dsh-voice-mode 0.7.21 → 0.7.22

package/README.en.md

@@ -340,6 +340,7 @@ Also: the on-device recording fixture used for debugging is **off by default**;
 | Mic click does nothing, red hint in the status bar | The browser denied mic permission: allow it in the address bar and retry |
 | Status bar stuck on `Loading model… x%` | Check the network; the model is large (160 MB) — `npm run prefetch` first; on mainland networks set `modelHost` to `https://hf-mirror.com` |
 | Status bar shows `Model download failed (<file>)` | Both mirrors are unreachable: check network/proxy and re-enter voice mode (resumable) |
+| Local Kokoro on the official desktop app shows "failed" and asks for Node.js | The desktop host is Electron, where Kokoro's native add-on cannot run: install Node.js >= 18 on PATH (or set `DSHVM_NODE` to its path) and retry, or use the default Edge / local VITS (unaffected on desktop) |
 | Caption appears (overlay) but no sound | Check system volume/output; if autoplay is blocked, click anywhere on the page and retry |
 | Status bar shows `朗读连接失败：正在重试…` | Edge TTS unreachable (overseas service), auto-retries; if it persists, check network/proxy |
 | Poor recognition | Get closer to the mic, reduce ambient noise; if echo remains, raise the interrupt sensitivity by one step |

package/README.md

@@ -313,6 +313,7 @@ flowchart LR
 | 唤醒词完全没反应 | ① 确认交互模式是 `toggle`（`hold` 与手动打断下唤醒词不生效）；② 确认 `bargeInMode`：`manual`（或 `detect` 在本机无原生回声消除时自动落 manual）会关闭常驻聆听 → 唤醒词失效，改回 `auto` 或按住麦克风/Ctrl；③ 朗读期说唤醒词不触发（先等 AI 说完）；④ 看待机态实时转写，确认它听到了什么（同音字 1 字内已容错，建议 3-4 字词） |
 | 按住说话松手后没反应 | 确认交互模式为「按住说」且按住期间按钮高亮；松手后识别定稿约 1 秒内进入草稿 |
 | 打不断（朗读中开口无反应） | 调高 `interruptLevel`（更敏感档 = 0 或 1）或检查麦克风权限；若 VAD 持续不触发可临时切到「手动打断」（`mode: hold` / 唤醒词定时延后）；**不要**调 `echoGateDb`——真机 3.1 分钟朗读期里 Silero 0/777 帧把回声判成语音，原生 AEC 生效时此阈值从未被执行（详见 ADR-0006） |
+| 官方桌面端上选本地 Kokoro，设置里显示「加载失败」并提示需要 Node.js | 桌面端宿主是 Electron，Kokoro 的原生 addon 在其中不可用：在系统 PATH 上安装 Node.js ≥18（或设环境变量 `DSHVM_NODE` 指向它）后重试；或改用默认的 Edge / 本地 VITS（桌面端不受影响） |
 | 字幕被输入框挡住 | 默认 `captionMaxWidth=1`（70vw）+ `captionFontSize=0`（12px）在窄屏可能与底部输入框重叠；调高档位或点字幕浮层「×」收起 |
 | 让位行为异常（朗读期说「嗯」不停 / 真话被打断） | 「嗯/对」类短词触发让位 1.5s 后继续朗读；继续说真话会走硬打断；不要时关 `backchannelYield` 即可恢复改造前行为（ADR-0008） |
 | 朗读期说「嗯」没让位 | 确认 `backchannelYield=true`（默认开）；hold 模式松手后让位 1.5s 内继续说话会变硬打断 |

package/lib/client.js

@@ -1534,6 +1534,7 @@ var zh = {
   engineLoading: "\u52A0\u8F7D\u4E2D\u2026",
   engineReady: "\u5C31\u7EEA",
   engineError: "\u52A0\u8F7D\u5931\u8D25",
+  errKokoroNeedsNode: "\u672C\u5730 Kokoro \u65E0\u6CD5\u5728\u5B98\u65B9\u684C\u9762\u7AEF\uFF08Electron\uFF09\u5185\u76F4\u63A5\u8FD0\u884C\uFF1A\u8BF7\u5728\u7CFB\u7EDF PATH \u4E0A\u5B89\u88C5 Node.js \u226518\uFF08\u6216\u8BBE\u73AF\u5883\u53D8\u91CF DSHVM_NODE \u6307\u5411\u5B83\uFF09\uFF0C\u6216\u628A\u6717\u8BFB\u5F15\u64CE\u5207\u6362\u4E3A Edge / \u672C\u5730 VITS\u3002",
   ttsModelsMissing: "\u672C\u5730\u6A21\u578B\u672A\u5C31\u7EEA",
   ttsDownload: "\u4E0B\u8F7D",
   ttsDelete: "\u5220\u9664",
@@ -1763,6 +1764,7 @@ var en = {
   engineLoading: "loading\u2026",
   engineReady: "ready",
   engineError: "failed",
+  errKokoroNeedsNode: "Local Kokoro cannot run directly inside the official desktop app (Electron): install Node.js >= 18 on PATH (or set DSHVM_NODE to its path), or switch the read-aloud engine to Edge / local VITS.",
   ttsModelsMissing: "local models missing",
   ttsDownload: "Download",
   ttsDelete: "Delete",
@@ -2749,7 +2751,7 @@ function EngineStatusInline() {
         children: acting ? acting === "clean" ? t("ttsDeleting") : t("ttsDownloading") : action === "clean" ? t("ttsDelete") : t("ttsDownload")
       }
     ),
-    tts.error && /* @__PURE__ */ (0, import_jsx_runtime.jsx)("span", { style: { fontSize: 11, color: "var(--dsw-alias-state-error-primary)", flexBasis: "100%" }, children: tts.error })
+    tts.error && /* @__PURE__ */ (0, import_jsx_runtime.jsx)("span", { style: { fontSize: 11, color: "var(--dsw-alias-state-error-primary)", flexBasis: "100%" }, children: tts.errorCode === "kokoro_needs_node" ? t("errKokoroNeedsNode") : tts.error })
   ] });
 }
 function ModelStatusView() {
@@ -3099,7 +3101,7 @@ var TELEMETRY_VIEW = [
   { stage: "first-tts-chunk", key: "telFirstChunk" },
   { stage: "first-audio-played", key: "telFirstPlayed" }
 ];
-var BUILD_TAG = "b51f379";
+var BUILD_TAG = "b05db39";
 var TELEMETRY_FLAG = "dsh-voice-mode.telemetry";
 var telemetryEnabled = typeof localStorage !== "undefined" && localStorage.getItem(TELEMETRY_FLAG) === "1";
 console.log("[dsh-voice] build=" + BUILD_TAG);

