X-Git-Url: https://jfr.im/git/yt-dlp.git/blobdiff_plain/0f0ac87be3fc55cab8fec767c446431a8ce085f3..24f3097ea9a470a984d0454dc013cafa2325f5f8:/yt_dlp/extractor/cbc.py diff --git a/yt_dlp/extractor/cbc.py b/yt_dlp/extractor/cbc.py index 5e4526c53..1522b08e2 100644 --- a/yt_dlp/extractor/cbc.py +++ b/yt_dlp/extractor/cbc.py @@ -1,19 +1,22 @@ -# coding: utf-8 -from __future__ import unicode_literals - +import base64 +import json import re +import time +import urllib.parse +import xml.etree.ElementTree from .common import InfoExtractor -from ..compat import ( - compat_str, -) from ..utils import ( + ExtractorError, + int_or_none, + join_nonempty, js_to_json, - smuggle_url, - try_get, orderedSet, + parse_iso8601, + smuggle_url, strip_or_none, - ExtractorError, + traverse_obj, + try_get, ) @@ -61,6 +64,7 @@ class CBCIE(InfoExtractor): 'uploader': 'CBCC-NEW', 'timestamp': 255977160, }, + 'skip': '404 Not Found', }, { # multiple iframes 'url': 'http://www.cbc.ca/natureofthings/blog/birds-eye-view-from-vancouvers-burrard-street-bridge-how-we-got-the-shot', @@ -92,7 +96,7 @@ class CBCIE(InfoExtractor): # multiple CBC.APP.Caffeine.initInstance(...) 'url': 'http://www.cbc.ca/news/canada/calgary/dog-indoor-exercise-winter-1.3928238', 'info_dict': { - 'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks', + 'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks', # FIXME: actual title includes " | CBC News" 'id': 'dog-indoor-exercise-winter-1.3928238', 'description': 'md5:c18552e41726ee95bd75210d1ca9194c', }, @@ -101,7 +105,7 @@ class CBCIE(InfoExtractor): @classmethod def suitable(cls, url): - return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url) + return False if CBCPlayerIE.suitable(url) else super().suitable(url) def _extract_player_init(self, player_init, display_id): player_info = self._parse_json(player_init, display_id, js_to_json) @@ -109,22 +113,22 @@ def _extract_player_init(self, player_init, display_id): if not media_id: clip_id = player_info['clipId'] feed = self._download_json( - 'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id, + f'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={{:mpsReleases}}{{{clip_id}}}', clip_id, fatal=False) if feed: - media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str) + media_id = try_get(feed, lambda x: x['entries'][0]['guid'], str) if not media_id: media_id = self._download_json( 'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id, clip_id)['entries'][0]['id'].split('/')[-1] - return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id) + return self.url_result(f'cbcplayer:{media_id}', 'CBCPlayer', media_id) def _real_extract(self, url): display_id = self._match_id(url) webpage = self._download_webpage(url, display_id) - title = self._og_search_title(webpage, default=None) or self._html_search_meta( - 'twitter:title', webpage, 'title', default=None) or self._html_search_regex( - r'([^<]+)', webpage, 'title', fatal=False) + title = (self._og_search_title(webpage, default=None) + or self._html_search_meta('twitter:title', webpage, 'title', default=None) + or self._html_extract_title(webpage)) entries = [ self._extract_player_init(player_init, display_id) for player_init in re.findall(r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage)] @@ -135,7 +139,7 @@ def _real_extract(self, url): r'guid["\']\s*:\s*["\'](\d+)'): media_ids.extend(re.findall(media_id_re, webpage)) entries.extend([ - self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id) + self.url_result(f'cbcplayer:{media_id}', 'CBCPlayer', media_id) for media_id in orderedSet(media_ids)]) return self.playlist_result( entries, display_id, strip_or_none(title), @@ -144,7 +148,7 @@ def _real_extract(self, url): class CBCPlayerIE(InfoExtractor): IE_NAME = 'cbc.ca:player' - _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/|i/caffeine/syndicate/\?mediaId=))(?P\d+)' + _VALID_URL = r'(?:cbcplayer:|https?://(?:www\.)?cbc\.ca/(?:player/play/(?:video/)?|i/caffeine/syndicate/\?mediaId=))(?P(?:\d\.)?\d+)' _TESTS = [{ 'url': 'http://www.cbc.ca/player/play/2683190193', 'md5': '64d25f841ddf4ddb28a235338af32e2c', @@ -157,10 +161,53 @@ class CBCPlayerIE(InfoExtractor): 'upload_date': '20160210', 'uploader': 'CBCC-NEW', }, - 'skip': 'Geo-restricted to Canada', + 'skip': 'Geo-restricted to Canada and no longer available', + }, { + 'url': 'http://www.cbc.ca/i/caffeine/syndicate/?mediaId=2657631896', + 'md5': 'e5e708c34ae6fca156aafe17c43e8b75', + 'info_dict': { + 'id': '2657631896', + 'ext': 'mp3', + 'title': 'CBC Montreal is organizing its first ever community hackathon!', + 'description': 'md5:dd3b692f0a139b0369943150bd1c46a9', + 'timestamp': 1425704400, + 'upload_date': '20150307', + 'uploader': 'CBCC-NEW', + 'thumbnail': 'http://thumbnails.cbc.ca/maven_legacy/thumbnails/sonali-karnick-220.jpg', + 'chapters': [], + 'duration': 494.811, + 'categories': ['AudioMobile/All in a Weekend Montreal'], + 'tags': 'count:8', + 'location': 'Quebec', + 'series': 'All in a Weekend Montreal', + 'season': 'Season 2015', + 'season_number': 2015, + 'media_type': 'Excerpt', + }, + }, { + 'url': 'http://www.cbc.ca/i/caffeine/syndicate/?mediaId=2164402062', + 'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6', + 'info_dict': { + 'id': '2164402062', + 'ext': 'mp4', + 'title': 'Cancer survivor four times over', + 'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.', + 'timestamp': 1320410746, + 'upload_date': '20111104', + 'uploader': 'CBCC-NEW', + 'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/277/67/cancer_852x480_2164412612.jpg', + 'chapters': [], + 'duration': 186.867, + 'series': 'CBC News: Windsor at 6:00', + 'categories': ['News/Canada/Windsor'], + 'location': 'Windsor', + 'tags': ['cancer'], + 'creators': ['Allison Johnson'], + 'media_type': 'Excerpt', + }, }, { # Redirected from http://www.cbc.ca/player/AudioMobile/All%20in%20a%20Weekend%20Montreal/ID/2657632011/ - 'url': 'http://www.cbc.ca/player/play/2657631896', + 'url': 'https://www.cbc.ca/player/play/1.2985700', 'md5': 'e5e708c34ae6fca156aafe17c43e8b75', 'info_dict': { 'id': '2657631896', @@ -170,9 +217,19 @@ class CBCPlayerIE(InfoExtractor): 'timestamp': 1425704400, 'upload_date': '20150307', 'uploader': 'CBCC-NEW', + 'thumbnail': 'http://thumbnails.cbc.ca/maven_legacy/thumbnails/sonali-karnick-220.jpg', + 'chapters': [], + 'duration': 494.811, + 'categories': ['AudioMobile/All in a Weekend Montreal'], + 'tags': 'count:8', + 'location': 'Quebec', + 'series': 'All in a Weekend Montreal', + 'season': 'Season 2015', + 'season_number': 2015, + 'media_type': 'Excerpt', }, }, { - 'url': 'http://www.cbc.ca/player/play/2164402062', + 'url': 'https://www.cbc.ca/player/play/1.1711287', 'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6', 