If I am building uniqueID based on first initial + last name + numeric and the user's name is something like O'Brien, is there an easy way to remove all punctuation, or do I simply need to use the replace tokens and specify the likely punctuation?

Seems like a xpath expression might get it done more efficiently?