Commit graph CultLib/tests/GameCult.Caching.Tests
Author SHA1 Message Date
Meta Vens
58712e3992 Payload depth tests: the largest fixmap and a container header cut short at the limit
Stryker survivors of the first scoped run: the fixmap boundary code and the truncated
header return. All commits build.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 08:58:50 +02:00
Meta Vens
0789ec8327 Payload depth walk: tests of every header and scalar width at the rule's own seam
Each case is a two-slot document whose chain puts the payload at the limit (accepted)
or one past (refused), so a scalar or container header read at the wrong width flips
one of the two. Also pins the message, truncated payloads and the first-value rule.
All commits build.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 08:45:00 +02:00
Meta Vens
abf624ab04 C# reference: one payload nesting bound (256) on read and write, nil arm slots refused, bounds vectors
CultPayloadDepth.Check walks MessagePack headers iteratively and refuses the first
container past CultMessagePackSecurity.MaximumPayloadDepth, read or skipped, called by
SerializeUntyped (on the bytes it produced) and DeserializeUntyped (before decoding).
UntrustedData's 500 is replaced by the same value as a backstop. CultUnionGuard refuses
[k, nil] arms. New vectors in tests/vectors/union-parity, with bounds-outcomes.msgpack
as their data; the gate now runs every Check-calling test body and fingerprints bytes
and write times. All commits build.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 08:33:15 +02:00
Meta Vens
b4edbb1eeb Union golden: the store's clock is an internal seam, the writer pins it, the gate test runs under any switch
CultCache mints storedAt from an internal Clock (default the wall clock); the vector writer pins it, so
union-surface.cc is regenerated in process and byte-compared on every run. The gate test sets the switch
itself and writes only a probe, so it neither skips nor touches a committed vector.

