Nastran OP2 and BDF

I’m trying to work with nx nastran bdf and op2 files using julia. There’s a library in python called pyNastran that can parse these nastran files really well and save them as hdf5 files. My plan is to read these hdf5 files into julia and do the number crunching from there. Is this the best way to work with these files? I only need to parse the op2 and bdfs once and then i can work with the hdf5 files for the rest of the work so python will only be used initially. Is using pycall better than having a separate python file for this case? Is there another route I should try instead?

(These files range from 1-50gbs so I want to be as efficient as possible)

Have you seen: GitHub - mgcth/NastranResults.jl: Nastran OP4 and PCH loader.

This may be a good starting point for solution in Julia.

If imagine either python for pre-processing or PyCall/PythonCall approaches should work similarly.