[yt-dlp.git] / youtube_dl / extractor / ard.py

# coding: utf-8
from __future__ import unicode_literals

import re

from .common import InfoExtractor
from .generic import GenericIE
from ..utils import (
    determine_ext,
    ExtractorError,
    int_or_none,
    parse_duration,
    qualities,
    str_or_none,
    try_get,
    unified_strdate,
    unified_timestamp,
    update_url_query,
    url_or_none,
    xpath_text,
)
from ..compat import compat_etree_fromstring


class ARDMediathekIE(InfoExtractor):
    IE_NAME = 'ARD:mediathek'
    _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'

    _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': {
            'id': '44726822',
            'ext': 'mp4',
            'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
            'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
            'duration': 1740,
        },
        'params': {
            # m3u8 download
            'skip_download': True,
        }
    }, {
        'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
        'only_matching': 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,
    }, {
        'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
        'only_matching': True,
    }]

    @classmethod
    def suitable(cls, url):
        return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)

    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 url_or_none(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<width>\d+)x(?P<height>\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:
            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'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
             r'<meta name="dcterms\.title" content="(.*?)"/>',
             r'<h4 class="headline">(.*?)</h4>',
             r'<title[^>]*>(.*?)</title>'],
            webpage, 'title')
        description = self._html_search_meta(
            'dcterms.abstract', webpage, 'description', default=None)
        if description is None:
            description = self._html_search_meta(
                'description', webpage, 'meta description', default=None)
        if description is None:
            description = self._html_search_regex(
                r'<p\s+class="teasertext">(.+?)</p>',
                webpage, 'teaser text', default=None)

        # 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,
            }
        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)

        info.update({
            'id': video_id,
            'title': self._live_title(title) if info.get('is_live') else title,
            'description': description,
            'thumbnail': thumbnail,
        })

        return info


class ARDIE(InfoExtractor):
    _VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[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': 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,
        }


class ARDBetaMediathekIE(InfoExtractor):
    _VALID_URL = r'https://(?:beta|www)\.ardmediathek\.de/[^/]+/(?:player|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
    _TESTS = [{
        'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
        'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
        'info_dict': {
            'display_id': 'die-robuste-roswita',
            'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
            'title': 'Tatort: Die robuste Roswita',
            'description': r're:^Der Mord.*trüber ist als die Ilm.',
            'duration': 5316,
            'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
            'upload_date': '20180826',
            'ext': 'mp4',
        },
    }, {
        'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
        'only_matching': True,
    }, {
        'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
        'only_matching': True,
    }]

    def _real_extract(self, url):
        mobj = re.match(self._VALID_URL, url)
        video_id = mobj.group('video_id')
        display_id = mobj.group('display_id') or video_id

        webpage = self._download_webpage(url, display_id)
        data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
        data = self._parse_json(data_json, display_id)

        res = {
            'id': video_id,
            'display_id': display_id,
        }
        formats = []
        subtitles = {}
        geoblocked = False
        for widget in data.values():
            if widget.get('_geoblocked') is True:
                geoblocked = True
            if '_duration' in widget:
                res['duration'] = int_or_none(widget['_duration'])
            if 'clipTitle' in widget:
                res['title'] = widget['clipTitle']
            if '_previewImage' in widget:
                res['thumbnail'] = widget['_previewImage']
            if 'broadcastedOn' in widget:
                res['timestamp'] = unified_timestamp(widget['broadcastedOn'])
            if 'synopsis' in widget:
                res['description'] = widget['synopsis']
            subtitle_url = url_or_none(widget.get('_subtitleUrl'))
            if subtitle_url:
                subtitles.setdefault('de', []).append({
                    'ext': 'ttml',
                    'url': subtitle_url,
                })
            if '_quality' in widget:
                format_url = url_or_none(try_get(
                    widget, lambda x: x['_stream']['json'][0]))
                if not format_url:
                    continue
                ext = determine_ext(format_url)
                if ext == 'f4m':
                    formats.extend(self._extract_f4m_formats(
                        format_url + '?hdcore=3.11.0',
                        video_id, f4m_id='hds', fatal=False))
                elif ext == 'm3u8':
                    formats.extend(self._extract_m3u8_formats(
                        format_url, video_id, 'mp4', m3u8_id='hls',
                        fatal=False))
                else:
                    # HTTP formats are not available when geoblocked is True,
                    # other formats are fine though
                    if geoblocked:
                        continue
                    quality = str_or_none(widget.get('_quality'))
                    formats.append({
                        'format_id': ('http-' + quality) if quality else 'http',
                        'url': format_url,
                        'preference': 10,  # Plain HTTP, that's nice
                    })

