yt_dlp/extractor/ard.py

   1 import json
   2 import re
   3
   4 from .common import InfoExtractor
   5 from .generic import GenericIE
   6 from ..utils import (
   7     determine_ext,
   8     ExtractorError,
   9     int_or_none,
  10     parse_duration,
  11     qualities,
  12     str_or_none,
  13     try_get,
  14     unified_strdate,
  15     unified_timestamp,
  16     update_url_query,
  17     url_or_none,
  18     xpath_text,
  19 )
  20 from ..compat import compat_etree_fromstring
  21
  22
  23 class ARDMediathekBaseIE(InfoExtractor):
  24     _GEO_COUNTRIES = ['DE']
  25
  26     def _extract_media_info(self, media_info_url, webpage, video_id):
  27         media_info = self._download_json(
  28             media_info_url, video_id, 'Downloading media JSON')
  29         return self._parse_media_info(media_info, video_id, '"fsk"' in webpage)
  30
  31     def _parse_media_info(self, media_info, video_id, fsk):
  32         formats = self._extract_formats(media_info, video_id)
  33
  34         if not formats:
  35             if fsk:
  36                 self.raise_no_formats(
  37                     'This video is only available after 20:00', expected=True)
  38             elif media_info.get('_geoblocked'):
  39                 self.raise_geo_restricted(
  40                     'This video is not available due to geoblocking',
  41                     countries=self._GEO_COUNTRIES, metadata_available=True)
  42
  43         subtitles = {}
  44         subtitle_url = media_info.get('_subtitleUrl')
  45         if subtitle_url:
  46             subtitles['de'] = [{
  47                 'ext': 'ttml',
  48                 'url': subtitle_url,
  49             }]
  50
  51         return {
  52             'id': video_id,
  53             'duration': int_or_none(media_info.get('_duration')),
  54             'thumbnail': media_info.get('_previewImage'),
  55             'is_live': media_info.get('_isLive') is True,
  56             'formats': formats,
  57             'subtitles': subtitles,
  58         }
  59
  60     def _ARD_extract_episode_info(self, title):
  61         """Try to extract season/episode data from the title."""
  62         res = {}
  63         if not title:
  64             return res
  65
  66         for pattern in [
  67             # Pattern for title like "Homo sapiens (S06/E07) - Originalversion"
  68             # from: https://www.ardmediathek.de/one/sendung/doctor-who/Y3JpZDovL3dkci5kZS9vbmUvZG9jdG9yIHdobw
  69             r'.*(?P<ep_info> \(S(?P<season_number>\d+)/E(?P<episode_number>\d+)\)).*',
  70             # E.g.: title="Fritjof aus Norwegen (2) (AD)"
  71             # from: https://www.ardmediathek.de/ard/sammlung/der-krieg-und-ich/68cMkqJdllm639Skj4c7sS/
  72             r'.*(?P<ep_info> \((?:Folge |Teil )?(?P<episode_number>\d+)(?:/\d+)?\)).*',
  73             r'.*(?P<ep_info>Folge (?P<episode_number>\d+)(?:\:| -|) )\"(?P<episode>.+)\".*',
  74             # E.g.: title="Folge 25/42: Symmetrie"
  75             # from: https://www.ardmediathek.de/ard/video/grips-mathe/folge-25-42-symmetrie/ard-alpha/Y3JpZDovL2JyLmRlL3ZpZGVvLzMyYzI0ZjczLWQ1N2MtNDAxNC05ZmZhLTFjYzRkZDA5NDU5OQ/
  76             # E.g.: title="Folge 1063 - Vertrauen"
  77             # from: https://www.ardmediathek.de/ard/sendung/die-fallers/Y3JpZDovL3N3ci5kZS8yMzAyMDQ4/
  78             r'.*(?P<ep_info>Folge (?P<episode_number>\d+)(?:/\d+)?(?:\:| -|) ).*',
  79         ]:
  80             m = re.match(pattern, title)
  81             if m:
  82                 groupdict = m.groupdict()
  83                 res['season_number'] = int_or_none(groupdict.get('season_number'))
  84                 res['episode_number'] = int_or_none(groupdict.get('episode_number'))
  85                 res['episode'] = str_or_none(groupdict.get('episode'))
  86                 # Build the episode title by removing numeric episode information:
  87                 if groupdict.get('ep_info') and not res['episode']:
  88                     res['episode'] = str_or_none(
  89                         title.replace(groupdict.get('ep_info'), ''))
  90                 if res['episode']:
  91                     res['episode'] = res['episode'].strip()
  92                 break
  93
  94         # As a fallback use the whole title as the episode name:
  95         if not res.get('episode'):
  96             res['episode'] = title.strip()
  97         return res
  98
  99     def _extract_formats(self, media_info, video_id):
 100         type_ = media_info.get('_type')
