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