yt_dlp/extractor/zdf.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..compat import compat_str
   8 from ..utils import (
   9     determine_ext,
  10     float_or_none,
  11     int_or_none,
  12     join_nonempty,
  13     merge_dicts,
  14     NO_DEFAULT,
  15     orderedSet,
  16     parse_codecs,
  17     qualities,
  18     traverse_obj,
  19     try_get,
  20     unified_timestamp,
  21     update_url_query,
  22     url_or_none,
  23     urljoin,
  24 )
  25
  26
  27 class ZDFBaseIE(InfoExtractor):
  28     _GEO_COUNTRIES = ['DE']
  29     _QUALITIES = ('auto', 'low', 'med', 'high', 'veryhigh', 'hd')
  30
  31     def _call_api(self, url, video_id, item, api_token=None, referrer=None):
  32         headers = {}
  33         if api_token:
  34             headers['Api-Auth'] = 'Bearer %s' % api_token
  35         if referrer:
  36             headers['Referer'] = referrer
  37         return self._download_json(
  38             url, video_id, 'Downloading JSON %s' % item, headers=headers)
  39
  40     @staticmethod
  41     def _extract_subtitles(src):
  42         subtitles = {}
  43         for caption in try_get(src, lambda x: x['captions'], list) or []:
  44             subtitle_url = url_or_none(caption.get('uri'))
  45             if subtitle_url:
  46                 lang = caption.get('language', 'deu')
  47                 subtitles.setdefault(lang, []).append({
  48                     'url': subtitle_url,
  49                 })
  50         return subtitles
  51
  52     def _extract_format(self, video_id, formats, format_urls, meta):
  53         format_url = url_or_none(meta.get('url'))
  54         if not format_url or format_url in format_urls:
  55             return
  56         format_urls.add(format_url)
  57
  58         mime_type, ext = meta.get('mimeType'), determine_ext(format_url)
  59         if mime_type == 'application/x-mpegURL' or ext == 'm3u8':
  60             new_formats = self._extract_m3u8_formats(
  61                 format_url, video_id, 'mp4', m3u8_id='hls',
  62                 entry_protocol='m3u8_native', fatal=False)
  63         elif mime_type == 'application/f4m+xml' or ext == 'f4m':
  64             new_formats = self._extract_f4m_formats(
  65                 update_url_query(format_url, {'hdcore': '3.7.0'}), video_id, f4m_id='hds', fatal=False)
  66         else:
  67             f = parse_codecs(meta.get('mimeCodec'))
  68             if not f and meta.get('type'):
  69                 data = meta['type'].split('_')
  70                 if try_get(data, lambda x: x[2]) == ext:
  71                     f = {'vcodec': data[0], 'acodec': data[1]}
  72             f.update({
  73                 'url': format_url,
  74                 'format_id': join_nonempty('http', meta.get('type'), meta.get('quality')),
  75             })
  76             new_formats = [f]
  77         formats.extend(merge_dicts(f, {
  78             'format_note': join_nonempty('quality', 'class', from_dict=meta, delim=', '),
  79             'language': meta.get('language'),
  80             'language_preference': 10 if meta.get('class') == 'main' else -10 if meta.get('class') == 'ad' else -1,
  81             'quality': qualities(self._QUALITIES)(meta.get('quality')),
  82         }) for f in new_formats)
  83
  84     def _extract_ptmd(self, ptmd_url, video_id, api_token, referrer):
  85         ptmd = self._call_api(
  86             ptmd_url, video_id, 'metadata', api_token, referrer)
  87
  88         content_id = ptmd.get('basename') or ptmd_url.split('/')[-1]
  89
  90         formats = []
  91         track_uris = set()
  92         for p in ptmd['priorityList']:
  93             formitaeten = p.get('formitaeten')
  94             if not isinstance(formitaeten, list):
  95                 continue
  96             for f in formitaeten:
  97                 f_qualities = f.get('qualities')
  98                 if not isinstance(f_qualities, list):
  99                     continue
 100                 for quality in f_qualities:
 101                     tracks = try_get(quality, lambda x: x['audio']['tracks'], list)
 102                     if not tracks:
 103                         continue
 104                     for track in tracks:
 105                         self._extract_format(
 106                             content_id, formats, track_uris, {
 107                                 'url': track.get('uri'),
