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         self._sort_formats(formats)
  44
  45         subtitles = {}
  46         subtitle_url = media_info.get('_subtitleUrl')
  47         if subtitle_url:
  48             subtitles['de'] = [{
  49                 'ext': 'ttml',
  50                 'url': subtitle_url,
  51             }]
  52
  53         return {
  54             'id': video_id,
  55             'duration': int_or_none(media_info.get('_duration')),
  56             'thumbnail': media_info.get('_previewImage'),
  57             'is_live': media_info.get('_isLive') is True,
  58             'formats': formats,
  59             'subtitles': subtitles,
  60         }
  61
  62     def _ARD_extract_episode_info(self, title):
  63         """Try to extract season/episode data from the title."""
  64         res = {}
  65         if not title:
  66             return res
  67
  68         for pattern in [
  69             # Pattern for title like "Homo sapiens (S06/E07) - Originalversion"
  70             # from: https://www.ardmediathek.de/one/sendung/doctor-who/Y3JpZDovL3dkci5kZS9vbmUvZG9jdG9yIHdobw
  71             r'.*(?P<ep_info> \(S(?P<season_number>\d+)/E(?P<episode_number>\d+)\)).*',
  72             # E.g.: title="Fritjof aus Norwegen (2) (AD)"
  73             # from: https://www.ardmediathek.de/ard/sammlung/der-krieg-und-ich/68cMkqJdllm639Skj4c7sS/
  74             r'.*(?P<ep_info> \((?:Folge |Teil )?(?P<episode_number>\d+)(?:/\d+)?\)).*',
  75             r'.*(?P<ep_info>Folge (?P<episode_number>\d+)(?:\:| -|) )\"(?P<episode>.+)\".*',
  76             # E.g.: title="Folge 25/42: Symmetrie"
  77             # from: https://www.ardmediathek.de/ard/video/grips-mathe/folge-25-42-symmetrie/ard-alpha/Y3JpZDovL2JyLmRlL3ZpZGVvLzMyYzI0ZjczLWQ1N2MtNDAxNC05ZmZhLTFjYzRkZDA5NDU5OQ/
  78             # E.g.: title="Folge 1063 - Vertrauen"
  79             # from: https://www.ardmediathek.de/ard/sendung/die-fallers/Y3JpZDovL3N3ci5kZS8yMzAyMDQ4/
  80             r'.*(?P<ep_info>Folge (?P<episode_number>\d+)(?:/\d+)?(?:\:| -|) ).*',
  81         ]:
  82             m = re.match(pattern, title)
  83             if m:
  84                 groupdict = m.groupdict()
  85                 res['season_number'] = int_or_none(groupdict.get('season_number'))
  86                 res['episode_number'] = int_or_none(groupdict.get('episode_number'))
  87                 res['episode'] = str_or_none(groupdict.get('episode'))
  88                 # Build the episode title by removing numeric episode information:
  89                 if groupdict.get('ep_info') and not res['episode']:
  90                     res['episode'] = str_or_none(
  91                         title.replace(groupdict.get('ep_info'), ''))
  92                 if res['episode']:
  93                     res['episode'] = res['episode'].strip()
  94                 break
  95
  96         # As a fallback use the whole title as the episode name:
  97         if not res.get('episode'):
  98             res['episode'] = title.strip()
  99         return res
 100
 101     def _extract_formats(self, media_info, video_id):
 102         type_ = media_info.get('_type')
 103         media_array = media_info.get('_mediaArray', [])
 104         formats = []
 105         for num, media in enumerate(media_array):
 106             for stream in media.get('_mediaStreamArray', []):
 107                 stream_urls = stream.get('_stream')
 108                 if not stream_urls:
 109                     continue
 110                 if not isinstance(stream_urls, list):
 111                     stream_urls = [stream_urls]
 112                 quality = stream.get('_quality')
 113                 server = stream.get('_server')
 114                 for stream_url in stream_urls:
 115                     if not url_or_none(stream_url):
 116                         continue
 117                     ext = determine_ext(stream_url)
 118                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 119                         continue
 120                     if ext == 'f4m':
 121                         formats.extend(self._extract_f4m_formats(
 122                             update_url_query(stream_url, {
 123                                 'hdcore': '3.1.1',
 124                                 'plugin': 'aasp-3.1.1.69.124'
 125                             }), video_id, f4m_id='hds', fatal=False))
 126                     elif ext == 'm3u8':
 127                         formats.extend(self._extract_m3u8_formats(
 128                             stream_url, video_id, 'mp4', 'm3u8_native',
 129                             m3u8_id='hls', fatal=False))
 130                     else:
 131                         if server and server.startswith('rtmp'):
