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     url_or_none,
  19     urljoin,
  20 )
  21
  22
  23 class ZDFBaseIE(InfoExtractor):
  24     def _call_api(self, url, player, referrer, video_id, item):
  25         return self._download_json(
  26             url, video_id, 'Downloading JSON %s' % item,
  27             headers={
  28                 'Referer': referrer,
  29                 'Api-Auth': 'Bearer %s' % player['apiToken'],
  30             })
  31
  32     def _extract_player(self, webpage, video_id, fatal=True):
  33         return self._parse_json(
  34             self._search_regex(
  35                 r'(?s)data-zdfplayer-jsb=(["\'])(?P<json>{.+?})\1', webpage,
  36                 'player JSON', default='{}' if not fatal else NO_DEFAULT,
  37                 group='json'),
  38             video_id)
  39
  40
  41 class ZDFIE(ZDFBaseIE):
  42     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?]+)\.html'
  43     _QUALITIES = ('auto', 'low', 'med', 'high', 'veryhigh')
  44
  45     _TESTS = [{
  46         'url': 'https://www.zdf.de/dokumentation/terra-x/die-magie-der-farben-von-koenigspurpur-und-jeansblau-100.html',
  47         'info_dict': {
  48             'id': 'die-magie-der-farben-von-koenigspurpur-und-jeansblau-100',
  49             'ext': 'mp4',
  50             'title': 'Die Magie der Farben (2/2)',
  51             'description': 'md5:a89da10c928c6235401066b60a6d5c1a',
  52             'duration': 2615,
  53             'timestamp': 1465021200,
  54             'upload_date': '20160604',
  55         },
  56     }, {
  57         'url': 'https://www.zdf.de/service-und-hilfe/die-neue-zdf-mediathek/zdfmediathek-trailer-100.html',
  58         'only_matching': True,
  59     }, {
  60         'url': 'https://www.zdf.de/filme/taunuskrimi/die-lebenden-und-die-toten-1---ein-taunuskrimi-100.html',
  61         'only_matching': True,
  62     }, {
  63         'url': 'https://www.zdf.de/dokumentation/planet-e/planet-e-uebersichtsseite-weitere-dokumentationen-von-planet-e-100.html',
  64         'only_matching': True,
  65     }]
  66
  67     @staticmethod
  68     def _extract_subtitles(src):
  69         subtitles = {}
  70         for caption in try_get(src, lambda x: x['captions'], list) or []:
  71             subtitle_url = url_or_none(caption.get('uri'))
  72             if subtitle_url:
  73                 lang = caption.get('language', 'deu')
  74                 subtitles.setdefault(lang, []).append({
  75                     'url': subtitle_url,
  76                 })
  77         return subtitles
  78
  79     def _extract_format(self, video_id, formats, format_urls, meta):
  80         format_url = url_or_none(meta.get('url'))
  81         if not format_url:
  82             return
  83         if format_url in format_urls:
  84             return
  85         format_urls.add(format_url)
  86         mime_type = meta.get('mimeType')
  87         ext = determine_ext(format_url)
  88         if mime_type == 'application/x-mpegURL' or ext == 'm3u8':
  89             formats.extend(self._extract_m3u8_formats(
  90                 format_url, video_id, 'mp4', m3u8_id='hls',
  91                 entry_protocol='m3u8_native', fatal=False))
  92         elif mime_type == 'application/f4m+xml' or ext == 'f4m':
  93             formats.extend(self._extract_f4m_formats(
  94                 update_url_query(format_url, {'hdcore': '3.7.0'}), video_id, f4m_id='hds', fatal=False))
  95         else:
  96             f = parse_codecs(meta.get('mimeCodec'))
  97             format_id = ['http']
  98             for p in (meta.get('type'), meta.get('quality')):
  99                 if p and isinstance(p, compat_str):
 100                     format_id.append(p)
 101             f.update({
 102                 'url': format_url,
 103                 'format_id': '-'.join(format_id),
 104                 'format_note': meta.get('quality'),
 105                 'language': meta.get('language'),
 106                 'quality': qualities(self._QUALITIES)(meta.get('quality')),
 107                 'preference': -10,
 108             })
 109             formats.append(f)
 110
 111     def _extract_entry(self, url, player, content, video_id):
 112         title = content.get('title') or content['teaserHeadline']
 113
