]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/kaltura.py
Update README.md
[youtubedl] / youtube_dl / extractor / kaltura.py
index ddf1165ffb021005119622d3f162cbce9c637b55..49d13460df7f0edd4d2a08f97deaf831ba9d6a46 100644 (file)
@@ -23,11 +23,11 @@ class KalturaIE(InfoExtractor):
                 (?:
                     kaltura:(?P<partner_id>\d+):(?P<id>[0-9a-z_]+)|
                     https?://
-                        (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/
+                        (:?(?:www|cdnapi(?:sec)?)\.)?kaltura\.com(?::\d+)?/
                         (?:
                             (?:
                                 # flash player
-                                index\.php/kwidget|
+                                index\.php/(?:kwidget|extwidget/preview)|
                                 # html5 player
                                 html5/html5lib/[^/]+/mwEmbedFrame\.php
                             )
@@ -36,6 +36,12 @@ class KalturaIE(InfoExtractor):
                 '''
     _SERVICE_URL = 'http://cdnapi.kaltura.com'
     _SERVICE_BASE = '/api_v3/index.php'
+    # See https://github.com/kaltura/server/blob/master/plugins/content/caption/base/lib/model/enums/CaptionType.php
+    _CAPTION_TYPES = {
+        1: 'srt',
+        2: 'ttml',
+        3: 'vtt',
+    }
     _TESTS = [
         {
             'url': 'kaltura:269692:1_1jc2y3e4',
@@ -67,42 +73,101 @@ class KalturaIE(InfoExtractor):
             # video with subtitles
             'url': 'kaltura:111032:1_cw786r8q',
             'only_matching': True,
+        },
+        {
+            # video with ttml subtitles (no fileExt)
+            'url': 'kaltura:1926081:0_l5ye1133',
+            'info_dict': {
+                'id': '0_l5ye1133',
+                'ext': 'mp4',
+                'title': 'What Can You Do With Python?',
+                'upload_date': '20160221',
+                'uploader_id': 'stork',
+                'thumbnail': 're:^https?://.*/thumbnail/.*',
+                'timestamp': int,
+                'subtitles': {
+                    'en': [{
+                        'ext': 'ttml',
+                    }],
+                },
+            },
+            'skip': 'Gone. Maybe https://www.safaribooksonline.com/library/tutorials/introduction-to-python-anon/3469/',
+            'params': {
+                'skip_download': True,
+            },
+        },
+        {
+            'url': 'https://www.kaltura.com/index.php/extwidget/preview/partner_id/1770401/uiconf_id/37307382/entry_id/0_58u8kme7/embed/iframe?&flashvars[streamerType]=auto',
+            'only_matching': True,
+        },
+        {
+            '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',
+            'only_matching': True,
+        },
+        {
+            # unavailable source format
+            'url': 'kaltura:513551:1_66x4rg7o',
+            'only_matching': True,
         }
     ]
 
