feat: use ValueError instead of TypeError for failed Operator.operation

[ycexiao]

What problem does this PR address?

Closes #72

Use ValueError instead of TypeError for incorrect calculations.

What should the reviewer(s) do?

Please check the modifications.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 67.05%. Comparing base (b94a3a9) to head (2c448e0).
⚠️ Report is 12 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #129   +/-   ##
=======================================
  Coverage   67.05%   67.05%           
=======================================
  Files          25       25           
  Lines        3160     3160           
=======================================
  Hits         2119     2119           
  Misses       1041     1041           

Files with missing lines	Coverage Δ
tests/test_literals.py	99.17% <100.00%> (ø)
tests/test_visitors.py	99.19% <100.00%> (ø)

🚀 New features to boost your workflow:

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

[ycexiao]

@sbillinge, it's ready for review.

[sbillinge]

please see comments

[sbillinge]

could we have this as an f-string for greater readability?

[sbillinge]

I think this should be TypeError maybe? Turning any exception into a valueError may lead to confusion?

[ycexiao]

Yes. After a second thought, I think maybe we should not modify it at all.

The test-case-bad uses np.add but only gives it one argument. We want to use ValueError instead of TypeError in this bad case, because the input type is correct. But we can't customize all possible Errors that a user might encounter in their practice. I think the best approach is to leave the original error message and let the user see where they are wrong directly.

So, I think it is better to raise whatever the external function raises. In this case, we can close #72 as not planned.

[sbillinge]

I read the original issue and it was that get_value returned TypeError instead of ValueError when it op._value was set to None, so it looks to me that you may be fixing the problem incorrectly? Not sure, but I would expect to see something like

if self._value is None:
    raise ValueError

somewhere inside getValue()?

[ycexiao]

I read the original issue and it was that get_value returned TypeError instead of ValueError when it op._value was set to None

Oh, I see. I thought we need fix both TypeError

[ycexiao]

Closed as the issue will be addressed in another cleaner PR.