This commit does not pass: union-surface.cc still holds the old wall-clock stored-at; the next commit
regenerates it. Builds.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 02:58:54 +02:00
Meta Vens
3c75cae4f0 Union parity vectors: the recursive-union fixture, refusal and extra-slot cases, written only under CULTCACHE_WRITE_VECTORS=1
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 00:26:15 +02:00
Meta Vens
a0e0cceaaa Union guard tests: pin the Int32 bounds and full-width encodings of an arm key
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-11 00:07:37 +02:00
Meta Vens
3042757701 Union guard: name an out-of-range or non-integer arm key, pin interface unions, name the schema on a variant refusal, changelog the exception type
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 23:48:49 +02:00
Meta Vens
506ec8c540 Union guard: the store refusal is readable from its outermost message; a consumer formatter for the union is wrapped, not replaced
Manual mutants survived: the record key and schema id were only asserted through the
cause chain, and an inner built from StandardResolver alone passed every test.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 22:56:53 +02:00
Meta Vens
62a0e9ac98 Union guard tests: pin the composite's resolver order and that a union owns its keys
Stryker survivors on the Compose line and on the inherit argument.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 22:41:04 +02:00
Meta Vens
28d79b190a Union guard tests: forge a directory-store page the way its manifest commits to it
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 21:50:53 +02:00
Meta Vens
29b21a0f33 Union guard tests: the store refusal names the record's schema id
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 21:35:37 +02:00
Meta Vens
afaae23c8a Union guard tests: helper names no longer collide with the arm classes
6254da0c does not build the tests.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 21:34:56 +02:00
Meta Vens
6254da0c88 CultCache: refuse an unknown union arm on read and an unnamed subtype on write
A guard ahead of MessagePack-CSharp's union formatter decides both; the store
decode site names the record and schema.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-10 21:17:56 +02:00
Meta Vens
c996ae7892 Caching: a single-file refusal echoes a header only in its known shape
RequireSingleFileFormat echoed any header string. It now names one only in
the shape cultcache.store.v<digits> and otherwise gives its byte length.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-30 18:35:20 +02:00
Meta Vens
2d7a3c78eb Caching tests: an empty store file is refused; a missing file or [0x90] is empty
Pins the C# side of the cross-runtime rule the Rust, Python and TS runtimes
now share. C# already refused a zero-byte file (the snapshot reader runs out
of bytes at the array header); this test keeps it that way. Its exception is
still EndOfStreamException, an IOException: that overlap belongs to R3's C#
work.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-30 17:05:15 +02:00
Meta Vens
486244b60f Caching: JSON converter and MessagePack formatter for CultPhasor
The Phacelle merge added the public CultMath value type CultPhasor without
its Caching adapters, failing the every-public-type tests. Register it the
way CultCellular is: {"cos","sin"} objects in JSON (which also gives the
dictionary-key form) and a [cos, sin] array in MessagePack. The adapters are
Caching code and stay MIT; only CultPhasor.cs itself is MPL-2.0.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-09-30 02:46:29 +02:00
Meta Vens
592a2d49d4 Journal registration counts as disposed only once removal succeeds; a throwing hold body surfaces journal failures
AddJournal's Disposable.Create counted a refused dispose (from inside a journal
or another cache's hold) as done, leaving the journal registered forever. A
private registration now flips only after the removal succeeded. Held's catch
path dropped the hold's journal failures; it now throws an AggregateException
with the body's exception first. The store-composition contract states exactly
when the second getter read runs on a conditional commit. Adds the direct test
that a load is journaled.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 20:35:50 +02:00
Meta Vens
36acb4295a Reach the empty admission through a commit that removes nothing
Attaching a store with no file never calls Loaded, so it did not exercise the
rule; a commit removing a key the cache does not hold admits nothing.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 20:04:59 +02:00
Meta Vens
281e1b238c Keep the test frame from holding the subscription it checks the cache released
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 20:03:08 +02:00
Meta Vens
5b3885a1e8 Detect an unsubscribed observer by what the cache still holds, and pin the empty admission
A disposed R3 subscription ignores OnNext, so a no-op removal from the cache's
observer array cannot be seen by what a receiver gets; the test now checks the
cache stops holding the subscriber. An admission that changed nothing (a store
attached with no records) is not journaled.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 19:47:38 +02:00
Meta Vens
44ad5e8948 Journal admissions under the cache gate, in Sequence order
CultCache.AddJournal registers an ordered admission journal. It is called under
the gate right after Apply, once per admission that changed something, with the
admission's changes (dependents included) in Sequence order, before any
observer. A journal that throws joins the hold's failures: the admission stands,
the other journals run, the change is published and the writer hears of it
afterwards. Any hold entered while a journal runs throws. A hold now carries its
own changes and failures. Order stays data; no delivery thread or scheduler is
added.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 19:40:02 +02:00
Meta Vens
be91ea837a Pin that a failed hold publishes what it minted and an unsubscribed observer stops
A hold whose body throws after it admitted changes still publishes them, and
Held now assigns its result up front so it compiles under Stryker Safe Mode
(CS0165 had dropped every mutant in it). Disposing a Watch subscription removes
the observer from the cache.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 19:36:41 +02:00
Meta Vens
2a1b8a8cda Merge CultCache publication Cuts 1-2: project before mint, isolated observers
Soul verdict: merge. Before Cut 4 relies on sequence density: pin the
publish-on-throw guard in Held and observer removal on unsubscribe.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 19:30:04 +02:00
Meta Vens
69bd45ff14 Merge canonical CultMath serialization and the Geometry vector collapse (P4)
Every public CultMath value type serializes: MessagePack in the default resolver
chain (byte-identical to Aetheria's and StreamPixels' copies across 653 values
and all 115 float fields of the real Aetheria.cc) and System.Text.Json converters
(STJ-parity number handling, math types as dictionary keys). CultVec2/3 and
CultRect are gone with bytes unchanged. The cultlib Unity package depends on
cultmath; the release gate refuses until CultMath 0.3.0 carries every type.
Three Soul passes.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 18:34:09 +02:00
Meta Vens
4e5327c61c Bring the CultMath JSON converters to System.Text.Json parity and remove their allocations
Named literals now follow AllowReadingFromString and WriteAsString as
System.Text.Json's float does; quoted numbers refuse whitespace;
UnmappedMemberHandling.Disallow and AllowDuplicateProperties=false are honoured
and a repeated property replaces the whole value; dictionary keys read only the
canonical spelling. Components live in stack spans, property names are
pre-encoded, and there is no per-property closure, so a float3 read allocates
nothing (424 bytes before). Tests pin unknown nested values, de-DE and sv-SE
style cultures, and a distinct negative-zero key.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 18:31:16 +02:00
Meta Vens
0a543c9fb2 Kill the Stryker survivors in the projection and observer cuts
Delete the dead removed-ancestor branch and the duplicated name-read in the
variant name rule (resolution already refuses a missing or removed base and a
variant shares its base type). Add tests for the index rule counting a record
the same commit removes, blank names not colliding, and Dispose rethrowing
escaped completions after every observer completed.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 17:46:36 +02:00
Meta Vens
26a4f4240b Pin that a zero or missing Random state decodes to the generator's default seed
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 17:29:02 +02:00
Meta Vens
817c2627b8 Rebuild the CultMath JSON converters on a streaming reader with STJ number semantics
One Scalar plus Node description per type now drives reading, writing and the
dictionary-key spelling, replacing the per-value JsonDocument parse. NaN and the
infinities follow JsonNumberHandling (named literals or JsonException), floats
beyond float.MaxValue are refused instead of becoming infinity, out-of-range
components throw JsonException instead of FormatException, property matching
honours PropertyNameCaseInsensitive, and every type works as a dictionary key
spelled as comma-joined components.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 17:29:02 +02:00
Meta Vens
20c30224fd Add canonical System.Text.Json converters for every CultMath value type
CultMathJson is the JSON twin of CultMathResolver: vectors, quaternion and
Color32 are objects keyed by field name, matrices are arrays of row objects,
rect is {min,max}. Consumers add them with AddCultMathConverters(); CultLib owns
no JSON options. CultCircle and CultSphere serialize and deserialize through
them, with [JsonConstructor] so the struct constructor is used. The Caching
README documents each MessagePack and JSON shape in one table.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:57:31 +02:00
Meta Vens
77602d5c0e Pin every CultMath MessagePack shape and decode a missing Color32 alpha as opaque
Golden bytes now cover double2, double3, int3, int4, bool3, bool4, quaternion,
Color32, Random, float3x3 and CultCellular. A missing Color32 alpha decodes as
255, the constructor default. A test proves a consumer resolver registered for
the assembly wins over CultMathResolver. The redundant nil check is deleted;
ReadArrayHeader already refuses nil.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:57:31 +02:00
Meta Vens
ecee41e8ff CultCache: own the observer list and isolate every observer
The cache no longer fans changes out through an R3 Subject. It keeps a
copy-on-write observer array and Publish catches per observer: every observer
receives every change, an exception that escapes (a fail-fast R3 handler) joins
the OnUpdate failures rethrown to the writer after delivery, and the observer
stays subscribed. Dispose completes observers, and a disposed cache completes
new ones at once. Deletes the false "OnNext does not throw" comment.

