feat(voice): add optional speech-to-text input and read-aloud TTS

Adds a push-to-talk mic button in the composer and a read-aloud button on
assistant messages. Both are opt-in and hidden unless a voice backend is
configured via VOICE_SIDECAR_URL.

The auth-gated /api/voice proxy forwards to a configurable backend exposing
/transcribe and /tts (provider-agnostic); the frontend probes /api/voice/health
and hides the controls when disabled. Adds i18n keys and docs/voice.md.

Includes a local, no-API-key reference backend in voice-sidecar/ (faster-whisper
for STT, Kokoro-82M for TTS, both CPU-capable).
This commit is contained in:
newsbubbles
2026-06-08 00:47:14 +01:00
parent af3a28abc7
commit d05585e1f4
17 changed files with 720 additions and 0 deletions

View File

@@ -0,0 +1,106 @@
import { useCallback, useRef, useState } from 'react';
import { authenticatedFetch } from '../../../utils/api';
// Mobile-safe recording: iOS Safari 18.4+ supports webm/opus; older iOS needs mp4.
const MIME_CANDIDATES = [
'audio/webm;codecs=opus',
'audio/webm',
'audio/mp4',
'audio/ogg;codecs=opus',
'audio/ogg',
];
function pickMime(): string {
for (const t of MIME_CANDIDATES) {
try {
if (typeof MediaRecorder !== 'undefined' && MediaRecorder.isTypeSupported(t)) return t;
} catch {
/* isTypeSupported can throw on some iOS versions */
}
}
return '';
}
export type VoiceInputState = 'idle' | 'recording' | 'transcribing';
/**
* Push-to-talk dictation. Records the mic, uploads to /api/voice/transcribe
* (faster-whisper sidecar via the Express proxy), returns text via onTranscript.
* Ported from tooler's VoiceInput.js.
*/
export function useVoiceInput(onTranscript: (text: string) => void, onError?: (msg: string) => void) {
const [state, setState] = useState<VoiceInputState>('idle');
const recorderRef = useRef<MediaRecorder | null>(null);
const chunksRef = useRef<Blob[]>([]);
const streamRef = useRef<MediaStream | null>(null);
const stopTracks = () => {
streamRef.current?.getTracks().forEach((t) => t.stop());
streamRef.current = null;
};
const start = useCallback(async () => {
try {
const stream = await navigator.mediaDevices.getUserMedia({
audio: { echoCancellation: true, noiseSuppression: true },
});
streamRef.current = stream;
const mimeType = pickMime();
const rec = mimeType ? new MediaRecorder(stream, { mimeType }) : new MediaRecorder(stream);
recorderRef.current = rec;
chunksRef.current = [];
rec.ondataavailable = (e) => {
if (e.data.size > 0) chunksRef.current.push(e.data);
};
rec.onstop = async () => {
stopTracks();
const type = rec.mimeType || 'audio/webm';
const blob = new Blob(chunksRef.current, { type });
if (blob.size < 800) {
setState('idle');
onError?.('Recording too short');
return;
}
setState('transcribing');
try {
const ext = type.includes('mp4') ? 'm4a' : type.includes('ogg') ? 'ogg' : 'webm';
const fd = new FormData();
fd.append('audio', blob, `recording.${ext}`);
const res = await authenticatedFetch('/api/voice/transcribe', { method: 'POST', body: fd });
if (!res.ok) throw new Error(`transcribe ${res.status}`);
const data = await res.json();
const text = String(data?.text || '').trim();
if (text) onTranscript(text);
else onError?.('No speech detected');
} catch (e) {
onError?.(`Transcription failed: ${e instanceof Error ? e.message : String(e)}`);
} finally {
setState('idle');
}
};
rec.start();
setState('recording');
} catch (e) {
const err = e as { name?: string; message?: string };
let msg = `Mic error: ${err?.message || e}`;
if (err?.name === 'NotAllowedError') msg = 'Microphone access denied.';
else if (err?.name === 'NotFoundError') msg = 'No microphone found.';
onError?.(msg);
setState('idle');
}
}, [onTranscript, onError]);
const stop = useCallback(() => {
if (recorderRef.current && state === 'recording') recorderRef.current.stop();
}, [state]);
const toggle = useCallback(() => {
if (state === 'recording') stop();
else if (state === 'idle') start();
}, [state, start, stop]);
return { state, toggle };
}