \
 704             self._process_legacy_playlist_url(url, playlist_id)
 705         self._sort_formats(formats)
 706         return {
 707             'id': programme_id,
 708             'title': title,
 709             'description': description,
 710             'duration': duration,
 711             'timestamp': timestamp,
 712             'formats': formats,
 713             'subtitles': subtitles,
 714         }
 715
 716     def _real_extract(self, url):
 717         playlist_id = self._match_id(url)
 718
 719         webpage = self._download_webpage(url, playlist_id)
 720
 721         timestamp = None
 722         playlist_title = None
 723         playlist_description = None
 724
 725         ld = self._parse_json(
 726             self._search_regex(
 727                 r'(?s)<script type="application/ld\+json">(.+?)</script>',
 728                 webpage, 'ld json', default='{}'),
 729             playlist_id, fatal=False)
 730         if ld:
 731             timestamp = parse_iso8601(ld.get('datePublished'))
 732             playlist_title = ld.get('headline')
 733             playlist_description = ld.get('articleBody')
 734
 735         if not timestamp:
 736             timestamp = parse_iso8601(self._search_regex(
 737                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 738                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 739                  r'"datePublished":\s*"([^"]+)'],
 740                 webpage, 'date', default=None))
 741
 742         entries = []
 743
 744         # article with multiple videos embedded with playlist.sxml (e.g.
 745         # http://www.bbc.com/sport/0/football/34475836)
 746         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 747         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 748         if playlists:
 749             entries = [
 750                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 751                 for playlist_url in playlists]
 752
 753         # news article with multiple videos embedded with data-playable
 754         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 755         if data_playables:
 756             for _, data_playable_json in data_playables:
 757                 data_playable = self._parse_json(
 758                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 759                 if not data_playable:
 760                     continue
 761                 settings = data_playable.get('settings', {})
 762                 if settings:
 763                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 764                     # http://www.bbc.com/news/world-us-canada-34473351)
 765                     playlist_object = settings.get('playlistObject', {})
 766                     if playlist_object:
 767                         items = playlist_object.get('items')
 768                         if items and isinstance(items, list):
 769                             title = playlist_object['title']
 770                             description = playlist_object.get('summary')
 771                             duration = int_or_none(items[0].get('duration'))
 772                             programme_id = items[0].get('vpid')
 773                             formats, subtitles = self._download_media_selector(programme_id)
 774                             self._sort_formats(formats)
 775                             entries.append({
 776                                 'id': programme_id,
 777                                 'title': title,
 778                                 'description': description,
 779                                 'timestamp': timestamp,
 780                                 'duration': duration,
 781                                 'formats': formats,
 782                                 'subtitles': subtitles,
 783                             })
 784                     else:
 785                         # data-playable without vpid but with a playlist.sxml URLs
 786                         # in otherSettings.playlist (e.g.
 787                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 788                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 789                         if playlist:
 790                             entries.append(self._extract_from_playlist_sxml(
 791                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
 792
 793         if entries:
 794             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
 795             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
 796             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 797
 798         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 799         programme_id = self._search_regex(
 800             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
 801              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 802              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 803             webpage, 'vpid', default=None)
 804
 805         if programme_id:
 806             formats, subtitles = self._download_media_selector(programme_id)
 807             self._sort_formats(formats)
 808             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 809             digital_data = self._parse_json(
 810                 self._search_regex(
 811                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 812                 programme_id, fatal=False)
 813             page_info = digital_data.get('page', {}).get('pageInfo', {})
 814             title = page_info.get('pageName') or self._og_search_title(webpage)
 815             description = page_info.get('description') or self._og_search_description(webpage)
 816             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 817             return {
 818                 'id': programme_id,
 819                 'title': title,
 820                 'description': description,
 821                 'timestamp': timestamp,
 822                 'formats': formats,
 823                 'subtitles': subtitles,
 824             }
 825
 826         playlist_title = self._html_search_regex(
 827             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
 828         playlist_description = self._og_search_description(webpage, default=None)
 829
 830         def extract_all(pattern):
 831             return list(filter(None, map(
 832                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 833                 re.findall(pattern, webpage))))
 834
 835         # Multiple video article (e.g.
 836         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 837         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 838         entries = []
 839         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 840             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 841             if embed_url and re.match(EMBED_URL, embed_url):
 842                 entries.append(embed_url)
 843         entries.extend(re.findall(
 844             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 845         if entries:
 846             return self.playlist_result(
 847                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 848                 playlist_id, playlist_title, playlist_description)
 849
 850         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 851         medias = extract_all(r"data-media-meta='({[^']+})'")
 852
 853         if not medias:
 854             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 855             media_asset = self._search_regex(
 856                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 857                 webpage, 'media asset', default=None)
 858             if media_asset:
 859                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 860                 medias = []
 861                 for video in media_asset_page.get('videos', {}).values():
 862                     medias.extend(video.values())
 863
 864         if not medias:
 865             # Multiple video playlist with single `now playing` entry (e.g.
 866             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 867             vxp_playlist = self._parse_json(
 868                 self._search_regex(
 869                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 870                     webpage, 'playlist data'),
 871                 playlist_id)
 872             playlist_medias = []
 873             for item in vxp_playlist:
 874                 media = item.get('media')
 875                 if not media:
 876                     continue
 877                 playlist_medias.append(media)
 878                 # Download single video if found media with asset id matching the video id from URL
 879                 if item.get('advert', {}).get('assetId') == playlist_id:
 880                     medias = [media]
 881                     break
 882             # Fallback to the whole playlist
 883             if not medias:
 884                 medias = playlist_medias
 885
 886         entries = []
 887         for num, media_meta in enumerate(medias, start=1):
 888             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 889             if not formats:
 890                 continue
 891             self._sort_formats(formats)
 892
 893             video_id = media_meta.get('externalId')
 894             if not video_id:
 895                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 896
 897             title = media_meta.get('caption')
 898             if not title:
 899                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 900
 901             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 902
 903             images = []
 904             for image in media_meta.get('images', {}).values():
 905                 images.extend(image.values())
 906             if 'image' in media_meta:
 907                 images.append(media_meta['image'])
 908
 909             thumbnails = [{
 910                 'url': image.get('href'),
 911                 'width': int_or_none(image.get('width')),
 912                 'height': int_or_none(image.get('height')),
 913             } for image in images]
 914
 915             entries.append({
 916                 'id': video_id,
 917                 'title': title,
 918                 'thumbnails': thumbnails,
 919                 'duration': duration,
 920                 'timestamp': timestamp,
 921                 'formats': formats,
 922                 'subtitles': subtitles,
 923             })
 924
 925         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 926
 927
 928 class BBCCoUkArticleIE(InfoExtractor):
 929     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 930     IE_NAME = 'bbc.co.uk:article'
 931     IE_DESC = 'BBC articles'
 932
 933     _TEST = {
 934         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 935         'info_dict': {
 936             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 937             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 938             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 939         },
 940         'playlist_count': 4,
 941         'add_ie': ['BBCCoUk'],
 942     }
 943
 944     def _real_extract(self, url):
 945         playlist_id = self._match_id(url)
 946
 947         webpage = self._download_webpage(url, playlist_id)
 948
 949         title = self._og_search_title(webpage)
 950         description = self._og_search_description(webpage).strip()
 951
 952         entries = [self.url_result(programme_url) for programme_url in re.findall(
 953             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 954
 955         return self.playlist_result(entries, playlist_id, title, description)