For Unicode code point in the highest range, 1 bit shift operation on the 4th byte is missing, giving a wrong encoding