 132                             f = {
 133                                 'url': server,
 134                                 'play_path': stream_url,
 135                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 136                             }
 137                         else:
 138                             f = {
 139                                 'url': stream_url,
 140                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 141                             }
 142                         m = re.search(
 143                             r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$',
 144                             stream_url)
 145                         if m:
 146                             f.update({
 147                                 'width': int(m.group('width')),
 148                                 'height': int(m.group('height')),
 149                             })
 150                         if type_ == 'audio':
 151                             f['vcodec'] = 'none'
 152                         formats.append(f)
 153         return formats
 154
 155
 156 class ARDMediathekIE(ARDMediathekBaseIE):
 157     IE_NAME = 'ARD:mediathek'
 158     _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
 159
 160     _TESTS = [{
 161         # available till 26.07.2022
 162         'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
 163         'info_dict': {
 164             'id': '44726822',
 165             'ext': 'mp4',
 166             'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
 167             'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
 168             'duration': 1740,
 169         },
 170         'params': {
 171             # m3u8 download
 172             'skip_download': True,
 173         }
 174     }, {
 175         'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
 176         'only_matching': True,
 177     }, {
 178         # audio
 179         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
 180         'only_matching': True,
 181     }, {
 182         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
 183         'only_matching': True,
 184     }, {
 185         # audio
 186         'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
 187         'only_matching': True,
 188     }, {
 189         'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
 190         'only_matching': True,
 191     }]
 192
 193     @classmethod
 194     def suitable(cls, url):
 195         return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
 196
 197     def _real_extract(self, url):
 198         # determine video id from url
 199         m = self._match_valid_url(url)
 200
 201         document_id = None
 202
 203         numid = re.search(r'documentId=([0-9]+)', url)
 204         if numid:
 205             document_id = video_id = numid.group(1)
 206         else:
 207             video_id = m.group('video_id')
 208
 209         webpage = self._download_webpage(url, video_id)
 210
 211         ERRORS = (
 212             ('>Leider liegt eine Störung vor.', 'Video %s is unavailable'),
 213             ('>Der gewünschte Beitrag ist nicht mehr verfügbar.<',
 214              'Video %s is no longer available'),
 215         )
 216
 217         for pattern, message in ERRORS:
 218             if pattern in webpage:
 219                 raise ExtractorError(message % video_id, expected=True)
 220
 221         if re.search(r'[\?&]rss($|[=&])', url):
 222             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 223             if doc.tag == 'rss':
 224                 return GenericIE()._extract_rss(url, video_id, doc)
 225
 226         title = self._og_search_title(webpage, default=None) or self._html_search_regex(
 227             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 228              r'<meta name="dcterms\.title" content="(.*?)"/>',
 229              r'<h4 class="headline">(.*?)</h4>',
 230              r'<title[^>]*>(.*?)</title>'],
 231             webpage, 'title')
 232         description = self._og_search_description(webpage, default=None) or self._html_search_meta(
 233             'dcterms.abstract', webpage, 'description', default=None)
 234         if description is None:
 235             description = self._html_search_meta(
 236                 'description', webpage, 'meta description', default=None)
 237         if description is None:
 238             description = self._html_search_regex(
 239                 r'<p\s+class="teasertext">(.+?)</p>',
 240                 webpage, 'teaser text', default=None)
 241
 242         # Thumbnail is sometimes not present.
 243         # It is in the mobile version, but that seems to use a different URL
 244         # structure altogether.