     @staticmethod
     def _extract_url(webpage):
-        mobj = (
-            re.search(
+        urls = KalturaIE._extract_urls(webpage)
+        return urls[0] if urls else None
+
+    @staticmethod
+    def _extract_urls(webpage):
+        # Embed codes: https://knowledge.kaltura.com/embedding-kaltura-media-players-your-site
+        finditer = (
+            re.finditer(
                 r"""(?xs)
                     kWidget\.(?:thumb)?[Ee]mbed\(
                     \{.*?
-                        (?P<q1>['\"])wid(?P=q1)\s*:\s*
-                        (?P<q2>['\"])_?(?P<partner_id>[^'\"]+)(?P=q2),.*?
-                        (?P<q3>['\"])entry_?[Ii]d(?P=q3)\s*:\s*
-                        (?P<q4>['\"])(?P<id>[^'\"]+)(?P=q4),
-                """, webpage) or
-            re.search(
+                        (?P<q1>['"])wid(?P=q1)\s*:\s*
+                        (?P<q2>['"])_?(?P<partner_id>(?:(?!(?P=q2)).)+)(?P=q2),.*?
+                        (?P<q3>['"])entry_?[Ii]d(?P=q3)\s*:\s*
+                        (?P<q4>['"])(?P<id>(?:(?!(?P=q4)).)+)(?P=q4)(?:,|\s*\})
+                """, webpage)
+            or re.finditer(
                 r'''(?xs)
-                    (?P<q1>["\'])
-                        (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com/.*?(?:p|partner_id)/(?P<partner_id>\d+).*?
+                    (?P<q1>["'])
+                        (?:https?:)?//cdnapi(?:sec)?\.kaltura\.com(?::\d+)?/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)(?:(?!(?P=q1)).)*
                     (?P=q1).*?
                     (?:
-                        entry_?[Ii]d|
-                        (?P<q2>["\'])entry_?[Ii]d(?P=q2)
-                    )\s*:\s*
-                    (?P<q3>["\'])(?P<id>.+?)(?P=q3)
-                ''', webpage))
-        if mobj:
+                        (?:
+                            entry_?[Ii]d|
+                            (?P<q2>["'])entry_?[Ii]d(?P=q2)
+                        )\s*:\s*|
+                        \[\s*(?P<q2_1>["'])entry_?[Ii]d(?P=q2_1)\s*\]\s*=\s*
+                    )
+                    (?P<q3>["'])(?P<id>(?:(?!(?P=q3)).)+)(?P=q3)
+                ''', webpage)
+            or re.finditer(
+                r'''(?xs)
+                    <(?:iframe[^>]+src|meta[^>]+\bcontent)=(?P<q1>["'])
+                      (?:https?:)?//(?:(?:www|cdnapi(?:sec)?)\.)?kaltura\.com/(?:(?!(?P=q1)).)*\b(?:p|partner_id)/(?P<partner_id>\d+)
+                      (?:(?!(?P=q1)).)*
+                      [?&;]entry_id=(?P<id>(?:(?!(?P=q1))[^&])+)
+                      (?:(?!(?P=q1)).)*
+                    (?P=q1)
+                ''', webpage)
+        )
+        urls = []
+        for mobj in finditer:
             embed_info = mobj.groupdict()
+            for k, v in embed_info.items():
+                if v:
+                    embed_info[k] = v.strip()
             url = 'kaltura:%(partner_id)s:%(id)s' % embed_info
             escaped_pid = re.escape(embed_info['partner_id'])
-            service_url = re.search(
-                r'<script[^>]+src=["\']((?:https?:)?//.+?)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
+            service_mobj = re.search(
+                r'<script[^>]+src=(["\'])(?P<id>(?:https?:)?//(?:(?!\1).)+)/p/%s/sp/%s00/embedIframeJs' % (escaped_pid, escaped_pid),
                 webpage)
-            if service_url:
-                url = smuggle_url(url, {'service_url': service_url.group(1)})
-            return url
+            if service_mobj:
+                url = smuggle_url(url, {'service_url': service_mobj.group('id')})
+            urls.append(url)
+        return urls
 
     def _kaltura_api_call(self, video_id, actions, service_url=None, *args, **kwargs):
         params = actions[0]
@@ -122,18 +187,6 @@ class KalturaIE(InfoExtractor):
 