'info_dict': { 'id': '2164402062', @@ -182,25 +239,129 @@ class CBCPlayerIE(InfoExtractor): 'timestamp': 1320410746, 'upload_date': '20111104', 'uploader': 'CBCC-NEW', + 'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/277/67/cancer_852x480_2164412612.jpg', + 'chapters': [], + 'duration': 186.867, + 'series': 'CBC News: Windsor at 6:00', + 'categories': ['News/Canada/Windsor'], + 'location': 'Windsor', + 'tags': ['cancer'], + 'creators': ['Allison Johnson'], + 'media_type': 'Excerpt', }, + }, { + # Has subtitles + # These broadcasts expire after ~1 month, can find new test URL here: + # https://www.cbc.ca/player/news/TV%20Shows/The%20National/Latest%20Broadcast + 'url': 'https://www.cbc.ca/player/play/1.7159484', + 'md5': '6ed6cd0fc2ef568d2297ba68a763d455', + 'info_dict': { + 'id': '2324213316001', + 'ext': 'mp4', + 'title': 'The National | School boards sue social media giants', + 'description': 'md5:4b4db69322fa32186c3ce426da07402c', + 'timestamp': 1711681200, + 'duration': 2743.400, + 'subtitles': {'eng': [{'ext': 'vtt', 'protocol': 'm3u8_native'}]}, + 'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/607/559/thumbnail.jpeg', + 'uploader': 'CBCC-NEW', + 'chapters': 'count:5', + 'upload_date': '20240329', + 'categories': 'count:4', + 'series': 'The National - Full Show', + 'tags': 'count:1', + 'creators': ['News'], + 'location': 'Canada', + 'media_type': 'Full Program', + }, + }, { + 'url': 'https://www.cbc.ca/player/play/video/1.7194274', + 'md5': '188b96cf6bdcb2540e178a6caa957128', + 'info_dict': { + 'id': '2334524995812', + 'ext': 'mp4', + 'title': '#TheMoment a rare white spirit moose was spotted in Alberta', + 'description': 'md5:18ae269a2d0265c5b0bbe4b2e1ac61a3', + 'timestamp': 1714788791, + 'duration': 77.678, + 'subtitles': {'eng': [{'ext': 'vtt', 'protocol': 'm3u8_native'}]}, + 'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/201/543/THE_MOMENT.jpg', + 'uploader': 'CBCC-NEW', + 'chapters': 'count:0', + 'upload_date': '20240504', + 'categories': 'count:3', + 'series': 'The National', + 'tags': 'count:15', + 'creators': ['encoder'], + 'location': 'Canada', + 'media_type': 'Excerpt', + }, + }, { + 'url': 'cbcplayer:1.7159484', + 'only_matching': True, + }, { + 'url': 'cbcplayer:2164402062', + 'only_matching': True, + }, { + 'url': 'http://www.cbc.ca/player/play/2657631896', + 'only_matching': True, }] def _real_extract(self, url): video_id = self._match_id(url) + if '.' in video_id: + webpage = self._download_webpage(f'https://www.cbc.ca/player/play/{video_id}', video_id) + video_id = self._search_json( + r'window\.__INITIAL_STATE__\s*=', webpage, + 'initial state', video_id)['video']['currentClip']['mediaId'] + return { '_type': 'url_transparent', 'ie_key': 'ThePlatform', 'url': smuggle_url( - 'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/%s?mbr=true&formats=MPEG4,FLV,MP3' % video_id, { - 'force_smil_url': True + f'http://link.theplatform.com/s/ExhSPC/media/guid/2655402169/{video_id}?mbr=true&formats=MPEG4,FLV,MP3', { + 'force_smil_url': True, }), 'id': video_id, + '_format_sort_fields': ('res', 'proto'), # Prioritize direct http formats over HLS } +class CBCPlayerPlaylistIE(InfoExtractor): + IE_NAME = 'cbc.ca:player:playlist' + _VALID_URL = r'https?://(?:www\.)?cbc\.ca/(?:player/)(?!play/)(?P[^?