]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/kaltura.py
debian/watch: Skip pgpsigurlmangle for the moment.
[youtubedl] / youtube_dl / extractor / kaltura.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import re
5 import base64
6
7 from .common import InfoExtractor
8 from ..compat import (
9 compat_urllib_parse,
10 compat_urlparse,
11 )
12 from ..utils import (
13 clean_html,
14 ExtractorError,
15 int_or_none,
16 unsmuggle_url,
17 )
18
19
20 class KalturaIE(InfoExtractor):
21 _VALID_URL = r'''(?x)
22 (?:
23 kaltura:(?P<partner_id_s>\d+):(?P<id_s>[0-9a-z_]+)|
24 https?://
25 (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/
26 (?:
27 (?:
28 # flash player
29 index\.php/kwidget/
30 (?:[^/]+/)*?wid/_(?P<partner_id>\d+)/
31 (?:[^/]+/)*?entry_id/(?P<id>[0-9a-z_]+)|
32 # html5 player
33 html5/html5lib/
34 (?:[^/]+/)*?entry_id/(?P<id_html5>[0-9a-z_]+)
35 .*\?.*\bwid=_(?P<partner_id_html5>\d+)
36 )
37 )
38 )
39 '''
40 _API_BASE = 'http://cdnapi.kaltura.com/api_v3/index.php?'
41 _TESTS = [
42 {
43 'url': 'kaltura:269692:1_1jc2y3e4',
44 'md5': '3adcbdb3dcc02d647539e53f284ba171',
45 'info_dict': {
46 'id': '1_1jc2y3e4',
47 'ext': 'mp4',
48 'title': 'Straight from the Heart',
49 'upload_date': '20131219',
50 'uploader_id': 'mlundberg@wolfgangsvault.com',
51 'description': 'The Allman Brothers Band, 12/16/1981',
52 'thumbnail': 're:^https?://.*/thumbnail/.*',
53 'timestamp': int,
54 },
55 },
56 {
57 'url': 'http://www.kaltura.com/index.php/kwidget/cache_st/1300318621/wid/_269692/uiconf_id/3873291/entry_id/1_1jc2y3e4',
58 'only_matching': True,
59 },
60 {
61 'url': 'https://cdnapisec.kaltura.com/index.php/kwidget/wid/_557781/uiconf_id/22845202/entry_id/1_plr1syf3',
62 'only_matching': True,
63 },
64 {
65 'url': 'https://cdnapisec.kaltura.com/html5/html5lib/v2.30.2/mwEmbedFrame.php/p/1337/uiconf_id/20540612/entry_id/1_sf5ovm7u?wid=_243342',
66 'only_matching': True,
67 }
68 ]
69
70 def _kaltura_api_call(self, video_id, actions, *args, **kwargs):
71 params = actions[0]
72 if len(actions) > 1:
73 for i, a in enumerate(actions[1:], start=1):
74 for k, v in a.items():
75 params['%d:%s' % (i, k)] = v
76
77 query = compat_urllib_parse.urlencode(params)
78 url = self._API_BASE + query
79 data = self._download_json(url, video_id, *args, **kwargs)
80
81 status = data if len(actions) == 1 else data[0]
82 if status.get('objectType') == 'KalturaAPIException':
83 raise ExtractorError(
84 '%s said: %s' % (self.IE_NAME, status['message']))
85
86 return data
87
88 def _get_kaltura_signature(self, video_id, partner_id):
89 actions = [{
90 'apiVersion': '3.1',
91 'expiry': 86400,
92 'format': 1,
93 'service': 'session',
94 'action': 'startWidgetSession',
95 'widgetId': '_%s' % partner_id,
96 }]
97 return self._kaltura_api_call(
98 video_id, actions, note='Downloading Kaltura signature')['ks']
99
100 def _get_video_info(self, video_id, partner_id):
101 signature = self._get_kaltura_signature(video_id, partner_id)
102 actions = [
103 {
104 'action': 'null',
105 'apiVersion': '3.1.5',
106 'clientTag': 'kdp:v3.8.5',
107 'format': 1, # JSON, 2 = XML, 3 = PHP
108 'service': 'multirequest',
109 'ks': signature,
110 },
111 {
112 'action': 'get',
113 'entryId': video_id,
114 'service': 'baseentry',
115 'version': '-1',
116 },
117 {
118 'action': 'getbyentryid',
119 'entryId': video_id,
120 'service': 'flavorAsset',
121 },
122 ]
123 return self._kaltura_api_call(
124 video_id, actions, note='Downloading video info JSON')
125
126 def _real_extract(self, url):
127 url, smuggled_data = unsmuggle_url(url, {})
128
129 mobj = re.match(self._VALID_URL, url)
130 partner_id = mobj.group('partner_id_s') or mobj.group('partner_id') or mobj.group('partner_id_html5')
131 entry_id = mobj.group('id_s') or mobj.group('id') or mobj.group('id_html5')
132
133 info, flavor_assets = self._get_video_info(entry_id, partner_id)
134
135 source_url = smuggled_data.get('source_url')
136 if source_url:
137 referrer = base64.b64encode(
138 '://'.join(compat_urlparse.urlparse(source_url)[:2])
139 .encode('utf-8')).decode('utf-8')
140 else:
141 referrer = None
142
143 formats = []
144 for f in flavor_assets:
145 # Continue if asset is not ready
146 if f['status'] != 2:
147 continue
148 video_url = '%s/flavorId/%s' % (info['dataUrl'], f['id'])
149 if referrer:
150 video_url += '?referrer=%s' % referrer
151 formats.append({
152 'format_id': '%(fileExt)s-%(bitrate)s' % f,
153 'ext': f.get('fileExt'),
154 'tbr': int_or_none(f['bitrate']),
155 'fps': int_or_none(f.get('frameRate')),
156 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
157 'container': f.get('containerFormat'),
158 'vcodec': f.get('videoCodecId'),
159 'height': int_or_none(f.get('height')),
160 'width': int_or_none(f.get('width')),
161 'url': video_url,
162 })
163 m3u8_url = info['dataUrl'].replace('format/url', 'format/applehttp')
164 if referrer:
165 m3u8_url += '?referrer=%s' % referrer
166 formats.extend(self._extract_m3u8_formats(
167 m3u8_url, entry_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False))
168
169 self._check_formats(formats, entry_id)
170 self._sort_formats(formats)
171
172 return {
173 'id': entry_id,
174 'title': info['name'],
175 'formats': formats,
176 'description': clean_html(info.get('description')),
177 'thumbnail': info.get('thumbnailUrl'),
178 'duration': info.get('duration'),
179 'timestamp': info.get('createdAt'),
180 'uploader_id': info.get('userId'),
181 'view_count': info.get('plays'),
182 }