yt_dlp/extractor/cbc.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import json
   6 import base64
   7 import time
   8
   9 from .common import InfoExtractor
  10 from ..compat import (
  11     compat_str,
  12 )
  13 from ..utils import (
  14     js_to_json,
  15     smuggle_url,
  16     try_get,
  17     orderedSet,
  18     strip_or_none,
  19     ExtractorError,
  20 )
  21
  22
  23 class CBCIE(InfoExtractor):
  24     IE_NAME = 'cbc.ca'
  25     _VALID_URL = r'https?://(?:www\.)?cbc\.ca/(?!player/)(?:[^/]+/)+(?P<id>[^/?#]+)'
  26     _TESTS = [{
  27         # with mediaId
  28         'url': 'http://www.cbc.ca/22minutes/videos/clips-season-23/don-cherry-play-offs',
  29         'md5': '97e24d09672fc4cf56256d6faa6c25bc',
  30         'info_dict': {
  31             'id': '2682904050',
  32             'ext': 'mp4',
  33             'title': 'Don Cherry – All-Stars',
  34             'description': 'Don Cherry has a bee in his bonnet about AHL player John Scott because that guy’s got heart.',
  35             'timestamp': 1454463000,
  36             'upload_date': '20160203',
  37             'uploader': 'CBCC-NEW',
  38         },
  39         'skip': 'Geo-restricted to Canada',
  40     }, {
  41         # with clipId, feed available via tpfeed.cbc.ca and feed.theplatform.com
  42         'url': 'http://www.cbc.ca/22minutes/videos/22-minutes-update/22-minutes-update-episode-4',
  43         'md5': '162adfa070274b144f4fdc3c3b8207db',
  44         'info_dict': {
  45             'id': '2414435309',
  46             'ext': 'mp4',
  47             'title': '22 Minutes Update: What Not To Wear Quebec',
  48             'description': "This week's latest Canadian top political story is What Not To Wear Quebec.",
  49             'upload_date': '20131025',
  50             'uploader': 'CBCC-NEW',
  51             'timestamp': 1382717907,
  52         },
  53         'skip': 'No longer available',
  54     }, {
  55         # with clipId, feed only available via tpfeed.cbc.ca
  56         'url': 'http://www.cbc.ca/archives/entry/1978-robin-williams-freestyles-on-90-minutes-live',
  57         'md5': '0274a90b51a9b4971fe005c63f592f12',
  58         'info_dict': {
  59             'id': '2487345465',
  60             'ext': 'mp4',
  61             'title': 'Robin Williams freestyles on 90 Minutes Live',
  62             'description': 'Wacky American comedian Robin Williams shows off his infamous "freestyle" comedic talents while being interviewed on CBC\'s 90 Minutes Live.',
  63             'upload_date': '19780210',
  64             'uploader': 'CBCC-NEW',
  65             'timestamp': 255977160,
  66         },
  67     }, {
  68         # multiple iframes
  69         'url': 'http://www.cbc.ca/natureofthings/blog/birds-eye-view-from-vancouvers-burrard-street-bridge-how-we-got-the-shot',
  70         'playlist': [{
  71             'md5': '377572d0b49c4ce0c9ad77470e0b96b4',
  72             'info_dict': {
  73                 'id': '2680832926',
  74                 'ext': 'mp4',
  75                 'title': 'An Eagle\'s-Eye View Off Burrard Bridge',
  76                 'description': 'Hercules the eagle flies from Vancouver\'s Burrard Bridge down to a nearby park with a mini-camera strapped to his back.',
  77                 'upload_date': '20160201',
  78                 'timestamp': 1454342820,
  79                 'uploader': 'CBCC-NEW',
  80             },
  81         }, {
  82             'md5': '415a0e3f586113894174dfb31aa5bb1a',
  83             'info_dict': {
  84                 'id': '2658915080',
  85                 'ext': 'mp4',
  86                 'title': 'Fly like an eagle!',
  87                 'description': 'Eagle equipped with a mini camera flies from the world\'s tallest tower',
  88                 'upload_date': '20150315',
  89                 'timestamp': 1426443984,
  90                 'uploader': 'CBCC-NEW',
  91             },
  92         }],
  93         'skip': 'Geo-restricted to Canada',
  94     }, {
  95         # multiple CBC.APP.Caffeine.initInstance(...)
