diff options
| -rw-r--r-- | yt_dlp/extractor/rtp.py | 226 |
1 files changed, 154 insertions, 72 deletions
diff --git a/yt_dlp/extractor/rtp.py b/yt_dlp/extractor/rtp.py index 26aec2e4c..03e985940 100644 --- a/yt_dlp/extractor/rtp.py +++ b/yt_dlp/extractor/rtp.py @@ -3,12 +3,20 @@ import json import re import urllib.parse -from .common import InfoExtractor -from ..utils import js_to_json +from .common import InfoExtractor, Request +from ..utils import ( + determine_ext, + int_or_none, + js_to_json, + parse_duration, + parse_iso8601, + url_or_none, +) +from ..utils.traversal import traverse_obj class RTPIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.)?rtp\.pt/play/(?:(?:estudoemcasa|palco|zigzag)/)?p(?P<program_id>[0-9]+)/(?P<id>[^/?#]+)' + _VALID_URL = r'https?://(?:www\.)?rtp\.pt/play/(?:[^/#?]+/)?p(?P<program_id>\d+)/(?P<id>e\d+)' _TESTS = [{ 'url': 'http://www.rtp.pt/play/p405/e174042/paixoes-cruzadas', 'md5': 'e736ce0c665e459ddb818546220b4ef8', @@ -16,99 +24,173 @@ class RTPIE(InfoExtractor): 'id': 'e174042', 'ext': 'mp3', 'title': 'Paixões Cruzadas', - 'description': 'As paixões musicais de António Cartaxo e António Macedo', + 'description': 'md5:af979e58ba0ab73f78435fc943fdb070', 'thumbnail': r're:^https?://.*\.jpg', + 'series': 'Paixões Cruzadas', + 'duration': 2950.0, + 'modified_timestamp': 1553693464, + 'modified_date': '20190327', + 'timestamp': 1417219200, + 'upload_date': '20141129', }, }, { 'url': 'https://www.rtp.pt/play/zigzag/p13166/e757904/25-curiosidades-25-de-abril', - 'md5': '9a81ed53f2b2197cfa7ed455b12f8ade', + 'md5': '5b4859940e3adef61247a77dfb76046a', 'info_dict': { 'id': 'e757904', 'ext': 'mp4', - 'title': '25 Curiosidades, 25 de Abril', - 'description': 'Estudar ou não estudar - Em cada um dos episódios descobrimos uma curiosidade acerca de como era viver em Portugal antes da revolução do 25 de abr', + 'title': 'Estudar ou não estudar', + 'description': 'md5:3bfd7eb8bebfd5711a08df69c9c14c35', 'thumbnail': r're:^https?://.*\.jpg', + 'timestamp': 1711958401, + 'duration': 146.0, + 'upload_date': '20240401', + 'modified_timestamp': 1712242991, + 'series': '25 Curiosidades, 25 de Abril', + 'episode_number': 2, + 'episode': 'Estudar ou não estudar', + 'modified_date': '20240404', }, }, { - 'url': 'http://www.rtp.pt/play/p831/a-quimica-das-coisas', - 'only_matching': True, - }, { - 'url': 'https://www.rtp.pt/play/estudoemcasa/p7776/portugues-1-ano', - 'only_matching': True, - }, { - 'url': 'https://www.rtp.pt/play/palco/p13785/l7nnon', - 'only_matching': True, + # Episode not accessible through API + 'url': 'https://www.rtp.pt/play/estudoemcasa/p7776/e500050/portugues-1-ano', + 'md5': '57660c0b46db9f22118c52cbd65975e4', + 'info_dict': { + 'id': 'e500050', + 'ext': 'mp4', + 'title': 'Português - 1.