Strange behaviour of REGEX in a Wordpress filter (trying to suppress emtpy paragraphs)

Question

I've been trying to add a filter in Wordpress to suppress empty paragraphs. I found a number of alternatives online (here is one of them https://gomakethings.com/remove-empty-paragraphs-added-by-wordpress/ ) but, so far, have failed miserably. After testing various hypothesis, I've found this strange regex behaviour happening:

I've created and tested this small script by itself (all it does is replace a space with a dash):

<?php 
$content = 'abc def';
$patterns = array('#\s#');
$replacements = array('-');
$new_content = preg_replace($patterns, $replacements, $content); 
echo $new_content;

The result, predictably was:

abc-def

Than I've added the same code as a filter in wordpress:

add_filter('the_content', array($this,'empty_paragraph_fix'),20,1);
function empty_paragraph($content) {
    $patterns = array('#\s*#');
    $replacements = array('-');
    $content = preg_replace($patterns, $replacements, $content);
    return $content;
}

But this time it gave me some very strange results. On a content that originally had something like this:

<div class="campus-list">
<ul class= ...

It returned this:

-<-d-i-v--c-l-a-s-s-=-"-c-a-m-p-u-s---l-i-s-t--"-->--<-u-l--c-l-a-s-s-=-

What am I missing here? Have you ever seen anything like this?

Answer 1

In your first example, your regex is \s, which will match a whitespace character. In the second example, you have \s*, which will match zero or more whitespace characters. You're probably looking for \s+, one or more whitespace characters.