youtube_dl/extractor/ard.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from .generic import GenericIE
   8 from ..utils import (
   9     determine_ext,
  10     ExtractorError,
  11     qualities,
  12     int_or_none,
  13     parse_duration,
  14     unified_strdate,
  15     xpath_text,
  16     update_url_query,
  17 )
  18 from ..compat import compat_etree_fromstring
  19
  20
  21 class ARDMediathekIE(InfoExtractor):
  22     IE_NAME = 'ARD:mediathek'
  23     _VALID_URL = r'^https?://(?:(?:www\.)?ardmediathek\.de|mediathek\.daserste\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
  24
  25     _TESTS = [{
  26         'url': 'http://www.ardmediathek.de/tv/Dokumentation-und-Reportage/Ich-liebe-das-Leben-trotzdem/rbb-Fernsehen/Video?documentId=29582122&bcastId=3822114',
  27         'info_dict': {
  28             'id': '29582122',
  29             'ext': 'mp4',
  30             'title': 'Ich liebe das Leben trotzdem',
  31             'description': 'md5:45e4c225c72b27993314b31a84a5261c',
  32             'duration': 4557,
  33         },
  34         'params': {
  35             # m3u8 download
  36             'skip_download': True,
  37         },
  38         'skip': 'HTTP Error 404: Not Found',
  39     }, {
  40         'url': 'http://www.ardmediathek.de/tv/Tatort/Tatort-Scheinwelten-H%C3%B6rfassung-Video/Das-Erste/Video?documentId=29522730&bcastId=602916',
  41         'md5': 'f4d98b10759ac06c0072bbcd1f0b9e3e',
  42         'info_dict': {
  43             'id': '29522730',
  44             'ext': 'mp4',
  45             'title': 'Tatort: Scheinwelten - Hörfassung (Video tgl. ab 20 Uhr)',
  46             'description': 'md5:196392e79876d0ac94c94e8cdb2875f1',
  47             'duration': 5252,
  48         },
  49         'skip': 'HTTP Error 404: Not Found',
  50     }, {
  51         # audio
  52         'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
  53         'md5': '219d94d8980b4f538c7fcb0865eb7f2c',
  54         'info_dict': {
  55             'id': '28488308',
  56             'ext': 'mp3',
  57             'title': 'Tod eines Fußballers',
  58             'description': 'md5:f6e39f3461f0e1f54bfa48c8875c86ef',
  59             'duration': 3240,
  60         },
  61         'skip': 'HTTP Error 404: Not Found',
  62     }, {
  63         'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
  64         'only_matching': True,
  65     }]
  66
  67     def _extract_media_info(self, media_info_url, webpage, video_id):
  68         media_info = self._download_json(
  69             media_info_url, video_id, 'Downloading media JSON')
  70
  71         formats = self._extract_formats(media_info, video_id)
  72
  73         if not formats:
  74             if '"fsk"' in webpage:
  75                 raise ExtractorError(
  76                     'This video is only available after 20:00', expected=True)
  77             elif media_info.get('_geoblocked'):
  78                 raise ExtractorError('This video is not available due to geo restriction', expected=True)
  79
  80         self._sort_formats(formats)
  81
  82         duration = int_or_none(media_info.get('_duration'))
  83         thumbnail = media_info.get('_previewImage')
  84
  85         subtitles = {}
  86         subtitle_url = media_info.get('_subtitleUrl')
  87         if subtitle_url:
  88             subtitles['de'] = [{
  89                 'ext': 'ttml',
  90                 'url': subtitle_url,
  91             }]
  92
  93         return {
  94             'id': video_id,
  95             'duration': duration,
  96             'thumbnail': thumbnail,
  97             'formats': formats,
  98             'subtitles': subtitles,
  99         }
 100
 101     def _extract_formats(self, media_info, video_id):
 102         type_ = media_info.get('_type')
 103         media_array = media_info.get('_mediaArray', [])
 104         formats = []
 105         for num, media in enumerate(media_array):
 106             for stream in media.get('_mediaStreamArray', []):
 107                 stream_urls = stream.get('_stream')
 108                 if not stream_urls:
 109                     continue
 110                 if not isinstance(stream_urls, list):
 111                     stream_urls = [stream_urls]
 112                 quality = stream.get('_quality')
 113                 server = stream.get('_server')
 114                 for stream_url in stream_urls:
 115                     ext = determine_ext(stream_url)
 116                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
 117                         continue
 118                     if ext == 'f4m':
 119                         formats.extend(self._extract_f4m_formats(
 120                             update_url_query(stream_url, {
 121                                 'hdcore': '3.1.1',
 122                                 'plugin': 'aasp-3.1.1.69.124'
 123                             }),
 124                             video_id, f4m_id='hds', fatal=False))
 125                     elif ext == 'm3u8':
 126                         formats.extend(self._extract_m3u8_formats(
 127                             stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
 128                     else:
 129                         if server and server.startswith('rtmp'):
 130                             f = {
 131                                 'url': server,
 132                                 'play_path': stream_url,
 133                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
 134                             }
 135                         elif stream_url.startswith('http'):
 136                             f = {
 137                                 'url': stream_url,
 138                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
 139                             }
 140                         else:
 141                             continue
 142                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
 143                         if m:
 144                             f.update({
 145                                 'width': int(m.group('width')),
 146                                 'height': int(m.group('height')),
 147                             })
 148                         if type_ == 'audio':
 149                             f['vcodec'] = 'none'
 150                         formats.append(f)
 151         return formats
 152
 153     def _real_extract(self, url):
 154         # determine video id from url
 155         m = re.match(self._VALID_URL, url)
 156
 157         numid = re.search(r'documentId=([0-9]+)', url)
 158         if numid:
 159             video_id = numid.group(1)
 160         else:
 161             video_id = m.group('video_id')
 162
 163         webpage = self._download_webpage(url, video_id)
 164
 165         if '>Der gewünschte Beitrag ist nicht mehr verfügbar.<' in webpage:
 166             raise ExtractorError('Video %s is no longer available' % video_id, expected=True)
 167
 168         if 'Diese Sendung ist für Jugendliche unter 12 Jahren nicht geeignet. Der Clip ist deshalb nur von 20 bis 6 Uhr verfügbar.' in webpage:
 169             raise ExtractorError('This program is only suitable for those aged 12 and older. Video %s is therefore only available between 20 pm and 6 am.' % video_id, expected=True)
 170
 171         if re.search(r'[\?&]rss($|[=&])', url):
 172             doc = compat_etree_fromstring(webpage.encode('utf-8'))
 173             if doc.tag == 'rss':
 174                 return GenericIE()._extract_rss(url, video_id, doc)
 175
 176         title = self._html_search_regex(
 177             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
 178              r'<meta name="dcterms.title" content="(.*?)"/>',
 179              r'<h4 class="headline">(.*?)</h4>'],
 180             webpage, 'title')
 181         description = self._html_search_meta(
 182             'dcterms.abstract', webpage, 'description', default=None)
 183         if description is None:
 184             description = self._html_search_meta(
 185                 'description', webpage, 'meta description')
 186
 187         # Thumbnail is sometimes not present.
