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, item):
  24         return self._download_json(
  25             url, video_id, 'Downloading JSON %s' % item,
  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/dokumentation/terra-x/die-magie-der-farben-von-koenigspurpur-und-jeansblau-100.html',
  46         'info_dict': {
  47             'id': 'die-magie-der-farben-von-koenigspurpur-und-jeansblau-100',
  48             'ext': 'mp4',
  49             'title': 'Die Magie der Farben (2/2)',
  50             'description': 'md5:a89da10c928c6235401066b60a6d5c1a',
  51             'duration': 2615,
  52             'timestamp': 1465021200,
  53             'upload_date': '20160604',
  54         },
  55     }, {
  56         'url': 'https://www.zdf.de/service-und-hilfe/die-neue-zdf-mediathek/zdfmediathek-trailer-100.html',
  57         'only_matching': True,
  58     }, {
  59         'url': 'https://www.zdf.de/filme/taunuskrimi/die-lebenden-und-die-toten-1---ein-taunuskrimi-100.html',
  60         'only_matching': True,
  61     }, {
  62         'url': 'https://www.zdf.de/dokumentation/planet-e/planet-e-uebersichtsseite-weitere-dokumentationen-von-planet-e-100.html',
  63         'only_matching': True,
  64     }]
  65
  66     @staticmethod
  67     def _extract_subtitles(src):
  68         subtitles = {}
  69         for caption in try_get(src, lambda x: x['captions'], list) or []:
  70             subtitle_url = caption.get('uri')
  71             if subtitle_url and isinstance(subtitle_url, compat_str):
  72                 lang = caption.get('language', 'deu')
  73                 subtitles.setdefault(lang, []).append({
  74                     'url': subtitle_url,
  75                 })
  76         return subtitles
  77
  78     def _extract_format(self, video_id, formats, format_urls, meta):
  79         format_url = meta.get('url')
  80         if not format_url or not isinstance(format_url, compat_str):
  81             return
  82         if format_url in format_urls:
  83             return
  84         format_urls.add(format_url)
  85         mime_type = meta.get('mimeType')
  86         ext = determine_ext(format_url)
  87         if mime_type == 'application/x-mpegURL' or ext == 'm3u8':
  88             formats.extend(self._extract_m3u8_formats(
  89                 format_url, video_id, 'mp4', m3u8_id='hls',
  90                 entry_protocol='m3u8_native', fatal=False))
  91         elif mime_type == 'application/f4m+xml' or ext == 'f4m':
  92             formats.extend(self._extract_f4m_formats(
  93                 update_url_query(format_url, {'hdcore': '3.7.0'}), video_id, f4m_id='hds', fatal=False))
  94         else:
  95             f = parse_codecs(meta.get('mimeCodec'))
  96             format_id = ['http']
  97             for p in (meta.get('type'), meta.get('quality')):
  98                 if p and isinstance(p, compat_str):
  99                     format_id.append(p)
 100             f.update({
 101                 'url': format_url,
 102                 'format_id': '-'.join(format_id),
 103                 'format_note': meta.get('quality'),
 104                 'language': meta.get('language'),
 105                 'quality': qualities(self._QUALITIES)(meta.get('quality')),
 106                 'preference': -10,
 107             })
 108             formats.append(f)
 109
 110     def _extract_entry(self, url, player, content, video_id):
 111         title = content.get('title') or content['teaserHeadline']
 112
 113         t = content['mainVideoContent']['http://zdf.de/rels/target']
 114
 115         ptmd_path = t.get('http://zdf.de/rels/streams/ptmd')
 116
 117         if not ptmd_path:
 118             ptmd_path = t[
 119                 'http://zdf.de/rels/streams/ptmd-template'].replace(
 120                 '{playerId}', 'portal')
 121
 122         ptmd = self._call_api(
 123             urljoin(url, ptmd_path), player, url, video_id, 'metadata')
 124
 125         formats = []
 126         track_uris = set()
 127         for p in ptmd['priorityList']:
 128             formitaeten = p.get('formitaeten')
 129             if not isinstance(formitaeten, list):
 130                 continue
 131             for f in formitaeten:
 132                 f_qualities = f.get('qualities')
 133                 if not isinstance(f_qualities, list):
 134                     continue
 135                 for quality in f_qualities:
 136                     tracks = try_get(quality, lambda x: x['audio']['tracks'], list)
 137                     if not tracks:
 138                         continue
 139                     for track in tracks:
 140                         self._extract_format(
 141                             video_id, formats, track_uris, {
 142                                 'url': track.get('uri'),
 143                                 'type': f.get('type'),
 144                                 'mimeType': f.get('mimeType'),
 145                                 'quality': quality.get('quality'),
 146                                 'language': track.get('language'),
 147                             })
 148         self._sort_formats(formats)
 149
 150         thumbnails = []
 151         layouts = try_get(
 152             content, lambda x: x['teaserImageRef']['layouts'], dict)
 153         if layouts:
 154             for layout_key, layout_url in layouts.items():
 155                 if not isinstance(layout_url, compat_str):
 156                     continue
 157                 thumbnail = {
 158                     'url': layout_url,
 159                     'format_id': layout_key,
 160                 }
 161                 mobj = re.search(r'(?P<width>\d+)x(?P<height>\d+)', layout_key)
 162                 if mobj:
 163                     thumbnail.update({
 164                         'width': int(mobj.group('width')),
 165                         'height': int(mobj.group('height')),
 166                     })
 167                 thumbnails.append(thumbnail)
 168
 169         return {
 170             'id': video_id,
 171             'title': title,
