Unsafe HTML by Feed

Hi @ulicgn,

Thank you for using Feedzy!

We have a helpful document that explains how you can preserve HTML in your feed items’ content. You can find it here: https://docs.themeisle.com/article/542-how-to-keep-html-in-feed-items-content.

Please feel free to reach out if you have any further questions or if there’s anything else we can assist you with.

@andreearadacina21 thank you for this pointer – I was not aware of this documentation. What I had tried was taken from https://docs.themeisle.com/article/942-in-feedzy-how-do-i#remove-html-tags, and it did not work for me ( since it was probably meant for the shortcode usage) Looks quite similar though, with minor differences.

So I will certainly try it and report back how it worked, however I can do that not before next Tuesday. Until then, I would leave this topic in unresolved state. Thanks and best regards,

Uli

@andreearadacina21 I disabled the FEEDZY_ALLOW_UNSAFE_HTML define and tried your code as suggested, with minor modifications. To identify the filter, I added “(unfiltered)” or “(filtered)” to the output:

function kdefd_summary_input( $description, $content, $feedURL ) {
    //If you want to use the item content as the description. If not, then remove this line
    //If feed don't have content meta, $content is already equal to $description
    $description = $content;

   if (strpos($feedURL, "my-trusted-feed-domain.de") !== false) {
        $description = $description . " (unfiltered)";
        } else {
    //Remove the item HTML tags (as in the default hook) if not in the above list
        $description = trim( strip_tags( $description ) ) . " (filtered) ";
    }

    //Remove hellip (as in the default hook)
    //Keep in mind that it will be added later in the plugin render
    $description = trim( chop( $description, '[…]' ) );

    return $description;
}

In the source feed, there is a CDATA description section containing some HTML tags:

<item>
 <guid>https://www.example.com/5540093</guid>
 <link>https://www.example.com/5540093</link>
 <pubDate>Wed, 21 Jun 2023 14:16:46 +0200</pubDate>
 <title><![CDATA[ My Title as CDATA ]]></title>
 <description><![CDATA[K?ln (ots) - 
SOME LONG TEXT ....

<div class="pp-image" data-src="https://example.com/story_big/20bdf4d9-c4f2-4bf3-be90-3a766f048dd8/Foto1.jpg" data-caption="Foto 1">
<span class="pp-caption">Foto 1</span><br><img src="https://example.com/story_big/20bdf4d9-c4f2-4bf3-be90-3a766f048dd8/Foto1.jpg" alt="Foto 1"/></div>

<div class="pp-image" data-src="https://example.com/story_big/f3e2b762-4387-4d87-877e-7ecb8246db2a/Foto2.jpg" data-caption="Foto 2">
<span class="pp-caption">Foto 2</span><br><img src="https://example.com/story_big/f3e2b762-4387-4d87-877e-7ecb8246db2a/Foto2.jpg" alt="Foto 2"/></div>

<div class="pp-source">Quelle:<a >https://www.example.com/5540093</a></div>

 ]]></description></item>

In the Map-Content section, I configured the [#item_description] placeholder for the import into posts.

The resulting post contains the “(unfiltered)” text, so the filter is indeed executed. However, the <img src… tags are missing, only the <a href…> tags are there. :

Output in the Post:
<div class="pp-image"><span class="pp-caption">Foto 1</span></div>
<div class="pp-image"><span class="pp-caption">Foto 2</span></div>
<div class="pp-source">Quelle:<a href="valid-URL" rel="noopener">....</a></div>
(unfiltered)

So far, the only way to keep the <img src seems to be the FEEDZY_ALLOW_UNSAFE_HTML define …

Regards
Ulrich

• This reply was modified 1 year, 4 months ago by ulicgn.
• This reply was modified 1 year, 4 months ago by ulicgn.

Hi @ulicgn!

Could you please provide us the feed URL so we can test it on our end?

Thank you!

Thanks for following up!! While the feed is technically “public” ( i.e. not protected ), the URL is not to be published in a forum. How could I send you the feed URL ( or a sample file – whatever you prefer) safely? Regards, Ulrich

@luciamarinescu I made a Demo feeed that is structurally identical to the original one and whose URL can be published. See here:

https://cdn01.koeln.de/uploads/files/demofeed.xml ( like the original feed, this feed validates when checked with the w3c online validator)

There is one item with img src in CDATA. With the code mentioned by @andreearadacina21 , the img src tags are wiped out, with FEEDZY_ALLOW_UNSAFE_HTML they are kept.

• This reply was modified 1 year, 4 months ago by ulicgn.

Hi @ulicgn!

Indeed, it should possible to display those tags with this line of code.

define( ‘FEEDZY_ALLOW_UNSAFE_HTML’, true );

Unfortunately, our support team can’t help with customizations such as creating a filter to allow unsafe html only for a certain feed as it’s beyond the level of assistance we provide. The good news is that other users from the community can see the thread and might want to help in this regard.

Thanks for understanding and have a nice day ahead!

What a pity, I had the impression that we were more than half way there. Obviously $description gets processed and tag-stripped once more later on, so the effects of this filter are overridden.

Even though I don’t think that someone will show up and solve this, I will leave this topic open for another couple of days before closing it.