]>
jfr.im git - yt-dlp.git/blob - youtube_dl/extractor/kaltura.py
2 from __future__
import unicode_literals
7 from .common
import InfoExtractor
21 class KalturaIE(InfoExtractor
):
24 kaltura:(?P<partner_id>\d+):(?P<id>[0-9a-z_]+)|
26 (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/
32 html5/html5lib/[^/]+/mwEmbedFrame\.php
34 )(?:/(?P<path>[^?]+))?(?:\?(?P<query>.*))?
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
47 'url': 'kaltura:269692:1_1jc2y3e4',
48 'md5': '3adcbdb3dcc02d647539e53f284ba171',
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/.*',
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,
65 'url': 'https://cdnapisec.kaltura.com/index.php/kwidget/wid/_557781/uiconf_id/22845202/entry_id/1_plr1syf3',
66 'only_matching': True,
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,
73 # video with subtitles
74 'url': 'kaltura:111032:1_cw786r8q',
75 'only_matching': True,
78 # video with ttml subtitles (no fileExt)
79 'url': 'kaltura:1926081:0_l5ye1133',
83 'title': 'What Can You Do With Python?',
84 'upload_date': '20160221',
85 'uploader_id': 'stork',
86 'thumbnail': 're:^https?://.*/thumbnail/.*',
95 'skip_download': True,
101 def _extract_url(webpage
):
105 kWidget\.(?:thumb)?[Ee]mbed\(
107 (?P<q1>['\"])wid(?P
=q1
)\s
*:\s
*
108 (?P
<q2
>['\"])_?(?P<partner_id>(?:(?!(?P=q2)).)+)(?P=q2),.*?
109 (?P<q3>['\"])entry_?
[Ii
]d(?P
=q3
)\s
*:\s
*
110 (?P
<q4
>['\"])(?P<id>(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\})
115 (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com/(?:(?!(?P=q1)).)*(?:p|partner_id)/(?P<partner_id>\d+)(?:(?!(?P=q1)).)*
119 (?P<q2>["\'])entry_?[Ii]d(?P=q2)
121 (?P<q3>["\'])(?P<id>(?:(?!(?P=q3)).)+)(?P=q3)
124 embed_info = mobj.groupdict()
125 url = 'kaltura
:%(partner_id)s:%(id)s' % embed_info
126 escaped_pid = re.escape(embed_info['partner_id
'])
127 service_url = re.search(
128 r'<script
[^
>]+src
=["\']((?:https?:)?//.+?)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
131 url = smuggle_url(url, {'service_url': service_url.group(1)})
134 def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
137 for i, a in enumerate(actions[1:], start=1):
138 for k, v in a.items():
139 params['%d:%s' % (i, k)] = v
141 data = self._download_json(
142 (service_url or self._SERVICE_URL) + self._SERVICE_BASE,
143 video_id, query=params, *args, **kwargs)
145 status = data if len(actions) == 1 else data[0]
146 if status.get('objectType') == 'KalturaAPIException':
147 raise ExtractorError(
148 '%s said: %s' % (self.IE_NAME, status['message']))
152 def _get_video_info(self, video_id, partner_id, service_url=None):
156 'apiVersion': '3.1.5',
157 'clientTag': 'kdp:v3.8.5',
158 'format': 1, # JSON, 2 = XML, 3 = PHP
159 'service': 'multirequest',
163 'service': 'session',
164 'action': 'startWidgetSession',
165 'widgetId': '_%s' % partner_id,
170 'service': 'baseentry',
171 'ks': '{1:result:ks}',
174 'action': 'getbyentryid',
176 'service': 'flavorAsset',
177 'ks': '{1:result:ks}',
181 'filter:entryIdEqual': video_id,
182 'service': 'caption_captionasset',
183 'ks': '{1:result:ks}',
186 return self._kaltura_api_call(
187 video_id, actions, service_url, note='Downloading video info JSON')
189 def _real_extract(self, url):
190 url, smuggled_data = unsmuggle_url(url, {})
192 mobj = re.match(self._VALID_URL, url)
193 partner_id, entry_id = mobj.group('partner_id', 'id')
196 if partner_id and entry_id:
197 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
199 path, query = mobj.group('path', 'query')
200 if not path and not query:
201 raise ExtractorError('Invalid URL', expected=True)
