feat(audio): add offline local neural voice via Piper (sv + en)

Announcements can use Piper, a free local neural TTS, instead of the browser's system
voice. No account, no paid service, and it keeps working with no internet at the track.

- npm run tts:install (or Settings -> Audio -> Install) downloads the standalone Piper
  release for Linux/Windows/macOS plus Swedish and English voices into piper/ once
- server/tts.js generates each line once, one at a time, and caches it as WAV under
  data/tts/; a new line takes ~0.3 s with the standalone build
- src/piper_speech.js plays it and falls back to the browser voice if Piper is missing,
  errors, or takes over 2.5 s, so no announcement is lost and timing is never held up
- "Pre-generate phrases and driver names" renders every fixed phrase and name in both
  languages before race day

Browser voice stays the default. Generation routes fall under the admin PIN.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

server.js

@@ -9,6 +9,8 @@ const Database = require("better-sqlite3");
 const publicTrainingProfile = require("./server/public_training_profile");
 const { createBackupService } = require("./server/backup_service");
 const { createAuth, COOKIE_NAME: AUTH_COOKIE_NAME, TOKEN_MAX_AGE_MS } = require("./server/auth");
+const { createTts, resolvePiperBinary } = require("./server/tts");
+const { installAll: installPiperAll } = require("./server/tts_install");
 
 const PORT = Number(process.env.PORT || 8081);
 const HOST = process.env.HOST || "0.0.0.0";
@@ -16,6 +18,11 @@ const DB_DIR = process.env.RC_TIMING_DATA_DIR ? path.resolve(process.env.RC_TIMI
 const DB_PATH = path.join(DB_DIR, "rc_timing.sqlite");
 const AMMC_CONFIG_PATH = path.join(DB_DIR, "ammc_config.json");
 const auth = createAuth({ dataDir: DB_DIR, allowedCidrs: process.env.RC_ADMIN_ALLOWED_CIDRS || "" });
+const tts = createTts({
+  dataDir: DB_DIR,
+  projectDir: __dirname,
+  voiceDirs: [process.env.RC_PIPER_VOICES_DIR, path.join(DB_DIR, "piper-voices"), path.join(__dirname, "piper", "voices"), path.join(os.homedir(), ".local", "share", "piper-voices")].filter(Boolean),
+});
 const PID_FILE = process.env.RC_TIMING_PID_FILE || "";
 const SRC_DIR = path.join(__dirname, "src");
 const RACE_SOUND_DIR = path.join(__dirname, "rc-race-sound-pack");
@@ -199,6 +206,38 @@ app.post("/api/auth/disable", (_req, res) => {
   res.json({ ok: true, enabled: false });
 });
 
+app.get("/api/tts/status", (_req, res) => {
+  res.json({ ...tts.status(), install: ttsInstall });
+});
+
+// POST (not GET) so generating audio falls under the admin PIN like every other write.
+app.post("/api/tts/speech", async (req, res) => {
+  try {
+    const file = await tts.speech({ text: req.body?.text, language: req.body?.language, rate: req.body?.rate });
+    res.type("audio/wav").sendFile(file);
+  } catch (error) {
+    res.status(error.code === "unavailable" ? 503 : error.code === "bad_request" ? 400 : 500).json({ error: error.message });
+  }
+});
+
+// One-time download of Piper + Swedish/English voices, run in the background; progress is
+// polled through /api/tts/status. Needs internet only while it runs.
+const ttsInstall = { running: false, log: [], error: "" };
+app.post("/api/tts/install", (_req, res) => {
+  if (!ttsInstall.running) {
+    Object.assign(ttsInstall, { running: true, log: [], error: "" });
+    installPiperAll({ projectDir: __dirname, hasBinary: Boolean(resolvePiperBinary(process.env.RC_PIPER_BIN || "", __dirname)), log: (line) => ttsInstall.log.push(line) })
+      .catch((error) => { ttsInstall.error = error.message; })
+      .finally(() => { ttsInstall.running = false; });
+  }
+  res.status(202).json(ttsInstall);
+});
+
+app.post("/api/tts/pregenerate", (req, res) => {
+  tts.pregenerate(req.body?.items);
+  res.status(202).json(tts.status().pregenerate);
+});
+
 app.get("/api/health", (_req, res) => {
   res.json({ ok: true, dbPath: DB_PATH });
 });