chore(deps): Bump com.cedarsoftware:java-util from 4.80.0 to 4.81.0

[dependabot]

Bumps com.cedarsoftware:java-util from 4.80.0 to 4.81.0.

Changelog

Sourced from com.cedarsoftware:java-util's changelog.

4.81.0 - 2025-01-10

• ARCHITECTURE: CompactMap - Replaced runtime Java compilation with pre-compiled bytecode template
  • No longer requires JDK: The builder API (CompactMap.builder().build()) now works on JRE, not just JDK
  • How it works: A pre-compiled bytecode template is patched at runtime with the configuration hash, then static fields are injected with configuration values (case sensitivity, compact size, ordering, etc.)
  • Performance: Eliminates JavaCompiler overhead - template classes are created via MethodHandles.Lookup.defineClass() with simple byte array patching
  • Same functionality: All existing builder options work exactly as before (case sensitivity, ordering, compact size, custom map types, etc.)
  • Backward compatible: Subclassing CompactMap continues to work unchanged
  • Disabled obsolete CompactMapMethodsTest.testGetJavaFileForOutputAndOpenOutputStream (tested old JavaFileManager approach)
• BUG FIX: Converter - Fixed char[]/byte[] cross-conversion returning null after isConversionSupportedFor() was called
  • Root cause: VoidConversions::toNull was used as a placeholder in CONVERSION_DB to "advertise" that char[] ↔ byte[] conversions are supported
  • When isConversionSupportedFor(char[].class, byte[].class) was called, this placeholder got cached
  • Subsequent convert(char[], byte[].class) found the cached placeholder and used it, returning null instead of performing actual conversion
  • Fix: Replaced placeholder entries with actual converters that call ArrayConversions.arrayToArray()
  • Added tests: testIsConversionSupportedForDoesNotBreakConvert, testArrayCrossConversionWithAndWithoutSupportCheck
• BUG FIX: AbstractConcurrentNullSafeMap.computeIfAbsent() - Fixed contention causing hangs under concurrent load
  • Affects: ConcurrentHashMapNullSafe and ConcurrentNavigableMapNullSafe (both extend AbstractConcurrentNullSafeMap)
  • Root cause: Used ConcurrentHashMap.compute() which holds a lock on the hash bin for the entire duration of the mapping function execution. Under high concurrency, threads pile up waiting for bin locks, causing hangs.
  • Key insight: The common case (no null values stored) doesn't need special sentinel handling during computation
  • Fix: Fast path now delegates directly to ConcurrentHashMap.computeIfAbsent() - this is lock-free for cache hits and only briefly locks for insertions. The slow path (key mapped to null, which is rare) uses optimistic locking with putIfAbsent()/replace().
  • Result: Performance is now virtually identical to plain ConcurrentHashMap for typical use cases (caching, memoization, etc.)
• TEST FIX: LRUCacheTest - Added proper cleanup to prevent OutOfMemoryError during deployment tests
  • The testSpeed and testCacheBlast tests create 10M-entry caches (~800MB each)
  • Added @AfterEach cleanup that clears and nullifies caches, then suggests GC
  • Added explicit cache.clear() at the end of large tests to free memory before next test
• IMPROVED: Scheduler lifecycle management for ThreadedLRUCacheStrategy and TTLCache
  • Both classes now have proper shutdown methods that await termination (up to 5 seconds graceful, then 1 second forced)
  • Schedulers are now lazily recreatable - if shut down, creating new cache instances automatically restarts them
  • ThreadedLRUCacheStrategy.shutdownScheduler() - new static method for explicit cleanup
  • TTLCache.shutdown() - now returns boolean indicating clean termination, properly awaits termination
  • All schedulers use daemon threads, so they won't prevent JVM shutdown even without explicit cleanup
• IMPROVED: ThreadedLRUCacheStrategy - Complete algorithm redesign with zone-based eviction and sample-15 approximate LRU
  • Memory guarantee: Cache never exceeds 2x capacity, allowing predictable worst-case memory sizing
  • Four-zone eviction strategy:
    • Zone A (0 to 1x capacity): Normal operation, no eviction
    • Zone B (1x to 1.5x capacity): Background cleanup only (every 500ms)
    • Zone C (1.5x to 2x capacity): Probabilistic inline eviction (0% at 1.5x → 100% at 2x)
    • Zone D (2x+ capacity): Hard cap with evict-before-insert
  • Sample-15 eviction: Instead of sorting all entries O(n log n), samples 15 entries and evicts the oldest. Based on Redis research, this provides ~99% LRU accuracy with O(1) cost per eviction.
  • No more sorting: Eliminated all O(n log n) sort operations - entire algorithm is now O(1) for inline operations
  • Concurrent race fix: Zone D eviction now uses a while-loop to guarantee hard cap enforcement. Previously, multiple threads could all check size < hardCap, then all insert, causing unbounded growth (check-then-act race condition). Now inserts first, then loops eviction until under hard cap.
  • Fixes deployment hang: Heavy write load (500k unique keys into 5k cache) no longer causes hangs due to expensive sorting or unbounded cache growth

Commits

• 56134d2 ThreadedLRUCacheStrategy: Zone-based eviction with sample-15 approximate LRU
• c612d51 Docs: Fix Javadoc to use {@​code } blocks instead of HTML entities
• 8b0a2f2 Remove JDK requirement warnings from CompactMap/CompactSet docs
• acc51a0 Minor formatting cleanup in CompactMap.injectStaticFields()
• 3ce3502 Refactor CompactMap to use ByteUtilities and ReflectionUtils
• 08b16bd Simplify CompactSet to always use CompactMap.builder()
• a0f004c Replace CompactMap runtime compilation with bytecode template manipulation
• 5d582d2 updated date on changelog.md
• b3e7344 update jar size info
• 267b2c6 update jar size
• See full diff in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot merge will merge this PR after your CI passes on it
• @dependabot squash and merge will squash and merge this PR after your CI passes on it
• @dependabot cancel merge will cancel a previously requested merge and block automerging
• @dependabot reopen will reopen this PR if it is closed
• @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

[github-actions]

🌟 Thank you for your contribution to the Apache Camel project! 🌟

🤖 CI automation will test this PR automatically.

🐫 Apache Camel Committers, please review the following items:

• First-time contributors require MANUAL approval for the GitHub Actions to run

• You can use the command /component-test (camel-)component-name1 (camel-)component-name2.. to request a test from the test bot.

• You can label PRs using build-all, build-dependents, skip-tests and test-dependents to fine-tune the checks executed by this PR.

• Build and test logs are available in the Summary page. Only Apache Camel committers have access to the summary.

• ⚠️ Be careful when sharing logs. Review their contents before sharing them publicly.