Remove support for Ruby 3.1 and aligned JRuby 9.4

[faisal]

Per https://endoflife.date/ruby, Ruby 3.1 is EOL. This PR prepares for when we want to drop support.

Note: this PR changes the JRuby CI tests from 9.4 (Ruby 3.1) to 10 (Ruby 3.4), which Rubycritic does not yet fully support.

• rubycritic.gemspec: update minimum Ruby version to 3.2, and improve support for JRuby.
• .rubocop.yml: update target Ruby version to 3.2
• CI: drop Ruby 3.1, change jruby from 9.4 (Ruby 3.1) to 10 (Ruby 3.4), and set continue-on-error for steps running on all experimental rubies (jruby and ruby-head)

Check list:

• Add an entry to the changelog (updating an existing line_
• Squash all commits into a single one
• Describe your PR, link issues, etc.

[faisal]

The JRuby failures make me think that in practice this PR needs to land on top of #508.

[faisal]

bundle exec rake features and bundle exec rake test now pass for me cleanly on jruby on my local machine.

[faisal]

Note that this includes some of the JRuby support fixes from #508.

[faisal]

This could probably use more review than "we finally got CI to pass."

[faisal]

This patch only differs from the prior attempt in that the CHANGELOG was updated. This failure is the same failure as from two runs ago. I do not see it on my machine. I could use help investigating.

[faisal]

I can now intermittently reproduce this failure locally.

[faisal]

This drops Ruby 3.1 support, which means dropping JRuby 9 support, which means using JRuby 10, on which Rubycritic doesn't quite work yet. Is that an acceptable tradeoff?

[JuanVqz]

This drops Ruby 3.1 support, which means dropping JRuby 9 support, which means using JRuby 10, on which Rubycritic doesn't quite work yet. Is that an acceptable tradeoff?

I didn't see this PR before; even though Ruby 3.1 is EOL, IMO, it is fine keeping support on it since we want/need to support as many Ruby versions as we can (as long as it is not torture), If it is hard to make their tests pass we can move it to experimental in CI.

But I would love to know what @etagwerker thinks about it.

[faisal]

This drops Ruby 3.1 support, which means dropping JRuby 9 support, which means using JRuby 10, on which Rubycritic doesn't quite work yet. Is that an acceptable tradeoff?

I didn't see this PR before; even though Ruby 3.1 is EOL, IMO, it is fine keeping support on it since we want/need to support as many Ruby versions as we can (as long as it is not torture), If it is hard to make their tests pass we can move it to experimental in CI.

But I would love to know what @etagwerker thinks about it.

I don't have a particular urgency to kill 3.1 support, but am trying to keep this PR current so we have the option when we decide to drop the support.

Also see #482 for additional discussion of when we should be in the habit of doing such things.