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         urls = KalturaIE._extract_urls(webpage)
 117         return urls[0] if urls else None
 118
 119     @staticmethod
 120     def _extract_urls(webpage):
 121         # Embed codes: https://knowledge.kaltura.com/embedding-kaltura-media-players-your-site
 122         finditer = (
 123             re.finditer(
 124                 r"""(?xs)
 125                     kWidget\.(?:thumb)?[Ee]mbed\(
 126                     \{.*?
 127                         (?P<q1>['"])wid(?P=q1)\s*:\s*
 128                         (?P<q2>['"])_?(?P<partner_id>(?:(?!(?P=q2)).)+)(?P=q2),.*?
 129                         (?P<q3>['"])entry_?[Ii]d(?P=q3)\s*:\s*
 130                         (?P<q4>['"])(?P<id>(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\})
 131                 """, webpage)
 132             or re.finditer(
 133                 r'''(?xs)
 134                     (?P<q1>["'])
 135                         (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com(?::\d+)?/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)(?:(?!(?P=q1)).)*
 136                     (?P=q1).*?
 137                     (?:
 138                         (?:
 139                             entry_?[Ii]d|
 140                             (?P<q2>["'])entry_?[Ii]d(?P=q2)
 141                         )\s*:\s*|
 142                         \[\s*(?P<q2_1>["'])entry_?[Ii]d(?P=q2_1)\s*\]\s*=\s*
 143                     )
 144                     (?P<q3>["'])(?P<id>(?:(?!(?P=q3)).)+)(?P=q3)
 145                 ''', webpage)
 146             or re.finditer(
 147                 r'''(?xs)
 148                     <(?:iframe[^>]+src|meta[^>]+\bcontent)=(?P<q1>["'])
 149                       (?:https?:)?//(?:(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)
 150                       (?:(?!(?P=q1)).)*
 151                       [?&;]entry_id=(?P<id>(?:(?!(?P=q1))[^&])+)
 152                       (?:(?!(?P=q1)).)*
 153                     (?P=q1)
 154                 ''', webpage)
 155         )
 156         urls = []
 157         for mobj in finditer:
 158             embed_info = mobj.groupdict()
 159             for k, v in embed_info.items():
 160                 if v:
 161                     embed_info[k] = v.strip()
 162             url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
 163             escaped_pid = re.escape(embed_info['partner_id'])
 164             service_mobj = re.search(
 165                 r'<script[^>]+src=(["\'])(?P<id>(?:https?:)?//(?:(?!\1).)+)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
 166                 webpage)
 167             if service_mobj:
 168                 url = smuggle_url(url, {'service_url': service_mobj.group('id')})
 169             urls.append(url)
 170         return urls
 171
 172     def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
 173         params = actions[0]
 174         if len(actions) > 1:
 175             for i, a in enumerate(actions[1:], start=1):
 176                 for k, v in a.items():
 177                     params['%d:%s' % (i, k)] = v
 178
 179         data = self._download_json(
 180             (service_url or self._SERVICE_URL) + self._SERVICE_BASE,
 181             video_id, query=params, *args, **kwargs)
 182
 183         status = data if len(actions) == 1 else data[0]
 184         if status.get('objectType') == 'KalturaAPIException':
 185             raise ExtractorError(
 186                 '%s said: %s' % (self.IE_NAME, status['message']))
 187
 188         return data
 189
 190     def _get_video_info(self, video_id, partner_id, service_url=None):
 191         actions = [
 192             {
 193                 'action': 'null',
 194                 'apiVersion': '3.1.5',
 195                 'clientTag': 'kdp:v3.8.5',
 196                 'format': 1,  # JSON, 2 = XML, 3 = PHP
 197                 'service': 'multirequest',
 198             },
 199             {
 200                 'expiry': 86400,
 201                 'service': 'session',
 202                 'action': 'startWidgetSession',
 203                 'widgetId': '_%s' % partner_id,
 204             },
 205             {
 206                 'action': 'get',
 207                 'entryId': video_id,
 208                 'service': 'baseentry',
 209                 'ks': '{1:result:ks}',
 210                 'responseProfile:fields': 'createdAt,dataUrl,duration,name,plays,thumbnailUrl,userId',
 211                 'responseProfile:type': 1,
 212             },
 213             {
 214                 'action': 'getbyentryid',
 215                 'entryId': video_id,
 216                 'service': 'flavorAsset',
 217                 'ks': '{1:result:ks}',
 218             },
 219             {
 220                 'action': 'list',
 221                 'filter:entryIdEqual': video_id,
 222                 'service': 'caption_captionasset',
 223                 'ks': '{1:result:ks}',
 224             },
 225         ]
 226         return self._kaltura_api_call(
 227             video_id, actions, service_url, note='Downloading video info JSON')
 228
 229     def _real_extract(self, url):
 230         url, smuggled_data = unsmuggle_url(url, {})
 231
 232         mobj = re.match(self._VALID_URL, url)
 233         partner_id, entry_id = mobj.group('partner_id', 'id')
 234         ks = None
 235         captions = None
 236         if partner_id and entry_id:
 237             _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id, smuggled_data.get('service_url'))
 238         else:
 239             path, query = mobj.group('path', 'query')
 240             if not path and not query:
 241                 raise ExtractorError('Invalid URL', expected=True)
 242             params = {}
 243             if query:
 244                 params = compat_parse_qs(query)
 245             if path:
 246                 splitted_path = path.split('/')
 247                 params.update(dict((zip(splitted_path[::2], [[v] for v in splitted_path[1::2]]))))
 248             if 'wid' in params:
 249                 partner_id = params['wid'][0][1:]
 250             elif 'p' in params:
 251                 partner_id = params['p'][0]
