More Email ereg Validation

I'm trying to parse the different "parts" of an email address both for validation and munging. I have the following regex: ([a-z0-9_\.\-]+)@([a-z0-9\.-]+).([a-z]{3}) This works great, except it doesn't accept the country code domains (.au, etc). So I changed the number of characters to {2,3} at the end of the regex: ([a-z0-9_\.\-]+)@([a-z0-9\.-]+).([a-z]{2,3}) But then this works for the country code domains, but cuts off a character of the 3 char domains because it matches the 2 characters as well. So if I'm using the my@domain.com email address I end up with $1 = "my", $2 = "domain" and $3 = "om" What do I do to catch both cases correctly? Thanks in advance! -Dan