]> gitweb @ CieloNegro.org - youtube-dl.git/blob - youtube_dl/extractor/cbc.py
[ard] Improve _VALID_URL (closes #25134) (#25198)
[youtube-dl.git] / youtube_dl / extractor / cbc.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import hashlib
5 import json
6 import re
7 from xml.sax.saxutils import escape
8
9 from .common import InfoExtractor
10 from ..compat import (
11     compat_str,
12     compat_HTTPError,
13 )
14 from ..utils import (
15     js_to_json,
16     smuggle_url,
17     try_get,
18     xpath_text,
19     xpath_element,
20     xpath_with_ns,
21     find_xpath_attr,
22     orderedSet,
23     parse_duration,
24     parse_iso8601,
25     parse_age_limit,
26     strip_or_none,
27     int_or_none,
28     ExtractorError,
29 )
30
31
32 class CBCIE(InfoExtractor):
33     IE_NAME = 'cbc.ca'
34     _VALID_URL = r'https?://(?:www\.)?cbc\.ca/(?!player/)(?:[^/]+/)+(?P<id>[^/?#]+)'
35     _TESTS = [{
36         # with mediaId
37         'url': 'http://www.cbc.ca/22minutes/videos/clips-season-23/don-cherry-play-offs',
38         'md5': '97e24d09672fc4cf56256d6faa6c25bc',
39         'info_dict': {
40             'id': '2682904050',
41             'ext': 'mp4',
42             'title': 'Don Cherry – All-Stars',
43             'description': 'Don Cherry has a bee in his bonnet about AHL player John Scott because that guy’s got heart.',
44             'timestamp': 1454463000,
45             'upload_date': '20160203',
46             'uploader': 'CBCC-NEW',
47         },
48         'skip': 'Geo-restricted to Canada',
49     }, {
50         # with clipId, feed available via tpfeed.cbc.ca and feed.theplatform.com
51         'url': 'http://www.cbc.ca/22minutes/videos/22-minutes-update/22-minutes-update-episode-4',
52         'md5': '162adfa070274b144f4fdc3c3b8207db',
53         'info_dict': {
54             'id': '2414435309',
55             'ext': 'mp4',
56             'title': '22 Minutes Update: What Not To Wear Quebec',
57             'description': "This week's latest Canadian top political story is What Not To Wear Quebec.",
58             'upload_date': '20131025',
59             'uploader': 'CBCC-NEW',
60             'timestamp': 1382717907,
61         },
62     }, {
63         # with clipId, feed only available via tpfeed.cbc.ca
64         'url': 'http://www.cbc.ca/archives/entry/1978-robin-williams-freestyles-on-90-minutes-live',
65         'md5': '0274a90b51a9b4971fe005c63f592f12',
66         'info_dict': {
67             'id': '2487345465',
68             'ext': 'mp4',
69             'title': 'Robin Williams freestyles on 90 Minutes Live',
70             'description': 'Wacky American comedian Robin Williams shows off his infamous "freestyle" comedic talents while being interviewed on CBC\'s 90 Minutes Live.',
71             'upload_date': '19780210',
72             'uploader': 'CBCC-NEW',
73             'timestamp': 255977160,
74         },
75     }, {
76         # multiple iframes
77         'url': 'http://www.cbc.ca/natureofthings/blog/birds-eye-view-from-vancouvers-burrard-street-bridge-how-we-got-the-shot',
78         'playlist': [{
79             'md5': '377572d0b49c4ce0c9ad77470e0b96b4',
80             'info_dict': {
81                 'id': '2680832926',
82                 'ext': 'mp4',
83                 'title': 'An Eagle\'s-Eye View Off Burrard Bridge',
84                 'description': 'Hercules the eagle flies from Vancouver\'s Burrard Bridge down to a nearby park with a mini-camera strapped to his back.',
85                 'upload_date': '20160201',
86                 'timestamp': 1454342820,
87                 'uploader': 'CBCC-NEW',
88             },
89         }, {
90             'md5': '415a0e3f586113894174dfb31aa5bb1a',
91             'info_dict': {
92                 'id': '2658915080',
93                 'ext': 'mp4',
94                 'title': 'Fly like an eagle!',
95                 'description': 'Eagle equipped with a mini camera flies from the world\'s tallest tower',
96                 'upload_date': '20150315',
97                 'timestamp': 1426443984,
98                 'uploader': 'CBCC-NEW',
99             },
100         }],
101         'skip': 'Geo-restricted to Canada',
102     }, {
103         # multiple CBC.APP.Caffeine.initInstance(...)
