preg_replace() syntax

[sciencenerd]

Hey, I'm creating a website that needs to allow users to submit content.  However, I'm trying to "clean" that content to a large degree because it will show on the page, so if there's any malicious formatting or scripts I need to remove those.  The code I'm about to post is *supposed* to remove all tags from $string except for <b >, <i > and <a > and additionally ensure that <b > and <i > have no arguments, while <a > only has the href="" argument and no others (note: spaces in tags to prevent the forum from rendering them).  The code:

 

        $string = strip_tags($string, '<b><i><a>' . $exceptions);

        $pattern[0] = '/(\<b)(.*)(\>)/e';

        $pattern[1] = '/(\<i)(.+?)(\>)/e';

        $pattern[2] = '/(\<a)(.+?)(href=\")(.+?)(\")(.+?)(\>)/e';

        $replacement[0] = '\\1\\3';

        $replacement[1] = '\\1\\3';

        $replacement[2] = '\\1\\3\\4\\5\\7';

        $string = preg_replace($pattern, $replacement, $string);

 

When I try to run this on the string '<b >test</b >', for example, the following errors are generated by the compiler:

 

Parse error: parse error, unexpected '<' in /home/www/URL of site/standardfunctions.php(56) : regexp code on line 1

 

Fatal error: preg_replace(): Failed evaluating code: <b > *SPACE ADDED* in /home/www/URL of site/standardfunctions.php on line 56

 

Line 56 in this case corresponds to the final line of my posted code.  Thanks in advance for any insight on this annoying problem!

[btherl]

That code doesn't produce errors here.  Can you post your entire script please?

[sciencenerd]

Okay, here's the whole function (I commented out everything else in the file to see if it would help, but no dice.  My problem has to be inside the function I think).

 

<?php

 

function CleanText($string, $class=null, $exceptions=null)

{

    if ($class == 'strict')

    {

        $string = strip_tags($string, '<b ><i ><a >' . $exceptions);

        $pattern[0] = '/(\<b)(.*)(\>)/e';

        $pattern[1] = '/(\<i)(.+?)(\>)/e';

        $pattern[2] = '/(\<a)(.+?)(href=\")(.+?)(\")(.+?)(\>)/e';

        $replacement[0] = '\\1\\3';

        $replacement[1] = '\\1\\3';

        $replacement[2] = '\\1\\3\\4\\5\\7';

        $string = preg_replace($pattern, $replacement, $string);

    }

    elseif ($class == 'lax')

    {

        $string = strip_tags($string, '<b ><i ><a ><span ><div ><table ><tr ><td ><h1 ><h2 ><h3 ><h4 ><h5 ><h6 ><br ><br /><ol ><ul ><li ><img ><img />' . $exceptions);

    }

    elseif ($class == null)

    {

        $string = strip_tags($string, $exceptions);

        $string = preg_replace('/\"/', '', $string);

    }

return $string;

}

 

echo CleanText('<b test>is a test</b test>', 'strict');

 

?>

[MadTechie]

no syntax errors in thats code..

[sciencenerd]

Ok, my mistake.  The reason the previous code compiled correctly is because of the spaces I put in the HTML tags to keep them from formatting my text when copying it to the form.  Now that I've found the "code" attribute on the forum, I can copy-paste directly.  Here's the code without the spaces in tags.  This is the code I've tested EXACTLY that produces the errors.  My apologies for the earlier ones that didn't.

<?php

function CleanText($string, $class=null, $exceptions=null)
{
    if ($class == 'strict')
    {
        $string = strip_tags($string, '<b><i><a>' . $exceptions);
        $pattern[0] = '/(\<b)(.*)(\>)/e';
        $pattern[1] = '/(\<i)(.+?)(\>)/e';
        $pattern[2] = '/(\<a)(.+?)(href=\")(.+?)(\")(.+?)(\>)/e';
        $replacement[0] = '\\1\\3';
        $replacement[1] = '\\1\\3';
        $replacement[2] = '\\1\\3\\4\\5\\7';
        $string = preg_replace($pattern, $replacement, $string);
    }
    elseif ($class == 'lax')
    {
        $string = strip_tags($string, '<b><i><a><span><div><table><tr><td><h1><h2><h3><h4><h5><h6><br><ol><ul><li><img><img />' . $exceptions);
    }
    elseif ($class == null)
    {
        $string = strip_tags($string, $exceptions);
        $string = preg_replace('/\"/', '', $string);
    }
return $string;
}

echo CleanText('<b test>is a test</b test>', 'strict');

?>

[sciencenerd]

Ok, my partner found the issue with my code, for some reason when I removed the "e" modifier from the pattern string it works out fine.  Thanks for trying to help though, if I have any more questions throughout the development I'll surely ask them here!