The composition contract now states the observer policy and that a throwing
name or index getter refuses a change set, loads included.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:51:33 +02:00
Meta Vens
6b0cf66f0e Count serializer reads apart in the once-per-landed-record projection test
Resolving a variant serializes its base, and the serializer reads the same
getters; the test measured them together and expected too few.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:50:29 +02:00
Meta Vens
7362257f08 CultCache: read names and indexes once before minting; a throwing getter refuses the whole change set
Resolve now ends in Project, the only place an admission runs a [CultName] or
[CultIndex] getter on a landing record. It runs before land and before any
Sequence is minted, so a throwing getter refuses the admission (loads
included) with nothing written, minted or left in memory. The variant name and
index rules and the indexer read the projection; Apply runs no user code.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:48:19 +02:00
Meta Vens
14621b3893 Reference CultMath from Caching.MessagePack and scope the span helpers
The first commit's resolver needs the CultMath project reference and scoped
span parameters to compile; the golden test file also had two literal slips.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:15:29 +02:00
Meta Vens
139deb0e28 Add the canonical MessagePack encoding for every public CultMath value type
CultMathResolver formats float2/3/4, double2/3, int2/3/4, bool2/3/4,
quaternion, Color32, Random, float2x2, float3x3, rect and CultCellular as
component arrays, and sits in the default CultCache resolver chain so no
consumer registers anything. It lives in GameCult.Caching.MessagePack, which
already owns that chain; CultMath core stays dependency-free.

Vector, rect and Aetheria-formatter byte shapes are pinned by golden tests, a
reflection test fails when a public CultMath value type has no formatter, and
decode skips extra elements and refuses nil.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 16:13:22 +02:00
Meta Vens
b9cca38dbf Cover valid surrogate pairs and a second unpaired surrogate in node ids
Stryker on the escaping showed the pair-skip branch uncovered and the run
offset unpinned; the kinds now include a valid pair, a replacement character
before a lone low surrogate, a reversed pair, and two unpaired surrogates.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 15:49:51 +02:00
Meta Vens
7defd82c20 Pin node-id escaping, case-distinct nodes and the flat ungrouped list
Ids that differ only by %2F versus /, keys that differ only by case, and
unpaired surrogates each get a test; the ungrouped list stays flat.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 14:00:47 +02:00
Meta Vens
4f7d8821af Pin the create-in-group node check against a value that fits another grouping
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 13:29:09 +02:00
Meta Vens
439efd1550 Drop the inert inherit flag from the global check and its derived-type case
[CultGlobal] is not inherited, so the flag changed nothing and no test can tell.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 13:16:11 +02:00
Meta Vens
30e1f3fb95 Cover ungrouped types, null strings, derived globals and id separators
Stryker survivors on the grouping diff: the no-declaration path, the
null-string identity, the inherited [CultGlobal] check and the id separator.
GroupLevel takes text and tie as optional arguments so the scalar kinds
stop passing constant strings that could be swapped without any effect.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 13:15:03 +02:00
x
84c9a79972 Pin string case ordering and node id spelling in the grouping tests
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 12:52:05 +02:00
x
3d07ba0d20 Give the overriding variant its own name in the grouping test
A variant must override its name.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 12:21:28 +02:00
x
79e9318698 Add model-owned record grouping: declaration, tree, create-in-group, ref labels
[CultInspectorGroupBy(nameof(A), nameof(B))] on a document class names the
members its record list groups by, outermost first. CultInspectorModel owns
the rest: GroupingOf validates the declaration (a refused one gives a flat
list and a notice), GroupRecords partitions RecordCandidates by value
identity with occupied values only, CreateInGroup builds the record a node
holds, RecordRefLabel owns None and Missing <key>. A variant groups by its
resolved Document.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 12:09:25 +02:00
x
8ad0691bdd Declare the label fixture's key on its base member
The registry requires an override to repeat or omit the base declaration's [Key].

