yt_dlp/extractor/cbc.py

   1 import re
   2 import json
   3 import base64
   4 import time
   5
   6 from .common import InfoExtractor
   7 from ..compat import (
   8     compat_str,
   9 )
  10 from ..utils import (
  11     int_or_none,
  12     join_nonempty,
  13     js_to_json,
  14     orderedSet,
  15     smuggle_url,
  16     strip_or_none,
  17     try_get,
  18     ExtractorError,
  19 )
  20
  21
  22 class CBCIE(InfoExtractor):
  23     IE_NAME = 'cbc.ca'
  24     _VALID_URL = r'https?://(?:www\.)?cbc\.ca/(?!player/)(?:[^/]+/)+(?P<id>[^/?#]+)'
  25     _TESTS = [{
  26         # with mediaId
  27         'url': 'http://www.cbc.ca/22minutes/videos/clips-season-23/don-cherry-play-offs',
  28         'md5': '97e24d09672fc4cf56256d6faa6c25bc',
  29         'info_dict': {
  30             'id': '2682904050',
  31             'ext': 'mp4',
  32             'title': 'Don Cherry – All-Stars',
  33             'description': 'Don Cherry has a bee in his bonnet about AHL player John Scott because that guy’s got heart.',
  34             'timestamp': 1454463000,
  35             'upload_date': '20160203',
  36             'uploader': 'CBCC-NEW',
  37         },
  38         'skip': 'Geo-restricted to Canada',
  39     }, {
  40         # with clipId, feed available via tpfeed.cbc.ca and feed.theplatform.com
  41         'url': 'http://www.cbc.ca/22minutes/videos/22-minutes-update/22-minutes-update-episode-4',
  42         'md5': '162adfa070274b144f4fdc3c3b8207db',
  43         'info_dict': {
  44             'id': '2414435309',
  45             'ext': 'mp4',
  46             'title': '22 Minutes Update: What Not To Wear Quebec',
  47             'description': "This week's latest Canadian top political story is What Not To Wear Quebec.",
  48             'upload_date': '20131025',
  49             'uploader': 'CBCC-NEW',
  50             'timestamp': 1382717907,
  51         },
  52         'skip': 'No longer available',
  53     }, {
  54         # with clipId, feed only available via tpfeed.cbc.ca
  55         'url': 'http://www.cbc.ca/archives/entry/1978-robin-williams-freestyles-on-90-minutes-live',
  56         'md5': '0274a90b51a9b4971fe005c63f592f12',
  57         'info_dict': {
  58             'id': '2487345465',
  59             'ext': 'mp4',
  60             'title': 'Robin Williams freestyles on 90 Minutes Live',
  61             'description': 'Wacky American comedian Robin Williams shows off his infamous "freestyle" comedic talents while being interviewed on CBC\'s 90 Minutes Live.',
  62             'upload_date': '19780210',
  63             'uploader': 'CBCC-NEW',
  64             'timestamp': 255977160,
  65         },
  66     }, {
  67         # multiple iframes
  68         'url': 'http://www.cbc.ca/natureofthings/blog/birds-eye-view-from-vancouvers-burrard-street-bridge-how-we-got-the-shot',
  69         'playlist': [{
  70             'md5': '377572d0b49c4ce0c9ad77470e0b96b4',
  71             'info_dict': {
  72                 'id': '2680832926',
  73                 'ext': 'mp4',
  74                 'title': 'An Eagle\'s-Eye View Off Burrard Bridge',
  75                 'description': 'Hercules the eagle flies from Vancouver\'s Burrard Bridge down to a nearby park with a mini-camera strapped to his back.',
  76                 'upload_date': '20160201',
  77                 'timestamp': 1454342820,
  78                 'uploader': 'CBCC-NEW',
  79             },
  80         }, {
  81             'md5': '415a0e3f586113894174dfb31aa5bb1a',
  82             'info_dict': {
  83                 'id': '2658915080',
  84                 'ext': 'mp4',
  85                 'title': 'Fly like an eagle!',
  86                 'description': 'Eagle equipped with a mini camera flies from the world\'s tallest tower',
  87                 'upload_date': '20150315',
  88                 'timestamp': 1426443984,
  89                 'uploader': 'CBCC-NEW',
  90             },
  91         }],
  92         'skip': 'Geo-restricted to Canada',
  93     }, {
  94         # multiple CBC.APP.Caffeine.initInstance(...)
