Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/arte.py

   1 # encoding: utf-8
   2 import re
   3 import json
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     ExtractorError,
   8     find_xpath_attr,
   9     unified_strdate,
  10     determine_ext,
  11     get_element_by_id,
  12     compat_str,
  13 )
  14
  15 # There are different sources of video in arte.tv, the extraction process
  16 # is different for each one. The videos usually expire in 7 days, so we can't
  17 # add tests.
  18
  19 class ArteTvIE(InfoExtractor):
  20     _VIDEOS_URL = r'(?:http://)?videos.arte.tv/(?P<lang>fr|de)/.*-(?P<id>.*?).html'
  21     _LIVEWEB_URL = r'(?:http://)?liveweb.arte.tv/(?P<lang>fr|de)/(?P<subpage>.+?)/(?P<name>.+)'
  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._VIDEOS_URL, cls._LIVEWEB_URL))
  29
  30     # TODO implement Live Stream
  31     # from ..utils import compat_urllib_parse
  32     # def extractLiveStream(self, url):
  33     #     video_lang = url.split('/')[-4]
  34     #     info = self.grep_webpage(
  35     #         url,
  36     #         r'src="(.*?/videothek_js.*?\.js)',
  37     #         0,
  38     #         [
  39     #             (1, 'url', u'Invalid URL: %s' % url)
  40     #         ]
  41     #     )
  42     #     http_host = url.split('/')[2]
  43     #     next_url = 'http://%s%s' % (http_host, compat_urllib_parse.unquote(info.get('url')))
  44     #     info = self.grep_webpage(
  45     #         next_url,
  46     #         r'(s_artestras_scst_geoFRDE_' + video_lang + '.*?)\'.*?' +
  47     #             '(http://.*?\.swf).*?' +
  48     #             '(rtmp://.*?)\'',
  49     #         re.DOTALL,
  50     #         [
  51     #             (1, 'path',   u'could not extract video path: %s' % url),
  52     #             (2, 'player', u'could not extract video player: %s' % url),
  53     #             (3, 'url',    u'could not extract video url: %s' % url)
  54     #         ]
  55     #     )
  56     #     video_url = u'%s/%s' % (info.get('url'), info.get('path'))
  57
  58     def _real_extract(self, url):
  59         mobj = re.match(self._VIDEOS_URL, url)
  60         if mobj is not None:
  61             id = mobj.group('id')
  62             lang = mobj.group('lang')
  63             return self._extract_video(url, id, lang)
  64
  65         mobj = re.match(self._LIVEWEB_URL, url)
  66         if mobj is not None:
  67             name = mobj.group('name')
  68             lang = mobj.group('lang')
  69             return self._extract_liveweb(url, name, lang)
  70
  71         if re.search(self._LIVE_URL, url) 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_video(self, url, video_id, lang):
  77         """Extract from videos.arte.tv"""
  78         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
  79         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
  80         ref_xml_doc = self._download_xml(ref_xml_url, video_id, note=u'Downloading metadata')
  81         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
  82         config_xml_url = config_node.attrib['ref']
  83         config_xml = self._download_webpage(config_xml_url, video_id, note=u'Downloading configuration')
  84
  85         video_urls = list(re.finditer(r'<url quality="(?P<quality>.*?)">(?P<url>.*?)</url>', config_xml))
  86         def _key(m):
  87             quality = m.group('quality')
  88             if quality == 'hd':
  89                 return 2
  90             else:
  91                 return 1
  92         # We pick the best quality
  93         video_urls = sorted(video_urls, key=_key)
  94         video_url = list(video_urls)[-1].group('url')
  95
  96         title = self._html_search_regex(r'<name>(.*?)</name>', config_xml, 'title')
  97         thumbnail = self._html_search_regex(r'<firstThumbnailUrl>(.*?)</firstThumbnailUrl>',
  98                                             config_xml, 'thumbnail')
  99         return {'id': video_id,
 100                 'title': title,
 101                 'thumbnail': thumbnail,
 102                 'url': video_url,
 103                 'ext': 'flv',
 104                 }
 105
 106     def _extract_liveweb(self, url, name, lang):
 107         """Extract form http://liveweb.arte.tv/"""
 108         webpage = self._download_webpage(url, name)
 109         video_id = self._search_regex(r'eventId=(\d+?)("|&)', webpage, u'event id')
 110         config_doc = self._download_xml('http://download.liveweb.arte.tv/o21/liveweb/events/event-%s.xml' % video_id,
 111                                             video_id, u'Downloading information')
 112         event_doc = config_doc.find('event')
 113         url_node = event_doc.find('video').find('urlHd')
 114         if url_node is None:
 115             url_node = event_doc.find('urlSd')
 116
 117         return {'id': video_id,
 118                 'title': event_doc.find('name%s' % lang.capitalize()).text,
 119                 'url': url_node.text.replace('MP4', 'mp4'),
 120                 'ext': 'flv',
 121                 'thumbnail': self._og_search_thumbnail(webpage),
 122                 }
 123
 124
 125 class ArteTVPlus7IE(InfoExtractor):
 126     IE_NAME = u'arte.tv:+7'
 127     _VALID_URL = r'https?://www\.arte.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
 128
 129     @classmethod
 130     def _extract_url_info(cls, url):
 131         mobj = re.match(cls._VALID_URL, url)
 132         lang = mobj.group('lang')
 133         # This is not a real id, it can be for example AJT for the news
 134         # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
 135         video_id = mobj.group('id')
