perf(ai): cache rendered message segments

packages/app-tests/aiMessageRender.test.ts

@@ -0,0 +1,42 @@
+import assert from "node:assert/strict";
+import test from "node:test";
+import { createAiMessageRenderer } from "../../apps/desktop/src/lib/aiMessageRender.ts";
+
+test("reuses rendered AI message segments for unchanged content", () => {
+  let markdownCalls = 0;
+  const renderer = createAiMessageRenderer({
+    markdown: (text) => {
+      markdownCalls++;
+      return `<p>${text}</p>`;
+    },
+  });
+
+  const first = renderer.render("hello **dbx**\n```sql\nSELECT 1\n```");
+  const second = renderer.render("hello **dbx**\n```sql\nSELECT 1\n```");
+
+  assert.equal(markdownCalls, 1);
+  assert.strictEqual(second, first);
+  assert.deepEqual(second, [
+    { type: "text", content: "hello **dbx**", html: "<p>hello **dbx**</p>" },
+    { type: "code", content: "SELECT 1", lang: "SQL" },
+  ]);
+});
+
+test("evicts older rendered AI message cache entries", () => {
+  let markdownCalls = 0;
+  const renderer = createAiMessageRenderer({
+    maxEntries: 2,
+    markdown: (text) => {
+      markdownCalls++;
+      return text;
+    },
+  });
+
+  renderer.render("one");
+  renderer.render("two");
+  renderer.render("one");
+  renderer.render("three");
+  renderer.render("two");
+
+  assert.equal(markdownCalls, 4);
+});