]>
Commit | Line | Data |
---|---|---|
69a0c470 | 1 | # encoding: utf-8 |
3798eadc PH |
2 | from __future__ import unicode_literals |
3 | ||
d5822b96 | 4 | import re |
75c94812 | 5 | import json |
d5822b96 PH |
6 | |
7 | from .common import InfoExtractor | |
8 | from ..utils import ( | |
d5822b96 | 9 | ExtractorError, |
df50a412 | 10 | find_xpath_attr, |
d5822b96 | 11 | unified_strdate, |
c40f5cf4 | 12 | determine_ext, |
69a0c470 | 13 | get_element_by_id, |
566d4e04 | 14 | compat_str, |
56a8ab7d | 15 | get_element_by_attribute, |
d5822b96 PH |
16 | ) |
17 | ||
c40f5cf4 JMF |
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 | |
20 | # add tests. | |
21 | ||
d5822b96 | 22 | |
515bbe4b | 23 | class ArteTvIE(InfoExtractor): |
878d11ec | 24 | _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html' |
3798eadc | 25 | IE_NAME = 'arte.tv' |
d5822b96 | 26 | |
d5822b96 | 27 | def _real_extract(self, url): |
8de64cac PH |
28 | ref_xml_url = url.replace('/videos/', '/do_delegate/videos/') |
29 | ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml') | |
a4ff6c47 PH |
30 | ref_xml_doc = self._download_xml( |
31 | ref_xml_url, video_id, note='Downloading metadata') | |
df50a412 | 32 | config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang) |
8de64cac | 33 | config_xml_url = config_node.attrib['ref'] |
878d11ec | 34 | config = self._download_xml( |
a4ff6c47 | 35 | config_xml_url, video_id, note='Downloading configuration') |
37b6a661 | 36 | |
878d11ec PH |
37 | formats = [{ |
38 | 'forma_id': q.attrib['quality'], | |
39 | 'url': q.text, | |
40 | 'quality': 2 if q.attrib['quality'] == 'hd' else 1, | |
41 | } for q in config.findall('.//quality')] | |
42 | self._sort_formats(formats) | |
43 | ||
44 | title = config.find('.//name').text | |
45 | thumbnail = config.find('.//firstThumbnailUrl').text | |
46 | return { | |
47 | 'id': video_id, | |
48 | 'title': title, | |
49 | 'thumbnail': thumbnail, | |
50 | 'url': video_url, | |
51 | 'ext': 'flv', | |
515bbe4b | 52 | } |
c40f5cf4 JMF |
53 | |
54 | ||
55 | class ArteTVPlus7IE(InfoExtractor): | |
3798eadc | 56 | IE_NAME = 'arte.tv:+7' |
f66ede43 | 57 | _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?' |
c40f5cf4 | 58 | |
69a0c470 JMF |
59 | @classmethod |
60 | def _extract_url_info(cls, url): | |
61 | mobj = re.match(cls._VALID_URL, url) | |
c40f5cf4 JMF |
62 | lang = mobj.group('lang') |
63 | # This is not a real id, it can be for example AJT for the news | |
64 | # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal | |
65 | video_id = mobj.group('id') | |
69a0c470 | 66 | return video_id, lang |
c40f5cf4 | 67 | |
69a0c470 JMF |
68 | def _real_extract(self, url): |
69 | video_id, lang = self._extract_url_info(url) | |
c40f5cf4 | 70 | webpage = self._download_webpage(url, video_id) |
69a0c470 JMF |
71 | return self._extract_from_webpage(webpage, video_id, lang) |
72 | ||
73 | def _extract_from_webpage(self, webpage, video_id, lang): | |
c40f5cf4 | 74 | json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url') |
56a8ab7d | 75 | return self._extract_from_json_url(json_url, video_id, lang) |
c40f5cf4 | 76 | |
56a8ab7d | 77 | def _extract_from_json_url(self, json_url, video_id, lang): |
c40f5cf4 JMF |
78 | json_info = self._download_webpage(json_url, video_id, 'Downloading info json') |
79 | self.report_extraction(video_id) | |
80 | info = json.loads(json_info) | |
81 | player_info = info['videoJsonPlayer'] | |
82 | ||
83 | info_dict = { | |
84 | 'id': player_info['VID'], | |
85 | 'title': player_info['VTI'], | |
86 | 'description': player_info.get('VDE'), | |
87 | 'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]), | |
88 | 'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'), | |
89 | } | |
90 | ||
21c924f4 JMF |
91 | all_formats = player_info['VSR'].values() |
92 | # Some formats use the m3u8 protocol | |
93 | all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats)) | |
c40f5cf4 JMF |
94 | def _match_lang(f): |
95 | if f.get('versionCode') is None: | |
96 | return True | |
97 | # Return true if that format is in the language of the url | |
98 | if lang == 'fr': | |
99 | l = 'F' | |
100 | elif lang == 'de': | |
101 | l = 'A' | |
102 | regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l] | |
103 | return any(re.match(r, f['versionCode']) for r in regexes) | |
104 | # Some formats may not be in the same language as the url | |
21c924f4 | 105 | formats = filter(_match_lang, all_formats) |
182a1078 | 106 | formats = list(formats) # in python3 filter returns an iterator |
21c924f4 JMF |
107 | if not formats: |
108 | # Some videos are only available in the 'Originalversion' | |
109 | # they aren't tagged as being in French or German | |
110 | if all(f['versionCode'] == 'VO' for f in all_formats): | |
111 | formats = all_formats | |
112 | else: | |
113 | raise ExtractorError(u'The formats list is empty') | |
f470c6c8 | 114 | |
182a1078 | 115 | if re.match(r'[A-Z]Q', formats[0]['quality']) is not None: |