  96         'url': 'http://www.cbc.ca/news/canada/calgary/dog-indoor-exercise-winter-1.3928238',
  97         'info_dict': {
  98             'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks',
  99             'id': 'dog-indoor-exercise-winter-1.3928238',
 100             'description': 'md5:c18552e41726ee95bd75210d1ca9194c',
 101         },
 102         'playlist_mincount': 6,
 103     }]
 104
 105     @classmethod
 106     def suitable(cls, url):
 107         return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url)
 108
 109     def _extract_player_init(self, player_init, display_id):
 110         player_info = self._parse_json(player_init, display_id, js_to_json)
 111         media_id = player_info.get('mediaId')
 112         if not media_id:
 113             clip_id = player_info['clipId']
 114             feed = self._download_json(
 115                 'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
 116                 clip_id, fatal=False)
 117             if feed:
 118                 media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
 119             if not media_id:
 120                 media_id = self._download_json(
 121                     'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
 122                     clip_id)['entries'][0]['id'].split('/')[-1]
 123         return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 124
 125     def _real_extract(self, url):
 126         display_id = self._match_id(url)
 127         webpage = self._download_webpage(url, display_id)
 128         title = self._og_search_title(webpage, default=None) or self._html_search_meta(
 129             'twitter:title', webpage, 'title', default=None) or self._html_search_regex(
 130                 r'<title>([^<]+)</title>', webpage, 'title', fatal=False)
 131         entries = [
 132             self._extract_player_init(player_init, display_id)
 133             for player_init in re.findall(r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage)]
 134         media_ids = []
 135         for media_id_re in (
 136                 r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"',
 137                 r'<div[^>]+\bid=["\']player-(\d+)',
 138                 r'guid["\']\s*:\s*["\'](\d+)'):
 139             media_ids.extend(re.findall(media_id_re, webpage))
 140         entries.extend([
 141             self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 142             for media_id in orderedSet(media_ids)])
 143         return self.playlist_result(
 144             entries, display_id, strip_or_none(title),
 145             self._og_search_description(webpage))
 146
 147
 148 class CBCPlayerIE(InfoExtractor):
 149     IE_NAME = 'cbc.ca:player'
 150     _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P<id>\d+)'
 151     _TESTS = [{
 152         'url': 'http://www.cbc.ca/player/play/2683190193',
 153         'md5': '64d25f841ddf4ddb28a235338af32e2c',
 154         'info_dict': {
 155             'id': '2683190193',
 156             'ext': 'mp4',
 157             'title': 'Gerry Runs a Sweat Shop',
 158             'description': 'md5:b457e1c01e8ff408d9d801c1c2cd29b0',
 159             'timestamp': 1455071400,
 160             'upload_date': '20160210',
 161             'uploader': 'CBCC-NEW',
 162         },
 163         'skip': 'Geo-restricted to Canada',
 164     }, {
 165         # Redirected from http://www.cbc.ca/player/AudioMobile/All%20in%20a%20Weekend%20Montreal/ID/2657632011/
 166         'url': 'http://www.cbc.ca/player/play/2657631896',
 167         'md5': 'e5e708c34ae6fca156aafe17c43e8b75',
 168         'info_dict': {
 169             'id': '2657631896',
 170             'ext': 'mp3',
 171             'title': 'CBC Montreal is organizing its first ever community hackathon!',
 172             '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.',
 173             'timestamp': 1425704400,
 174             'upload_date': '20150307',
 175             'uploader': 'CBCC-NEW',
 176         },
 177     }, {
 178         'url': 'http://www.cbc.ca/player/play/2164402062',
 179         'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6',
 180         'info_dict': {
 181             'id': '2164402062',
 182             'ext': 'mp4',
 183             'title': 'Cancer survivor four times over',
 184             'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.',
 185             'timestamp': 1320410746,
 186             'upload_date': '20111104',
 187             'uploader': 'CBCC-NEW',
 188         },
 189     }]
 190
 191     def _real_extract(self, url):