 136         return video_id, lang
 137
 138     def _real_extract(self, url):
 139         video_id, lang = self._extract_url_info(url)
 140         webpage = self._download_webpage(url, video_id)
 141         return self._extract_from_webpage(webpage, video_id, lang)
 142
 143     def _extract_from_webpage(self, webpage, video_id, lang):
 144         json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url')
 145
 146         json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
 147         self.report_extraction(video_id)
 148         info = json.loads(json_info)
 149         player_info = info['videoJsonPlayer']
 150
 151         info_dict = {
 152             'id': player_info['VID'],
 153             'title': player_info['VTI'],
 154             'description': player_info.get('VDE'),
 155             'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]),
 156             'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
 157         }
 158
 159         all_formats = player_info['VSR'].values()
 160         # Some formats use the m3u8 protocol
 161         all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats))
 162         def _match_lang(f):
 163             if f.get('versionCode') is None:
 164                 return True
 165             # Return true if that format is in the language of the url
 166             if lang == 'fr':
 167                 l = 'F'
 168             elif lang == 'de':
 169                 l = 'A'
 170             regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
 171             return any(re.match(r, f['versionCode']) for r in regexes)
 172         # Some formats may not be in the same language as the url
 173         formats = filter(_match_lang, all_formats)
 174         formats = list(formats) # in python3 filter returns an iterator
 175         if not formats:
 176             # Some videos are only available in the 'Originalversion'
 177             # they aren't tagged as being in French or German
 178             if all(f['versionCode'] == 'VO' for f in all_formats):
 179                 formats = all_formats
 180             else:
 181                 raise ExtractorError(u'The formats list is empty')
 182
 183         if re.match(r'[A-Z]Q', formats[0]['quality']) is not None:
 184             def sort_key(f):
 185                 return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality'])
 186         else:
 187             def sort_key(f):
 188                 return (
 189                     # Sort first by quality
 190                     int(f.get('height',-1)),
 191                     int(f.get('bitrate',-1)),
 192                     # The original version with subtitles has lower relevance
 193                     re.match(r'VO-ST(F|A)', f.get('versionCode', '')) is None,
 194                     # The version with sourds/mal subtitles has also lower relevance
 195                     re.match(r'VO?(F|A)-STM\1', f.get('versionCode', '')) is None,
 196                 )
 197         formats = sorted(formats, key=sort_key)
 198         def _format(format_info):
 199             quality = ''
 200             height = format_info.get('height')
 201             if height is not None:
 202                 quality = compat_str(height)
 203             bitrate = format_info.get('bitrate')
 204             if bitrate is not None:
 205                 quality += '-%d' % bitrate
 206             if format_info.get('versionCode') is not None:
 207                 format_id = u'%s-%s' % (quality, format_info['versionCode'])
 208             else:
 209                 format_id = quality
 210             info = {
 211                 'format_id': format_id,
 212                 'format_note': format_info.get('versionLibelle'),
 213                 'width': format_info.get('width'),
 214                 'height': height,
 215             }
 216             if format_info['mediaType'] == u'rtmp':
 217                 info['url'] = format_info['streamer']
 218                 info['play_path'] = 'mp4:' + format_info['url']
 219                 info['ext'] = 'flv'
 220             else:
 221                 info['url'] = format_info['url']
 222                 info['ext'] = determine_ext(info['url'])
 223             return info
 224         info_dict['formats'] = [_format(f) for f in formats]
 225
 226         return info_dict
 227
 228
 229 # It also uses the arte_vp_url url from the webpage to extract the information
 230 class ArteTVCreativeIE(ArteTVPlus7IE):
 231     IE_NAME = u'arte.tv:creative'
 232     _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/magazine?/(?P<id>.+)'
 233
 234     _TEST = {
 235         u'url': u'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
 236         u'file': u'050489-002.mp4',
 237         u'info_dict': {
 238             u'title': u'Agentur Amateur / Agence Amateur #2 : Corporate Design',
 239         },
 240     }
 241
 242
 243 class ArteTVFutureIE(ArteTVPlus7IE):
 244     IE_NAME = u'arte.tv:future'
 245     _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
 246
 247     _TEST = {
 248         u'url': u'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
 249         u'file': u'050940-003.mp4',
 250         u'info_dict': {
 251             u'title': u'Les champignons au secours de la planète',
 252         },
 253     }
 254
 255     def _real_extract(self, url):
 256         anchor_id, lang = self._extract_url_info(url)
 257         webpage = self._download_webpage(url, anchor_id)
 258         row = get_element_by_id(anchor_id, webpage)
 259         return self._extract_from_webpage(row, anchor_id, lang)