_ 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     unescapeHTML,
  14 )
  15 from ..compat import (
  16     compat_etree_fromstring,
  17     compat_HTTPError,
  18 )
  19
  20
  21 class BBCCoUkIE(InfoExtractor):
  22     IE_NAME = 'bbc.co.uk'
  23     IE_DESC = 'BBC iPlayer'
  24     _ID_REGEX = r'[pb][\da-z]{7}'
  25     _VALID_URL = r'''(?x)
  26                     https?://
  27                         (?:www\.)?bbc\.co\.uk/
  28                         (?:
  29                             programmes/(?!articles/)|
  30                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
  31                             music/clips[/#]|
  32                             radio/player/
  33                         )
  34                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
  35                     ''' % _ID_REGEX
  36
  37     _MEDIASELECTOR_URLS = [
  38         # Provides HQ HLS streams with even better quality that pc mediaset but fails
  39         # with geolocation in some cases when it's even not geo restricted at all (e.g.
  40         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
  41         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
  42         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
  43     ]
  44
  45     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
  46     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
  47
  48     _NAMESPACES = (
  49         _MEDIASELECTION_NS,
  50         _EMP_PLAYLIST_NS,
  51     )
  52
  53     _TESTS = [
  54         {
  55             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
  56             'info_dict': {
  57                 'id': 'b039d07m',
  58                 'ext': 'flv',
  59                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
  60                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
  61             },
  62             'params': {
  63                 # rtmp download
  64                 'skip_download': True,
  65             }
  66         },
  67         {
  68             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
  69             'info_dict': {
  70                 'id': 'b00yng1d',
  71                 'ext': 'flv',
  72                 'title': 'The Man in Black: Series 3: The Printed Name',
  73                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
  74                 'duration': 1800,
  75             },
  76             'params': {
  77                 # rtmp download
  78                 'skip_download': True,
  79             },
  80             'skip': 'Episode is no longer available on BBC iPlayer Radio',
  81         },
  82         {
  83             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
  84             'info_dict': {
  85                 'id': 'b00yng1d',
  86                 'ext': 'flv',
  87                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
  88                 '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.',
  89                 'duration': 5100,
  90             },
  91             'params': {
  92                 # rtmp download
  93                 'skip_download': True,
  94             },
  95             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
  96         },
  97         {
  98             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
  99             'info_dict': {
 100                 'id': 'b03k3pb7',
 101                 'ext': 'flv',
 102                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
 103                 'description': '2. Invasion',
 104                 'duration': 3600,
 105             },
 106             'params': {
 107                 # rtmp download
 108                 'skip_download': True,
 109             },
 110             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
 111         }, {
 112             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
 113             'info_dict': {
 114                 'id': 'b04v209v',
 115                 'ext': 'flv',
 116                 'title': 'Pete Tong, The Essential New Tune Special',
 117                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
 118                 'duration': 10800,
 119             },
 120             'params': {
 121                 # rtmp download
 122                 'skip_download': True,
 123             },
 124             'skip': 'Episode is no longer available on BBC iPlayer Radio',
 125         }, {
 126             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
 127             'note': 'Audio',
 128             'info_dict': {
 129                 'id': 'p022h44j',
 130                 'ext': 'flv',
 131                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
 132                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
 133                 'duration': 227,
 134             },
 135             'params': {
 136                 # rtmp download
 137                 'skip_download': True,
 138             }
 139         }, {
 140             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
 141             'note': 'Video',
 142             'info_dict': {
 143                 'id': 'p025c103',
 144                 'ext': 'flv',
 145                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
 146                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
 147                 'duration': 226,
 148             },
 149             'params': {
 150                 # rtmp download
 151                 'skip_download': True,
 152             }
 153         }, {
 154             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
 155             'info_dict': {
 156                 'id': 'p02n76xf',
