5cde90c5b23d1f92a7709c766102bb298e63cadb
[ytdl] / youtube_dl / extractor / arte.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..compat import (
8     compat_parse_qs,
9     compat_urllib_parse_urlparse,
10 )
11 from ..utils import (
12     ExtractorError,
13     find_xpath_attr,
14     get_element_by_attribute,
15     int_or_none,
16     NO_DEFAULT,
17     qualities,
18     unified_strdate,
19 )
20
21 # There are different sources of video in arte.tv, the extraction process
22 # is different for each one. The videos usually expire in 7 days, so we can't
23 # add tests.
24
25
26 class ArteTvIE(InfoExtractor):
27     _VALID_URL = r'https?://videos\.arte\.tv/(?P<lang>fr|de|en|es)/.*-(?P<id>.*?)\.html'
28     IE_NAME = 'arte.tv'
29
30     def _real_extract(self, url):
31         mobj = re.match(self._VALID_URL, url)
32         lang = mobj.group('lang')
33         video_id = mobj.group('id')
34
35         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
36         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
37         ref_xml_doc = self._download_xml(
38             ref_xml_url, video_id, note='Downloading metadata')
39         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
40         config_xml_url = config_node.attrib['ref']
41         config = self._download_xml(
42             config_xml_url, video_id, note='Downloading configuration')
43
44         formats = [{
45             'format_id': q.attrib['quality'],
46             # The playpath starts at 'mp4:', if we don't manually
47             # split the url, rtmpdump will incorrectly parse them
48             'url': q.text.split('mp4:', 1)[0],
49             'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
50             'ext': 'flv',
51             'quality': 2 if q.attrib['quality'] == 'hd' else 1,
52         } for q in config.findall('./urls/url')]
53         self._sort_formats(formats)
54
55         title = config.find('.//name').text
56         thumbnail = config.find('.//firstThumbnailUrl').text
57         return {
58             'id': video_id,
59             'title': title,
60             'thumbnail': thumbnail,
61             'formats': formats,
62         }
63
64
65 class ArteTVBaseIE(InfoExtractor):
66     @classmethod
67     def _extract_url_info(cls, url):
68         mobj = re.match(cls._VALID_URL, url)
69         lang = mobj.group('lang')
70         query = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
71         if 'vid' in query:
72             video_id = query['vid'][0]
73         else:
74             # This is not a real id, it can be for example AJT for the news
75             # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
76             video_id = mobj.group('id')
77         return video_id, lang
78
79     def _extract_from_json_url(self, json_url, video_id, lang, title=None):
80         info = self._download_json(json_url, video_id)
81         player_info = info['videoJsonPlayer']
82
83         vsr = player_info['VSR']
84
85         if not vsr:
86             raise ExtractorError(
87                 'Video %s is not available' % player_info.get('VID') or video_id,
88                 expected=True)
89
90         upload_date_str = player_info.get('shootingDate')
91         if not upload_date_str:
92             upload_date_str = (player_info.get('VRA') or player_info.get('VDA') or '').split(' ')[0]
93
94         title = (player_info.get('VTI') or title or player_info['VID']).strip()
95         subtitle = player_info.get('VSU', '').strip()
96         if subtitle:
97             title += ' - %s' % subtitle
98
99         info_dict = {
100             'id': player_info['VID'],
101             'title': title,
102             'description': player_info.get('VDE'),
103             'upload_date': unified_strdate(upload_date_str),
104             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
105         }
