Okay, I got it done. Now--don't laugh--my list of curses is very G-rated. I wouldn't really check for those!
This is very simple implementation just to give you a direction to go in. When user clicks the OK button it checks the length of the username (just for kicks) and it looks for any instance of a bad word in a string, even if its appearance in a longer word is perfectly innocent. For example, it will flag "damnable" because it contains "damn". Also, if there are two "curse" words, like "goshdarnit" the error message comes up twice. You could set a flag if you only want it to come up once, or better yet use a counter to say how many disallowed words appear.
EDIT: I just couldn't leave it the way it was. I added a couple of events to keep track of the number of "curse" words found.