 108                                 'type': f.get('type'),
 109                                 'mimeType': f.get('mimeType'),
 110                                 'quality': quality.get('quality'),
 111                                 'class': track.get('class'),
 112                                 'language': track.get('language'),
 113                             })
 114         self._sort_formats(formats, ('hasaud', 'res', 'quality', 'language_preference'))
 115
 116         duration = float_or_none(try_get(
 117             ptmd, lambda x: x['attributes']['duration']['value']), scale=1000)
 118
 119         return {
 120             'extractor_key': ZDFIE.ie_key(),
 121             'id': content_id,
 122             'duration': duration,
 123             'formats': formats,
 124             'subtitles': self._extract_subtitles(ptmd),
 125         }
 126
 127     def _extract_player(self, webpage, video_id, fatal=True):
 128         return self._parse_json(
 129             self._search_regex(
 130                 r'(?s)data-zdfplayer-jsb=(["\'])(?P<json>{.+?})\1', webpage,
 131                 'player JSON', default='{}' if not fatal else NO_DEFAULT,
 132                 group='json'),
 133             video_id)
 134
 135
 136 class ZDFIE(ZDFBaseIE):
 137     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)\.html'
 138     _TESTS = [{
 139         'url': 'https://www.zdf.de/nachrichten/heute-journal/heute-journal-vom-30-12-2021-100.html',
 140         'info_dict': {
 141             'id': '211230_sendung_hjo',
 142             'ext': 'mp4',
 143             'description': 'md5:47dff85977bde9fb8cba9e9c9b929839',
 144             'duration': 1890.0,
 145             'upload_date': '20211230',
 146             'chapters': list,
 147             'thumbnail': 'md5:e65f459f741be5455c952cd820eb188e',
 148             'title': 'heute journal vom 30.12.2021',
 149             'timestamp': 1640897100,
 150         }
 151     }, {
 152         'url': 'https://www.zdf.de/dokumentation/terra-x/die-magie-der-farben-von-koenigspurpur-und-jeansblau-100.html',
 153         'info_dict': {
 154             'id': '151025_magie_farben2_tex',
 155             'ext': 'mp4',
 156             'title': 'Die Magie der Farben (2/2)',
 157             'description': 'md5:a89da10c928c6235401066b60a6d5c1a',
 158             'duration': 2615,
 159             'timestamp': 1465021200,
 160             'upload_date': '20160604',
 161             'thumbnail': 'https://www.zdf.de/assets/mauve-im-labor-100~768x432?cb=1464909117806',
 162         },
 163     }, {
 164         'url': 'https://www.zdf.de/funk/druck-11790/funk-alles-ist-verzaubert-102.html',
 165         'md5': '3d6f1049e9682178a11c54b91f3dd065',
 166         'info_dict': {
 167             'ext': 'mp4',
 168             'id': 'video_funk_1770473',
 169             'duration': 1278,
 170             'description': 'Die Neue an der Schule verdreht Ismail den Kopf.',
 171             'title': 'Alles ist verzaubert',
 172             'timestamp': 1635520560,
 173             'upload_date': '20211029',
 174             'thumbnail': 'https://www.zdf.de/assets/teaser-funk-alles-ist-verzaubert-100~1920x1080?cb=1636466431799',
 175         },
 176     }, {
 177         # Same as https://www.phoenix.de/sendungen/dokumentationen/gesten-der-maechtigen-i-a-89468.html?ref=suche
 178         'url': 'https://www.zdf.de/politik/phoenix-sendungen/die-gesten-der-maechtigen-100.html',
 179         'only_matching': True,
 180     }, {
 181         # Same as https://www.3sat.de/film/spielfilm/der-hauptmann-100.html
 182         'url': 'https://www.zdf.de/filme/filme-sonstige/der-hauptmann-112.html',
 183         'only_matching': True,
 184     }, {
 185         # Same as https://www.3sat.de/wissen/nano/nano-21-mai-2019-102.html, equal media ids
 186         'url': 'https://www.zdf.de/wissen/nano/nano-21-mai-2019-102.html',
 187         'only_matching': True,
 188     }, {
 189         'url': 'https://www.zdf.de/service-und-hilfe/die-neue-zdf-mediathek/zdfmediathek-trailer-100.html',
 190         'only_matching': True,
 191     }, {
 192         'url': 'https://www.zdf.de/filme/taunuskrimi/die-lebenden-und-die-toten-1---ein-taunuskrimi-100.html',
 193         'only_matching': True,
 194     }, {
 195         'url': 'https://www.zdf.de/dokumentation/planet-e/planet-e-uebersichtsseite-weitere-dokumentationen-von-planet-e-100.html',
