N-Gram-Based Text Categorization library to guess the language that input text
is written in. Fundamentally an adaption of of wiseguys libtextcat extended to
be UTF-8 aware.