import re from .common import InfoExtractor from ..utils import ( ExtractorError, GeoRestrictedError, int_or_none, parse_iso8601, parse_qs, strip_or_none, traverse_obj, url_or_none, ) class ArteTVBaseIE(InfoExtractor): _ARTE_LANGUAGES = 'fr|de|en|es|it|pl' _API_BASE = 'https://api.arte.tv/api/player/v2' class ArteTVIE(ArteTVBaseIE): _VALID_URL = r'''(?x) (?:https?:// (?: (?:www\.)?arte\.tv/(?P%(langs)s)/videos| api\.arte\.tv/api/player/v\d+/config/(?P%(langs)s) ) |arte://program) /(?P\d{6}-\d{3}-[AF]|LIVE) ''' % {'langs': ArteTVBaseIE._ARTE_LANGUAGES} _TESTS = [{ 'url': 'https://www.arte.tv/en/videos/088501-000-A/mexico-stealing-petrol-to-survive/', 'only_matching': True, }, { 'url': 'https://www.arte.tv/pl/videos/100103-000-A/usa-dyskryminacja-na-porodowce/', 'info_dict': { 'id': '100103-000-A', 'title': 'USA: Dyskryminacja na porodówce', 'description': 'md5:242017b7cce59ffae340a54baefcafb1', 'alt_title': 'ARTE Reportage', 'upload_date': '20201103', 'duration': 554, 'thumbnail': r're:https://api-cdn\.arte\.tv/.+940x530', 'timestamp': 1604417980, 'ext': 'mp4', }, 'params': {'skip_download': 'm3u8'} }, { 'url': 'https://api.arte.tv/api/player/v2/config/de/100605-013-A', 'only_matching': True, }, { 'url': 'https://api.arte.tv/api/player/v2/config/de/LIVE', 'only_matching': True, }] _GEO_BYPASS = True _LANG_MAP = { # ISO639 -> French abbreviations 'fr': 'F', 'de': 'A', 'en': 'E[ANG]', 'es': 'E[ESP]', 'it': 'E[ITA]', 'pl': 'E[POL]', # XXX: probably means mixed; # uses this code for audio that happens to be in Ukrainian, but the manifest uses the ISO code 'mul' (mixed) 'mul': 'EU', } _VERSION_CODE_RE = re.compile(r'''(?x) V (?PO?) (?P[FA]|E\[[A-Z]+\]|EU)? (?PAUD|) (?: (?P-ST) (?PM?) (?P[FA]|E\[[A-Z]+\]|EU) )? ''') # all obtained by exhaustive testing _COUNTRIES_MAP = { 'DE_FR': { 'BL', 'DE', 'FR', 'GF', 'GP', 'MF', 'MQ', 'NC', 'PF', 'PM', 'RE', 'WF', 'YT', }, # with both of the below 'BE' sometimes works, sometimes doesn't 'EUR_DE_FR': { 'AT', 'BL', 'CH', 'DE', 'FR', 'GF', 'GP', 'LI', 'MC', 'MF', 'MQ', 'NC', 'PF', 'PM', 'RE', 'WF', 'YT', }, 'SAT': { 'AD', 'AT', 'AX', 'BG', 'BL', 'CH', 'CY', 'CZ', 'DE', 'DK', 'EE', 'ES', 'FI', 'FR', 'GB', 'GF', 'GR', 'HR', 'HU', 'IE', 'IS', 'IT', 'KN', 'LI', 'LT', 'LU', 'LV', 'MC', 'MF', 'MQ', 'MT', 'NC', 'NL', 'NO', 'PF', 'PL', 'PM', 'PT', 'RE', 'RO', 'SE', 'SI', 'SK', 'SM', 'VA', 'WF', 'YT', }, } def _real_extract(self, url): mobj = self._match_valid_url(url) video_id = mobj.group('id') lang = mobj.group('lang') or mobj.group('lang_2') langauge_code = self._LANG_MAP.get(lang) config = self._download_json(f'{self._API_BASE}/config/{lang}/{video_id}', video_id) geoblocking = traverse_obj(config, ('data', 'attributes', 'restriction', 'geoblocking')) or {} if geoblocking.get('restrictedArea'): raise GeoRestrictedError(f'Video restricted to {geoblocking["code"]!r}', countries=self._COUNTRIES_MAP.get(geoblocking['code'], ('DE', 'FR'))) if not traverse_obj(config, ('data', 'attributes', 'rights')): # Eg: https://www.arte.tv/de/videos/097407-215-A/28-minuten # Eg: https://www.arte.tv/es/videos/104351-002-A/serviteur-du-peuple-1-23 raise ExtractorError( 'Video is not available in this language edition of Arte or broadcast