 192         video_id = self._match_id(url)
 193         return {
 194             '_type': 'url_transparent',
 195             'ie_key': 'ThePlatform',
 196             'url': smuggle_url(
 197                 'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/%s?mbr=true&formats=MPEG4,FLV,MP3' % video_id, {
 198                     'force_smil_url': True
 199                 }),
 200             'id': video_id,
 201         }
 202
 203
 204 class CBCGemIE(InfoExtractor):
 205     IE_NAME = 'gem.cbc.ca'
 206     _VALID_URL = r'https?://gem\.cbc\.ca/media/(?P<id>[0-9a-z-]+/s[0-9]+[a-z][0-9]+)'
 207     _TESTS = [{
 208         # This is a normal, public, TV show video
 209         'url': 'https://gem.cbc.ca/media/schitts-creek/s06e01',
 210         'md5': '93dbb31c74a8e45b378cf13bd3f6f11e',
 211         'info_dict': {
 212             'id': 'schitts-creek/s06e01',
 213             'ext': 'mp4',
 214             'title': 'Smoke Signals',
 215             'description': 'md5:929868d20021c924020641769eb3e7f1',
 216             'thumbnail': 'https://images.radio-canada.ca/v1/synps-cbc/episode/perso/cbc_schitts_creek_season_06e01_thumbnail_v01.jpg?im=Resize=(Size)',
 217             'duration': 1314,
 218             'categories': ['comedy'],
 219             'series': 'Schitt\'s Creek',
 220             'season': 'Season 6',
 221             'season_number': 6,
 222             'episode': 'Smoke Signals',
 223             'episode_number': 1,
 224             'episode_id': 'schitts-creek/s06e01',
 225         },
 226         'params': {'format': 'bv'},
 227         'skip': 'Geo-restricted to Canada',
 228     }, {
 229         # This video requires an account in the browser, but works fine in yt-dlp
 230         'url': 'https://gem.cbc.ca/media/schitts-creek/s01e01',
 231         'md5': '297a9600f554f2258aed01514226a697',
 232         'info_dict': {
 233             'id': 'schitts-creek/s01e01',
 234             'ext': 'mp4',
 235             'title': 'The Cup Runneth Over',
 236             'description': 'md5:9bca14ea49ab808097530eb05a29e797',
 237             'thumbnail': 'https://images.radio-canada.ca/v1/synps-cbc/episode/perso/cbc_schitts_creek_season_01e01_thumbnail_v01.jpg?im=Resize=(Size)',
 238             'series': 'Schitt\'s Creek',
 239             'season_number': 1,
 240             'season': 'Season 1',
 241             'episode_number': 1,
 242             'episode': 'The Cup Runneth Over',
 243             'episode_id': 'schitts-creek/s01e01',
 244             'duration': 1309,
 245             'categories': ['comedy'],
 246         },
 247         'params': {'format': 'bv'},
 248         'skip': 'Geo-restricted to Canada',
 249     }]
 250
 251     _GEO_COUNTRIES = ['CA']
 252     _TOKEN_API_KEY = '3f4beddd-2061-49b0-ae80-6f1f2ed65b37'
 253     _NETRC_MACHINE = 'cbcgem'
 254     _claims_token = None
 255
 256     def _new_claims_token(self, email, password):
 257         data = json.dumps({
 258             'email': email,
 259             'password': password,
 260         }).encode()
 261         headers = {'content-type': 'application/json'}
 262         query = {'apikey': self._TOKEN_API_KEY}
 263         resp = self._download_json('https://api.loginradius.com/identity/v2/auth/login',
 264                                    None, data=data, headers=headers, query=query)
 265         access_token = resp['access_token']
 266
 267         query = {
 268             'access_token': access_token,
 269             'apikey': self._TOKEN_API_KEY,
 270             'jwtapp': 'jwt',
 271         }
 272         resp = self._download_json('https://cloud-api.loginradius.com/sso/jwt/api/token',
 273                                    None, headers=headers, query=query)
 274         sig = resp['signature']
 275
 276         data = json.dumps({'jwt': sig}).encode()
 277         headers = {'content-type': 'application/json', 'ott-device-type': 'web'}
 278         resp = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/token',
 279                                    None, data=data, headers=headers)
 280         cbc_access_token = resp['accessToken']
 281
 282         headers = {'content-type': 'application/json', 'ott-device-type': 'web', 'ott-access-token': cbc_access_token}
 283         resp = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/profile',
 284                                    None, headers=headers)
 285         return resp['claimsToken']
 286
 287     def _get_claims_token_expiry(self):
 288         # Token is a JWT
 289         # JWT is decoded here and 'exp' field is extracted
