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(?::\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             # unavailable source format
 109             'url': 'kaltura:513551:1_66x4rg7o',
 110             'only_matching': True,
 111         }
 112     ]
 113
 114     @staticmethod
 115     def _extract_url(webpage):
 116         # Embed codes: https://knowledge.kaltura.com/embedding-kaltura-media-players-your-site
 117         mobj = (
 118             re.search(
 119                 r"""(?xs)
 120                     kWidget\.(?:thumb)?[Ee]mbed\(
 121                     \{.*?
 122                         (?P<q1>['"])wid(?P=q1)\s*:\s*
 123                         (?P<q2>['"])_?(?P<partner_id>(?:(?!(?P=q2)).)+)(?P=q2),.*?
 124                         (?P<q3>['"])entry_?[Ii]d(?P=q3)\s*:\s*
 125                         (?P<q4>['"])(?P<id>(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\})
 126                 """, webpage)
 127             or re.search(
 128                 r'''(?xs)
 129                     (?P<q1>["'])
 130                         (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com(?::\d+)?/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)(?:(?!(?P=q1)).)*
 131                     (?P=q1).*?
 132                     (?:
 133                         (?:
 134                             entry_?[Ii]d|
 135                             (?P<q2>["'])entry_?[Ii]d(?P=q2)
 136                         )\s*:\s*|
 137                         \[\s*(?P<q2_1>["'])entry_?[Ii]d(?P=q2_1)\s*\]\s*=\s*
 138                     )
 139                     (?P<q3>["'])(?P<id>(?:(?!(?P=q3)).)+)(?P=q3)
 140                 ''', webpage)
 141             or re.search(
 142                 r'''(?xs)
 143                     <(?:iframe[^>]+src|meta[^>]+\bcontent)=(?P<q1>["'])
 144                       (?:https?:)?//(?:(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)
 145                       (?:(?!(?P=q1)).)*
 146                       [?&;]entry_id=(?P<id>(?:(?!(?P=q1))[^&])+)
 147                       (?:(?!(?P=q1)).)*
 148                     (?P=q1)
 149                 ''', webpage)
 150         )
 151         if mobj:
 152             embed_info = mobj.groupdict()
 153             for k, v in embed_info.items():
 154                 if v:
 155                     embed_info[k] = v.strip()
 156             url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
 157             escaped_pid = re.escape(embed_info['partner_id'])
 158             service_mobj = re.search(
 159                 r'<script[^>]+src=(["\'])(?P<id>(?:https?:)?//(?:(?!\1).)+)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
 160                 webpage)
 161             if service_mobj:
 162                 url = smuggle_url(url, {'service_url': service_mobj.group('id')})
 163             return url
 164
 165     def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
 166         params = actions[0]
 167         if len(actions) > 1:
 168             for i, a in enumerate(actions[1:], start=1):
 169                 for k, v in a.items():
 170                     params['%d:%s' % (i, k)] = v
 171
 172         data = self._download_json(
 173             (service_url or self._SERVICE_URL) + self._SERVICE_BASE,
 174             video_id, query=params, *args, **kwargs)
 175
 176         status = data if len(actions) == 1 else data[0]
 177         if status.get('objectType') == 'KalturaAPIException':
 178             raise ExtractorError(
 179                 '%s said: %s' % (self.IE_NAME, status['message']))
 180
 181         return data
 182
 183     def _get_video_info(self, video_id, partner_id, service_url=None):
 184         actions = [
 185             {
 186                 'action': 'null',
 187                 'apiVersion': '3.1.5',
 188                 'clientTag': 'kdp:v3.8.5',
 189                 'format': 1,  # JSON, 2 = XML, 3 = PHP
 190                 'service': 'multirequest',
 191             },
 192             {
 193                 'expiry': 86400,
 194                 'service': 'session',
 195                 'action': 'startWidgetSession',
 196                 'widgetId': '_%s' % partner_id,
 197             },
 198             {
 199                 'action': 'get',
 200                 'entryId': video_id,
 201                 'service': 'baseentry',
 202                 'ks': '{1:result:ks}',
 203                 'responseProfile:fields': 'createdAt,dataUrl,duration,name,plays,thumbnailUrl,userId',
 204                 'responseProfile:type': 1,
 205             },
 206             {
 207                 'action': 'getbyentryid',
 208                 'entryId': video_id,
 209                 'service': 'flavorAsset',
 210                 'ks': '{1:result:ks}',
 211             },
 212             {
 213                 'action': 'list',
 214                 'filter:entryIdEqual': video_id,
 215                 'service': 'caption_captionasset',
 216                 'ks': '{1:result:ks}',
 217             },
 218         ]
 219         return self._kaltura_api_call(
 220             video_id, actions, service_url, note='Downloading video info JSON')
 221
 222     def _real_extract(self, url):
 223         url, smuggled_data = unsmuggle_url(url, {})
 224
 225         mobj = re.match(self._VALID_URL, url)
 226         partner_id, entry_id = mobj.group('partner_id', 'id')
 227         ks = None
 228         captions = None
 229         if partner_id and entry_id:
 230             _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
 231         else:
 232             path, query = mobj.group('path', 'query')
 233             if not path and not query:
 234                 raise ExtractorError('Invalid URL', expected=True)
 235             params = {}
 236             if query:
 237                 params = compat_parse_qs(query)
 238             if path:
 239                 splitted_path = path.split('/')
 240                 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
 241             if 'wid' in params:
 242                 partner_id = params['wid'][0][1:]
 243             elif 'p' in params:
 244                 partner_id = params['p'][0]
