test(cpp): ✅ migrate unit tests to Catch2

[robertodr]

Since the unification of Python binding and C++ test compilation #152, the build times of the test step have increased. This is an experiment to check what's the impact of our unit test framework on compilation times in CI.

🤖 AI text below 🤖

Summary

Replace Boost.Test with Catch2 v3 for the C++ unit suite while preserving per-case CTest discovery and lockstep whole-suite MPI execution. This removes the Boost.Test binary dependency without removing Boost production utilities.

Changes

• Port C++ test cases, fixtures, assertions, matchers, and generators to Catch2 v3.
• Use system-first Catch2 discovery with a pinned v3.15.3 CPM fallback.
• Preserve per-case serial CTest labels and whole-suite MPI rank variants, with deterministic ordering across ranks.
• Add Catch2 installation support and remove Boost.Test provisioning from development and CI environments.
• Update contributor and testing documentation.

Testing

• Standard Python/workspace suite: 591 passed.
• Serial C++ suite: 208 passed, 2 expected single-rank skips.
• Two-rank MPI C++ suite: passed.
• Wide TermIndex configuration: 618 Python tests and 209 C++ tests passed, with 2 expected skips.
• Commit hooks: passed.
• Documentation build was not run because npm is unavailable in the development container.

Checklist

• Tests added or updated to cover the changes
• Documentation updated (docstrings, docs/, CONTRIBUTING.md) if needed
• CHANGELOG / release notes updated if applicable (not applicable: test infrastructure only)

AI/LLM disclosure

• I did not use LLM tooling, or used it only privately for ideation
• I used GitHub Copilot to help write this PR description
• I used GitHub Copilot to generate or modify code

Important

By opening this PR I confirm that I have read CONTRIBUTING.md and I agree to the terms of the Contributor License Agreement.

Warning

If you're contributing on behalf of your employer, contact cla@algorithmiq.fi to arrange a Corporate CLA.

[github-actions]

Docs preview: https://pr-264.monoprop-docs.pages.dev

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.70%. Comparing base (2953349) to head (636657b).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #264   +/-   ##
=======================================
  Coverage   97.70%   97.70%           
=======================================
  Files          14       14           
  Lines         742      742           
  Branches       98       98           
=======================================
  Hits          725      725           
  Misses         12       12           
  Partials        5        5           

Flag	Coverage Δ
cpp	97.70% <ø> (ø)

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

[robertodr]

It doesn't seem that using Catch2 vs Boost.Test makes that much of a difference in compile times. I believe this is a necessary evil due to templating in the library. I will try to compile the unit tests themselves as -O1 (instead of -O3) and see if that helps compile times.

[robertodr]

Going to down to -O1 for the compilation of the unit tests (83c8182) cuts about 1 minute from the Install package step. Of course the execution is a bit slower, which more or less compensates.

TBH, I would merge this one. Catch2 is less bloated than Boost.Test as a unit test framework and I found its documentation more readable. Up to the reviewer.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud