For extra credit:
Support Unicode characters.
Write a second function (possibly as a wrapper to the first) which detects inexact palindromes, i.e. phrases that are palindromes if white-space and punctuation is ignored and case-insensitive comparison is used.