 157                 'ext': 'flv',
 158                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
 159                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
 160                 'duration': 3540,
 161             },
 162             'params': {
 163                 # rtmp download
 164                 'skip_download': True,
 165             },
 166             'skip': 'geolocation',
 167         }, {
 168             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
 169             'info_dict': {
 170                 'id': 'b05zmgw1',
 171                 'ext': 'flv',
 172                 '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.',
 173                 'title': 'Royal Academy Summer Exhibition',
 174                 'duration': 3540,
 175             },
 176             'params': {
 177                 # rtmp download
 178                 'skip_download': True,
 179             },
 180             'skip': 'geolocation',
 181         }, {
 182             # iptv-all mediaset fails with geolocation however there is no geo restriction
 183             # for this programme at all
 184             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
 185             'info_dict': {
 186                 'id': 'b06rkms3',
 187                 'ext': 'flv',
 188                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
 189                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
 190             },
 191             'params': {
 192                 # rtmp download
 193                 'skip_download': True,
 194             },
 195             'skip': 'Now it\'s really geo-restricted',
 196         }, {
 197             # compact player (https://github.com/rg3/youtube-dl/issues/8147)
 198             'url': 'http://www.bbc.co.uk/programmes/p028bfkf/player',
 199             'info_dict': {
 200                 'id': 'p028bfkj',
 201                 'ext': 'flv',
 202                 'title': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 203                 'description': 'Extract from BBC documentary Look Stranger - Giant Leeks and Magic Brews',
 204             },
 205             'params': {
 206                 # rtmp download
 207                 'skip_download': True,
 208             },
 209         }, {
 210             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
 211             'only_matching': True,
 212         }, {
 213             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
 214             'only_matching': True,
 215         }, {
 216             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
 217             'only_matching': True,
 218         }, {
 219             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
 220             'only_matching': True,
 221         }
 222     ]
 223
 224     class MediaSelectionError(Exception):
 225         def __init__(self, id):
 226             self.id = id
 227
 228     def _extract_asx_playlist(self, connection, programme_id):
 229         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
 230         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
 231
 232     def _extract_items(self, playlist):
 233         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
 234
 235     def _findall_ns(self, element, xpath):
 236         elements = []
 237         for ns in self._NAMESPACES:
 238             elements.extend(element.findall(xpath % ns))
 239         return elements
 240
 241     def _extract_medias(self, media_selection):
 242         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
 243         if error is None:
 244             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
 245         if error is not None:
 246             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
 247         return self._findall_ns(media_selection, './{%s}media')
 248
 249     def _extract_connections(self, media):
 250         return self._findall_ns(media, './{%s}connection')
 251
 252     def _get_subtitles(self, media, programme_id):
 253         subtitles = {}
 254         for connection in self._extract_connections(media):
 255             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
 256             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
 257             subtitles[lang] = [
 258                 {
 259                     'url': connection.get('href'),
 260                     'ext': 'ttml',
 261                 },
 262             ]
 263         return subtitles
 264
 265     def _raise_extractor_error(self, media_selection_error):
 266         raise ExtractorError(
 267             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
 268             expected=True)
 269
 270     def _download_media_selector(self, programme_id):
 271         last_exception = None
 272         for mediaselector_url in self._MEDIASELECTOR_URLS:
 273             try:
 274                 return self._download_media_selector_url(
 275                     mediaselector_url % programme_id, programme_id)
 276             except BBCCoUkIE.MediaSelectionError as e:
 277                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
 278                     last_exception = e
 279                     continue
 280                 self._raise_extractor_error(e)
 281         self._raise_extractor_error(last_exception)
 282
 283     def _download_media_selector_url(self, url, programme_id=None):
 284         try:
 285             media_selection = self._download_xml(
 286                 url, programme_id, 'Downloading media selection XML')
 287         except ExtractorError as ee:
 288             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
