Add local MeloTTS support

2026-05-03 01:56:09 +09:00
parent 3360015179
commit ad357a6ede
13 changed files with 396 additions and 3 deletions
--- a/src/services/melo-tts.ts
+++ b/src/services/melo-tts.ts
@@ -0,0 +1,113 @@
+import { spawn } from "node:child_process";
+import { randomUUID } from "node:crypto";
+import { mkdir, rm } from "node:fs/promises";
+import path from "node:path";
+
+import type { AppConfig } from "../config.js";
+import type { Logger } from "../logger.js";
+import { playWavFile } from "./audio-playback.js";
+
+async function run(command: string, args: string[], stdio: "ignore" | "inherit" = "ignore"): Promise<void> {
+  await new Promise<void>((resolve, reject) => {
+    const child = spawn(command, args, {
+      stdio: ["ignore", stdio, "inherit"],
+      windowsHide: true,
+    });
+
+    child.on("error", reject);
+    child.on("exit", (code) => {
+      if (code === 0) {
+        resolve();
+        return;
+      }
+      reject(new Error(`${command} ${args.join(" ")} exited with code ${code ?? "null"}`));
+    });
+  });
+}
+
+export class MeloTtsService {
+  constructor(
+    private readonly config: AppConfig,
+    private readonly logger: Logger,
+  ) {}
+
+  async warmup(): Promise<void> {
+    await mkdir(path.resolve(process.cwd(), this.config.TTS_CACHE_DIR), { recursive: true });
+    await mkdir(path.resolve(process.cwd(), this.config.TTS_OUTPUT_DIR), { recursive: true });
+
+    await run("docker", ["--version"]);
+    await run("docker", ["image", "inspect", this.config.TTS_IMAGE]);
+  }
+
+  async speak(text: string): Promise<void> {
+    const trimmed = text.trim();
+    if (!trimmed) {
+      return;
+    }
+
+    const fileName = `tts-${Date.now()}-${randomUUID()}.wav`;
+    const targetPath = path.resolve(process.cwd(), this.config.TTS_OUTPUT_DIR, fileName);
+
+    try {
+      await this.synthesizeToFile(trimmed, targetPath);
+      await playWavFile(targetPath);
+    } finally {
+      await rm(targetPath, { force: true }).catch(() => undefined);
+    }
+  }
+
+  async synthesizeToFile(text: string, targetPath: string): Promise<void> {
+    await this.warmup();
+
+    const outputDir = path.dirname(targetPath);
+    const cacheDir = path.resolve(process.cwd(), this.config.TTS_CACHE_DIR);
+    const fileName = path.basename(targetPath);
+
+    await mkdir(outputDir, { recursive: true });
+
+    const args = [
+      "run",
+      "--rm",
+      "-v",
+      `${outputDir}:/work/output`,
+      "-v",
+      `${cacheDir}:/cache`,
+      "-e",
+      "HF_HOME=/cache/huggingface",
+      "-e",
+      "HF_HUB_CACHE=/cache/huggingface/hub",
+      "-e",
+      "TRANSFORMERS_CACHE=/cache/transformers",
+    ];
+
+    if (this.config.TTS_DEVICE !== "cpu") {
+      args.push("--gpus", "all");
+    }
+
+    args.push(
+      this.config.TTS_IMAGE,
+      "--text",
+      text,
+      "--output",
+      `/work/output/${fileName}`,
+      "--language",
+      this.config.TTS_LANGUAGE,
+      "--speaker",
+      this.config.TTS_SPEAKER,
+      "--speed",
+      String(this.config.TTS_SPEED),
+      "--device",
+      this.config.TTS_DEVICE,
+    );
+
+    this.logger.info("Starting MeloTTS synthesis", {
+      image: this.config.TTS_IMAGE,
+      language: this.config.TTS_LANGUAGE,
+      speaker: this.config.TTS_SPEAKER,
+      speed: this.config.TTS_SPEED,
+      device: this.config.TTS_DEVICE,
+    });
+
+    await run("docker", args, "inherit");
+  }
+}