feat(dictate): add voice-reactive listening waveform

extensions/dictate/index.ts

@@ -2,7 +2,7 @@ import type { ExtensionAPI, ExtensionContext } from "@earendil-works/pi-coding-a
 import { createConnection } from "node:net";
 
 export const DICTATE_SOCKET_PATH = "/run/user/1000/pi-dictate.sock";
-export const PROTOCOL_VERSION = 1;
+export const PROTOCOL_VERSION = 2;
 export const MAX_FRAME_BYTES = 64 * 1024;
 export const MAX_TRANSCRIPT_BYTES = 32 * 1024;
 export const MAX_ID_BYTES = 128;
@@ -10,6 +10,8 @@ export const MAX_ID_BYTES = 128;
 const STATUS_KEY = "dictate";
 const MAX_ERROR_DETAIL_BYTES = 2 * 1024;
 const UTF8_DECODER = new TextDecoder("utf-8", { fatal: true });
+const WAVEFORM_BARS = ["▁", "▂", "▃", "▄", "▅", "▆", "▇", "█"] as const;
+const WAVEFORM_SAMPLES = 12;
 
 type DictateState =
   | "idle"
@@ -97,6 +99,7 @@ export function createDictateExtension(options: DictateExtensionOptions = {}) {
     let receiveBuffer = Buffer.alloc(0);
     let nextId = 0;
     let resultDelivered = false;
+    let audioLevels: number[] = [];
 
     const newId = (kind: string) => `${kind}-${++nextId}`;
 
@@ -104,6 +107,25 @@ export function createDictateExtension(options: DictateExtensionOptions = {}) {
       activeContext?.ui.setStatus(STATUS_KEY, status);
     };
 
+    const setListeningStatus = () => {
+      if (!activeContext) return;
+      const theme = activeContext.ui.theme;
+      const waveform = audioLevels
+        .map(
+          (level) =>
+            WAVEFORM_BARS[
+              Math.min(
+                WAVEFORM_BARS.length - 1,
+                Math.floor(level * WAVEFORM_BARS.length),
+              )
+            ],
+        )
+        .join("");
+      setStatus(
+        `${theme.fg("error", "●")} ${theme.fg("dim", "listening")}  ${theme.fg("accent", waveform)}`,
+      );
+    };
+
     const closeSocket = (current: DictateSocket, force: boolean) => {
       try {
         if (force) current.destroy();
@@ -121,6 +143,7 @@ export function createDictateExtension(options: DictateExtensionOptions = {}) {
       negotiated = false;
       receiveBuffer = Buffer.alloc(0);
       resultDelivered = false;
+      audioLevels = [];
       state = "idle";
       setStatus(undefined);
       activeContext = null;
@@ -183,7 +206,8 @@ export function createDictateExtension(options: DictateExtensionOptions = {}) {
       }
       if (serviceState === "listening" && state === "connecting") {
         state = "recording";
-        setStatus("listening...");
+        audioLevels = Array<number>(WAVEFORM_SAMPLES).fill(0);
+        setListeningStatus();
         return true;
       }
       if (serviceState === "stopping" && state === "stopping") {
@@ -267,6 +291,22 @@ export function createDictateExtension(options: DictateExtensionOptions = {}) {
         return;
       }
 
+      if (value.type === "level") {
+        if (
+          state !== "recording" ||
+          typeof value.level !== "number" ||
+          !Number.isFinite(value.level) ||
+          value.level < 0 ||
+          value.level > 1
+        ) {
+          protocolError("audio level is invalid for the current state.");
+          return;
+        }
+        audioLevels = [...audioLevels.slice(1), value.level];
+        setListeningStatus();
+        return;
+      }
+
       if (value.type === "result") {
         if (state !== "transcribing" || typeof value.transcript !== "string") {
           protocolError("result is invalid for the current state.");
@@ -349,6 +389,7 @@ export function createDictateExtension(options: DictateExtensionOptions = {}) {
       helloRequestId = newId("hello");
       negotiated = false;
       resultDelivered = false;
+      audioLevels = [];
       receiveBuffer = Buffer.alloc(0);
 
       let current: DictateSocket;