refactor(event cache): allow deref state lock read/write guards to the state

[bnjbvr]

This makes it possible to common out the implementations of functions that should be available to both kinds.

Also, moves a bit of code that could easily live in the RoomEventCacheState impl block, instead of being free functions.

Only code motion, no changes in functionality.

• I've documented the public API Changes in the appropriate CHANGELOG.md files.
• This PR was made with the help of AI. (is rust-analyzer AI?)

[codecov]

Codecov Report

❌ Patch coverage is 83.13253% with 14 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.81%. Comparing base (81bed18) to head (4fd97b4).
⚠️ Report is 9 commits behind head on main.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
...es/matrix-sdk/src/event_cache/caches/room/state.rs	85.13%	5 Missing and 6 partials ⚠️
crates/matrix-sdk/src/event_cache/caches/lock.rs	66.66%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6314      +/-   ##
==========================================
- Coverage   89.83%   89.81%   -0.03%     
==========================================
  Files         376      376              
  Lines      102469   102455      -14     
  Branches   102469   102455      -14     
==========================================
- Hits        92058    92017      -41     
- Misses       6840     6856      +16     
- Partials     3571     3582      +11     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[codspeed-hq]

Merging this PR will improve performance by 98.51%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
✅ 49 untouched benchmarks

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	Restore session [memory store]	222.4 ms	112.1 ms	+98.51%

---

_{Comparing bnjbvr/refactor-event-cache-lock-guards (4fd97b4) with main (81bed18)}

[Hywan]

Looks sensible to me. Left a comment about the Deref (extended to DerefMut) implementations, not a blocker.

[Hywan]

Since state is pub, is it really useful to have a Deref implementation here? My fear is that it can create a bit of confusion when reading the code (despite being simpler in appearance).

[bnjbvr]

"jump to definition" and every other actions seem to work fine with Deref/DerefMut. Not having the Deref impl would mean that every current caller of state_guard.room_linked_chunk() would need to be replaced with state_guard.state.room_linked_chunk().

As a thought experiment, I've tried to remove the Deref impls, and I get around 40 compilation errors, which shows that the Deref is then used in multiple places. Since Deref is rather idiomatic in Rust, especially expected for a lock guard (the stdlib has such a Deref impl for its own lock guards), and there are many uses of this, with the related refactors, I'll keep it for now. We can definitely revisit this later, if needs be.