yt_dlp/extractor/zdf.py

   1 import re
   2
   3 from .common import InfoExtractor
   4 from ..compat import compat_str
   5 from ..utils import (
   6     NO_DEFAULT,
   7     ExtractorError,
   8     determine_ext,
   9     extract_attributes,
  10     float_or_none,
  11     int_or_none,
  12     join_nonempty,
  13     merge_dicts,
  14     parse_codecs,
  15     qualities,
  16     traverse_obj,
  17     try_get,
  18     unified_timestamp,
  19     update_url_query,
  20     url_or_none,
  21     urljoin,
  22 )
  23
  24
  25 class ZDFBaseIE(InfoExtractor):
  26     _GEO_COUNTRIES = ['DE']
  27     _QUALITIES = ('auto', 'low', 'med', 'high', 'veryhigh', 'hd', 'fhd', 'uhd')
  28
  29     def _call_api(self, url, video_id, item, api_token=None, referrer=None):
  30         headers = {}
  31         if api_token:
  32             headers['Api-Auth'] = 'Bearer %s' % api_token
  33         if referrer:
  34             headers['Referer'] = referrer
  35         return self._download_json(
  36             url, video_id, 'Downloading JSON %s' % item, headers=headers)
  37
  38     @staticmethod
  39     def _extract_subtitles(src):
  40         subtitles = {}
  41         for caption in try_get(src, lambda x: x['captions'], list) or []:
  42             subtitle_url = url_or_none(caption.get('uri'))
  43             if subtitle_url:
  44                 lang = caption.get('language', 'deu')
  45                 subtitles.setdefault(lang, []).append({
  46                     'url': subtitle_url,
  47                 })
  48         return subtitles
  49
  50     def _extract_format(self, video_id, formats, format_urls, meta):
  51         format_url = url_or_none(meta.get('url'))
  52         if not format_url or format_url in format_urls:
  53             return
  54         format_urls.add(format_url)
  55
  56         mime_type, ext = meta.get('mimeType'), determine_ext(format_url)
  57         if mime_type == 'application/x-mpegURL' or ext == 'm3u8':
  58             new_formats = self._extract_m3u8_formats(
  59                 format_url, video_id, 'mp4', m3u8_id='hls',
  60                 entry_protocol='m3u8_native', fatal=False)
  61         elif mime_type == 'application/f4m+xml' or ext == 'f4m':
  62             new_formats = self._extract_f4m_formats(
  63                 update_url_query(format_url, {'hdcore': '3.7.0'}), video_id, f4m_id='hds', fatal=False)
  64         elif ext == 'mpd':
  65             new_formats = self._extract_mpd_formats(
  66                 format_url, video_id, mpd_id='dash', fatal=False)
  67         else:
  68             f = parse_codecs(meta.get('mimeCodec'))
  69             if not f and meta.get('type'):
  70                 data = meta['type'].split('_')
  71                 if try_get(data, lambda x: x[2]) == ext:
  72                     f = {'vcodec': data[0], 'acodec': data[1]}
  73             f.update({
  74                 'url': format_url,
  75                 'format_id': join_nonempty('http', meta.get('type'), meta.get('quality')),
  76                 'tbr': int_or_none(self._search_regex(r'_(\d+)k_', format_url, 'tbr', default=None))
  77             })
  78             new_formats = [f]
  79         formats.extend(merge_dicts(f, {
  80             'format_note': join_nonempty('quality', 'class', from_dict=meta, delim=', '),
  81             'language': meta.get('language'),
  82             'language_preference': 10 if meta.get('class') == 'main' else -10 if meta.get('class') == 'ad' else -1,
  83             'quality': qualities(self._QUALITIES)(meta.get('quality')),
  84         }) for f in new_formats)
  85
  86     def _extract_ptmd(self, ptmd_url, video_id, api_token, referrer):
  87         ptmd = self._call_api(
  88             ptmd_url, video_id, 'metadata', api_token, referrer)
  89
  90         content_id = ptmd.get('basename') or ptmd_url.split('/')[-1]
  91
  92         formats = []
  93         track_uris = set()
  94         for p in ptmd['priorityList']:
  95             formitaeten = p.get('formitaeten')
  96             if not isinstance(formitaeten, list):
  97                 continue
