update node image to the latest and most secure (#81)

Co-authored-by: Guilherme Rauen <g.rauen@cognigy.com>

.gitignore

@@ -1,6 +1,7 @@
 # Logs
 logs
 *.log
+run-tests.sh
 
 # Runtime data
 pids

Dockerfile

@@ -1,4 +1,4 @@
-FROM --platform=linux/amd64 node:18.9.0-alpine3.16 as base
+FROM --platform=linux/amd64 node:18.12.1-alpine3.16 as base
 
 RUN apk --update --no-cache add --virtual .builds-deps build-base python3
 

lib/utils/speech-utils.js

@@ -2,6 +2,8 @@ const ttsGoogle = require('@google-cloud/text-to-speech');
 const sttGoogle = require('@google-cloud/speech').v1p1beta1;
 const Polly = require('aws-sdk/clients/polly');
 const AWS = require('aws-sdk');
+const { Deepgram } = require('@deepgram/sdk');
+const sdk = require('microsoft-cognitiveservices-speech-sdk');
 const bent = require('bent');
 const fs = require('fs');
 
@@ -32,6 +34,65 @@ const testGoogleStt = async(logger, credentials) => {
   }
 };
 
+const testDeepgramStt = async(logger, credentials) => {
+  const {api_key} = credentials;
+  const deepgram = new Deepgram(api_key);
+
+  const mimetype = 'audio/wav';
+  const source = {
+    buffer: fs.readFileSync(`${__dirname}/../../data/test_audio.wav`),
+    mimetype: mimetype
+  };
+
+  return new Promise((resolve, reject) => {
+    // Send the audio to Deepgram and get the response
+    deepgram.transcription
+      .preRecorded(source, {punctuate: true})
+      .then((response) => {
+        //logger.debug({response}, 'got transcript');
+        if (response?.results?.channels[0]?.alternatives?.length > 0) resolve(response);
+        else reject(new Error('no transcript returned'));
+        return;
+      })
+      .catch((err) => {
+        logger.info({err}, 'failed to get deepgram transcript');
+        reject(err);
+      });
+  });
+};
+
+const testMicrosoftStt = async(logger, credentials) => {
+  const {api_key, region} = credentials;
+
+  const speechConfig = sdk.SpeechConfig.fromSubscription(api_key, region);
+  const audioConfig = sdk.AudioConfig.fromWavFileInput(fs.readFileSync(`${__dirname}/../../data/test_audio.wav`));
+  speechConfig.speechRecognitionLanguage = 'en-US';
+  const speechRecognizer = new sdk.SpeechRecognizer(speechConfig, audioConfig);
+
+  return new Promise((resolve, reject) => {
+    speechRecognizer.recognizeOnceAsync((result) => {
+      switch (result.reason) {
+        case sdk.ResultReason.RecognizedSpeech:
+          resolve();
+          break;
+        case sdk.ResultReason.NoMatch:
+          reject('Speech could not be recognized.');
+          break;
+        case sdk.ResultReason.Canceled:
+          const cancellation = sdk.CancellationDetails.fromResult(result);
+          logger.info(`CANCELED: Reason=${cancellation.reason}`);
+          if (cancellation.reason == sdk.CancellationReason.Error) {
+            logger.info(`CANCELED: ErrorCode=${cancellation.ErrorCode}`);
+            logger.info(`CANCELED: ErrorDetails=${cancellation.errorDetails}`);
+          }
+          reject(cancellation.reason);
+          break;
+      }
+      speechRecognizer.close();
+    });
+  });
+};
+
 const testAwsTts = (logger, credentials) => {
   const polly = new Polly(credentials);
   return new Promise((resolve, reject) => {
@@ -89,11 +150,6 @@ const testMicrosoftTts = async(logger, credentials) => {
   }
 };
 
-const testMicrosoftStt = async(logger, credentials) => {
-  //TODO
-  return true;
-};
-
 const testWellSaidTts = async(logger, credentials) => {
   const {api_key} = credentials;
   try {
@@ -127,4 +183,5 @@ module.exports = {
   testMicrosoftTts,
   testMicrosoftStt,
   testWellSaidStt,
+  testDeepgramStt
 };