f470c6c8 JMF |
116 | def sort_key(f): |
117 | return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality']) | |
182a1078 | 118 | else: |
f470c6c8 JMF |
119 | def sort_key(f): |
120 | return ( | |
121 | # Sort first by quality | |
122 | int(f.get('height',-1)), | |
123 | int(f.get('bitrate',-1)), | |
124 | # The original version with subtitles has lower relevance | |
125 | re.match(r'VO-ST(F|A)', f.get('versionCode', '')) is None, | |
126 | # The version with sourds/mal subtitles has also lower relevance | |
127 | re.match(r'VO?(F|A)-STM\1', f.get('versionCode', '')) is None, | |
4966a0b2 JMF |
128 | # Prefer http downloads over m3u8 |
129 | 0 if f['url'].endswith('m3u8') else 1, | |
f470c6c8 | 130 | ) |
182a1078 | 131 | formats = sorted(formats, key=sort_key) |
c40f5cf4 | 132 | def _format(format_info): |
566d4e04 JMF |
133 | quality = '' |
134 | height = format_info.get('height') | |
135 | if height is not None: | |
136 | quality = compat_str(height) | |
137 | bitrate = format_info.get('bitrate') | |
138 | if bitrate is not None: | |
139 | quality += '-%d' % bitrate | |
182a1078 | 140 | if format_info.get('versionCode') is not None: |
3798eadc | 141 | format_id = '%s-%s' % (quality, format_info['versionCode']) |
182a1078 JMF |
142 | else: |
143 | format_id = quality | |
c40f5cf4 | 144 | info = { |
182a1078 JMF |
145 | 'format_id': format_id, |
146 | 'format_note': format_info.get('versionLibelle'), | |
c40f5cf4 | 147 | 'width': format_info.get('width'), |
566d4e04 | 148 | 'height': height, |
c40f5cf4 | 149 | } |
3798eadc | 150 | if format_info['mediaType'] == 'rtmp': |
c40f5cf4 JMF |
151 | info['url'] = format_info['streamer'] |
152 | info['play_path'] = 'mp4:' + format_info['url'] | |
153 | info['ext'] = 'flv' | |
154 | else: | |
155 | info['url'] = format_info['url'] | |
156 | info['ext'] = determine_ext(info['url']) | |
157 | return info | |
158 | info_dict['formats'] = [_format(f) for f in formats] | |
c40f5cf4 JMF |
159 | |
160 | return info_dict | |
161 | ||
162 | ||
163 | # It also uses the arte_vp_url url from the webpage to extract the information | |
164 | class ArteTVCreativeIE(ArteTVPlus7IE): | |
3798eadc | 165 | IE_NAME = 'arte.tv:creative' |
c40f5cf4 JMF |
166 | _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/magazine?/(?P<id>.+)' |
167 | ||
168 | _TEST = { | |
3798eadc | 169 | 'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design', |
3798eadc | 170 | 'info_dict': { |
39a743fb JMF |
171 | 'id': '050489-002', |
172 | 'ext': 'mp4', | |
3798eadc | 173 | 'title': 'Agentur Amateur / Agence Amateur #2 : Corporate Design', |
c40f5cf4 JMF |
174 | }, |
175 | } | |
176 | ||
69a0c470 JMF |
177 | |
178 | class ArteTVFutureIE(ArteTVPlus7IE): | |
3798eadc | 179 | IE_NAME = 'arte.tv:future' |
69a0c470 JMF |
180 | _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)' |
181 | ||
182 | _TEST = { | |
3798eadc | 183 | 'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081', |
3798eadc | 184 | 'info_dict': { |
39a743fb JMF |
185 | 'id': '050940-003', |
186 | 'ext': 'mp4', | |
3798eadc | 187 | 'title': 'Les champignons au secours de la planète', |
69a0c470 JMF |
188 | }, |
189 | } | |
190 | ||
191 | def _real_extract(self, url): | |
192 | anchor_id, lang = self._extract_url_info(url) | |
193 | webpage = self._download_webpage(url, anchor_id) | |
194 | row = get_element_by_id(anchor_id, webpage) | |
195 | return self._extract_from_webpage(row, anchor_id, lang) | |
56a8ab7d | 196 | |
ac5118bc | 197 | |
56a8ab7d | 198 | class ArteTVDDCIE(ArteTVPlus7IE): |
3798eadc | 199 | IE_NAME = 'arte.tv:ddc' |
39a743fb | 200 | _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)' |
56a8ab7d | 201 | |
56a8ab7d CD |
202 | def _real_extract(self, url): |
203 | video_id, lang = self._extract_url_info(url) | |
204 | if lang == 'folge': | |
205 | lang = 'de' | |
206 | elif lang == 'emission': | |
207 | lang = 'fr' | |
208 | webpage = self._download_webpage(url, video_id) | |
209 | scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage) | |
210 | script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url') | |
211 | javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator') | |
212 | json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url') | |
213 | return self._extract_from_json_url(json_url, video_id, lang) | |
4966a0b2 JMF |
214 | |
215 | ||
216 | class ArteTVConcertIE(ArteTVPlus7IE): | |
217 | IE_NAME = 'arte.tv:concert' | |
218 | _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)' | |
219 | ||
220 | _TEST = { | |
221 | 'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde', | |
222 | 'md5': '9ea035b7bd69696b67aa2ccaaa218161', | |
223 | 'info_dict': { | |
224 | 'id': '186', | |
225 | 'ext': 'mp4', | |
226 | 'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"', | |
227 | 'upload_date': '20140128', | |
a9c2896e | 228 | 'description': 'md5:486eb08f991552ade77439fe6d82c305', |
4966a0b2 JMF |
229 | }, |
230 | } |