106         qfunc = qualities(['HQ', 'MQ', 'EQ', 'SQ'])
107
108         LANGS = {
109             'fr': 'F',
110             'de': 'A',
111             'en': 'E[ANG]',
112             'es': 'E[ESP]',
113         }
114
115         langcode = LANGS.get(lang, lang)
116
117         formats = []
118         for format_id, format_dict in vsr.items():
119             f = dict(format_dict)
120             versionCode = f.get('versionCode')
121             l = re.escape(langcode)
122
123             # Language preference from most to least priority
124             # Reference: section 5.6.3 of
125             # http://www.arte.tv/sites/en/corporate/files/complete-technical-guidelines-arte-geie-v1-05.pdf
126             PREFERENCES = (
127                 # original version in requested language, without subtitles
128                 r'VO{0}$'.format(l),
129                 # original version in requested language, with partial subtitles in requested language
130                 r'VO{0}-ST{0}$'.format(l),
131                 # original version in requested language, with subtitles for the deaf and hard-of-hearing in requested language
132                 r'VO{0}-STM{0}$'.format(l),
133                 # non-original (dubbed) version in requested language, without subtitles
134                 r'V{0}$'.format(l),
135                 # non-original (dubbed) version in requested language, with subtitles partial subtitles in requested language
136                 r'V{0}-ST{0}$'.format(l),
137                 # non-original (dubbed) version in requested language, with subtitles for the deaf and hard-of-hearing in requested language
138                 r'V{0}-STM{0}$'.format(l),
139                 # original version in requested language, with partial subtitles in different language
140                 r'VO{0}-ST(?!{0}).+?$'.format(l),
141                 # original version in requested language, with subtitles for the deaf and hard-of-hearing in different language
142                 r'VO{0}-STM(?!{0}).+?$'.format(l),
143                 # original version in different language, with partial subtitles in requested language
144                 r'VO(?:(?!{0}).+?)?-ST{0}$'.format(l),
145                 # original version in different language, with subtitles for the deaf and hard-of-hearing in requested language
146                 r'VO(?:(?!{0}).+?)?-STM{0}$'.format(l),
147                 # original version in different language, without subtitles
148                 r'VO(?:(?!{0}))?$'.format(l),
149                 # original version in different language, with partial subtitles in different language
150                 r'VO(?:(?!{0}).+?)?-ST(?!{0}).+?$'.format(l),
151                 # original version in different language, with subtitles for the deaf and hard-of-hearing in different language
152                 r'VO(?:(?!{0}).+?)?-STM(?!{0}).+?$'.format(l),
153             )
154
155             for pref, p in enumerate(PREFERENCES):
156                 if re.match(p, versionCode):
157                     lang_pref = len(PREFERENCES) - pref
158                     break
159             else:
160                 lang_pref = -1
161
162             format = {
163                 'format_id': format_id,
164                 'preference': -10 if f.get('videoFormat') == 'M3U8' else None,
165                 'language_preference': lang_pref,
166                 'format_note': '%s, %s' % (f.get('versionCode'), f.get('versionLibelle')),
167                 'width': int_or_none(f.get('width')),
168                 'height': int_or_none(f.get('height')),
169                 'tbr': int_or_none(f.get('bitrate')),
170                 'quality': qfunc(f.get('quality')),
171             }
172
173             if f.get('mediaType') == 'rtmp':
174                 format['url'] = f['streamer']
175                 format['play_path'] = 'mp4:' + f['url']
176                 format['ext'] = 'flv'