 290         # It is a Unix timestamp for when the token expires
 291         b64_data = self._claims_token.split('.')[1]
 292         data = base64.urlsafe_b64decode(b64_data + "==")
 293         return json.loads(data)['exp']
 294
 295     def claims_token_expired(self):
 296         exp = self._get_claims_token_expiry()
 297         if exp - time.time() < 10:
 298             # It will expire in less than 10 seconds, or has already expired
 299             return True
 300         return False
 301
 302     def claims_token_valid(self):
 303         return self._claims_token is not None and not self.claims_token_expired()
 304
 305     def _get_claims_token(self, email, password):
 306         if not self.claims_token_valid():
 307             self._claims_token = self._new_claims_token(email, password)
 308             self._downloader.cache.store(self._NETRC_MACHINE, 'claims_token', self._claims_token)
 309         return self._claims_token
 310
 311     def _real_initialize(self):
 312         if self.claims_token_valid():
 313             return
 314         self._claims_token = self._downloader.cache.load(self._NETRC_MACHINE, 'claims_token')
 315
 316     def _real_extract(self, url):
 317         video_id = self._match_id(url)
 318         video_info = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/assets/' + video_id, video_id)
 319
 320         email, password = self._get_login_info()
 321         if email and password:
 322             claims_token = self._get_claims_token(email, password)
 323             headers = {'x-claims-token': claims_token}
 324         else:
 325             headers = {}
 326         m3u8_info = self._download_json(video_info['playSession']['url'], video_id, headers=headers)
 327         m3u8_url = m3u8_info.get('url')
 328
 329         if m3u8_info.get('errorCode') == 1:
 330             self.raise_geo_restricted(countries=['CA'])
 331         elif m3u8_info.get('errorCode') == 35:
 332             self.raise_login_required(method='password')
 333         elif m3u8_info.get('errorCode') != 0:
 334             raise ExtractorError(f'{self.IE_NAME} said: {m3u8_info.get("errorCode")} - {m3u8_info.get("message")}')
 335
 336         formats = self._extract_m3u8_formats(m3u8_url, video_id, m3u8_id='hls')
 337         self._remove_duplicate_formats(formats)
 338
 339         for format in formats:
 340             if format.get('vcodec') == 'none':
 341                 if format.get('ext') is None:
 342                     format['ext'] = 'm4a'
 343                 if format.get('acodec') is None:
 344                     format['acodec'] = 'mp4a.40.2'
 345
 346                 # Put described audio at the beginning of the list, so that it
 347                 # isn't chosen by default, as most people won't want it.
 348                 if 'descriptive' in format['format_id'].lower():
 349                     format['preference'] = -2
 350
 351         self._sort_formats(formats)
 352
 353         return {
 354             'id': video_id,
 355             'title': video_info['title'],
 356             'description': video_info.get('description'),
 357             'thumbnail': video_info.get('image'),
 358             'series': video_info.get('series'),
 359             'season_number': video_info.get('season'),
 360             'season': f'Season {video_info.get("season")}',
 361             'episode_number': video_info.get('episode'),
 362             'episode': video_info.get('title'),
 363             'episode_id': video_id,
 364             'duration': video_info.get('duration'),
 365             'categories': [video_info.get('category')],
 366             'formats': formats,
 367             'release_timestamp': video_info.get('airDate'),
 368             'timestamp': video_info.get('availableDate'),
 369         }
 370
 371
 372 class CBCGemPlaylistIE(InfoExtractor):
 373     IE_NAME = 'gem.cbc.ca:playlist'
 374     _VALID_URL = r'https?://gem\.cbc\.ca/media/(?P<id>(?P<show>[0-9a-z-]+)/s(?P<season>[0-9]+))/?(?:[?#]|$)'
 375     _TESTS = [{
 376         # TV show playlist, all public videos
 377         'url': 'https://gem.cbc.ca/media/schitts-creek/s06',
 378         'playlist_count': 16,
 379         'info_dict': {
 380             'id': 'schitts-creek/s06',
 381             'title': 'Season 6',
 382             'description': 'md5:6a92104a56cbeb5818cc47884d4326a2',
 383         },
 384     }]
 385     _API_BASE = 'https://services.radio-canada.ca/ott/cbc-api/v2/shows/'
 386
 387     def _real_extract(self, url):
 388         match = self._match_valid_url(url)
