updated avatars and voice mode
All checks were successful
Deploy FluentGerman.ai / deploy (push) Successful in 51s
All checks were successful
Deploy FluentGerman.ai / deploy (push) Successful in 51s
This commit is contained in:
@@ -1,12 +1,11 @@
|
||||
/* FluentGerman.ai — Voice module (Web Speech API + API mode) */
|
||||
/* FluentGerman.ai — Voice module (API-only TTS, browser + API STT) */
|
||||
|
||||
class VoiceManager {
|
||||
constructor() {
|
||||
this.mode = 'browser'; // will be set from server config
|
||||
this.mode = 'browser';
|
||||
this.recognition = null;
|
||||
this.synthesis = window.speechSynthesis;
|
||||
this.isRecording = false;
|
||||
this.isDisabled = false; // true when no STT method is available
|
||||
this.isDisabled = false;
|
||||
this.lastInputWasVoice = false;
|
||||
this.mediaRecorder = null;
|
||||
this.audioChunks = [];
|
||||
@@ -17,10 +16,8 @@ class VoiceManager {
|
||||
}
|
||||
|
||||
async init() {
|
||||
// Check browser STT support
|
||||
this._initBrowserSTT();
|
||||
|
||||
// Fetch voice mode from server
|
||||
try {
|
||||
const response = await api('/voice/config');
|
||||
if (response?.ok) {
|
||||
@@ -34,24 +31,17 @@ class VoiceManager {
|
||||
this.mode = 'browser';
|
||||
}
|
||||
|
||||
// Determine the best available mode
|
||||
// Determine best STT method
|
||||
if (this.mode === 'browser' && !this.browserSTTSupported) {
|
||||
if (this.apiAvailable) {
|
||||
console.log('[Voice] Browser STT not supported, falling back to API mode');
|
||||
this.mode = 'api';
|
||||
showToast('Using cloud voice recognition — your browser doesn\'t support built-in speech recognition.', 'info');
|
||||
} else {
|
||||
// Neither method works
|
||||
console.warn('[Voice] No STT method available — disabling voice input');
|
||||
this.isDisabled = true;
|
||||
}
|
||||
} else if (this.mode === 'api' && !this.apiAvailable) {
|
||||
// Server says API but API isn't actually configured
|
||||
if (this.browserSTTSupported) {
|
||||
console.log('[Voice] API STT not configured, using browser STT');
|
||||
this.mode = 'browser';
|
||||
} else {
|
||||
console.warn('[Voice] No STT method available — disabling voice input');
|
||||
this.isDisabled = true;
|
||||
}
|
||||
}
|
||||
@@ -62,7 +52,6 @@ class VoiceManager {
|
||||
_initBrowserSTT() {
|
||||
const SpeechRecognition = window.SpeechRecognition || window.webkitSpeechRecognition;
|
||||
if (!SpeechRecognition) {
|
||||
console.warn('[Voice] SpeechRecognition API not available in this browser');
|
||||
this.browserSTTSupported = false;
|
||||
return;
|
||||
}
|
||||
@@ -75,35 +64,32 @@ class VoiceManager {
|
||||
|
||||
this.recognition.onresult = (event) => {
|
||||
const text = event.results[0][0].transcript;
|
||||
console.log('[Voice] Browser STT result:', text);
|
||||
console.log('[Voice] STT result:', text);
|
||||
this.lastInputWasVoice = true;
|
||||
if (this.onResult) this.onResult(text);
|
||||
};
|
||||
|
||||
this.recognition.onend = () => {
|
||||
console.log('[Voice] Browser STT ended');
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
};
|
||||
|
||||
this.recognition.onerror = (event) => {
|
||||
console.error('[Voice] Browser STT error:', event.error);
|
||||
console.error('[Voice] STT error:', event.error);
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
|
||||
if (event.error === 'not-allowed') {
|
||||
showToast('Microphone access denied. Please allow microphone in browser settings.', 'error');
|
||||
showToast('Microphone access denied. Allow it in browser settings.', 'error');
|
||||
} else if (event.error === 'no-speech') {
|
||||
showToast('No speech detected. Try again.', 'error');
|
||||
}
|
||||
};
|
||||
|
||||
console.log('[Voice] Browser STT initialized');
|
||||
}
|
||||
|
||||
async startRecording() {
|
||||
if (this.isDisabled) {
|
||||
showToast('Voice input requires Chrome or Edge (with HTTPS). Firefox is not supported.', 'error');
|
||||
showToast('Voice requires Chrome or Edge (HTTPS).', 'error');
