test: replace streaming source freezes

docs/reviews/2026-08-18-r3-test-truth-ledger.md

@@ -38,12 +38,12 @@ pwsh ./tools/audit-test-inventory.ps1
 The generated JSON lives under ignored `artifacts/test-audit/`; it is not a
 second 6.8 MB checked-in source of truth. The script and this reviewed ledger
 are durable, while paths and line numbers regenerate after every batch. The
-inventory refreshed through Batch W reports:
+inventory refreshed through Batch X reports:
 
 | Syntax signal | Current count |
 |---|---:|
 | Tracked/parsed C# test files | 1,254 / 1,254 |
-| Attributed test methods (not expanded theory cases) | 11,416 |
+| Attributed test methods (not expanded theory cases) | 11,417 |
 | Exact duplicate-data rows | 0 |
 | Reviewed body-equivalent groups / methods | 11 / 27 |
 | Methods containing at least one empty `return;` | 138 |
@@ -56,7 +56,7 @@ inventory refreshed through Batch W reports:
 | Methods directly using `Thread.Sleep` / `Task.Delay` | 15 / 14 |
 | Cancellable-infinite-only / elapsed-time methods | 5 / 24 |
 | Methods directly reading environment variables | 47 |
-| Methods directly / through same-file helpers reading `.cs` source text | 63 / 98 |
+| Methods directly / through same-file helpers reading `.cs` source text | 57 / 92 |
 
 The three remaining prerequisite-return candidates are all reviewed branch
 false positives: the two Windows/Linux factory assertions and the launcher's
@@ -140,7 +140,7 @@ identifiable in the lane report.
 | T-009 wall-clock double-click tests | resolved in batch E | Four sleeps were replaced by a deterministic test clock behind an internal factory overload. The production factory still reads `Environment.TickCount64` exactly as before. |
 | T-010 two useless cases | high-confidence cleanup batch A | Delete `SmokeTest.TestProject_IsWired` and `ChaseCameraTests.ImplementsICamera`; compilation already proves both claims. |
 | T-011 diagnostic-only methods | resolved in batches C, L, and R | The reviewed current set is 82 methods / 103 cases. All carry `Purpose=Diagnostic`, preserving the apparatus while removing it from release pass totals. Batch R catches investigations whose only assertion validates fixture/DAT availability, which the original mechanical output-only scan could not distinguish from an oracle. |
-| T-012 source-text freezes | direct-read map completed in batch I, helper-mediated gap corrected in batch U, and staged replacement active | Seventeen whole-tree architecture rules and five cross-artifact contracts stay. Batch W retires the first nine literal implementation freezes beside stronger behavior or compiled-metadata evidence; 76 staged replacements remain. |
+| T-012 source-text freezes | direct-read map completed in batch I, helper-mediated gap corrected in batch U, and staged replacement active | Seventeen whole-tree architecture rules and five cross-artifact contracts stay. Batches W/X retire the first 15 literal implementation freezes beside stronger behavior or compiled-metadata evidence; 70 staged replacements remain. |
 | T-013 controller self-comparison | high-confidence cleanup batch A | Capture the first controller next to the first body and compare every retry with that reference. |
 | T-014 seven load-sensitive tests | six mechanisms repaired in batch F; one product defect classified | Virtual/wall-clock mixing, tiered-JIT allocation noise, a live controller clock, and a ThreadPool-start timing oracle were removed without weakening behavioral contracts. `DatSoundCache` #321 is a real in-flight-entry race and now runs as `Status=KnownFailure` pending a product fix. |
 | T-015 four non-prerequisite skips | resolved in batches A/B | PVS scaffold deleted with rationale preserved; redundant chat/radar generators deleted in favor of the comprehensive Manual lane; tower oracle is `Status=KnownFailure`. |
@@ -1085,3 +1085,67 @@ Verification:
   source readers from 67 to 63 and total direct/helper readers from 107 to 98.
   The remaining 98 reconcile to the 22 approved retained policies/contracts
   and 76 staged replacements.
+
+## Batch X streaming publication and live-entity source-freeze replacement
+
+Batch X converts the four approved streaming-publication freezes and both
+live-entity composition freezes. No product source changes. The tests now
+inspect the built dependency graph, execute the real origin publishers, or
+assert the typed owner/reset calls instead of matching C# fragments.
+
+The conversion found one materially misleading old assertion. The physics
+publisher test rejected the literal strings
+`_physicsDataCache.RemoveCellsForLandblock` and
+`_physicsDataCache.RemoveBuildingsForLandblock`. Those fragments were absent,
+but the real publisher correctly performs both removals through
+`publication.StagingCache`; the test therefore passed without establishing its
+intended generation-isolation rule. The replacement verifies the compiled
+`AdvanceBeginOne` path obtains the staging cache before each removal, commits
+through `RuntimePhysicsState.CommitCollisionGeneration`, does not restart the
+retained-owner capture, and keeps the old publication bodies out of
+`GameWindow`.
+
+The other exact decisions are:
+
+- the render publisher boundary now scans every compiled GameWindow method and
+  constructor for forbidden terrain, cell-visibility, and EnvCell publication
+  calls, and reflects the removed mesh-preparation/body ownership;
+- the former umbrella origin test is narrowed to the claim its fixture could
+  not observe: production `LivePresentationCompositionPhase` constructs the
+  render, physics, and static publishers before the one presentation pipeline,
+  while GameWindow retains only that pipeline. Captured origin itself remains
+  covered by the real `LandblockBuildFactory.BuildFar` and
+  `LandblockRenderPublisher.BeginPublication` behavior tests;
+- recenter ordering is now its own named compiled-call contract: old-window
+  retirement completion precedes `LiveWorldOriginState.Recenter`, which
+  precedes destination commit. The extensive recenter transaction suite still
+  owns retry, replacement, session-reset, and retirement behavior;
+- the shutdown test reads the compiled operation manifest, including delegate
+  targets, to prove Runtime session stop precedes streamer disposal. It also
+  proves `GameRuntime.StopSession` disposes the live session before checking
+  its completion barrier and retains the re-entrant diagnostic message;
+- live animation/projectile composition now reflects the exact typed fields
+  and proves `CompletePresentation` constructs `DatProjectileSetupResolver`
+  before `ProjectileController`, with no window closure or remote-motion-view
+  owner; and
+- the session reset test follows compiled calls through
+  `LiveSessionRuntimeFactory`, `PlayerModeController`, the network-update
+  owner, and `LiveWorldOriginState`, rather than pinning four statement strings.
+
+`CompiledCallGraph` gained reusable readers for every declared method,
+delegate method targets, and compiled string literals. These are test-only
+metadata facilities used when invoking the real process/window graph would
+mutate external resources.
+
+Verification:
+
+- the four directly affected owner suites pass 111/111;
+- the complete locked Release build covers all 44 projects with zero warnings
+  and zero errors;
+- the no-retry complete hermetic Release gate passes 14,349/14,349 with zero
+  skips or failures across all 12 test assemblies. The one-case increase is the
+  newly separated recenter-order contract; and
+- the regenerated 1,254-file inventory parses every file, reports 11,417
+  attributed methods, and reduces direct/total source readers from 63/98 to
+  57/92. The remaining 92 reconcile to 22 approved retained
+  policies/contracts and 70 staged replacements.