Code formatting and performance optimizations

[naftaly]

Summary

• Add .swiftformat configuration file for consistent code style
• Apply code formatting across the codebase
• Run memory and disk cache operations concurrently in forEachResource using async let
• Remove unnecessary async from cost() and costLimit() methods
• Optimize purgeUnowned() to use single-pass instead of multiple array iterations
• Add test for purgeUnowned behavior

Test plan

• Run existing test suite (33 tests passing)
• Verify formatting is consistent
• Verify optimizations don't change behavior

[github-actions]

🚀 Arsenal Cache Performance Benchmarks

Multi-layer caching with LRU eviction and disk persistence

Test Hardware: Apple M1 (Virtual), 7 GB RAM

Memory Cache Operations

Operation	Items	Time	Per-Op	Status
Get	1000	166.0ms	166.0 μs	⚠️ OK
Set	1000	172.0ms	172.0 μs	⚠️ OK
Set (with purge)	500	200.0ms	400.0 μs	⚠️ OK

Disk Cache Operations

Operation	Items	Time	Per-Op	Status
Get	100	154.0ms	1.54 ms	✅ Excellent
Set	100	132.0ms	1.32 ms	✅ Excellent

Combined Cache (Memory + Disk)

Operation	Items	Time	Per-Op	Status
Get (disk→memory promotion)	100	126.0ms	1.26 ms	✅ Excellent
Set (both layers)	100	667.0ms	6.67 ms	✅ Good

Large Items (1 MB each)

Storage	Items	Time	Per-Item	Status
Disk	20	155ms	7.8ms	✅ Excellent
Memory	50	147ms	2.9ms	✅ Excellent

Throughput (Mixed Read/Write)

Operation	Ops	Time	Ops/sec	Status
Memory (33% write, 67% read)	5,000	201ms	24,876	⚠️ OK

Clear Operations

Storage	Items	Time	Status
Disk	100	152ms	✅ Good
Memory	1000	179ms	✅ Good

Performance Characteristics

Status Legend

• ✅ Excellent/Good: Optimal performance
• ⚠️ OK: Acceptable, monitor in production
• ❌ Review: May need optimization

Architecture

• Memory cache: LRU eviction with cost-based limits
• Disk cache: File-based with staleness and cost eviction
• Combined: Automatic disk→memory promotion on read
• Thread safety: @globalActor isolation via ArsenalActor

---

Total benchmarks: 12 passed | Generated 2025-11-30 17:15:28