        if not formats and geoblocked:
            self.raise_geo_restricted(
                msg='This video is not available due to geoblocking',
                countries=['DE'])

        self._sort_formats(formats)
        res.update({
            'subtitles': subtitles,
            'formats': formats,
        })

        return res
Commit	Line	Data
f9b85496 PH	1	# coding: utf-8
	2	from __future__ import unicode_literals
	3
d5822b96 PH	4	import re
	5
	6	from .common import InfoExtractor
3741302a	7	from .generic import GenericIE
d5822b96	8	from ..utils import (
f9b85496	9	determine_ext,
d5822b96	10	ExtractorError,
6d3d3fc0 PH	11	int_or_none,
6d3d3fc0 PH	12	parse_duration,
75258218 S	13	qualities,
	14	str_or_none,
	15	try_get,
6d3d3fc0	16	unified_strdate,
75258218	17	unified_timestamp,
31eeab9f	18	update_url_query,
3052a30d	19	url_or_none,
75258218	20	xpath_text,
d5822b96	21	)
f7854627	22	from ..compat import compat_etree_fromstring
d5822b96	23
f9b85496	24
6d3d3fc0 PH	25	class ARDMediathekIE(InfoExtractor):
6d3d3fc0 PH	26	IE_NAME = 'ARD:mediathek'
59c39401	27	_VALID_URL = r'^https?://(?:(?:(?:www\|classic)\.)?ardmediathek\.de\|mediathek\.(?:daserste\|rbb-online)\.de\|one\.ard\.de)/(?:./)(?P<video_id>[0-9]+\|[^0-9][^/\?]+)[^/\?](?:\?.*)?'
f9b85496	28
29546b34	29	_TESTS = [{
d6a03502	30	# available till 26.07.2022
ad29ef04 W	31	'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
	32	'info_dict': {
	33	'id': '44726822',
	34	'ext': 'mp4',
	35	'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
	36	'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
	37	'duration': 1740,
	38	},
	39	'params': {
	40	# m3u8 download
	41	'skip_download': True,
	42	}
0b87e884 L	43	}, {
	44	'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
	45	'only_matching': True,
86b4e98a S	46	}, {
	47	# audio
	48	'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
d6a03502	49	'only_matching': True,
769efa16 S	50	}, {
	51	'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
	52	'only_matching': True,
a66a73ee S	53	}, {
	54	# audio
	55	'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
d6a03502	56	'only_matching': True,
59c39401 S	57	}, {
	58	'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
	59	'only_matching': True,
29546b34	60	}]
d5822b96	61
1c821227 S	62	@classmethod
	63	def suitable(cls, url):
	64	return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
	65
e37c92ec S	66	def _extract_media_info(self, media_info_url, webpage, video_id):
	67	media_info = self._download_json(
	68	media_info_url, video_id, 'Downloading media JSON')
	69
	70	formats = self._extract_formats(media_info, video_id)
	71
	72	if not formats:
	73	if '"fsk"' in webpage:
	74	raise ExtractorError(
	75	'This video is only available after 20:00', expected=True)
	76	elif media_info.get('_geoblocked'):
	77	raise ExtractorError('This video is not available due to geo restriction', expected=True)
	78
	79	self._sort_formats(formats)
	80
	81	duration = int_or_none(media_info.get('_duration'))
	82	thumbnail = media_info.get('_previewImage')
ca127ab2	83	is_live = media_info.get('_isLive') is True
e37c92ec S	84
	85	subtitles = {}
	86	subtitle_url = media_info.get('_subtitleUrl')
	87	if subtitle_url:
	88	subtitles['de'] = [{
ffa2cecf	89	'ext': 'ttml',
e37c92ec S	90	'url': subtitle_url,
	91	}]
	92
	93	return {
	94	'id': video_id,
	95	'duration': duration,
	96	'thumbnail': thumbnail,
ca127ab2	97	'is_live': is_live,
e37c92ec S	98	'formats': formats,
	99	'subtitles': subtitles,
	100	}
	101
	102	def _extract_formats(self, media_info, video_id):
	103	type_ = media_info.get('_type')
