X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/af478477605bdf3f5d57562035885cfee905f379..87046f03007c9b8a95c6297423e5660f795b186d:/youtube_dl/extractor/ard.py diff --git a/youtube_dl/extractor/ard.py b/youtube_dl/extractor/ard.py index b88f71b..86951d9 100644 --- a/youtube_dl/extractor/ard.py +++ b/youtube_dl/extractor/ard.py @@ -4,80 +4,281 @@ from __future__ import unicode_literals import re from .common import InfoExtractor +from .generic import GenericIE +from ..compat import compat_str from ..utils import ( determine_ext, ExtractorError, + qualities, + int_or_none, + parse_duration, + unified_strdate, + xpath_text, + update_url_query, ) +from ..compat import compat_etree_fromstring -class ARDIE(InfoExtractor): - _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P[^/\?]+)(?:\?.*)?' +class ARDMediathekIE(InfoExtractor): + IE_NAME = 'ARD:mediathek' + _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de)/(?:.*/)(?P[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?' - _TEST = { - 'url': 'http://www.ardmediathek.de/das-erste/guenther-jauch/edward-snowden-im-interview-held-oder-verraeter?documentId=19288786', - 'file': '19288786.mp4', - 'md5': '515bf47ce209fb3f5a61b7aad364634c', + _TESTS = [{ + # available till 26.07.2022 + 'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822', 'info_dict': { - 'title': 'Edward Snowden im Interview - Held oder Verräter?', - 'description': 'Edward Snowden hat alles aufs Spiel gesetzt, um die weltweite \xdcberwachung durch die Geheimdienste zu enttarnen. Nun stellt sich der ehemalige NSA-Mitarbeiter erstmals weltweit in einem TV-Interview den Fragen eines NDR-Journalisten. Die Sendung vom Sonntagabend.', - 'thumbnail': 'http://www.ardmediathek.de/ard/servlet/contentblob/19/28/87/90/19288790/bild/2250037', + 'id': '44726822', + 'ext': 'mp4', + 'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?', + 'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5', + 'duration': 1740, }, - 'skip': 'Blocked outside of Germany', - } + 'params': { + # m3u8 download + 'skip_download': True, + } + }, { + # audio + 'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086', + 'only_matching': True, + }, { + 'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht', + 'only_matching': True, + }, { + # audio + 'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158', + 'only_matching': True, + }] + + def _extract_media_info(self, media_info_url, webpage, video_id): + media_info = self._download_json( + media_info_url, video_id, 'Downloading media JSON') + + formats = self._extract_formats(media_info, video_id) + + if not formats: + if '"fsk"' in webpage: + raise ExtractorError( + 'This video is only available after 20:00', expected=True) + elif media_info.get('_geoblocked'): + raise ExtractorError('This video is not available due to geo restriction', expected=True) + + self._sort_formats(formats) + + duration = int_or_none(media_info.get('_duration')) + thumbnail = media_info.get('_previewImage') + is_live = media_info.get('_isLive') is True + + subtitles = {} + subtitle_url = media_info.get('_subtitleUrl') + if subtitle_url: + subtitles['de'] = [{ + 'ext': 'ttml', + 'url': subtitle_url, + }] + + return { + 'id': video_id, + 'duration': duration, + 'thumbnail': thumbnail, + 'is_live': is_live, + 'formats': formats, + 'subtitles': subtitles, + } + + def _extract_formats(self, media_info, video_id): + type_ = media_info.get('_type') + media_array = media_info.get('_mediaArray', []) + formats = [] + for num, media in enumerate(media_array): + for stream in media.get('_mediaStreamArray', []): + stream_urls = stream.get('_stream') + if not stream_urls: + continue + if not isinstance(stream_urls, list): + stream_urls = [stream_urls] + quality = stream.get('_quality') + server = stream.get('_server') + for stream_url in stream_urls: + if not isinstance(stream_url, compat_str) or '//' not in stream_url: + continue + ext = determine_ext(stream_url) + if quality != 'auto' and ext in ('f4m', 'm3u8'): + continue + if ext == 'f4m': + formats.extend(self._extract_f4m_formats( + update_url_query(stream_url, { + 'hdcore': '3.1.1', + 'plugin': 'aasp-3.1.1.69.124' + }), + video_id, f4m_id='hds', fatal=False)) + elif ext == 'm3u8': + formats.extend(self._extract_m3u8_formats( + stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False)) + else: + if server and server.startswith('rtmp'): + f = { + 'url': server, + 'play_path': stream_url, + 'format_id': 'a%s-rtmp-%s' % (num, quality), + } + else: + f = { + 'url': stream_url, + 'format_id': 'a%s-%s-%s' % (num, ext, quality) + } + m = re.search(r'_(?P\d+)x(?P\d+)\.mp4$', stream_url) + if m: + f.update({ + 'width': int(m.group('width')), + 'height': int(m.group('height')), + }) + if type_ == 'audio': + f['vcodec'] = 'none' + formats.append(f) + return formats def _real_extract(self, url): # determine video id from url m = re.match(self._VALID_URL, url) + document_id = None + numid = re.search(r'documentId=([0-9]+)', url) if numid: - video_id = numid.group(1) + document_id = video_id = numid.group(1) else: video_id = m.group('video_id') webpage = self._download_webpage(url, video_id) + ERRORS = ( + ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'), + ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<', + 'Video %s is no longer available'), + ) + + for pattern, message in ERRORS: + if pattern in webpage: + raise ExtractorError(message % video_id, expected=True) + + if re.search(r'[\?&]rss($|[=&])', url): + doc = compat_etree_fromstring(webpage.encode('utf-8')) + if doc.tag == 'rss': + return GenericIE()._extract_rss(url, video_id, doc) + title = self._html_search_regex( - r'(.*?)', webpage, 'title') + [r'(.*?)', + r'', + r'

(.*?)