  98             for f in formitaeten:
  99                 f_qualities = f.get('qualities')
 100                 if not isinstance(f_qualities, list):
 101                     continue
 102                 for quality in f_qualities:
 103                     tracks = try_get(quality, lambda x: x['audio']['tracks'], list)
 104                     if not tracks:
 105                         continue
 106                     for track in tracks:
 107                         self._extract_format(
 108                             content_id, formats, track_uris, {
 109                                 'url': track.get('uri'),
 110                                 'type': f.get('type'),
 111                                 'mimeType': f.get('mimeType'),
 112                                 'quality': quality.get('quality'),
 113                                 'class': track.get('class'),
 114                                 'language': track.get('language'),
 115                             })
 116
 117         duration = float_or_none(try_get(
 118             ptmd, lambda x: x['attributes']['duration']['value']), scale=1000)
 119
 120         return {
 121             'extractor_key': ZDFIE.ie_key(),
 122             'id': content_id,
 123             'duration': duration,
 124             'formats': formats,
 125             'subtitles': self._extract_subtitles(ptmd),
 126             '_format_sort_fields': ('tbr', 'res', 'quality', 'language_preference'),
 127         }
 128
 129     def _extract_player(self, webpage, video_id, fatal=True):
 130         return self._parse_json(
 131             self._search_regex(
 132                 r'(?s)data-zdfplayer-jsb=(["\'])(?P<json>{.+?})\1', webpage,
 133                 'player JSON', default='{}' if not fatal else NO_DEFAULT,
 134                 group='json'),
 135             video_id)
 136
 137
 138 class ZDFIE(ZDFBaseIE):
 139     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)\.html'
 140     _TESTS = [{
 141         # Same as https://www.phoenix.de/sendungen/ereignisse/corona-nachgehakt/wohin-fuehrt-der-protest-in-der-pandemie-a-2050630.html
 142         'url': 'https://www.zdf.de/politik/phoenix-sendungen/wohin-fuehrt-der-protest-in-der-pandemie-100.html',
 143         'md5': '34ec321e7eb34231fd88616c65c92db0',
 144         'info_dict': {
 145             'id': '210222_phx_nachgehakt_corona_protest',
 146             'ext': 'mp4',
 147             'title': 'Wohin führt der Protest in der Pandemie?',
 148             'description': 'md5:7d643fe7f565e53a24aac036b2122fbd',
 149             'duration': 1691,
 150             'timestamp': 1613948400,
 151             'upload_date': '20210221',
 152         },
 153         'skip': 'No longer available: "Diese Seite wurde leider nicht gefunden"',
 154     }, {
 155         # Same as https://www.3sat.de/film/ab-18/10-wochen-sommer-108.html
 156         'url': 'https://www.zdf.de/dokumentation/ab-18/10-wochen-sommer-102.html',
 157         'md5': '0aff3e7bc72c8813f5e0fae333316a1d',
 158         'info_dict': {
 159             'id': '141007_ab18_10wochensommer_film',
 160             'ext': 'mp4',
 161             'title': 'Ab 18! - 10 Wochen Sommer',
 162             'description': 'md5:8253f41dc99ce2c3ff892dac2d65fe26',
 163             'duration': 2660,
 164             'timestamp': 1608604200,
 165             'upload_date': '20201222',
 166         },
 167         'skip': 'No longer available: "Diese Seite wurde leider nicht gefunden"',
 168     }, {
 169         'url': 'https://www.zdf.de/nachrichten/heute-journal/heute-journal-vom-30-12-2021-100.html',
 170         'info_dict': {
 171             'id': '211230_sendung_hjo',
 172             'ext': 'mp4',
 173             'description': 'md5:47dff85977bde9fb8cba9e9c9b929839',
 174             'duration': 1890.0,
 175             'upload_date': '20211230',
 176             'chapters': list,
 177             'thumbnail': 'md5:e65f459f741be5455c952cd820eb188e',
 178             'title': 'heute journal vom 30.12.2021',
 179             'timestamp': 1640897100,
 180         },
 181         'skip': 'No longer available: "Diese Seite wurde leider nicht gefunden"',