 289                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
 290             else:
 291                 raise
 292         return self._process_media_selector(media_selection, programme_id)
 293
 294     def _process_media_selector(self, media_selection, programme_id):
 295         formats = []
 296         subtitles = None
 297
 298         for media in self._extract_medias(media_selection):
 299             kind = media.get('kind')
 300             if kind in ('video', 'audio'):
 301                 bitrate = int_or_none(media.get('bitrate'))
 302                 encoding = media.get('encoding')
 303                 service = media.get('service')
 304                 width = int_or_none(media.get('width'))
 305                 height = int_or_none(media.get('height'))
 306                 file_size = int_or_none(media.get('media_file_size'))
 307                 for connection in self._extract_connections(media):
 308                     conn_kind = connection.get('kind')
 309                     protocol = connection.get('protocol')
 310                     supplier = connection.get('supplier')
 311                     href = connection.get('href')
 312                     transfer_format = connection.get('transferFormat')
 313                     format_id = supplier or conn_kind or protocol
 314                     if service:
 315                         format_id = '%s_%s' % (service, format_id)
 316                     # ASX playlist
 317                     if supplier == 'asx':
 318                         for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
 319                             formats.append({
 320                                 'url': ref,
 321                                 'format_id': 'ref%s_%s' % (i, format_id),
 322                             })
 323                     elif transfer_format == 'dash':
 324                         formats.extend(self._extract_mpd_formats(
 325                             href, programme_id, mpd_id=format_id, fatal=False))
 326                     elif transfer_format == 'hls':
 327                         formats.extend(self._extract_m3u8_formats(
 328                             href, programme_id, ext='mp4', entry_protocol='m3u8_native',
 329                             m3u8_id=format_id, fatal=False))
 330                     elif transfer_format == 'hds':
 331                         formats.extend(self._extract_f4m_formats(
 332                             href, programme_id, f4m_id=format_id, fatal=False))
 333                     else:
 334                         fmt = {
 335                             'format_id': format_id,
 336                             'filesize': file_size,
 337                         }
 338                         if kind == 'video':
 339                             fmt.update({
 340                                 'width': width,
 341                                 'height': height,
 342                                 'vbr': bitrate,
 343                                 'vcodec': encoding,
 344                             })
 345                         else:
 346                             fmt.update({
 347                                 'abr': bitrate,
 348                                 'acodec': encoding,
 349                                 'vcodec': 'none',
 350                             })
 351                         if protocol == 'http':
 352                             # Direct link
 353                             fmt.update({
 354                                 'url': href,
 355                             })
 356                         elif protocol == 'rtmp':
 357                             application = connection.get('application', 'ondemand')
 358                             auth_string = connection.get('authString')
 359                             identifier = connection.get('identifier')
 360                             server = connection.get('server')
 361                             fmt.update({
 362                                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
 363                                 'play_path': identifier,
 364                                 'app': '%s?%s' % (application, auth_string),
 365                                 'page_url': 'http://www.bbc.co.uk',
 366                                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
 367                                 'rtmp_live': False,
 368                                 'ext': 'flv',
 369                             })
 370                         formats.append(fmt)
 371             elif kind == 'captions':
 372                 subtitles = self.extract_subtitles(media, programme_id)
 373         return formats, subtitles
 374
 375     def _download_playlist(self, playlist_id):
 376         try:
 377             playlist = self._download_json(
 378                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
 379                 playlist_id, 'Downloading playlist JSON')
 380
 381             version = playlist.get('defaultAvailableVersion')
 382             if version:
 383                 smp_config = version['smpConfig']
 384                 title = smp_config['title']
 385                 description = smp_config['summary']
 386                 for item in smp_config['items']:
 387                     kind = item['kind']
 388                     if kind != 'programme' and kind != 'radioProgramme':
 389                         continue
 390                     programme_id = item.get('vpid')
 391                     duration = int_or_none(item.get('duration'))
 392                     formats, subtitles = self._download_media_selector(programme_id)
 393                 return programme_id, title, description, duration, formats, subtitles
