+
Rt STT
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+ enumerateDevices();
+ connect(socket);
+ };
+
+ function copyToClipboard(id) {
+ var textarea = document.getElementById(id);
+ textarea.select();
+ document.execCommand('copy');
+ }
+
+
+
+
+
\ No newline at end of file
diff --git a/web/server.js b/web/server.js
index 5ab311e..efb66a5 100644
--- a/web/server.js
+++ b/web/server.js
@@ -13,18 +13,49 @@ console.log(process.env)
console.log(process.env.TTS_BACKEND_URL)
console.log(process.env.WS_URL)
+let language = "en";
+let storeRecordings = false;
+let queueCounter = 0;
+
+const storage = require('node-persist');
+storage.init().then(() => {
+ storage.getItem('language').then((value) => {
+ if (value != undefined) { language = value; console.log('language: ' + language); }
+ else { storage.setItem('language', language).then(() => { console.log('language set to ' + language + "(default)"); }); }
+ });
+
+ storage.getItem('storeRecordings').then((value) => {
+ if (value != undefined) { storeRecordings = value; console.log('storeRecordings: ' + storeRecordings); }
+ else { storage.setItem('storeRecordings', storeRecordings).then(() => { console.log('storeRecordings set to ' + storeRecordings + "(default)"); }); }
+ });
+});
+
+
+
//we use https://hub.docker.com/r/onerahmet/openai-whisper-asr-webservice to transcribe the audio
//docker run -p 9009:9009 -d onerahmet/openai-whisper-asr-webservice
-wss.on('connection', (ws) => {
- console.log('Client ' + ws._socket.remoteAddress + ' connected');
+const sessions = new Map(); // Store session data
+
+wss.on('connection', (ws, req) => {
+ ws.sessionId = Math.random().toString(36).slice(2);
+ sessions.set(ws.sessionId, { language: 'en' });
+ console.log('Client ' + ws._socket.remoteAddress + ' connected with session id ' + ws.sessionId);
+ //send cookie to client
+ ws.send(JSON.stringify({ sessionId: ws.sessionId, language: language, storeRecordings: storeRecordings }));
ws.on('message', (data) => {
+ let webSocket = ws;
+ const sessionData = sessions.get(webSocket.sessionId);
+ if (!sessionData) {
+ console.log('No session data found for session id ' + webSocket.sessionId);
+ }
+ let language = sessionData?.language || 'en';
//show the size of the audio data as 0.000 MB
- console.log('Received data from client: ' + (data.length / 1024 / 1024).toFixed(3) + ' MB');
+ console.log('(queue ' + queueCounter + ') Received ' + (data.length / 1024 / 1024).toFixed(3) + ' MB audio from client. Crrent language: ' + language);
var request = require('request');
var formData = {
task: 'transcribe',
- language: 'en-US', //bg-BG|en-US
+ language: sessionData.language,
output: 'json',
audio_file: {
value: data,
@@ -34,31 +65,36 @@ wss.on('connection', (ws) => {
}
}
};
+
+ storeRecordings = sessionData?.storeRecordings || storeRecordings;
+ if (storeRecordings) {
+ //"yyyymmdd-hhMMss"
+ var timestampfilename = Date.now("yyyymmdd-hhMMss");
+ var fs = require('fs');
+ fs.mkdir('rec', { recursive: true }, (err) => {
+ if (err) throw err;
+ });
+ fs.writeFile('rec/audio' + timestampfilename + '.ogg', data, function (err) {
+ if (err) {
+ return console.log(err);
+ }
+ console.log('Audio data saved to audio.ogg');
+ });
+ }
- //"yyyymmdd-hhMMss"
- var timestampfilename = Date.now("yyyymmdd-hhMMss");
-
- //save the audio data to a file to /rec subfolder
- var fs = require('fs');
- fs.mkdir('rec', { recursive: true }, (err) => {
- if (err) throw err;
- });
-
- fs.writeFile('rec/audio' + timestampfilename + '.ogg', data, function (err) {
- if (err) {
- return console.log(err);
- }
- console.log('Audio data saved to audio.ogg');
- });
-
-
-
+ //record start time
+ var start = new Date().getTime();
+ queueCounter++;
request.post({ url: process.env.TTS_BACKEND_URL, formData: formData }, function optionalCallback(err, httpResponse, body) {
+ queueCounter--;
if (err) {
return console.error('upload failed:', err);
}
- console.log('Whisper decoded:', body);
- ws.send(body);
+ //duration of the transcribe in 0.00s
+ var duration = new Date().getTime() - start;
+ //console.log('decoded (' + duration + 'ms):', body);
+ console.log('decoded (' + (duration / 1000).toFixed(2) + 's):', body);
+ webSocket.send("(" + queueCounter + ") " + body);
});
});
});
@@ -72,7 +108,11 @@ function transcribeAudio(audioData) {
// --- web server that servers client.html
const express = require('express');
+const bodyParser = require('body-parser');
const app = express();
+app.use(bodyParser.json());
+// app.use(bodyParser.urlencoded({ extended: false })); // Parse request body as URL-encoded
+
const path = require('path');
app.get('/', (req, res) => {
@@ -84,6 +124,56 @@ app.get('/', (req, res) => {
app.get('/wsurl', (req, res) => {
res.send(process.env.WS_URL, 200, { 'Content-Type': 'text/plain' });
});
+//GET used to store default settings for all clients
+app.get('/settings', (req, res) => {
+ if (req.query.language != undefined) {
+ language = req.query.language;
+ storage.setItem('language', language).then(() => { console.log('language set to ' + language); });
+ }
+ if (req.query.storeRecordings != undefined) {
+ storeRecordings = req.query.storeRecordings;
+ storage.setItem('storeRecordings', storeRecordings).then(() => { console.log('storeRecordings set to ' + storeRecordings); });
+ }
+ //send back the current settings as json
+ res.send(JSON.stringify({ language: language, storeRecordings: storeRecordings }), 200, { 'Content-Type': 'text/plain' });
+});
+
+//POST used to store settings for a specific client
+app.post('/settings', (req, res) => {
+ //get the language from the json body ( { language: language, sessionId: sessionId })
+ const body = req.body;
+ const sid = body.sessionId;
+ const sessionData = sessions.get(sid);
+ if (body.language != undefined) {
+ sessionData.language = body.language;
+ console.log(`Session ${sid}: language set to ${sessionData.language}`);
+ }
+ if(body.storeRecordings != undefined) {
+ sessionData.storeRecordings = body.storeRecordings;
+ console.log(`Session ${sid}: storeRecordings set to ${sessionData.storeRecordings}`);
+ }
+ res.send('OK', 200, { 'Content-Type': 'text/plain' });
+});
+
+
+//save the audio file
+app.post('/upload', (req, res) => {
+ try {
+ //save the audio file
+ var timestampfilename = Date.now("yyyymmdd-hhMMss");
+ var fs = require('fs');
+ fs.mkdir('rec', { recursive: true }, (err) => {
+ if (err) throw err;
+ });
+ var file = fs.createWriteStream('rec/audio_slice_' + timestampfilename + '.ogg');
+ req.pipe(file);
+ res.send('OK', 200, { 'Content-Type': 'text/plain' });
+ } catch (err) {
+ console.log(err);
+ res.send('ERROR', 500, { 'Content-Type': 'text/plain' });
+ }
+});
+
app.listen(8080, () => {
console.log('Server listening on port 8080');