A library for manipulating multidimensional geometries that are defined in terms of rational number coordinates.
Mostly, this library is for taking intersections, unions, and differences between polytopes in two and three dimensions. While there are math libraries already capable of doing similar things, they all assume they must deal with floating point number coordinates, leading to completely different kinds of calculations under the hood. This library is intended to start with the assumption that all values calculated with represent exact, infinitely precise numbers.
No assumption is made (hopefully) about what kind of rational numbers are being used, but a rational number type is included in the library. While this rational number type may well be useful in other places, it has a different set of caveats than someone might be used to if they're used to a programming language's primitive types (e.g. int, float, etc).