 101         media_array = media_info.get('_mediaArray', [])
 102         formats = []
 103         for num, media in enumerate(media_array):
 104             for stream in media.get('_mediaStreamArray', []):
 105                 stream_urls = stream.get('_stream')
 106                 if not stream_urls:
 107                     continue
 108                 if not isinstance(stream_urls, list):
 109                     stream_urls = [stream_urls]
 110                 quality = stream.get('_quality')
 111                 server = stream.get('_server')
 112                 for stream_url in stream_urls:
 113                     if not url_or_none(stream_url):
 114                         continue
 115                     ext = determine_ext(stream_url)
 116                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 117                         continue
 118                     if ext == 'f4m':
 119                         formats.extend(self._extract_f4m_formats(
 120                             update_url_query(stream_url, {
 121                                 'hdcore': '3.1.1',
 122                                 'plugin': 'aasp-3.1.1.69.124'
 123                             }), video_id, f4m_id='hds', fatal=False))
 124                     elif ext == 'm3u8':
 125                         formats.extend(self._extract_m3u8_formats(
 126                             stream_url, video_id, 'mp4', 'm3u8_native',
 127                             m3u8_id='hls', fatal=False))
 128                     else:
 129                         if server and server.startswith('rtmp'):
 130                             f = {
 131                                 'url': server,
 132                                 'play_path': stream_url,
 133                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 134                             }
 135                         else:
 136                             f = {
 137                                 'url': stream_url,
 138                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 139                             }
 140                         m = re.search(
 141                             r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$',
 142                             stream_url)
 143                         if m:
 144                             f.update({
 145                                 'width': int(m.group('width')),
 146                                 'height': int(m.group('height')),
 147                             })
 148                         if type_ == 'audio':
 149                             f['vcodec'] = 'none'
 150                         formats.append(f)
 151         return formats
 152
 153
 154 class ARDMediathekIE(ARDMediathekBaseIE):
 155     IE_NAME = 'ARD:mediathek'
 156     _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
 157
 158     _TESTS = [{
 159         # available till 26.07.2022
 160         'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
 161         'info_dict': {
 162             'id': '44726822',
 163             'ext': 'mp4',
 164             'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
 165             'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
 166             'duration': 1740,
 167         },
 168         'params': {
 169             # m3u8 download
 170             'skip_download': True,
 171         }
 172     }, {
 173         'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
 174         'only_matching': True,
 175     }, {
 176         # audio
 177         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
 178         'only_matching': True,
 179     }, {
 180         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
 181         'only_matching': True,
 182     }, {
 183         # audio
 184         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
 185         'only_matching': True,
 186     }, {
 187         'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
 188         'only_matching': True,
 189     }]
 190
 191     @classmethod
 192     def suitable(cls, url):
 193         return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
 194
 195     def _real_extract(self, url):
 196         # determine video id from url
 197         m = self._match_valid_url(url)
 198
 199         document_id = None
 200
 201         numid = re.search(r'documentId=([0-9]+)', url)
 202         if numid:
 203             document_id = video_id = numid.group(1)
 204         else:
