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