 245         thumbnail = self._og_search_thumbnail(webpage, default=None)
 246
 247         media_streams = re.findall(r'''(?x)
 248             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 249             "([^"]+)"''', webpage)
 250
 251         if media_streams:
 252             QUALITIES = qualities(['lo', 'hi', 'hq'])
 253             formats = []
 254             for furl in set(media_streams):
 255                 if furl.endswith('.f4m'):
 256                     fid = 'f4m'
 257                 else:
 258                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 259                     fid = fid_m.group(1) if fid_m else None
 260                 formats.append({
 261                     'quality': QUALITIES(fid),
 262                     'format_id': fid,
 263                     'url': furl,
 264                 })
 265             self._sort_formats(formats)
 266             info = {
 267                 'formats': formats,
 268             }
 269         else:  # request JSON file
 270             if not document_id:
 271                 video_id = self._search_regex(
 272                     (r'/play/(?:config|media|sola)/(\d+)', r'contentId["\']\s*:\s*(\d+)'),
 273                     webpage, 'media id', default=None)
 274             info = self._extract_media_info(
 275                 'http://www.ardmediathek.de/play/media/%s' % video_id,
 276                 webpage, video_id)
 277
 278         info.update({
 279             'id': video_id,
 280             'title': title,
 281             'description': description,
 282             'thumbnail': thumbnail,
 283         })
 284         info.update(self._ARD_extract_episode_info(info['title']))
 285
 286         return info
 287
 288
 289 class ARDIE(InfoExtractor):
 290     _VALID_URL = r'(?P<mainurl>https?://(?:www\.)?daserste\.de/(?:[^/?#&]+/)+(?P<id>[^/?#&]+))\.html'
 291     _TESTS = [{
 292         # available till 7.01.2022
 293         'url': 'https://www.daserste.de/information/talk/maischberger/videos/maischberger-die-woche-video100.html',
 294         'md5': '867d8aa39eeaf6d76407c5ad1bb0d4c1',
 295         'info_dict': {
 296             'id': 'maischberger-die-woche-video100',
 297             'display_id': 'maischberger-die-woche-video100',
 298             'ext': 'mp4',
 299             'duration': 3687.0,
 300             'title': 'maischberger. die woche vom 7. Januar 2021',
 301             'upload_date': '20210107',
 302             'thumbnail': r're:^https?://.*\.jpg$',
 303         },
 304     }, {
 305         'url': 'https://www.daserste.de/information/politik-weltgeschehen/morgenmagazin/videosextern/dominik-kahun-aus-der-nhl-direkt-zur-weltmeisterschaft-100.html',
 306         'only_matching': True,
 307     }, {
 308         'url': 'https://www.daserste.de/information/nachrichten-wetter/tagesthemen/videosextern/tagesthemen-17736.html',
 309         'only_matching': True,
 310     }, {
 311         'url': 'https://www.daserste.de/unterhaltung/serie/in-aller-freundschaft-die-jungen-aerzte/videos/diversity-tag-sanam-afrashteh100.html',
 312         'only_matching': True,
 313     }, {
 314         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 315         'only_matching': True,
 316     }, {
 317         'url': 'https://www.daserste.de/unterhaltung/serie/in-aller-freundschaft-die-jungen-aerzte/Drehpause-100.html',
 318         'only_matching': True,
 319     }, {
 320         'url': 'https://www.daserste.de/unterhaltung/film/filmmittwoch-im-ersten/videos/making-ofwendezeit-video-100.html',
 321         'only_matching': True,
 322     }]
 323
 324     def _real_extract(self, url):
 325         mobj = self._match_valid_url(url)
 326         display_id = mobj.group('id')
 327
 328         player_url = mobj.group('mainurl') + '~playerXml.xml'
 329         doc = self._download_xml(player_url, display_id)
 330         video_node = doc.find('./video')
 331         upload_date = unified_strdate(xpath_text(
 332             video_node, './broadcastDate'))
 333         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 334
 335         formats = []
 336         for a in video_node.findall('.//asset'):
 337             file_name = xpath_text(a, './fileName', default=None)
 338             if not file_name:
 339                 continue
 340             format_type = a.attrib.get('type')
 341             format_url = url_or_none(file_name)
 342             if format_url:
 343                 ext = determine_ext(file_name)
 344                 if ext == 'm3u8':
 345                     formats.extend(self._extract_m3u8_formats(
