Raise test coverage to 98% and drop dead SaveChangesFailed handling

[GabrielMarquezMatte]

What does this change?

Raises line coverage of the three packages from 87.6% to 98.5% (unit + integration; unit alone 86.4% → 97.0%), mostly in the EF Core invalidation interceptors, which were at 61%. It also removes the SaveChangesFailed handling from QueryCacheInvalidation: it reset state that the next SavingChanges always overwrites, so it had no effect. No public API or behavior change. The PR also carries the v0.1.0 public API promotion (PublicAPI.Unshipped.txt → PublicAPI.Shipped.txt) merged into develop in #1.

Package	Unit	Unit + integration
QueryCache.Core	97.0%	98.8%
QueryCache.Dapper	100%	100%
QueryCache.EFCore	96.0%	97.6%

New tests:

• EF Core: synchronous SaveChanges, synchronous and asynchronous commit and rollback, a failed SaveChanges keeps cached entries, ToDictionaryCachedAsync with a key comparer, FirstCachedAsync with a row, a string filter in the query.
• PostgreSQL: completing a TransactionScope invalidates what another connection cached while the scope was open.
• HybridCache: a synchronous SaveChanges returns only after the invalidation reached the distributed cache, and a value invalidated during its fill is never written there. The fake distributed cache in the tests now writes asynchronously and records its writes, so both can be checked.
• Core and Dapper: lazy (iterator) results, empty and non-empty; QueryKey object equality and ToString; DapperCacheQuery equality, hash code and ToString; DynamicParameters filled with Add.

CA1849 is suppressed in the unit test project only: those tests call SaveChanges, Commit and Rollback synchronously on purpose, to cover the synchronous interceptors.

The 5 lines still uncovered are defensive or depend on a race: null context or missing state in the interceptors, other EF expression node types in TableTags, the LRU update callback, and an invalidation landing exactly during the HybridCache write.

Checklist

• dotnet build QueryCache.slnx --configuration Release builds clean (warnings are errors)
• dotnet test --project tests/QueryCache.Tests/QueryCache.Tests.csproj --configuration Release passes (79 tests)
• Integration tests pass (31 tests: SQL Server, PostgreSQL, Redis)
• Public API unchanged
• Tests added/updated for the behavior change
• One focused change

Test plan

• Ran both suites with coverage and compared line coverage per package against master.
• Broke 7 code paths one at a time: synchronous SavedChanges, synchronous commit, the synchronous wait on HybridCache invalidation, the HybridCache check that skips storing an invalidated value, the TransactionScope handler, and two more. The new tests caught all 7.
• Ran the unit suite 3 times to rule out flakiness.

[codecov-commenter]

Welcome to Codecov 🎉

Once you merge this PR into your default branch, you're all set! Codecov will compare coverage reports and display results in all future pull requests.

Thanks for integrating Codecov - We've got you covered ☂️

[github-actions]

Benchmark Results

No published baseline was found, so these are absolute numbers only. They come from a shared ubuntu-latest runner and cannot be compared against another machine's.

Full results per group

QueryCache.Benchmarks.CacheBenchmarks

BenchmarkDotNet v0.15.8, Linux Ubuntu 24.04.5 LTS (Noble Numbat)
INTEL XEON PLATINUM 8573C 2.30GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 10.0.401
  [Host]     : .NET 10.0.12 (10.0.12, 10.0.1226.42308), X64 RyuJIT x86-64-v4
  Job-MEHJPP : .NET 10.0.12 (10.0.12, 10.0.1226.42308), X64 RyuJIT x86-64-v4

IterationCount=5  WarmupCount=1  

Method	Categories	Rows	Mean	Error	StdDev	Ratio	RatioSD	Gen0	Allocated	Alloc Ratio
Dapper_Direct	Dapper	10	14,744.5 ns	259.69 ns	67.44 ns	1.00	0.01	0.0305	3312 B	1.00
Dapper_CacheHit	Dapper	10	398.5 ns	10.66 ns	1.65 ns	0.03	0.00	0.0081	688 B	0.21
Dapper_CacheMiss	Dapper	10	16,687.0 ns	315.50 ns	81.93 ns	1.13	0.01	0.0305	5008 B	1.51
Dapper_Direct	Dapper	1000	853,711.9 ns	5,976.73 ns	924.91 ns	1.000	0.00	0.9766	138384 B	1.000
Dapper_CacheHit	Dapper	1000	400.0 ns	4.53 ns	0.70 ns	0.000	0.00	0.0081	688 B	0.005
Dapper_CacheMiss	Dapper	1000	866,642.5 ns	12,114.07 ns	3,145.98 ns	1.015	0.00	0.9766	140080 B	1.012
Ef_Direct	EF	10	25,399.3 ns	15,304.11 ns	2,368.33 ns	1.01	0.11	-	9536 B	1.00
Ef_CacheHit	EF	10	6,307.1 ns	111.79 ns	17.30 ns	0.25	0.02	0.0534	4872 B	0.51
Ef_CacheMiss	EF	10	52,799.9 ns	46,830.11 ns	12,161.63 ns	2.09	0.47	0.2441	26867 B	2.82
Ef_Direct	EF	1000	853,436.3 ns	29,632.65 ns	7,695.50 ns	1.000	0.01	1.9531	287168 B	1.00
Ef_CacheHit	EF	1000	6,312.4 ns	84.35 ns	21.91 ns	0.007	0.00	0.0534	4872 B	0.02
Ef_CacheMiss	EF	1000	882,220.1 ns	26,817.98 ns	4,150.11 ns	1.034	0.01	1.9531	304523 B	1.06

QueryCache.Benchmarks.EfMissBreakdownBenchmarks

BenchmarkDotNet v0.15.8, Linux Ubuntu 24.04.5 LTS (Noble Numbat)
Intel Xeon 6973P-C 2.60GHz, 1 CPU, 4 logical and 2 physical cores
.NET SDK 10.0.401
  [Host]     : .NET 10.0.12 (10.0.12, 10.0.1226.42308), X64 RyuJIT x86-64-v4
  Job-MEHJPP : .NET 10.0.12 (10.0.12, 10.0.1226.42308), X64 RyuJIT x86-64-v4

IterationCount=5  WarmupCount=1  

Method	Mean	Error	StdDev	Ratio	RatioSD	Gen0	Allocated	Alloc Ratio
Direct	15.917 μs	0.1772 μs	0.0274 μs	1.00	0.00	0.0610	9.31 KB	1.00
CommandAndKey	4.169 μs	0.0172 μs	0.0045 μs	0.26	0.00	0.0534	4.6 KB	0.49
TableTagsOnly	1.989 μs	0.0329 μs	0.0085 μs	0.12	0.00	0.0420	3.48 KB	0.37
InvalidateOnly	5.646 μs	0.0218 μs	0.0057 μs	0.35	0.00	0.0916	7.8 KB	0.84
StoreMissAroundDirect	17.040 μs	0.3276 μs	0.0507 μs	1.07	0.00	0.0610	9.93 KB	1.07
DirectAfterCommand	23.002 μs	15.2394 μs	3.9576 μs	1.45	0.23	0.1221	13.41 KB	1.44
DirectVarying	28.295 μs	33.2187 μs	5.1406 μs	1.78	0.29	0.1221	13.22 KB	1.42
MissVarying	43.160 μs	91.4899 μs	14.1582 μs	2.71	0.80	0.2441	23.29 KB	2.50
FullMiss	42.547 μs	51.7588 μs	13.4416 μs	2.67	0.77	0.2441	26.32 KB	2.83