Metaprogramming day

So it's day three, and today I kept extending the prototype. By now it is possible to add fmpz's, in good cases. Things are not completely optimized yet, and both code formatting and layout (and comments) still have a long way to go, but this feels like progress. The metaprogramming style I employ very heavily relies on the compiler being able to optimize it, and I feel like I should verify that. Thus tomorrow I'll extend the code to the extent that we can add arbitrary fmpz's, and then I'll start analysing (and testing) the generated code. This should be interesting.