177             else:
178                 format['url'] = f['url']
179
180             formats.append(format)
181
182         self._check_formats(formats, video_id)
183         self._sort_formats(formats)
184
185         info_dict['formats'] = formats
186         return info_dict
187
188
189 class ArteTVPlus7IE(ArteTVBaseIE):
190     IE_NAME = 'arte.tv:+7'
191     _VALID_URL = r'https?://(?:(?:www|sites)\.)?arte\.tv/(?:[^/]+/)?(?P<lang>fr|de|en|es)/(?:videos/)?(?:[^/]+/)*(?P<id>[^/?#&]+)'
192
193     _TESTS = [{
194         'url': 'http://www.arte.tv/guide/de/sendungen/XEN/xenius/?vid=055918-015_PLUS7-D',
195         'only_matching': True,
196     }, {
197         'url': 'http://sites.arte.tv/karambolage/de/video/karambolage-22',
198         'only_matching': True,
199     }, {
200         'url': 'http://www.arte.tv/de/videos/048696-000-A/der-kluge-bauch-unser-zweites-gehirn',
201         'only_matching': True,
202     }]
203
204     @classmethod
205     def suitable(cls, url):
206         return False if ArteTVPlaylistIE.suitable(url) else super(ArteTVPlus7IE, cls).suitable(url)
207
208     def _real_extract(self, url):
209         video_id, lang = self._extract_url_info(url)
210         webpage = self._download_webpage(url, video_id)
211         return self._extract_from_webpage(webpage, video_id, lang)
212
213     def _extract_from_webpage(self, webpage, video_id, lang):
214         patterns_templates = (r'arte_vp_url=["\'](.*?%s.*?)["\']', r'data-url=["\']([^"]+%s[^"]+)["\']')
215         ids = (video_id, '')
216         # some pages contain multiple videos (like
217         # http://www.arte.tv/guide/de/sendungen/XEN/xenius/?vid=055918-015_PLUS7-D),
218         # so we first try to look for json URLs that contain the video id from
219         # the 'vid' parameter.
220         patterns = [t % re.escape(_id) for _id in ids for t in patterns_templates]
221         json_url = self._html_search_regex(
222             patterns, webpage, 'json vp url', default=None)
223         if not json_url:
224             def find_iframe_url(webpage, default=NO_DEFAULT):
225                 return self._html_search_regex(
226                     r'<iframe[^>]+src=(["\'])(?P<url>.+\bjson_url=.+?)\1',
227                     webpage, 'iframe url', group='url', default=default)
228
229             iframe_url = find_iframe_url(webpage, None)
230             if not iframe_url:
231                 embed_url = self._html_search_regex(
232                     r'arte_vp_url_oembed=\'([^\']+?)\'', webpage, 'embed url', default=None)
233                 if embed_url:
234                     player = self._download_json(
235                         embed_url, video_id, 'Downloading player page')
236                     iframe_url = find_iframe_url(player['html'])
237             # en and es URLs produce react-based pages with different layout (e.g.
238             # http://www.arte.tv/guide/en/053330-002-A/carnival-italy?zone=world)
239             if not iframe_url:
240                 program = self._search_regex(
241                     r'program\s*:\s*({.+?["\']embed_html["\'].+?}),?\s*\n',
242                     webpage, 'program', default=None)
243                 if program:
244                     embed_html = self._parse_json(program, video_id)
245                     if embed_html:
246                         iframe_url = find_iframe_url(embed_html['embed_html'])
247             if iframe_url:
248                 json_url = compat_parse_qs(
249                     compat_urllib_parse_urlparse(iframe_url).query)['json_url'][0]
250         if json_url:
251             title = self._search_regex(
252                 r'<h3[^>]+title=(["\'])(?P<title>.+?)\1',
253                 webpage, 'title', default=None, group='title')
254             return self._extract_from_json_url(json_url, video_id, lang, title=title)
255         # Different kind of embed URL (e.g.
