]> gitweb @ CieloNegro.org - youtube-dl.git/blob - youtube_dl/extractor/bbc.py
Merge pull request #8130 from dyn888/master
[youtube-dl.git] / youtube_dl / extractor / bbc.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..utils import (
8     ExtractorError,
9     float_or_none,
10     int_or_none,
11     parse_duration,
12     parse_iso8601,
13     remove_end,
14     unescapeHTML,
15 )
16 from ..compat import (
17     compat_etree_fromstring,
18     compat_HTTPError,
19 )
20
21
22 class BBCCoUkIE(InfoExtractor):
23     IE_NAME = 'bbc.co.uk'
24     IE_DESC = 'BBC iPlayer'
25     _ID_REGEX = r'[pb][\da-z]{7}'
26     _VALID_URL = r'''(?x)
27                     https?://
28                         (?:www\.)?bbc\.co\.uk/
29                         (?:
30                             programmes/(?!articles/)|
31                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
32                             music/clips[/#]|
33                             radio/player/
34                         )
35                         (?P<id>%s)
36                     ''' % _ID_REGEX
37
38     _MEDIASELECTOR_URLS = [
39         # Provides HQ HLS streams with even better quality that pc mediaset but fails
40         # with geolocation in some cases when it's even not geo restricted at all (e.g.
41         # http://www.bbc.co.uk/programmes/b06bp7lf). Also may fail with selectionunavailable.
42         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/iptv-all/vpid/%s',
43         'http://open.live.bbc.co.uk/mediaselector/5/select/version/2.0/mediaset/pc/vpid/%s',
44     ]
45
46     _MEDIASELECTION_NS = 'http://bbc.co.uk/2008/mp/mediaselection'
47     _EMP_PLAYLIST_NS = 'http://bbc.co.uk/2008/emp/playlist'
48
49     _NAMESPACES = (
50         _MEDIASELECTION_NS,
51         _EMP_PLAYLIST_NS,
52     )
53
54     _TESTS = [
55         {
56             'url': 'http://www.bbc.co.uk/programmes/b039g8p7',
57             'info_dict': {
58                 'id': 'b039d07m',
59                 'ext': 'flv',
60                 'title': 'Leonard Cohen, Kaleidoscope - BBC Radio 4',
61                 'description': 'The Canadian poet and songwriter reflects on his musical career.',
62             },
63             'params': {
64                 # rtmp download
65                 'skip_download': True,
66             }
67         },
68         {
69             'url': 'http://www.bbc.co.uk/iplayer/episode/b00yng5w/The_Man_in_Black_Series_3_The_Printed_Name/',
70             'info_dict': {
71                 'id': 'b00yng1d',
72                 'ext': 'flv',
73                 'title': 'The Man in Black: Series 3: The Printed Name',
74                 'description': "Mark Gatiss introduces Nicholas Pierpan's chilling tale of a writer's devilish pact with a mysterious man. Stars Ewan Bailey.",
75                 'duration': 1800,
76             },
77             'params': {
78                 # rtmp download
79                 'skip_download': True,
80             },
81             'skip': 'Episode is no longer available on BBC iPlayer Radio',
82         },
83         {
84             'url': 'http://www.bbc.co.uk/iplayer/episode/b03vhd1f/The_Voice_UK_Series_3_Blind_Auditions_5/',
85             'info_dict': {
86                 'id': 'b00yng1d',
87                 'ext': 'flv',
88                 'title': 'The Voice UK: Series 3: Blind Auditions 5',
89                 'description': "Emma Willis and Marvin Humes present the fifth set of blind auditions in the singing competition, as the coaches continue to build their teams based on voice alone.",
90                 'duration': 5100,
91             },
92             'params': {
93                 # rtmp download
94                 'skip_download': True,
95             },
96             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
97         },
98         {
99             'url': 'http://www.bbc.co.uk/iplayer/episode/p026c7jt/tomorrows-worlds-the-unearthly-history-of-science-fiction-2-invasion',
100             'info_dict': {
101                 'id': 'b03k3pb7',
102                 'ext': 'flv',
103                 'title': "Tomorrow's Worlds: The Unearthly History of Science Fiction",
104                 'description': '2. Invasion',
105                 'duration': 3600,
106             },
107             'params': {
108                 # rtmp download
109                 'skip_download': True,
110             },
111             'skip': 'Currently BBC iPlayer TV programmes are available to play in the UK only',
112         }, {
113             'url': 'http://www.bbc.co.uk/programmes/b04v20dw',
114             'info_dict': {
115                 'id': 'b04v209v',
116                 'ext': 'flv',
117                 'title': 'Pete Tong, The Essential New Tune Special',
