I'm trying to write (or just find an existing) PHP method that can take a link and extract the url. The trick is, it needs to hold under the weight of strange looking domains like:

www.champa.kku.ac.th 

Looking at this one myself with human eyes, I still guessed it incorrectly: thought the domain would be kku.ac.th but that gives a dns error when visiting.

So anyone knows of a good way to reliably extract the domain from url:

http://site.com/hello.php
http://site.com.uk/hello.php
http://subdomain.site.com/hello.php
http://subdomain.site.com.uk/hello.php
http://www.champa.kku.ac.th/hello.php // and even the one I couldn't tell