正则表达式转换htmlspecialchars转义后的字符

Question

I take user input on my website and I allow formatting through a WYSIWYG editor.

Now I run the entire post I get in PHP through htmlspecialchars with ENT_quotes.

From there I would like to run a regex to convert my allowed rules back to real html.

So that

<p>TEXT</p> 

Thanks!

becomes

<p>TEXT</p>

Is there way to do this with regex? I am not familair with it. But basically look for the escaped tags and replace them with real tags, while keeping everything in the middle?

Edit: I want the regex to make the valid HTML codes. First I run the entire thing through htmlspecialchars to be 100% secure and get entities version of all < & > characters. Then I want a regex to convert ONLY the tags I want back to normal. I do not want to decode all entities back to their regular, I only want to decode the tags I want. so for example a regex search for <p>TEXT</p> back to <p>TEXT</p> - I don't want to reconvert just the <, &, > themselves. I want to reconvert only the tags I want. Which in this example is the paragraph tags.

Answer 1

I think this regex should work for what you've decribed:

/<(\/?)(allowed_tags_here_seperated_by_|)>/

Regex Demo: https://regex101.com/r/uX7cT1/1
PHP Demo: https://eval.in/579884

PHP Usage:

$allowed_tags = array('p', 'h1');
$regex = '/<(\/?)(' . implode('|', $allowed_tags) . ')>/';
$test_string = '&lt;p&gt;TEXT&lt;/p&gt; This & that don\'t < decode > these though <h2> &lt;h1&gt; &lt;p class=&quot;test&quot;&gt;';
echo preg_replace($regex, '<$1$2>', $test_string);

Oh, this also doesn't take into account self closing elements. <br />, <hr />, <img stuff />, etc.