118                 'description': "Pete has a very special mix - all of 2014's Essential New Tunes!",
119                 'duration': 10800,
120             },
121             'params': {
122                 # rtmp download
123                 'skip_download': True,
124             },
125             'skip': 'Episode is no longer available on BBC iPlayer Radio',
126         }, {
127             'url': 'http://www.bbc.co.uk/music/clips/p022h44b',
128             'note': 'Audio',
129             'info_dict': {
130                 'id': 'p022h44j',
131                 'ext': 'flv',
132                 'title': 'BBC Proms Music Guides, Rachmaninov: Symphonic Dances',
133                 'description': "In this Proms Music Guide, Andrew McGregor looks at Rachmaninov's Symphonic Dances.",
134                 'duration': 227,
135             },
136             'params': {
137                 # rtmp download
138                 'skip_download': True,
139             }
140         }, {
141             'url': 'http://www.bbc.co.uk/music/clips/p025c0zz',
142             'note': 'Video',
143             'info_dict': {
144                 'id': 'p025c103',
145                 'ext': 'flv',
146                 'title': 'Reading and Leeds Festival, 2014, Rae Morris - Closer (Live on BBC Three)',
147                 'description': 'Rae Morris performs Closer for BBC Three at Reading 2014',
148                 'duration': 226,
149             },
150             'params': {
151                 # rtmp download
152                 'skip_download': True,
153             }
154         }, {
155             'url': 'http://www.bbc.co.uk/iplayer/episode/b054fn09/ad/natural-world-20152016-2-super-powered-owls',
156             'info_dict': {
157                 'id': 'p02n76xf',
158                 'ext': 'flv',
159                 'title': 'Natural World, 2015-2016: 2. Super Powered Owls',
160                 'description': 'md5:e4db5c937d0e95a7c6b5e654d429183d',
161                 'duration': 3540,
162             },
163             'params': {
164                 # rtmp download
165                 'skip_download': True,
166             },
167             'skip': 'geolocation',
168         }, {
169             'url': 'http://www.bbc.co.uk/iplayer/episode/b05zmgwn/royal-academy-summer-exhibition',
170             'info_dict': {
171                 'id': 'b05zmgw1',
172                 'ext': 'flv',
173                 'description': 'Kirsty Wark and Morgan Quaintance visit the Royal Academy as it prepares for its annual artistic extravaganza, meeting people who have come together to make the show unique.',
174                 'title': 'Royal Academy Summer Exhibition',
175                 'duration': 3540,
176             },
177             'params': {
178                 # rtmp download
179                 'skip_download': True,
180             },
181             'skip': 'geolocation',
182         }, {
183             # iptv-all mediaset fails with geolocation however there is no geo restriction
184             # for this programme at all
185             'url': 'http://www.bbc.co.uk/programmes/b06rkn85',
186             'info_dict': {
187                 'id': 'b06rkms3',
188                 'ext': 'flv',
189                 'title': "Best of the Mini-Mixes 2015: Part 3, Annie Mac's Friday Night - BBC Radio 1",
190                 'description': "Annie has part three in the Best of the Mini-Mixes 2015, plus the year's Most Played!",
191             },
192             'params': {
193                 # rtmp download
194                 'skip_download': True,
195             },
196         }, {
197             'url': 'http://www.bbc.co.uk/iplayer/playlist/p01dvks4',
198             'only_matching': True,
199         }, {
200             'url': 'http://www.bbc.co.uk/music/clips#p02frcc3',
201             'only_matching': True,
202         }, {
203             'url': 'http://www.bbc.co.uk/iplayer/cbeebies/episode/b0480276/bing-14-atchoo',
204             'only_matching': True,
205         }, {
206             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
207             'only_matching': True,
208         }
209     ]
210
211     class MediaSelectionError(Exception):
212         def __init__(self, id):
213             self.id = id
214
215     def _extract_asx_playlist(self, connection, programme_id):
216         asx = self._download_xml(connection.get('href'), programme_id, 'Downloading ASX playlist')
217         return [ref.get('href') for ref in asx.findall('./Entry/ref')]
218
219     def _extract_connection(self, connection, programme_id):
220         formats = []
221         kind = connection.get('kind')
222         protocol = connection.get('protocol')
223         supplier = connection.get('supplier')
