youtube_dl/extractor/br.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..utils import (
   9     determine_ext,
  10     ExtractorError,
  11     int_or_none,
  12     parse_duration,
  13     parse_iso8601,
  14     xpath_element,
  15     xpath_text,
  16 )
  17
  18
  19 class BRIE(InfoExtractor):
  20     IE_DESC = 'Bayerischer Rundfunk'
  21     _VALID_URL = r'(?P<base_url>https?://(?:www\.)?br(?:-klassik)?\.de)/(?:[a-z0-9\-_]+/)+(?P<id>[a-z0-9\-_]+)\.html'
  22
  23     _TESTS = [
  24         {
  25             'url': 'http://www.br.de/mediathek/video/sendungen/abendschau/betriebliche-altersvorsorge-104.html',
  26             'md5': '83a0477cf0b8451027eb566d88b51106',
  27             'info_dict': {
  28                 'id': '48f656ef-287e-486f-be86-459122db22cc',
  29                 'ext': 'mp4',
  30                 'title': 'Die böse Überraschung',
  31                 'description': 'md5:ce9ac81b466ce775b8018f6801b48ac9',
  32                 'duration': 180,
  33                 'uploader': 'Reinhard Weber',
  34                 'upload_date': '20150422',
  35             },
  36             'skip': '404 not found',
  37         },
  38         {
  39             'url': 'http://www.br.de/nachrichten/oberbayern/inhalt/muenchner-polizeipraesident-schreiber-gestorben-100.html',
  40             'md5': 'af3a3a4aa43ff0ce6a89504c67f427ef',
  41             'info_dict': {
  42                 'id': 'a4b83e34-123d-4b81-9f4e-c0d3121a4e05',
  43                 'ext': 'flv',
  44                 'title': 'Manfred Schreiber ist tot',
  45                 'description': 'md5:b454d867f2a9fc524ebe88c3f5092d97',
  46                 'duration': 26,
  47             },
  48             'skip': '404 not found',
  49         },
  50         {
  51             'url': 'https://www.br-klassik.de/audio/peeping-tom-premierenkritik-dance-festival-muenchen-100.html',
  52             'md5': '8b5b27c0b090f3b35eac4ab3f7a73d3d',
  53             'info_dict': {
  54                 'id': '74c603c9-26d3-48bb-b85b-079aeed66e0b',
  55                 'ext': 'aac',
  56                 'title': 'Kurzweilig und sehr bewegend',
  57                 'description': 'md5:0351996e3283d64adeb38ede91fac54e',
  58                 'duration': 296,
  59             },
  60             'skip': '404 not found',
  61         },
  62         {
  63             'url': 'http://www.br.de/radio/bayern1/service/team/videos/team-video-erdelt100.html',
  64             'md5': 'dbab0aef2e047060ea7a21fc1ce1078a',
  65             'info_dict': {
  66                 'id': '6ba73750-d405-45d3-861d-1ce8c524e059',
  67                 'ext': 'mp4',
  68                 'title': 'Umweltbewusster Häuslebauer',
  69                 'description': 'md5:d52dae9792d00226348c1dbb13c9bae2',
  70                 'duration': 116,
  71             }
  72         },
  73         {
  74             'url': 'http://www.br.de/fernsehen/br-alpha/sendungen/kant-fuer-anfaenger/kritik-der-reinen-vernunft/kant-kritik-01-metaphysik100.html',
  75             'md5': '23bca295f1650d698f94fc570977dae3',
  76             'info_dict': {
  77                 'id': 'd982c9ce-8648-4753-b358-98abb8aec43d',
  78                 'ext': 'mp4',
  79                 'title': 'Folge 1 - Metaphysik',
  80                 'description': 'md5:bb659990e9e59905c3d41e369db1fbe3',
  81                 'duration': 893,
  82                 'uploader': 'Eva Maria Steimle',
  83                 'upload_date': '20170208',
  84             }
  85         },
  86     ]
  87
  88     def _real_extract(self, url):
  89         base_url, display_id = re.search(self._VALID_URL, url).groups()
  90         page = self._download_webpage(url, display_id)
  91         xml_url = self._search_regex(
  92             r"return BRavFramework\.register\(BRavFramework\('avPlayer_(?:[a-f0-9-]{36})'\)\.setup\({dataURL:'(/(?:[a-z0-9\-]+/)+[a-z0-9/~_.-]+)'}\)\);", page, 'XMLURL')
  93         xml = self._download_xml(base_url + xml_url, display_id)
  94
  95         medias = []
  96
  97         for xml_media in xml.findall('video') + xml.findall('audio'):
  98             media_id = xml_media.get('externalId')
  99             media = {
 100                 'id': media_id,
 101                 'title': xpath_text(xml_media, 'title', 'title', True),
 102                 'duration': parse_duration(xpath_text(xml_media, 'duration')),
 103                 'formats': self._extract_formats(xpath_element(
 104                     xml_media, 'assets'), media_id),
 105                 'thumbnails': self._extract_thumbnails(xpath_element(
 106                     xml_media, 'teaserImage/variants'), base_url),
 107                 'description': xpath_text(xml_media, 'desc'),
 108                 'webpage_url': xpath_text(xml_media, 'permalink'),
 109                 'uploader': xpath_text(xml_media, 'author'),
 110             }
 111             broadcast_date = xpath_text(xml_media, 'broadcastDate')
 112             if broadcast_date:
 113                 media['upload_date'] = ''.join(reversed(broadcast_date.split('.')))