'], + webpage, 'title') description = self._html_search_meta( - 'dcterms.abstract', webpage, 'description') - thumbnail = self._og_search_thumbnail(webpage) - - streams = [ - mo.groupdict() - for mo in re.finditer( - r'mediaCollection\.addMediaStream\((?P\d+), (?P\d+), "(?P[^"]*)", "(?P[^"]*)", "[^"]*"\)', webpage)] - if not streams: - if '"fsk"' in webpage: - raise ExtractorError('This video is only available after 20:00') + 'dcterms.abstract', webpage, 'description', default=None) + if description is None: + description = self._html_search_meta( + 'description', webpage, 'meta description') - formats = [] - for s in streams: - format = { - 'quality': int(s['quality']), + # Thumbnail is sometimes not present. + # It is in the mobile version, but that seems to use a different URL + # structure altogether. + thumbnail = self._og_search_thumbnail(webpage, default=None) + + media_streams = re.findall(r'''(?x) + mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s* + "([^"]+)"''', webpage) + + if media_streams: + QUALITIES = qualities(['lo', 'hi', 'hq']) + formats = [] + for furl in set(media_streams): + if furl.endswith('.f4m'): + fid = 'f4m' + else: + fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl) + fid = fid_m.group(1) if fid_m else None + formats.append({ + 'quality': QUALITIES(fid), + 'format_id': fid, + 'url': furl, + }) + self._sort_formats(formats) + info = { + 'formats': formats, } - if s.get('rtmp_url'): - format['protocol'] = 'rtmp' - format['url'] = s['rtmp_url'] - format['playpath'] = s['video_url'] - else: - format['url'] = s['video_url'] + else: # request JSON file + if not document_id: + video_id = self._search_regex( + r'/play/(?:config|media)/(\d+)', webpage, 'media id') + info = self._extract_media_info( + 'http://www.ardmediathek.de/play/media/%s' % video_id, + webpage, video_id) - quality_name = self._search_regex( - r'[,.]([a-zA-Z0-9_-]+),?\.mp4', format['url'], - 'quality name', default='NA') - format['format_id'] = '%s-%s-%s-%s' % ( - determine_ext(format['url']), quality_name, s['media_type'], - s['quality']) + info.update({ + 'id': video_id, + 'title': self._live_title(title) if info.get('is_live') else title, + 'description': description, + 'thumbnail': thumbnail, + }) + + return info - formats.append(format) +class ARDIE(InfoExtractor): + _VALID_URL = r'(?Phttps?://(www\.)?daserste\.de/[^?#]+/videos/(?P[^/?#]+)-(?P[0-9]+))\.html' + _TESTS = [{ + # available till 14.02.2019 + 'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html', + 'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49', + 'info_dict': { + 'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video', + 'id': '102', + 'ext': 'mp4', + 'duration': 4435.0, + 'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?', + 'upload_date': '20180214', + 'thumbnail': r're:^https?://.*\.jpg$', + }, + }, { + 'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html', + 'only_matching': True, + }] + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + display_id = mobj.group('display_id') + + player_url = mobj.group('mainurl') + '~playerXml.xml' + doc = self._download_xml(player_url, display_id) + video_node = doc.find('./video') + upload_date = unified_strdate(xpath_text( + video_node, './broadcastDate')) + thumbnail = xpath_text(video_node, './/teaserImage//variant/url') + + formats = [] + for a in video_node.findall('.//asset'): + f = { + 'format_id': a.attrib['type'], + 'width': int_or_none(a.find('./frameWidth').text), + 'height': int_or_none(a.find('./frameHeight').text), + 'vbr': int_or_none(a.find('./bitrateVideo').text), + 'abr': int_or_none(a.find('./bitrateAudio').text), + 'vcodec': a.find('./codecVideo').text, + 'tbr': int_or_none(a.find('./totalBitrate').text), + } + if a.find('./serverPrefix').text: + f['url'] = a.find('./serverPrefix').text + f['playpath'] = a.find('./fileName').text + else: + f['url'] = a.find('./fileName').text + formats.append(f) self._sort_formats(formats) return { - 'id': video_id, - 'title': title, - 'description': description, + 'id': mobj.group('id'), 'formats': formats, + 'display_id': display_id, + 'title': video_node.find('./title').text, + 'duration': parse_duration(video_node.find('./duration').text), + 'upload_date': upload_date, 'thumbnail': thumbnail, }