youtube_dl/extractor/arte.py

   1 import re
   2 import json
   3 import xml.etree.ElementTree
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     # This is used by the not implemented extractLiveStream method
   8     compat_urllib_parse,
   9
  10     ExtractorError,
  11     unified_strdate,
  12 )
  13
  14 class ArteTvIE(InfoExtractor):
  15     """
  16     There are two sources of video in arte.tv: videos.arte.tv and
  17     www.arte.tv/guide, the extraction process is different for each one.
  18     The videos expire in 7 days, so we can't add tests.
  19     """
  20     _EMISSION_URL = r'(?:http://)?www\.arte.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
  21     _VIDEOS_URL = r'(?:http://)?videos.arte.tv/(?P<lang>fr|de)/.*-(?P<id>.*?).html'
  22     _LIVE_URL = r'index-[0-9]+\.html$'
  23
  24     IE_NAME = u'arte.tv'
  25
  26     @classmethod
  27     def suitable(cls, url):
  28         return any(re.match(regex, url) for regex in (cls._EMISSION_URL, cls._VIDEOS_URL))
  29
  30     # TODO implement Live Stream
  31     # def extractLiveStream(self, url):
  32     #     video_lang = url.split('/')[-4]
  33     #     info = self.grep_webpage(
  34     #         url,
  35     #         r'src="(.*?/videothek_js.*?\.js)',
  36     #         0,
  37     #         [
  38     #             (1, 'url', u'Invalid URL: %s' % url)
  39     #         ]
  40     #     )
  41     #     http_host = url.split('/')[2]
  42     #     next_url = 'http://%s%s' % (http_host, compat_urllib_parse.unquote(info.get('url')))
  43     #     info = self.grep_webpage(
  44     #         next_url,
  45     #         r'(s_artestras_scst_geoFRDE_' + video_lang + '.*?)\'.*?' +
  46     #             '(http://.*?\.swf).*?' +
  47     #             '(rtmp://.*?)\'',
  48     #         re.DOTALL,
  49     #         [
  50     #             (1, 'path',   u'could not extract video path: %s' % url),
  51     #             (2, 'player', u'could not extract video player: %s' % url),
  52     #             (3, 'url',    u'could not extract video url: %s' % url)
  53     #         ]
  54     #     )
  55     #     video_url = u'%s/%s' % (info.get('url'), info.get('path'))
  56
  57     def _real_extract(self, url):
  58         mobj = re.match(self._EMISSION_URL, url)
  59         if mobj is not None:
  60             name = mobj.group('name')
  61             lang = mobj.group('lang')
  62             # This is not a real id, it can be for example AJT for the news
  63             # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
  64             video_id = mobj.group('id')
  65             return self._extract_emission(url, video_id, lang)
  66
  67         mobj = re.match(self._VIDEOS_URL, url)
  68         if mobj is not None:
  69             id = mobj.group('id')
  70             lang = mobj.group('lang')
  71             return self._extract_video(url, id, lang)
  72
  73         if re.search(self._LIVE_URL, video_id) is not None:
  74             raise ExtractorError(u'Arte live streams are not yet supported, sorry')
  75             # self.extractLiveStream(url)
  76             # return
  77
  78     def _extract_emission(self, url, video_id, lang):
  79         """Extract from www.arte.tv/guide"""
  80         webpage = self._download_webpage(url, video_id)
  81         json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url')
  82
  83         json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
  84         self.report_extraction(video_id)
  85         info = json.loads(json_info)
  86         player_info = info['videoJsonPlayer']
  87
  88         info_dict = {'id': player_info['VID'],
  89                      'title': player_info['VTI'],
  90                      'description': player_info['VDE'],
  91                      'upload_date': unified_strdate(player_info['VDA'].split(' ')[0]),
  92                      'thumbnail': player_info['programImage'],
  93                      'ext': 'flv',
  94                      }
  95
  96         formats = player_info['VSR'].values()
  97         def _match_lang(f):
  98             # Return true if that format is in the language of the url
  99             if lang == 'fr':
 100                 l = 'F'
 101             elif lang == 'de':
 102                 l = 'A'
 103             regexes = [r'VO?%s' % l, r'V%s-ST.' % l]
 104             return any(re.match(r, f['versionCode']) for r in regexes)
 105         # Some formats may not be in the same language as the url
 106         formats = filter(_match_lang, formats)
 107         # We order the formats by quality
 108         formats = sorted(formats, key=lambda f: int(f['height']))
 109         # Pick the best quality
 110         format_info = formats[-1]
 111         if format_info['mediaType'] == u'rtmp':
 112             info_dict['url'] = format_info['streamer']
 113             info_dict['play_path'] = 'mp4:' + format_info['url']
 114         else:
 115             info_dict['url'] = format_info['url']
 116
 117         return info_dict
 118
 119     def _extract_video(self, url, video_id, lang):
 120         """Extract from videos.arte.tv"""
 121         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
 122         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
 123         ref_xml = self._download_webpage(ref_xml_url, video_id, note=u'Downloading metadata')
 124         ref_xml_doc = xml.etree.ElementTree.fromstring(ref_xml)
 125         config_node = ref_xml_doc.find('.//video[@lang="%s"]' % lang)
 126         config_xml_url = config_node.attrib['ref']
 127         config_xml = self._download_webpage(config_xml_url, video_id, note=u'Downloading configuration')
 128
 129         video_urls = list(re.finditer(r'<url quality="(?P<quality>.*?)">(?P<url>.*?)</url>', config_xml))
 130         def _key(m):
 131             quality = m.group('quality')
 132             if quality == 'hd':
 133                 return 2
 134             else:
 135                 return 1
 136         # We pick the best quality
 137         video_urls = sorted(video_urls, key=_key)
 138         video_url = list(video_urls)[-1].group('url')
 139
 140         title = self._html_search_regex(r'<name>(.*?)</name>', config_xml, 'title')
 141         thumbnail = self._html_search_regex(r'<firstThumbnailUrl>(.*?)</firstThumbnailUrl>',
 142                                             config_xml, 'thumbnail')
 143         return {'id': video_id,
 144                 'title': title,
 145                 'thumbnail': thumbnail,
 146                 'url': video_url,
 147                 'ext': 'flv',
 148                 }