plugins/feeds: Add optional filter_function

This optional callable is called for every body and summary that is
added to a feed_entry.
This can be used to remove unwanted elements from the feed (e.g. using
bs4).

Signed-off-by: Chris Fiege <cfi@pengutronix.de>

doc/content/plugins/feeds.rst

@@ -91,6 +91,15 @@ Usage
             # of every feed entry
             'published':
                 lambda content: content['date'].strftime('%Y-%m-%d %H:%M:%S+01:00'),
+
+            # This optional callable is called for every body and summary that is added to a feed entry.
+            # This can be used to remove unwanted elements from the feed (e.g. by using beautifulsoup4).
+            # The callable must follow this signature and return the filtered html as string:
+            # |def _filter_function(html_input):
+            # |     # do the filtering
+            # |     return html_output
+            'html_filter': _callable,
+            '
         },
     ]
 

flamingo/plugins/feeds.py

@@ -122,6 +122,8 @@ def pre_build(self, context):
 
                     if i['content_body']:
                         exitcode, output = context.pre_render(i)
+                        if 'html_filter' in feed_config:
+                            output = feed_config['html_filter'](output)
                         fe.content(output, type='html')
 
                     if i['authors']:
@@ -131,7 +133,10 @@ def pre_build(self, context):
                             })
 
                     if i['summary']:
-                        fe.summary(str(i['summary']))
+                        summary = str(i['summary'])
+                        if 'html_filter' in feed_config:
+                            summary = feed_config['html_filter'](summary)
+                        fe.summary(summary)
 
                     if feed_config['type'] == 'podcast':
                         fe.enclosure(fe_podcast_url, str(fe_podcast_size), fe_podcast_type)