224         if protocol == 'http':
225             href = connection.get('href')
226             transfer_format = connection.get('transferFormat')
227             # ASX playlist
228             if supplier == 'asx':
229                 for i, ref in enumerate(self._extract_asx_playlist(connection, programme_id)):
230                     formats.append({
231                         'url': ref,
232                         'format_id': 'ref%s_%s' % (i, supplier),
233                     })
234             # Skip DASH until supported
235             elif transfer_format == 'dash':
236                 pass
237             elif transfer_format == 'hls':
238                 formats.extend(self._extract_m3u8_formats(
239                     href, programme_id, ext='mp4', entry_protocol='m3u8_native',
240                     m3u8_id=supplier, fatal=False))
241             # Direct link
242             else:
243                 formats.append({
244                     'url': href,
245                     'format_id': supplier or kind or protocol,
246                 })
247         elif protocol == 'rtmp':
248             application = connection.get('application', 'ondemand')
249             auth_string = connection.get('authString')
250             identifier = connection.get('identifier')
251             server = connection.get('server')
252             formats.append({
253                 'url': '%s://%s/%s?%s' % (protocol, server, application, auth_string),
254                 'play_path': identifier,
255                 'app': '%s?%s' % (application, auth_string),
256                 'page_url': 'http://www.bbc.co.uk',
257                 'player_url': 'http://www.bbc.co.uk/emp/releases/iplayer/revisions/617463_618125_4/617463_618125_4_emp.swf',
258                 'rtmp_live': False,
259                 'ext': 'flv',
260                 'format_id': supplier,
261             })
262         return formats
263
264     def _extract_items(self, playlist):
265         return playlist.findall('./{%s}item' % self._EMP_PLAYLIST_NS)
266
267     def _findall_ns(self, element, xpath):
268         elements = []
269         for ns in self._NAMESPACES:
270             elements.extend(element.findall(xpath % ns))
271         return elements
272
273     def _extract_medias(self, media_selection):
274         error = media_selection.find('./{%s}error' % self._MEDIASELECTION_NS)
275         if error is None:
276             media_selection.find('./{%s}error' % self._EMP_PLAYLIST_NS)
277         if error is not None:
278             raise BBCCoUkIE.MediaSelectionError(error.get('id'))
279         return self._findall_ns(media_selection, './{%s}media')
280
281     def _extract_connections(self, media):
282         return self._findall_ns(media, './{%s}connection')
283
284     def _extract_video(self, media, programme_id):
285         formats = []
286         vbr = int_or_none(media.get('bitrate'))
287         vcodec = media.get('encoding')
288         service = media.get('service')
289         width = int_or_none(media.get('width'))
290         height = int_or_none(media.get('height'))
291         file_size = int_or_none(media.get('media_file_size'))
292         for connection in self._extract_connections(media):
293             conn_formats = self._extract_connection(connection, programme_id)
294             for format in conn_formats:
295                 format.update({
296                     'width': width,
297                     'height': height,
298                     'vbr': vbr,
299                     'vcodec': vcodec,
300                     'filesize': file_size,
301                 })
302                 if service:
303                     format['format_id'] = '%s_%s' % (service, format['format_id'])
304             formats.extend(conn_formats)
305         return formats
306
307     def _extract_audio(self, media, programme_id):
308         formats = []
309         abr = int_or_none(media.get('bitrate'))
310         acodec = media.get('encoding')
311         service = media.get('service')
312         for connection in self._extract_connections(media):
313             conn_formats = self._extract_connection(connection, programme_id)
314             for format in conn_formats:
315                 format.update({
316                     'format_id': '%s_%s' % (service, format['format_id']),
317                     'abr': abr,
318                     'acodec': acodec,
319                 })
320             formats.extend(conn_formats)
321         return formats
322
323     def _get_subtitles(self, media, programme_id):
324         subtitles = {}
325         for connection in self._extract_connections(media):
326             captions = self._download_xml(connection.get('href'), programme_id, 'Downloading captions')
