yt_dlp/extractor/br.py

   1 import json
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     determine_ext,
   6     ExtractorError,
   7     int_or_none,
   8     parse_duration,
   9     parse_iso8601,
  10     xpath_element,
  11     xpath_text,
  12 )
  13
  14
  15 class BRIE(InfoExtractor):
  16     IE_DESC = 'Bayerischer Rundfunk'
  17     _VALID_URL = r'(?P<base_url>https?://(?:www\.)?br(?:-klassik)?\.de)/(?:[a-z0-9\-_]+/)+(?P<id>[a-z0-9\-_]+)\.html'
  18
  19     _TESTS = [
  20         {
  21             'url': 'http://www.br.de/mediathek/video/sendungen/abendschau/betriebliche-altersvorsorge-104.html',
  22             'md5': '83a0477cf0b8451027eb566d88b51106',
  23             'info_dict': {
  24                 'id': '48f656ef-287e-486f-be86-459122db22cc',
  25                 'ext': 'mp4',
  26                 'title': 'Die böse Überraschung',
  27                 'description': 'md5:ce9ac81b466ce775b8018f6801b48ac9',
  28                 'duration': 180,
  29                 'uploader': 'Reinhard Weber',
  30                 'upload_date': '20150422',
  31             },
  32             'skip': '404 not found',
  33         },
  34         {
  35             'url': 'http://www.br.de/nachrichten/oberbayern/inhalt/muenchner-polizeipraesident-schreiber-gestorben-100.html',
  36             'md5': 'af3a3a4aa43ff0ce6a89504c67f427ef',
  37             'info_dict': {
  38                 'id': 'a4b83e34-123d-4b81-9f4e-c0d3121a4e05',
  39                 'ext': 'flv',
  40                 'title': 'Manfred Schreiber ist tot',
  41                 'description': 'md5:b454d867f2a9fc524ebe88c3f5092d97',
  42                 'duration': 26,
  43             },
  44             'skip': '404 not found',
  45         },
  46         {
  47             'url': 'https://www.br-klassik.de/audio/peeping-tom-premierenkritik-dance-festival-muenchen-100.html',
  48             'md5': '8b5b27c0b090f3b35eac4ab3f7a73d3d',
  49             'info_dict': {
  50                 'id': '74c603c9-26d3-48bb-b85b-079aeed66e0b',
  51                 'ext': 'aac',
  52                 'title': 'Kurzweilig und sehr bewegend',
  53                 'description': 'md5:0351996e3283d64adeb38ede91fac54e',
  54                 'duration': 296,
  55             },
  56             'skip': '404 not found',
  57         },
  58         {
  59             'url': 'http://www.br.de/radio/bayern1/service/team/videos/team-video-erdelt100.html',
  60             'md5': 'dbab0aef2e047060ea7a21fc1ce1078a',
  61             'info_dict': {
  62                 'id': '6ba73750-d405-45d3-861d-1ce8c524e059',
  63                 'ext': 'mp4',
  64                 'title': 'Umweltbewusster Häuslebauer',
  65                 'description': 'md5:d52dae9792d00226348c1dbb13c9bae2',
  66                 'duration': 116,
  67             }
  68         },
  69         {
  70             'url': 'http://www.br.de/fernsehen/br-alpha/sendungen/kant-fuer-anfaenger/kritik-der-reinen-vernunft/kant-kritik-01-metaphysik100.html',
  71             'md5': '23bca295f1650d698f94fc570977dae3',
  72             'info_dict': {
  73                 'id': 'd982c9ce-8648-4753-b358-98abb8aec43d',
  74                 'ext': 'mp4',
  75                 'title': 'Folge 1 - Metaphysik',
  76                 'description': 'md5:bb659990e9e59905c3d41e369db1fbe3',
  77                 'duration': 893,
  78                 'uploader': 'Eva Maria Steimle',
  79                 'upload_date': '20170208',
  80             }
  81         },
  82     ]
  83
  84     def _real_extract(self, url):
  85         base_url, display_id = self._match_valid_url(url).groups()
  86         page = self._download_webpage(url, display_id)
  87         xml_url = self._search_regex(
  88             r"return BRavFramework\.register\(BRavFramework\('avPlayer_(?:[a-f0-9-]{36})'\)\.setup\({dataURL:'(/(?:[a-z0-9\-]+/)+[a-z0-9/~_.-]+)'}\)\);", page, 'XMLURL')
  89         xml = self._download_xml(base_url + xml_url, display_id)
  90
  91         medias = []
  92
  93         for xml_media in xml.findall('video') + xml.findall('audio'):
  94             media_id = xml_media.get('externalId')
  95             media = {
  96                 'id': media_id,
  97                 'title': xpath_text(xml_media, 'title', 'title', True),
  98                 'duration': parse_duration(xpath_text(xml_media, 'duration')),
  99                 'formats': self._extract_formats(xpath_element(
 100                     xml_media, 'assets'), media_id),
 101                 'thumbnails': self._extract_thumbnails(xpath_element(
 102                     xml_media, 'teaserImage/variants'), base_url),
 103                 'description': xpath_text(xml_media, 'desc'),
 104                 'webpage_url': xpath_text(xml_media, 'permalink'),
 105                 'uploader': xpath_text(xml_media, 'author'),
 106             }
 107             broadcast_date = xpath_text(xml_media, 'broadcastDate')
 108             if broadcast_date:
 109                 media['upload_date'] = ''.join(reversed(broadcast_date.split('.')))
