prevent decoding of html entities in textareas

preface: As far as i know, it is not possible to prevent the decoding of html entities in textaresas! The article describes my workaround for that issue.

what for: In my case, i use a <textarea> as basic html editor. Raw html snippets are stored in the database and are loaded into a <textarea>.

But without modification, the raw html ...

<span>&quot;usage: %s &lt;argument1&gt; &lt;argument2&gt;&quot;</span>

... is displayed as ...

<span>"usage: %s <argument1> <argument2>"</span>

... in the <textarea>. If you would store that back to the database, all html entities are lost.

workaround: I use a single regular expression to "prevent" the decoding.

Python example code:

#!/usr/bin/python
import re
raw_html = '<span>&quot;usage: %s &lt;argument1&gt; &lt;argument2&gt;&quot;</span>'
to_textarea = re.sub('&([#\w]+;)', '&amp;\g<1>', raw_html )
print to_textarea

PHP example code:

<?php
$raw_html = '<span>&quot;usage: %s &lt;argument1&gt; &lt;argument2&gt;&quot;</span>';
$to_textarea = preg_replace('/&([#\w]+;)/', '&amp;$1', $raw_html );
echo $to_textarea;
?>

The scripts above are replacing the "&" sign in every html entitie code with "&amp;", and returning the following string:

<span>&amp;quot;usage: %s &amp;lt;argument1&amp;gt; &amp;lt;argument2&amp;gt;&amp;quot;</span>

... If this string is loaded to the <textarea> rather than the raw one, you can store it back to database without destroying your content.