package.json

@@ -18,10 +18,11 @@
     "url": "https://github.com/jambonz/jambonz-api-server.git"
   },
   "dependencies": {
-    "@google-cloud/speech": "^4.10.2",
+    "@deepgram/sdk": "^1.10.2",
-    "@google-cloud/text-to-speech": "^3.4.0",
+    "@google-cloud/speech": "^5.1.0",
-    "@jambonz/db-helpers": "^0.6.19",
+    "@google-cloud/text-to-speech": "^4.0.3",
-    "@jambonz/realtimedb-helpers": "^0.4.35",
+    "@jambonz/db-helpers": "^0.7.3",
+    "@jambonz/realtimedb-helpers": "^0.5.9",
     "@jambonz/time-series": "^0.2.5",
     "argon2-ffi": "^2.0.0",
     "aws-sdk": "^2.1152.0",
@@ -35,6 +36,7 @@
     "helmet": "^5.1.0",
     "jsonwebtoken": "^8.5.1",
     "mailgun.js": "^3.7.3",
+    "microsoft-cognitiveservices-speech-sdk": "^1.24.1",
     "mysql2": "^2.3.3",
     "passport": "^0.6.0",
     "passport-http-bearer": "^1.0.1",

test/speech-credentials.js

@@ -30,7 +30,9 @@ test('speech credentials tests', async(t) => {
       json: true,
       body: {
         vendor: 'google',
-        service_key: jsonKey
+        service_key: jsonKey,
+        use_for_tts: true,
+        use_for_stt: true
       }
     });
     t.ok(result.statusCode === 201, 'successfully added a speech credential to service provider');
@@ -61,7 +63,9 @@ test('speech credentials tests', async(t) => {
       json: true,
       body: {
         vendor: 'google',
-        service_key: jsonKey
+        service_key: jsonKey,
+        use_for_tts: true,
+        use_for_stt: true
       }
     });
     t.ok(result.statusCode === 201, 'successfully added speech credential');
@@ -110,20 +114,20 @@ test('speech credentials tests', async(t) => {
     });
     t.ok(result.statusCode === 204, 'successfully deleted speech credential');
 
-    /* add a credential for microsoft */
+    /* add / test a credential for google */
-    if (process.env.MICROSOFT_API_KEY && process.env.MICROSOFT_REGION) {
+    if (process.env.GCP_JSON_KEY) {
       result = await request.post(`/Accounts/${account_sid}/SpeechCredentials`, {
         resolveWithFullResponse: true,
         auth: authUser,
         json: true,
         body: {
-          vendor: 'microsoft',
+          vendor: 'google',
           use_for_tts: true,
-          api_key: process.env.MICROSOFT_API_KEY,
+          use_for_stt: true,
-          region: process.env.MICROSOFT_REGION
+          service_key: process.env.GCP_JSON_KEY
         }
       });
-      t.ok(result.statusCode === 201, 'successfully added speech credential');
+      t.ok(result.statusCode === 201, 'successfully added speech credential for google');
       const ms_sid = result.body.sid;
 
       /* test the speech credential */
@@ -132,7 +136,37 @@ test('speech credentials tests', async(t) => {
         auth: authUser,
         json: true,   
       });
-      console.log(JSON.stringify(result));
+      //console.log(JSON.stringify(result));
+      t.ok(result.statusCode === 200 && result.body.tts.status === 'ok', 'successfully tested speech credential for google tts');
+      t.ok(result.statusCode === 200 && result.body.stt.status === 'ok', 'successfully tested speech credential for google stt');
+    }
+
+    /* add / test a credential for microsoft */
+    if (process.env.MICROSOFT_API_KEY && process.env.MICROSOFT_REGION) {
+      result = await request.post(`/Accounts/${account_sid}/SpeechCredentials`, {
+        resolveWithFullResponse: true,
+        auth: authUser,
+        json: true,
+        body: {
+          vendor: 'microsoft',
+          use_for_tts: true,
+          use_for_stt: true,
+          api_key: process.env.MICROSOFT_API_KEY,
+          region: process.env.MICROSOFT_REGION
+        }
+      });
+      t.ok(result.statusCode === 201, 'successfully added speech credential for microsoft');
+      const ms_sid = result.body.sid;
+
+      /* test the speech credential */
+      result = await request.get(`/Accounts/${account_sid}/SpeechCredentials/${ms_sid}/test`, {
+        resolveWithFullResponse: true,
+        auth: authUser,
+        json: true,   
+      });
+      //console.log(JSON.stringify(result));
+      t.ok(result.statusCode === 200 && result.body.tts.status === 'ok', 'successfully tested speech credential for microsoft tts');
+      t.ok(result.statusCode === 200 && result.body.stt.status === 'ok', 'successfully tested speech credential for microsoft stt');
     }
 
     /* add a credential for wellsaid */
@@ -156,7 +190,8 @@ test('speech credentials tests', async(t) => {
         auth: authUser,
         json: true,   
       });
-      console.log(JSON.stringify(result));
+      //console.log(JSON.stringify(result));
+      t.ok(result.statusCode === 200 && result.body.tts.status === 'ok', 'successfully tested speech credential for wellsaid');
 
       /* delete the credential */
       result = await request.delete(`/Accounts/${account_sid}/SpeechCredentials/${ms_sid}`, {