 114         t = content['mainVideoContent']['http://zdf.de/rels/target']
 115
 116         ptmd_path = t.get('http://zdf.de/rels/streams/ptmd')
 117
 118         if not ptmd_path:
 119             ptmd_path = t[
 120                 'http://zdf.de/rels/streams/ptmd-template'].replace(
 121                 '{playerId}', 'portal')
 122
 123         ptmd = self._call_api(
 124             urljoin(url, ptmd_path), player, url, video_id, 'metadata')
 125
 126         formats = []
 127         track_uris = set()
 128         for p in ptmd['priorityList']:
 129             formitaeten = p.get('formitaeten')
 130             if not isinstance(formitaeten, list):
 131                 continue
 132             for f in formitaeten:
 133                 f_qualities = f.get('qualities')
 134                 if not isinstance(f_qualities, list):
 135                     continue
 136                 for quality in f_qualities:
 137                     tracks = try_get(quality, lambda x: x['audio']['tracks'], list)
 138                     if not tracks:
 139                         continue
 140                     for track in tracks:
 141                         self._extract_format(
 142                             video_id, formats, track_uris, {
 143                                 'url': track.get('uri'),
 144                                 'type': f.get('type'),
 145                                 'mimeType': f.get('mimeType'),
 146                                 'quality': quality.get('quality'),
 147                                 'language': track.get('language'),
 148                             })
 149         self._sort_formats(formats)
 150
 151         thumbnails = []
 152         layouts = try_get(
 153             content, lambda x: x['teaserImageRef']['layouts'], dict)
 154         if layouts:
 155             for layout_key, layout_url in layouts.items():
 156                 layout_url = url_or_none(layout_url)
 157                 if not layout_url:
 158                     continue
 159                 thumbnail = {
 160                     'url': layout_url,
 161                     'format_id': layout_key,
 162                 }
 163                 mobj = re.search(r'(?P<width>\d+)x(?P<height>\d+)', layout_key)
 164                 if mobj:
 165                     thumbnail.update({
 166                         'width': int(mobj.group('width')),
 167                         'height': int(mobj.group('height')),
 168                     })
 169                 thumbnails.append(thumbnail)
 170
 171         return {
 172             'id': video_id,
 173             'title': title,
 174             'description': content.get('leadParagraph') or content.get('teasertext'),
 175             'duration': int_or_none(t.get('duration')),
 176             'timestamp': unified_timestamp(content.get('editorialDate')),
 177             'thumbnails': thumbnails,
 178             'subtitles': self._extract_subtitles(ptmd),
 179             'formats': formats,
 180         }
 181
 182     def _extract_regular(self, url, player, video_id):
 183         content = self._call_api(
 184             player['content'], player, url, video_id, 'content')
 185         return self._extract_entry(player['content'], player, content, video_id)
 186
 187     def _extract_mobile(self, video_id):
 188         document = self._download_json(
 189             'https://zdf-cdn.live.cellular.de/mediathekV2/document/%s' % video_id,
 190             video_id)['document']
 191
 192         title = document['titel']
 193
 194         formats = []
 195         format_urls = set()
 196         for f in document['formitaeten']:
 197             self._extract_format(video_id, formats, format_urls, f)
 198         self._sort_formats(formats)
 199
 200         thumbnails = []
 201         teaser_bild = document.get('teaserBild')
 202         if isinstance(teaser_bild, dict):
 203             for thumbnail_key, thumbnail in teaser_bild.items():
 204                 thumbnail_url = try_get(
 205                     thumbnail, lambda x: x['url'], compat_str)
 206                 if thumbnail_url:
 207                     thumbnails.append({
 208                         'url': thumbnail_url,
 209                         'id': thumbnail_key,
 210                         'width': int_or_none(thumbnail.get('width')),
