Support multiple junit-platform.properties

[holly-cummins]

Ingest properties from all junit-platform.properties files on the classpath, rather than just the first. Where the same property exists in multiple files, maintain the current behaviour.

Resolves #2794

This ended up being a fairly straightforward change, because all the testing was already in place, and just needed a slight extension for the new behaviour. What I've done is loaded all properties files with the expected name. I reverse the order so that if the same property is present in multiple files, the first occurrence 'wins'.

I also kept the warning (but with amended wording, of course). If we wanted to be really fancy we could only warn if there actually is a conflict, but I wasn't sure if the extra effort was warranted.

I strengthened the assertions in the test to equals rather than contains, since it's easy to accidentally use similar values in test properties files.

I looked through the docs but couldn't find anywhere where it says you can only have one properties file, so I don't think there's anything to change there.

---

I hereby agree to the terms of the JUnit Contributor License Agreement.

---

Definition of Done

• There are no TODOs left in the code
• Method preconditions are checked and documented in the method's Javadoc
• Coding conventions (e.g. for logging) have been followed
• Change is covered by automated tests including corner cases, errors, and exception handling
• NOT APPLICABLE Public API has Javadoc and @API annotations
• Change is documented in the User Guide and Release Notes

[sormuras]

Hi Holly, I started the CI checks - and expect only the "Cross-Version / OpenJDK 27 (ea)" one to fail (due to #5605).

I note that the underlying issue

• Support multiple junit-platform.properties on the classpath #2794

is still under discussion.

[testlens-app]

✅ All tests passed ✅

🏷️ Commit: ec5e192
▶️ Tests: 27689 executed
⚪️ Checks: 16/16 completed

---

Learn more about TestLens at testlens.app.

[holly-cummins]

Oh dear, it looks like the Leyden job is failing as well (https://github.com/junit-team/junit-framework/actions/runs/23615551087/job/71287894863?pr=5559). It looks like either a hang or excessive slowness. I can't see anything obviously related to my changes in the thread dumps.

(Thanks for kicking the jobs off, @sormuras!)