104         'url': 'http://www.cbc.ca/news/canada/calgary/dog-indoor-exercise-winter-1.3928238',
105         'info_dict': {
106             'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks',
107             'id': 'dog-indoor-exercise-winter-1.3928238',
108             'description': 'md5:c18552e41726ee95bd75210d1ca9194c',
109         },
110         'playlist_mincount': 6,
111     }]
112
113     @classmethod
114     def suitable(cls, url):
115         return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url)
116
117     def _extract_player_init(self, player_init, display_id):
118         player_info = self._parse_json(player_init, display_id, js_to_json)
119         media_id = player_info.get('mediaId')
120         if not media_id:
121             clip_id = player_info['clipId']
122             feed = self._download_json(
123                 'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
124                 clip_id, fatal=False)
125             if feed:
126                 media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
127             if not media_id:
128                 media_id = self._download_json(
129                     'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
130                     clip_id)['entries'][0]['id'].split('/')[-1]
131         return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
132
133     def _real_extract(self, url):
134         display_id = self._match_id(url)
135         webpage = self._download_webpage(url, display_id)
136         title = self._og_search_title(webpage, default=None) or self._html_search_meta(
137             'twitter:title', webpage, 'title', default=None) or self._html_search_regex(
138                 r'<title>([^<]+)</title>', webpage, 'title', fatal=False)
139         entries = [
140             self._extract_player_init(player_init, display_id)
141             for player_init in re.findall(r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage)]
142         media_ids = []
143         for media_id_re in (
144                 r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"',
145                 r'<div[^>]+\bid=["\']player-(\d+)',
146                 r'guid["\']\s*:\s*["\'](\d+)'):
147             media_ids.extend(re.findall(media_id_re, webpage))
148         entries.extend([
149             self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
150             for media_id in orderedSet(media_ids)])
151         return self.playlist_result(
152             entries, display_id, strip_or_none(title),
153             self._og_search_description(webpage))
154
155
156 class CBCPlayerIE(InfoExtractor):
157     IE_NAME = 'cbc.ca:player'
158     _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P<id>\d+)'
159     _TESTS = [{
160         'url': 'http://www.cbc.ca/player/play/2683190193',
161         'md5': '64d25f841ddf4ddb28a235338af32e2c',
162         'info_dict': {
163             'id': '2683190193',
164             'ext': 'mp4',
165             'title': 'Gerry Runs a Sweat Shop',
166             'description': 'md5:b457e1c01e8ff408d9d801c1c2cd29b0',
167             'timestamp': 1455071400,
168             'upload_date': '20160210',
169             'uploader': 'CBCC-NEW',
170         },
171         'skip': 'Geo-restricted to Canada',
172     }, {
173         # Redirected from http://www.cbc.ca/player/AudioMobile/All%20in%20a%20Weekend%20Montreal/ID/2657632011/
174         'url': 'http://www.cbc.ca/player/play/2657631896',
175         'md5': 'e5e708c34ae6fca156aafe17c43e8b75',
176         'info_dict': {
177             'id': '2657631896',
178             'ext': 'mp3',
179             'title': 'CBC Montreal is organizing its first ever community hackathon!',
180             'description': 'The modern technology we tend to depend on so heavily, is never without it\'s share of hiccups and headaches. Next weekend - CBC Montreal will be getting members of the public for its first Hackathon.',
181             'timestamp': 1425704400,
182             'upload_date': '20150307',
183             'uploader': 'CBCC-NEW',
184         },
185     }, {
186         'url': 'http://www.cbc.ca/player/play/2164402062',
187         'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6',
188         'info_dict': {
189             'id': '2164402062',
190             'ext': 'mp4',
191             'title': 'Cancer survivor four times over',
192             'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.',
193             'timestamp': 1320410746,
194             'upload_date': '20111104',
195             'uploader': 'CBCC-NEW',
196         },
197     }]
198
199     def _real_extract(self, url):
200         video_id = self._match_id(url)
201         return {
202             '_type': 'url_transparent',
203             'ie_key': 'ThePlatform',
204             'url': smuggle_url(
205                 'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/%s?mbr=true&formats=MPEG4,FLV,MP3' % video_id, {
206                     'force_smil_url': True
207                 }),
208             'id': video_id,
209         }
210
211
212 class CBCWatchBaseIE(InfoExtractor):
213     _device_id = None
214     _device_token = None
215     _API_BASE_URL = 'https://api-cbc.cloud.clearleap.com/cloffice/client/'
216     _NS_MAP = {
217         'media': 'http://search.yahoo.com/mrss/',
218         'clearleap': 'http://www.clearleap.com/namespace/clearleap/1.0/',
219     }
220     _GEO_COUNTRIES = ['CA']
221     _LOGIN_URL = 'https://api.loginradius.com/identity/v2/auth/login'
222     _TOKEN_URL = 'https://cloud-api.loginradius.com/sso/jwt/api/token'
223     _API_KEY = '3f4beddd-2061-49b0-ae80-6f1f2ed65b37'
224     _NETRC_MACHINE = 'cbcwatch'
225
226     def _signature(self, email, password):
227         data = json.dumps({
228             'email': email,
229             'password': password,
230         }).encode()
231         headers = {'content-type': 'application/json'}
232         query = {'apikey': self._API_KEY}
233         resp = self._download_json(self._LOGIN_URL, None, data=data, headers=headers, query=query)
234         access_token = resp['access_token']
235
236         # token
237         query = {
238             'access_token': access_token,
239             'apikey': self._API_KEY,
240             'jwtapp': 'jwt',
241         }
242         resp = self._download_json(self._TOKEN_URL, None, headers=headers, query=query)
243         return resp['signature']
244
245     def _call_api(self, path, video_id):
246         url = path if path.startswith('http') else self._API_BASE_URL + path
247         for _ in range(2):
248             try:
249                 result = self._download_xml(url, video_id, headers={
250                     'X-Clearleap-DeviceId': self._device_id,
251                     'X-Clearleap-DeviceToken': self._device_token,
252                 })
253             except ExtractorError as e:
254                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
255                     # Device token has expired, re-acquiring device token
256                     self._register_device()
257                     continue
258                 raise
259         error_message = xpath_text(result, 'userMessage') or xpath_text(result, 'systemMessage')
260         if error_message:
261             raise ExtractorError('%s said: %s' % (self.IE_NAME, error_message))
262         return result
263
264     def _real_initialize(self):
265         if self._valid_device_token():
266             return
267         device = self._downloader.cache.load(
268             'cbcwatch', self._cache_device_key()) or {}
269         self._device_id, self._device_token = device.get('id'), device.get('token')
270         if self._valid_device_token():
271             return
272         self._register_device()
273
274     def _valid_device_token(self):
275         return self._device_id and self._device_token
276
277     def _cache_device_key(self):
278         email, _ = self._get_login_info()
279         return '%s_device' % hashlib.sha256(email.encode()).hexdigest() if email else 'device'
280
281     def _register_device(self):
282         result = self._download_xml(
283             self._API_BASE_URL + 'device/register',
284             None, 'Acquiring device token',
285             data=b'<device><type>web</type></device>')
286         self._device_id = xpath_text(result, 'deviceId', fatal=True)
287         email, password = self._get_login_info()
288         if email and password:
289             signature = self._signature(email, password)
290             data = '<login><token>{0}</token><device><deviceId>{1}</deviceId><type>web</type></device></login>'.format(
291                 escape(signature), escape(self._device_id)).encode()
292             url = self._API_BASE_URL + 'device/login'
293             result = self._download_xml(
294                 url, None, data=data,
295                 headers={'content-type': 'application/xml'})
296             self._device_token = xpath_text(result, 'token', fatal=True)
297         else:
298             self._device_token = xpath_text(result, 'deviceToken', fatal=True)
299         self._downloader.cache.store(
300             'cbcwatch', self._cache_device_key(), {
301                 'id': self._device_id,
302                 'token': self._device_token,
303             })
304
305     def _parse_rss_feed(self, rss):
306         channel = xpath_element(rss, 'channel', fatal=True)
307
308         def _add_ns(path):
309             return xpath_with_ns(path, self._NS_MAP)
310
311         entries = []
312         for item in channel.findall('item'):
313             guid = xpath_text(item, 'guid', fatal=True)
314             title = xpath_text(item, 'title', fatal=True)
315
316             media_group = xpath_element(item, _add_ns('media:group'), fatal=True)
317             content = xpath_element(media_group, _add_ns('media:content'), fatal=True)
318             content_url = content.attrib['url']
319
320             thumbnails = []
321             for thumbnail in media_group.findall(_add_ns('media:thumbnail')):
322                 thumbnail_url = thumbnail.get('url')
323                 if not thumbnail_url:
324                     continue
325                 thumbnails.append({
326                     'id': thumbnail.get('profile'),
327                     'url': thumbnail_url,
328                     'width': int_or_none(thumbnail.get('width')),
329                     'height': int_or_none(thumbnail.get('height')),
330                 })
331
332             timestamp = None
333             release_date = find_xpath_attr(
334                 item, _add_ns('media:credit'), 'role', 'releaseDate')
335             if release_date is not None:
336                 timestamp = parse_iso8601(release_date.text)
337
338             entries.append({
339                 '_type': 'url_transparent',
340                 'url': content_url,
341                 'id': guid,
342                 'title': title,
343                 'description': xpath_text(item, 'description'),
344                 'timestamp': timestamp,
345                 'duration': int_or_none(content.get('duration')),
346                 'age_limit': parse_age_limit(xpath_text(item, _add_ns('media:rating'))),
347                 'episode': xpath_text(item, _add_ns('clearleap:episode')),
348                 'episode_number': int_or_none(xpath_text(item, _add_ns('clearleap:episodeInSeason'))),
349                 'series': xpath_text(item, _add_ns('clearleap:series')),
350                 'season_number': int_or_none(xpath_text(item, _add_ns('clearleap:season'))),
351                 'thumbnails': thumbnails,
352                 'ie_key': 'CBCWatchVideo',
353             })
354
355         return self.playlist_result(
356             entries, xpath_text(channel, 'guid'),
357             xpath_text(channel, 'title'),
358             xpath_text(channel, 'description'))
359
360
361 class CBCWatchVideoIE(CBCWatchBaseIE):
362     IE_NAME = 'cbc.ca:watch:video'
363     _VALID_URL = r'https?://api-cbc\.cloud\.clearleap\.com/cloffice/client/web/play/?\?.*?\bcontentId=(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
364     _TEST = {
365         # geo-restricted to Canada, bypassable
366         'url': 'https://api-cbc.cloud.clearleap.com/cloffice/client/web/play/?contentId=3c84472a-1eea-4dee-9267-2655d5055dcf&categoryId=ebc258f5-ee40-4cca-b66b-ba6bd55b7235',
367         'only_matching': True,
368     }
369
370     def _real_extract(self, url):
371         video_id = self._match_id(url)
372         result = self._call_api(url, video_id)
373
374         m3u8_url = xpath_text(result, 'url', fatal=True)
375         formats = self._extract_m3u8_formats(re.sub(r'/([^/]+)/[^/?]+\.m3u8', r'/\1/\1.m3u8', m3u8_url), video_id, 'mp4', fatal=False)
376         if len(formats) < 2:
377             formats = self._extract_m3u8_formats(m3u8_url, video_id, 'mp4')
378         for f in formats:
379             format_id = f.get('format_id')
380             if format_id.startswith('AAC'):
381                 f['acodec'] = 'aac'
382             elif format_id.startswith('AC3'):
383                 f['acodec'] = 'ac-3'
384         self._sort_formats(formats)
385
386         info = {
387             'id': video_id,
388             'title': video_id,
389             'formats': formats,
390         }
391
392         rss = xpath_element(result, 'rss')
393         if rss:
394             info.update(self._parse_rss_feed(rss)['entries'][0])
395             del info['url']
396             del info['_type']