#]+)' + _TESTS = [{ + 'url': 'https://www.cbc.ca/player/news/TV%20Shows/The%20National/Latest%20Broadcast', + 'playlist_mincount': 25, + 'info_dict': { + 'id': 'news/tv shows/the national/latest broadcast', + }, + }, { + 'url': 'https://www.cbc.ca/player/news/Canada/North', + 'playlist_mincount': 25, + 'info_dict': { + 'id': 'news/canada/north', + }, + }] + + def _real_extract(self, url): + playlist_id = urllib.parse.unquote(self._match_id(url)).lower() + webpage = self._download_webpage(url, playlist_id) + json_content = self._search_json( + r'window\.__INITIAL_STATE__\s*=', webpage, 'initial state', playlist_id) + + def entries(): + for video_id in traverse_obj(json_content, ( + 'video', 'clipsByCategory', lambda k, _: k.lower() == playlist_id, 'items', ..., 'id', + )): + yield self.url_result(f'https://www.cbc.ca/player/play/{video_id}', CBCPlayerIE) + + return self.playlist_result(entries(), playlist_id) + + class CBCGemIE(InfoExtractor): IE_NAME = 'gem.cbc.ca' - _VALID_URL = r'https?://gem\.cbc\.ca/media/(?P[0-9a-z-]+/s[0-9]+[a-z][0-9]+)' + _VALID_URL = r'https?://gem\.cbc\.ca/(?:media/)?(?P[0-9a-z-]+/s[0-9]+[a-z][0-9]+)' _TESTS = [{ # This is a normal, public, TV show video 'url': 'https://gem.cbc.ca/media/schitts-creek/s06e01', @@ -243,50 +404,143 @@ class CBCGemIE(InfoExtractor): }, 'params': {'format': 'bv'}, 'skip': 'Geo-restricted to Canada', + }, { + 'url': 'https://gem.cbc.ca/nadiyas-family-favourites/s01e01', + 'only_matching': True, }] - _API_BASE = 'https://services.radio-canada.ca/ott/cbc-api/v2/assets/' + + _GEO_COUNTRIES = ['CA'] + _TOKEN_API_KEY = '3f4beddd-2061-49b0-ae80-6f1f2ed65b37' + _NETRC_MACHINE = 'cbcgem' + _claims_token = None + + def _new_claims_token(self, email, password): + data = json.dumps({ + 'email': email, + 'password': password, + }).encode() + headers = {'content-type': 'application/json'} + query = {'apikey': self._TOKEN_API_KEY} + resp = self._download_json('https://api.loginradius.com/identity/v2/auth/login', + None, data=data, headers=headers, query=query) + access_token = resp['access_token'] + + query = { + 'access_token': access_token, + 'apikey': self._TOKEN_API_KEY, + 'jwtapp': 'jwt', + } + resp = self._download_json('https://cloud-api.loginradius.com/sso/jwt/api/token', + None, headers=headers, query=query) + sig = resp['signature'] + + data = json.dumps({'jwt': sig}).encode() + headers = {'content-type': 'application/json', 'ott-device-type': 'web'} + resp = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/token', + None, data=data, headers=headers, expected_status=426) + cbc_access_token = resp['accessToken'] + + headers = {'content-type': 'application/json', 'ott-device-type': 'web', 'ott-access-token': cbc_access_token} + resp = self._download_json('https://services.radio-canada.ca/ott/cbc-api/v2/profile', + None, headers=headers, expected_status=426) + return resp['claimsToken'] + + def _get_claims_token_expiry(self): + # Token is a JWT + # JWT is decoded here and 'exp' field is extracted + # It is a Unix timestamp for when the token expires + b64_data = self._claims_token.split('.')