  95         'url': 'http://www.cbc.ca/news/canada/calgary/dog-indoor-exercise-winter-1.3928238',
  96         'info_dict': {
  97             'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks',
  98             'id': 'dog-indoor-exercise-winter-1.3928238',
  99             'description': 'md5:c18552e41726ee95bd75210d1ca9194c',
 100         },
 101         'playlist_mincount': 6,
 102     }]
 103
 104     @classmethod
 105     def suitable(cls, url):
 106         return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url)
 107
 108     def _extract_player_init(self, player_init, display_id):
 109         player_info = self._parse_json(player_init, display_id, js_to_json)
 110         media_id = player_info.get('mediaId')
 111         if not media_id:
 112             clip_id = player_info['clipId']
 113             feed = self._download_json(
 114                 'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
 115                 clip_id, fatal=False)
 116             if feed:
 117                 media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
 118             if not media_id:
 119                 media_id = self._download_json(
 120                     'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
 121                     clip_id)['entries'][0]['id'].split('/')[-1]
 122         return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 123
 124     def _real_extract(self, url):
 125         display_id = self._match_id(url)
 126         webpage = self._download_webpage(url, display_id)
 127         title = (self._og_search_title(webpage, default=None)
 128                  or self._html_search_meta('twitter:title', webpage, 'title', default=None)
 129                  or self._html_extract_title(webpage))
 130         entries = [
 131             self._extract_player_init(player_init, display_id)
 132             for player_init in re.findall(r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage)]
 133         media_ids = []
 134         for media_id_re in (
 135                 r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"',
 136                 r'<div[^>]+\bid=["\']player-(\d+)',
 137                 r'guid["\']\s*:\s*["\'](\d+)'):
 138             media_ids.extend(re.findall(media_id_re, webpage))
 139         entries.extend([
 140             self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
 141             for media_id in orderedSet(media_ids)])
 142         return self.playlist_result(
 143             entries, display_id, strip_or_none(title),
 144             self._og_search_description(webpage))
 145
 146
 147 class CBCPlayerIE(InfoExtractor):
 148     IE_NAME = 'cbc.ca:player'
 149     _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P<id>\d+)'
 150     _TESTS = [{
 151         'url': 'http://www.cbc.ca/player/play/2683190193',
 152         'md5': '64d25f841ddf4ddb28a235338af32e2c',
 153         'info_dict': {
 154             'id': '2683190193',
 155             'ext': 'mp4',
 156             'title': 'Gerry Runs a Sweat Shop',
 157             'description': 'md5:b457e1c01e8ff408d9d801c1c2cd29b0',
 158             'timestamp': 1455071400,
 159             'upload_date': '20160210',
 160             'uploader': 'CBCC-NEW',
 161         },
 162         'skip': 'Geo-restricted to Canada',
 163     }, {
 164         # Redirected from http://www.cbc.ca/player/AudioMobile/All%20in%20a%20Weekend%20Montreal/ID/2657632011/
 165         'url': 'http://www.cbc.ca/player/play/2657631896',
 166         'md5': 'e5e708c34ae6fca156aafe17c43e8b75',
 167         'info_dict': {
 168             'id': '2657631896',
 169             'ext': 'mp3',
 170             'title': 'CBC Montreal is organizing its first ever community hackathon!',
 171             '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.',
 172             'timestamp': 1425704400,
 173             'upload_date': '20150307',
 174             'uploader': 'CBCC-NEW',
 175         },
 176     }, {
 177         'url': 'http://www.cbc.ca/player/play/2164402062',
 178         'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6',
 179         'info_dict': {
 180             'id': '2164402062',
 181             'ext': 'mp4',
 182             'title': 'Cancer survivor four times over',
 183             'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.',
 184             'timestamp': 1320410746,
 185             'upload_date': '20111104',
 186             'uploader': 'CBCC-NEW',
 187         },
 188     }]
 189
 190     def _real_extract(self, url):