º ano', + 'duration': 1669.0, + 'description': 'md5:be68925c81269f8c6886589f25fe83ea', + 'upload_date': '20201020', + 'timestamp': 1603180799, + 'thumbnail': 'https://cdn-images.rtp.pt/EPG/imagens/39482_59449_64850.png?v=3&w=860', + }, }] + _USER_AGENT = 'rtpplay/2.0.66 (pt.rtp.rtpplay; build:2066; iOS 15.8.3) Alamofire/5.9.1' + _AUTH_TOKEN = None + + def _fetch_auth_token(self): + if self._AUTH_TOKEN: + return self._AUTH_TOKEN + self._AUTH_TOKEN = traverse_obj(self._download_json(Request( + 'https://rtpplayapi.rtp.pt/play/api/2/token-manager', + headers={ + 'Accept': '*/*', + 'rtp-play-auth': 'RTPPLAY_MOBILE_IOS', + 'rtp-play-auth-hash': 'fac9c328b2f27e26e03d7f8942d66c05b3e59371e16c2a079f5c83cc801bd3ee', + 'rtp-play-auth-timestamp': '2145973229682', + 'User-Agent': self._USER_AGENT, + }, extensions={'keep_header_casing': True}), None, + note='Fetching guest auth token', errnote='Could not fetch guest auth token', + fatal=False), ('token', 'token', {str})) + return self._AUTH_TOKEN + + @staticmethod + def _cleanup_media_url(url): + if urllib.parse.urlparse(url).netloc == 'streaming-ondemand.rtp.pt': + return None + return url.replace('/drm-fps/', '/hls/').replace('/drm-dash/', '/dash/') + + def _extract_formats(self, media_urls, episode_id): + formats = [] + subtitles = {} + for media_url in set(traverse_obj(media_urls, (..., {url_or_none}, {self._cleanup_media_url}))): + ext = determine_ext(media_url) + if ext == 'm3u8': + fmts, subs = self._extract_m3u8_formats_and_subtitles( + media_url, episode_id, m3u8_id='hls', fatal=False) + formats.extend(fmts) + self._merge_subtitles(subs, target=subtitles) + elif ext == 'mpd': + fmts, subs = self._extract_mpd_formats_and_subtitles( + media_url, episode_id, mpd_id='dash', fatal=False) + formats.extend(fmts) + self._merge_subtitles(subs, target=subtitles) + else: + formats.append({ + 'url': media_url, + 'format_id': 'http', + }) + return formats, subtitles + + def _extract_from_api(self, program_id, episode_id): + auth_token = self._fetch_auth_token() + if not auth_token: + return + episode_data = traverse_obj(self._download_json( + f'https://www.rtp.pt/play/api/1/get-episode/{program_id}/{episode_id[1:]}', episode_id, + query={'include_assets': 'true', 'include_webparams': 'true'}, + headers={ + 'Accept': '*/*', + 'Authorization': f'Bearer {auth_token}', + 'User-Agent': self._USER_AGENT, + }, fatal=False), 'result', {dict}) + if not episode_data: + return + asset_urls = traverse_obj(episode_data, ('assets', 0, 'asset_url', {dict})) + media_urls = traverse_obj(asset_urls, ( + ((('hls', 'dash'), 'stream_url'), ('multibitrate', ('url_hls', 'url_dash'))),)) + formats, subtitles = self._extract_formats(media_urls, episode_id) + + for sub_data in traverse_obj(asset_urls, ('subtitles', 'vtt_list', lambda _, v: url_or_none(v['file']))): + subtitles.setdefault(sub_data.get('code') or 'pt', []).append({ + 'url': sub_data['file'], + 'name': sub_data.get('language'), + }) + + return { + 'id': episode_id, + 'formats': formats, + 'subtitles': subtitles, + 'thumbnail': traverse_obj(episode_data, ('assets', 0, 'asset_thumbnail', {url_or_none})), + **traverse_obj(episode_data, ('episode', { + 'title': (('episode_title', 'program_title'), {str}, filter, any), + 'alt_title': ('episode_subtitle', {str}, filter), + 'description': (('episode_description', 'episode_summary'), {str}, filter, any), + 'timestamp': ('episode_air_date', {parse_iso8601(delimiter=' ')}), + 'modified_timestamp': ('episode_lastchanged', {parse_iso8601(delimiter=' ')}), + 'duration': ('episode_duration_complete', {parse_duration}), + 'episode': ('episode_title', {str}, filter), + 'episode_number': ('episode_number', {int_or_none}), + 'season': ('program_season', {str}, filter), + 'series': ('program_title', {str}, filter), + })), + } + _RX_OBFUSCATION = re.compile(r'''(?xs) atob\s*\(\s*decodeURIComponent\s*\(\s* (\[[0-9A-Za-z%,'"]*\]) \s*\.