	104	media_array = media_info.get('_mediaArray', [])
	105	formats = []
	106	for num, media in enumerate(media_array):
	107	for stream in media.get('_mediaStreamArray', []):
	108	stream_urls = stream.get('_stream')
	109	if not stream_urls:
	110	continue
	111	if not isinstance(stream_urls, list):
	112	stream_urls = [stream_urls]
	113	quality = stream.get('_quality')
	114	server = stream.get('_server')
	115	for stream_url in stream_urls:
3052a30d	116	if not url_or_none(stream_url):
91328f26	117	continue
e37c92ec	118	ext = determine_ext(stream_url)
1fc0b47f	119	if quality != 'auto' and ext in ('f4m', 'm3u8'):
1fc0b47f	120	continue
e37c92ec	121	if ext == 'f4m':
7e5edcfd	122	formats.extend(self._extract_f4m_formats(
31eeab9f RA	123	update_url_query(stream_url, {
	124	'hdcore': '3.1.1',
	125	'plugin': 'aasp-3.1.1.69.124'
	126	}),
	127	video_id, f4m_id='hds', fatal=False))
e37c92ec	128	elif ext == 'm3u8':
7e5edcfd	129	formats.extend(self._extract_m3u8_formats(
31eeab9f	130	stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
e37c92ec S	131	else:
	132	if server and server.startswith('rtmp'):
	133	f = {
	134	'url': server,
	135	'play_path': stream_url,
	136	'format_id': 'a%s-rtmp-%s' % (num, quality),
	137	}
91328f26	138	else:
e37c92ec S	139	f = {
	140	'url': stream_url,
	141	'format_id': 'a%s-%s-%s' % (num, ext, quality)
	142	}
e37c92ec S	143	m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
	144	if m:
	145	f.update({
	146	'width': int(m.group('width')),
	147	'height': int(m.group('height')),
	148	})
	149	if type_ == 'audio':
	150	f['vcodec'] = 'none'
	151	formats.append(f)
	152	return formats
	153
d5822b96 PH	154	def _real_extract(self, url):
	155	# determine video id from url
	156	m = re.match(self._VALID_URL, url)
	157
ca127ab2 S	158	document_id = None
ca127ab2 S	159
d5822b96 PH	160	numid = re.search(r'documentId=([0-9]+)', url)
d5822b96 PH	161	if numid:
ca127ab2	162	document_id = video_id = numid.group(1)
d5822b96 PH	163	else:
	164	video_id = m.group('video_id')
	165
5622f29a	166	webpage = self._download_webpage(url, video_id)
f9b85496	167
3791d84a S	168	ERRORS = (
	169	('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
	170	('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
	171	'Video %s is no longer available'),
3791d84a S	172	)
	173
	174	for pattern, message in ERRORS:
	175	if pattern in webpage:
	176	raise ExtractorError(message % video_id, expected=True)
39aa42ff	177
bfd91588	178	if re.search(r'[\?&]rss($\|[=&])', url):
f7854627	179	doc = compat_etree_fromstring(webpage.encode('utf-8'))
3741302a OE	180	if doc.tag == 'rss':
	181	return GenericIE()._extract_rss(url, video_id, doc)
	182
f9b85496	183	title = self._html_search_regex(
0f97c9a0	184	[r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
197224b7	185	r'<meta name="dcterms\.title" content="(.*?)"/>',
8c587971 AS	186	r'<h4 class="headline">(.*?)</h4>',
8c587971 AS	187	r'<title[^>]>(.?)</title>'],
0f97c9a0	188	webpage, 'title')
f9b85496	189	description = self._html_search_meta(
29546b34 PH	190	'dcterms.abstract', webpage, 'description', default=None)
	191	if description is None:
	192	description = self._html_search_meta(
8c587971 AS	193	'description', webpage, 'meta description', default=None)
	194	if description is None:
	195	description = self._html_search_regex(
	196	r'<p\s+class="teasertext">(.+?)</p>',
	197	webpage, 'teaser text', default=None)
29546b34 PH	198
	199	# Thumbnail is sometimes not present.
	200	# It is in the mobile version, but that seems to use a different URL
	201	# structure altogether.
	202	thumbnail = self._og_search_thumbnail(webpage, default=None)
