ENH: Cubic Crystal Dislocation Quadrupoles

[thomas-rocke]

No description provided.

[thomas-rocke]

Overhaul of the test_dislocation.py framework for testing the CubicCrystalDislocation and CubicCrystalDissociatedDislocation classes.

• pytest-based class solution
• Multiple tests per dislocation type, and use of subtests means maximum coverage, and minimum skipped tests.
• Per-dislocation tests achieved using pytest.mark.parametrize
• On-the-fly discovery of dislocations for parametrize -> automatic coverage of new dislocation types
• Full benchmarking of all displacement methods, for all dislocation types, automatically.

[jameskermode]

Looking nearly good to go! Can you please add 1-2 bullet points to ChangeLog.md?

[thomas-rocke]

Changelog modified, so should be good to go, pending tests passing.

Bit of a temp fix for view_quad - the dislocation lines are incorrectly placed (because the cell makes the coordinate system for the ngl objects a bit messy), so I've disabled them for now. Ideally this should be fixed and the lines added back in, but it's not impactful enough to warrant a massive delay on the PR. Will try to figure this out in the coming weeks.

[thomas-rocke]

Output of the pytest parametrized tests is not ideal for all test_* files. Will look for a quick fix, or revert my change to conftest.py