 346                         format_url, display_id, 'mp4', entry_protocol='m3u8_native',
 347                         m3u8_id=format_type or 'hls', fatal=False))
 348                     continue
 349                 elif ext == 'f4m':
 350                     formats.extend(self._extract_f4m_formats(
 351                         update_url_query(format_url, {'hdcore': '3.7.0'}),
 352                         display_id, f4m_id=format_type or 'hds', fatal=False))
 353                     continue
 354             f = {
 355                 'format_id': format_type,
 356                 'width': int_or_none(xpath_text(a, './frameWidth')),
 357                 'height': int_or_none(xpath_text(a, './frameHeight')),
 358                 'vbr': int_or_none(xpath_text(a, './bitrateVideo')),
 359                 'abr': int_or_none(xpath_text(a, './bitrateAudio')),
 360                 'vcodec': xpath_text(a, './codecVideo'),
 361                 'tbr': int_or_none(xpath_text(a, './totalBitrate')),
 362             }
 363             server_prefix = xpath_text(a, './serverPrefix', default=None)
 364             if server_prefix:
 365                 f.update({
 366                     'url': server_prefix,
 367                     'playpath': file_name,
 368                 })
 369             else:
 370                 if not format_url:
 371                     continue
 372                 f['url'] = format_url
 373             formats.append(f)
 374         self._sort_formats(formats)
 375
 376         _SUB_FORMATS = (
 377             ('./dataTimedText', 'ttml'),
 378             ('./dataTimedTextNoOffset', 'ttml'),
 379             ('./dataTimedTextVtt', 'vtt'),
 380         )
 381
 382         subtitles = {}
 383         for subsel, subext in _SUB_FORMATS:
 384             for node in video_node.findall(subsel):
 385                 subtitles.setdefault('de', []).append({
 386                     'url': node.attrib['url'],
 387                     'ext': subext,
 388                 })
 389
 390         return {
 391             'id': xpath_text(video_node, './videoId', default=display_id),
 392             'formats': formats,
 393             'subtitles': subtitles,
 394             'display_id': display_id,
 395             'title': video_node.find('./title').text,
 396             'duration': parse_duration(video_node.find('./duration').text),
 397             'upload_date': upload_date,
 398             'thumbnail': thumbnail,
 399         }
 400
 401
 402 class ARDBetaMediathekIE(ARDMediathekBaseIE):
 403     _VALID_URL = r'''(?x)https://
 404         (?:(?:beta|www)\.)?ardmediathek\.de/
 405         (?:(?P<client>[^/]+)/)?
 406         (?:player|live|video|(?P<playlist>sendung|sammlung))/
 407         (?:(?P<display_id>(?(playlist)[^?#]+?|[^?#]+))/)?
 408         (?P<id>(?(playlist)|Y3JpZDovL)[a-zA-Z0-9]+)
 409         (?(playlist)/(?P<season>\d+)?/?(?:[?#]|$))'''
 410
 411     _TESTS = [{
 412         'url': 'https://www.ardmediathek.de/mdr/video/die-robuste-roswita/Y3JpZDovL21kci5kZS9iZWl0cmFnL2Ntcy84MWMxN2MzZC0wMjkxLTRmMzUtODk4ZS0wYzhlOWQxODE2NGI/',
 413         'md5': 'a1dc75a39c61601b980648f7c9f9f71d',
 414         'info_dict': {
 415             'display_id': 'die-robuste-roswita',
 416             'id': '78566716',
 417             'title': 'Die robuste Roswita',
 418             'description': r're:^Der Mord.*totgeglaubte Ehefrau Roswita',
 419             'duration': 5316,
 420             'thumbnail': 'https://img.ardmediathek.de/standard/00/78/56/67/84/575672121/16x9/960?mandant=ard',
 421             'timestamp': 1596658200,
 422             'upload_date': '20200805',
 423             'ext': 'mp4',
 424         },
 425         'skip': 'Error',
 426     }, {
 427         'url': 'https://www.ardmediathek.de/video/tagesschau-oder-tagesschau-20-00-uhr/das-erste/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhZ2Vzc2NoYXUvZmM4ZDUxMjgtOTE0ZC00Y2MzLTgzNzAtNDZkNGNiZWJkOTll',
 428         'md5': 'f1837e563323b8a642a8ddeff0131f51',
 429         'info_dict': {
 430             'id': '10049223',
 431             'ext': 'mp4',
 432             'title': 'tagesschau, 20:00 Uhr',
 433             'timestamp': 1636398000,
 434             'description': 'md5:39578c7b96c9fe50afdf5674ad985e6b',
 435             'upload_date': '20211108',
 436         },
 437     }, {
 438         'url': 'https://www.ardmediathek.de/sendung/beforeigners/beforeigners/staffel-1/Y3JpZDovL2Rhc2Vyc3RlLmRlL2JlZm9yZWlnbmVycw/1',
 439         'playlist_count': 6,
 440         'info_dict': {
 441             'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL2JlZm9yZWlnbmVycw',
