small problem with preg_match for url validation

[Tijmen]

Hi,

I got the following code to check if the submitted url is valid.

 if(!preg_match('/(http):\/\/([a-z0-9]+)\.[a-z]{2,4}/i', $_POST['website'])) {
     echo 'wrong';
 } else {
     echo 'oke';
 }

When i enter

@$.#$ = wrong
aa.aa = fine
aa.a = wrong which makes sense, but when i enter aa.aaaaaa it's also fine. This {2,4} tells it that it needs to be between 2 and 4 characters, if i enter 1 character it comes up with the error, but 5 characters is fine?

Another thing is this
aa.aa.com will validate, but aa.aa.c will also validate. Which it shouldn't do, so i try'd this.

if(!preg_match("/(http):\/\/([a-z0-9]*)?\.([a-z0-9-_])+\.([a-z]{2,4})/i", $_POST['website'])) {

If i understand it correctly it should check the for a-z0-9 after the http:// and then their might be another value after the. And after that you have the .com/.net But it doesn't really work..

Any idea's how i could fix these problems?

[szigeti]

Try this:

if (preg_match('#^https?://[a-z0-9-._]+\.[a-z0-9]{2,4}$#i', $_POST['website']))
{
    echo 'OK, that\'s a valid URL';
}
else
{
    echo 'invalid URL';
} 

This {2,4} tells it that it needs to be between 2 and 4 characters, if i enter 1 character it comes up with the error, but 5 characters is fine?

In that case you should put a $ at the end of the regex, which is an "end of line". (^ means "start of line" by the way). The bug was that it has found 2--4 characters, but you didn't told the regex engine that the URL should end there.

There's a really useful page about regular expressions here: http://regular-expression.info/