Math::AnyNum - Arbitrary size precision for integers, rationals, floating-points and complex numbers.