 394         except ExtractorError as ee:
 395             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
 396                 raise
 397
 398         # fallback to legacy playlist
 399         return self._process_legacy_playlist(playlist_id)
 400
 401     def _process_legacy_playlist_url(self, url, display_id):
 402         playlist = self._download_legacy_playlist_url(url, display_id)
 403         return self._extract_from_legacy_playlist(playlist, display_id)
 404
 405     def _process_legacy_playlist(self, playlist_id):
 406         return self._process_legacy_playlist_url(
 407             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
 408
 409     def _download_legacy_playlist_url(self, url, playlist_id=None):
 410         return self._download_xml(
 411             url, playlist_id, 'Downloading legacy playlist XML')
 412
 413     def _extract_from_legacy_playlist(self, playlist, playlist_id):
 414         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
 415         if no_items is not None:
 416             reason = no_items.get('reason')
 417             if reason == 'preAvailability':
 418                 msg = 'Episode %s is not yet available' % playlist_id
 419             elif reason == 'postAvailability':
 420                 msg = 'Episode %s is no longer available' % playlist_id
 421             elif reason == 'noMedia':
 422                 msg = 'Episode %s is not currently available' % playlist_id
 423             else:
 424                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
 425             raise ExtractorError(msg, expected=True)
 426
 427         for item in self._extract_items(playlist):
 428             kind = item.get('kind')
 429             if kind != 'programme' and kind != 'radioProgramme':
 430                 continue
 431             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
 432             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
 433             description = description_el.text if description_el is not None else None
 434
 435             def get_programme_id(item):
 436                 def get_from_attributes(item):
 437                     for p in('identifier', 'group'):
 438                         value = item.get(p)
 439                         if value and re.match(r'^[pb][\da-z]{7}$', value):
 440                             return value
 441                 get_from_attributes(item)
 442                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
 443                 if mediator is not None:
 444                     return get_from_attributes(mediator)
 445
 446             programme_id = get_programme_id(item)
 447             duration = int_or_none(item.get('duration'))
 448
 449             if programme_id:
 450                 formats, subtitles = self._download_media_selector(programme_id)
 451             else:
 452                 formats, subtitles = self._process_media_selector(item, playlist_id)
 453                 programme_id = playlist_id
 454
 455         return programme_id, title, description, duration, formats, subtitles
 456
 457     def _real_extract(self, url):
 458         group_id = self._match_id(url)
 459
 460         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 461
 462         programme_id = None
 463         duration = None
 464
 465         tviplayer = self._search_regex(
 466             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
 467             webpage, 'player', default=None)
 468
 469         if tviplayer:
 470             player = self._parse_json(tviplayer, group_id).get('player', {})
 471             duration = int_or_none(player.get('duration'))
 472             programme_id = player.get('vpid')
 473
 474         if not programme_id:
 475             programme_id = self._search_regex(
 476                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
 477
 478         if programme_id:
 479             formats, subtitles = self._download_media_selector(programme_id)
 480             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 481                 (r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>',
 482                  r'<div[^>]+class="info"[^>]*>\s*<h1>(.+?)</h1>'), webpage, 'title')
 483             description = self._search_regex(
 484                 (r'<p class="[^"]*medium-description[^"]*">([^<]+)</p>',
 485                  r'<div[^>]+class="info_+synopsis"[^>]*>([^<]+)</div>'),
 486                 webpage, 'description', default=None)
 487             if not description:
 488                 description = self._html_search_meta('description', webpage)
 489         else:
 490             programme_id, title, description, duration, formats, subtitles = self._download_playlist(group_id)
 491
 492         self._sort_formats(formats)
 493
 494         return {
 495             'id': programme_id,
 496             'title': title,
 497             'description': description,
 498             'thumbnail': self._og_search_thumbnail(webpage, default=None),
 499             'duration': duration,
 500             'formats': formats,
 501             'subtitles': subtitles,
 502         }
 503
 504
 505 class BBCIE(BBCCoUkIE):
 506     IE_NAME = 'bbc'
 507     IE_DESC = 'BBC'
 508     _VALID_URL = r'https?://(?:www\.)?bbc\.(?:com|co\.uk)/(?:[^/]+/)+(?P<id>[^/#?]+)'
 509
 510     _MEDIASELECTOR_URLS = [
 511         # Provides HQ HLS streams but fails with geolocation in some cases when it's
 512         # even not geo restricted at all
 513         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
 514         # Provides more formats, namely direct mp4 links, but fails on some videos with
 515         # notukerror for non UK (?) users (e.g.