|
||||
return;
|
||||
}
|
||||
|
||||
@@ -112,79 +98,66 @@ class VoiceManager {
|
||||
if (this.onStateChange) this.onStateChange(true);
|
||||
|
||||
if (this.mode === 'api') {
|
||||
// API mode — record audio via MediaRecorder, send to Whisper
|
||||
try {
|
||||
const stream = await navigator.mediaDevices.getUserMedia({ audio: true });
|
||||
this.audioChunks = [];
|
||||
this.mediaRecorder = new MediaRecorder(stream);
|
||||
|
||||
this.mediaRecorder.ondataavailable = (event) => {
|
||||
this.audioChunks.push(event.data);
|
||||
};
|
||||
this.mediaRecorder.ondataavailable = (e) => this.audioChunks.push(e.data);
|
||||
|
||||
this.mediaRecorder.onstop = async () => {
|
||||
stream.getTracks().forEach(track => track.stop());
|
||||
stream.getTracks().forEach(t => t.stop());
|
||||
const blob = new Blob(this.audioChunks, { type: 'audio/webm' });
|
||||
await this._transcribeAPI(blob);
|
||||
};
|
||||
|
||||
this.mediaRecorder.start();
|
||||
console.log('[Voice] API recording started');
|
||||
} catch (e) {
|
||||
console.error('[Voice] Microphone access error:', e);
|
||||
showToast('Microphone access denied', 'error');
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
}
|
||||
} else {
|
||||
// Browser mode — use Web Speech API
|
||||
if (this.recognition) {
|
||||
try {
|
||||
this.recognition.start();
|
||||
console.log('[Voice] Browser STT started');
|
||||
} catch (e) {
|
||||
console.error('[Voice] Failed to start recognition:', e);
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
showToast('Voice recognition failed to start. Try again.', 'error');
|
||||
showToast('Voice recognition failed. Try again.', 'error');
|
||||
}
|
||||
} else {
|
||||
console.warn('[Voice] No speech recognition available');
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
showToast('Voice input requires Chrome or Edge (with HTTPS).', 'error');
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
stopRecording() {
|
||||
console.log('[Voice] Stopping recording...');
|
||||
if (this.mode === 'api') {
|
||||
if (this.mediaRecorder && this.mediaRecorder.state === 'recording') {
|
||||
if (this.mediaRecorder?.state === 'recording') {
|
||||
this.mediaRecorder.stop();
|
||||
} else {
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
}
|
||||
} else {
|
||||
if (this.recognition) {
|
||||
try {
|
||||
this.recognition.stop();
|
||||
} catch (e) {
|
||||
// Already stopped
|
||||
}
|
||||
}
|
||||
try { this.recognition?.stop(); } catch (e) { /* already stopped */ }
|
||||
this.isRecording = false;
|
||||
if (this.onStateChange) this.onStateChange(false);
|
||||
}
|
||||
}
|
||||
|
||||
toggleRecording() {
|
||||
if (this.isDisabled) {
|
||||
showToast('Voice requires Chrome or Edge (HTTPS).', 'error');
|
||||
return;
|
||||
}
|
||||
this.isRecording ? this.stopRecording() : this.startRecording();
|
||||
}
|
||||
|
||||
async _transcribeAPI(blob) {
|
||||
try {
|
||||
const formData = new FormData();
|
||||
formData.append('audio', blob, 'recording.webm');
|
||||
|
||||
console.log('[Voice] Sending audio to API for transcription...');
|
||||
const response = await api('/voice/transcribe', {
|
||||
method: 'POST',
|
||||
body: formData,
|
||||
@@ -192,14 +165,12 @@ class VoiceManager {
|
||||
|
||||
if (response?.ok) {
|
||||
const data = await response.json();
|
||||
console.log('[Voice] API transcription result:', data.text);
|
||||
this.lastInputWasVoice = true;
|
||||
if (this.onResult) this.onResult(data.text);
|
||||
} else {
|
||||
showToast('Transcription failed. Please try again.', 'error');
|
||||
showToast('Transcription failed.', 'error');
|
||||
}
|
||||
} catch (e) {
|
||||
console.error('[Voice] API transcription error:', e);
|
||||
showToast('Transcription error', 'error');
|
||||
} finally {
|
||||
this.isRecording = false;
|
||||
@@ -207,33 +178,20 @@ class VoiceManager {
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Speak text via API TTS only. No browser fallback.