	203
	204	media_streams = re.findall(r'''(?x)
	205	mediaCollection\.addMediaStream\([0-9]+,\s[0-9]+,\s"[^"]",\s
	206	"([^"]+)"''', webpage)
	207
	208	if media_streams:
	209	QUALITIES = qualities(['lo', 'hi', 'hq'])
	210	formats = []
	211	for furl in set(media_streams):
	212	if furl.endswith('.f4m'):
	213	fid = 'f4m'
	214	else:
	215	fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
	216	fid = fid_m.group(1) if fid_m else None
	217	formats.append({
	218	'quality': QUALITIES(fid),
	219	'format_id': fid,
	220	'url': furl,
	221	})
e37c92ec S	222	self._sort_formats(formats)
	223	info = {
	224	'formats': formats,
	225	}
29546b34	226	else: # request JSON file
ca127ab2 S	227	if not document_id:
	228	video_id = self._search_regex(
	229	r'/play/(?:config\|media)/(\d+)', webpage, 'media id')
e37c92ec	230	info = self._extract_media_info(
ca127ab2 S	231	'http://www.ardmediathek.de/play/media/%s' % video_id,
ca127ab2 S	232	webpage, video_id)
f9b85496	233
e37c92ec	234	info.update({
f9b85496	235	'id': video_id,
ca127ab2	236	'title': self._live_title(title) if info.get('is_live') else title,
f9b85496	237	'description': description,
f9b85496	238	'thumbnail': thumbnail,
e37c92ec S	239	})
	240
	241	return info
6d3d3fc0 PH	242
	243
	244	class ARDIE(InfoExtractor):
25042f73	245	_VALID_URL = r'(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
ad29ef04	246	_TESTS = [{
d6a03502	247	# available till 14.02.2019
ad29ef04 W	248	'url': 'http://www.daserste.de/information/talk/maischberger/videos/das-groko-drama-zerlegen-sich-die-volksparteien-video-102.html',
	249	'md5': '8e4ec85f31be7c7fc08a26cdbc5a1f49',
	250	'info_dict': {
	251	'display_id': 'das-groko-drama-zerlegen-sich-die-volksparteien-video',
	252	'id': '102',
	253	'ext': 'mp4',
	254	'duration': 4435.0,
	255	'title': 'Das GroKo-Drama: Zerlegen sich die Volksparteien?',
	256	'upload_date': '20180214',
	257	'thumbnail': r're:^https?://.*\.jpg$',
	258	},
d6a03502	259	}, {
6d3d3fc0	260	'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
d6a03502	261	'only_matching': True,
ad29ef04	262	}]
6d3d3fc0 PH	263
	264	def _real_extract(self, url):
	265	mobj = re.match(self._VALID_URL, url)
	266	display_id = mobj.group('display_id')
	267
	268	player_url = mobj.group('mainurl') + '~playerXml.xml'
	269	doc = self._download_xml(player_url, display_id)
	270	video_node = doc.find('./video')
bf0ff932 PH	271	upload_date = unified_strdate(xpath_text(
	272	video_node, './broadcastDate'))
	273	thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
6d3d3fc0 PH	274
	275	formats = []
	276	for a in video_node.findall('.//asset'):
	277	f = {
	278	'format_id': a.attrib['type'],
	279	'width': int_or_none(a.find('./frameWidth').text),
	280	'height': int_or_none(a.find('./frameHeight').text),
	281	'vbr': int_or_none(a.find('./bitrateVideo').text),
	282	'abr': int_or_none(a.find('./bitrateAudio').text),
	283	'vcodec': a.find('./codecVideo').text,
	284	'tbr': int_or_none(a.find('./totalBitrate').text),
	285	}
	286	if a.find('./serverPrefix').text:
	287	f['url'] = a.find('./serverPrefix').text
	288	f['playpath'] = a.find('./fileName').text
	289	else:
	290	f['url'] = a.find('./fileName').text
	291	formats.append(f)
	292	self._sort_formats(formats)
	293
	294	return {
	295	'id': mobj.group('id'),
	296	'formats': formats,
	297	'display_id': display_id,
	298	'title': video_node.find('./title').text,
	299	'duration': parse_duration(video_node.find('./duration').text),
	300	'upload_date': upload_date,
	301	'thumbnail': thumbnail,
	302	}
c1a37eb2 PH	303
	304
	305	class ARDBetaMediathekIE(InfoExtractor):