327             lang = captions.get('{http://www.w3.org/XML/1998/namespace}lang', 'en')
328             subtitles[lang] = [
329                 {
330                     'url': connection.get('href'),
331                     'ext': 'ttml',
332                 },
333             ]
334         return subtitles
335
336     def _raise_extractor_error(self, media_selection_error):
337         raise ExtractorError(
338             '%s returned error: %s' % (self.IE_NAME, media_selection_error.id),
339             expected=True)
340
341     def _download_media_selector(self, programme_id):
342         last_exception = None
343         for mediaselector_url in self._MEDIASELECTOR_URLS:
344             try:
345                 return self._download_media_selector_url(
346                     mediaselector_url % programme_id, programme_id)
347             except BBCCoUkIE.MediaSelectionError as e:
348                 if e.id in ('notukerror', 'geolocation', 'selectionunavailable'):
349                     last_exception = e
350                     continue
351                 self._raise_extractor_error(e)
352         self._raise_extractor_error(last_exception)
353
354     def _download_media_selector_url(self, url, programme_id=None):
355         try:
356             media_selection = self._download_xml(
357                 url, programme_id, 'Downloading media selection XML')
358         except ExtractorError as ee:
359             if isinstance(ee.cause, compat_HTTPError) and ee.cause.code in (403, 404):
360                 media_selection = compat_etree_fromstring(ee.cause.read().decode('utf-8'))
361             else:
362                 raise
363         return self._process_media_selector(media_selection, programme_id)
364
365     def _process_media_selector(self, media_selection, programme_id):
366         formats = []
367         subtitles = None
368
369         for media in self._extract_medias(media_selection):
370             kind = media.get('kind')
371             if kind == 'audio':
372                 formats.extend(self._extract_audio(media, programme_id))
373             elif kind == 'video':
374                 formats.extend(self._extract_video(media, programme_id))
375             elif kind == 'captions':
376                 subtitles = self.extract_subtitles(media, programme_id)
377         return formats, subtitles
378
379     def _download_playlist(self, playlist_id):
380         try:
381             playlist = self._download_json(
382                 'http://www.bbc.co.uk/programmes/%s/playlist.json' % playlist_id,
383                 playlist_id, 'Downloading playlist JSON')
384
385             version = playlist.get('defaultAvailableVersion')
386             if version:
387                 smp_config = version['smpConfig']
388                 title = smp_config['title']
389                 description = smp_config['summary']
390                 for item in smp_config['items']:
391                     kind = item['kind']
392                     if kind != 'programme' and kind != 'radioProgramme':
393                         continue
394                     programme_id = item.get('vpid')
395                     duration = int_or_none(item.get('duration'))
396                     formats, subtitles = self._download_media_selector(programme_id)
397                 return programme_id, title, description, duration, formats, subtitles
398         except ExtractorError as ee:
399             if not (isinstance(ee.cause, compat_HTTPError) and ee.cause.code == 404):
400                 raise
401
402         # fallback to legacy playlist
403         return self._process_legacy_playlist(playlist_id)
404
405     def _process_legacy_playlist_url(self, url, display_id):
406         playlist = self._download_legacy_playlist_url(url, display_id)
407         return self._extract_from_legacy_playlist(playlist, display_id)
408
409     def _process_legacy_playlist(self, playlist_id):
410         return self._process_legacy_playlist_url(
411             'http://www.bbc.co.uk/iplayer/playlist/%s' % playlist_id, playlist_id)
412
413     def _download_legacy_playlist_url(self, url, playlist_id=None):
414         return self._download_xml(
415             url, playlist_id, 'Downloading legacy playlist XML')
416
417     def _extract_from_legacy_playlist(self, playlist, playlist_id):
418         no_items = playlist.find('./{%s}noItems' % self._EMP_PLAYLIST_NS)
419         if no_items is not None:
420             reason = no_items.get('reason')
421             if reason == 'preAvailability':
422                 msg = 'Episode %s is not yet available' % playlist_id
423             elif reason == 'postAvailability':
424                 msg = 'Episode %s is no longer available' % playlist_id