 114             medias.append(media)
 115
 116         if len(medias) > 1:
 117             self._downloader.report_warning(
 118                 'found multiple medias; please '
 119                 'report this with the video URL to http://yt-dl.org/bug')
 120         if not medias:
 121             raise ExtractorError('No media entries found')
 122         return medias[0]
 123
 124     def _extract_formats(self, assets, media_id):
 125         formats = []
 126         for asset in assets.findall('asset'):
 127             format_url = xpath_text(asset, ['downloadUrl', 'url'])
 128             asset_type = asset.get('type')
 129             if asset_type.startswith('HDS'):
 130                 formats.extend(self._extract_f4m_formats(
 131                     format_url + '?hdcore=3.2.0', media_id, f4m_id='hds', fatal=False))
 132             elif asset_type.startswith('HLS'):
 133                 formats.extend(self._extract_m3u8_formats(
 134                     format_url, media_id, 'mp4', 'm3u8_native', m3u8_id='hds', fatal=False))
 135             else:
 136                 format_info = {
 137                     'ext': xpath_text(asset, 'mediaType'),
 138                     'width': int_or_none(xpath_text(asset, 'frameWidth')),
 139                     'height': int_or_none(xpath_text(asset, 'frameHeight')),
 140                     'tbr': int_or_none(xpath_text(asset, 'bitrateVideo')),
 141                     'abr': int_or_none(xpath_text(asset, 'bitrateAudio')),
 142                     'vcodec': xpath_text(asset, 'codecVideo'),
 143                     'acodec': xpath_text(asset, 'codecAudio'),
 144                     'container': xpath_text(asset, 'mediaType'),
 145                     'filesize': int_or_none(xpath_text(asset, 'size')),
 146                 }
 147                 format_url = self._proto_relative_url(format_url)
 148                 if format_url:
 149                     http_format_info = format_info.copy()
 150                     http_format_info.update({
 151                         'url': format_url,
 152                         'format_id': 'http-%s' % asset_type,
 153                     })
 154                     formats.append(http_format_info)
 155                 server_prefix = xpath_text(asset, 'serverPrefix')
 156                 if server_prefix:
 157                     rtmp_format_info = format_info.copy()
 158                     rtmp_format_info.update({
 159                         'url': server_prefix,
 160                         'play_path': xpath_text(asset, 'fileName'),
 161                         'format_id': 'rtmp-%s' % asset_type,
 162                     })
 163                     formats.append(rtmp_format_info)
 164         self._sort_formats(formats)
 165         return formats
 166
 167     def _extract_thumbnails(self, variants, base_url):
 168         thumbnails = [{
 169             'url': base_url + xpath_text(variant, 'url'),
 170             'width': int_or_none(xpath_text(variant, 'width')),
 171             'height': int_or_none(xpath_text(variant, 'height')),
 172         } for variant in variants.findall('variant') if xpath_text(variant, 'url')]
 173         thumbnails.sort(key=lambda x: x['width'] * x['height'], reverse=True)
 174         return thumbnails
 175
 176
 177 class BRMediathekIE(InfoExtractor):
 178     IE_DESC = 'Bayerischer Rundfunk Mediathek'
 179     _VALID_URL = r'https?://(?:www\.)?br\.de/mediathek/video/[^/?&#]*?-(?P<id>av:[0-9a-f]{24})'
 180
 181     _TESTS = [{
 182         'url': 'https://www.br.de/mediathek/video/gesundheit-die-sendung-vom-28112017-av:5a1e6a6e8fce6d001871cc8e',
 183         'md5': 'fdc3d485835966d1622587d08ba632ec',
 184         'info_dict': {
 185             'id': 'av:5a1e6a6e8fce6d001871cc8e',
 186             'ext': 'mp4',
 187             'title': 'Die Sendung vom 28.11.2017',
 188             'description': 'md5:6000cdca5912ab2277e5b7339f201ccc',
 189             'timestamp': 1511942766,
 190             'upload_date': '20171129',
 191         }
 192     }]
 193
 194     def _real_extract(self, url):
 195         clip_id = self._match_id(url)
 196
 197         clip = self._download_json(
 198             'https://proxy-base.master.mango.express/graphql',
 199             clip_id, data=json.dumps({
 200                 "query": """{
 201   viewer {
 202     clip(id: "%s") {
 203       title
 204       description
 205       duration
 206       createdAt
 207       ageRestriction
 208       videoFiles {
 209         edges {
 210           node {
 211             publicLocation
 212             fileSize
 213             videoProfile {
 214               width
 215               height
 216               bitrate
 217               encoding
 218             }
 219           }
 220         }
 221       }
 222       captionFiles {
 223         edges {
 224           node {
 225             publicLocation
 226           }
 227         }
 228       }
 229       teaserImages {
 230         edges {
 231           node {
 232             imageFiles {
 233               edges {
 234                 node {
 235                   publicLocation
 236                   width
 237                   height
 238                 }
 239               }
 240             }
 241           }
 242         }
 243       }
 244     }
 245   }
 246 }""" % clip_id}).encode(), headers={
 247                 'Content-Type': 'application/json',
 248             })['data']['viewer']['clip']
 249         title = clip['title']
 250
 251         formats = []
 252         for edge in clip.get('videoFiles', {}).get('edges', []):
 253             node = edge.get('node', {})
 254             n_url = node.get('publicLocation')
 255             if not n_url:
 256                 continue
 257             ext = determine_ext(n_url)
 258             if ext == 'm3u8':
 259                 formats.extend(self._extract_m3u8_formats(
 260                     n_url, clip_id, 'mp4', 'm3u8_native',
 261                     m3u8_id='hls', fatal=False))
 262             else:
 263                 video_profile = node.get('videoProfile', {})
 264                 tbr = int_or_none(video_profile.get('bitrate'))
 265                 format_id = 'http'
 266                 if tbr:
 267                     format_id += '-%d' % tbr
 268                 formats.append({
 269                     'format_id': format_id,
 270                     'url': n_url,
 271                     'width': int_or_none(video_profile.get('width')),
 272                     'height': int_or_none(video_profile.get('height')),
 273                     'tbr': tbr,
 274                     'filesize': int_or_none(node.get('fileSize')),
 275                 })
 276         self._sort_formats(formats)
 277
 278         subtitles = {}
 279         for edge in clip.get('captionFiles', {}).get('edges', []):
 280             node = edge.get('node', {})
 281             n_url = node.get('publicLocation')
 282             if not n_url:
 283                 continue
 284             subtitles.setdefault('de', []).append({
 285                 'url': n_url,
 286             })
 287
 288         thumbnails = []
 289         for edge in clip.get('teaserImages', {}).get('edges', []):
 290             for image_edge in edge.get('node', {}).get('imageFiles', {}).get('edges', []):
 291                 node = image_edge.get('node', {})
 292                 n_url = node.get('publicLocation')
 293                 if not n_url:
 294                     continue
 295                 thumbnails.append({
 296                     'url': n_url,
 297                     'width': int_or_none(node.get('width')),
 298                     'height': int_or_none(node.get('height')),
 299                 })
 300
 301         return {
 302             'id': clip_id,
 303             'title': title,
 304             'description': clip.get('description'),
 305             'duration': int_or_none(clip.get('duration')),
 306             'timestamp': parse_iso8601(clip.get('createdAt')),
 307             'age_limit': int_or_none(clip.get('ageRestriction')),
 308             'formats': formats,
 309             'subtitles': subtitles,
 310             'thumbnails': thumbnails,
 311         }