3 import xml.etree.ElementTree
5 from .common import InfoExtractor
12 class ArteTvIE(InfoExtractor):
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.
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 _LIVEWEB_URL = r'(?:http://)?liveweb.arte.tv/(?P<lang>fr|de)/(?P<subpage>.+?)/(?P<name>.+)'
21 _LIVE_URL = r'index-[0-9]+\.html$'
26 def suitable(cls, url):
27 return any(re.match(regex, url) for regex in (cls._EMISSION_URL, cls._VIDEOS_URL, cls._LIVEWEB_URL))
29 # TODO implement Live Stream
30 # from ..utils import compat_urllib_parse
31 # def extractLiveStream(self, url):
32 # video_lang = url.split('/')[-4]
33 # info = self.grep_webpage(
35 # r'src="(.*?/videothek_js.*?\.js)',
38 # (1, 'url', u'Invalid URL: %s' % url)
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(
45 # r'(s_artestras_scst_geoFRDE_' + video_lang + '.*?)\'.*?' +
46 # '(http://.*?\.swf).*?' +
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)
55 # video_url = u'%s/%s' % (info.get('url'), info.get('path'))
57 def _real_extract(self, url):
58 mobj = re.match(self._EMISSION_URL, url)
60 lang = mobj.group('lang')
61 # This is not a real id, it can be for example AJT for the news
62 # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
63 video_id = mobj.group('id')
64 return self._extract_emission(url, video_id, lang)
66 mobj = re.match(self._VIDEOS_URL, url)
69 lang = mobj.group('lang')
70 return self._extract_video(url, id, lang)
72 mobj = re.match(self._LIVEWEB_URL, url)
74 name = mobj.group('name')
75 lang = mobj.group('lang')
76 return self._extract_liveweb(url, name, lang)
78 if re.search(self._LIVE_URL, video_id) is not None:
79 raise ExtractorError(u'Arte live streams are not yet supported, sorry')
80 # self.extractLiveStream(url)
83 def _extract_emission(self, url, video_id, lang):
84 """Extract from www.arte.tv/guide"""
85 webpage = self._download_webpage(url, video_id)
86 json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url')
88 json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
89 self.report_extraction(video_id)
90 info = json.loads(json_info)
91 player_info = info['videoJsonPlayer']
93 info_dict = {'id': player_info['VID'],
94 'title': player_info['VTI'],
95 'description': player_info.get('VDE'),
96 'upload_date': unified_strdate(player_info['VDA'].split(' ')[0]),
97 'thumbnail': player_info['programImage'],
101 formats = player_info['VSR'].values()
103 # Return true if that format is in the language of the url
108 regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
109 return any(re.match(r, f['versionCode']) for r in regexes)
110 # Some formats may not be in the same language as the url
111 formats = filter(_match_lang, formats)
112 # Some formats use the m3u8 protocol
113 formats = filter(lambda f: f['videoFormat'] != 'M3U8', formats)
114 # We order the formats by quality
115 formats = sorted(formats, key=lambda f: int(f['height']))
116 # Prefer videos without subtitles in the same language
117 formats = sorted(formats, key=lambda f: re.match(r'VO(F|A)-STM\1', f['versionCode']) is None)
118 # Pick the best quality
119 def _format(format_info):
120 info = {'ext': 'flv',
121 'width': format_info.get('width'),
122 'height': format_info.get('height'),
124 if format_info['mediaType'] == u'rtmp':
125 info['url'] = format_info['streamer']
126 info['play_path'] = 'mp4:' + format_info['url']
128 info_dict['url'] = format_info['url']
130 info_dict['formats'] = [_format(f) for f in formats]
131 # TODO: Remove when #980 has been merged
132 info_dict.update(info_dict['formats'][-1])
136 def _extract_video(self, url, video_id, lang):
137 """Extract from videos.arte.tv"""
138 ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
139 ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
140 ref_xml = self._download_webpage(ref_xml_url, video_id, note=u'Downloading metadata')
141 ref_xml_doc = xml.etree.ElementTree.fromstring(ref_xml)
142 config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
143 config_xml_url = config_node.attrib['ref']
144 config_xml = self._download_webpage(config_xml_url, video_id, note=u'Downloading configuration')
146 video_urls = list(re.finditer(r'<url quality="(?P<quality>.*?)">(?P<url>.*?)</url>', config_xml))
148 quality = m.group('quality')
153 # We pick the best quality
154 video_urls = sorted(video_urls, key=_key)
155 video_url = list(video_urls)[-1].group('url')
157 title = self._html_search_regex(r'<name>(.*?)</name>', config_xml, 'title')
158 thumbnail = self._html_search_regex(r'<firstThumbnailUrl>(.*?)</firstThumbnailUrl>',
159 config_xml, 'thumbnail')
160 return {'id': video_id,
162 'thumbnail': thumbnail,
167 def _extract_liveweb(self, url, name, lang):
168 """Extract form http://liveweb.arte.tv/"""
169 webpage = self._download_webpage(url, name)
170 video_id = self._search_regex(r'eventId=(\d+?)("|&)', webpage, u'event id')
171 config_xml = self._download_webpage('http://download.liveweb.arte.tv/o21/liveweb/events/event-%s.xml' % video_id,
172 video_id, u'Downloading information')
173 config_doc = xml.etree.ElementTree.fromstring(config_xml.encode('utf-8'))
174 event_doc = config_doc.find('event')
175 url_node = event_doc.find('video').find('urlHd')
177 url_node = video_doc.find('urlSd')
179 return {'id': video_id,
180 'title': event_doc.find('name%s' % lang.capitalize()).text,
181 'url': url_node.text.replace('MP4', 'mp4'),
183 'thumbnail': self._og_search_thumbnail(webpage),