PlaneWaveModel and kpoints names about dftk.jl

[mfherbst]

How about PWModel ... it is a model after al ... that's how we ended up with this name.

Re kpoints: Really ... that can only be leftovers in the tests or so, definitely not intentional

[antoine-levitt]

I think the fact we took to calling it basis is a strong indication it's best thought as a basis than a model. I get your point and I know that's what we settled on, but usually we don't think of discretizing a model as creating a model - it's just an approximation of the model. We could call it PWDiscretization but basis is simpler I think.

RE kpoints, it's mostly in examples and in bz handling

[mfherbst]

Ok, let's go with PWBasis then.

That's all things I did not modify a lot during the last restructuring pass. I tried to make sure it's clear that kcoords is a plain vector and kpt or kpoints KPoint objects, but I might have missed something.