 516         # http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 517         'http://open.live.bbc.co.uk/mediaselector/4/mtis/stream/%s',
 518         # Provides fewer formats, but works everywhere for everybody (hopefully)
 519         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/journalism-pc/vpid/%s',
 520     ]
 521
 522     _TESTS = [{
 523         # article with multiple videos embedded with data-playable containing vpids
 524         'url': 'http://www.bbc.com/news/world-europe-32668511',
 525         'info_dict': {
 526             'id': 'world-europe-32668511',
 527             'title': 'Russia stages massive WW2 parade despite Western boycott',
 528             'description': 'md5:00ff61976f6081841f759a08bf78cc9c',
 529         },
 530         'playlist_count': 2,
 531     }, {
 532         # article with multiple videos embedded with data-playable (more videos)
 533         'url': 'http://www.bbc.com/news/business-28299555',
 534         'info_dict': {
 535             'id': 'business-28299555',
 536             'title': 'Farnborough Airshow: Video highlights',
 537             'description': 'BBC reports and video highlights at the Farnborough Airshow.',
 538         },
 539         'playlist_count': 9,
 540         'skip': 'Save time',
 541     }, {
 542         # article with multiple videos embedded with `new SMP()`
 543         # broken
 544         'url': 'http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460',
 545         'info_dict': {
 546             'id': '3662a707-0af9-3149-963f-47bea720b460',
 547             'title': 'BUGGER',
 548         },
 549         'playlist_count': 18,
 550     }, {
 551         # single video embedded with data-playable containing vpid
 552         'url': 'http://www.bbc.com/news/world-europe-32041533',
 553         'info_dict': {
 554             'id': 'p02mprgb',
 555             'ext': 'mp4',
 556             'title': 'Aerial footage showed the site of the crash in the Alps - courtesy BFM TV',
 557             'description': 'md5:2868290467291b37feda7863f7a83f54',
 558             'duration': 47,
 559             'timestamp': 1427219242,
 560             'upload_date': '20150324',
 561         },
 562         'params': {
 563             # rtmp download
 564             'skip_download': True,
 565         }
 566     }, {
 567         # article with single video embedded with data-playable containing XML playlist
 568         # with direct video links as progressiveDownloadUrl (for now these are extracted)
 569         # and playlist with f4m and m3u8 as streamingUrl
 570         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
 571         'info_dict': {
 572             'id': '150615_telabyad_kentin_cogu',
 573             'ext': 'mp4',
 574             'title': "Tel Abyad'da IŞİD bayrağı indirildi YPG bayrağı çekildi",
 575             'description': 'md5:33a4805a855c9baf7115fcbde57e7025',
 576             'timestamp': 1434397334,
 577             'upload_date': '20150615',
 578         },
 579         'params': {
 580             'skip_download': True,
 581         }
 582     }, {
 583         # single video embedded with data-playable containing XML playlists (regional section)
 584         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
 585         'info_dict': {
 586             'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
 587             'ext': 'mp4',
 588             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
 589             'description': 'md5:1525f17448c4ee262b64b8f0c9ce66c8',
 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': 'Jurgen Klopp: Furious football from a witty and winning coach',
 659             'description': 'Fast-paced football, wit, wisdom and a ready smile - why Liverpool fans should come to love new boss Jurgen Klopp.',
 660         },
 661         'playlist_count': 3,
 662     }, {
 663         # school report article with single video
 664         'url': 'http://www.bbc.co.uk/schoolreport/35744779',
 665         'info_dict': {
 666             'id': '35744779',
 667             'title': 'School which breaks down barriers in Jerusalem',
 668         },
 669         'playlist_count': 1,
 670     }, {
 671         # single video with playlist URL from weather section
 672         'url': 'http://www.bbc.com/weather/features/33601775',
 673         'only_matching': True,
 674     }, {
 675         # custom redirection to www.bbc.com
 676         'url': 'http://www.bbc.co.uk/news/science-environment-33661876',
 677         'only_matching': True,
 678     }, {
 679         # single video article embedded with data-media-vpid
 680         'url': 'http://www.bbc.co.uk/sport/rowing/35908187',
 681         'only_matching': True,
 682     }]
 683
 684     @classmethod
 685     def suitable(cls, url):
 686         EXCLUDE_IE = (BBCCoUkIE, BBCCoUkArticleIE, BBCCoUkIPlayerPlaylistIE, BBCCoUkPlaylistIE)
 687         return (False if any(ie.suitable(url) for ie in EXCLUDE_IE)
 688                 else super(BBCIE, cls).suitable(url))
 689
 690     def _extract_from_media_meta(self, media_meta, video_id):
 691         # Direct links to media in media metadata (e.g.