 245             elif 'partner_id' in params:
 246                 partner_id = params['partner_id'][0]
 247             else:
 248                 raise ExtractorError('Invalid URL', expected=True)
 249             if 'entry_id' in params:
 250                 entry_id = params['entry_id'][0]
 251                 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
 252             elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
 253                 reference_id = params['flashvars[referenceId]'][0]
 254                 webpage = self._download_webpage(url, reference_id)
 255                 entry_data = self._parse_json(self._search_regex(
 256                     r'window\.kalturaIframePackageData\s*=\s*({.*});',
 257                     webpage, 'kalturaIframePackageData'),
 258                     reference_id)['entryResult']
 259                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
 260                 entry_id = info['id']
 261                 # Unfortunately, data returned in kalturaIframePackageData lacks
 262                 # captions so we will try requesting the complete data using
 263                 # regular approach since we now know the entry_id
 264                 try:
 265                     _, info, flavor_assets, captions = self._get_video_info(
 266                         entry_id, partner_id)
 267                 except ExtractorError:
 268                     # Regular scenario failed but we already have everything
 269                     # extracted apart from captions and can process at least
 270                     # with this
 271                     pass
 272             else:
 273                 raise ExtractorError('Invalid URL', expected=True)
 274             ks = params.get('flashvars[ks]', [None])[0]
 275
 276         source_url = smuggled_data.get('source_url')
 277         if source_url:
 278             referrer = base64.b64encode(
 279                 '://'.join(compat_urlparse.urlparse(source_url)[:2])
 280                 .encode('utf-8')).decode('utf-8')
 281         else:
 282             referrer = None
 283
 284         def sign_url(unsigned_url):
 285             if ks:
 286                 unsigned_url += '/ks/%s' % ks
 287             if referrer:
 288                 unsigned_url += '?referrer=%s' % referrer
 289             return unsigned_url
 290
 291         data_url = info['dataUrl']
 292         if '/flvclipper/' in data_url:
 293             data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
 294
 295         formats = []
 296         for f in flavor_assets:
 297             # Continue if asset is not ready
 298             if f.get('status') != 2:
 299                 continue
 300             # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g)
 301             # skip for now.
 302             if f.get('fileExt') == 'chun':
 303                 continue
 304             # DRM-protected video, cannot be decrypted
 305             if f.get('fileExt') == 'wvm':
 306                 continue
 307             if not f.get('fileExt'):
 308                 # QT indicates QuickTime; some videos have broken fileExt
 309                 if f.get('containerFormat') == 'qt':
 310                     f['fileExt'] = 'mov'
 311                 else:
 312                     f['fileExt'] = 'mp4'
 313             video_url = sign_url(
 314                 '%s/flavorId/%s' % (data_url, f['id']))
 315             format_id = '%(fileExt)s-%(bitrate)s' % f
 316             # Source format may not be available (e.g. kaltura:513551:1_66x4rg7o)
 317             if f.get('isOriginal') is True and not self._is_valid_url(
 318                     video_url, entry_id, format_id):
 319                 continue
 320             # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g
 321             # -f mp4-56)
 322             vcodec = 'none' if 'videoCodecId' not in f and f.get(
 323                 'frameRate') == 0 else f.get('videoCodecId')
 324             formats.append({
 325                 'format_id': format_id,
 326                 'ext': f.get('fileExt'),
 327                 'tbr': int_or_none(f['bitrate']),
 328                 'fps': int_or_none(f.get('frameRate')),
 329                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
 330                 'container': f.get('containerFormat'),
 331                 'vcodec': vcodec,
 332                 'height': int_or_none(f.get('height')),
 333                 'width': int_or_none(f.get('width')),
 334                 'url': video_url,
 335             })
 336         if '/playManifest/' in data_url:
 337             m3u8_url = sign_url(data_url.replace(
 338                 'format/url', 'format/applehttp'))
 339             formats.extend(self._extract_m3u8_formats(
 340                 m3u8_url, entry_id, 'mp4', 'm3u8_native',
 341                 m3u8_id='hls', fatal=False))
 342
 343         self._sort_formats(formats)
 344
 345         subtitles = {}
 346         if captions:
 347             for caption in captions.get('objects', []):
 348                 # Continue if caption is not ready
 349                 if caption.get('status') != 2:
 350                     continue
 351                 if not caption.get('id'):
 352                     continue
 353                 caption_format = int_or_none(caption.get('format'))
 354                 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
 355                     'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
 356                     'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml',
 357                 })
 358
 359         return {
 360             'id': entry_id,
 361             'title': info['name'],
 362             'formats': formats,
 363             'subtitles': subtitles,
 364             'description': clean_html(info.get('description')),
 365             'thumbnail': info.get('thumbnailUrl'),
 366             'duration': info.get('duration'),
 367             'timestamp': info.get('createdAt'),
 368             'uploader_id': info.get('userId') if info.get('userId') != 'None' else None,
 369             'view_count': info.get('plays'),
 370         }