AFQMC converter update

[camelto2]

This PR updates the AFQMC interface to QMCPACK to properly handle hamiltonians with kpoints and the correct format of the wave function.

Changes are as follows

• Ion-Ion energy was incorrect for supercells tiled by kpoints. Need to scale ion-ion energy by nkpts
• Hamiltonian dimensions need to use Nup and Ndown of the tiled supercell. I count the electrons by reading the kpoint occupancies
• The NOMSD wave function was only writing the information for one kpoint, whereas we need the entire trial wavefunction for the tiled supercell. This is now handled correctly in write_afqmc_waves function.
• I put an abort if the calculation is noncollinear as I'm unsure if the matrix elements work for noncollinear yet.

If anything is unclear or you want changed let me know, but this now agrees with some reference inputs I have.
Sorry this took so long.