e369959e3ca0702721c80024c04f7df990efdbf4
[ytdl] / 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_urlparse,
10     compat_parse_qs,
11 )
12 from ..utils import (
13     clean_html,
14     ExtractorError,
15     int_or_none,
16     unsmuggle_url,
17     smuggle_url,
18 )
19
20
21 class KalturaIE(InfoExtractor):
22     _VALID_URL = r'''(?x)
23                 (?:
24                     kaltura:(?P<partner_id>\d+):(?P<id>[0-9a-z_]+)|
25                     https?://
26                         (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com(?::\d+)?/
27                         (?:
28                             (?:
29                                 # flash player
30                                 index\.php/(?:kwidget|extwidget/preview)|
31                                 # html5 player
32                                 html5/html5lib/[^/]+/mwEmbedFrame\.php
33                             )
34                         )(?:/(?P<path>[^?]+))?(?:\?(?P<query>.*))?
35                 )
36                 '''
37     _SERVICE_URL = 'http://cdnapi.kaltura.com'
38     _SERVICE_BASE = '/api_v3/index.php'
39     # See https://github.com/kaltura/server/blob/master/plugins/content/caption/base/lib/model/enums/CaptionType.php
40     _CAPTION_TYPES = {
41         1: 'srt',
42         2: 'ttml',
43         3: 'vtt',
44     }
45     _TESTS = [
46         {
47             'url': 'kaltura:269692:1_1jc2y3e4',
48             'md5': '3adcbdb3dcc02d647539e53f284ba171',
49             'info_dict': {
50                 'id': '1_1jc2y3e4',
51                 'ext': 'mp4',
52                 'title': 'Straight from the Heart',
53                 'upload_date': '20131219',
54                 'uploader_id': 'mlundberg@wolfgangsvault.com',
55                 'description': 'The Allman Brothers Band, 12/16/1981',
56                 'thumbnail': 're:^https?://.*/thumbnail/.*',
57                 'timestamp': int,
58             },
59         },
60         {
61             'url': 'http://www.kaltura.com/index.php/kwidget/cache_st/1300318621/wid/_269692/uiconf_id/3873291/entry_id/1_1jc2y3e4',
62             'only_matching': True,
63         },
64         {
65             'url': 'https://cdnapisec.kaltura.com/index.php/kwidget/wid/_557781/uiconf_id/22845202/entry_id/1_plr1syf3',
66             'only_matching': True,
67         },
68         {
69             'url': 'https://cdnapisec.kaltura.com/html5/html5lib/v2.30.2/mwEmbedFrame.php/p/1337/uiconf_id/20540612/entry_id/1_sf5ovm7u?wid=_243342',
70             'only_matching': True,
71         },
72         {
73             # video with subtitles
74             'url': 'kaltura:111032:1_cw786r8q',
75             'only_matching': True,
76         },
77         {
78             # video with ttml subtitles (no fileExt)
79             'url': 'kaltura:1926081:0_l5ye1133',
80             'info_dict': {
81                 'id': '0_l5ye1133',
82                 'ext': 'mp4',
83                 'title': 'What Can You Do With Python?',
84                 'upload_date': '20160221',
85                 'uploader_id': 'stork',
86                 'thumbnail': 're:^https?://.*/thumbnail/.*',
87                 'timestamp': int,
88                 'subtitles': {
89                     'en': [{
90                         'ext': 'ttml',
91                     }],
92                 },
93             },
94             'skip': 'Gone. Maybe https://www.safaribooksonline.com/library/tutorials/introduction-to-python-anon/3469/',
95             'params': {
96                 'skip_download': True,
97             },
98         },
99         {
100             'url': 'https://www.kaltura.com/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto',
101             'only_matching': True,
102         },
103         {
104             'url': 'https://www.kaltura.com:443/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto',
105             'only_matching': True,
106         }
107     ]
108
109     @staticmethod
110     def _extract_url(webpage):
111         # Embed codes: https://knowledge.kaltura.com/embedding-kaltura-media-players-your-site
112         mobj = (
113             re.search(
114                 r"""(?xs)
115                     kWidget\.(?:thumb)?[Ee]mbed\(
116                     \{.*?
117                         (?P<q1>['"])wid(?P=q1)\s*:\s*
118                         (?P<q2>['"])_?(?P<partner_id>(?:(?!(?P=q2)).)+)(?P=q2),.*?