256         # http://www.arte.tv/magazine/trepalium/fr/episode-0406-replay-trepalium)
257         entries = [
258             self.url_result(url)
259             for _, url in re.findall(r'<iframe[^>]+src=(["\'])(?P<url>.+?)\1', webpage)]
260         return self.playlist_result(entries)
261
262
263 # It also uses the arte_vp_url url from the webpage to extract the information
264 class ArteTVCreativeIE(ArteTVPlus7IE):
265     IE_NAME = 'arte.tv:creative'
266     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de|en|es)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
267
268     _TESTS = [{
269         'url': 'http://creative.arte.tv/fr/episode/osmosis-episode-1',
270         'info_dict': {
271             'id': '057405-001-A',
272             'ext': 'mp4',
273             'title': 'OSMOSIS - N\'AYEZ PLUS PEUR D\'AIMER (1)',
274             'upload_date': '20150716',
275         },
276     }, {
277         'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
278         'playlist_count': 11,
279         'add_ie': ['Youtube'],
280     }, {
281         'url': 'http://creative.arte.tv/de/episode/agentur-amateur-4-der-erste-kunde',
282         'only_matching': True,
283     }]
284
285
286 class ArteTVInfoIE(ArteTVPlus7IE):
287     IE_NAME = 'arte.tv:info'
288     _VALID_URL = r'https?://info\.arte\.tv/(?P<lang>fr|de|en|es)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
289
290     _TESTS = [{
291         'url': 'http://info.arte.tv/fr/service-civique-un-cache-misere',
292         'info_dict': {
293             'id': '067528-000-A',
294             'ext': 'mp4',
295             'title': 'Service civique, un cache misère ?',
296             'upload_date': '20160403',
297         },
298     }]
299
300
301 class ArteTVFutureIE(ArteTVPlus7IE):
302     IE_NAME = 'arte.tv:future'
303     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
304
305     _TESTS = [{
306         'url': 'http://future.arte.tv/fr/info-sciences/les-ecrevisses-aussi-sont-anxieuses',
307         'info_dict': {
308             'id': '050940-028-A',
309             'ext': 'mp4',
310             'title': 'Les écrevisses aussi peuvent être anxieuses',
311             'upload_date': '20140902',
312         },
313     }, {
314         'url': 'http://future.arte.tv/fr/la-science-est-elle-responsable',
315         'only_matching': True,
316     }]
317
318
319 class ArteTVDDCIE(ArteTVPlus7IE):
320     IE_NAME = 'arte.tv:ddc'
321     _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>[^/?#&]+)'
322
323     _TESTS = []
324
325     def _real_extract(self, url):
326         video_id, lang = self._extract_url_info(url)
327         if lang == 'folge':
328             lang = 'de'
329         elif lang == 'emission':
330             lang = 'fr'
331         webpage = self._download_webpage(url, video_id)
332         scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
333         script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
334         javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
335         json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
336         return self._extract_from_json_url(json_url, video_id, lang)
337
338
339 class ArteTVConcertIE(ArteTVPlus7IE):
340     IE_NAME = 'arte.tv:concert'
341     _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
342
343     _TESTS = [{
344         'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
345         'md5': '9ea035b7bd69696b67aa2ccaaa218161',
346         'info_dict': {
347             'id': '186',
348             'ext': 'mp4',
349             'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
350             'upload_date': '20140128',
351             'description': 'md5:486eb08f991552ade77439fe6d82c305',
352         },
353     }]
354
355
356 class ArteTVCinemaIE(ArteTVPlus7IE):
357     IE_NAME = 'arte.tv:cinema'
358     _VALID_URL = r'https?://cinema\.arte\.tv/(?P<lang>fr|de|en|es)/(?P<id>.+)'
359
360     _TESTS = [{
361         'url': 'http://cinema.arte.tv/fr/article/les-ailes-du-desir-de-julia-reck',
362         'md5': 'a5b9dd5575a11d93daf0e3f404f45438',
363         'info_dict': {
364             'id': '062494-000-A',
365             'ext': 'mp4',
366             'title': 'Film lauréat du concours web - "Les ailes du désir" de Julia Reck',
367             'upload_date': '20150807',
368         },
369     }]
370
371
372 class ArteTVMagazineIE(ArteTVPlus7IE):
373     IE_NAME = 'arte.tv:magazine'
374     _VALID_URL = r'https?://(?:www\.)?arte\.tv/magazine/[^/]+/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
375
376     _TESTS = [{
377         # Embedded via <iframe src="http://www.arte.tv/arte_vp/index.php?json_url=..."
