_ Git - youtube-dl/blob - youtube_dl/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     int_or_none,
  11     NO_DEFAULT,
  12     orderedSet,
  13     parse_codecs,
  14     qualities,
  15     try_get,
  16     unified_timestamp,
  17     update_url_query,
  18     urljoin,
  19 )
  20
  21
  22 class ZDFBaseIE(InfoExtractor):
  23     def _call_api(self, url, player, referrer, video_id):
  24         return self._download_json(
  25             url, video_id, 'Downloading JSON content',
  26             headers={
  27                 'Referer': referrer,
  28                 'Api-Auth': 'Bearer %s' % player['apiToken'],
  29             })
  30
  31     def _extract_player(self, webpage, video_id, fatal=True):
  32         return self._parse_json(
  33             self._search_regex(
  34                 r'(?s)data-zdfplayer-jsb=(["\'])(?P<json>{.+?})\1', webpage,
  35                 'player JSON', default='{}' if not fatal else NO_DEFAULT,
  36                 group='json'),
  37             video_id)
  38
  39
  40 class ZDFIE(ZDFBaseIE):
  41     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?]+)\.html'
  42     _QUALITIES = ('auto', 'low', 'med', 'high', 'veryhigh')
  43
  44     _TESTS = [{
  45         'url': 'https://www.zdf.de/service-und-hilfe/die-neue-zdf-mediathek/zdfmediathek-trailer-100.html',
  46         'info_dict': {
  47             'id': 'zdfmediathek-trailer-100',
  48             'ext': 'mp4',
  49             'title': 'Die neue ZDFmediathek',
  50             'description': 'md5:3003d36487fb9a5ea2d1ff60beb55e8d',
  51             'duration': 30,
  52             'timestamp': 1477627200,
  53             'upload_date': '20161028',
  54         }
  55     }, {
  56         'url': 'https://www.zdf.de/filme/taunuskrimi/die-lebenden-und-die-toten-1---ein-taunuskrimi-100.html',
  57         'only_matching': True,
  58     }, {
  59         'url': 'https://www.zdf.de/dokumentation/planet-e/planet-e-uebersichtsseite-weitere-dokumentationen-von-planet-e-100.html',
  60         'only_matching': True,
  61     }]
  62
  63     @staticmethod
  64     def _extract_subtitles(src):
  65         subtitles = {}
  66         for caption in try_get(src, lambda x: x['captions'], list) or []:
  67             subtitle_url = caption.get('uri')
  68             if subtitle_url and isinstance(subtitle_url, compat_str):
  69                 lang = caption.get('language', 'deu')
  70                 subtitles.setdefault(lang, []).append({
  71                     'url': subtitle_url,
  72                 })
  73         return subtitles
  74
  75     def _extract_format(self, video_id, formats, format_urls, meta):
  76         format_url = meta.get('url')
  77         if not format_url or not isinstance(format_url, compat_str):
  78             return
  79         if format_url in format_urls:
  80             return
  81         format_urls.add(format_url)
  82         mime_type = meta.get('mimeType')
  83         ext = determine_ext(format_url)
  84         if mime_type == 'application/x-mpegURL' or ext == 'm3u8':
  85             formats.extend(self._extract_m3u8_formats(
  86                 format_url, video_id, 'mp4', m3u8_id='hls',
  87                 entry_protocol='m3u8_native', fatal=False))
  88         elif mime_type == 'application/f4m+xml' or ext == 'f4m':
  89             formats.extend(self._extract_f4m_formats(
  90                 update_url_query(format_url, {'hdcore': '3.7.0'}), video_id, f4m_id='hds', fatal=False))
  91         else:
  92             f = parse_codecs(meta.get('mimeCodec'))
  93             format_id = ['http']
  94             for p in (meta.get('type'), meta.get('quality')):
  95                 if p and isinstance(p, compat_str):
  96                     format_id.append(p)
  97             f.update({
  98                 'url': format_url,
  99                 'format_id': '-'.join(format_id),
 100                 'format_note': meta.get('quality'),
 101                 'language': meta.get('language'),
 102                 'quality': qualities(self._QUALITIES)(meta.get('quality')),
 103                 'preference': -10,
 104             })
 105             formats.append(f)
 106
 107     def _extract_entry(self, url, content, video_id):
 108         title = content.get('title') or content['teaserHeadline']