1c821227	306	_VALID_URL = r'https://(?:beta\|www)\.ardmediathek\.de/[^/]+/(?:player\|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
c1a37eb2 PH	307	_TESTS = [{
	308	'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
	309	'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
	310	'info_dict': {
	311	'display_id': 'die-robuste-roswita',
	312	'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
	313	'title': 'Tatort: Die robuste Roswita',
	314	'description': r're:^Der Mord.*trüber ist als die Ilm.',
	315	'duration': 5316,
	316	'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
	317	'upload_date': '20180826',
	318	'ext': 'mp4',
	319	},
1c821227 S	320	}, {
	321	'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
	322	'only_matching': True,
	323	}, {
	324	'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
	325	'only_matching': True,
c1a37eb2 PH	326	}]
	327
	328	def _real_extract(self, url):
	329	mobj = re.match(self._VALID_URL, url)
	330	video_id = mobj.group('video_id')
1c821227	331	display_id = mobj.group('display_id') or video_id
c1a37eb2 PH	332
c1a37eb2 PH	333	webpage = self._download_webpage(url, display_id)
ed6919e7	334	data_json = self._search_regex(r'window\.__APOLLO_STATE__\s=\s(\{.*);\n', webpage, 'json')
c1a37eb2 PH	335	data = self._parse_json(data_json, display_id)
	336
	337	res = {
	338	'id': video_id,
	339	'display_id': display_id,
	340	}
	341	formats = []
75258218 S	342	subtitles = {}
75258218 S	343	geoblocked = False
c1a37eb2	344	for widget in data.values():
75258218 S	345	if widget.get('_geoblocked') is True:
75258218 S	346	geoblocked = True
c1a37eb2	347	if '_duration' in widget:
75258218	348	res['duration'] = int_or_none(widget['_duration'])
c1a37eb2 PH	349	if 'clipTitle' in widget:
	350	res['title'] = widget['clipTitle']
	351	if '_previewImage' in widget:
	352	res['thumbnail'] = widget['_previewImage']
	353	if 'broadcastedOn' in widget:
75258218	354	res['timestamp'] = unified_timestamp(widget['broadcastedOn'])
c1a37eb2 PH	355	if 'synopsis' in widget:
c1a37eb2 PH	356	res['description'] = widget['synopsis']
75258218 S	357	subtitle_url = url_or_none(widget.get('_subtitleUrl'))
	358	if subtitle_url:
	359	subtitles.setdefault('de', []).append({
c1a37eb2	360	'ext': 'ttml',
75258218 S	361	'url': subtitle_url,
75258218 S	362	})
c1a37eb2	363	if '_quality' in widget:
75258218 S	364	format_url = url_or_none(try_get(
	365	widget, lambda x: x['_stream']['json'][0]))
	366	if not format_url:
	367	continue
	368	ext = determine_ext(format_url)
	369	if ext == 'f4m':
7f2611cb RA	370	formats.extend(self._extract_f4m_formats(
	371	format_url + '?hdcore=3.11.0',
	372	video_id, f4m_id='hds', fatal=False))
75258218	373	elif ext == 'm3u8':
2b83da24	374	formats.extend(self._extract_m3u8_formats(
75258218 S	375	format_url, video_id, 'mp4', m3u8_id='hls',
75258218 S	376	fatal=False))
2b83da24	377	else:
75258218 S	378	# HTTP formats are not available when geoblocked is True,
	379	# other formats are fine though
	380	if geoblocked:
	381	continue
	382	quality = str_or_none(widget.get('_quality'))
2b83da24	383	formats.append({
75258218	384	'format_id': ('http-' + quality) if quality else 'http',
2b83da24 PH	385	'url': format_url,
	386	'preference': 10, # Plain HTTP, that's nice
	387	})
c1a37eb2	388
75258218 S	389	if not formats and geoblocked:
	390	self.raise_geo_restricted(
	391	msg='This video is not available due to geoblocking',
	392	countries=['DE'])
	393
c1a37eb2	394	self._sort_formats(formats)
75258218 S	395	res.update({
	396	'subtitles': subtitles,
	397	'formats': formats,
	398	})
c1a37eb2 PH	399
c1a37eb2 PH	400	return res