Problem:    Comparing utf-8 sequences does not handle different byte sizes
            correctly.
Solution:   Get the byte size of each character. (Dominique Pelle)