\s*join\(\s*(?:""|'')\s*\)\s*\)\s*\) ''') - def __unobfuscate(self, data, *, video_id): - if data.startswith('{'): - data = self._RX_OBFUSCATION.sub( - lambda m: json.dumps( - base64.b64decode(urllib.parse.unquote( - ''.join(self._parse_json(m.group(1), video_id)), - )).decode('iso-8859-1')), - data) - return js_to_json(data) + def __unobfuscate(self, data): + return self._RX_OBFUSCATION.sub( + lambda m: json.dumps( + base64.b64decode(urllib.parse.unquote( + ''.join(json.loads(m.group(1))), + )).decode('iso-8859-1')), + data) - def _real_extract(self, url): - video_id = self._match_id(url) - - webpage = self._download_webpage(url, video_id) - title = self._html_search_meta( - 'twitter:title', webpage, display_name='title', fatal=True) - - f, config = self._search_regex( - r'''(?sx) - (?:var\s+f\s*=\s*(?P<f>".*?"|{[^;]+?});\s*)? - var\s+player1\s+=\s+new\s+RTPPlayer\s*\((?P<config>{(?:(?!\*/).)+?})\);(?!\s*\*/) - ''', webpage, - 'player config', group=('f', 'config')) - - config = self._parse_json( - config, video_id, - lambda data: self.__unobfuscate(data, video_id=video_id)) - f = config['file'] if not f else self._parse_json( - f, video_id, - lambda data: self.__unobfuscate(data, video_id=video_id)) + def _extract_from_html(self, url, episode_id): + webpage = self._download_webpage(url, episode_id) formats = [] - if isinstance(f, dict): - f_hls = f.get('hls') - if f_hls is not None: - formats.extend(self._extract_m3u8_formats( - f_hls, video_id, 'mp4', 'm3u8_native', m3u8_id='hls')) - - f_dash = f.get('dash') - if f_dash is not None: - formats.extend(self._extract_mpd_formats(f_dash, video_id, mpd_id='dash')) - else: - formats.append({ - 'format_id': 'f', - 'url': f, - 'vcodec': 'none' if config.get('mediaType') == 'audio' else None, - }) - subtitles = {} - - vtt = config.get('vtt') - if vtt is not None: - for lcode, lname, url in vtt: - subtitles.setdefault(lcode, []).append({ - 'name': lname, - 'url': url, - }) + media_urls = traverse_obj(re.findall(r'(?:var\s+f\s*=|RTPPlayer\({[^}]+file:)\s*({[^}]+}|"[^"]+")', webpage), ( + -1, (({self.__unobfuscate}, {js_to_json}, {json.loads}, {dict.values}, ...), {json.loads}))) + formats, subtitles = self._extract_formats(media_urls, episode_id) return { - 'id': video_id, - 'title': title, + 'id': episode_id, 'formats': formats, - 'description': self._html_search_meta(['description', 'twitter:description'], webpage), - 'thumbnail': config.get('poster') or self._og_search_thumbnail(webpage), 'subtitles': subtitles, + 'description': self._html_search_meta(['og:description', 'twitter:description'], webpage, default=None), + 'thumbnail': self._html_search_meta(['og:image', 'twitter:image'], webpage, default=None), + **self._search_json_ld(webpage, episode_id, default={}), + 'title': self._html_search_meta(['og:title', 'twitter:title'], webpage, default=None), } + + def _real_extract(self, url): + program_id, episode_id = self._match_valid_url(url).group('program_id', 'id') + return self._extract_from_api(program_id, episode_id) or self._extract_from_html(url, episode_id) |