 205             video_id = m.group('video_id')
 206
 207         webpage = self._download_webpage(url, video_id)
 208
 209         ERRORS = (
 210             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
 211             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
 212              'Video %s is no longer available'),
 213         )
 214
 215         for pattern, message in ERRORS:
 216             if pattern in webpage:
 217                 raise ExtractorError(message % video_id, expected=True)
 218
 219         if re.search(r'[\?&]rss($|[=&])', url):
 220             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 221             if doc.tag == 'rss':
 222                 return GenericIE()._extract_rss(url, video_id, doc)
 223
 224         title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 225             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 226              r'<meta name="dcterms\.title" content="(.*?)"/>',
 227              r'<h4 class="headline">(.*?)</h4>',
 228              r'<title[^>]*>(.*?)</title>'],
 229             webpage, 'title')
 230         description = self._og_search_description(webpage, default=None) or self._html_search_meta(
 231             'dcterms.abstract', webpage, 'description', default=None)
 232         if description is None:
 233             description = self._html_search_meta(
 234                 'description', webpage, 'meta description', default=None)
 235         if description is None:
 236             description = self._html_search_regex(
 237                 r'<p\s+class="teasertext">(.+?)</p>',
 238                 webpage, 'teaser text', default=None)
 239
 240         # Thumbnail is sometimes not present.
 241         # It is in the mobile version, but that seems to use a different URL
 242         # structure altogether.
 243         thumbnail = self._og_search_thumbnail(webpage, default=None)
 244
 245         media_streams = re.findall(r'''(?x)
 246             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 247             "([^"]+)"''', webpage)
 248
 249         if media_streams:
 250             QUALITIES = qualities(['lo', 'hi', 'hq'])
 251             formats = []
 252             for furl in set(media_streams):
 253                 if furl.endswith('.f4m'):
 254                     fid = 'f4m'
 255                 else:
 256                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 257                     fid = fid_m.group(1) if fid_m else None
 258                 formats.append({
 259                     'quality': QUALITIES(fid),
 260                     'format_id': fid,
 261                     'url': furl,
 262                 })
 263             info = {
 264                 'formats': formats,
 265             }
 266         else:  # request JSON file
 267             if not document_id:
 268                 video_id = self._search_regex(
 269                     (r'/play/(?:config|media|sola)/(\d+)', r'contentId["\']\s*:\s*(\d+)'),
 270                     webpage, 'media id', default=None)
 271             info = self._extract_media_info(
 272                 'http://www.ardmediathek.de/play/media/%s' % video_id,
 273                 webpage, video_id)
 274
 275         info.update({
 276             'id': video_id,
 277             'title': title,
 278             'description': description,
 279             'thumbnail': thumbnail,
 280         })
 281         info.update(self._ARD_extract_episode_info(info['title']))
 282
 283         return info
 284
 285
 286 class ARDIE(InfoExtractor):
 287     _VALID_URL = r'(?P<mainurl>https?://(?:www\.)?daserste\.de/(?:[^/?#&]+/)+(?P<id>[^/?#&]+))\.html'
 288     _TESTS = [{
 289         # available till 7.01.2022
 290         'url': 'https://www.daserste.de/information/talk/maischberger/videos/maischberger-die-woche-video100.html',
 291         'md5': '867d8aa39eeaf6d76407c5ad1bb0d4c1',
 292         'info_dict': {
 293             'id': 'maischberger-die-woche-video100',
 294             'display_id': 'maischberger-die-woche-video100',
 295             'ext': 'mp4',
 296             'duration': 3687.0,
 297             'title': 'maischberger. die woche vom 7. Januar 2021',
 298             'upload_date': '20210107',
 299             'thumbnail': r're:^https?://.*\.jpg$',
 300         },
 301     }, {
 302         'url': 'https://www.daserste.de/information/politik-weltgeschehen/morgenmagazin/videosextern/dominik-kahun-aus-der-nhl-direkt-zur-weltmeisterschaft-100.html',
