[SOLVED] regex for replacing underscores

[xnowandtheworldx]

Basically, I have a script where it takes a users input, and will replace !@#$% etc. etc. with an _. Now I need a regex expression that will remove all the extra _'s say for instance....

 

!hello*#world(*! would then be turned into "_hello__world___", now what I need to do is take that output and turn it into hello_world. Is there any possible way to do this with regex? I suck with regex, so any help would be nice! Thanks for any future help some of you may provide.

[effigy]

Have you original regex match multiple characters.

[xnowandtheworldx]

I'm using str_replace() currently to replace all the special symbols. So how might I go about to check for multiple instances of the _'s with regex? Thank you for any help.

[corbin]

preg_replace('/[_]+/', '', $str);

 

 

Or, just:

 

preg_replace('/[!@#$%]+/', '_', $str);

[xnowandtheworldx]

alright, i've got it fixed. Heres what i used.

 

$string = preg_replace( "/\_{1,}/", "_", $string); 

 

thanks for all your help guys!

[effigy]

Why run two replacements when you can do it once with the second expression?

[xnowandtheworldx]

Even if I did that expression wouldn't i still have to remove the extra underscores? basically, what i needed was a regex command that would remove the _'s from the beginning and the end, and any double _'s there may be. Plus I don't know how to do a regex command for all the symbols...i need it for symbols !@#$%^&*()-+={[}]|\;;"'<,>.?/

[effigy]

Try this:

 

<pre>
<?php	
echo $str = 'ABC!@#$%^&*()-+={[}]|\;;"\'<,>.?/DEF';
echo '<hr>';
echo preg_replace('/[\p{P}\p{S}]+/', '_', $str);
?>
</pre>