Add support for verbosity keyword

[jthorton]

Description

This PR implements and fixes #73 by adding support for a verbosity keyword which when set to muted will stop temporary files from being made which causes a failure when running many parallel xtb calculations.

Changelog description

The harness now accepts a verbosity keyword which can be a string or int corresponding to the level of verbosity, the accepted strings and associated ints are full=2, minimal=1, muted=0.

Status

Ready to go

[awvwgk]

Thanks for sharing. I left a few comments.

[awvwgk]

Maybe it is worth to add the string keywords directly in the interface layer at

xtb-python/xtb/interface.py

Lines 219 to 221 in d72c2ec

	def set_verbosity(self, verbosity: int) -> None:
	"""Set verbosity of calculation output"""
	_lib.xtb_setVerbosity(self._env, verbosity)

This would keep the harness clean from this logic and allow to reuse it easily in the ASE calculator as well as any other dependent interface.

[jthorton]

Thats a great idea would you want to support ints still or only the valid strings I was thinking something like this to support both and we can return the int value the verbosity is set to?

def set_verbosity(self, verbosity: Union[Literal["full", "minimal", "muted"], int]) ->int:

[awvwgk]

For backwards compatibility it would be preferable to still have the int support.

[codecov]

Codecov Report

Merging #74 (7fa18d1) into main (0a6fd5c) will increase coverage by 0.05%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main      #74      +/-   ##
==========================================
+ Coverage   97.04%   97.10%   +0.05%     
==========================================
  Files           8        8              
  Lines         440      449       +9     
==========================================
+ Hits          427      436       +9     
  Misses         13       13              

Impacted Files	Coverage Δ
xtb/interface.py	97.36% <100.00%> (+0.04%)	⬆️
xtb/qcschema/harness.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0a6fd5c...7fa18d1. Read the comment docs.

[jthorton]

@awvwgk I think this is now ready apart from the mypy warnings, any ideas how I should get around these?

[awvwgk]

Seems like a false positive to me, we should be able to skip the statement with a # type: ignore comment.