2 from __future__ import unicode_literals
6 from .common import InfoExtractor
14 get_element_by_attribute,
17 # There are different sources of video in arte.tv, the extraction process
18 # is different for each one. The videos usually expire in 7 days, so we can't
22 class ArteTvIE(InfoExtractor):
23 _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
26 def _real_extract(self, url):
27 mobj = re.match(self._VALID_URL, url)
28 lang = mobj.group('lang')
29 video_id = mobj.group('id')
31 ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
32 ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
33 ref_xml_doc = self._download_xml(
34 ref_xml_url, video_id, note='Downloading metadata')
35 config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
36 config_xml_url = config_node.attrib['ref']
37 config = self._download_xml(
38 config_xml_url, video_id, note='Downloading configuration')
41 'forma_id': q.attrib['quality'],
42 # The playpath starts at 'mp4:', if we don't manually
43 # split the url, rtmpdump will incorrectly parse them
44 'url': q.text.split('mp4:', 1)[0],
45 'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
47 'quality': 2 if q.attrib['quality'] == 'hd' else 1,
48 } for q in config.findall('./urls/url')]
49 self._sort_formats(formats)
51 title = config.find('.//name').text
52 thumbnail = config.find('.//firstThumbnailUrl').text
56 'thumbnail': thumbnail,
61 class ArteTVPlus7IE(InfoExtractor):
62 IE_NAME = 'arte.tv:+7'
63 _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
66 def _extract_url_info(cls, url):
67 mobj = re.match(cls._VALID_URL, url)
68 lang = mobj.group('lang')
69 # This is not a real id, it can be for example AJT for the news
70 # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
71 video_id = mobj.group('id')
74 def _real_extract(self, url):
75 video_id, lang = self._extract_url_info(url)
76 webpage = self._download_webpage(url, video_id)
77 return self._extract_from_webpage(webpage, video_id, lang)
79 def _extract_from_webpage(self, webpage, video_id, lang):
80 json_url = self._html_search_regex(
81 [r'arte_vp_url=["\'](.*?)["\']', r'data-url=["\']([^"]+)["\']'],
82 webpage, 'json vp url')
83 return self._extract_from_json_url(json_url, video_id, lang)
85 def _extract_from_json_url(self, json_url, video_id, lang):
86 info = self._download_json(json_url, video_id)
87 player_info = info['videoJsonPlayer']
90 'id': player_info['VID'],
91 'title': player_info['VTI'],
92 'description': player_info.get('VDE'),
93 'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]),
94 'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
97 all_formats = player_info['VSR'].values()
98 # Some formats use the m3u8 protocol
99 all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats))
101 if f.get('versionCode') is None:
103 # Return true if that format is in the language of the url
110 regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
111 return any(re.match(r, f['versionCode']) for r in regexes)
112 # Some formats may not be in the same language as the url
113 # TODO: Might want not to drop videos that does not match requested language
114 # but to process those formats with lower precedence
115 formats = filter(_match_lang, all_formats)
116 formats = list(formats) # in python3 filter returns an iterator
118 # Some videos are only available in the 'Originalversion'
119 # they aren't tagged as being in French or German
120 # Sometimes there are neither videos of requested lang code
121 # nor original version videos available
122 # For such cases we just take all_formats as is
123 formats = all_formats
125 raise ExtractorError('The formats list is empty')
127 if re.match(r'[A-Z]Q', formats[0]['quality']) is not None:
129 return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality'])
132 versionCode = f.get('versionCode')
133 if versionCode is None:
136 # Sort first by quality
137 int(f.get('height', -1)),
138 int(f.get('bitrate', -1)),
139 # The original version with subtitles has lower relevance
140 re.match(r'VO-ST(F|A)', versionCode) is None,
141 # The version with sourds/mal subtitles has also lower relevance
142 re.match(r'VO?(F|A)-STM\1', versionCode) is None,
143 # Prefer http downloads over m3u8
144 0 if f['url'].endswith('m3u8') else 1,
146 formats = sorted(formats, key=sort_key)
147 def _format(format_info):
149 height = format_info.get('height')
150 if height is not None:
151 quality = compat_str(height)
152 bitrate = format_info.get('bitrate')
153 if bitrate is not None:
154 quality += '-%d' % bitrate
155 if format_info.get('versionCode') is not None:
156 format_id = '%s-%s' % (quality, format_info['versionCode'])
160 'format_id': format_id,
161 'format_note': format_info.get('versionLibelle'),
162 'width': format_info.get('width'),
165 if format_info['mediaType'] == 'rtmp':
166 info['url'] = format_info['streamer']
167 info['play_path'] = 'mp4:' + format_info['url']
170 info['url'] = format_info['url']
171 info['ext'] = determine_ext(info['url'])
173 info_dict['formats'] = [_format(f) for f in formats]
178 # It also uses the arte_vp_url url from the webpage to extract the information
179 class ArteTVCreativeIE(ArteTVPlus7IE):
180 IE_NAME = 'arte.tv:creative'
181 _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/(?:magazine?/)?(?P<id>[^?#]+)'
184 'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
188 'title': 'Folge 2 - Corporate Design',
189 'upload_date': '20131004',
192 'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
196 'title': 'Monty Python live (mostly)',
197 'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
198 'upload_date': '20140805',
203 class ArteTVFutureIE(ArteTVPlus7IE):
204 IE_NAME = 'arte.tv:future'
205 _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'
208 'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
212 'title': 'Les champignons au secours de la planète',
213 'upload_date': '20131101',
217 def _real_extract(self, url):
218 anchor_id, lang = self._extract_url_info(url)
219 webpage = self._download_webpage(url, anchor_id)
220 row = get_element_by_id(anchor_id, webpage)
221 return self._extract_from_webpage(row, anchor_id, lang)
224 class ArteTVDDCIE(ArteTVPlus7IE):
225 IE_NAME = 'arte.tv:ddc'
226 _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'
228 def _real_extract(self, url):
229 video_id, lang = self._extract_url_info(url)
232 elif lang == 'emission':
234 webpage = self._download_webpage(url, video_id)
235 scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
236 script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
237 javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
238 json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
239 return self._extract_from_json_url(json_url, video_id, lang)
242 class ArteTVConcertIE(ArteTVPlus7IE):
243 IE_NAME = 'arte.tv:concert'
244 _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'
247 'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
248 'md5': '9ea035b7bd69696b67aa2ccaaa218161',
252 'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
253 'upload_date': '20140128',
254 'description': 'md5:486eb08f991552ade77439fe6d82c305',
259 class ArteTVEmbedIE(ArteTVPlus7IE):
260 IE_NAME = 'arte.tv:embed'
261 _VALID_URL = r'''(?x)
263 /playerv2/embed\.php\?json_url=
265 http://arte\.tv/papi/tvguide/videos/stream/player/
266 (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
270 def _real_extract(self, url):
271 mobj = re.match(self._VALID_URL, url)
272 video_id = mobj.group('id')
273 lang = mobj.group('lang')
274 json_url = mobj.group('json_url')
275 return self._extract_from_json_url(json_url, video_id, lang)