 692         # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 693         # TODO: there are also f4m and m3u8 streams incorporated in playlist.sxml
 694         source_files = media_meta.get('sourceFiles')
 695         if source_files:
 696             return [{
 697                 'url': f['url'],
 698                 'format_id': format_id,
 699                 'ext': f.get('encoding'),
 700                 'tbr': float_or_none(f.get('bitrate'), 1000),
 701                 'filesize': int_or_none(f.get('filesize')),
 702             } for format_id, f in source_files.items() if f.get('url')], []
 703
 704         programme_id = media_meta.get('externalId')
 705         if programme_id:
 706             return self._download_media_selector(programme_id)
 707
 708         # Process playlist.sxml as legacy playlist
 709         href = media_meta.get('href')
 710         if href:
 711             playlist = self._download_legacy_playlist_url(href)
 712             _, _, _, _, formats, subtitles = self._extract_from_legacy_playlist(playlist, video_id)
 713             return formats, subtitles
 714
 715         return [], []
 716
 717     def _extract_from_playlist_sxml(self, url, playlist_id, timestamp):
 718         programme_id, title, description, duration, formats, subtitles = \
 719             self._process_legacy_playlist_url(url, playlist_id)
 720         self._sort_formats(formats)
 721         return {
 722             'id': programme_id,
 723             'title': title,
 724             'description': description,
 725             'duration': duration,
 726             'timestamp': timestamp,
 727             'formats': formats,
 728             'subtitles': subtitles,
 729         }
 730
 731     def _real_extract(self, url):
 732         playlist_id = self._match_id(url)
 733
 734         webpage = self._download_webpage(url, playlist_id)
 735
 736         json_ld_info = self._search_json_ld(webpage, playlist_id, default=None)
 737         timestamp = json_ld_info.get('timestamp')
 738
 739         playlist_title = json_ld_info.get('title')
 740         if not playlist_title:
 741             playlist_title = self._og_search_title(
 742                 webpage, default=None) or self._html_search_regex(
 743                 r'<title>(.+?)</title>', webpage, 'playlist title', default=None)
 744             if playlist_title:
 745                 playlist_title = re.sub(r'(.+)\s*-\s*BBC.*?$', r'\1', playlist_title).strip()
 746
 747         playlist_description = json_ld_info.get(
 748             'description') or self._og_search_description(webpage, default=None)
 749
 750         if not timestamp:
 751             timestamp = parse_iso8601(self._search_regex(
 752                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
 753                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
 754                  r'"datePublished":\s*"([^"]+)'],
 755                 webpage, 'date', default=None))
 756
 757         entries = []
 758
 759         # article with multiple videos embedded with playlist.sxml (e.g.
 760         # http://www.bbc.com/sport/0/football/34475836)
 761         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
 762         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
 763         if playlists:
 764             entries = [
 765                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
 766                 for playlist_url in playlists]
 767
 768         # news article with multiple videos embedded with data-playable
 769         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
 770         if data_playables:
 771             for _, data_playable_json in data_playables:
 772                 data_playable = self._parse_json(
 773                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
 774                 if not data_playable:
 775                     continue
 776                 settings = data_playable.get('settings', {})
 777                 if settings:
 778                     # data-playable with video vpid in settings.playlistObject.items (e.g.
 779                     # http://www.bbc.com/news/world-us-canada-34473351)
 780                     playlist_object = settings.get('playlistObject', {})
 781                     if playlist_object:
 782                         items = playlist_object.get('items')
 783                         if items and isinstance(items, list):
 784                             title = playlist_object['title']
 785                             description = playlist_object.get('summary')
 786                             duration = int_or_none(items[0].get('duration'))
 787                             programme_id = items[0].get('vpid')
 788                             formats, subtitles = self._download_media_selector(programme_id)
 789                             self._sort_formats(formats)
 790                             entries.append({
 791                                 'id': programme_id,
 792                                 'title': title,
 793                                 'description': description,
 794                                 'timestamp': timestamp,
 795                                 'duration': duration,
 796                                 'formats': formats,
 797                                 'subtitles': subtitles,
 798                             })
 799                     else:
 800                         # data-playable without vpid but with a playlist.sxml URLs
 801                         # in otherSettings.playlist (e.g.
 802                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
 803                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
 804                         if playlist:
 805                             entry = None
 806                             for key in ('streaming', 'progressiveDownload'):
 807                                 playlist_url = playlist.get('%sUrl' % key)
 808                                 if not playlist_url:
 809                                     continue
 810                                 try:
 811                                     info = self._extract_from_playlist_sxml(
 812                                         playlist_url, playlist_id, timestamp)
 813                                     if not entry:
 814                                         entry = info
 815                                     else:
 816                                         entry['title'] = info['title']
 817                                         entry['formats'].extend(info['formats'])
 818                                 except Exception as e:
 819                                     # Some playlist URL may fail with 500, at the same time
 820                                     # the other one may work fine (e.g.
