Raphaël G. Git Repositories - youtubedl/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             # video with subtitles
  68             'url': 'kaltura:111032:1_cw786r8q',
  69             'only_matching': True,
  70         }
  71     ]
  72
  73     @staticmethod
  74     def _extract_url(webpage):
  75         mobj = (
  76             re.search(
  77                 r"""(?xs)
  78                     kWidget\.(?:thumb)?[Ee]mbed\(
  79                     \{.*?
  80                         (?P<q1>['\"])wid(?P=q1)\s*:\s*
  81                         (?P<q2>['\"])_?(?P<partner_id>[^'\"]+)(?P=q2),.*?
  82                         (?P<q3>['\"])entry_?[Ii]d(?P=q3)\s*:\s*
  83                         (?P<q4>['\"])(?P<id>[^'\"]+)(?P=q4),
  84                 """, webpage) or
  85             re.search(
  86                 r'''(?xs)
  87                     (?P<q1>["\'])
  88                         (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com/.*?(?:p|partner_id)/(?P<partner_id>\d+).*?
  89                     (?P=q1).*?
  90                     (?:
  91                         entry_?[Ii]d|
  92                         (?P<q2>["\'])entry_?[Ii]d(?P=q2)
  93                     )\s*:\s*
  94                     (?P<q3>["\'])(?P<id>.+?)(?P=q3)
  95                 ''', webpage))
  96         if mobj:
  97             embed_info = mobj.groupdict()
  98             url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
  99             escaped_pid = re.escape(embed_info['partner_id'])
 100             service_url = re.search(
 101                 r'<script[^>]+src=["\']((?:https?:)?//.+?)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
 102                 webpage)
 103             if service_url:
 104                 url = smuggle_url(url, {'service_url': service_url.group(1)})
 105             return url
 106
 107     def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
 108         params = actions[0]
 109         if len(actions) > 1:
 110             for i, a in enumerate(actions[1:], start=1):
 111                 for k, v in a.items():
 112                     params['%d:%s' % (i, k)] = v
 113
 114         data = self._download_json(
 115             (service_url or self._SERVICE_URL) + self._SERVICE_BASE,
 116             video_id, query=params, *args, **kwargs)
 117
 118         status = data if len(actions) == 1 else data[0]
 119         if status.get('objectType') == 'KalturaAPIException':
 120             raise ExtractorError(
 121                 '%s said: %s' % (self.IE_NAME, status['message']))
 122
 123         return data
 124
 125     def _get_kaltura_signature(self, video_id, partner_id, service_url=None):
 126         actions = [{
 127             'apiVersion': '3.1',
 128             'expiry': 86400,
 129             'format': 1,
 130             'service': 'session',
 131             'action': 'startWidgetSession',
 132             'widgetId': '_%s' % partner_id,
 133         }]
 134         return self._kaltura_api_call(
 135             video_id, actions, service_url, note='Downloading Kaltura signature')['ks']
 136
 137     def _get_video_info(self, video_id, partner_id, service_url=None):
 138         actions = [
 139             {
 140                 'action': 'null',
 141                 'apiVersion': '3.1.5',
 142                 'clientTag': 'kdp:v3.8.5',
 143                 'format': 1,  # JSON, 2 = XML, 3 = PHP
 144                 'service': 'multirequest',
 145             },
 146             {
 147                 'expiry': 86400,
 148                 'service': 'session',
 149                 'action': 'startWidgetSession',
 150                 'widgetId': '_%s' % partner_id,
 151             },
 152             {
 153                 'action': 'get',
 154                 'entryId': video_id,
 155                 'service': 'baseentry',
 156                 'ks': '{1:result:ks}',
 157             },
 158             {
 159                 'action': 'getbyentryid',
 160                 'entryId': video_id,
 161                 'service': 'flavorAsset',
 162                 'ks': '{1:result:ks}',
 163             },
 164             {
 165                 'action': 'list',
 166                 'filter:entryIdEqual': video_id,
 167                 'service': 'caption_captionasset',
 168                 'ks': '{1:result:ks}',
 169             },
 170         ]
 171         return self._kaltura_api_call(
 172             video_id, actions, service_url, note='Downloading video info JSON')
 173
 174     def _real_extract(self, url):
 175         url, smuggled_data = unsmuggle_url(url, {})
 176
 177         mobj = re.match(self._VALID_URL, url)
 178         partner_id, entry_id = mobj.group('partner_id', 'id')
 179         ks = None
 180         captions = None
