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