[1] + data = base64.urlsafe_b64decode(b64_data + '==') + return json.loads(data)['exp'] + + def claims_token_expired(self): + exp = self._get_claims_token_expiry() + # It will expire in less than 10 seconds, or has already expired + return exp - time.time() < 10 + + def claims_token_valid(self): + return self._claims_token is not None and not self.claims_token_expired() + + def _get_claims_token(self, email, password): + if not self.claims_token_valid(): + self._claims_token = self._new_claims_token(email, password) + self.cache.store(self._NETRC_MACHINE, 'claims_token', self._claims_token) + return self._claims_token + + def _real_initialize(self): + if self.claims_token_valid(): + return + self._claims_token = self.cache.load(self._NETRC_MACHINE, 'claims_token') + + def _find_secret_formats(self, formats, video_id): + """ Find a valid video url and convert it to the secret variant """ + base_format = next((f for f in formats if f.get('vcodec') != 'none'), None) + if not base_format: + return + + base_url = re.sub(r'(Manifest\(.*?),filter=[\w-]+(.*?\))', r'\1\2', base_format['url']) + url = re.sub(r'(Manifest\(.*?),format=[\w-]+(.*?\))', r'\1\2', base_url) + + secret_xml = self._download_xml(url, video_id, note='Downloading secret XML', fatal=False) + if not isinstance(secret_xml, xml.etree.ElementTree.Element): + return + + for child in secret_xml: + if child.attrib.get('Type') != 'video': + continue + for video_quality in child: + bitrate = int_or_none(video_quality.attrib.get('Bitrate')) + if not bitrate or 'Index' not in video_quality.attrib: + continue + height = int_or_none(video_quality.attrib.get('MaxHeight')) + + yield { + **base_format, + 'format_id': join_nonempty('sec', height), + # Note: \g<1> is necessary instead of \1 since bitrate is a number + 'url': re.sub(r'(QualityLevels\()\d+(\))', fr'\g<1>{bitrate}\2', base_url), + 'width': int_or_none(video_quality.attrib.get('MaxWidth')), + 'tbr': bitrate / 1000.0, + 'height': height, + } def _real_extract(self, url): video_id = self._match_id(url) - video_info = self._download_json(self._API_BASE + video_id, video_id) - - last_error = None - attempt = -1 - retries = self.get_param('extractor_retries', 15) - while attempt < retries: - attempt += 1 - if last_error: - self.report_warning('%s. Retrying ...' % last_error) - m3u8_info = self._download_json( - video_info['playSession']['url'], video_id, - note='Downloading JSON metadata%s' % f' (attempt {attempt})') - m3u8_url = m3u8_info.get('url') - if m3u8_url: - break - elif m3u8_info.get('errorCode') == 1: - self.raise_geo_restricted(countries=['CA']) - else: - last_error = f'{self.IE_NAME} said: {m3u8_info.get("errorCode")} - {m3u8_info.get("message")}' - # 35 means media unavailable, but retries work - if m3u8_info.get('errorCode') != 35 or attempt >= retries: - raise ExtractorError(last_error) + video_info = self._download_json( + f'https://services.radio-canada.ca/ott/cbc-api/v2/assets/{video_id}', + video_id, expected_status=426) + + email, password = self._get_login_info() + if email and password: + claims_token = self._get_claims_token(email, password) + headers = {'x-claims-token': claims_token} + else: + headers = {} + m3u8_info = self._download_json(video_info['playSession']['url'], video_id, headers=headers) + m3u8_url = m3u8_info.get('url') + + if m3u8_info.get('errorCode') == 1: + self.raise_geo_restricted(countries=['CA']) + elif m3u8_info.get('errorCode') == 35: + self.raise_login_required(method='password') + elif m3u8_info.get('errorCode') != 0: + raise ExtractorError(f'{self.IE_NAME} said: {m3u8_info.get("errorCode")} - {m3u8_info.get("message")}') formats = self._extract_m3u8_formats(m3u8_url, video_id, m3u8_id='hls') self._remove_duplicate_formats(formats) + formats.extend(self._find_secret_formats(formats, video_id)) - for i, format in enumerate(formats): - if format.get('vcodec') == 'none': - if format.get('ext') is None: - format['ext'] = 'm4a' - if format.get('acodec') is None: - format['acodec'] = 'mp4a.40.2' + for fmt in formats: + if fmt.get('vcodec') == 'none': + if fmt.get('ext') is None: + fmt['ext'] = 'm4a' + if fmt.get('acodec') is None: + fmt['acodec'] = 'mp4a.40.2' # Put described audio at the beginning of the list, so that it # isn't chosen by default, as most people won't want it. - if 'descriptive' in format['format_id'].lower(): - format['preference'] = -2 - - self._sort_formats(formats) + if 'descriptive' in fmt['format_id'].lower(): + fmt['preference'] = -2 return { 'id': video_id, @@ -309,7 +563,7 @@ def _real_extract(self, url): class CBCGemPlaylistIE(InfoExtractor): IE_NAME = 'gem.cbc.ca:playlist' - _VALID_URL = r'https?://gem\.cbc\.ca/media/(?P(?P[0-9a-z-]+)/s(?P[0-9]+))/?(?:[?#]|$)' + _VALID_URL = r'https?://gem\.cbc\.ca/(?:media/)?(?P(?P[0-9a-z-]+)/s(?P[0-9]+))/?(?:[?#]|$)' _TESTS = [{ # TV show playlist, all public videos 'url': 'https://gem.cbc.ca/media/schitts-creek/s06', @@ -318,7 +572,14 @@ class CBCGemPlaylistIE(InfoExtractor): 'id': 'schitts-creek/s06', 'title': 'Season 6', 'description': 'md5:6a92104a56cbeb5818cc47884d4326a2', + 'series': 'Schitt\'s Creek', + 'season_number': 6, + 'season': 'Season 6', + 'thumbnail': 'https://images.radio-canada.ca/v1/synps-cbc/season/perso/cbc_schitts_creek_season_06_carousel_v03.jpg?impolicy=ott&im=Resize=(_Size_)&quality=75', }, + }, { + 'url': 'https://gem.cbc.ca/schitts-creek/s06', + 'only_matching': True, }] _API_BASE = 'https://services.radio-canada.ca/ott/cbc-api/v2/shows/' @@ -326,9 +587,10 @@ def _real_extract(self, url): match = self._match_valid_url(url) season_id = match.group('id') show = match.group('show') - show_info = self._download_json(self._API_BASE + show, season_id) + show_info = self._download_json(self._API_BASE + show, season_id, expected_status=426) season = int(match.group('season')) - season_info = try_get(show_info, lambda x: x['seasons'][season - 1]) + + season_info = next((s for s in show_info['seasons'] if s.get('season') == season), None) if season_info is None: raise ExtractorError(f'Couldn\'t find season {season} of {show}') @@ -377,49 +639,90 @@ def _real_extract(self, url): class CBCGemLiveIE(InfoExtractor): IE_NAME = 'gem.cbc.ca:live' - _VALID_URL = r'https?://gem\.cbc\.ca/live/(?P[0-9]{12})' - _TEST = { - 'url': 'https://gem.cbc.ca/live/920604739687', - 'info_dict': { - 'title': 'Ottawa', - 'description': 'The live TV channel and local programming from Ottawa', - 'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/CBC_OTT_VMS/Live_Channel_Static_Images/Ottawa_2880x1620.jpg', - 'is_live': True, - 'id': 'AyqZwxRqh8EH', - 'ext': 'mp4', - 'timestamp': 1492106160, - 'upload_date': '20170413', - 'uploader': 'CBCC-NEW', + _VALID_URL = r'https?://gem\.cbc\.ca/live(?:-event)?