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