 182     }, {
 183         'url': 'https://www.zdf.de/dokumentation/terra-x/die-magie-der-farben-von-koenigspurpur-und-jeansblau-100.html',
 184         'info_dict': {
 185             'id': '151025_magie_farben2_tex',
 186             'ext': 'mp4',
 187             'title': 'Die Magie der Farben (2/2)',
 188             'description': 'md5:a89da10c928c6235401066b60a6d5c1a',
 189             'duration': 2615,
 190             'timestamp': 1465021200,
 191             'upload_date': '20160604',
 192             'thumbnail': 'https://www.zdf.de/assets/mauve-im-labor-100~768x432?cb=1464909117806',
 193         },
 194     }, {
 195         'url': 'https://www.zdf.de/funk/druck-11790/funk-alles-ist-verzaubert-102.html',
 196         'md5': '57af4423db0455a3975d2dc4578536bc',
 197         'info_dict': {
 198             'ext': 'mp4',
 199             'id': 'video_funk_1770473',
 200             'duration': 1278,
 201             'description': 'Die Neue an der Schule verdreht Ismail den Kopf.',
 202             'title': 'Alles ist verzaubert',
 203             'timestamp': 1635520560,
 204             'upload_date': '20211029',
 205             'thumbnail': 'https://www.zdf.de/assets/teaser-funk-alles-ist-verzaubert-102~1920x1080?cb=1663848412907',
 206         },
 207     }, {
 208         # Same as https://www.phoenix.de/sendungen/dokumentationen/gesten-der-maechtigen-i-a-89468.html?ref=suche
 209         'url': 'https://www.zdf.de/politik/phoenix-sendungen/die-gesten-der-maechtigen-100.html',
 210         'only_matching': True,
 211     }, {
 212         # Same as https://www.3sat.de/film/spielfilm/der-hauptmann-100.html
 213         'url': 'https://www.zdf.de/filme/filme-sonstige/der-hauptmann-112.html',
 214         'only_matching': True,
 215     }, {
 216         # Same as https://www.3sat.de/wissen/nano/nano-21-mai-2019-102.html, equal media ids
 217         'url': 'https://www.zdf.de/wissen/nano/nano-21-mai-2019-102.html',
 218         'only_matching': True,
 219     }, {
 220         'url': 'https://www.zdf.de/service-und-hilfe/die-neue-zdf-mediathek/zdfmediathek-trailer-100.html',
 221         'only_matching': True,
 222     }, {
 223         'url': 'https://www.zdf.de/filme/taunuskrimi/die-lebenden-und-die-toten-1---ein-taunuskrimi-100.html',
 224         'only_matching': True,
 225     }, {
 226         'url': 'https://www.zdf.de/dokumentation/planet-e/planet-e-uebersichtsseite-weitere-dokumentationen-von-planet-e-100.html',
 227         'only_matching': True,
 228     }, {
 229         'url': 'https://www.zdf.de/arte/todliche-flucht/page-video-artede-toedliche-flucht-16-100.html',
 230         'info_dict': {
 231             'id': 'video_artede_083871-001-A',
 232             'ext': 'mp4',
 233             'title': 'Tödliche Flucht (1/6)',
 234             'description': 'md5:e34f96a9a5f8abd839ccfcebad3d5315',
 235             'duration': 3193.0,
 236             'timestamp': 1641355200,
 237             'upload_date': '20220105',
 238         },
 239         'skip': 'No longer available "Diese Seite wurde leider nicht gefunden"'
 240     }, {
 241         'url': 'https://www.zdf.de/serien/soko-stuttgart/das-geld-anderer-leute-100.html',
 242         'info_dict': {
 243             'id': '191205_1800_sendung_sok8',
 244             'ext': 'mp4',
 245             'title': 'Das Geld anderer Leute',
 246             'description': 'md5:cb6f660850dc5eb7d1ab776ea094959d',
 247             'duration': 2581.0,
 248             'timestamp': 1675160100,
 249             'upload_date': '20230131',
 250             'thumbnail': 'https://epg-image.zdf.de/fotobase-webdelivery/images/e2d7e55a-09f0-424e-ac73-6cac4dd65f35?layout=2400x1350',
 251         },
 252     }, {
 253         'url': 'https://www.zdf.de/dokumentation/terra-x/unser-gruener-planet-wuesten-doku-100.html',
 254         'info_dict': {
 255             'id': '220605_dk_gruener_planet_wuesten_tex',
 256             'ext': 'mp4',
 257             'title': 'Unser grüner Planet - Wüsten',
 258             'description': 'md5:4fc647b6f9c3796eea66f4a0baea2862',
 259             'duration': 2613.0,