 110             medias.append(media)
 111
 112         if len(medias) > 1:
 113             self.report_warning(
 114                 'found multiple medias; please '
 115                 'report this with the video URL to http://yt-dl.org/bug')
 116         if not medias:
 117             raise ExtractorError('No media entries found')
 118         return medias[0]
 119
 120     def _extract_formats(self, assets, media_id):
 121         formats = []
 122         for asset in assets.findall('asset'):
 123             format_url = xpath_text(asset, ['downloadUrl', 'url'])
 124             asset_type = asset.get('type')
 125             if asset_type.startswith('HDS'):
 126                 formats.extend(self._extract_f4m_formats(
 127                     format_url + '?hdcore=3.2.0', media_id, f4m_id='hds', fatal=False))
 128             elif asset_type.startswith('HLS'):
 129                 formats.extend(self._extract_m3u8_formats(
 130                     format_url, media_id, 'mp4', 'm3u8_native', m3u8_id='hds', fatal=False))
 131             else:
 132                 format_info = {
 133                     'ext': xpath_text(asset, 'mediaType'),
 134                     'width': int_or_none(xpath_text(asset, 'frameWidth')),
 135                     'height': int_or_none(xpath_text(asset, 'frameHeight')),
 136                     'tbr': int_or_none(xpath_text(asset, 'bitrateVideo')),
 137                     'abr': int_or_none(xpath_text(asset, 'bitrateAudio')),
 138                     'vcodec': xpath_text(asset, 'codecVideo'),
 139                     'acodec': xpath_text(asset, 'codecAudio'),
 140                     'container': xpath_text(asset, 'mediaType'),
 141                     'filesize': int_or_none(xpath_text(asset, 'size')),
 142                 }
 143                 format_url = self._proto_relative_url(format_url)
 144                 if format_url:
 145                     http_format_info = format_info.copy()
 146                     http_format_info.update({
 147                         'url': format_url,
 148                         'format_id': 'http-%s' % asset_type,
 149                     })
 150                     formats.append(http_format_info)
 151                 server_prefix = xpath_text(asset, 'serverPrefix')
 152                 if server_prefix:
 153                     rtmp_format_info = format_info.copy()
 154                     rtmp_format_info.update({
 155                         'url': server_prefix,
 156                         'play_path': xpath_text(asset, 'fileName'),
 157                         'format_id': 'rtmp-%s' % asset_type,
 158                     })
 159                     formats.append(rtmp_format_info)
 160         return formats
 161
 162     def _extract_thumbnails(self, variants, base_url):
 163         thumbnails = [{
 164             'url': base_url + xpath_text(variant, 'url'),
 165             'width': int_or_none(xpath_text(variant, 'width')),
 166             'height': int_or_none(xpath_text(variant, 'height')),
 167         } for variant in variants.findall('variant') if xpath_text(variant, 'url')]
 168         thumbnails.sort(key=lambda x: x['width'] * x['height'], reverse=True)
 169         return thumbnails
 170
 171
 172 class BRMediathekIE(InfoExtractor):
 173     IE_DESC = 'Bayerischer Rundfunk Mediathek'
 174     _VALID_URL = r'https?://(?:www\.)?br\.de/mediathek//?video/(?:[^/?&#]+?-)?(?P<id>av:[0-9a-f]{24})'
 175
 176     _TESTS = [{
 177         'url': 'https://www.br.de/mediathek/video/gesundheit-die-sendung-vom-28112017-av:5a1e6a6e8fce6d001871cc8e',
 178         'md5': 'fdc3d485835966d1622587d08ba632ec',
 179         'info_dict': {
 180             'id': 'av:5a1e6a6e8fce6d001871cc8e',
 181             'ext': 'mp4',
 182             'title': 'Die Sendung vom 28.11.2017',
 183             'description': 'md5:6000cdca5912ab2277e5b7339f201ccc',
 184             'timestamp': 1511942766,
 185             'upload_date': '20171129',
 186         }
 187     }, {
 188         'url': 'https://www.br.de/mediathek//video/av:61b0db581aed360007558c12',
 189         'only_matching': True,
 190     }]