119                         (?P<q3>['"])entry_?[Ii]d(?P=q3)\s*:\s*
120                         (?P<q4>['"])(?P<id>(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\})
121                 """, webpage) or
122             re.search(
123                 r'''(?xs)
124                     (?P<q1>["'])
125                         (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com(?::\d+)?/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)(?:(?!(?P=q1)).)*
126                     (?P=q1).*?
127                     (?:
128                         (?:
129                             entry_?[Ii]d|
130                             (?P<q2>["'])entry_?[Ii]d(?P=q2)
131                         )\s*:\s*|
132                         \[\s*(?P<q2_1>["'])entry_?[Ii]d(?P=q2_1)\s*\]?\s*=\s*
133                     )
134                     (?P<q3>["'])(?P<id>(?:(?!(?P=q3)).)+)(?P=q3)
135                 ''', webpage) or
136             re.search(
137                 r'''(?xs)
138                     <iframe[^>]+src=(?P<q1>["'])
139                       (?:https?:)?//(?:www\.)?kaltura\.com/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)
140                       (?:(?!(?P=q1)).)*
141                       [?&]entry_id=(?P<id>(?:(?!(?P=q1))[^&])+)
142                     (?P=q1)
143                 ''', webpage)
144         )
145         if mobj:
146             embed_info = mobj.groupdict()
147             url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
148             escaped_pid = re.escape(embed_info['partner_id'])
149             service_url = re.search(
150                 r'<script[^>]+src=["\']((?:https?:)?//.+?)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
151                 webpage)
152             if service_url:
153                 url = smuggle_url(url, {'service_url': service_url.group(1)})
154             return url
155
156     def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
157         params = actions[0]
158         if len(actions) > 1:
159             for i, a in enumerate(actions[1:], start=1):
160                 for k, v in a.items():
161                     params['%d:%s' % (i, k)] = v
162
163         data = self._download_json(
164             (service_url or self._SERVICE_URL) + self._SERVICE_BASE,
165             video_id, query=params, *args, **kwargs)
166
167         status = data if len(actions) == 1 else data[0]
168         if status.get('objectType') == 'KalturaAPIException':
169             raise ExtractorError(
170                 '%s said: %s' % (self.IE_NAME, status['message']))
171
172         return data
173
174     def _get_video_info(self, video_id, partner_id, service_url=None):
175         actions = [
176             {
177                 'action': 'null',
178                 'apiVersion': '3.1.5',
179                 'clientTag': 'kdp:v3.8.5',
180                 'format': 1,  # JSON, 2 = XML, 3 = PHP
181                 'service': 'multirequest',
182             },
183             {
184                 'expiry': 86400,
185                 'service': 'session',
186                 'action': 'startWidgetSession',
187                 'widgetId': '_%s' % partner_id,
188             },
189             {
190                 'action': 'get',
191                 'entryId': video_id,
192                 'service': 'baseentry',
193                 'ks': '{1:result:ks}',
194             },
195             {
196                 'action': 'getbyentryid',
197                 'entryId': video_id,
198                 'service': 'flavorAsset',
199                 'ks': '{1:result:ks}',
200             },
201             {
202                 'action': 'list',
203                 'filter:entryIdEqual': video_id,
204                 'service': 'caption_captionasset',
205                 'ks': '{1:result:ks}',
206             },
207         ]
208         return self._kaltura_api_call(
209             video_id, actions, service_url, note='Downloading video info JSON')
210
211     def _real_extract(self, url):
212         url, smuggled_data = unsmuggle_url(url, {})
213
214         mobj = re.match(self._VALID_URL, url)
215         partner_id, entry_id = mobj.group('partner_id', 'id')
216         ks = None
217         captions = None
218         if partner_id and entry_id:
219             _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
220         else:
221             path, query = mobj.group('path', 'query')
222             if not path and not query:
223                 raise ExtractorError('Invalid URL', expected=True)
224             params = {}
225             if query:
226                 params = compat_parse_qs(query)
227             if path:
228                 splitted_path = path.split('/')
229                 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
230             if 'wid' in params:
231                 partner_id = params['wid'][0][1:]
232             elif 'p' in params:
233                 partner_id = params['p'][0]
234             elif 'partner_id' in params:
235                 partner_id = params['partner_id'][0]
236             else:
237                 raise ExtractorError('Invalid URL', expected=True)
238             if 'entry_id' in params:
239                 entry_id = params['entry_id'][0]
240                 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
241             elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
