This PR restructures and extends the math.mc library. It moves most functions from math.mc into wither float.mc or int.mc.
Additions and changes to float.mc include the following.
Addition of a isNaN function that checks if a value is a nan. This required a change in the code generation for OCaml so that eqf t t will return false if t is a nan (this was not the case before).
Addition of a cmpf function that compares floats. This function equates nan with itself and orders floats so that nan is less than negative infinity.
Updates to several implementations to handle nan arguments.
This PR restructures and extends the
math.mclibrary. It moves most functions frommath.mcinto witherfloat.mcorint.mc.Additions and changes to
float.mcinclude the following.isNaNfunction that checks if a value is anan. This required a change in the code generation for OCaml so thateqf t twill returnfalseiftis anan(this was not the case before).cmpffunction that compares floats. This function equatesnanwith itself and orders floats so thatnanis less than negative infinity.nanarguments.https://github.com/miking-lang/miking-dppl/pull/161 should be merged together with this PR as this PR exposes a bug in the the cppl-to-mexpr compiler.