         return data
 
-    def _get_kaltura_signature(self, video_id, partner_id, service_url=None):
-        actions = [{
-            'apiVersion': '3.1',
-            'expiry': 86400,
-            'format': 1,
-            'service': 'session',
-            'action': 'startWidgetSession',
-            'widgetId': '_%s' % partner_id,
-        }]
-        return self._kaltura_api_call(
-            video_id, actions, service_url, note='Downloading Kaltura signature')['ks']
-
     def _get_video_info(self, video_id, partner_id, service_url=None):
         actions = [
             {
@@ -154,6 +207,8 @@ class KalturaIE(InfoExtractor):
                 'entryId': video_id,
                 'service': 'baseentry',
                 'ks': '{1:result:ks}',
+                'responseProfile:fields': 'createdAt,dataUrl,duration,name,plays,thumbnailUrl,userId',
+                'responseProfile:type': 1,
             },
             {
                 'action': 'getbyentryid',
@@ -194,6 +249,8 @@ class KalturaIE(InfoExtractor):
                 partner_id = params['wid'][0][1:]
             elif 'p' in params:
                 partner_id = params['p'][0]
+            elif 'partner_id' in params:
+                partner_id = params['partner_id'][0]
             else:
                 raise ExtractorError('Invalid URL', expected=True)
             if 'entry_id' in params:
@@ -208,6 +265,17 @@ class KalturaIE(InfoExtractor):
                     reference_id)['entryResult']
                 info, flavor_assets = entry_data['meta'], entry_data['contextData']['flavorAssets']
                 entry_id = info['id']
+                # Unfortunately, data returned in kalturaIframePackageData lacks
+                # captions so we will try requesting the complete data using
+                # regular approach since we now know the entry_id
+                try:
+                    _, info, flavor_assets, captions = self._get_video_info(
+                        entry_id, partner_id)
+                except ExtractorError:
+                    # Regular scenario failed but we already have everything
+                    # extracted apart from captions and can process at least
+                    # with this
+                    pass
             else:
                 raise ExtractorError('Invalid URL', expected=True)
             ks = params.get('flashvars[ks]', [None])[0]
@@ -236,16 +304,38 @@ class KalturaIE(InfoExtractor):
             # Continue if asset is not ready
             if f.get('status') != 2:
                 continue
+            # Original format that's not available (e.g. kaltura:1926081:0_c03e1b5g)
+            # skip for now.
+            if f.get('fileExt') == 'chun':
+                continue
+            # DRM-protected video, cannot be decrypted
+            if f.get('fileExt') == 'wvm':
+                continue
+            if not f.get('fileExt'):
+                # QT indicates QuickTime; some videos have broken fileExt
+                if f.get('containerFormat') == 'qt':
+                    f['fileExt'] = 'mov'
+                else:
+                    f['fileExt'] = 'mp4'
             video_url = sign_url(
                 '%s/flavorId/%s' % (data_url, f['id']))
+            format_id = '%(fileExt)s-%(bitrate)s' % f
+            # Source format may not be available (e.g. kaltura:513551:1_66x4rg7o)
+            if f.get('isOriginal') is True and not self._is_valid_url(
+                    video_url, entry_id, format_id):
+                continue
+            # audio-only has no videoCodecId (e.g. kaltura:1926081:0_c03e1b5g
+            # -f mp4-56)
+            vcodec = 'none' if 'videoCodecId' not in f and f.get(
+                'frameRate') == 0 else f.get('videoCodecId')
             formats.append({
-                'format_id': '%(fileExt)s-%(bitrate)s' % f,
+                'format_id': format_id,
                 'ext': f.get('fileExt'),
                 'tbr': int_or_none(f['bitrate']),
                 'fps': int_or_none(f.get('frameRate')),
                 'filesize_approx': int_or_none(f.get('size'), invscale=1024),
                 'container': f.get('containerFormat'),
-                'vcodec': f.get('videoCodecId'),
+                'vcodec': vcodec,
                 'height': int_or_none(f.get('height')),
                 'width': int_or_none(f.get('width')),
                 'url': video_url,
@@ -263,11 +353,14 @@ class KalturaIE(InfoExtractor):
         if captions:
             for caption in captions.get('objects', []):
                 # Continue if caption is not ready
-                if f.get('status') != 2:
+                if caption.get('status') != 2:
+                    continue
+                if not caption.get('id'):
                     continue
+                caption_format = int_or_none(caption.get('format'))
                 subtitles.setdefault(caption.get('languageCode') or caption.get('language'), []).append({
                     'url': '%s/api_v3/service/caption_captionasset/action/serve/captionAssetId/%s' % (self._SERVICE_URL, caption['id']),
-                    'ext': caption.get('fileExt'),
+                    'ext': caption.get('fileExt') or self._CAPTION_TYPES.get(caption_format) or 'ttml',
                 })
 
         return {
@@ -279,6 +372,6 @@ class KalturaIE(InfoExtractor):
             'thumbnail': info.get('thumbnailUrl'),
             'duration': info.get('duration'),
             'timestamp': info.get('createdAt'),
-            'uploader_id': info.get('userId'),
+            'uploader_id': info.get('userId') if info.get('userId') != 'None' else None,
             'view_count': info.get('plays'),
         }