 109
 110         t = content['mainVideoContent']['http://zdf.de/rels/target']
 111
 112         ptmd_path = t.get('http://zdf.de/rels/streams/ptmd')
 113
 114         if not ptmd_path:
 115             ptmd_path = t[
 116                 'http://zdf.de/rels/streams/ptmd-template'].replace(
 117                 '{playerId}', 'portal')
 118
 119         ptmd = self._download_json(urljoin(url, ptmd_path), video_id)
 120
 121         formats = []
 122         track_uris = set()
 123         for p in ptmd['priorityList']:
 124             formitaeten = p.get('formitaeten')
 125             if not isinstance(formitaeten, list):
 126                 continue
 127             for f in formitaeten:
 128                 f_qualities = f.get('qualities')
 129                 if not isinstance(f_qualities, list):
 130                     continue
 131                 for quality in f_qualities:
 132                     tracks = try_get(quality, lambda x: x['audio']['tracks'], list)
 133                     if not tracks:
 134                         continue
 135                     for track in tracks:
 136                         self._extract_format(
 137                             video_id, formats, track_uris, {
 138                                 'url': track.get('uri'),
 139                                 'type': f.get('type'),
 140                                 'mimeType': f.get('mimeType'),
 141                                 'quality': quality.get('quality'),
 142                                 'language': track.get('language'),
 143                             })
 144         self._sort_formats(formats)
 145
 146         thumbnails = []
 147         layouts = try_get(
 148             content, lambda x: x['teaserImageRef']['layouts'], dict)
 149         if layouts:
 150             for layout_key, layout_url in layouts.items():
 151                 if not isinstance(layout_url, compat_str):
 152                     continue
 153                 thumbnail = {
 154                     'url': layout_url,
 155                     'format_id': layout_key,
 156                 }
 157                 mobj = re.search(r'(?P<width>\d+)x(?P<height>\d+)', layout_key)
 158                 if mobj:
 159                     thumbnail.update({
 160                         'width': int(mobj.group('width')),
 161                         'height': int(mobj.group('height')),
 162                     })
 163                 thumbnails.append(thumbnail)
 164
 165         return {
 166             'id': video_id,
 167             'title': title,
 168             'description': content.get('leadParagraph') or content.get('teasertext'),
 169             'duration': int_or_none(t.get('duration')),
 170             'timestamp': unified_timestamp(content.get('editorialDate')),
 171             'thumbnails': thumbnails,
 172             'subtitles': self._extract_subtitles(ptmd),
 173             'formats': formats,
 174         }
 175
 176     def _extract_regular(self, url, player, video_id):
 177         content = self._call_api(player['content'], player, url, video_id)
 178         return self._extract_entry(player['content'], content, video_id)
 179
 180     def _extract_mobile(self, video_id):
 181         document = self._download_json(
 182             'https://zdf-cdn.live.cellular.de/mediathekV2/document/%s' % video_id,
 183             video_id)['document']
 184
 185         title = document['titel']
 186
 187         formats = []
 188         format_urls = set()
 189         for f in document['formitaeten']:
 190             self._extract_format(video_id, formats, format_urls, f)
 191         self._sort_formats(formats)
 192
 193         thumbnails = []
 194         teaser_bild = document.get('teaserBild')
 195         if isinstance(teaser_bild, dict):
 196             for thumbnail_key, thumbnail in teaser_bild.items():
 197                 thumbnail_url = try_get(
 198                     thumbnail, lambda x: x['url'], compat_str)
 199                 if thumbnail_url:
 200                     thumbnails.append({
 201                         'url': thumbnail_url,
 202                         'id': thumbnail_key,
 203                         'width': int_or_none(thumbnail.get('width')),
 204                         'height': int_or_none(thumbnail.get('height')),