 196         'only_matching': True,
 197     }, {
 198         # Same as https://www.phoenix.de/sendungen/ereignisse/corona-nachgehakt/wohin-fuehrt-der-protest-in-der-pandemie-a-2050630.html
 199         'url': 'https://www.zdf.de/politik/phoenix-sendungen/wohin-fuehrt-der-protest-in-der-pandemie-100.html',
 200         'only_matching': True
 201     }, {
 202         # Same as https://www.3sat.de/film/ab-18/10-wochen-sommer-108.html
 203         'url': 'https://www.zdf.de/dokumentation/ab-18/10-wochen-sommer-102.html',
 204         'only_matching': True
 205     }]
 206
 207     def _extract_entry(self, url, player, content, video_id):
 208         title = content.get('title') or content['teaserHeadline']
 209
 210         t = content['mainVideoContent']['http://zdf.de/rels/target']
 211
 212         ptmd_path = t.get('http://zdf.de/rels/streams/ptmd')
 213
 214         if not ptmd_path:
 215             ptmd_path = traverse_obj(
 216                 t, ('streams', 'default', 'http://zdf.de/rels/streams/ptmd-template'),
 217                 'http://zdf.de/rels/streams/ptmd-template').replace(
 218                 '{playerId}', 'ngplayer_2_4')
 219
 220         info = self._extract_ptmd(
 221             urljoin(url, ptmd_path), video_id, player['apiToken'], url)
 222
 223         thumbnails = []
 224         layouts = try_get(
 225             content, lambda x: x['teaserImageRef']['layouts'], dict)
 226         if layouts:
 227             for layout_key, layout_url in layouts.items():
 228                 layout_url = url_or_none(layout_url)
 229                 if not layout_url:
 230                     continue
 231                 thumbnail = {
 232                     'url': layout_url,
 233                     'format_id': layout_key,
 234                 }
 235                 mobj = re.search(r'(?P<width>\d+)x(?P<height>\d+)', layout_key)
 236                 if mobj:
 237                     thumbnail.update({
 238                         'width': int(mobj.group('width')),
 239                         'height': int(mobj.group('height')),
 240                     })
 241                 thumbnails.append(thumbnail)
 242
 243         chapter_marks = t.get('streamAnchorTag') or []
 244         chapter_marks.append({'anchorOffset': int_or_none(t.get('duration'))})
 245         chapters = [{
 246             'start_time': chap.get('anchorOffset'),
 247             'end_time': next_chap.get('anchorOffset'),
 248             'title': chap.get('anchorLabel')
 249         } for chap, next_chap in zip(chapter_marks, chapter_marks[1:])]
 250
 251         return merge_dicts(info, {
 252             'title': title,
 253             'description': content.get('leadParagraph') or content.get('teasertext'),
 254             'duration': int_or_none(t.get('duration')),
 255             'timestamp': unified_timestamp(content.get('editorialDate')),
 256             'thumbnails': thumbnails,
 257             'chapters': chapters or None
 258         })
 259
 260     def _extract_regular(self, url, player, video_id):
 261         content = self._call_api(
 262             player['content'], video_id, 'content', player['apiToken'], url)
 263         return self._extract_entry(player['content'], player, content, video_id)
 264
 265     def _extract_mobile(self, video_id):
 266         video = self._download_json(
 267             'https://zdf-cdn.live.cellular.de/mediathekV2/document/%s' % video_id,
 268             video_id)
 269
 270         document = video['document']
 271
 272         title = document['titel']
 273         content_id = document['basename']
 274
 275         formats = []
 276         format_urls = set()
 277         for f in document['formitaeten']:
 278             self._extract_format(content_id, formats, format_urls, f)
 279         self._sort_formats(formats)
 280
 281         thumbnails = []
 282         teaser_bild = document.get('teaserBild')
 283         if isinstance(teaser_bild, dict):
 284             for thumbnail_key, thumbnail in teaser_bild.items():
 285                 thumbnail_url = try_get(
 286                     thumbnail, lambda x: x['url'], compat_str)
 287                 if thumbnail_url:
 288                     thumbnails.append({
 289                         'url': thumbnail_url,
 290                         'id': thumbnail_key,
 291                         'width': int_or_none(thumbnail.get('width')),