 191         video_id = self._match_id(url)
 192         return {
 193             '_type': 'url_transparent',
 194             'ie_key': 'ThePlatform',
 195             'url': smuggle_url(
 196                 'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/%s?mbr=true&formats=MPEG4,FLV,MP3' % video_id, {
 197                     'force_smil_url': True
 198                 }),
 199             'id': video_id,
 200         }
 201
 202
 203 class CBCGemIE(InfoExtractor):
 204     IE_NAME = 'gem.cbc.ca'
 205     _VALID_URL = r'https?://gem\.cbc\.ca/media/(?P<id>[0-9a-z-]+/s[0-9]+[a-z][0-9]+)'
 206     _TESTS = [{
 207         # This is a normal, public, TV show video
 208         'url': 'https://gem.cbc.ca/media/schitts-creek/s06e01',
 209         'md5': '93dbb31c74a8e45b378cf13bd3f6f11e',
 210         'info_dict': {
 211             'id': 'schitts-creek/s06e01',
 212             'ext': 'mp4',
 213             'title': 'Smoke Signals',
 214             'description': 'md5:929868d20021c924020641769eb3e7f1',
 215             'thumbnail': 'https://images.radio-canada.ca/v1/synps-cbc/episode/perso/cbc_schitts_creek_season_06e01_thumbnail_v01.jpg?im=Resize=(Size)',
 216             'duration': 1314,
 217             'categories': ['comedy'],
 218             'series': 'Schitt\'s Creek',
 219             'season': 'Season 6',
 220             'season_number': 6,
 221             'episode': 'Smoke Signals',
 222             'episode_number': 1,
 223             'episode_id': 'schitts-creek/s06e01',
 224         },
 225         'params': {'format': 'bv'},
 226         'skip': 'Geo-restricted to Canada',
 227     }, {
 228         # This video requires an account in the browser, but works fine in yt-dlp
 229         'url': 'https://gem.cbc.ca/media/schitts-creek/s01e01',
 230         'md5': '297a9600f554f2258aed01514226a697',
 231         'info_dict': {
 232             'id': 'schitts-creek/s01e01',
 233             'ext': 'mp4',
 234             'title': 'The Cup Runneth Over',
 235             'description': 'md5:9bca14ea49ab808097530eb05a29e797',
 236             'thumbnail': 'https://images.radio-canada.ca/v1/synps-cbc/episode/perso/cbc_schitts_creek_season_01e01_thumbnail_v01.jpg?im=Resize=(Size)',
 237             'series': 'Schitt\'s Creek',
 238             'season_number': 1,
 239             'season': 'Season 1',
 240             'episode_number': 1,
 241             'episode': 'The Cup Runneth Over',
 242             'episode_id': 'schitts-creek/s01e01',
 243             'duration': 1309,
 244             'categories': ['comedy'],
 245         },
 246         'params': {'format': 'bv'},
 247         'skip': 'Geo-restricted to Canada',
 248     }]
 249
 250     _GEO_COUNTRIES = ['CA']
 251     _TOKEN_API_KEY = '3f4beddd-2061-49b0-ae80-6f1f2ed65b37'
 252     _NETRC_MACHINE = 'cbcgem'
 253     _claims_token = None
 254
 255     def _new_claims_token(self, email, password):
 256         data = json.dumps({
 257             'email': email,
 258             'password': password,
 259         }).encode()
 260         headers = {'content-type': 'application/json'}
 261         query = {'apikey': self._TOKEN_API_KEY}
 262         resp = self._download_json('https://api.loginradius.com/identity/v2/auth/login',
 263                                    None, data=data, headers=headers, query=query)
 264         access_token = resp['access_token']
 265
 266         query = {
 267             'access_token': access_token,
 268             'apikey': self._TOKEN_API_KEY,
 269             'jwtapp': 'jwt',
 270         }
 271         resp = self._download_json('https://cloud-api.loginradius.com/sso/jwt/api/token',
 272                                    None, headers=headers, query=query)
 273         sig = resp['signature']
 274
 275         data = json.dumps({'jwt': sig}).encode()
 276         headers = {'content-type': 'application/json', 'ott-device-type': 'web'}
 277         resp = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/token',
 278                                    None, data=data, headers=headers)
 279         cbc_access_token = resp['accessToken']
 280
 281         headers = {'content-type': 'application/json', 'ott-device-type': 'web', 'ott-access-token': cbc_access_token}
 282         resp = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/profile',
 283                                    None, headers=headers)
 284         return resp['claimsToken']
 285
 286     def _get_claims_token_expiry(self):
 287         # Token is a JWT