 821                                     # http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu)
 822                                     if isinstance(e.cause, compat_HTTPError) and e.cause.code == 500:
 823                                         continue
 824                                     raise
 825                             if entry:
 826                                 self._sort_formats(entry['formats'])
 827                                 entries.append(entry)
 828
 829         if entries:
 830             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 831
 832         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
 833         programme_id = self._search_regex(
 834             [r'data-(?:video-player|media)-vpid="(%s)"' % self._ID_REGEX,
 835              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
 836              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
 837             webpage, 'vpid', default=None)
 838
 839         if programme_id:
 840             formats, subtitles = self._download_media_selector(programme_id)
 841             self._sort_formats(formats)
 842             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
 843             digital_data = self._parse_json(
 844                 self._search_regex(
 845                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
 846                 programme_id, fatal=False)
 847             page_info = digital_data.get('page', {}).get('pageInfo', {})
 848             title = page_info.get('pageName') or self._og_search_title(webpage)
 849             description = page_info.get('description') or self._og_search_description(webpage)
 850             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
 851             return {
 852                 'id': programme_id,
 853                 'title': title,
 854                 'description': description,
 855                 'timestamp': timestamp,
 856                 'formats': formats,
 857                 'subtitles': subtitles,
 858             }
 859
 860         def extract_all(pattern):
 861             return list(filter(None, map(
 862                 lambda s: self._parse_json(s, playlist_id, fatal=False),
 863                 re.findall(pattern, webpage))))
 864
 865         # Multiple video article (e.g.
 866         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
 867         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
 868         entries = []
 869         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
 870             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
 871             if embed_url and re.match(EMBED_URL, embed_url):
 872                 entries.append(embed_url)
 873         entries.extend(re.findall(
 874             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
 875         if entries:
 876             return self.playlist_result(
 877                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
 878                 playlist_id, playlist_title, playlist_description)
 879
 880         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
 881         medias = extract_all(r"data-media-meta='({[^']+})'")
 882
 883         if not medias:
 884             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
 885             media_asset = self._search_regex(
 886                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
 887                 webpage, 'media asset', default=None)
 888             if media_asset:
 889                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
 890                 medias = []
 891                 for video in media_asset_page.get('videos', {}).values():
 892                     medias.extend(video.values())
 893
 894         if not medias:
 895             # Multiple video playlist with single `now playing` entry (e.g.
 896             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
 897             vxp_playlist = self._parse_json(
 898                 self._search_regex(
 899                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
 900                     webpage, 'playlist data'),
 901                 playlist_id)
 902             playlist_medias = []
 903             for item in vxp_playlist:
 904                 media = item.get('media')
 905                 if not media:
 906                     continue
 907                 playlist_medias.append(media)
 908                 # Download single video if found media with asset id matching the video id from URL
 909                 if item.get('advert', {}).get('assetId') == playlist_id:
 910                     medias = [media]
 911                     break
 912             # Fallback to the whole playlist
 913             if not medias:
 914                 medias = playlist_medias
 915
 916         entries = []
 917         for num, media_meta in enumerate(medias, start=1):
 918             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
 919             if not formats:
 920                 continue
 921             self._sort_formats(formats)
 922
 923             video_id = media_meta.get('externalId')
 924             if not video_id:
 925                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
 926
 927             title = media_meta.get('caption')
 928             if not title:
 929                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
 930
 931             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
 932
 933             images = []