 442             'title': 'beforeigners/beforeigners/staffel-1',
 443         },
 444     }, {
 445         'url': 'https://beta.ardmediathek.de/ard/video/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
 446         'only_matching': True,
 447     }, {
 448         'url': 'https://ardmediathek.de/ard/video/saartalk/saartalk-gesellschaftsgift-haltung-gegen-hass/sr-fernsehen/Y3JpZDovL3NyLW9ubGluZS5kZS9TVF84MTY4MA/',
 449         'only_matching': True,
 450     }, {
 451         'url': 'https://www.ardmediathek.de/ard/video/trailer/private-eyes-s01-e01/one/Y3JpZDovL3dkci5kZS9CZWl0cmFnLTE1MTgwYzczLWNiMTEtNGNkMS1iMjUyLTg5MGYzOWQxZmQ1YQ/',
 452         'only_matching': True,
 453     }, {
 454         'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3N3ci5kZS9hZXgvbzEwNzE5MTU/',
 455         'only_matching': True,
 456     }, {
 457         'url': 'https://www.ardmediathek.de/swr/live/Y3JpZDovL3N3ci5kZS8xMzQ4MTA0Mg',
 458         'only_matching': True,
 459     }, {
 460         # playlist of type 'sendung'
 461         'url': 'https://www.ardmediathek.de/ard/sendung/doctor-who/Y3JpZDovL3dkci5kZS9vbmUvZG9jdG9yIHdobw/',
 462         'only_matching': True,
 463     }, {
 464         # playlist of type 'sammlung'
 465         'url': 'https://www.ardmediathek.de/ard/sammlung/team-muenster/5JpTzLSbWUAK8184IOvEir/',
 466         'only_matching': True,
 467     }, {
 468         'url': 'https://www.ardmediathek.de/video/coronavirus-update-ndr-info/astrazeneca-kurz-lockdown-und-pims-syndrom-81/ndr/Y3JpZDovL25kci5kZS84NzE0M2FjNi0wMWEwLTQ5ODEtOTE5NS1mOGZhNzdhOTFmOTI/',
 469         'only_matching': True,
 470     }, {
 471         'url': 'https://www.ardmediathek.de/ard/player/Y3JpZDovL3dkci5kZS9CZWl0cmFnLWQ2NDJjYWEzLTMwZWYtNGI4NS1iMTI2LTU1N2UxYTcxOGIzOQ/tatort-duo-koeln-leipzig-ihr-kinderlein-kommet',
 472         'only_matching': True,
 473     }]
 474
 475     def _ARD_load_playlist_snipped(self, playlist_id, display_id, client, mode, pageNumber):
 476         """ Query the ARD server for playlist information
 477         and returns the data in "raw" format """
 478         if mode == 'sendung':
 479             graphQL = json.dumps({
 480                 'query': '''{
 481                     showPage(
 482                         client: "%s"
 483                         showId: "%s"
 484                         pageNumber: %d
 485                     ) {
 486                         pagination {
 487                             pageSize
 488                             totalElements
 489                         }
 490                         teasers {        # Array
 491                             mediumTitle
 492                             links { target { id href title } }
 493                             type
 494                         }
 495                     }}''' % (client, playlist_id, pageNumber),
 496             }).encode()
 497         else:  # mode == 'sammlung'
 498             graphQL = json.dumps({
 499                 'query': '''{
 500                     morePage(
 501                         client: "%s"
 502                         compilationId: "%s"
 503                         pageNumber: %d
 504                     ) {
 505                         widget {
 506                             pagination {
 507                                 pageSize
 508                                 totalElements
 509                             }
 510                             teasers {        # Array
 511                                 mediumTitle
 512                                 links { target { id href title } }
 513                                 type
 514                             }
 515                         }
 516                     }}''' % (client, playlist_id, pageNumber),
 517             }).encode()
 518         # Ressources for ARD graphQL debugging:
 519         # https://api-test.ardmediathek.de/public-gateway
 520         show_page = self._download_json(
 521             'https://api.ardmediathek.de/public-gateway',
 522             '[Playlist] %s' % display_id,
 523             data=graphQL,
 524             headers={'Content-Type': 'application/json'})['data']
 525         # align the structure of the returned data:
 526         if mode == 'sendung':
 527             show_page = show_page['showPage']
 528         else:  # mode == 'sammlung'
 529             show_page = show_page['morePage']['widget']
 530         return show_page
 531
 532     def _ARD_extract_playlist(self, url, playlist_id, display_id, client, mode):
 533         """ Collects all playlist entries and returns them as info dict.