package/lib/index.js

@@ -1313,6 +1313,13 @@ var probeNode = (command, env) => {
   }
 };
 var NATIVE_RUNTIME_UNAVAILABLE = 'Local Kokoro cannot run inside the Electron-based desktop host (native add-on blocked: "External buffers are not allowed"). Install Node.js >= 18 on PATH (or set DSHVM_NODE to its path), or switch the read-aloud engine to Edge or VITS.';
+var NativeRuntimeUnavailableError = class extends Error {
+  code = "kokoro_needs_node";
+  constructor() {
+    super(NATIVE_RUNTIME_UNAVAILABLE);
+    this.name = "NativeRuntimeUnavailableError";
+  }
+};
 var cached;
 function resolveNativeRuntime(opts = {}) {
   const env = opts.env ?? process.env;
@@ -1660,6 +1667,7 @@ function createSherpaLocalEngine(options) {
   let ready = null;
   let engineLoading = false;
   let engineError;
+  let engineErrorCode;
   let nextId = 1;
   const pending = /* @__PURE__ */ new Map();
   const call = (msg) => new Promise((resolve, reject) => {
@@ -1681,6 +1689,7 @@ function createSherpaLocalEngine(options) {
       ready = (async () => {
         engineLoading = true;
         engineError = void 0;
+        engineErrorCode = void 0;
         downloadProgress = null;
         try {
           if (!child || !childInit) {
@@ -1708,7 +1717,7 @@ function createSherpaLocalEngine(options) {
             }
             if (!child) {
               const runtime = options.kind === "kokoro" ? resolveNativeRuntime() : { env: process.env };
-              if (!runtime) throw new Error(NATIVE_RUNTIME_UNAVAILABLE);
+              if (!runtime) throw new NativeRuntimeUnavailableError();
               child = fork(workerPath, [], {
                 stdio: ["ignore", "ignore", "pipe", "ipc"],
                 ...runtime.execPath ? { execPath: runtime.execPath, env: runtime.env } : {}
@@ -1754,6 +1763,7 @@ function createSherpaLocalEngine(options) {
           broadcast("tts-ready", { engine: options.kind, worker: true });
         } catch (e) {
           engineError = e instanceof Error ? e.message : String(e);
+          engineErrorCode = e instanceof NativeRuntimeUnavailableError ? e.code : void 0;
           throw e;
         } finally {
           engineLoading = false;
@@ -1790,6 +1800,7 @@ function createSherpaLocalEngine(options) {
         ready: childInit === true,
         loading: engineLoading,
         error: engineError,
+        errorCode: engineErrorCode,
         progress: downloadProgress ?? void 0,
         local: {
           repo: repoName,
@@ -2064,6 +2075,7 @@ var en = {
   engineLoading: "loading\u2026",
   engineReady: "ready",
   engineError: "failed",
+  errKokoroNeedsNode: "Local Kokoro cannot run directly inside the official desktop app (Electron): install Node.js >= 18 on PATH (or set DSHVM_NODE to its path), or switch the read-aloud engine to Edge / local VITS.",
   ttsModelsMissing: "local models missing",
   ttsDownload: "Download",
   ttsDelete: "Delete",
@@ -2284,7 +2296,7 @@ var name = "voice-mode";
 var NS_VOICE_MODE = "voice-mode";
 var BASE_PATH = "/voice-mode";
 var PREVIEW_NETWORK_PATTERN = /fetch failed|ECONN|ENOTFOUND|getaddrinfo|ETIMEDOUT|EAI_AGAIN|network|unreachable|socket hang up|aborted/i;
-var PREVIEW_ENGINE_PATTERN = /model download|model verify|init failed|child exited|tts child|local TTS|prepare|sherpa/i;
+var PREVIEW_ENGINE_PATTERN = /model download|model verify|init failed|child exited|tts child|local TTS|prepare|sherpa|local Kokoro/i;
 var PREVIEW_TEXT_PATTERN = /empty or invalid audio|invalid audio|invalid text|too long|truncat/i;
 function classifyPreviewError(msg) {
   if (PREVIEW_TEXT_PATTERN.test(msg)) return "text";

package/package.json

@@ -2,7 +2,7 @@
   "name": "dsh-voice-mode",
   "description": "Full-duplex voice plugin for DeepSeek Harness: local zipformer2 streaming ASR (no API key) → editable draft; Edge TTS or local VITS / Kokoro read-aloud with live captions; true barge-in; hardened HTTP surface + model SHA256 pinning; compatible with all dsh releases since 0.1.1-rc.2 (verified 0.1.1-rc.2 → 0.2.0-rc.2).",
   "icon": "./icon.svg",
-  "version": "0.7.21",
+  "version": "0.7.22",
   "type": "module",
   "main": "lib/index.js",
   "repository": {