Merge branch 'master' of github.com-rndusr:rg3/youtube-dl into fix/str-item-assignment
[youtube-dl] / youtube_dl / extractor / condenast.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5
6 from .common import InfoExtractor
7 from ..compat import (
8     compat_urllib_parse_urlparse,
9     compat_urlparse,
10 )
11 from ..utils import (
12     determine_ext,
13     extract_attributes,
14     int_or_none,
15     js_to_json,
16     mimetype2ext,
17     orderedSet,
18     parse_iso8601,
19     remove_end,
20 )
21
22
23 class CondeNastIE(InfoExtractor):
24     """
25     Condé Nast is a media group, some of its sites use a custom HTML5 player
26     that works the same in all of them.
27     """
28
29     # The keys are the supported sites and the values are the name to be shown
30     # to the user and in the extractor description.
31     _SITES = {
32         'allure': 'Allure',
33         'architecturaldigest': 'Architectural Digest',
34         'arstechnica': 'Ars Technica',
35         'bonappetit': 'Bon Appétit',
36         'brides': 'Brides',
37         'cnevids': 'Condé Nast',
38         'cntraveler': 'Condé Nast Traveler',
39         'details': 'Details',
40         'epicurious': 'Epicurious',
41         'glamour': 'Glamour',
42         'golfdigest': 'Golf Digest',
43         'gq': 'GQ',
44         'newyorker': 'The New Yorker',
45         'self': 'SELF',
46         'teenvogue': 'Teen Vogue',
47         'vanityfair': 'Vanity Fair',
48         'vogue': 'Vogue',
49         'wired': 'WIRED',
50         'wmagazine': 'W Magazine',
51     }
52
53     _VALID_URL = r'https?://(?:video|www|player)\.(?P<site>%s)\.com/(?P<type>watch|series|video|embed(?:js)?)/(?P<id>[^/?#]+)' % '|'.join(_SITES.keys())
54     IE_DESC = 'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values()))
55
56     EMBED_URL = r'(?:https?:)?//player\.(?P<site>%s)\.com/(?P<type>embed(?:js)?)/.+?' % '|'.join(_SITES.keys())
57
58     _TESTS = [{
59         'url': 'http://video.wired.com/watch/3d-printed-speakers-lit-with-led',
60         'md5': '1921f713ed48aabd715691f774c451f7',
61         'info_dict': {
62             'id': '5171b343c2b4c00dd0c1ccb3',
63             'ext': 'mp4',
64             'title': '3D Printed Speakers Lit With LED',
65             'description': 'Check out these beautiful 3D printed LED speakers.  You can\'t actually buy them, but LumiGeek is working on a board that will let you make you\'re own.',
66             'uploader': 'wired',
67             'upload_date': '20130314',
68             'timestamp': 1363219200,
69         }
70     }, {
71         'url': 'http://video.gq.com/watch/the-closer-with-keith-olbermann-the-only-true-surprise-trump-s-an-idiot?c=series',
72         'info_dict': {
73             'id': '58d1865bfd2e6126e2000015',
74             'ext': 'mp4',
75             'title': 'The Only True Surprise? Trump’s an Idiot',
76             'uploader': 'gq',
77             'upload_date': '20170321',
78             'timestamp': 1490126427,
79         },
80     }, {
81         # JS embed
82         'url': 'http://player.cnevids.com/embedjs/55f9cf8b61646d1acf00000c/5511d76261646d5566020000.js',
83         'md5': 'f1a6f9cafb7083bab74a710f65d08999',
84         'info_dict': {
85             'id': '55f9cf8b61646d1acf00000c',
86             'ext': 'mp4',
87             'title': '3D printed TSA Travel Sentry keys really do open TSA locks',
88             'uploader': 'arstechnica',
89             'upload_date': '20150916',
90             'timestamp': 1442434955,
91         }
92     }]
93
94     def _extract_series(self, url, webpage):
95         title = self._html_search_regex(
96             r'(?s)<div class="cne-series-info">.*?<h1>(.+?)</h1>',
97             webpage, 'series title')
98         url_object = compat_urllib_parse_urlparse(url)
99         base_url = '%s://%s' % (url_object.scheme, url_object.netloc)
100         m_paths = re.finditer(
101             r'(?s)<p class="cne-thumb-title">.*?<a href="(/watch/.+?)["\?]', webpage)
102         paths = orderedSet(m.group(1) for m in m_paths)
103         build_url = lambda path: compat_urlparse.urljoin(base_url, path)
104         entries = [self.url_result(build_url(path), 'CondeNast') for path in paths]
