_ 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             # 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_video_info(self, video_id, partner_id, service_url=None):
 126         actions = [
 127             {
 128                 'action': 'null',
 129                 'apiVersion': '3.1.5',
 130                 'clientTag': 'kdp:v3.8.5',
 131                 'format': 1,  # JSON, 2 = XML, 3 = PHP
 132                 'service': 'multirequest',
 133             },
 134             {
 135                 'expiry': 86400,
 136                 'service': 'session',
 137                 'action': 'startWidgetSession',
 138                 'widgetId': '_%s' % partner_id,
 139             },
 140             {
 141                 'action': 'get',
 142                 'entryId': video_id,
 143                 'service': 'baseentry',
 144                 'ks': '{1:result:ks}',
 145             },
 146             {
 147                 'action': 'getbyentryid',
 148                 'entryId': video_id,
 149                 'service': 'flavorAsset',
 150                 'ks': '{1:result:ks}',
 151             },
 152             {
 153                 'action': 'list',
 154                 'filter:entryIdEqual': video_id,
 155                 'service': 'caption_captionasset',
 156                 'ks': '{1:result:ks}',
 157             },
 158         ]
 159         return self._kaltura_api_call(
 160             video_id, actions, service_url, note='Downloading video info JSON')
 161
 162     def _real_extract(self, url):
 163         url, smuggled_data = unsmuggle_url(url, {})
 164
 165         mobj = re.match(self._VALID_URL, url)
 166         partner_id, entry_id = mobj.group('partner_id', 'id')
 167         ks = None
 168         captions = None
 169         if partner_id and entry_id:
 170             _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
 171         else:
 172             path, query = mobj.group('path', 'query')
 173             if not path and not query:
 174                 raise ExtractorError('Invalid URL', expected=True)
 175             params = {}
 176             if query:
 177                 params = compat_parse_qs(query)
 178             if path:
 179                 splitted_path = path.split('/')
 180                 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
 181             if 'wid' in params:
 182                 partner_id = params['wid'][0][1:]
 183             elif 'p' in params:
 184                 partner_id = params['p'][0]
 185             else:
 186                 raise ExtractorError('Invalid URL', expected=True)
 187             if 'entry_id' in params:
 188                 entry_id = params['entry_id'][0]
 189                 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
 190             elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
 191                 reference_id = params['flashvars[referenceId]'][0]
 192                 webpage = self._download_webpage(url, reference_id)
 193                 entry_data = self._parse_json(self._search_regex(
 194                     r'window\.kalturaIframePackageData\s*=\s*({.*});',
 195                     webpage, 'kalturaIframePackageData'),
 196                     reference_id)['entryResult']
 197                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
 198                 entry_id = info['id']
 199             else:
 200                 raise ExtractorError('Invalid URL', expected=True)
 201             ks = params.get('flashvars[ks]', [None])[0]
 202
 203         source_url = smuggled_data.get('source_url')
 204         if source_url:
 205             referrer = base64.b64encode(
 206                 '://'.join(compat_urlparse.urlparse(source_url)[:2])
 207                 .encode('utf-8')).decode('utf-8')
 208         else:
 209             referrer = None
 210
 211         def sign_url(unsigned_url):
 212             if ks:
 213                 unsigned_url += '/ks/%s' % ks
 214             if referrer:
 215                 unsigned_url += '?referrer=%s' % referrer
 216             return unsigned_url
 217
 218         data_url = info['dataUrl']
 219         if '/flvclipper/' in data_url:
 220             data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
 221
 222         formats = []
 223         for f in flavor_assets:
 224             # Continue if asset is not ready
 225             if f.get('status') != 2:
 226                 continue
 227             video_url = sign_url(
 228                 '%s/flavorId/%s' % (data_url, f['id']))
 229             formats.append({
 230                 'format_id': '%(fileExt)s-%(bitrate)s' % f,
 231                 'ext': f.get('fileExt'),
 232                 'tbr': int_or_none(f['bitrate']),
 233                 'fps': int_or_none(f.get('frameRate')),
 234                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
 235                 'container': f.get('containerFormat'),
 236                 'vcodec': f.get('videoCodecId'),
 237                 'height': int_or_none(f.get('height')),
 238                 'width': int_or_none(f.get('width')),
 239                 'url': video_url,
 240             })
 241         if '/playManifest/' in data_url:
 242             m3u8_url = sign_url(data_url.replace(
 243                 'format/url', 'format/applehttp'))
 244             formats.extend(self._extract_m3u8_formats(
 245                 m3u8_url, entry_id, 'mp4', 'm3u8_native',
 246                 m3u8_id='hls', fatal=False))
 247
 248         self._sort_formats(formats)
 249
 250         subtitles = {}
 251         if captions:
 252             for caption in captions.get('objects', []):
 253                 # Continue if caption is not ready
 254                 if f.get('status') != 2:
 255                     continue
 256                 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
 257                     'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
 258                     'ext': caption.get('fileExt'),
 259                 })
 260
 261         return {
 262             'id': entry_id,
 263             'title': info['name'],
 264             'formats': formats,
 265             'subtitles': subtitles,
 266             'description': clean_html(info.get('description')),
 267             'thumbnail': info.get('thumbnailUrl'),
 268             'duration': info.get('duration'),
 269             'timestamp': info.get('createdAt'),
 270             'uploader_id': info.get('userId'),
 271             'view_count': info.get('plays'),
 272         }