 191
 192     def _real_extract(self, url):
 193         clip_id = self._match_id(url)
 194
 195         clip = self._download_json(
 196             'https://proxy-base.master.mango.express/graphql',
 197             clip_id, data=json.dumps({
 198                 "query": """{
 199   viewer {
 200     clip(id: "%s") {
 201       title
 202       description
 203       duration
 204       createdAt
 205       ageRestriction
 206       videoFiles {
 207         edges {
 208           node {
 209             publicLocation
 210             fileSize
 211             videoProfile {
 212               width
 213               height
 214               bitrate
 215               encoding
 216             }
 217           }
 218         }
 219       }
 220       captionFiles {
 221         edges {
 222           node {
 223             publicLocation
 224           }
 225         }
 226       }
 227       teaserImages {
 228         edges {
 229           node {
 230             imageFiles {
 231               edges {
 232                 node {
 233                   publicLocation
 234                   width
 235                   height
 236                 }
 237               }
 238             }
 239           }
 240         }
 241       }
 242     }
 243   }
 244 }""" % clip_id}).encode(), headers={
 245                 'Content-Type': 'application/json',
 246             })['data']['viewer']['clip']
 247         title = clip['title']
 248
 249         formats = []
 250         for edge in clip.get('videoFiles', {}).get('edges', []):
 251             node = edge.get('node', {})
 252             n_url = node.get('publicLocation')
 253             if not n_url:
 254                 continue
 255             ext = determine_ext(n_url)
 256             if ext == 'm3u8':
 257                 formats.extend(self._extract_m3u8_formats(
 258                     n_url, clip_id, 'mp4', 'm3u8_native',
 259                     m3u8_id='hls', fatal=False))
 260             else:
 261                 video_profile = node.get('videoProfile', {})
 262                 tbr = int_or_none(video_profile.get('bitrate'))
 263                 format_id = 'http'
 264                 if tbr:
 265                     format_id += '-%d' % tbr
 266                 formats.append({
 267                     'format_id': format_id,
 268                     'url': n_url,
 269                     'width': int_or_none(video_profile.get('width')),
 270                     'height': int_or_none(video_profile.get('height')),
 271                     'tbr': tbr,
 272                     'filesize': int_or_none(node.get('fileSize')),
 273                 })
 274
 275         subtitles = {}
 276         for edge in clip.get('captionFiles', {}).get('edges', []):
 277             node = edge.get('node', {})
 278             n_url = node.get('publicLocation')
 279             if not n_url:
 280                 continue
 281             subtitles.setdefault('de', []).append({
 282                 'url': n_url,
 283             })
 284
 285         thumbnails = []
 286         for edge in clip.get('teaserImages', {}).get('edges', []):
 287             for image_edge in edge.get('node', {}).get('imageFiles', {}).get('edges', []):
 288                 node = image_edge.get('node', {})
 289                 n_url = node.get('publicLocation')
 290                 if not n_url:
 291                     continue
 292                 thumbnails.append({
 293                     'url': n_url,
 294                     'width': int_or_none(node.get('width')),
 295                     'height': int_or_none(node.get('height')),
 296                 })
 297
 298         return {
 299             'id': clip_id,
 300             'title': title,
 301             'description': clip.get('description'),
 302             'duration': int_or_none(clip.get('duration')),
 303             'timestamp': parse_iso8601(clip.get('createdAt')),
 304             'age_limit': int_or_none(clip.get('ageRestriction')),
 305             'formats': formats,
 306             'subtitles': subtitles,
 307             'thumbnails': thumbnails,
 308         }