Nested MODX tags are not removed from extract correctly

[Boddlnagg]

When there are MODX tags in the result extract, they are filtered out, but if these tags contain other tags, e.g. [[!foo? &blah=[[!+ph? &blah=Value]] &bar=Value]], the parameters of the outer tag after the first ]] are shown in the extract, which results in the following to be visible to users in the search results: &bar=`Value

It does not happen in all cases, but I could reproduce it when the inner tag was either an uncached placeholder ([[!+...]]) or a ressource content field ([[*...]]).

[hugopeek]

Can confirm that it happens when a nested tag contains a [[*...]] field.

[jopberkhout]

I'm am experiencing this as well with our MODX site: https://ctmeter.nl/algemeen/zoekresultaten?search=downloads&id=319. As @hugopeek says, it happens when a nested tag contains a [[*...]] field.

[Mark-H]

Maybe applying an output filter to the extract that strips out MODX tags can help? Kind of a tricky problem though.

[hugopeek]

Thanks for the suggestion @Mark-H. I circumvented it a little by loading the introtext instead, if the extract contains the string =`:

[[[[+extract:contains=`=``:then=`+introtext:stripTags=`span``:else=`+extract`]]]]

Not an ideal / permanent solution though..

[tillilab]

I'm having same issue, maybe adding an option to do a major cleanup of content?

for example in model/simplesearch/simplesearch.class.php

add in the createextract function, just after sanitize:

$regx = '/[\w\W^]+/i'; // strip anything between ticks including ticks (``)
$text = preg_replace( $regx, '', $text );
$regx = '/&(amp;)?[a-zA-Z0-9_-.?^=]+=/i'; // strip remaining tag qualifiers ( &tag= )
$text = preg_replace( $regx, '', $text );
`
I took the idea/code from this post https://forums.modx.com/thread/87390/strip-modx-tags-from-simplesearch-results#dis-post-533583