 292                         'height': int_or_none(thumbnail.get('height')),
 293                     })
 294
 295         return {
 296             'id': content_id,
 297             'title': title,
 298             'description': document.get('beschreibung'),
 299             'duration': int_or_none(document.get('length')),
 300             'timestamp': unified_timestamp(document.get('date')) or unified_timestamp(
 301                 try_get(video, lambda x: x['meta']['editorialDate'], compat_str)),
 302             'thumbnails': thumbnails,
 303             'subtitles': self._extract_subtitles(document),
 304             'formats': formats,
 305         }
 306
 307     def _real_extract(self, url):
 308         video_id = self._match_id(url)
 309
 310         webpage = self._download_webpage(url, video_id, fatal=False)
 311         if webpage:
 312             player = self._extract_player(webpage, url, fatal=False)
 313             if player:
 314                 return self._extract_regular(url, player, video_id)
 315
 316         return self._extract_mobile(video_id)
 317
 318
 319 class ZDFChannelIE(ZDFBaseIE):
 320     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 321     _TESTS = [{
 322         'url': 'https://www.zdf.de/sport/das-aktuelle-sportstudio',
 323         'info_dict': {
 324             'id': 'das-aktuelle-sportstudio',
 325             'title': 'das aktuelle sportstudio | ZDF',
 326         },
 327         'playlist_mincount': 23,
 328     }, {
 329         'url': 'https://www.zdf.de/dokumentation/planet-e',
 330         'info_dict': {
 331             'id': 'planet-e',
 332             'title': 'planet e.',
 333         },
 334         'playlist_mincount': 50,
 335     }, {
 336         'url': 'https://www.zdf.de/filme/taunuskrimi/',
 337         'only_matching': True,
 338     }]
 339
 340     @classmethod
 341     def suitable(cls, url):
 342         return False if ZDFIE.suitable(url) else super(ZDFChannelIE, cls).suitable(url)
 343
 344     def _real_extract(self, url):
 345         channel_id = self._match_id(url)
 346
 347         webpage = self._download_webpage(url, channel_id)
 348
 349         entries = [
 350             self.url_result(item_url, ie=ZDFIE.ie_key())
 351             for item_url in orderedSet(re.findall(
 352                 r'data-plusbar-url=["\'](http.+?\.html)', webpage))]
 353
 354         return self.playlist_result(
 355             entries, channel_id, self._og_search_title(webpage, fatal=False))
 356
 357         r"""
 358         player = self._extract_player(webpage, channel_id)
 359
 360         channel_id = self._search_regex(
 361             r'docId\s*:\s*(["\'])(?P<id>(?!\1).+?)\1', webpage,
 362             'channel id', group='id')
 363
 364         channel = self._call_api(
 365             'https://api.zdf.de/content/documents/%s.json' % channel_id,
 366             player, url, channel_id)
 367
 368         items = []
 369         for module in channel['module']:
 370             for teaser in try_get(module, lambda x: x['teaser'], list) or []:
 371                 t = try_get(
 372                     teaser, lambda x: x['http://zdf.de/rels/target'], dict)
 373                 if not t:
 374                     continue
 375                 items.extend(try_get(
 376                     t,
 377                     lambda x: x['resultsWithVideo']['http://zdf.de/rels/search/results'],
 378                     list) or [])
 379             items.extend(try_get(
 380                 module,
 381                 lambda x: x['filterRef']['resultsWithVideo']['http://zdf.de/rels/search/results'],
 382                 list) or [])
 383
 384         entries = []
 385         entry_urls = set()
 386         for item in items:
 387             t = try_get(item, lambda x: x['http://zdf.de/rels/target'], dict)
 388             if not t:
 389                 continue
 390             sharing_url = t.get('http://zdf.de/rels/sharing-url')
 391             if not sharing_url or not isinstance(sharing_url, compat_str):
 392                 continue
 393             if sharing_url in entry_urls:
 394                 continue
 395             entry_urls.add(sharing_url)
 396             entries.append(self.url_result(
 397                 sharing_url, ie=ZDFIE.ie_key(), video_id=t.get('id')))
 398
 399         return self.playlist_result(entries, channel_id, channel.get('title'))
 400         """