 260             'timestamp': 1654450200,
 261             'upload_date': '20220605',
 262             'format_note': 'uhd, main',
 263             'thumbnail': 'https://www.zdf.de/assets/saguaro-kakteen-102~3840x2160?cb=1655910690796',
 264         },
 265     }]
 266
 267     def _extract_entry(self, url, player, content, video_id):
 268         title = content.get('title') or content['teaserHeadline']
 269
 270         t = content['mainVideoContent']['http://zdf.de/rels/target']
 271         ptmd_path = traverse_obj(t, (
 272             (('streams', 'default'), None),
 273             ('http://zdf.de/rels/streams/ptmd', 'http://zdf.de/rels/streams/ptmd-template')
 274         ), get_all=False)
 275         if not ptmd_path:
 276             raise ExtractorError('Could not extract ptmd_path')
 277
 278         info = self._extract_ptmd(
 279             urljoin(url, ptmd_path.replace('{playerId}', 'android_native_5')), video_id, player['apiToken'], url)
 280
 281         thumbnails = []
 282         layouts = try_get(
 283             content, lambda x: x['teaserImageRef']['layouts'], dict)
 284         if layouts:
 285             for layout_key, layout_url in layouts.items():
 286                 layout_url = url_or_none(layout_url)
 287                 if not layout_url:
 288                     continue
 289                 thumbnail = {
 290                     'url': layout_url,
 291                     'format_id': layout_key,
 292                 }
 293                 mobj = re.search(r'(?P<width>\d+)x(?P<height>\d+)', layout_key)
 294                 if mobj:
 295                     thumbnail.update({
 296                         'width': int(mobj.group('width')),
 297                         'height': int(mobj.group('height')),
 298                     })
 299                 thumbnails.append(thumbnail)
 300
 301         chapter_marks = t.get('streamAnchorTag') or []
 302         chapter_marks.append({'anchorOffset': int_or_none(t.get('duration'))})
 303         chapters = [{
 304             'start_time': chap.get('anchorOffset'),
 305             'end_time': next_chap.get('anchorOffset'),
 306             'title': chap.get('anchorLabel')
 307         } for chap, next_chap in zip(chapter_marks, chapter_marks[1:])]
 308
 309         return merge_dicts(info, {
 310             'title': title,
 311             'description': content.get('leadParagraph') or content.get('teasertext'),
 312             'duration': int_or_none(t.get('duration')),
 313             'timestamp': unified_timestamp(content.get('editorialDate')),
 314             'thumbnails': thumbnails,
 315             'chapters': chapters or None
 316         })
 317
 318     def _extract_regular(self, url, player, video_id):
 319         content = self._call_api(
 320             player['content'], video_id, 'content', player['apiToken'], url)
 321         return self._extract_entry(player['content'], player, content, video_id)
 322
 323     def _extract_mobile(self, video_id):
 324         video = self._download_json(
 325             'https://zdf-cdn.live.cellular.de/mediathekV2/document/%s' % video_id,
 326             video_id)
 327
 328         formats = []
 329         formitaeten = try_get(video, lambda x: x['document']['formitaeten'], list)
 330         document = formitaeten and video['document']
 331         if formitaeten:
 332             title = document['titel']
 333             content_id = document['basename']
 334
 335             format_urls = set()
 336             for f in formitaeten or []:
 337                 self._extract_format(content_id, formats, format_urls, f)
 338
 339         thumbnails = []
 340         teaser_bild = document.get('teaserBild')
 341         if isinstance(teaser_bild, dict):
 342             for thumbnail_key, thumbnail in teaser_bild.items():
 343                 thumbnail_url = try_get(
 344                     thumbnail, lambda x: x['url'], compat_str)
 345                 if thumbnail_url:
 346                     thumbnails.append({
 347                         'url': thumbnail_url,
 348                         'id': thumbnail_key,
