Integrate with AtomsBase 0.3

[mfherbst]

Closes #26, additionally:

• Integrates with arbitrary property interface from AtomsBase 0.3
• Restructures tests
• Updates CI workflow
• Updated the readme
• Bumps version to 0.1.10

[jameskermode]

This looks good, thank you!

My one reservation is that I don't like the use of hard-coded lists of properties and the consequent fixed mapping of particular string names to units. This is OK for the truly mandatory ExtXYZ properties (basically just lattice, positions, chemical species) and by extension to the other 'standard' AtomsBase properties but I'm not sure how to deal with anything beyond this.

I suppose just leaving them unitless as you have done is the only thing we can do.

[mfherbst]

If you confirm u"eV^0.5/u^0.5" in the other PR, then I'll use that here as well.

My one reservation is that I don't like the use of hard-coded lists of properties

yeah this is tricky. My reasoning is that: If there is a preferred name that has a well-adjusted meaning, then it ought to have a well-defined unit as well. So you are bound to check this explicitly. In the long run (when the list gets longer) we could provide proper constants in AtomsBase that essentially export both the system / atom keywords along with the associated unit. Then one does not have to hard-code this in here (and everywhere else).

[mfherbst]

@jameskermode I changed the units as discussed in the other PR. Let me know if you have other things to be done before merging this.

[jameskermode]

Happy to merge now. Thanks for the contribution!

[jameskermode]

Release 0.1.10 tagged and will show up in General registry shortly.

[mfherbst]

Thanks!