 252             elif 'partner_id' in params:
 253                 partner_id = params['partner_id'][0]
 254             else:
 255                 raise ExtractorError('Invalid URL', expected=True)
 256             if 'entry_id' in params:
 257                 entry_id = params['entry_id'][0]
 258                 _, info, flavor_assets, captions = self._get_video_info(entry_id, partner_id)
 259             elif 'uiconf_id' in params and 'flashvars[referenceId]' in params:
 260                 reference_id = params['flashvars[referenceId]'][0]
 261                 webpage = self._download_webpage(url, reference_id)
 262                 entry_data = self._parse_json(self._search_regex(
 263                     r'window\.kalturaIframePackageData\s*=\s*({.*});',
 264                     webpage, 'kalturaIframePackageData'),
 265                     reference_id)['entryResult']
 266                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
 267                 entry_id = info['id']
 268                 # Unfortunately, data returned in kalturaIframePackageData lacks
 269                 # captions so we will try requesting the complete data using
 270                 # regular approach since we now know the entry_id
 271                 try:
 272                     _, info, flavor_assets, captions = self._get_video_info(
 273                         entry_id, partner_id)
 274                 except ExtractorError:
 275                     # Regular scenario failed but we already have everything
 276                     # extracted apart from captions and can process at least
 277                     # with this
 278                     pass
 279             else:
 280                 raise ExtractorError('Invalid URL', expected=True)
 281             ks = params.get('flashvars[ks]', [None])[0]
 282
 283         source_url = smuggled_data.get('source_url')
 284         if source_url:
 285             referrer = base64.b64encode(
 286                 '://'.join(compat_urlparse.urlparse(source_url)[:2])
 287                 .encode('utf-8')).decode('utf-8')
 288         else:
 289             referrer = None
 290
 291         def sign_url(unsigned_url):
 292             if ks:
 293                 unsigned_url += '/ks/%s' % ks
 294             if referrer:
 295                 unsigned_url += '?referrer=%s' % referrer
 296             return unsigned_url
 297
 298         data_url = info['dataUrl']
 299         if '/flvclipper/' in data_url:
 300             data_url = re.sub(r'/flvclipper/.*', '/serveFlavor', data_url)
 301
 302         formats = []
 303         for f in flavor_assets:
 304             # Continue if asset is not ready
 305             if f.get('status') != 2:
 306                 continue
 307             # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g)
 308             # skip for now.
 309             if f.get('fileExt') == 'chun':
 310                 continue
 311             # DRM-protected video, cannot be decrypted
 312             if f.get('fileExt') == 'wvm':
 313                 continue
 314             if not f.get('fileExt'):
 315                 # QT indicates QuickTime; some videos have broken fileExt
 316                 if f.get('containerFormat') == 'qt':
 317                     f['fileExt'] = 'mov'
 318                 else:
 319                     f['fileExt'] = 'mp4'
 320             video_url = sign_url(
 321                 '%s/flavorId/%s' % (data_url, f['id']))
 322             format_id = '%(fileExt)s-%(bitrate)s' % f
 323             # Source format may not be available (e.g. kaltura:513551:1_66x4rg7o)
 324             if f.get('isOriginal') is True and not self._is_valid_url(
 325                     video_url, entry_id, format_id):
 326                 continue
 327             # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g
 328             # -f mp4-56)
 329             vcodec = 'none' if 'videoCodecId' not in f and f.get(
 330                 'frameRate') == 0 else f.get('videoCodecId')
 331             formats.append({
 332                 'format_id': format_id,
 333                 'ext': f.get('fileExt'),
 334                 'tbr': int_or_none(f['bitrate']),
 335                 'fps': int_or_none(f.get('frameRate')),
 336                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
 337                 'container': f.get('containerFormat'),
 338                 'vcodec': vcodec,
 339                 'height': int_or_none(f.get('height')),
 340                 'width': int_or_none(f.get('width')),
 341                 'url': video_url,
 342             })
 343         if '/playManifest/' in data_url:
 344             m3u8_url = sign_url(data_url.replace(
 345                 'format/url', 'format/applehttp'))
 346             formats.extend(self._extract_m3u8_formats(
 347                 m3u8_url, entry_id, 'mp4', 'm3u8_native',
 348                 m3u8_id='hls', fatal=False))
 349
 350         self._sort_formats(formats)
 351
 352         subtitles = {}
 353         if captions:
 354             for caption in captions.get('objects', []):
 355                 # Continue if caption is not ready
 356                 if caption.get('status') != 2:
 357                     continue
 358                 if not caption.get('id'):
 359                     continue
 360                 caption_format = int_or_none(caption.get('format'))
 361                 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
 362                     'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
 363                     'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml',
 364                 })
 365
 366         return {
 367             'id': entry_id,
 368             'title': info['name'],
 369             'formats': formats,
 370             'subtitles': subtitles,
 371             'description': clean_html(info.get('description')),
 372             'thumbnail': info.get('thumbnailUrl'),
 373             'duration': info.get('duration'),
 374             'timestamp': info.get('createdAt'),
 375             'uploader_id': info.get('userId') if info.get('userId') != 'None' else None,
 376             'view_count': info.get('plays'),
 377         }