WIP: Renamed sun to su2, no changes yet

Created by: tobiaslcmueller

PR to discuss exact plans for U(2) implementation (should be a draft PR, but not available in private repos). Open questions:

• Spin-S: Which convention to use? Baez/Reuther has 2S for every site summation, leads to large vertices. Alternative: rescaling vertices with 1/2S. I would prefer latter convention,despite more factors in the code.
• Model vs symmetry: Giving this a bit more thought, I think we should get rid of the symmetry parameter exposed to the user. Reason being, that it could have twofold meaning: a) which Symmetry group the spin operators are representations of (i.e. SU(N)) b) which part of the symmetry group survives for a give model (e.g. U1xZ2 for XXZ models, but still the spin operators are SU(2) generators). The moddel should select the correctly symmetrized action.