@flowingspring/dsh-voco 0.3.7 → 0.3.8

package/lib/plugins/voice-local.js

@@ -72,6 +72,9 @@ var LocalSession = class {
 	async start() {
 		await this.backend.start((event) => this.receive(event));
 	}
+	beginManualUtterance() {
+		this.backend.beginManualUtterance();
+	}
 	appendAudio(audio) {
 		this.backend.appendAudio(audio);
 	}
@@ -314,6 +317,8 @@ var NodeSpeechBackend = class {
 	preRoll = [];
 	preRollBytes = 0;
 	active;
+	manual = false;
+	manualMaxed = false;
 	recognitionQueue = Promise.resolve();
 	synthesisQueue = Promise.resolve();
 	synthesisGeneration = 0;
@@ -345,8 +350,16 @@ var NodeSpeechBackend = class {
 		emit({ type: "ready" });
 		return Promise.resolve();
 	}
+	beginManualUtterance() {
+		if (this.closed) return;
+		this.active = void 0;
+		this.preRoll.splice(0);
+		this.preRollBytes = 0;
+		this.manual = true;
+		this.manualMaxed = false;
+	}
 	appendAudio(audio) {
-		if (this.closed || audio.byteLength === 0) return;
+		if (this.closed || audio.byteLength === 0 || this.manualMaxed) return;
 		if (audio.byteLength % 2 !== 0) {
 			this.emit?.({
 				type: "error",
@@ -369,14 +382,19 @@ var NodeSpeechBackend = class {
 			this.active.silenceBytes = 0;
 			if (!this.active.confirmed && this.active.voicedBytes >= this.bytesFor(this.config.minSpeechDurationMs ?? 250)) this.confirmUtterance();
 		} else this.active.silenceBytes += frame.byteLength;
-		if (!this.active.confirmed && this.active.silenceBytes >= this.bytesFor(this.config.minSpeechDurationMs ?? 250)) {
+		if (!this.manual && !this.active.confirmed && this.active.silenceBytes >= this.bytesFor(this.config.minSpeechDurationMs ?? 250)) {
 			this.active = void 0;
 			return;
 		}
-		if (this.active.silenceBytes >= this.bytesFor(this.config.silenceDurationMs) || this.active.totalBytes >= this.bytesFor(this.config.maxUtteranceMs)) this.finishUtterance();
+		if (!this.manual && this.active.silenceBytes >= this.bytesFor(this.config.silenceDurationMs) || this.active.totalBytes >= this.bytesFor(this.config.maxUtteranceMs)) {
+			this.finishUtterance();
+			if (this.manual) this.manualMaxed = true;
+		}
 	}
 	commitAudio() {
 		this.finishUtterance();
+		this.manual = false;
+		this.manualMaxed = false;
 	}
 	synthesize(responseId, text) {
 		const generation = this.synthesisGeneration;
@@ -435,6 +453,8 @@ var NodeSpeechBackend = class {
 		this.synthesisGeneration += 1;
 		this.synthesisResponses.clear();
 		this.active = void 0;
+		this.manual = false;
+		this.manualMaxed = false;
 		this.preRoll.splice(0);
 		await Promise.allSettled([this.recognitionQueue, this.synthesisQueue]);
 		this.emit?.({

package/lib/plugins/voice-web.js

@@ -213,7 +213,8 @@ async function attach(ctx, socket, rawUrl) {
 				return;
 			}
 			const type = control.type;
-			if (type === "audio.commit") voice.commitAudio(session.id);
+			if (type === "audio.push-to-talk.start") voice.beginManualUtterance(session.id);
+			else if (type === "audio.commit") voice.commitAudio(session.id);
 			else if (type === "response.interrupt") voice.interruptResponse(session.id);
 			else if (type === "playback.ended") voice.playbackEnded(session.id);
 			else if (type === "task.cancel") {

package/lib/plugins/voice.js

@@ -169,6 +169,10 @@ var VoiceRuntime = class extends Service {
 	appendAudio(id, audio) {
 		this.requireAttachedSession(id).provider.appendAudio(audio);
 	}
+	/** Begin a manually committed push-to-talk utterance. */
+	beginManualUtterance(id) {
+		this.requireAttachedSession(id).provider.beginManualUtterance();
+	}
 	/**
 	* Commit current microphone input.
 	* @param id - voice session.

package/package.json

@@ -1,7 +1,7 @@
-{
+{
   "name": "@flowingspring/dsh-voco",
   "description": "Persistent voice conversations for DSH with cloud speech recognition, Edge TTS, and background Agent delegation",
-  "version": "0.3.7",
+  "version": "0.3.8",
   "keywords": [
     "deepseek-harness",
     "dsh",
@@ -15,18 +15,18 @@
   "bugs": {
     "url": "https://github.com/lgquan/dsh-voco/issues"
   },
-  "publishConfig": {
-    "access": "public"
-  },
-  "repository": {
-    "type": "git",
-    "url": "git+https://github.com/lgquan/dsh-voco.git",
-    "directory": "packages/voice-app"
-  },
-  "type": "module",
-  "main": "lib/index.js",
-  "types": "lib/types/index.d.ts",
-  "exports": {
+  "publishConfig": {
+    "access": "public"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/lgquan/dsh-voco.git",
+    "directory": "packages/voice-app"
+  },
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
     ".": {
       "types": "./lib/types/index.d.ts",
       "default": "./lib/index.js"
@@ -45,17 +45,17 @@
     "./voice-web": "./lib/plugins/voice-web.js",
     "./cordis.patch.yml": "./cordis.patch.yml",
     "./package.json": "./package.json"
-  },
-  "files": [
+  },
+  "files": [
     "lib/index.js",
     "lib/invariant.js",
     "lib/client.js",
     "lib/client.js.map",
     "lib/plugins/*.js",
     "cordis.patch.yml",
-    "lib/types/**/*.d.ts"
-  ],
-  "license": "MIT",
+    "lib/types/**/*.d.ts"
+  ],
+  "license": "MIT",
   "dsh": {
     "bundle": {
       "patch": "./cordis.patch.yml"
@@ -131,15 +131,15 @@
     "@deepseek-ai/dsh-system-prompt": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-tools": "^0.1.1-rc.1",
     "@deepseek-ai/dsh-workspace": "^0.1.1-rc.1",
-    "@flowingspring/dsh-client-ui-voice": "workspace:*",
-    "@flowingspring/dsh-llm-tool-call-compat": "workspace:*",
-    "@flowingspring/dsh-voice": "workspace:*",
-    "@flowingspring/dsh-voice-assistant": "workspace:*",
-    "@flowingspring/dsh-voice-local": "workspace:*",
-    "@flowingspring/dsh-voice-web": "workspace:*",
     "@types/react": "~18.3.1",
     "@types/ws": "^8.18.1",
     "react": "^18.2.0",
-    "react-dom": "^18.2.0"
+    "react-dom": "^18.2.0",
+    "@flowingspring/dsh-voice": "0.1.2",
+    "@flowingspring/dsh-voice-local": "0.1.0",
+    "@flowingspring/dsh-llm-tool-call-compat": "0.1.0",
+    "@flowingspring/dsh-voice-web": "0.1.2",
+    "@flowingspring/dsh-client-ui-voice": "0.1.2",
+    "@flowingspring/dsh-voice-assistant": "0.1.2"
   }
-}
+}