perf(compression): avoid cloning memo misses twice (#11727)

Obrigado! Validado em lote combinado (8 PRs, release/v3.8.51):

- TDD claro: sem o fix, o teste focado falha nas asserções de contagem exata de lookup sync/async; com o fix, `tests/unit/compression/result-memo.test.ts` passa (34/34).
- Remove supressão eslint agora obsoleta (`no-unused-vars` no arquivo de teste).
- `⚠️ base-red inherited: #11449` reconhecido e verificado — não é responsabilidade desta PR (confirmado via probe-worktree do tip puro).
- Gates estáticos do lote OK.

config/quality/eslint-suppressions.json

@@ -4539,11 +4539,6 @@
       "count": 1
     }
   },
-  "tests/unit/compression/result-memo.test.ts": {
-    "@typescript-eslint/no-unused-vars": {
-      "count": 1
-    }
-  },
   "tests/unit/compression/rtk-grouping.test.ts": {
     "@typescript-eslint/no-unused-vars": {
       "count": 1

open-sse/services/compression/resultMemo.ts

@@ -4,6 +4,7 @@ import type { CompressionConfig, CompressionMode, CompressionResult } from "./ty
 export const MEMO_CAP = 5_000;
 
 const memoMap = new Map<string, CompressionResult>();
+let lookupCountForTests = 0;
 
 // Opt-IN whitelist (NOT opt-out): cache only engines proven pure + STATELESS across
 // requests. Excluded on purpose: `ccr` and `session-dedup` write to the cross-request
@@ -94,6 +95,7 @@ function boundedSet(key: string, value: CompressionResult): void {
 }
 
 export function memoLookup(key: string): CompressionResult | null {
+  lookupCountForTests++;
   const hit = memoMap.get(key);
   if (!hit) return null;
   // Return a clone so downstream mutation cannot corrupt the cached value.
@@ -110,4 +112,10 @@ export function memoStore(key: string, result: CompressionResult): void {
 /** For tests only — clears the in-process memo store. */
 export function clearMemoStore(): void {
   memoMap.clear();
+  lookupCountForTests = 0;
 }
+export const resultMemoForTests = {
+  get lookupCount(): number {
+    return lookupCountForTests;
+  },
+};

open-sse/services/compression/strategySelector.ts

@@ -332,7 +332,7 @@ function runCompression(
       config: { ...options.config, memoizeCompressionResults: false },
     });
     memoStore(key, result);
-    return memoLookup(key)!;
+    return result;
   }
   if (mode === "rtk") {
     return applyRtkCompression(body, {
@@ -565,7 +565,7 @@ async function runCompressionAsync(
       config: { ...options.config, memoizeCompressionResults: false },
     });
     memoStore(key, result);
-    return memoLookup(key)!;
+    return result;
   }
   // Single-mode omniglyph (async-only) — resolution lives in engines/omniglyphSingleMode.ts.
   if (mode === "omniglyph") return applyOmniglyphSingleMode(body, options);

tests/unit/compression/result-memo.test.ts

@@ -6,11 +6,15 @@ import {
   makeMemoKey,
   isDeterministicMode,
   clearMemoStore,
+  resultMemoForTests,
   MEMO_CAP,
 } from "../../../open-sse/services/compression/resultMemo.ts";
 import type { CompressionResult } from "../../../open-sse/services/compression/types.ts";
 import { DEFAULT_COMPRESSION_CONFIG } from "../../../open-sse/services/compression/types.ts";
-import { applyCompression } from "../../../open-sse/services/compression/strategySelector.ts";
+import {
+  applyCompression,
+  applyCompressionAsync,
+} from "../../../open-sse/services/compression/strategySelector.ts";
 
 const baseBody = {
   messages: [{ role: "user", content: "hello world compress me please" }],
@@ -218,7 +222,6 @@ describe("applyCompression with memoization", () => {
   });
 
   it("flag OFF: two identical calls both compute (no caching path)", () => {
-    let callCount = 0;
     // We can't easily spy on internal engine, so we verify via deterministic output
     // equality between independent calls (proving cache isn't interfering).
     // Use a body that will be lightly compressed.
@@ -257,6 +260,38 @@ describe("applyCompression with memoization", () => {
     assert.notEqual(memoLookup(key), null);
   });
 
+  it("memo misses and hits cannot mutate the cached result", () => {
+    const body = {
+      messages: [{ role: "user", content: "Mutation isolation test content. ".repeat(15) }],
+      model: "gpt-4",
+    };
+    const miss = applyCompression(body, "lite", { config: memoConfig, principalId: "u1" });
+    const expected = structuredClone(miss.body);
+    miss.body.messages[0]!.content = "mutated miss";
+
+    const hit = applyCompression(body, "lite", { config: memoConfig, principalId: "u1" });
+    assert.deepEqual(hit.body, expected);
+    hit.body.messages[0]!.content = "mutated hit";
+
+    assert.deepEqual(
+      applyCompression(body, "lite", { config: memoConfig, principalId: "u1" }).body,
+      expected
+    );
+    assert.equal(resultMemoForTests.lookupCount, 3);
+  });
+
+  it("async memo misses and hits perform one lookup per call", async () => {
+    const body = {
+      messages: [{ role: "user", content: "Async lookup count test content. ".repeat(15) }],
+      model: "gpt-4",
+    };
+
+    await applyCompressionAsync(body, "lite", { config: memoConfig, principalId: "u1" });
+    assert.equal(resultMemoForTests.lookupCount, 1);
+    await applyCompressionAsync(body, "lite", { config: memoConfig, principalId: "u1" });
+    assert.equal(resultMemoForTests.lookupCount, 2);
+  });
+
   it("flag ON + deterministic mode: different principalId = MISS", () => {
     const body = {
       messages: [{ role: "user", content: "Cross-principal test content. ".repeat(10) }],