[WIP]Implement PSATD advance with the Picard implicit iterative solver

[dpgrote]

This implements use of the PSATD field advance coupled with the implicit Picard iterative solver, using a Strang split advance.

• Advect Maxwell using PSATD with no J, ½ step
• Advance particles along with dE/dt = -J, full step, iterating
• Advect Maxwell using PSATD with no J, ½ step

This requires the input parameter psatd.update_with_rho = 0.
With psatd.periodic_single_box_fft = 1, exact energy conservation is obtained. Otherwise good conservation is seen, but not exact ($\Delta E/E$ ~ 1.e-9 but will depend on parameters).
Convergence is found for wpedt <= 1.9 (compared to wpedt < 0.25 for FDTD).

Todo:

• Add CI test
• Implement specialized advance routines (as noted in source comments)

[JustinRayAngus]

Charge conservation?

[dpgrote]

Closed. Replaced by PR #5027.