 349                         'width': int_or_none(thumbnail.get('width')),
 350                         'height': int_or_none(thumbnail.get('height')),
 351                     })
 352
 353         return {
 354             'id': content_id,
 355             'title': title,
 356             'description': document.get('beschreibung'),
 357             'duration': int_or_none(document.get('length')),
 358             'timestamp': unified_timestamp(document.get('date')) or unified_timestamp(
 359                 try_get(video, lambda x: x['meta']['editorialDate'], compat_str)),
 360             'thumbnails': thumbnails,
 361             'subtitles': self._extract_subtitles(document),
 362             'formats': formats,
 363         }
 364
 365     def _real_extract(self, url):
 366         video_id = self._match_id(url)
 367
 368         webpage = self._download_webpage(url, video_id, fatal=False)
 369         if webpage:
 370             player = self._extract_player(webpage, url, fatal=False)
 371             if player:
 372                 return self._extract_regular(url, player, video_id)
 373
 374         return self._extract_mobile(video_id)
 375
 376
 377 class ZDFChannelIE(ZDFBaseIE):
 378     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 379     _TESTS = [{
 380         'url': 'https://www.zdf.de/sport/das-aktuelle-sportstudio',
 381         'info_dict': {
 382             'id': 'das-aktuelle-sportstudio',
 383             'title': 'das aktuelle sportstudio',
 384         },
 385         'playlist_mincount': 18,
 386     }, {
 387         'url': 'https://www.zdf.de/dokumentation/planet-e',
 388         'info_dict': {
 389             'id': 'planet-e',
 390             'title': 'planet e.',
 391         },
 392         'playlist_mincount': 50,
 393     }, {
 394         'url': 'https://www.zdf.de/gesellschaft/aktenzeichen-xy-ungeloest',
 395         'info_dict': {
 396             'id': 'aktenzeichen-xy-ungeloest',
 397             'title': 'Aktenzeichen XY... ungelöst',
 398             'entries': "lambda x: not any('xy580-fall1-kindermoerder-gesucht-100' in e['url'] for e in x)",
 399         },
 400         'playlist_mincount': 2,
 401     }, {
 402         'url': 'https://www.zdf.de/filme/taunuskrimi/',
 403         'only_matching': True,
 404     }]
 405
 406     @classmethod
 407     def suitable(cls, url):
 408         return False if ZDFIE.suitable(url) else super(ZDFChannelIE, cls).suitable(url)
 409
 410     def _og_search_title(self, webpage, fatal=False):
 411         title = super(ZDFChannelIE, self)._og_search_title(webpage, fatal=fatal)
 412         return re.split(r'\s+[-|]\s+ZDF(?:mediathek)?$', title or '')[0] or None
 413
 414     def _real_extract(self, url):
 415         channel_id = self._match_id(url)
 416
 417         webpage = self._download_webpage(url, channel_id)
 418
 419         matches = re.finditer(
 420             r'''<div\b[^>]*?\sdata-plusbar-id\s*=\s*(["'])(?P<p_id>[\w-]+)\1[^>]*?\sdata-plusbar-url=\1(?P<url>%s)\1''' % ZDFIE._VALID_URL,
 421             webpage)
 422
 423         if self._downloader.params.get('noplaylist', False):
 424             entry = next(
 425                 (self.url_result(m.group('url'), ie=ZDFIE.ie_key()) for m in matches),
 426                 None)
 427             self.to_screen('Downloading just the main video because of --no-playlist')
 428             if entry:
 429                 return entry
 430         else:
 431             self.to_screen('Downloading playlist %s - add --no-playlist to download just the main video' % (channel_id, ))
 432
 433         def check_video(m):
 434             v_ref = self._search_regex(
 435                 r'''(<a\b[^>]*?\shref\s*=[^>]+?\sdata-target-id\s*=\s*(["'])%s\2[^>]*>)''' % (m.group('p_id'), ),
 436                 webpage, 'check id', default='')
 437             v_ref = extract_attributes(v_ref)
 438             return v_ref.get('data-target-video-type') != 'novideo'
 439
 440         return self.playlist_from_matches(
 441             (m.group('url') for m in matches if check_video(m)),
 442             channel_id, self._og_search_title(webpage, fatal=False))