_ Git - youtube-dl/blob - 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/
  27                         (?:
  28                             (?:
  29                                 # flash player
  30                                 index\.php/kwidget|
  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     _TESTS = [
  40         {
  41             'url': 'kaltura:269692:1_1jc2y3e4',
  42             'md5': '3adcbdb3dcc02d647539e53f284ba171',
  43             'info_dict': {
  44                 'id': '1_1jc2y3e4',
  45                 'ext': 'mp4',
  46                 'title': 'Straight from the Heart',
  47                 'upload_date': '20131219',
  48                 'uploader_id': 'mlundberg@wolfgangsvault.com',
  49                 'description': 'The Allman Brothers Band, 12/16/1981',
  50                 'thumbnail': 're:^https?://.*/thumbnail/.*',
  51                 'timestamp': int,
  52             },
  53         },
  54         {
  55             'url': 'http://www.kaltura.com/index.php/kwidget/cache_st/1300318621/wid/_269692/uiconf_id/3873291/entry_id/1_1jc2y3e4',
  56             'only_matching': True,
  57         },
  58         {
  59             'url': 'https://cdnapisec.kaltura.com/index.php/kwidget/wid/_557781/uiconf_id/22845202/entry_id/1_plr1syf3',
  60             'only_matching': True,
  61         },
  62         {
  63             'url': 'https://cdnapisec.kaltura.com/html5/html5lib/v2.30.2/mwEmbedFrame.php/p/1337/uiconf_id/20540612/entry_id/1_sf5ovm7u?wid=_243342',
  64             'only_matching': True,
  65         }
  66     ]
  67
  68     @staticmethod
  69     def _extract_url(webpage):
  70         mobj = (
  71             re.search(
  72                 r"""(?xs)
  73                     kWidget\.(?:thumb)?[Ee]mbed\(
  74                     \{.*?
  75                         (?P<q1>['\"])wid(?P=q1)\s*:\s*
  76                         (?P<q2>['\"])_?(?P<partner_id>[^'\"]+)(?P=q2),.*?
  77                         (?P<q3>['\"])entry_?[Ii]d(?P=q3)\s*:\s*
  78                         (?P<q4>['\"])(?P<id>[^'\"]+)(?P=q4),
  79                 """, webpage) or
  80             re.search(
  81                 r'''(?xs)
  82                     (?P<q1>["\'])
  83                         (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com/.*?(?:p|partner_id)/(?P<partner_id>\d+).*?
  84                     (?P=q1).*?
  85                     (?:
  86                         entry_?[Ii]d|
  87                         (?P<q2>["\'])entry_?[Ii]d(?P=q2)
  88                     )\s*:\s*
  89                     (?P<q3>["\'])(?P<id>.+?)(?P=q3)
  90                 ''', webpage))
  91         if mobj:
  92             embed_info = mobj.groupdict()
  93             url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
  94             service_url = re.search(
  95                 '<script[^>]+src=(?:["\'])((?:https?:)?//.+?)/p/%(partner_id)s/sp/%(partner_id)s00/embedIframeJs' % embed_info,
  96                 webpage)
  97             if service_url:
  98                 url = smuggle_url(url, {'service_url': service_url.group(1)})
  99             return url
 100
 101     def _kaltura_api_call(self, video_id, actions, *args, **kwargs):
 102         params = actions[0]
 103         if len(actions) > 1:
 104             for i, a in enumerate(actions[1:], start=1):
 105                 for k, v in a.items():
 106                     params['%d:%s' % (i, k)] = v
 107
 108         data = self._download_json(
 109             self._SERVICE_URL + self._SERVICE_BASE,
 110             video_id, query=params, *args, **kwargs)
 111
 112         status = data if len(actions) == 1 else data[0]
 113         if status.get('objectType') == 'KalturaAPIException':
 114             raise ExtractorError(
 115                 '%s said: %s' % (self.IE_NAME, status['message']))
 116
 117         return data
 118
 119     def _get_kaltura_signature(self, video_id, partner_id):
 120         actions = [{
 121             'apiVersion': '3.1',
 122             'expiry': 86400,
 123             'format': 1,
 124             'service': 'session',
 125             'action': 'startWidgetSession',
 126             'widgetId': '_%s' % partner_id,
 127         }]
 128         return self._kaltura_api_call(
 129             video_id, actions, note='Downloading Kaltura signature')['ks']
 130
 131     def _get_video_info(self, video_id, partner_id):
 132         signature = self._get_kaltura_signature(video_id, partner_id)
 133         actions = [
 134             {
 135                 'action': 'null',
 136                 'apiVersion': '3.1.5',
 137                 'clientTag': 'kdp:v3.8.5',
 138                 'format': 1,  # JSON, 2 = XML, 3 = PHP
 139                 'service': 'multirequest',
 140                 'ks': signature,
 141             },
 142             {
 143                 'action': 'get',
 144                 'entryId': video_id,
 145                 'service': 'baseentry',
 146                 'version': '-1',
 147             },