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 12:03:28 +02:00
Meta Vens
a6382475e4 Collapse the ref-candidate predicate and read inherited member attributes
RecordCandidates and the new IsRecordCandidate share one rule, so a
picker's candidates and a drop's acceptance cannot disagree. Member
attributes are read with Attribute.GetCustomAttributes(member, true):
PropertyInfo.GetCustomAttributes ignores inherit, so a base abstract
property's inspector attributes never reached the override.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 11:55:41 +02:00
Meta Vens
2eec436850 Variants C1: the index rule is linear, a merge judges only when a variant is involved, duplicates are deterministic
- 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>
2026-09-29 10:12:20 +02:00
Meta Vens
282e337c2c Variants C1: the held index tells the truth; a merge is judged on what the file will hold
The held index kept one key per value, last writer winning, and RemoveKey dropped
a value outright. Two consequences: removing or changing one of two plain
holders left the other unfindable, and the variant index rule judged a variant
against that lossy view, so a variant could land on a value a plain record still
held and the store then refused to reopen.

- _indexes maps a value to the keys holding it, oldest first. Index, Unindex and
  the removal maintain the set. GetByIndex finds the most recently indexed
  holder of a value, so removing the winner finds the other (behaviour change for
  plain duplicates: it used to find nothing).
- One rule, RefuseVariantIndexSharing, serves both write orders. It reads the
  truthful index minus what the commit replaces or removes, plus what it lands.
  ResolveOne no longer judges indexes; ClaimIncomingIndexValues, HeldByVariants
  and RefusePlainSharingWithVariant are gone.
- A conditional commit merged onto the file is judged by the cache's own
  resolution (CacheBackingStore.Judging) on the merged set: records other writers
  added, records they removed, and the batch. Another writer's variant sharing a
  value with the batch, or a base removal that strands another writer's variant,
  is refused with the named refusal instead of writing an unopenable store.
- Shedding is one decision at load. ToStoredDocument resolves the schema once and
  returns the report with it; a variant's override of a slot the type lacks is
  dropped from the delta there, so flush needs no predicate and every store
  reports through the same surface (ReportVariantSheds and OverridesTheTypeHas
  are gone). The directory store refuses variants and loads through the same call.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 09:20:47 +02:00
Meta Vens
354f9743e6 Refuse a plain write that shares a variant's indexed value; a flush sheds overrides the type lacks
The index rule is about the commit's result, so it now runs for every commit that changes an indexed value: a plain record landing on a value a variant of the post-commit set holds is refused naming both keys and the index (the held index is never trusted; a plain write had already overwritten it). Plain-vs-plain is untouched. A variant sheds overrides of slots its type lacks when it is flushed, as a plain record sheds the member, and the load's report names the slot when the schema itself never had it.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 08:28:24 +02:00
Meta Vens
5d6847e220 Align the C1 tests with the unique-index rule and fix the live-subscription fixture
Hand-written and directory-store variants now override their indexed value or
name as the rules require; the subscription test writes its variant through a
committed batch so the file holds it, subscribes by key, and reports what it
saw when it times out.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 07:04:01 +02:00
Meta Vens
c66d79e248 Variant overrides follow the plain-record outcome; bad writes and shared indexes are refused
F3: an override drifts exactly as the same member would on a plain record. A
slot the type dropped is ignored and reported by the store's schema resolution
(LastSchemaMigrationReports), the one surface a plain record uses; a value that
no longer decodes refuses the load, naming the variant and member. The second
report surface (VariantMigrationReports, the per-variant report map, the
catalog slot-name plumbing) is deleted.

F4: an override for a slot the type does not have, or a value that does not
decode as the member, is refused at write. Drift is for type changes after the
fact, not for bad writes. F5: two overrides of one slot are refused.

F2: a variant that shares an indexed value with another record is refused,
naming both keys and the index (the R6 rule extended to [CultIndex]); plain
record duplicate handling is unchanged.

F7: a flatten now resolves against the same batch, so a base edited in that
batch is in the flattened record; and the directory store refuses to load a
variant page as it refuses to write one.

The C1 vector's big variant now overrides its indexed code, as the unique-index
rule requires.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-09-29 06:58:51 +02:00