 172             'description': content.get('leadParagraph') or content.get('teasertext'),
 173             'duration': int_or_none(t.get('duration')),
 174             'timestamp': unified_timestamp(content.get('editorialDate')),
 175             'thumbnails': thumbnails,
 176             'subtitles': self._extract_subtitles(ptmd),
 177             'formats': formats,
 178         }
 179
 180     def _extract_regular(self, url, player, video_id):
 181         content = self._call_api(
 182             player['content'], player, url, video_id, 'content')
 183         return self._extract_entry(player['content'], player, content, video_id)
 184
 185     def _extract_mobile(self, video_id):
 186         document = self._download_json(
 187             'https://zdf-cdn.live.cellular.de/mediathekV2/document/%s' % video_id,
 188             video_id)['document']
 189
 190         title = document['titel']
 191
 192         formats = []
 193         format_urls = set()
 194         for f in document['formitaeten']:
 195             self._extract_format(video_id, formats, format_urls, f)
 196         self._sort_formats(formats)
 197
 198         thumbnails = []
 199         teaser_bild = document.get('teaserBild')
 200         if isinstance(teaser_bild, dict):
 201             for thumbnail_key, thumbnail in teaser_bild.items():
 202                 thumbnail_url = try_get(
 203                     thumbnail, lambda x: x['url'], compat_str)
 204                 if thumbnail_url:
 205                     thumbnails.append({
 206                         'url': thumbnail_url,
 207                         'id': thumbnail_key,
 208                         'width': int_or_none(thumbnail.get('width')),
 209                         'height': int_or_none(thumbnail.get('height')),
 210                     })
 211
 212         return {
 213             'id': video_id,
 214             'title': title,
 215             'description': document.get('beschreibung'),
 216             'duration': int_or_none(document.get('length')),
 217             'timestamp': unified_timestamp(try_get(
 218                 document, lambda x: x['meta']['editorialDate'], compat_str)),
 219             'thumbnails': thumbnails,
 220             'subtitles': self._extract_subtitles(document),
 221             'formats': formats,
 222         }
 223
 224     def _real_extract(self, url):
 225         video_id = self._match_id(url)
 226
 227         webpage = self._download_webpage(url, video_id, fatal=False)
 228         if webpage:
 229             player = self._extract_player(webpage, url, fatal=False)
 230             if player:
 231                 return self._extract_regular(url, player, video_id)
 232
 233         return self._extract_mobile(video_id)
 234
 235
 236 class ZDFChannelIE(ZDFBaseIE):
 237     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 238     _TESTS = [{
 239         'url': 'https://www.zdf.de/sport/das-aktuelle-sportstudio',
 240         'info_dict': {
 241             'id': 'das-aktuelle-sportstudio',
 242             'title': 'das aktuelle sportstudio | ZDF',
 243         },
 244         'playlist_count': 21,
 245     }, {
 246         'url': 'https://www.zdf.de/dokumentation/planet-e',
 247         'info_dict': {
 248             'id': 'planet-e',
 249             'title': 'planet e.',
 250         },
 251         'playlist_count': 4,
 252     }, {
 253         'url': 'https://www.zdf.de/filme/taunuskrimi/',
 254         'only_matching': True,
 255     }]
 256
 257     @classmethod
 258     def suitable(cls, url):
 259         return False if ZDFIE.suitable(url) else super(ZDFChannelIE, cls).suitable(url)
 260
 261     def _real_extract(self, url):
 262         channel_id = self._match_id(url)
 263
 264         webpage = self._download_webpage(url, channel_id)
 265
 266         entries = [
 267             self.url_result(item_url, ie=ZDFIE.ie_key())
 268             for item_url in orderedSet(re.findall(
 269                 r'data-plusbar-url=["\'](http.+?\.html)', webpage))]
 270
 271         return self.playlist_result(
 272             entries, channel_id, self._og_search_title(webpage, fatal=False))
 273
 274         r"""
 275         player = self._extract_player(webpage, channel_id)
 276
 277         channel_id = self._search_regex(
 278             r'docId\s*:\s*(["\'])(?P<id>(?!\1).+?)\1', webpage,
 279             'channel id', group='id')
 280
 281         channel = self._call_api(
 282             'https://api.zdf.de/content/documents/%s.json' % channel_id,
 283             player, url, channel_id)
 284
 285         items = []
 286         for module in channel['module']:
 287             for teaser in try_get(module, lambda x: x['teaser'], list) or []:
 288                 t = try_get(
 289                     teaser, lambda x: x['http://zdf.de/rels/target'], dict)
 290                 if not t:
 291                     continue
 292                 items.extend(try_get(
 293                     t,
 294                     lambda x: x['resultsWithVideo']['http://zdf.de/rels/search/results'],
 295                     list) or [])
 296             items.extend(try_get(
 297                 module,
 298                 lambda x: x['filterRef']['resultsWithVideo']['http://zdf.de/rels/search/results'],
 299                 list) or [])
 300
 301         entries = []
 302         entry_urls = set()
 303         for item in items:
 304             t = try_get(item, lambda x: x['http://zdf.de/rels/target'], dict)
 305             if not t:
 306                 continue
 307             sharing_url = t.get('http://zdf.de/rels/sharing-url')
 308             if not sharing_url or not isinstance(sharing_url, compat_str):
 309                 continue
 310             if sharing_url in entry_urls:
 311                 continue
 312             entry_urls.add(sharing_url)
 313             entries.append(self.url_result(
 314                 sharing_url, ie=ZDFIE.ie_key(), video_id=t.get('id')))
 315
 316         return self.playlist_result(entries, channel_id, channel.get('title'))
 317         """