/(?P\d+)' + _TESTS = [ + { + 'url': 'https://gem.cbc.ca/live/920604739687', + 'info_dict': { + 'title': 'Ottawa', + 'description': 'The live TV channel and local programming from Ottawa', + 'thumbnail': 'https://thumbnails.cbc.ca/maven_legacy/thumbnails/CBC_OTT_VMS/Live_Channel_Static_Images/Ottawa_2880x1620.jpg', + 'is_live': True, + 'id': 'AyqZwxRqh8EH', + 'ext': 'mp4', + 'timestamp': 1492106160, + 'upload_date': '20170413', + 'uploader': 'CBCC-NEW', + }, + 'skip': 'Live might have ended', }, - 'skip': 'Live might have ended', - } - - # It's unclear where the chars at the end come from, but they appear to be - # constant. Might need updating in the future. - _API = 'https://tpfeed.cbc.ca/f/ExhSPC/t_t3UKJR6MAT' + { + 'url': 'https://gem.cbc.ca/live/44', + 'info_dict': { + 'id': '44', + 'ext': 'mp4', + 'is_live': True, + 'title': r're:^Ottawa [0-9\-: ]+', + 'description': 'The live TV channel and local programming from Ottawa', + 'live_status': 'is_live', + 'thumbnail': r're:https://images.gem.cbc.ca/v1/cbc-gem/live/.*', + }, + 'params': {'skip_download': True}, + 'skip': 'Live might have ended', + }, + { + 'url': 'https://gem.cbc.ca/live-event/10835', + 'info_dict': { + 'id': '10835', + 'ext': 'mp4', + 'is_live': True, + 'title': r're:^The National \| Biden’s trip wraps up, Paltrow testifies, Bird flu [0-9\-: ]+', + 'description': 'March 24, 2023 | President Biden’s Ottawa visit ends with big pledges from both countries. Plus, Gwyneth Paltrow testifies in her ski collision trial.', + 'live_status': 'is_live', + 'thumbnail': r're:https://images.gem.cbc.ca/v1/cbc-gem/live/.*', + 'timestamp': 1679706000, + 'upload_date': '20230325', + }, + 'params': {'skip_download': True}, + 'skip': 'Live might have ended', + }, + ] def _real_extract(self, url): video_id = self._match_id(url) - live_info = self._download_json(self._API, video_id)['entries'] - - video_info = None - for stream in live_info: - if stream.get('guid') == video_id: - video_info = stream - - if video_info is None: - raise ExtractorError( - 'Couldn\'t find video metadata, maybe this livestream is now offline', - expected=True) + webpage = self._download_webpage(url, video_id) + video_info = self._search_nextjs_data(webpage, video_id)['props']['pageProps']['data'] + + # Two types of metadata JSON + if not video_info.get('formattedIdMedia'): + video_info = traverse_obj( + video_info, (('freeTv', ('streams', ...)), 'items', lambda _, v: v['key'] == video_id, {dict}), + get_all=False, default={}) + + video_stream_id = video_info.get('formattedIdMedia') + if not video_stream_id: + raise ExtractorError('Couldn\'t find video metadata, maybe this livestream is now offline', expected=True) + + stream_data = self._download_json( + 'https://services.radio-canada.ca/media/validation/v2/', video_id, query={ + 'appCode': 'mpx', + 'connectionType': 'hd', + 'deviceType': 'ipad', + 'idMedia': video_stream_id, + 'multibitrate': 'true', + 'output': 'json', + 'tech': 'hls', + 'manifestType': 'desktop', + }) return { - '_type': 'url_transparent', - 'ie_key': 'ThePlatform', - 'url': video_info['content'][0]['url'], 'id': video_id, - 'title': video_info.get('title'), - 'description': video_info.get('description'), - 'tags': try_get(video_info, lambda x: x['keywords'].split(', ')), - 'thumbnail': video_info.get('cbc$staticImage'), + 'formats': self._extract_m3u8_formats(stream_data['url'], video_id, 'mp4', live=True), 'is_live': True, + **traverse_obj(video_info, { + 'title': 'title', + 'description': 'description', + 'thumbnail': ('images', 'card', 'url'), + 'timestamp': ('airDate', {parse_iso8601}), + }), }