chore: use computIfAbsent in `MarkerManager

[JarvisCraft]

This is a small change which utlizes a more domain-specific and potentially performant computeIfAbsent in MarkerManager.

Checklist

• Base your changes on 2.x branch if you are targeting Log4j 2; use main otherwise
• ./mvnw verify succeeds (if it fails due to code formatting issues reported by Spotless, simply run ./mvnw spotless:apply and retry)
• Non-trivial changes contain an entry file in the src/changelog/.2.x.x directory
• Tests for the changes are provided There are no functional changes
• Commits are signed (optional, but highly recommended)

[github-actions]

Job	Requested goals	Build Tool Version	Build Outcome	Build Scan®
build-macos-latest	clean install	3.9.8	✅
build-ubuntu-latest	clean install	3.9.8	✅
build-windows-latest	clean install	3.9.8	✅

Generated by gradle/develocity-actions

[ppkarwasz]

@JarvisCraft,

LGTM, thanks.

[vy]

@JarvisCraft, thanks so much for the contribution! 🙇 Curious: How did you encounter the problem? Do you observe a performance issue? If so, mind sharing some more details, please?

[JarvisCraft]

@vy, hi! I was just looking through the sources of MarkerManager (interested in its implementation details) and, when looking at this specific method, just saw the common pattern which is almost always replaceable with the computeIfAbsent.

While I did not need any practical performance improvement from it (our use case for Markers is initializing them once into a static final field) this generally should be better from microopt. point of view since Map implementation may (and commonly will) be able to perform single key lookup instead of multiple ones and, what's more important, ConcurrentHashMap explicitly guarantees its atomicity.

But what's also important IMO is that the code is now more explicit now in what it does.