Espen
Espen
When testing workchains one need to be able to test different output files based on some configuration in a workchain iteration. With the current `mock-vasp` that is not so easy...
The parser tests right uses fixtures some places and direct module import elsewhere. Make sure this is uniform.
Across the different workchains we reuse a bit of code to set `parser_settings` which should be factored to simplify maintenance.
We should have a feature that lets you remove potentials from the `vasp data vasp-potcar` interface.
We need to refactor code. The `io.vaspun` and `calcs.workchains.converge` are two cases that gets complaints from pylint (too many public methods).
Add a better help text explaining the tags etc. when for instance uploading or exporting potcar data.