RF: Cleanup Makefile to bring parity with CI, remove tox + nose references

[jacobdr]

Closes #1104

• Align on venv as virtual environment convention
• Update makefile test target to re-use same script used in CI to orchestrate testing
• Remove refererences to tox and nose
• Fix dev-requirements.txt and doc-requirements.txt to pin to versions that are non-breaking

[pep8speaks]

Hello @jacobdr, Thank you for updating!

Cheers! There are no style issues detected in this Pull Request. :beers: To test for issues locally, pip install flake8 and then run flake8 nibabel.

Comment last updated at 2022-05-07 19:05:52 UTC

[codecov[bot]]

Codecov Report

Merging #1105 (ad2cc68) into master (7cfaebf) will not change coverage. The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #1105   +/-   ##
=======================================
  Coverage   92.29%   92.29%           
=======================================
  Files         100      100           
  Lines       12247    12247           
  Branches     2393     2393           
=======================================
  Hits        11303    11303           
  Misses        621      621           
  Partials      323      323           

---

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 7cfaebf...ad2cc68. Read the comment docs.

[jacobdr]

@effigies all the pre-release tests failed for the same reason -- a string representation in a doctest for netcdf_file. I don't immediately see how my changes would have introduced this. Looking at this now, but any thoughts as to root cause?

[jacobdr]

Trying to narrow differences... since mmap is coming from the standard lib I suspected a change in Python versions and how repr(mmap_obj) or str(mmap_obj) is displayed... Its possible, though still smells to me so needs a sanity check.

As evidence, though, comparing against a prior (chose #1096 as the most recent unmerged green build I saw....)

Python version comparison

• 3.10.4 this MR vs 3.10.2 prior success
• 3.9.12 this MR vs 3.9.10 prior success
• the scheduled pre-release cron job against master is failing for the same reason.

Final note: the only pre-release job that passes is the one that does not have the PRE_PIP_FLAGS environment variable set....