425             elif reason == 'noMedia':
426                 msg = 'Episode %s is not currently available' % playlist_id
427             else:
428                 msg = 'Episode %s is not available: %s' % (playlist_id, reason)
429             raise ExtractorError(msg, expected=True)
430
431         for item in self._extract_items(playlist):
432             kind = item.get('kind')
433             if kind != 'programme' and kind != 'radioProgramme':
434                 continue
435             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
436             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
437             description = description_el.text if description_el is not None else None
438
439             def get_programme_id(item):
440                 def get_from_attributes(item):
441                     for p in('identifier', 'group'):
442                         value = item.get(p)
443                         if value and re.match(r'^[pb][\da-z]{7}$', value):
444                             return value
445                 get_from_attributes(item)
446                 mediator = item.find('./{%s}mediator' % self._EMP_PLAYLIST_NS)
447                 if mediator is not None:
448                     return get_from_attributes(mediator)
449
450             programme_id = get_programme_id(item)
451             duration = int_or_none(item.get('duration'))
452
453             if programme_id:
454                 formats, subtitles = self._download_media_selector(programme_id)
455             else:
456                 formats, subtitles = self._process_media_selector(item, playlist_id)
457                 programme_id = playlist_id
458
459         return programme_id, title, description, duration, formats, subtitles
460
461     def _real_extract(self, url):
462         group_id = self._match_id(url)
463
464         webpage = self._download_webpage(url, group_id, 'Downloading video page')
465
466         programme_id = None
467         duration = None
468
469         tviplayer = self._search_regex(
470             r'mediator\.bind\(({.+?})\s*,\s*document\.getElementById',
471             webpage, 'player', default=None)
472
473         if tviplayer:
474             player = self._parse_json(tviplayer, group_id).get('player', {})
475             duration = int_or_none(player.get('duration'))
476             programme_id = player.get('vpid')
477
478         if not programme_id:
479             programme_id = self._search_regex(
480                 r'"vpid"\s*:\s*"(%s)"' % self._ID_REGEX, webpage, 'vpid', fatal=False, default=None)
481
482         if programme_id:
483             formats, subtitles = self._download_media_selector(programme_id)
484             title = self._og_search_title(webpage, default=None) or self._html_search_regex(
485                 r'<h2[^>]+id="parent-title"[^>]*>(.+?)</h2>', webpage, 'title')
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         json_ld_info = self._search_json_ld(webpage, playlist_id, default=None)
722         timestamp = json_ld_info.get('timestamp')
723         playlist_title = json_ld_info.get('title')
724         playlist_description = json_ld_info.get('description')
725
726         if not timestamp:
727             timestamp = parse_iso8601(self._search_regex(
728                 [r'<meta[^>]+property="article:published_time"[^>]+content="([^"]+)"',
729                  r'itemprop="datePublished"[^>]+datetime="([^"]+)"',
730                  r'"datePublished":\s*"([^"]+)'],
731                 webpage, 'date', default=None))
732
733         entries = []
734
735         # article with multiple videos embedded with playlist.sxml (e.g.
736         # http://www.bbc.com/sport/0/football/34475836)
737         playlists = re.findall(r'<param[^>]+name="playlist"[^>]+value="([^"]+)"', webpage)
738         playlists.extend(re.findall(r'data-media-id="([^"]+/playlist\.sxml)"', webpage))
739         if playlists:
740             entries = [
741                 self._extract_from_playlist_sxml(playlist_url, playlist_id, timestamp)
742                 for playlist_url in playlists]
743
744         # news article with multiple videos embedded with data-playable
745         data_playables = re.findall(r'data-playable=(["\'])({.+?})\1', webpage)
746         if data_playables:
747             for _, data_playable_json in data_playables:
748                 data_playable = self._parse_json(
749                     unescapeHTML(data_playable_json), playlist_id, fatal=False)
750                 if not data_playable:
751                     continue
752                 settings = data_playable.get('settings', {})
753                 if settings:
754                     # data-playable with video vpid in settings.playlistObject.items (e.g.