rights expired', expected=True) formats, subtitles = [], {} for stream in config['data']['attributes']['streams']: # official player contains code like `e.get("versions")[0].eStat.ml5` stream_version = stream['versions'][0] stream_version_code = stream_version['eStat']['ml5'] lang_pref = -1 m = self._VERSION_CODE_RE.match(stream_version_code) if m: lang_pref = int(''.join('01'[x] for x in ( m.group('vlang') == langauge_code, # we prefer voice in the requested language not m.group('audio_desc'), # and not the audio description version bool(m.group('original_voice')), # but if voice is not in the requested language, at least choose the original voice m.group('sub_lang') == langauge_code, # if subtitles are present, we prefer them in the requested language not m.group('has_sub'), # but we prefer no subtitles otherwise not m.group('sdh_sub'), # and we prefer not the hard-of-hearing subtitles if there are subtitles ))) if stream['protocol'].startswith('HLS'): fmts, subs = self._extract_m3u8_formats_and_subtitles( stream['url'], video_id=video_id, ext='mp4', m3u8_id=stream_version_code, fatal=False) for fmt in fmts: fmt.update({ 'format_note': f'{stream_version.get("label", "unknown")} [{stream_version.get("shortLabel", "?")}]', 'language_preference': lang_pref, }) formats.extend(fmts) self._merge_subtitles(subs, target=subtitles) elif stream['protocol'] in ('HTTPS', 'RTMP'): formats.append({ 'format_id': f'{stream["protocol"]}-{stream_version_code}', 'url': stream['url'], 'format_note': f'{stream_version.get("label", "unknown")} [{stream_version.get("shortLabel", "?")}]', 'language_preference': lang_pref, # 'ext': 'mp4', # XXX: may or may not be necessary, at least for HTTPS }) else: self.report_warning(f'Skipping stream with unknown protocol {stream["protocol"]}') # TODO: chapters from stream['segments']? # The JS also looks for chapters in config['data']['attributes']['chapters'], # but I am yet to find a video having those self._sort_formats(formats) metadata = config['data']['attributes']['metadata'] return { 'id': metadata['providerId'], 'webpage_url': traverse_obj(metadata, ('link', 'url')), 'title': metadata.get('subtitle'), 'alt_title': metadata.get('title'), 'description': metadata.get('description'), 'duration': traverse_obj(metadata, ('duration', 'seconds')), 'language': metadata.get('language'), 'timestamp': traverse_obj(config, ('data', 'attributes', 'rights', 'begin'), expected_type=parse_iso8601), 'is_live': config['data']['attributes'].get('live', False), 'formats': formats, 'subtitles': subtitles, 'thumbnails': [ {'url': image['url'], 'id': image.get('caption')} for image in metadata.get('images') or [] if url_or_none(image.get('url')) ], } class ArteTVEmbedIE(InfoExtractor): _VALID_URL = r'https?://(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+' _TESTS = [{ 'url': 'https://www.arte.tv/player/v5/index.php?json_url=https%3A%2F%2Fapi.arte.tv%2Fapi%2Fplayer%2Fv2%2Fconfig%2Fde%2F100605-013-A&lang=de&autoplay=true&mute=0100605-013-A', 'info_dict': { 'id': '100605-013-A', 'ext': 'mp4', 'title': 'United we Stream November Lockdown Edition #13', 'description': 'md5:be40b667f45189632b78c1425c7c2ce1', 'upload_date': '20201116', }, 'skip': 'No video available' }, { 'url': 'https://www.arte.tv/player/v3/index.php?json_url=https://api.arte.tv/api/player/v2/config/de/100605-013-A', 'only_matching': True, }] @staticmethod def _extract_urls(webpage): return [url for _, url in re.findall( r'<(?:iframe|script)[^>]+src=(["\'])(?P(?:https?:)?