Switch to TestItems.jl

[icweaver]

Hey all, I've been noodling around with different test frameworks lately and have been liking TestItems.jl. Its filtering capabilities are pretty convenient, and I've found myself reaching for it when wanting to quickly test something in other PRs, e.g.,

julia> using TestItemRunner

julia> @run_package_tests filter = ti -> occursin( "aperture/circular", ti.name) verbose=true;
Test Summary:                            | Pass  Total  Time
Package                                  |   21     21  0.0s
  Photometry.jl/src/aperture/circular.jl |   21     21  0.0s
    aperture/circular: Apertures         |   13     13  0.0s
    aperture/circular: broadcasting      |    3      3  0.0s
    aperture/circular: CircularAperture  |    3      3  0.0s
    aperture/circular: CircularAnnulus   |    2      2  0.0s

I also like how they take a page out of Rust's book and allow for source code and tests to be located in the same file, but I know that's more a matter of personal taste

I'm not a VS Code user, but I've heard it's pretty nice for that too, which could potentially help to further reduce contributor friction

At any rate, I was wondering what folks thought about transitioning to this framework here, and if there are any trade-offs to consider

Thanks!

[codecov]

Codecov Report

❌ Patch coverage is 93.43434% with 13 lines in your changes missing coverage. Please review.
✅ Project coverage is 94.01%. Comparing base (fcc2e10) to head (0461fc5).
⚠️ Report is 21 commits behind head on main.

Files with missing lines	Patch %	Lines
src/aperture/Aperture.jl	93.97%	5 Missing ⚠️
src/aperture/overlap.jl	95.91%	2 Missing ⚠️
src/background/estimators.jl	91.30%	2 Missing ⚠️
src/detection/Detection.jl	88.23%	2 Missing ⚠️
src/aperture/plotting.jl	94.11%	1 Missing ⚠️
src/background/Background.jl	88.88%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #77      +/-   ##
==========================================
- Coverage   94.70%   94.01%   -0.70%     
==========================================
  Files          11       11              
  Lines         737      935     +198     
==========================================
+ Hits          698      879     +181     
- Misses         39       56      +17     

Flag	Coverage Δ
unittests	94.01% <93.43%> (-0.70%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[icweaver]

Ugh, Windows. Will take a look

Update: Welp, looks to be an inference issue specific to Julia v1.6. Side request, can we start dropping support for old Julia versions and swap to LTS (v1.10) instead 😅

I am curious though why this issue only shows itself when switching to this testing framework

[icweaver]

Fixed a type instability I accidentally introduced. Seemed to do the trick (update: plus bumping oldest supported version to LTS).

Looks like the last thing is something like this for nightly julia-vscode/TestItemRunner.jl#114. I don't think this should hold up the PR, especially if we end up going with something like JuliaAstro/JuliaAstro.github.io#157, so will go ahead merge after other tests complete

[icweaver]

codecov decrease looks to be mostly noise, will go ahead and merge