feat(chat): add Edge TTS playback

src/app/api/tts/edge/route.test.ts

@@ -0,0 +1,50 @@
+/**
+ * @jest-environment node
+ */
+
+import { POST } from "./route";
+
+const streamMock = jest.fn();
+
+jest.mock("edge-tts-ts", () => ({
+  Communicate: jest.fn().mockImplementation(() => ({
+    stream: streamMock,
+  })),
+}));
+
+describe("POST /api/tts/edge", () => {
+  beforeEach(() => {
+    streamMock.mockReset();
+  });
+
+  it("returns synthesized mp3 audio", async () => {
+    streamMock.mockImplementation(async function* () {
+      yield { type: "audio", data: new Uint8Array([1, 2]) };
+      yield { type: "SentenceBoundary", offset: 0, duration: 1, text: "测试" };
+      yield { type: "audio", data: new Uint8Array([3]) };
+    });
+
+    const response = await POST(
+      new Request("http://localhost/api/tts/edge", {
+        method: "POST",
+        body: JSON.stringify({ text: "测试文本" }),
+      }),
+    );
+
+    expect(response.status).toBe(200);
+    expect(response.headers.get("Content-Type")).toBe("audio/mpeg");
+    expect(Array.from(new Uint8Array(await response.arrayBuffer()))).toEqual([1, 2, 3]);
+  });
+
+  it("rejects empty text", async () => {
+    const response = await POST(
+      new Request("http://localhost/api/tts/edge", {
+        method: "POST",
+        body: JSON.stringify({ text: " " }),
+      }),
+    );
+
+    expect(response.status).toBe(400);
+    expect(await response.json()).toEqual({ error: "text is required" });
+  });
+});

src/app/api/tts/edge/route.ts

@@ -0,0 +1,76 @@
+import { NextResponse } from "next/server";
+import { Communicate } from "edge-tts-ts";
+
+export const runtime = "nodejs";
+export const dynamic = "force-dynamic";
+
+const DEFAULT_VOICE = process.env.EDGE_TTS_VOICE || "zh-CN-XiaoxiaoNeural";
+const MAX_TEXT_LENGTH = 12000;
+
+type EdgeTtsRequest = {
+  text?: unknown;
+  voice?: unknown;
+};
+
+const jsonError = (message: string, status: number) =>
+  NextResponse.json({ error: message }, { status });
+
+export async function POST(request: Request) {
+  let payload: EdgeTtsRequest;
+  try {
+    payload = (await request.json()) as EdgeTtsRequest;
+  } catch {
+    return jsonError("Invalid JSON body", 400);
+  }
+
+  const text = typeof payload.text === "string" ? payload.text.trim() : "";
+  if (!text) {
+    return jsonError("text is required", 400);
+  }
+  if (text.length > MAX_TEXT_LENGTH) {
+    return jsonError(`text must be ${MAX_TEXT_LENGTH} characters or fewer`, 413);
+  }
+
+  const voice =
+    typeof payload.voice === "string" && payload.voice.trim()
+      ? payload.voice.trim()
+      : DEFAULT_VOICE;
+
+  try {
+    const communicate = new Communicate(text, { voice });
+    const chunks: Uint8Array[] = [];
+    let byteLength = 0;
+
+    for await (const chunk of communicate.stream()) {
+      if (chunk.type !== "audio") continue;
+      chunks.push(chunk.data);
+      byteLength += chunk.data.byteLength;
+    }
+
+    if (byteLength === 0) {
+      return jsonError("Edge TTS returned empty audio", 502);
+    }
+
+    const audio = new Uint8Array(byteLength);
+    let offset = 0;
+    for (const chunk of chunks) {
+      audio.set(chunk, offset);
+      offset += chunk.byteLength;
+    }
+
+    const audioBuffer = audio.buffer.slice(
+      audio.byteOffset,
+      audio.byteOffset + audio.byteLength,
+    );
+
+    return new Response(audioBuffer, {
+      headers: {
+        "Content-Type": "audio/mpeg",
+        "Cache-Control": "no-store",
+      },
+    });
+  } catch (error) {
+    console.error("[EdgeTTS] Failed to synthesize speech:", error);
+    return jsonError("Failed to synthesize speech", 502);
+  }
+}