 148             {
 149                 'action': 'getbyentryid',
 150                 'entryId': video_id,
 151                 'service': 'flavorAsset',
 152             },
 153         ]
 154         return self._kaltura_api_call(
 155             video_id, actions, note='Downloading video info JSON')
 156
 157     def _real_extract(self, url):
 158         url, smuggled_data = unsmuggle_url(url, {})
 159         service_url = smuggled_data.get('service_url')
 160         if service_url:
 161             self._SERVICE_URL = service_url
 162
 163         mobj = re.match(self._VALID_URL, url)
 164         partner_id, entry_id = mobj.group('partner_id', 'id')
 165         ks = None
 166         if partner_id and entry_id:
 167             info, flavor_assets = self._get_video_info(entry_id, partner_id)
 168         else:
 169             path, query = mobj.group('path', 'query')
 170             if not path and not query:
 171                 raise ExtractorError('Invalid URL', expected=True)
 172             params = {}
 173             if query:
 174                 params = compat_parse_qs(query)
 175             if path:
 176                 splitted_path = path.split('/')
 177                 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
 178             if 'wid' in params:
 179                 partner_id = params['wid'][0][1:]
 180             elif 'p' in params:
 181                 partner_id = params['p'][0]
 182             else:
 183                 raise ExtractorError('Invalid URL', expected=True)
 184             if 'entry_id' in params:
 185                 entry_id = params['entry_id'][0]
 186                 info, flavor_assets = self._get_video_info(entry_id, partner_id)
 187             elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
 188                 reference_id = params['flashvars[referenceId]'][0]
 189                 webpage = self._download_webpage(url, reference_id)
 190                 entry_data = self._parse_json(self._search_regex(
 191                     r'window\.kalturaIframePackageData\s*=\s*({.*});',
 192                     webpage, 'kalturaIframePackageData'),
 193                     reference_id)['entryResult']
 194                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
 195                 entry_id = info['id']
 196             else:
 197                 raise ExtractorError('Invalid URL', expected=True)
 198             ks = params.get('flashvars[ks]', [None])[0]
 199
 200         source_url = smuggled_data.get('source_url')
 201         if source_url:
 202             referrer = base64.b64encode(
 203                 '://'.join(compat_urlparse.urlparse(source_url)[:2])
 204                 .encode('utf-8')).decode('utf-8')
 205         else:
 206             referrer = None
 207
 208         def sign_url(unsigned_url):
 209             if ks:
 210                 unsigned_url += '/ks/%s' % ks
 211             if referrer:
 212                 unsigned_url += '?referrer=%s' % referrer
 213             return unsigned_url
 214
 215         data_url = info['dataUrl']
 216         if '/flvclipper/' in data_url:
 217             data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
 218
 219         formats = []
 220         for f in flavor_assets:
 221             # Continue if asset is not ready
 222             if f['status'] != 2:
 223                 continue
 224             video_url = sign_url(
 225                 '%s/flavorId/%s' % (data_url, f['id']))
 226             formats.append({
 227                 'format_id': '%(fileExt)s-%(bitrate)s' % f,
 228                 'ext': f.get('fileExt'),
 229                 'tbr': int_or_none(f['bitrate']),
 230                 'fps': int_or_none(f.get('frameRate')),
 231                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
 232                 'container': f.get('containerFormat'),
 233                 'vcodec': f.get('videoCodecId'),
 234                 'height': int_or_none(f.get('height')),
 235                 'width': int_or_none(f.get('width')),
 236                 'url': video_url,
 237             })
 238         if '/playManifest/' in data_url:
 239             m3u8_url = sign_url(data_url.replace(
 240                 'format/url', 'format/applehttp'))
 241             formats.extend(self._extract_m3u8_formats(
 242                 m3u8_url, entry_id, 'mp4', 'm3u8_native',
 243                 m3u8_id='hls', fatal=False))
 244
 245         self._check_formats(formats, entry_id)
 246         self._sort_formats(formats)
 247
 248         return {
 249             'id': entry_id,
 250             'title': info['name'],
 251             'formats': formats,
 252             'description': clean_html(info.get('description')),
 253             'thumbnail': info.get('thumbnailUrl'),
 254             'duration': info.get('duration'),
 255             'timestamp': info.get('createdAt'),
 256             'uploader_id': info.get('userId'),
 257             'view_count': info.get('plays'),
 258         }