 205                     })
 206
 207         return {
 208             'id': video_id,
 209             'title': title,
 210             'description': document.get('beschreibung'),
 211             'duration': int_or_none(document.get('length')),
 212             'timestamp': unified_timestamp(try_get(
 213                 document, lambda x: x['meta']['editorialDate'], compat_str)),
 214             'thumbnails': thumbnails,
 215             'subtitles': self._extract_subtitles(document),
 216             'formats': formats,
 217         }
 218
 219     def _real_extract(self, url):
 220         video_id = self._match_id(url)
 221
 222         webpage = self._download_webpage(url, video_id, fatal=False)
 223         if webpage:
 224             player = self._extract_player(webpage, url, fatal=False)
 225             if player:
 226                 return self._extract_regular(url, player, video_id)
 227
 228         return self._extract_mobile(video_id)
 229
 230
 231 class ZDFChannelIE(ZDFBaseIE):
 232     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 233     _TESTS = [{
 234         'url': 'https://www.zdf.de/sport/das-aktuelle-sportstudio',
 235         'info_dict': {
 236             'id': 'das-aktuelle-sportstudio',
 237             'title': 'das aktuelle sportstudio | ZDF',
 238         },
 239         'playlist_count': 21,
 240     }, {
 241         'url': 'https://www.zdf.de/dokumentation/planet-e',
 242         'info_dict': {
 243             'id': 'planet-e',
 244             'title': 'planet e.',
 245         },
 246         'playlist_count': 4,
 247     }, {
 248         'url': 'https://www.zdf.de/filme/taunuskrimi/',
 249         'only_matching': True,
 250     }]
 251
 252     @classmethod
 253     def suitable(cls, url):
 254         return False if ZDFIE.suitable(url) else super(ZDFChannelIE, cls).suitable(url)
 255
 256     def _real_extract(self, url):
 257         channel_id = self._match_id(url)
 258
 259         webpage = self._download_webpage(url, channel_id)
 260
 261         entries = [
 262             self.url_result(item_url, ie=ZDFIE.ie_key())
 263             for item_url in orderedSet(re.findall(
 264                 r'data-plusbar-url=["\'](http.+?\.html)', webpage))]
 265
 266         return self.playlist_result(
 267             entries, channel_id, self._og_search_title(webpage, fatal=False))
 268
 269         r"""
 270         player = self._extract_player(webpage, channel_id)
 271
 272         channel_id = self._search_regex(
 273             r'docId\s*:\s*(["\'])(?P<id>(?!\1).+?)\1', webpage,
 274             'channel id', group='id')
 275
 276         channel = self._call_api(
 277             'https://api.zdf.de/content/documents/%s.json' % channel_id,
 278             player, url, channel_id)
 279
 280         items = []
 281         for module in channel['module']:
 282             for teaser in try_get(module, lambda x: x['teaser'], list) or []:
 283                 t = try_get(
 284                     teaser, lambda x: x['http://zdf.de/rels/target'], dict)
 285                 if not t:
 286                     continue
 287                 items.extend(try_get(
 288                     t,
 289                     lambda x: x['resultsWithVideo']['http://zdf.de/rels/search/results'],
 290                     list) or [])
 291             items.extend(try_get(
 292                 module,
 293                 lambda x: x['filterRef']['resultsWithVideo']['http://zdf.de/rels/search/results'],
 294                 list) or [])
 295
 296         entries = []
 297         entry_urls = set()
 298         for item in items:
 299             t = try_get(item, lambda x: x['http://zdf.de/rels/target'], dict)
 300             if not t:
 301                 continue
 302             sharing_url = t.get('http://zdf.de/rels/sharing-url')
 303             if not sharing_url or not isinstance(sharing_url, compat_str):
 304                 continue
 305             if sharing_url in entry_urls:
 306                 continue
 307             entry_urls.add(sharing_url)
 308             entries.append(self.url_result(
 309                 sharing_url, ie=ZDFIE.ie_key(), video_id=t.get('id')))
 310
 311         return self.playlist_result(entries, channel_id, channel.get('title'))
 312         """