204 params = compat_parse_qs(query)
206 splitted_path = path.split('/')
207 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
209 partner_id = params['wid'][0][1:]
211 partner_id = params['p'][0]
213 raise ExtractorError('Invalid URL', expected=True)
214 if 'entry_id' in params:
215 entry_id = params['entry_id'][0]
216 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
217 elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
218 reference_id = params['flashvars[referenceId]'][0]
219 webpage = self._download_webpage(url, reference_id)
220 entry_data = self._parse_json(self._search_regex(
221 r'window\.kalturaIframePackageData\s*=\s*({.*});',
222 webpage, 'kalturaIframePackageData'),
223 reference_id)['entryResult']
224 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
225 entry_id = info['id']
226 # Unfortunately, data returned in kalturaIframePackageData lacks
227 # captions so we will try requesting the complete data using
228 # regular approach since we now know the entry_id
230 _, info, flavor_assets, captions = self._get_video_info(
231 entry_id, partner_id)
232 except ExtractorError:
233 # Regular scenario failed but we already have everything
234 # extracted apart from captions and can process at least
238 raise ExtractorError('Invalid URL', expected=True)
239 ks = params.get('flashvars[ks]', [None])[0]
241 source_url = smuggled_data.get('source_url')
243 referrer = base64.b64encode(
244 '://'.join(compat_urlparse.urlparse(source_url)[:2])
245 .encode('utf-8')).decode('utf-8')
249 def sign_url(unsigned_url):
251 unsigned_url += '/ks/%s' % ks
253 unsigned_url += '?referrer=%s' % referrer
256 data_url = info['dataUrl']
257 if '/flvclipper/' in data_url:
258 data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
261 for f in flavor_assets:
262 # Continue if asset is not ready
263 if f.get('status') != 2:
265 # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g)
267 if f.get('fileExt') == 'chun':
269 if not f.get('fileExt') and f.get('containerFormat') == 'qt':
270 # QT indicates QuickTime; some videos have broken fileExt
272 video_url = sign_url(
273 '%s/flavorId/%s' % (data_url, f['id']))
274 # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g
276 vcodec = 'none' if 'videoCodecId' not in f and f.get(
277 'frameRate') == 0 else f.get('videoCodecId')
279 'format_id': '%(fileExt)s-%(bitrate)s' % f,
280 'ext': f.get('fileExt'),
281 'tbr': int_or_none(f['bitrate']),
282 'fps': int_or_none(f.get('frameRate')),
283 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
284 'container': f.get('containerFormat'),
286 'height': int_or_none(f.get('height')),
287 'width': int_or_none(f.get('width')),
290 if '/playManifest/' in data_url:
291 m3u8_url = sign_url(data_url.replace(
292 'format/url', 'format/applehttp'))
293 formats.extend(self._extract_m3u8_formats(
294 m3u8_url, entry_id, 'mp4', 'm3u8_native',
295 m3u8_id='hls', fatal=False))
297 self._sort_formats(formats)
301 for caption in captions.get('objects', []):
302 # Continue if caption is not ready
303 if f.get('status') != 2:
305 if not caption.get('id'):
307 caption_format = int_or_none(caption.get('format'))
308 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
309 'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
310 'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml',
315 'title': info['name'],
317 'subtitles': subtitles,
318 'description': clean_html(info.get('description')),
319 'thumbnail': info.get('thumbnailUrl'),
320 'duration': info.get('duration'),
321 'timestamp': info.get('createdAt'),
322 'uploader_id': info.get('userId'),
323 'view_count': info.get('plays'),