 303         'only_matching': True,
 304     }, {
 305         'url': 'https://www.daserste.de/information/nachrichten-wetter/tagesthemen/videosextern/tagesthemen-17736.html',
 306         'only_matching': True,
 307     }, {
 308         'url': 'https://www.daserste.de/unterhaltung/serie/in-aller-freundschaft-die-jungen-aerzte/videos/diversity-tag-sanam-afrashteh100.html',
 309         'only_matching': True,
 310     }, {
 311         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 312         'only_matching': True,
 313     }, {
 314         'url': 'https://www.daserste.de/unterhaltung/serie/in-aller-freundschaft-die-jungen-aerzte/Drehpause-100.html',
 315         'only_matching': True,
 316     }, {
 317         'url': 'https://www.daserste.de/unterhaltung/film/filmmittwoch-im-ersten/videos/making-ofwendezeit-video-100.html',
 318         'only_matching': True,
 319     }]
 320
 321     def _real_extract(self, url):
 322         mobj = self._match_valid_url(url)
 323         display_id = mobj.group('id')
 324
 325         player_url = mobj.group('mainurl') + '~playerXml.xml'
 326         doc = self._download_xml(player_url, display_id)
 327         video_node = doc.find('./video')
 328         upload_date = unified_strdate(xpath_text(
 329             video_node, './broadcastDate'))
 330         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 331
 332         formats = []
 333         for a in video_node.findall('.//asset'):
 334             file_name = xpath_text(a, './fileName', default=None)
 335             if not file_name:
 336                 continue
 337             format_type = a.attrib.get('type')
 338             format_url = url_or_none(file_name)
 339             if format_url:
 340                 ext = determine_ext(file_name)
 341                 if ext == 'm3u8':
 342                     formats.extend(self._extract_m3u8_formats(
 343                         format_url, display_id, 'mp4', entry_protocol='m3u8_native',
 344                         m3u8_id=format_type or 'hls', fatal=False))
 345                     continue
 346                 elif ext == 'f4m':
 347                     formats.extend(self._extract_f4m_formats(
 348                         update_url_query(format_url, {'hdcore': '3.7.0'}),
 349                         display_id, f4m_id=format_type or 'hds', fatal=False))
 350                     continue
 351             f = {
 352                 'format_id': format_type,
 353                 'width': int_or_none(xpath_text(a, './frameWidth')),
 354                 'height': int_or_none(xpath_text(a, './frameHeight')),
 355                 'vbr': int_or_none(xpath_text(a, './bitrateVideo')),
 356                 'abr': int_or_none(xpath_text(a, './bitrateAudio')),
 357                 'vcodec': xpath_text(a, './codecVideo'),
 358                 'tbr': int_or_none(xpath_text(a, './totalBitrate')),
 359             }
 360             server_prefix = xpath_text(a, './serverPrefix', default=None)
 361             if server_prefix:
 362                 f.update({
 363                     'url': server_prefix,
 364                     'playpath': file_name,
 365                 })
 366             else:
 367                 if not format_url:
 368                     continue
 369                 f['url'] = format_url
 370             formats.append(f)
 371
 372         _SUB_FORMATS = (
 373             ('./dataTimedText', 'ttml'),
 374             ('./dataTimedTextNoOffset', 'ttml'),
 375             ('./dataTimedTextVtt', 'vtt'),
 376         )
 377
 378         subtitles = {}
 379         for subsel, subext in _SUB_FORMATS:
 380             for node in video_node.findall(subsel):
 381                 subtitles.setdefault('de', []).append({
 382                     'url': node.attrib['url'],
 383                     'ext': subext,
 384                 })
 385
 386         return {
 387             'id': xpath_text(video_node, './videoId', default=display_id),
 388             'formats': formats,
 389             'subtitles': subtitles,
 390             'display_id': display_id,
 391             'title': video_node.find('./title').text,
 392             'duration': parse_duration(video_node.find('./duration').text),
 393             'upload_date': upload_date,
 394             'thumbnail': thumbnail,
 395         }
 396
 397
 398 class ARDBetaMediathekIE(ARDMediathekBaseIE):
 399     _VALID_URL = r'''(?x)https://
 400         (?:(?:beta|www)\.)?ardmediathek\.de/
 401         (?:(?P<client>[^/]+)/)?
