Bad Language Filter - Exact Words

[Petsmacker]

I'm trying to work on a posting system for my members but obviously want to be able to filter obscenities.
My problem is that the word 'ass' is a word I'd want to block. However if I say 'assess' - my filter would incorrectly block it.
I need help to create a filter that filters the EXACT word. I've found plenty of others on the net that don't do that and just do something like '***ess' (ass) and '****ake' (shit)

Here's my script in progress (yes its basic but bear with me):
[code]<?php
$cuss=mysql_query("SELECT word FROM swearwords");
$num8=mysql_numrows($cuss);$i8=0;
while ($i8 < $num8) {
$word=mysql_result($cuss,$i8,"word");
$countsws=substr_count(strtolower($postedmessage), strtolower($word));
if ($countsws > 0){$i8=$num8;$theword=$word;} // Halts loop and grabs swear word in $theword variable
++$i8;}?>[/code]

How can I get it to block the word and ONLY the word?

[Petsmacker]

You're telling me that of all the advanced problems people post on these forums, this one has no solvable solution?

[utexas_pjm]

Maybe I'm oversimplifying, but couldn't you just do something like:

[code]
str_replace(' ass ', ' *** ', $string);
[/code]

?

Best,

Patrick

[Petsmacker]

The idea had gone through my mind but if somebody wrote something like:
'Petsmacker is an ass.'
It wouldn't block it because the word 'ass' was next to a full stop. This also happens if somebody JUST wrote the word 'ass' on a post and then submitted.

[utexas_pjm]

I'm not a regex expert, but something like this might work:

[code]
<?php
$bad = 'ass';
$foo = 'ass asset bass ass bass ass brass class ass';
echo preg_replace('/(^'.$bad.')(\.|,|!| )|(\.|,|!| )('.$bad.')(\.|,|!| )|(\.|,|!| )('.$bad.'$)|(^'.$bad.'$)/', ' *** ', $foo);

// Outputs: *** asset bass *** bass *** brass class ***

?>
[/code]

[Petsmacker]

How could I then get it to stop the loop if an offending word was found
And then turn the offending word into a variable?

------------------------------

I've worked a little on my original script, it sorts the problem of spaces and punctuation besides real swear words, my only problem now is line breaks.

If I write [quote]My name is ass[/quote]
It would find it fine and stop it.

However if I write it like this:
[quote]My name is

ass[/quote]
Then it doesn't, it sees the word as 'isass'...anyone know how to fix that?

Script in question:
[code]<?php
$chunks = explode(" ", $postedmessage);
$chunks = preg_replace('~[^a-zA-Z]~', '', $chunks);
$countarr=count($chunks);$c=0;
while ($c < $countarr){
$cuss=mysql_query("SELECT word FROM swearwords");
$num8=mysql_numrows($cuss);$i8=0;
while ($i8 < $num8) {
$word=mysql_result($cuss,$i8,"word");
if ($chunks[$c] == $word){$c=$countarr;$i8=$num8;$theword=$word;$countsws=1;} // Halts everything and alerts filter with $countsws as 1
++$i8;}
++$c;}?>[/code]

-----------

If you/anybody could answer either of my questions it'd be really great!

[Petsmacker]

Heck, I wish it was possible to edit posts on here. I've managed to fix my own problem in my own script by changing
$chunks = explode(" ", $postedmessage);
to
$chunks = explode(" ", nl2br($postedmessage));

===

However, my way of doing it is reaaaally bad, I hope someone can answer my original question with an answer that doesn't use so much server brains.