refactor!: Cache instead of Arc<Cache> in public APIs

[m4tx]

Similarly to #432, this drops the Arc from Arc<Cache> in public APIs to make the interfaces more ergonomic and easier to use for the users.

[Copilot]

Pull request overview

This PR refactors the caching API to improve ergonomics by removing the need for users to wrap Cache in Arc. The Cache struct now internally wraps its implementation in Arc<CacheImpl>, making it directly cloneable while maintaining the same functionality.

Key changes:

• Modified Cache to contain an Arc<CacheImpl> wrapper around the store implementation
• Updated all public APIs that previously used Arc<Cache> to use Cache directly
• Changed internal field accesses to go through self.inner

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
cot/src/cache.rs	Refactored Cache struct to wrap CacheImpl in Arc internally, enabling direct cloning without requiring external Arc wrapping
cot/src/test.rs	Updated TestRequestBuilder and TestCache to use Cache instead of Arc<Cache> in field types and return values
cot/src/project.rs	Changed function signatures, type aliases, and method return types throughout to use Cache directly instead of Arc<Cache>

The refactoring is well-executed and maintains backward compatibility in functionality while improving the API ergonomics. All method implementations correctly access the inner fields through self.inner, and the change is applied consistently across all affected files.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[github-actions]

Bencher Report

Branch	arc-cache
Testbed	github-ubuntu-latest

Click to view all benchmark results

Benchmark	Latency	Benchmark Result microseconds (µs) (Result Δ%)	Upper Boundary microseconds (µs) (Limit %)
empty_router/empty_router	📈 view plot 🚷 view threshold	6,178.80 µs (+2.02%) Baseline: 6,056.57 µs	6,805.38 µs (90.79%)
json_api/json_api	📈 view plot 🚷 view threshold	1,058.10 µs (+3.36%) Baseline: 1,023.70 µs	1,132.61 µs (93.42%)
nested_routers/nested_routers	📈 view plot 🚷 view threshold	951.87 µs (+0.55%) Baseline: 946.70 µs	1,043.58 µs (91.21%)
single_root_route/single_root_route	📈 view plot 🚷 view threshold	910.80 µs (+0.52%) Baseline: 906.13 µs	996.11 µs (91.44%)
single_root_route_burst/single_root_route_burst	📈 view plot 🚷 view threshold	16,620.00 µs (-6.13%) Baseline: 17,705.59 µs	20,944.96 µs (79.35%)

🐰 View full continuous benchmarking report in Bencher

[ElijahAhianyo]

Curious about this. What's the benefit of doing it this way(boxing the store in another struct and wrapping it in an Arc)?