755                     # http://www.bbc.com/news/world-us-canada-34473351)
756                     playlist_object = settings.get('playlistObject', {})
757                     if playlist_object:
758                         items = playlist_object.get('items')
759                         if items and isinstance(items, list):
760                             title = playlist_object['title']
761                             description = playlist_object.get('summary')
762                             duration = int_or_none(items[0].get('duration'))
763                             programme_id = items[0].get('vpid')
764                             formats, subtitles = self._download_media_selector(programme_id)
765                             self._sort_formats(formats)
766                             entries.append({
767                                 'id': programme_id,
768                                 'title': title,
769                                 'description': description,
770                                 'timestamp': timestamp,
771                                 'duration': duration,
772                                 'formats': formats,
773                                 'subtitles': subtitles,
774                             })
775                     else:
776                         # data-playable without vpid but with a playlist.sxml URLs
777                         # in otherSettings.playlist (e.g.
778                         # http://www.bbc.com/turkce/multimedya/2015/10/151010_vid_ankara_patlama_ani)
779                         playlist = data_playable.get('otherSettings', {}).get('playlist', {})
780                         if playlist:
781                             entries.append(self._extract_from_playlist_sxml(
782                                 playlist.get('progressiveDownloadUrl'), playlist_id, timestamp))
783
784         if entries:
785             playlist_title = playlist_title or remove_end(self._og_search_title(webpage), ' - BBC News')
786             playlist_description = playlist_description or self._og_search_description(webpage, default=None)
787             return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
788
789         # single video story (e.g. http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret)
790         programme_id = self._search_regex(
791             [r'data-video-player-vpid="(%s)"' % self._ID_REGEX,
792              r'<param[^>]+name="externalIdentifier"[^>]+value="(%s)"' % self._ID_REGEX,
793              r'videoId\s*:\s*["\'](%s)["\']' % self._ID_REGEX],
794             webpage, 'vpid', default=None)
795
796         if programme_id:
797             formats, subtitles = self._download_media_selector(programme_id)
798             self._sort_formats(formats)
799             # digitalData may be missing (e.g. http://www.bbc.com/autos/story/20130513-hyundais-rock-star)
800             digital_data = self._parse_json(
801                 self._search_regex(
802                     r'var\s+digitalData\s*=\s*({.+?});?\n', webpage, 'digital data', default='{}'),
803                 programme_id, fatal=False)
804             page_info = digital_data.get('page', {}).get('pageInfo', {})
805             title = page_info.get('pageName') or self._og_search_title(webpage)
806             description = page_info.get('description') or self._og_search_description(webpage)
807             timestamp = parse_iso8601(page_info.get('publicationDate')) or timestamp
808             return {
809                 'id': programme_id,
810                 'title': title,
811                 'description': description,
812                 'timestamp': timestamp,
813                 'formats': formats,
814                 'subtitles': subtitles,
815             }
816
817         playlist_title = self._html_search_regex(
818             r'<title>(.*?)(?:\s*-\s*BBC [^ ]+)?</title>', webpage, 'playlist title')
819         playlist_description = self._og_search_description(webpage, default=None)
820
821         def extract_all(pattern):
822             return list(filter(None, map(
823                 lambda s: self._parse_json(s, playlist_id, fatal=False),
824                 re.findall(pattern, webpage))))
825
826         # Multiple video article (e.g.
827         # http://www.bbc.co.uk/blogs/adamcurtis/entries/3662a707-0af9-3149-963f-47bea720b460)
828         EMBED_URL = r'https?://(?:www\.)?bbc\.co\.uk/(?:[^/]+/)+%s(?:\b[^"]+)?' % self._ID_REGEX
829         entries = []
830         for match in extract_all(r'new\s+SMP\(({.+?})\)'):
831             embed_url = match.get('playerSettings', {}).get('externalEmbedUrl')
832             if embed_url and re.match(EMBED_URL, embed_url):
833                 entries.append(embed_url)
834         entries.extend(re.findall(
835             r'setPlaylist\("(%s)"\)' % EMBED_URL, webpage))
836         if entries:
837             return self.playlist_result(
838                 [self.url_result(entry, 'BBCCoUk') for entry in entries],
839                 playlist_id, playlist_title, playlist_description)
840
841         # Multiple video article (e.g. http://www.bbc.com/news/world-europe-32668511)
842         medias = extract_all(r"data-media-meta='({[^']+})'")
843
844         if not medias:
845             # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international)
846             media_asset = self._search_regex(
847                 r'mediaAssetPage\.init\(\s*({.+?}), "/',
848                 webpage, 'media asset', default=None)
849             if media_asset:
850                 media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False)
851                 medias = []
852                 for video in media_asset_page.get('videos', {}).values():
853                     medias.extend(video.values())
854
855         if not medias:
856             # Multiple video playlist with single `now playing` entry (e.g.
