Bugfix: fix versioning for user pip install

[tomaroberts]

As per #30 and #31, pip install was failing because of issues with dunamai implementation.

In short: when users pip installed nii2dcm it would fail in __init__.py because dunamai was attempting to determine a version based on the assumption that the files underneath are in a Git repo. However, on a user's computer, it was unlikely they were working in a local version of the nii2dcm repo, hence dunamai would fail.

Reading more, the dunamai instructions imply here and in this issue that it is incompatible with Python sdist builds and wheels/bdist should be used.

This PR:

• updates setup.py and the init.py versioning call
• updates the publish GitHub Actions so that only wheels are built

[tomaroberts]

@jcohen02 – I think I've fixed the dunamai issue. I've tested in a fresh venv using this TestPyPI version outside of my local nii2dcm repo folder and it worked for me. Any chance you could test locally? Can add you as PR reviewer even.

I'm going to build a nii2dcm Docker container as well, which ought to help with build testing too.

PS: I always find the TestPyPI install instruction fails, so I tend to use the following: pip install --index-url https://test.pypi.org/simple/ --extra-index-url https://pypi.org/simple/ nii2dcm==0.1.2-post.11

[jcohen02]

Thanks for the update @tomaroberts, happy to review/test.

[github-actions[bot]]

Coverage Report

File	Stmts	Miss	Cover	Missing
nii2dcm
__main__.py	34	34	0%	5–68
dcm.py	93	2	98%	165–166
dcm_writer.py	33	6	82%	89–95
nii.py	35	2	94%	34–35
run.py	30	2	93%	57, 66
utils.py	7	7	0%	8–36
nii2dcm/modules
mr_image.py	59	2	97%	39–40
TOTAL	431	55	87%

Tests	Skipped	Failures	Errors	Time
18	0 :zzz:	0 :x:	0 :fire:	4.987s :stopwatch:

[github-actions[bot]]

Coverage Report

File	Stmts	Miss	Cover	Missing
nii2dcm
__main__.py	34	34	0%	5–68
dcm.py	93	2	98%	165–166
dcm_writer.py	33	6	82%	89–95
nii.py	35	2	94%	34–35
run.py	30	2	93%	57, 66
utils.py	7	7	0%	8–36
nii2dcm/modules
mr_image.py	59	2	97%	39–40
TOTAL	431	55	87%

Tests	Skipped	Failures	Errors	Time
18	0 :zzz:	0 :x:	0 :fire:	4.931s :stopwatch:

[tomaroberts]

@jcohen02 – I've test this quite a bit now and fairly sure it's fixed. Going to do a release soon – if it's still broken somehow I will reinvestigate (!)