 288         # JWT is decoded here and 'exp' field is extracted
 289         # It is a Unix timestamp for when the token expires
 290         b64_data = self._claims_token.split('.')[1]
 291         data = base64.urlsafe_b64decode(b64_data + "==")
 292         return json.loads(data)['exp']
 293
 294     def claims_token_expired(self):
 295         exp = self._get_claims_token_expiry()
 296         if exp - time.time() < 10:
 297             # It will expire in less than 10 seconds, or has already expired
 298             return True
 299         return False
 300
 301     def claims_token_valid(self):
 302         return self._claims_token is not None and not self.claims_token_expired()
 303
 304     def _get_claims_token(self, email, password):
 305         if not self.claims_token_valid():
 306             self._claims_token = self._new_claims_token(email, password)
 307             self.cache.store(self._NETRC_MACHINE, 'claims_token', self._claims_token)
 308         return self._claims_token
 309
 310     def _real_initialize(self):
 311         if self.claims_token_valid():
 312             return
 313         self._claims_token = self.cache.load(self._NETRC_MACHINE, 'claims_token')
 314
 315     def _find_secret_formats(self, formats, video_id):
 316         """ Find a valid video url and convert it to the secret variant """
 317         base_format = next((f for f in formats if f.get('vcodec') != 'none'), None)
 318         if not base_format:
 319             return
 320
 321         base_url = re.sub(r'(Manifest\(.*?),filter=[\w-]+(.*?\))', r'\1\2', base_format['url'])
 322         url = re.sub(r'(Manifest\(.*?),format=[\w-]+(.*?\))', r'\1\2', base_url)
 323
 324         secret_xml = self._download_xml(url, video_id, note='Downloading secret XML', fatal=False)
 325         if not secret_xml:
 326             return
 327
 328         for child in secret_xml:
 329             if child.attrib.get('Type') != 'video':
 330                 continue
 331             for video_quality in child:
 332                 bitrate = int_or_none(video_quality.attrib.get('Bitrate'))
 333                 if not bitrate or 'Index' not in video_quality.attrib:
 334                     continue
 335                 height = int_or_none(video_quality.attrib.get('MaxHeight'))
 336
 337                 yield {
 338                     **base_format,
 339                     'format_id': join_nonempty('sec', height),
 340                     # Note: \g<1> is necessary instead of \1 since bitrate is a number
 341                     'url': re.sub(r'(QualityLevels\()\d+(\))', fr'\g<1>{bitrate}\2', base_url),
 342                     'width': int_or_none(video_quality.attrib.get('MaxWidth')),
 343                     'tbr': bitrate / 1000.0,
 344                     'height': height,
 345                 }
 346
 347     def _real_extract(self, url):
 348         video_id = self._match_id(url)
 349         video_info = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/assets/' + video_id, video_id)
 350
 351         email, password = self._get_login_info()
 352         if email and password:
 353             claims_token = self._get_claims_token(email, password)
 354             headers = {'x-claims-token': claims_token}
 355         else:
 356             headers = {}
 357         m3u8_info = self._download_json(video_info['playSession']['url'], video_id, headers=headers)
 358         m3u8_url = m3u8_info.get('url')
 359
 360         if m3u8_info.get('errorCode') == 1:
 361             self.raise_geo_restricted(countries=['CA'])
 362         elif m3u8_info.get('errorCode') == 35:
 363             self.raise_login_required(method='password')
 364         elif m3u8_info.get('errorCode') != 0:
 365             raise ExtractorError(f'{self.IE_NAME} said: {m3u8_info.get("errorCode")} - {m3u8_info.get("message")}')
 366
 367         formats = self._extract_m3u8_formats(m3u8_url, video_id, m3u8_id='hls')
 368         self._remove_duplicate_formats(formats)
 369         formats.extend(self._find_secret_formats(formats, video_id))
 370
 371         for format in formats:
 372             if format.get('vcodec') == 'none':
 373                 if format.get('ext') is None:
 374                     format['ext'] = 'm4a'
 375                 if format.get('acodec') is None:
 376                     format['acodec'] = 'mp4a.40.2'
 377
 378                 # Put described audio at the beginning of the list, so that it
 379                 # isn't chosen by default, as most people won't want it.