 211                         'height': int_or_none(thumbnail.get('height')),
 212                     })
 213
 214         return {
 215             'id': video_id,
 216             'title': title,
 217             'description': document.get('beschreibung'),
 218             'duration': int_or_none(document.get('length')),
 219             'timestamp': unified_timestamp(try_get(
 220                 document, lambda x: x['meta']['editorialDate'], compat_str)),
 221             'thumbnails': thumbnails,
 222             'subtitles': self._extract_subtitles(document),
 223             'formats': formats,
 224         }
 225
 226     def _real_extract(self, url):
 227         video_id = self._match_id(url)
 228
 229         webpage = self._download_webpage(url, video_id, fatal=False)
 230         if webpage:
 231             player = self._extract_player(webpage, url, fatal=False)
 232             if player:
 233                 return self._extract_regular(url, player, video_id)
 234
 235         return self._extract_mobile(video_id)
 236
 237
 238 class ZDFChannelIE(ZDFBaseIE):
 239     _VALID_URL = r'https?://www\.zdf\.de/(?:[^/]+/)*(?P<id>[^/?#&]+)'
 240     _TESTS = [{
 241         'url': 'https://www.zdf.de/sport/das-aktuelle-sportstudio',
 242         'info_dict': {
 243             'id': 'das-aktuelle-sportstudio',
 244             'title': 'das aktuelle sportstudio | ZDF',
 245         },
 246         'playlist_count': 21,
 247     }, {
 248         'url': 'https://www.zdf.de/dokumentation/planet-e',
 249         'info_dict': {
 250             'id': 'planet-e',
 251             'title': 'planet e.',
 252         },
 253         'playlist_count': 4,
 254     }, {
 255         'url': 'https://www.zdf.de/filme/taunuskrimi/',
 256         'only_matching': True,
 257     }]
 258
 259     @classmethod
 260     def suitable(cls, url):
 261         return False if ZDFIE.suitable(url) else super(ZDFChannelIE, cls).suitable(url)
 262
 263     def _real_extract(self, url):
 264         channel_id = self._match_id(url)
 265
 266         webpage = self._download_webpage(url, channel_id)
 267
 268         entries = [
 269             self.url_result(item_url, ie=ZDFIE.ie_key())
 270             for item_url in orderedSet(re.findall(
 271                 r'data-plusbar-url=["\'](http.+?\.html)', webpage))]
 272
 273         return self.playlist_result(
 274             entries, channel_id, self._og_search_title(webpage, fatal=False))
 275
 276         r"""
 277         player = self._extract_player(webpage, channel_id)
 278
 279         channel_id = self._search_regex(
 280             r'docId\s*:\s*(["\'])(?P<id>(?!\1).+?)\1', webpage,
 281             'channel id', group='id')
 282
 283         channel = self._call_api(
 284             'https://api.zdf.de/content/documents/%s.json' % channel_id,
 285             player, url, channel_id)
 286
 287         items = []
 288         for module in channel['module']:
 289             for teaser in try_get(module, lambda x: x['teaser'], list) or []:
 290                 t = try_get(
 291                     teaser, lambda x: x['http://zdf.de/rels/target'], dict)
 292                 if not t:
 293                     continue
 294                 items.extend(try_get(
 295                     t,
 296                     lambda x: x['resultsWithVideo']['http://zdf.de/rels/search/results'],
 297                     list) or [])
 298             items.extend(try_get(
 299                 module,
 300                 lambda x: x['filterRef']['resultsWithVideo']['http://zdf.de/rels/search/results'],
 301                 list) or [])
 302
 303         entries = []
 304         entry_urls = set()
 305         for item in items:
 306             t = try_get(item, lambda x: x['http://zdf.de/rels/target'], dict)
 307             if not t:
 308                 continue
 309             sharing_url = t.get('http://zdf.de/rels/sharing-url')
 310             if not sharing_url or not isinstance(sharing_url, compat_str):
 311                 continue
 312             if sharing_url in entry_urls:
 313                 continue
 314             entry_urls.add(sharing_url)
 315             entries.append(self.url_result(
 316                 sharing_url, ie=ZDFIE.ie_key(), video_id=t.get('id')))
 317
 318         return self.playlist_result(entries, channel_id, channel.get('title'))
 319         """