| Filename | Latest commit message | Latest commit date |
|---|---|---|
2eec436850
- RefuseVariantIndexSharing built its holder list per landing record by scanning every landing record, so opening a store with one variant was quadratic. It now groups what lands once and looks each (index, value) up once. Pinned by a count of index-accessor reads at two sizes, not a clock. - JudgeMerge decoded every unseen disk record before Resolve could exit early, so a plain-only store failed a conditional commit on another writer's schema it does not register. It now judges only when the batch, the disk (by the record's variant header, never its payload) or this cache holds a variant. - Plain duplicates find the holder with the smallest key (ordinal), so the choice is the same live, after a flush and after a reopen. The comment claiming "most recently indexed" was wrong across reload. - A dropped slot is reported once per record, not once by the schema diff and again by the variant. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> |
||
| .. | ||
| Fixtures | ||
| AssemblyInfo.cs | ||
| BackingStoreTests.cs | ||
| ConditionalCommitTests.cs | ||
| CultAssetGuidKeyTests.cs | ||
| CultDocumentRegistryTests.cs | ||
| CultDocumentSelectionSurfaceTests.cs | ||
| CultInspectorModelTests.cs | ||
| DirectoryStoreDurabilityTests.cs | ||
| DocumentShapeTests.cs | ||
| GameCult.Caching.Tests.csproj | ||
| PreCut2FixtureDocuments.cs | ||
| PreCut2StoreFormatTests.cs | ||
| README.md | ||
| SerializationOptionsTests.cs | ||
| StoreRoutingTests.cs | ||
| VariantTests.cs | ||
GameCult.Caching.Tests
GameCult.Caching.Tests contains NUnit tests for the caching library and its
MessagePack persistence stores.
Scope
The tests currently cover:
SingleFileMessagePackBackingStoreround-trips, external pulls, and the dirty-pull guardDirectoryMessagePackBackingStorewith thecultcache.store.v4.directory-content-addressed-pagesmanifest: page writes, clean flushes, concurrent writers under the commit lease, external deltas, unflushed local keys, orphaned pages, and observers running after the lease is releasedCultCacheMessagePack.OpenAsync, persisted globals, dirty state, and flush on dispose- cache lookups (
Get,GetAll, runtime-type upserts) CultCache.Commitbatches on a directory store: visibility after the store commits, abandoned stages, sealed batches, observers committing again- Cult document payloads, registry shape rejections, canonical schema fixtures, and schema drift reports
- hand-written MessagePack snapshot and record serialization
StoreRoutingTests: the store-composition contract (routing, attachment is hydration, read-only stores, global singletons, assignable lookups, batches, all-or-nothing loads)ConditionalCommitTests:Expect,ExpectUnchanged,TryCommit, the single-file lock, strictly increasingstoredAt, and the same conditions on a directory storeDirectoryStoreDurabilityTests: failed manifest replace, content-addressed pages, the commit lease, orphan pages and tampered pages, with faults from real file locksPreCut2StoreFormatTests: stores written before Cut 2 still open
Selective hydration, the directory store's stage probes, pre-v4 directory
formats, and the SoA table are no longer in the tree; the SoA table is parked
at parked/cultcache-soa (docs/parked-features.md).
Run
dotnet test tests\GameCult.Caching.Tests\GameCult.Caching.Tests.csproj