BUG: BandStructure not in data store anymore

[JaGeo]

@utf I think the bug described in #572 is not related to the workflow being dynamic. I just ran the tutorial example with fewer kpoints

from atomate2.vasp.flows.core import RelaxBandStructureMaker
from jobflow import run_locally
from pymatgen.core import Structure


mgo_structure = Structure(
    lattice=[[0, 2.13, 2.13], [2.13, 0, 2.13], [2.13, 2.13, 0]],
    species=["Mg", "O"],
    coords=[[0, 0, 0], [0.5, 0.5, 0.5]],
)

bandstructure_flow = RelaxBandStructureMaker().make(mgo_structure)

# run the job
run_locally(bandstructure_flow, create_folders=True)

And the band structure ends up in the MongoDB:

But again, not the DOS:

I actually don't have any idea where this could come from but it's not a Lobster-workflow related issue alone.

[JaGeo]

Could someone else test this independently from me?

[utf]

This is so strange and I don't understand why this is happening. It should work to either specify an MSONable class OR a dictionary key. That said, I'm happy to merge the change in #605 while we investigate further.

[JaGeo]

Thanks. I just fear that some of the other transfers to the data store might also be broken. I, however, haven't tested all yet.

[utf]

Ah, I figured it out! BandStructure is not MSONable for some reason.... https://github.com/materialsproject/pymatgen/blob/535f9252c15ea112969f2dba431442027a8a7e14/pymatgen/electronic_structure/bandstructure.py#L134

But BandStructureSymmLine is: https://github.com/materialsproject/pymatgen/blob/535f9252c15ea112969f2dba431442027a8a7e14/pymatgen/electronic_structure/bandstructure.py#L676

I think the proper fix is to make BandStructure MSONable but lets get this merged for now.