 380                 if 'descriptive' in format['format_id'].lower():
 381                     format['preference'] = -2
 382
 383         self._sort_formats(formats)
 384
 385         return {
 386             'id': video_id,
 387             'title': video_info['title'],
 388             'description': video_info.get('description'),
 389             'thumbnail': video_info.get('image'),
 390             'series': video_info.get('series'),
 391             'season_number': video_info.get('season'),
 392             'season': f'Season {video_info.get("season")}',
 393             'episode_number': video_info.get('episode'),
 394             'episode': video_info.get('title'),
 395             'episode_id': video_id,
 396             'duration': video_info.get('duration'),
 397             'categories': [video_info.get('category')],
 398             'formats': formats,
 399             'release_timestamp': video_info.get('airDate'),
 400             'timestamp': video_info.get('availableDate'),
 401         }
 402
 403
 404 class CBCGemPlaylistIE(InfoExtractor):
 405     IE_NAME = 'gem.cbc.ca:playlist'
 406     _VALID_URL = r'https?://gem\.cbc\.ca/media/(?P<id>(?P<show>[0-9a-z-]+)/s(?P<season>[0-9]+))/?(?:[?#]|$)'
 407     _TESTS = [{
 408         # TV show playlist, all public videos
 409         'url': 'https://gem.cbc.ca/media/schitts-creek/s06',
 410         'playlist_count': 16,
 411         'info_dict': {
 412             'id': 'schitts-creek/s06',
 413             'title': 'Season 6',
 414             'description': 'md5:6a92104a56cbeb5818cc47884d4326a2',
 415         },
 416     }]
 417     _API_BASE = 'https://services.radio-canada.ca/ott/cbc-api/v2/shows/'
 418
 419     def _real_extract(self, url):
 420         match = self._match_valid_url(url)
 421         season_id = match.group('id')
 422         show = match.group('show')
 423         show_info = self._download_json(self._API_BASE + show, season_id)
 424         season = int(match.group('season'))
 425
 426         season_info = next((s for s in show_info['seasons'] if s.get('season') == season), None)
 427
 428         if season_info is None:
 429             raise ExtractorError(f'Couldn\'t find season {season} of {show}')
 430
 431         episodes = []
 432         for episode in season_info['assets']:
 433             episodes.append({
 434                 '_type': 'url_transparent',
 435                 'ie_key': 'CBCGem',
 436                 'url': 'https://gem.cbc.ca/media/' + episode['id'],
 437                 'id': episode['id'],
 438                 'title': episode.get('title'),
 439                 'description': episode.get('description'),
 440                 'thumbnail': episode.get('image'),
 441                 'series': episode.get('series'),
 442                 'season_number': episode.get('season'),
 443                 'season': season_info['title'],
 444                 'season_id': season_info.get('id'),
 445                 'episode_number': episode.get('episode'),
 446                 'episode': episode.get('title'),
 447                 'episode_id': episode['id'],
 448                 'duration': episode.get('duration'),
 449                 'categories': [episode.get('category')],
 450             })
 451
 452         thumbnail = None
 453         tn_uri = season_info.get('image')
 454         # the-national was observed to use a "data:image/png;base64"
 455         # URI for their 'image' value. The image was 1x1, and is
 456         # probably just a placeholder, so it is ignored.
 457         if tn_uri is not None and not tn_uri.startswith('data:'):
 458             thumbnail = tn_uri
 459
 460         return {
 461             '_type': 'playlist',
 462             'entries': episodes,
 463             'id': season_id,
 464             'title': season_info['title'],
 465             'description': season_info.get('description'),
 466             'thumbnail': thumbnail,
 467             'series': show_info.get('title'),
 468             'season_number': season_info.get('season'),
 469             'season': season_info['title'],
 470         }
 471
 472
 473 class CBCGemLiveIE(InfoExtractor):
 474     IE_NAME = 'gem.cbc.ca:live'
 475     _VALID_URL = r'https?://gem\.cbc\.ca/live/(?P<id>\d+)'
 476     _TEST = {
 477         'url': 'https://gem.cbc.ca/live/920604739687',
 478         'info_dict': {
 479             'title': 'Ottawa',
 480             'description': 'The live TV channel and local programming from Ottawa',
 481             'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/CBC_OTT_VMS/Live_Channel_Static_Images/Ottawa_2880x1620.jpg',
 482             'is_live': True,
 483             'id': 'AyqZwxRqh8EH',
 484             'ext': 'mp4',
 485             'timestamp': 1492106160,
 486             'upload_date': '20170413',
 487             'uploader': 'CBCC-NEW',
 488         },
 489         'skip': 'Live might have ended',
 490     }
 491
 492     # It's unclear where the chars at the end come from, but they appear to be
 493     # constant. Might need updating in the future.
 494     # There are two URLs, some livestreams are in one, and some
 495     # in the other. The JSON schema is the same for both.
 496     _API_URLS = ['https://tpfeed.cbc.ca/f/ExhSPC/t_t3UKJR6MAT', 'https://tpfeed.cbc.ca/f/ExhSPC/FNiv9xQx_BnT']
 497
 498     def _real_extract(self, url):
 499         video_id = self._match_id(url)
 500
 501         for api_url in self._API_URLS:
 502             video_info = next((
 503                 stream for stream in self._download_json(api_url, video_id)['entries']
 504                 if stream.get('guid') == video_id), None)
 505             if video_info:
 506                 break
 507         else:
 508             raise ExtractorError('Couldn\'t find video metadata, maybe this livestream is now offline', expected=True)
 509
 510         return {
 511             '_type': 'url_transparent',
 512             'ie_key': 'ThePlatform',
 513             'url': video_info['content'][0]['url'],
 514             'id': video_id,
 515             'title': video_info.get('title'),
 516             'description': video_info.get('description'),
 517             'tags': try_get(video_info, lambda x: x['keywords'].split(', ')),
 518             'thumbnail': video_info.get('cbc$staticImage'),
 519             'is_live': True,
 520         }