 402         (?:player|live|video|(?P<playlist>sendung|sammlung))/
 403         (?:(?P<display_id>(?(playlist)[^?#]+?|[^?#]+))/)?
 404         (?P<id>(?(playlist)|Y3JpZDovL)[a-zA-Z0-9]+)
 405         (?(playlist)/(?P<season>\d+)?/?(?:[?#]|$))'''
 406
 407     _TESTS = [{
 408         'url': 'https://www.ardmediathek.de/mdr/video/die-robuste-roswita/Y3JpZDovL21kci5kZS9iZWl0cmFnL2Ntcy84MWMxN2MzZC0wMjkxLTRmMzUtODk4ZS0wYzhlOWQxODE2NGI/',
 409         'md5': 'a1dc75a39c61601b980648f7c9f9f71d',
 410         'info_dict': {
 411             'display_id': 'die-robuste-roswita',
 412             'id': '78566716',
 413             'title': 'Die robuste Roswita',
 414             'description': r're:^Der Mord.*totgeglaubte Ehefrau Roswita',
 415             'duration': 5316,
 416             'thumbnail': 'https://img.ardmediathek.de/standard/00/78/56/67/84/575672121/16x9/960?mandant=ard',
 417             'timestamp': 1596658200,
 418             'upload_date': '20200805',
 419             'ext': 'mp4',
 420         },
 421         'skip': 'Error',
 422     }, {
 423         'url': 'https://www.ardmediathek.de/video/tagesschau-oder-tagesschau-20-00-uhr/das-erste/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhZ2Vzc2NoYXUvZmM4ZDUxMjgtOTE0ZC00Y2MzLTgzNzAtNDZkNGNiZWJkOTll',
 424         'md5': 'f1837e563323b8a642a8ddeff0131f51',
 425         'info_dict': {
 426             'id': '10049223',
 427             'ext': 'mp4',
 428             'title': 'tagesschau, 20:00 Uhr',
 429             'timestamp': 1636398000,
 430             'description': 'md5:39578c7b96c9fe50afdf5674ad985e6b',
 431             'upload_date': '20211108',
 432         },
 433     }, {
 434         'url': 'https://www.ardmediathek.de/sendung/beforeigners/beforeigners/staffel-1/Y3JpZDovL2Rhc2Vyc3RlLmRlL2JlZm9yZWlnbmVycw/1',
 435         'playlist_count': 6,
 436         'info_dict': {
 437             'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL2JlZm9yZWlnbmVycw',
 438             'title': 'beforeigners/beforeigners/staffel-1',
 439         },
 440     }, {
 441         'url': 'https://beta.ardmediathek.de/ard/video/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
 442         'only_matching': True,
 443     }, {
 444         'url': 'https://ardmediathek.de/ard/video/saartalk/saartalk-gesellschaftsgift-haltung-gegen-hass/sr-fernsehen/Y3JpZDovL3NyLW9ubGluZS5kZS9TVF84MTY4MA/',
 445         'only_matching': True,
 446     }, {
 447         'url': 'https://www.ardmediathek.de/ard/video/trailer/private-eyes-s01-e01/one/Y3JpZDovL3dkci5kZS9CZWl0cmFnLTE1MTgwYzczLWNiMTEtNGNkMS1iMjUyLTg5MGYzOWQxZmQ1YQ/',
 448         'only_matching': True,
 449     }, {
 450         'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
 451         'only_matching': True,
 452     }, {
 453         'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
 454         'only_matching': True,
 455     }, {
 456         # playlist of type 'sendung'
 457         'url': 'https://www.ardmediathek.de/ard/sendung/doctor-who/Y3JpZDovL3dkci5kZS9vbmUvZG9jdG9yIHdobw/',
 458         'only_matching': True,
 459     }, {
 460         # playlist of type 'sammlung'
 461         'url': 'https://www.ardmediathek.de/ard/sammlung/team-muenster/5JpTzLSbWUAK8184IOvEir/',
