Edge cases and error handling for the squeeze morph

[Sparks29032]

Throw error when invalid input is given (e.g. non-numeric list).

Indicate to user that duplicated/repeated commas and trailing commas will be removed: e.g. 0,,1, becomes 0,1.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 99.07%. Comparing base (784b3df) to head (8503cce).
Report is 3 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #227      +/-   ##
==========================================
+ Coverage   99.04%   99.07%   +0.02%     
==========================================
  Files          21       21              
  Lines        1048     1079      +31     
==========================================
+ Hits         1038     1069      +31     
  Misses         10       10              

Files with missing lines	Coverage Δ
tests/test_morphapp.py	100.00% <100.00%> (ø)
tests/test_morphio.py	100.00% <ø> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[sbillinge]

one small question

[sbillinge]

why are we using sys.exit and not a proper exception? In general, this is to be avoided as it bypasses all the nice exception handling you get for free in Python?

[Sparks29032]

This is a test for the CLI, and the intended behavior is that once we receive a ValueError from the 'a' character, the parser raises an error, and the CLI program exits with value 1.

[Sparks29032]

The line parser.error(f"{coeff} could not be converted to float.") (line 549 on 'morphapp.py') is calling the in-built option parser function to handle the error once we detect a ValueError on the Python side.

[Sparks29032]

Hmm, but we should ensure that the test is capturing this particular error, and not necessarily system exiting elsewhere. Let me put a new commit in a bit.

[sbillinge]

I am happy to merge this but it is a bit overly complicated I think. We generally don't test the parser or handle improper inputs beyond what is native to argparse. That's fine though.....