|
||||
* Strips markdown formatting before sending.
|
||||
*/
|
||||
async speak(text) {
|
||||
if (this.mode === 'api' && this.apiAvailable) {
|
||||
return this._speakAPI(text);
|
||||
} else {
|
||||
return this._speakBrowser(text);
|
||||
if (!this.apiAvailable) {
|
||||
console.log('[Voice] API TTS not available, skipping speech');
|
||||
return;
|
||||
}
|
||||
}
|
||||
|
||||
_speakBrowser(text) {
|
||||
return new Promise((resolve) => {
|
||||
// Cancel any ongoing speech
|
||||
this.synthesis.cancel();
|
||||
const utterance = new SpeechSynthesisUtterance(text);
|
||||
utterance.lang = 'de-DE';
|
||||
utterance.rate = 0.95;
|
||||
utterance.onend = resolve;
|
||||
utterance.onerror = () => {
|
||||
console.warn('[Voice] Browser TTS error');
|
||||
resolve();
|
||||
};
|
||||
this.synthesis.speak(utterance);
|
||||
});
|
||||
}
|
||||
const clean = VoiceManager.stripMarkdown(text);
|
||||
|
||||
async _speakAPI(text) {
|
||||
try {
|
||||
const response = await api(`/voice/synthesize?text=${encodeURIComponent(text)}`, {
|
||||
const response = await api(`/voice/synthesize?text=${encodeURIComponent(clean)}`, {
|
||||
method: 'POST',
|
||||
});
|
||||
|
||||
@@ -242,26 +200,33 @@ class VoiceManager {
|
||||
const audioUrl = URL.createObjectURL(audioBlob);
|
||||
const audio = new Audio(audioUrl);
|
||||
await audio.play();
|
||||
return new Promise(resolve => {
|
||||
audio.onended = resolve;
|
||||
});
|
||||
return new Promise(resolve => { audio.onended = resolve; });
|
||||
}
|
||||
} catch (e) {
|
||||
console.warn('[Voice] API TTS failed, falling back to browser');
|
||||
console.warn('[Voice] API TTS failed:', e);
|
||||
}
|
||||
// Fallback to browser TTS
|
||||
return this._speakBrowser(text);
|
||||
}
|
||||
|
||||
toggleRecording() {
|
||||
if (this.isDisabled) {
|
||||
showToast('Voice input requires Chrome or Edge (with HTTPS). Firefox is not supported.', 'error');
|
||||
return;
|
||||
}
|
||||
if (this.isRecording) {
|
||||
this.stopRecording();
|
||||
} else {
|
||||
this.startRecording();
|
||||
}
|
||||
/**
|
||||
* Strip markdown formatting from text so TTS reads naturally.
|
||||
*/
|
||||
static stripMarkdown(text) {
|
||||
return text
|
||||
.replace(/```[\s\S]*?```/g, '') // code blocks
|
||||
.replace(/`([^`]+)`/g, '$1') // inline code
|
||||
.replace(/#{1,6}\s+/g, '') // headings
|
||||
.replace(/\*\*([^*]+)\*\*/g, '$1') // bold
|
||||
.replace(/\*([^*]+)\*/g, '$1') // italic
|
||||
.replace(/__([^_]+)__/g, '$1') // bold alt
|
||||
.replace(/_([^_]+)_/g, '$1') // italic alt
|
||||
.replace(/~~([^~]+)~~/g, '$1') // strikethrough
|
||||
.replace(/^\s*[-*+]\s+/gm, '') // unordered lists
|
||||
.replace(/^\s*\d+\.\s+/gm, '') // ordered lists
|
||||
.replace(/\[([^\]]+)\]\([^)]+\)/g, '$1')// links
|
||||
.replace(/!\[([^\]]*)\]\([^)]+\)/g, '') // images
|
||||
.replace(/>\s+/g, '') // blockquotes
|
||||
.replace(/\n{2,}/g, '. ') // paragraph breaks → pause
|
||||
.replace(/\n/g, ' ') // newlines → space
|
||||
.trim();
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user