 462         'only_matching': True,
 463     }, {
 464         'url': 'https://www.ardmediathek.de/video/coronavirus-update-ndr-info/astrazeneca-kurz-lockdown-und-pims-syndrom-81/ndr/Y3JpZDovL25kci5kZS84NzE0M2FjNi0wMWEwLTQ5ODEtOTE5NS1mOGZhNzdhOTFmOTI/',
 465         'only_matching': True,
 466     }, {
 467         'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3dkci5kZS9CZWl0cmFnLWQ2NDJjYWEzLTMwZWYtNGI4NS1iMTI2LTU1N2UxYTcxOGIzOQ/tatort-duo-koeln-leipzig-ihr-kinderlein-kommet',
 468         'only_matching': True,
 469     }]
 470
 471     def _ARD_load_playlist_snipped(self, playlist_id, display_id, client, mode, pageNumber):
 472         """ Query the ARD server for playlist information
 473         and returns the data in "raw" format """
 474         if mode == 'sendung':
 475             graphQL = json.dumps({
 476                 'query': '''{
 477                     showPage(
 478                         client: "%s"
 479                         showId: "%s"
 480                         pageNumber: %d
 481                     ) {
 482                         pagination {
 483                             pageSize
 484                             totalElements
 485                         }
 486                         teasers {        # Array
 487                             mediumTitle
 488                             links { target { id href title } }
 489                             type
 490                         }
 491                     }}''' % (client, playlist_id, pageNumber),
 492             }).encode()
 493         else:  # mode == 'sammlung'
 494             graphQL = json.dumps({
 495                 'query': '''{
 496                     morePage(
 497                         client: "%s"
 498                         compilationId: "%s"
 499                         pageNumber: %d
 500                     ) {
 501                         widget {
 502                             pagination {
 503                                 pageSize
 504                                 totalElements
 505                             }
 506                             teasers {        # Array
 507                                 mediumTitle
 508                                 links { target { id href title } }
 509                                 type
 510                             }
 511                         }
 512                     }}''' % (client, playlist_id, pageNumber),
 513             }).encode()
 514         # Ressources for ARD graphQL debugging:
 515         # https://api-test.ardmediathek.de/public-gateway
 516         show_page = self._download_json(
 517             'https://api.ardmediathek.de/public-gateway',
 518             '[Playlist] %s' % display_id,
 519             data=graphQL,
 520             headers={'Content-Type': 'application/json'})['data']
 521         # align the structure of the returned data:
 522         if mode == 'sendung':
 523             show_page = show_page['showPage']
 524         else:  # mode == 'sammlung'
 525             show_page = show_page['morePage']['widget']
 526         return show_page
 527
 528     def _ARD_extract_playlist(self, url, playlist_id, display_id, client, mode):
 529         """ Collects all playlist entries and returns them as info dict.