 181         if partner_id and entry_id:
 182             _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
 183         else:
 184             path, query = mobj.group('path', 'query')
 185             if not path and not query:
 186                 raise ExtractorError('Invalid URL', expected=True)
 187             params = {}
 188             if query:
 189                 params = compat_parse_qs(query)
 190             if path:
 191                 splitted_path = path.split('/')
 192                 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
 193             if 'wid' in params:
 194                 partner_id = params['wid'][0][1:]
 195             elif 'p' in params:
 196                 partner_id = params['p'][0]
 197             else:
 198                 raise ExtractorError('Invalid URL', expected=True)
 199             if 'entry_id' in params:
 200                 entry_id = params['entry_id'][0]
 201                 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
 202             elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
 203                 reference_id = params['flashvars[referenceId]'][0]
 204                 webpage = self._download_webpage(url, reference_id)
 205                 entry_data = self._parse_json(self._search_regex(
 206                     r'window\.kalturaIframePackageData\s*=\s*({.*});',
 207                     webpage, 'kalturaIframePackageData'),
 208                     reference_id)['entryResult']
 209                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
 210                 entry_id = info['id']
 211             else:
 212                 raise ExtractorError('Invalid URL', expected=True)
 213             ks = params.get('flashvars[ks]', [None])[0]
 214
 215         source_url = smuggled_data.get('source_url')
 216         if source_url:
 217             referrer = base64.b64encode(
 218                 '://'.join(compat_urlparse.urlparse(source_url)[:2])
 219                 .encode('utf-8')).decode('utf-8')
 220         else:
 221             referrer = None
 222
 223         def sign_url(unsigned_url):
 224             if ks:
 225                 unsigned_url += '/ks/%s' % ks
 226             if referrer:
 227                 unsigned_url += '?referrer=%s' % referrer
 228             return unsigned_url
 229
 230         data_url = info['dataUrl']
 231         if '/flvclipper/' in data_url:
 232             data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
 233
 234         formats = []
 235         for f in flavor_assets:
 236             # Continue if asset is not ready
 237             if f.get('status') != 2:
 238                 continue
 239             video_url = sign_url(
 240                 '%s/flavorId/%s' % (data_url, f['id']))
 241             formats.append({
 242                 'format_id': '%(fileExt)s-%(bitrate)s' % f,
 243                 'ext': f.get('fileExt'),
 244                 'tbr': int_or_none(f['bitrate']),
 245                 'fps': int_or_none(f.get('frameRate')),
 246                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
 247                 'container': f.get('containerFormat'),
 248                 'vcodec': f.get('videoCodecId'),
 249                 'height': int_or_none(f.get('height')),
 250                 'width': int_or_none(f.get('width')),
 251                 'url': video_url,
 252             })
 253         if '/playManifest/' in data_url:
 254             m3u8_url = sign_url(data_url.replace(
 255                 'format/url', 'format/applehttp'))
 256             formats.extend(self._extract_m3u8_formats(
 257                 m3u8_url, entry_id, 'mp4', 'm3u8_native',
 258                 m3u8_id='hls', fatal=False))
 259
 260         self._sort_formats(formats)
 261
 262         subtitles = {}
 263         if captions:
 264             for caption in captions.get('objects', []):
 265                 # Continue if caption is not ready
 266                 if f.get('status') != 2:
 267                     continue
 268                 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
 269                     'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
 270                     'ext': caption.get('fileExt'),
 271                 })
 272
 273         return {
 274             'id': entry_id,
 275             'title': info['name'],
 276             'formats': formats,
 277             'subtitles': subtitles,
 278             'description': clean_html(info.get('description')),
 279             'thumbnail': info.get('thumbnailUrl'),
 280             'duration': info.get('duration'),
 281             'timestamp': info.get('createdAt'),
 282             'uploader_id': info.get('userId'),
 283             'view_count': info.get('plays'),
 284         }