 534         Supports playlists of mode 'sendung' and 'sammlung', and also nested
 535         playlists. """
 536         entries = []
 537         pageNumber = 0
 538         while True:  # iterate by pageNumber
 539             show_page = self._ARD_load_playlist_snipped(
 540                 playlist_id, display_id, client, mode, pageNumber)
 541             for teaser in show_page['teasers']:  # process playlist items
 542                 if '/compilation/' in teaser['links']['target']['href']:
 543                     # alternativ cond.: teaser['type'] == "compilation"
 544                     # => This is an nested compilation, e.g. like:
 545                     # https://www.ardmediathek.de/ard/sammlung/die-kirche-bleibt-im-dorf/5eOHzt8XB2sqeFXbIoJlg2/
 546                     link_mode = 'sammlung'
 547                 else:
 548                     link_mode = 'video'
 549
 550                 item_url = 'https://www.ardmediathek.de/%s/%s/%s/%s/%s' % (
 551                     client, link_mode, display_id,
 552                     # perform HTLM quoting of episode title similar to ARD:
 553                     re.sub('^-|-$', '',  # remove '-' from begin/end
 554                            re.sub('[^a-zA-Z0-9]+', '-',  # replace special chars by -
 555                                   teaser['links']['target']['title'].lower()
 556                                   .replace('ä', 'ae').replace('ö', 'oe')
 557                                   .replace('ü', 'ue').replace('ß', 'ss'))),
 558                     teaser['links']['target']['id'])
 559                 entries.append(self.url_result(
 560                     item_url,
 561                     ie=ARDBetaMediathekIE.ie_key()))
 562
 563             if (show_page['pagination']['pageSize'] * (pageNumber + 1)
 564                >= show_page['pagination']['totalElements']):
 565                 # we've processed enough pages to get all playlist entries
 566                 break
 567             pageNumber = pageNumber + 1
 568
 569         return self.playlist_result(entries, playlist_id, playlist_title=display_id)
 570
 571     def _real_extract(self, url):
 572         video_id, display_id, playlist_type, client, season_number = self._match_valid_url(url).group(
 573             'id', 'display_id', 'playlist', 'client', 'season')
 574         display_id, client = display_id or video_id, client or 'ard'
 575
 576         if playlist_type:
 577             # TODO: Extract only specified season
 578             return self._ARD_extract_playlist(url, video_id, display_id, client, playlist_type)
 579
 580         player_page = self._download_json(
 581             'https://api.ardmediathek.de/public-gateway',
 582             display_id, data=json.dumps({
 583                 'query': '''{
 584   playerPage(client:"%s", clipId: "%s") {
 585     blockedByFsk
 586     broadcastedOn
 587     maturityContentRating
 588     mediaCollection {
 589       _duration
 590       _geoblocked
 591       _isLive
 592       _mediaArray {
 593         _mediaStreamArray {
 594           _quality
 595           _server
 596           _stream
 597         }
 598       }
 599       _previewImage
 600       _subtitleUrl
 601       _type
 602     }
 603     show {
 604       title
 605     }
 606     synopsis
 607     title
 608     tracking {
 609       atiCustomVars {
 610         contentId
 611       }
 612     }
 613   }
 614 }''' % (client, video_id),
 615             }).encode(), headers={
 616                 'Content-Type': 'application/json'
 617             })['data']['playerPage']
 618         title = player_page['title']
 619         content_id = str_or_none(try_get(
 620             player_page, lambda x: x['tracking']['atiCustomVars']['contentId']))
 621         media_collection = player_page.get('mediaCollection') or {}
 622         if not media_collection and content_id:
 623             media_collection = self._download_json(
 624                 'https://www.ardmediathek.de/play/media/' + content_id,
 625                 content_id, fatal=False) or {}
 626         info = self._parse_media_info(
 627             media_collection, content_id or video_id,
 628             player_page.get('blockedByFsk'))
 629         age_limit = None
 630         description = player_page.get('synopsis')
 631         maturity_content_rating = player_page.get('maturityContentRating')
 632         if maturity_content_rating:
 633             age_limit = int_or_none(maturity_content_rating.lstrip('FSK'))
 634         if not age_limit and description:
 635             age_limit = int_or_none(self._search_regex(
 636                 r'\(FSK\s*(\d+)\)\s*$', description, 'age limit', default=None))
 637         info.update({
 638             'age_limit': age_limit,
 639             'display_id': display_id,
 640             'title': title,
 641             'description': description,
 642             'timestamp': unified_timestamp(player_page.get('broadcastedOn')),
 643             'series': try_get(player_page, lambda x: x['show']['title']),
 644         })
 645         info.update(self._ARD_extract_episode_info(info['title']))
 646         return info