 389         season_id = match.group('id')
 390         show = match.group('show')
 391         show_info = self._download_json(self._API_BASE + show, season_id)
 392         season = int(match.group('season'))
 393         season_info = try_get(show_info, lambda x: x['seasons'][season - 1])
 394
 395         if season_info is None:
 396             raise ExtractorError(f'Couldn\'t find season {season} of {show}')
 397
 398         episodes = []
 399         for episode in season_info['assets']:
 400             episodes.append({
 401                 '_type': 'url_transparent',
 402                 'ie_key': 'CBCGem',
 403                 'url': 'https://gem.cbc.ca/media/' + episode['id'],
 404                 'id': episode['id'],
 405                 'title': episode.get('title'),
 406                 'description': episode.get('description'),
 407                 'thumbnail': episode.get('image'),
 408                 'series': episode.get('series'),
 409                 'season_number': episode.get('season'),
 410                 'season': season_info['title'],
 411                 'season_id': season_info.get('id'),
 412                 'episode_number': episode.get('episode'),
 413                 'episode': episode.get('title'),
 414                 'episode_id': episode['id'],
 415                 'duration': episode.get('duration'),
 416                 'categories': [episode.get('category')],
 417             })
 418
 419         thumbnail = None
 420         tn_uri = season_info.get('image')
 421         # the-national was observed to use a "data:image/png;base64"
 422         # URI for their 'image' value. The image was 1x1, and is
 423         # probably just a placeholder, so it is ignored.
 424         if tn_uri is not None and not tn_uri.startswith('data:'):
 425             thumbnail = tn_uri
 426
 427         return {
 428             '_type': 'playlist',
 429             'entries': episodes,
 430             'id': season_id,
 431             'title': season_info['title'],
 432             'description': season_info.get('description'),
 433             'thumbnail': thumbnail,
 434             'series': show_info.get('title'),
 435             'season_number': season_info.get('season'),
 436             'season': season_info['title'],
 437         }
 438
 439
 440 class CBCGemLiveIE(InfoExtractor):
 441     IE_NAME = 'gem.cbc.ca:live'
 442     _VALID_URL = r'https?://gem\.cbc\.ca/live/(?P<id>\d+)'
 443     _TEST = {
 444         'url': 'https://gem.cbc.ca/live/920604739687',
 445         'info_dict': {
 446             'title': 'Ottawa',
 447             'description': 'The live TV channel and local programming from Ottawa',
 448             'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/CBC_OTT_VMS/Live_Channel_Static_Images/Ottawa_2880x1620.jpg',
 449             'is_live': True,
 450             'id': 'AyqZwxRqh8EH',
 451             'ext': 'mp4',
 452             'timestamp': 1492106160,
 453             'upload_date': '20170413',
 454             'uploader': 'CBCC-NEW',
 455         },
 456         'skip': 'Live might have ended',
 457     }
 458
 459     # It's unclear where the chars at the end come from, but they appear to be
 460     # constant. Might need updating in the future.
 461     # There are two URLs, some livestreams are in one, and some
 462     # in the other. The JSON schema is the same for both.
 463     _API_URLS = ['https://tpfeed.cbc.ca/f/ExhSPC/t_t3UKJR6MAT', 'https://tpfeed.cbc.ca/f/ExhSPC/FNiv9xQx_BnT']
 464
 465     def _real_extract(self, url):
 466         video_id = self._match_id(url)
 467
 468         for api_url in self._API_URLS:
 469             video_info = next((
 470                 stream for stream in self._download_json(api_url, video_id)['entries']
 471                 if stream.get('guid') == video_id), None)
 472             if video_info:
 473                 break
 474         else:
 475             raise ExtractorError('Couldn\'t find video metadata, maybe this livestream is now offline', expected=True)
 476
 477         return {
 478             '_type': 'url_transparent',
 479             'ie_key': 'ThePlatform',
 480             'url': video_info['content'][0]['url'],
 481             'id': video_id,
 482             'title': video_info.get('title'),
 483             'description': video_info.get('description'),
 484             'tags': try_get(video_info, lambda x: x['keywords'].split(', ')),
 485             'thumbnail': video_info.get('cbc$staticImage'),
 486             'is_live': True,
 487         }