//(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+?)\1', webpage)] def _real_extract(self, url): qs = parse_qs(url) json_url = qs['json_url'][0] video_id = ArteTVIE._match_id(json_url) return self.url_result( json_url, ie=ArteTVIE.ie_key(), video_id=video_id) class ArteTVPlaylistIE(ArteTVBaseIE): _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P%s)/videos/(?PRC-\d{6})' % ArteTVBaseIE._ARTE_LANGUAGES _TESTS = [{ 'url': 'https://www.arte.tv/en/videos/RC-016954/earn-a-living/', 'only_matching': True, }, { 'url': 'https://www.arte.tv/pl/videos/RC-014123/arte-reportage/', 'playlist_mincount': 100, 'info_dict': { 'description': 'md5:84e7bf1feda248bc325ebfac818c476e', 'id': 'RC-014123', 'title': 'ARTE Reportage - najlepsze reportaże', }, }] def _real_extract(self, url): lang, playlist_id = self._match_valid_url(url).group('lang', 'id') playlist = self._download_json( f'{self._API_BASE}/playlist/{lang}/{playlist_id}', playlist_id)['data']['attributes'] entries = [{ '_type': 'url_transparent', 'url': video['config']['url'], 'ie_key': ArteTVIE.ie_key(), 'id': video.get('providerId'), 'title': video.get('title'), 'alt_title': video.get('subtitle'), 'thumbnail': url_or_none(traverse_obj(video, ('mainImage', 'url'))), 'duration': int_or_none(traverse_obj(video, ('duration', 'seconds'))), } for video in traverse_obj(playlist, ('items', lambda _, v: v['config']['url']))] return self.playlist_result(entries, playlist_id, traverse_obj(playlist, ('metadata', 'title')), traverse_obj(playlist, ('metadata', 'description'))) class ArteTVCategoryIE(ArteTVBaseIE): _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P%s)/videos/(?P[\w-]+(?:/[\w-]+)*)/?\s*$' % ArteTVBaseIE._ARTE_LANGUAGES _TESTS = [{ 'url': 'https://www.arte.tv/en/videos/politics-and-society/', 'info_dict': { 'id': 'politics-and-society', 'title': 'Politics and society', 'description': 'Investigative documentary series, geopolitical analysis, and international commentary', }, 'playlist_mincount': 13, }] @classmethod def suitable(cls, url): return ( not any(ie.suitable(url) for ie in (ArteTVIE, ArteTVPlaylistIE, )) and super().suitable(url)) def _real_extract(self, url): lang, playlist_id = self._match_valid_url(url).groups() webpage = self._download_webpage(url, playlist_id) items = [] for video in re.finditer( r']*?href\s*=\s*(?P"|\'|\b)(?Phttps?://www\.arte\.tv/%s/videos/[\w/-]+)(?P=q)' % lang, webpage): video = video.group('url') if video == url: continue if any(ie.suitable(video) for ie in (ArteTVIE, ArteTVPlaylistIE, )): items.append(video) title = (self._og_search_title(webpage, default=None) or self._html_search_regex(r']*>([^<]+)', default=None)) title = strip_or_none(title.rsplit('|', 1)[0]) or self._generic_title(url) return self.playlist_from_matches(items, playlist_id=playlist_id, playlist_title=title, description=self._og_search_description(webpage, default=None))