857             # http://www.bbc.com/news/video_and_audio/must_see/33767813)
858             vxp_playlist = self._parse_json(
859                 self._search_regex(
860                     r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>',
861                     webpage, 'playlist data'),
862                 playlist_id)
863             playlist_medias = []
864             for item in vxp_playlist:
865                 media = item.get('media')
866                 if not media:
867                     continue
868                 playlist_medias.append(media)
869                 # Download single video if found media with asset id matching the video id from URL
870                 if item.get('advert', {}).get('assetId') == playlist_id:
871                     medias = [media]
872                     break
873             # Fallback to the whole playlist
874             if not medias:
875                 medias = playlist_medias
876
877         entries = []
878         for num, media_meta in enumerate(medias, start=1):
879             formats, subtitles = self._extract_from_media_meta(media_meta, playlist_id)
880             if not formats:
881                 continue
882             self._sort_formats(formats)
883
884             video_id = media_meta.get('externalId')
885             if not video_id:
886                 video_id = playlist_id if len(medias) == 1 else '%s-%s' % (playlist_id, num)
887
888             title = media_meta.get('caption')
889             if not title:
890                 title = playlist_title if len(medias) == 1 else '%s - Video %s' % (playlist_title, num)
891
892             duration = int_or_none(media_meta.get('durationInSeconds')) or parse_duration(media_meta.get('duration'))
893
894             images = []
895             for image in media_meta.get('images', {}).values():
896                 images.extend(image.values())
897             if 'image' in media_meta:
898                 images.append(media_meta['image'])
899
900             thumbnails = [{
901                 'url': image.get('href'),
902                 'width': int_or_none(image.get('width')),
903                 'height': int_or_none(image.get('height')),
904             } for image in images]
905
906             entries.append({
907                 'id': video_id,
908                 'title': title,
909                 'thumbnails': thumbnails,
910                 'duration': duration,
911                 'timestamp': timestamp,
912                 'formats': formats,
913                 'subtitles': subtitles,
914             })
915
916         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
917
918
919 class BBCCoUkArticleIE(InfoExtractor):
920     _VALID_URL = 'http://www.bbc.co.uk/programmes/articles/(?P<id>[a-zA-Z0-9]+)'
921     IE_NAME = 'bbc.co.uk:article'
922     IE_DESC = 'BBC articles'
923
924     _TEST = {
925         'url': 'http://www.bbc.co.uk/programmes/articles/3jNQLTMrPlYGTBn0WV6M2MS/not-your-typical-role-model-ada-lovelace-the-19th-century-programmer',
926         'info_dict': {
927             'id': '3jNQLTMrPlYGTBn0WV6M2MS',
928             'title': 'Calculating Ada: The Countess of Computing - Not your typical role model: Ada Lovelace the 19th century programmer - BBC Four',
929             'description': 'Hannah Fry reveals some of her surprising discoveries about Ada Lovelace during filming.',
930         },
931         'playlist_count': 4,
932         'add_ie': ['BBCCoUk'],
933     }
934
935     def _real_extract(self, url):
936         playlist_id = self._match_id(url)
937
938         webpage = self._download_webpage(url, playlist_id)
939
940         title = self._og_search_title(webpage)
941         description = self._og_search_description(webpage).strip()
942
943         entries = [self.url_result(programme_url) for programme_url in re.findall(
944             r'<div[^>]+typeof="Clip"[^>]+resource="([^"]+)"', webpage)]
945
946         return self.playlist_result(entries, playlist_id, title, description)