 934             for image in media_meta.get('images', {}).values():
 935                 images.extend(image.values())
 936             if 'image' in media_meta:
 937                 images.append(media_meta['image'])
 938
 939             thumbnails = [{
 940                 'url': image.get('href'),
 941                 'width': int_or_none(image.get('width')),
 942                 'height': int_or_none(image.get('height')),
 943             } for image in images]
 944
 945             entries.append({
 946                 'id': video_id,
 947                 'title': title,
 948                 'thumbnails': thumbnails,
 949                 'duration': duration,
 950                 'timestamp': timestamp,
 951                 'formats': formats,
 952                 'subtitles': subtitles,
 953             })
 954
 955         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 956
 957
 958 class BBCCoUkArticleIE(InfoExtractor):
 959     _VALID_URL = r'https?://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
 960     IE_NAME = 'bbc.co.uk:article'
 961     IE_DESC = 'BBC articles'
 962
 963     _TEST = {
 964         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
 965         'info_dict': {
 966             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
 967             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
 968             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
 969         },
 970         'playlist_count': 4,
 971         'add_ie': ['BBCCoUk'],
 972     }
 973
 974     def _real_extract(self, url):
 975         playlist_id = self._match_id(url)
 976
 977         webpage = self._download_webpage(url, playlist_id)
 978
 979         title = self._og_search_title(webpage)
 980         description = self._og_search_description(webpage).strip()
 981
 982         entries = [self.url_result(programme_url) for programme_url in re.findall(
 983             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
 984
 985         return self.playlist_result(entries, playlist_id, title, description)
 986
 987
 988 class BBCCoUkPlaylistBaseIE(InfoExtractor):
 989     def _real_extract(self, url):
 990         playlist_id = self._match_id(url)
 991
 992         webpage = self._download_webpage(url, playlist_id)
 993
 994         entries = [
 995             self.url_result(self._URL_TEMPLATE % video_id, BBCCoUkIE.ie_key())
 996             for video_id in re.findall(
 997                 self._VIDEO_ID_TEMPLATE % BBCCoUkIE._ID_REGEX, webpage)]
 998
 999         title, description = self._extract_title_and_description(webpage)
1000
1001         return self.playlist_result(entries, playlist_id, title, description)
1002
1003
1004 class BBCCoUkIPlayerPlaylistIE(BBCCoUkPlaylistBaseIE):
1005     IE_NAME = 'bbc.co.uk:iplayer:playlist'
1006     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/iplayer/(?:episodes|group)/(?P<id>%s)' % BBCCoUkIE._ID_REGEX
1007     _URL_TEMPLATE = 'http://www.bbc.co.uk/iplayer/episode/%s'
1008     _VIDEO_ID_TEMPLATE = r'data-ip-id=["\'](%s)'
1009     _TESTS = [{
1010         'url': 'http://www.bbc.co.uk/iplayer/episodes/b05rcz9v',
1011         'info_dict': {
1012             'id': 'b05rcz9v',
1013             'title': 'The Disappearance',
1014             'description': 'French thriller serial about a missing teenager.',
1015         },
1016         'playlist_mincount': 6,
1017         'skip': 'This programme is not currently available on BBC iPlayer',
1018     }, {
1019         # Available for over a year unlike 30 days for most other programmes
1020         'url': 'http://www.bbc.co.uk/iplayer/group/p02tcc32',
1021         'info_dict': {
1022             'id': 'p02tcc32',
1023             'title': 'Bohemian Icons',
1024             'description': 'md5:683e901041b2fe9ba596f2ab04c4dbe7',
1025         },
1026         'playlist_mincount': 10,
1027     }]
1028
1029     def _extract_title_and_description(self, webpage):
1030         title = self._search_regex(r'<h1>([^<]+)</h1>', webpage, 'title', fatal=False)
1031         description = self._search_regex(
1032             r'<p[^>]+class=(["\'])subtitle\1[^>]*>(?P<value>[^<]+)</p>',
1033             webpage, 'description', fatal=False, group='value')
1034         return title, description
1035
1036
1037 class BBCCoUkPlaylistIE(BBCCoUkPlaylistBaseIE):
1038     IE_NAME = 'bbc.co.uk:playlist'
1039     _VALID_URL = r'https?://(?:www\.)?bbc\.co\.uk/programmes/(?P<id>%s)/(?:episodes|broadcasts|clips)' % BBCCoUkIE._ID_REGEX
1040     _URL_TEMPLATE = 'http://www.bbc.co.uk/programmes/%s'
1041     _VIDEO_ID_TEMPLATE = r'data-pid=["\'](%s)'
1042     _TESTS = [{
1043         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1044         'info_dict': {
1045             'id': 'b05rcz9v',
1046             'title': 'The Disappearance - Clips - BBC Four',
1047             'description': 'French thriller serial about a missing teenager.',
1048         },
1049         'playlist_mincount': 7,
1050     }, {
1051         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/broadcasts/2016/06',
1052         'only_matching': True,
1053     }, {
1054         'url': 'http://www.bbc.co.uk/programmes/b05rcz9v/clips',
1055         'only_matching': True,
1056     }, {
1057         'url': 'http://www.bbc.co.uk/programmes/b055jkys/episodes/player',
1058         'only_matching': True,
1059     }]
1060
1061     def _extract_title_and_description(self, webpage):
1062         title = self._og_search_title(webpage, fatal=False)
1063         description = self._og_search_description(webpage)
1064         return title, description