 530         Supports playlists of mode 'sendung' and 'sammlung', and also nested
 531         playlists. """
 532         entries = []
 533         pageNumber = 0
 534         while True:  # iterate by pageNumber
 535             show_page = self._ARD_load_playlist_snipped(
 536                 playlist_id, display_id, client, mode, pageNumber)
 537             for teaser in show_page['teasers']:  # process playlist items
 538                 if '/compilation/' in teaser['links']['target']['href']:
 539                     # alternativ cond.: teaser['type'] == "compilation"
 540                     # => This is an nested compilation, e.g. like:
 541                     # https://www.ardmediathek.de/ard/sammlung/die-kirche-bleibt-im-dorf/5eOHzt8XB2sqeFXbIoJlg2/
 542                     link_mode = 'sammlung'
 543                 else:
 544                     link_mode = 'video'
 545
 546                 item_url = 'https://www.ardmediathek.de/%s/%s/%s/%s/%s' % (
 547                     client, link_mode, display_id,
 548                     # perform HTLM quoting of episode title similar to ARD:
 549                     re.sub('^-|-$', '',  # remove '-' from begin/end
 550                            re.sub('[^a-zA-Z0-9]+', '-',  # replace special chars by -
 551                                   teaser['links']['target']['title'].lower()
 552                                   .replace('ä', 'ae').replace('ö', 'oe')
 553                                   .replace('ü', 'ue').replace('ß', 'ss'))),
 554                     teaser['links']['target']['id'])
 555                 entries.append(self.url_result(
 556                     item_url,
 557                     ie=ARDBetaMediathekIE.ie_key()))
 558
 559             if (show_page['pagination']['pageSize'] * (pageNumber + 1)
 560                >= show_page['pagination']['totalElements']):
 561                 # we've processed enough pages to get all playlist entries
 562                 break
 563             pageNumber = pageNumber + 1
 564
 565         return self.playlist_result(entries, playlist_id, playlist_title=display_id)
 566
 567     def _real_extract(self, url):
 568         video_id, display_id, playlist_type, client, season_number = self._match_valid_url(url).group(
 569             'id', 'display_id', 'playlist', 'client', 'season')
 570         display_id, client = display_id or video_id, client or 'ard'
 571
 572         if playlist_type:
 573             # TODO: Extract only specified season
 574             return self._ARD_extract_playlist(url, video_id, display_id, client, playlist_type)
 575
 576         player_page = self._download_json(
 577             'https://api.ardmediathek.de/public-gateway',
 578             display_id, data=json.dumps({
 579                 'query': '''{
 580   playerPage(client:"%s", clipId: "%s") {
 581     blockedByFsk
 582     broadcastedOn
 583     maturityContentRating
 584     mediaCollection {
 585       _duration
 586       _geoblocked
 587       _isLive
 588       _mediaArray {
 589         _mediaStreamArray {
 590           _quality
 591           _server
 592           _stream
 593         }
 594       }
 595       _previewImage
 596       _subtitleUrl
 597       _type
 598     }
 599     show {
 600       title
 601     }
 602     synopsis
 603     title
 604     tracking {
 605       atiCustomVars {
 606         contentId
 607       }
 608     }
 609   }
 610 }''' % (client, video_id),
 611             }).encode(), headers={
 612                 'Content-Type': 'application/json'
 613             })['data']['playerPage']
 614         title = player_page['title']
 615         content_id = str_or_none(try_get(
 616             player_page, lambda x: x['tracking']['atiCustomVars']['contentId']))
 617         media_collection = player_page.get('mediaCollection') or {}
 618         if not media_collection and content_id:
 619             media_collection = self._download_json(
 620                 'https://www.ardmediathek.de/play/media/' + content_id,
 621                 content_id, fatal=False) or {}
 622         info = self._parse_media_info(
 623             media_collection, content_id or video_id,
 624             player_page.get('blockedByFsk'))
 625         age_limit = None
 626         description = player_page.get('synopsis')
 627         maturity_content_rating = player_page.get('maturityContentRating')
 628         if maturity_content_rating:
 629             age_limit = int_or_none(maturity_content_rating.lstrip('FSK'))
 630         if not age_limit and description:
 631             age_limit = int_or_none(self._search_regex(
 632                 r'\(FSK\s*(\d+)\)\s*$', description, 'age limit', default=None))
 633         info.update({
 634             'age_limit': age_limit,
 635             'display_id': display_id,
 636             'title': title,
 637             'description': description,
 638             'timestamp': unified_timestamp(player_page.get('broadcastedOn')),
 639             'series': try_get(player_page, lambda x: x['show']['title']),
 640         })
 641         info.update(self._ARD_extract_episode_info(info['title']))
 642         return info