[core] Restore behavior with thread=1 setting

[adangel]

Describe the PR

Found this while upgrading the manual integration tests (see pmd/pmd-regression-tester#119): The errors where different when comparing to the baseline - but only the stacktrace. I could see, that the stacktraces are either from the main thread (including PmdCli) or from a worker thread (then they are shorter).
As we didn't specify the thread count for the regression tester baseline, it should be the default, which is "1" - and that should actually run the MultiThreadProcessor. Only with thread count=0 we should use MonoThreadProcessor. See the CLI documentation on threads:

Sets the number of threads used by PMD. Set threads to 0 to disable multi-threading processing.

Keeping to this behavior, it should make the error stacktraces more comparable regardless whether you run it with 4 threads or only with 1.
See also the old impl:

pmd/pmd-core/src/main/java/net/sourceforge/pmd/PmdAnalysis.java

Line 309 in ef34553

return configuration.getThreads() > 0 ? new MultiThreadProcessor(configuration)

Additionally, I took the chance to set the threads now when creating the baseline or running regression tester. Maybe it makes it a bit faster (don't know, how powerful the GitHub action instance runners are).

Ready?

• Added unit tests for fixed bug/feature
• Passing all unit tests
• Complete build ./mvnw clean verify passes (checked automatically by github actions)
• Added (in-code) documentation (if needed)

[pmd-test]

	1 Message
📖	Compared to master: This changeset changes 11 violations, introduces 1 new violations, 1 new errors and 0 new configuration errors, removes 297 violations, 0 errors and 0 configuration errors. Full report

Generated by 🚫 Danger