 188         # It is in the mobile version, but that seems to use a different URL
 189         # structure altogether.
 190         thumbnail = self._og_search_thumbnail(webpage, default=None)
 191
 192         media_streams = re.findall(r'''(?x)
 193             mediaCollection\.addMediaStream\([0-9]+,\s*[0-9]+,\s*"[^"]*",\s*
 194             "([^"]+)"''', webpage)
 195
 196         if media_streams:
 197             QUALITIES = qualities(['lo', 'hi', 'hq'])
 198             formats = []
 199             for furl in set(media_streams):
 200                 if furl.endswith('.f4m'):
 201                     fid = 'f4m'
 202                 else:
 203                     fid_m = re.match(r'.*\.([^.]+)\.[^.]+$', furl)
 204                     fid = fid_m.group(1) if fid_m else None
 205                 formats.append({
 206                     'quality': QUALITIES(fid),
 207                     'format_id': fid,
 208                     'url': furl,
 209                 })
 210             self._sort_formats(formats)
 211             info = {
 212                 'formats': formats,
 213             }
 214         else:  # request JSON file
 215             info = self._extract_media_info(
 216                 'http://www.ardmediathek.de/play/media/%s' % video_id, webpage, video_id)
 217
 218         info.update({
 219             'id': video_id,
 220             'title': title,
 221             'description': description,
 222             'thumbnail': thumbnail,
 223         })
 224
 225         return info
 226
 227
 228 class ARDIE(InfoExtractor):
 229     _VALID_URL = '(?P<mainurl>https?://(www\.)?daserste\.de/[^?#]+/videos/(?P<display_id>[^/?#]+)-(?P<id>[0-9]+))\.html'
 230     _TEST = {
 231         'url': 'http://www.daserste.de/information/reportage-dokumentation/dokus/videos/die-story-im-ersten-mission-unter-falscher-flagge-100.html',
 232         'md5': 'd216c3a86493f9322545e045ddc3eb35',
 233         'info_dict': {
 234             'display_id': 'die-story-im-ersten-mission-unter-falscher-flagge',
 235             'id': '100',
 236             'ext': 'mp4',
 237             'duration': 2600,
 238             'title': 'Die Story im Ersten: Mission unter falscher Flagge',
 239             'upload_date': '20140804',
 240             'thumbnail': 're:^https?://.*\.jpg$',
 241         },
 242         'skip': 'HTTP Error 404: Not Found',
 243     }
 244
 245     def _real_extract(self, url):
 246         mobj = re.match(self._VALID_URL, url)
 247         display_id = mobj.group('display_id')
 248
 249         player_url = mobj.group('mainurl') + '~playerXml.xml'
 250         doc = self._download_xml(player_url, display_id)
 251         video_node = doc.find('./video')
 252         upload_date = unified_strdate(xpath_text(
 253             video_node, './broadcastDate'))
 254         thumbnail = xpath_text(video_node, './/teaserImage//variant/url')
 255
 256         formats = []
 257         for a in video_node.findall('.//asset'):
 258             f = {
 259                 'format_id': a.attrib['type'],
 260                 'width': int_or_none(a.find('./frameWidth').text),
 261                 'height': int_or_none(a.find('./frameHeight').text),
 262                 'vbr': int_or_none(a.find('./bitrateVideo').text),
 263                 'abr': int_or_none(a.find('./bitrateAudio').text),
 264                 'vcodec': a.find('./codecVideo').text,
 265                 'tbr': int_or_none(a.find('./totalBitrate').text),
 266             }
 267             if a.find('./serverPrefix').text:
 268                 f['url'] = a.find('./serverPrefix').text
 269                 f['playpath'] = a.find('./fileName').text
 270             else:
 271                 f['url'] = a.find('./fileName').text
 272             formats.append(f)
 273         self._sort_formats(formats)
 274
 275         return {
 276             'id': mobj.group('id'),
 277             'formats': formats,
 278             'display_id': display_id,
 279             'title': video_node.find('./title').text,
 280             'duration': parse_duration(video_node.find('./duration').text),
 281             'upload_date': upload_date,
 282             'thumbnail': thumbnail,
 283         }