Fast arbitrary precision correctly-rounded decimal floating point arithmetic - View it on GitHub