v103: РАБОТАЕТ — транскрипция через ProxyAPI.ru (whisper-1), конвертация webm→WAV в браузере

This commit is contained in:
“Naeel”
2026-05-23 11:35:52 +04:00
parent c83e3d9da2
commit d8b3643da3
8 changed files with 257 additions and 148 deletions
+1 -1
View File
@@ -1,6 +1,6 @@
// ---------- Глобальные переменные и инициализация ----------
const GROQ_API_KEY = '';
const VERSION = 'v89';
const VERSION = 'v103';
let mediaRecorder, audioChunks = [], userAudioBlob = null, ttsText = '';
let isRecording = false, isAnalyzing = false, micAnalyser = null, micAnimId = null;
+50 -38
View File
@@ -1,46 +1,58 @@
// ---------- Groq API: транскрипция (HTTP-чанки <10KB — обход DPI) ----------
// ---------- Groq API: транскрипция (прямой POST через ProxyAPI.ru) ----------
async function blobToWav(blob) {
const ab = await blob.arrayBuffer();
const ctx = new AudioContext({ sampleRate: 16000 });
const buf = await ctx.decodeAudioData(ab);
await ctx.close();
const mono = new Float32Array(buf.length);
for (let c = 0; c < buf.numberOfChannels; c++) {
const ch = buf.getChannelData(c);
for (let i = 0; i < buf.length; i++) mono[i] += ch[i];
}
if (buf.numberOfChannels > 1) for (let i = 0; i < mono.length; i++) mono[i] /= buf.numberOfChannels;
const wavBuf = new ArrayBuffer(44 + mono.length * 2);
const v = new DataView(wavBuf);
const wr = (o, s) => { for (let i = 0; i < s.length; i++) v.setUint8(o + i, s.charCodeAt(i)); };
wr(0, 'RIFF'); v.setUint32(4, 36 + mono.length * 2, true);
wr(8, 'WAVE'); wr(12, 'fmt ');
v.setUint32(16, 16, true); v.setUint16(20, 1, true); v.setUint16(22, 1, true);
v.setUint32(24, 16000, true); v.setUint32(28, 32000, true); v.setUint16(32, 2, true); v.setUint16(34, 16, true);
wr(36, 'data'); v.setUint32(40, mono.length * 2, true);
let off = 44;
for (let i = 0; i < mono.length; i++) {
const s = Math.max(-1, Math.min(1, mono[i]));
v.setInt16(off, s < 0 ? s * 0x8000 : s * 0x7FFF, true); off += 2;
}
return new Blob([wavBuf], { type: 'audio/wav' });
}
async function transcribe(blob) {
if (!GROQ_API_KEY) return { text: '', error: 'no_key' };
const tStart = performance.now();
const base64 = await new Promise((resolve, reject) => {
const reader = new FileReader();
reader.onload = () => resolve(reader.result.split(',')[1]);
reader.onerror = reject;
reader.readAsDataURL(blob);
});
const CHUNK = 3000; // ~4KB base64 на чанк (POST <10KB — DPI не режет)
const total = Math.ceil(base64.length / CHUNK);
const mime = blob.type || 'audio/webm';
console.log('[CHUNK] blob=' + (blob.size/1024).toFixed(1) + 'KB b64=' + base64.length + ' chunks=' + total);
let sid = '';
try {
for (let i = 0; i < base64.length; i += CHUNK) {
const idx = Math.floor(i / CHUNK);
const body = JSON.stringify({
idx, total,
chunk: base64.slice(i, i + CHUNK),
mime, token: GROQ_API_KEY,
sid
});
const r = await fetch('https://lang.kube5s.ru/openai/v1/transcribe', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body
});
const data = await r.json();
if (idx === total - 1) {
console.log('[CHUNK] ← ' + ((performance.now()-tStart)/1000).toFixed(2) + 's ' + (data.text||'(empty)'));
return data.error ? { text: '', error: data.error } : { text: data.text || '', error: null };
}
sid = data.sid || '';
}
const wavBlob = await blobToWav(blob);
const fd = new FormData();
fd.append('file', wavBlob, 'audio.wav');
fd.append('model', 'whisper-1');
fd.append('language', 'it');
const ctrl = new AbortController();
const timer = setTimeout(() => ctrl.abort(), 30000);
const r = await fetch('https://api.proxyapi.ru/openai/v1/audio/transcriptions', {
method: 'POST',
headers: { 'Authorization': `Bearer ${GROQ_API_KEY}` },
body: fd,
signal: ctrl.signal
});
clearTimeout(timer);
const raw = await r.text();
console.log('[TRANSCRIBE] ← ' + r.status + ' ' + ((performance.now()-tStart)/1000).toFixed(2) + 's raw=' + raw.slice(0,200));
let data;
try { data = JSON.parse(raw); } catch(pe) { return { text: '', error: 'parse:' + raw.slice(0,80) }; }
return { text: data.text || '', error: data.error?.message || data.error || null };
} catch(e) {
console.log('[CHUNK] ERR', e);
return { text: '', error: 'network' };
console.log('[TRANSCRIBE] ERR', e);
return { text: '', error: e.message };
}
}
@@ -51,7 +63,7 @@ async function translateToRussian(text) {
method: 'POST',
headers: { 'Authorization': `Bearer ${GROQ_API_KEY}`, 'Content-Type': 'application/json' },
body: JSON.stringify({
model: 'llama-3.3-70b-versatile',
model: 'gpt-4o',
messages: [{ role: 'user', content: 'Переведи на русский одним словом или фразой: ' + text }],
max_tokens: 50,
temperature: 0