397             del info['ie_key']
398         return info
399
400
401 class CBCWatchIE(CBCWatchBaseIE):
402     IE_NAME = 'cbc.ca:watch'
403     _VALID_URL = r'https?://(?:gem|watch)\.cbc\.ca/(?:[^/]+/)+(?P<id>[0-9a-f-]+)'
404     _TESTS = [{
405         # geo-restricted to Canada, bypassable
406         'url': 'http://watch.cbc.ca/doc-zone/season-6/customer-disservice/38e815a-009e3ab12e4',
407         'info_dict': {
408             'id': '9673749a-5e77-484c-8b62-a1092a6b5168',
409             'ext': 'mp4',
410             'title': 'Customer (Dis)Service',
411             'description': 'md5:8bdd6913a0fe03d4b2a17ebe169c7c87',
412             'upload_date': '20160219',
413             'timestamp': 1455840000,
414         },
415         'params': {
416             # m3u8 download
417             'skip_download': True,
418             'format': 'bestvideo',
419         },
420     }, {
421         # geo-restricted to Canada, bypassable
422         'url': 'http://watch.cbc.ca/arthur/all/1ed4b385-cd84-49cf-95f0-80f004680057',
423         'info_dict': {
424             'id': '1ed4b385-cd84-49cf-95f0-80f004680057',
425             'title': 'Arthur',
426             'description': 'Arthur, the sweetest 8-year-old aardvark, and his pals solve all kinds of problems with humour, kindness and teamwork.',
427         },
428         'playlist_mincount': 30,
429     }, {
430         'url': 'https://gem.cbc.ca/media/this-hour-has-22-minutes/season-26/episode-20/38e815a-0108c6c6a42',
431         'only_matching': True,
432     }]
433
434     def _real_extract(self, url):
435         video_id = self._match_id(url)
436         rss = self._call_api('web/browse/' + video_id, video_id)
437         return self._parse_rss_feed(rss)
438
439
440 class CBCOlympicsIE(InfoExtractor):
441     IE_NAME = 'cbc.ca:olympics'
442     _VALID_URL = r'https?://olympics\.cbc\.ca/video/[^/]+/(?P<id>[^/?#]+)'
443     _TESTS = [{
444         'url': 'https://olympics.cbc.ca/video/whats-on-tv/olympic-morning-featuring-the-opening-ceremony/',
445         'only_matching': True,
446     }]
447
448     def _real_extract(self, url):
449         display_id = self._match_id(url)
450         webpage = self._download_webpage(url, display_id)
451         video_id = self._hidden_inputs(webpage)['videoId']
452         video_doc = self._download_xml(
453             'https://olympics.cbc.ca/videodata/%s.xml' % video_id, video_id)
454         title = xpath_text(video_doc, 'title', fatal=True)
455         is_live = xpath_text(video_doc, 'kind') == 'Live'
456         if is_live:
457             title = self._live_title(title)
458
459         formats = []
460         for video_source in video_doc.findall('videoSources/videoSource'):
461             uri = xpath_text(video_source, 'uri')
462             if not uri:
463                 continue
464             tokenize = self._download_json(
465                 'https://olympics.cbc.ca/api/api-akamai/tokenize',
466                 video_id, data=json.dumps({
467                     'VideoSource': uri,
468                 }).encode(), headers={
469                     'Content-Type': 'application/json',
470                     'Referer': url,
471                     # d3.VideoPlayer._init in https://olympics.cbc.ca/components/script/base.js
472                     'Cookie': '_dvp=TK:C0ObxjerU',  # AKAMAI CDN cookie
473                 }, fatal=False)
474             if not tokenize:
475                 continue
476             content_url = tokenize['ContentUrl']
477             video_source_format = video_source.get('format')
478             if video_source_format == 'IIS':
479                 formats.extend(self._extract_ism_formats(
480                     content_url, video_id, ism_id=video_source_format, fatal=False))
481             else:
482                 formats.extend(self._extract_m3u8_formats(
483                     content_url, video_id, 'mp4',
484                     'm3u8' if is_live else 'm3u8_native',
485                     m3u8_id=video_source_format, fatal=False))
486         self._sort_formats(formats)
487
488         return {
489             'id': video_id,
490             'display_id': display_id,
491             'title': title,
492             'description': xpath_text(video_doc, 'description'),
493             'thumbnail': xpath_text(video_doc, 'thumbnailUrl'),
494             'duration': parse_duration(xpath_text(video_doc, 'duration')),
495             'formats': formats,
496             'is_live': is_live,
497         }