378         'url': 'http://www.arte.tv/magazine/trepalium/fr/entretien-avec-le-realisateur-vincent-lannoo-trepalium',
379         'md5': '2a9369bcccf847d1c741e51416299f25',
380         'info_dict': {
381             'id': '065965-000-A',
382             'ext': 'mp4',
383             'title': 'Trepalium - Extrait Ep.01',
384             'upload_date': '20160121',
385         },
386     }, {
387         # Embedded via <iframe src="http://www.arte.tv/guide/fr/embed/054813-004-A/medium"
388         'url': 'http://www.arte.tv/magazine/trepalium/fr/episode-0406-replay-trepalium',
389         'md5': 'fedc64fc7a946110fe311634e79782ca',
390         'info_dict': {
391             'id': '054813-004_PLUS7-F',
392             'ext': 'mp4',
393             'title': 'Trepalium (4/6)',
394             'description': 'md5:10057003c34d54e95350be4f9b05cb40',
395             'upload_date': '20160218',
396         },
397     }, {
398         'url': 'http://www.arte.tv/magazine/metropolis/de/frank-woeste-german-paris-metropolis',
399         'only_matching': True,
400     }]
401
402
403 class ArteTVEmbedIE(ArteTVPlus7IE):
404     IE_NAME = 'arte.tv:embed'
405     _VALID_URL = r'''(?x)
406         http://www\.arte\.tv
407         /(?:playerv2/embed|arte_vp/index)\.php\?json_url=
408         (?P<json_url>
409             http://arte\.tv/papi/tvguide/videos/stream/player/
410             (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
411         )
412     '''
413
414     _TESTS = []
415
416     def _real_extract(self, url):
417         mobj = re.match(self._VALID_URL, url)
418         video_id = mobj.group('id')
419         lang = mobj.group('lang')
420         json_url = mobj.group('json_url')
421         return self._extract_from_json_url(json_url, video_id, lang)
422
423
424 class TheOperaPlatformIE(ArteTVPlus7IE):
425     IE_NAME = 'theoperaplatform'
426     _VALID_URL = r'https?://(?:www\.)?theoperaplatform\.eu/(?P<lang>fr|de|en|es)/(?P<id>[^/?#&]+)'
427
428     _TESTS = [{
429         'url': 'http://www.theoperaplatform.eu/de/opera/verdi-otello',
430         'md5': '970655901fa2e82e04c00b955e9afe7b',
431         'info_dict': {
432             'id': '060338-009-A',
433             'ext': 'mp4',
434             'title': 'Verdi - OTELLO',
435             'upload_date': '20160927',
436         },
437     }]
438
439
440 class ArteTVPlaylistIE(ArteTVBaseIE):
441     IE_NAME = 'arte.tv:playlist'
442     _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de|en|es)/[^#]*#collection/(?P<id>PL-\d+)'
443
444     _TESTS = [{
445         'url': 'http://www.arte.tv/guide/de/plus7/?country=DE#collection/PL-013263/ARTETV',
446         'info_dict': {
447             'id': 'PL-013263',
448             'title': 'Areva & Uramin',
449             'description': 'md5:a1dc0312ce357c262259139cfd48c9bf',
450         },
451         'playlist_mincount': 6,
452     }, {
453         'url': 'http://www.arte.tv/guide/de/playlists?country=DE#collection/PL-013190/ARTETV',
454         'only_matching': True,
455     }]
456
457     def _real_extract(self, url):
458         playlist_id, lang = self._extract_url_info(url)
459         collection = self._download_json(
460             'https://api.arte.tv/api/player/v1/collectionData/%s/%s?source=videos'
461             % (lang, playlist_id), playlist_id)
462         title = collection.get('title')
463         description = collection.get('shortDescription') or collection.get('teaserText')
464         entries = [
465             self._extract_from_json_url(
466                 video['jsonUrl'], video.get('programId') or playlist_id, lang)
467             for video in collection['videos'] if video.get('jsonUrl')]
468         return self.playlist_result(entries, playlist_id, title, description)