105         return self.playlist_result(entries, playlist_title=title)
106
107     def _extract_video(self, webpage, url_type):
108         query = {}
109         params = self._search_regex(
110             r'(?s)var params = {(.+?)}[;,]', webpage, 'player params', default=None)
111         if params:
112             query.update({
113                 'videoId': self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, 'video id'),
114                 'playerId': self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, 'player id'),
115                 'target': self._search_regex(r'target: [\'"](.+?)[\'"]', params, 'target'),
116             })
117         else:
118             params = extract_attributes(self._search_regex(
119                 r'(<[^>]+data-js="video-player"[^>]+>)',
120                 webpage, 'player params element'))
121             query.update({
122                 'videoId': params['data-video'],
123                 'playerId': params['data-player'],
124                 'target': params['id'],
125             })
126         video_id = query['videoId']
127         video_info = None
128         info_page = self._download_json(
129             'http://player.cnevids.com/player/video.js',
130             video_id, 'Downloading video info', fatal=False, query=query)
131         if info_page:
132             video_info = info_page.get('video')
133         if not video_info:
134             info_page = self._download_webpage(
135                 'http://player.cnevids.com/player/loader.js',
136                 video_id, 'Downloading loader info', query=query)
137             video_info = self._parse_json(
138                 self._search_regex(
139                     r'(?s)var\s+config\s*=\s*({.+?});', info_page, 'config'),
140                 video_id, transform_source=js_to_json)['video']
141
142         title = video_info['title']
143
144         formats = []
145         for fdata in video_info['sources']:
146             src = fdata.get('src')
147             if not src:
148                 continue
149             ext = mimetype2ext(fdata.get('type')) or determine_ext(src)
150             if ext == 'm3u8':
151                 formats.extend(self._extract_m3u8_formats(
152                     src, video_id, 'mp4', entry_protocol='m3u8_native',
153                     m3u8_id='hls', fatal=False))
154                 continue
155             quality = fdata.get('quality')
156             formats.append({
157                 'format_id': ext + ('-%s' % quality if quality else ''),
158                 'url': src,
159                 'ext': ext,
160                 'quality': 1 if quality == 'high' else 0,
161             })
162         self._sort_formats(formats)
163
164         info = self._search_json_ld(
165             webpage, video_id, fatal=False) if url_type != 'embed' else {}
166         info.update({
167             'id': video_id,
168             'formats': formats,
169             'title': title,
170             'thumbnail': video_info.get('poster_frame'),
171             'uploader': video_info.get('brand'),
172             'duration': int_or_none(video_info.get('duration')),
173             'tags': video_info.get('tags'),
174             'series': video_info.get('series_title'),
175             'season': video_info.get('season_title'),
176             'timestamp': parse_iso8601(video_info.get('premiere_date')),
177         })
178         return info
179
180     def _real_extract(self, url):
181         site, url_type, item_id = re.match(self._VALID_URL, url).groups()
182
183         # Convert JS embed to regular embed
184         if url_type == 'embedjs':
185             parsed_url = compat_urlparse.urlparse(url)
186             url = compat_urlparse.urlunparse(parsed_url._replace(
187                 path=remove_end(parsed_url.path, '.js').replace('/embedjs/', '/embed/')))
188             url_type = 'embed'
189
190         webpage = self._download_webpage(url, item_id)
191
192         if url_type == 'series':
193             return self._extract_series(url, webpage)
194         else:
195             return self._extract_video(webpage, url_type)