242                 reference_id = params['flashvars[referenceId]'][0]
243                 webpage = self._download_webpage(url, reference_id)
244                 entry_data = self._parse_json(self._search_regex(
245                     r'window\.kalturaIframePackageData\s*=\s*({.*});',
246                     webpage, 'kalturaIframePackageData'),
247                     reference_id)['entryResult']
248                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
249                 entry_id = info['id']
250                 # Unfortunately, data returned in kalturaIframePackageData lacks
251                 # captions so we will try requesting the complete data using
252                 # regular approach since we now know the entry_id
253                 try:
254                     _, info, flavor_assets, captions = self._get_video_info(
255                         entry_id, partner_id)
256                 except ExtractorError:
257                     # Regular scenario failed but we already have everything
258                     # extracted apart from captions and can process at least
259                     # with this
260                     pass
261             else:
262                 raise ExtractorError('Invalid URL', expected=True)
263             ks = params.get('flashvars[ks]', [None])[0]
264
265         source_url = smuggled_data.get('source_url')
266         if source_url:
267             referrer = base64.b64encode(
268                 '://'.join(compat_urlparse.urlparse(source_url)[:2])
269                 .encode('utf-8')).decode('utf-8')
270         else:
271             referrer = None
272
273         def sign_url(unsigned_url):
274             if ks:
275                 unsigned_url += '/ks/%s' % ks
276             if referrer:
277                 unsigned_url += '?referrer=%s' % referrer
278             return unsigned_url
279
280         data_url = info['dataUrl']
281         if '/flvclipper/' in data_url:
282             data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
283
284         formats = []
285         for f in flavor_assets:
286             # Continue if asset is not ready
287             if f.get('status') != 2:
288                 continue
289             # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g)
290             # skip for now.
291             if f.get('fileExt') == 'chun':
292                 continue
293             # DRM-protected video, cannot be decrypted
294             if f.get('fileExt') == 'wvm':
295                 continue
296             if not f.get('fileExt'):
297                 # QT indicates QuickTime; some videos have broken fileExt
298                 if f.get('containerFormat') == 'qt':
299                     f['fileExt'] = 'mov'
300                 else:
301                     f['fileExt'] = 'mp4'
302             video_url = sign_url(
303                 '%s/flavorId/%s' % (data_url, f['id']))
304             # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g
305             # -f mp4-56)
306             vcodec = 'none' if 'videoCodecId' not in f and f.get(
307                 'frameRate') == 0 else f.get('videoCodecId')
308             formats.append({
309                 'format_id': '%(fileExt)s-%(bitrate)s' % f,
310                 'ext': f.get('fileExt'),
311                 'tbr': int_or_none(f['bitrate']),
312                 'fps': int_or_none(f.get('frameRate')),
313                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
314                 'container': f.get('containerFormat'),
315                 'vcodec': vcodec,
316                 'height': int_or_none(f.get('height')),
317                 'width': int_or_none(f.get('width')),
318                 'url': video_url,
319             })
320         if '/playManifest/' in data_url:
321             m3u8_url = sign_url(data_url.replace(
322                 'format/url', 'format/applehttp'))
323             formats.extend(self._extract_m3u8_formats(
324                 m3u8_url, entry_id, 'mp4', 'm3u8_native',
325                 m3u8_id='hls', fatal=False))
326
327         self._sort_formats(formats)
328
329         subtitles = {}
330         if captions:
331             for caption in captions.get('objects', []):
332                 # Continue if caption is not ready
333                 if caption.get('status') != 2:
334                     continue
335                 if not caption.get('id'):
336                     continue
337                 caption_format = int_or_none(caption.get('format'))
338                 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
339                     'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
340                     'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml',
341                 })
342
343         return {
344             'id': entry_id,
345             'title': info['name'],
346             'formats': formats,
347             'subtitles': subtitles,
348             'description': clean_html(info.get('description')),
349             'thumbnail': info.get('thumbnailUrl'),
350             'duration': info.get('duration'),
351             'timestamp': info.get('createdAt'),
352             'uploader_id': info.get('userId') if info.get('userId') != 'None' else None,
353             'view_count': info.get('plays'),
354         }