update to latest speech packages

add functions to retrieve voices for google and tts

lib/get-tts-voices.js

@@ -1,9 +1,11 @@
 const assert = require('assert');
-const {noopLogger, createNuanceClient} = require('./utils');
+const {noopLogger, createNuanceClient, createKryptonClient} = require('./utils');
 const getNuanceAccessToken = require('./get-nuance-access-token');
 const {GetVoicesRequest, Voice} = require('../stubs/nuance/synthesizer_pb');
 const TextToSpeechV1 = require('ibm-watson/text-to-speech/v1');
 const { IamAuthenticator } = require('ibm-watson/auth');
+const ttsGoogle = require('@google-cloud/text-to-speech');
+const { PollyClient, DescribeVoicesCommand } = require('@aws-sdk/client-polly');
 
 const getIbmVoices = async(client, logger, credentials) => {
   const {tts_region, tts_api_key} = credentials;
@@ -21,15 +23,20 @@ const getIbmVoices = async(client, logger, credentials) => {
 };
 
 const getNuanceVoices = async(client, logger, credentials) => {
-  const {client_id: clientId, secret: secret} = credentials;
+  const {client_id: clientId, secret: secret, nuance_tts_uri} = credentials;
 
   return new Promise(async(resolve, reject) => {
     /* get a nuance access token */
     let token, nuanceClient;
     try {
-      const access_token = await getNuanceAccessToken(client, logger, clientId, secret, 'tts');
-      token = access_token.access_token;
-      nuanceClient = await createNuanceClient(token);
+      if (nuance_tts_uri) {
+        nuanceClient = await createKryptonClient(nuance_tts_uri);
+      }
+      else {
+        const access_token = await getNuanceAccessToken(client, logger, clientId, secret, 'tts');
+        token = access_token.access_token;
+        nuanceClient = await createNuanceClient(token);
+      }
     } catch (err) {
       logger.error({err}, 'getTtsVoices: error retrieving access token');
       return reject(err);
@@ -75,6 +82,30 @@ const getNuanceVoices = async(client, logger, credentials) => {
   });
 };
 
+const getGoogleVoices = async(_client, logger, credentials) => {
+  const client = new ttsGoogle.TextToSpeechClient({credentials});
+  return await client.listVoices();
+};
+
+const getAwsVoices = async(_client, logger, credentials) => {
+  try {
+    const {region, accessKeyId, secretAccessKey} = credentials;
+    const client = new PollyClient({
+      region,
+      credentials: {
+        accessKeyId,
+        secretAccessKey
+      }
+    });
+    const command = new DescribeVoicesCommand({LanguageCode: 'en-US'});
+    const response = await client.send(command);
+    return response;
+  } catch (err) {
+    logger.info({err}, 'testMicrosoftTts - failed to list voices for region ${region}');
+    throw err;
+  }
+};
+
 /**
  * Synthesize speech to an mp3 file, and also cache the generated speech
  * in redis (base64 format) for 24 hours so as to avoid unnecessarily paying
@@ -94,7 +125,7 @@ const getNuanceVoices = async(client, logger, credentials) => {
 async function getTtsVoices(client, logger, {vendor, credentials}) {
   logger = logger || noopLogger;
 
-  assert.ok(['nuance', 'ibm'].includes(vendor),
+  assert.ok(['nuance', 'ibm', 'google', 'aws', 'polly'].includes(vendor),
     `getTtsVoices not supported for vendor ${vendor}`);
 
   switch (vendor) {
@@ -102,6 +133,11 @@ async function getTtsVoices(client, logger, {vendor, credentials}) {
       return getNuanceVoices(client, logger, credentials);
     case 'ibm':
       return getIbmVoices(client, logger, credentials);
+    case 'google':
+      return getGoogleVoices(client, logger, credentials);
+    case 'aws':
+    case 'polly':
+      return getAwsVoices(client, logger, credentials);
     default:
       break;
   }

lib/synth-audio.js

@@ -16,7 +16,7 @@ const {
   CancellationDetails,
   SpeechSynthesisOutputFormat
 } = sdk;
-const {makeSynthKey, createNuanceClient, noopLogger, createRivaClient} = require('./utils');
+const {makeSynthKey, createNuanceClient, createKryptonClient, createRivaClient, noopLogger} = require('./utils');
 const getNuanceAccessToken = require('./get-nuance-access-token');
 const {
   SynthesisRequest,
@@ -32,7 +32,7 @@ const {
 const {SynthesizeSpeechRequest} = require('../stubs/riva/proto/riva_tts_pb');
 const {AudioEncoding} = require('../stubs/riva/proto/riva_audio_pb');
 const debug = require('debug')('jambonz:realtimedb-helpers');
-const EXPIRES = process.env.JAMBONES_TTS_CACHE_DURATION_MINS || 3600 * 24; // cache tts for 24 hours
+const EXPIRES = (process.env.JAMBONES_TTS_CACHE_DURATION_MINS || 24 * 60) * 60; // cache tts for 24 hours
 const TMP_FOLDER = '/tmp';
 
 /**
@@ -75,8 +75,10 @@ async function synthAudio(client, logger, stats, { account_sid,
   }
   else if ('nuance' === vendor) {
     assert.ok(voice, 'synthAudio requires voice when nuance is used');
-    assert.ok(credentials.client_id, 'synthAudio requires client_id in credentials when nuance is used');
-    assert.ok(credentials.secret, 'synthAudio requires client_id in credentials when nuance is used');
+    if (!credentials.nuance_tts_uri) {
+      assert.ok(credentials.client_id, 'synthAudio requires client_id in credentials when nuance is used');
+      assert.ok(credentials.secret, 'synthAudio requires client_id in credentials when nuance is used');
+    }
   }
   else if ('nvidia' === vendor) {
     assert.ok(voice, 'synthAudio requires voice when nvidia is used');
@@ -155,7 +157,8 @@ async function synthAudio(client, logger, stats, { account_sid,
         audioBuffer = await synthWellSaid(logger, {credentials, stats, language, voice, text, filePath});
         break;
       case vendor.startsWith('custom') ? vendor : 'cant_match_value':
-        audioBuffer = await synthCustomVendor(logger, {credentials, stats, language, voice, text});
+        ({ audioBuffer, filePath } = await synthCustomVendor(logger,
+          {credentials, stats, language, voice, text, filePath}));
         break;
       default:
         assert(`synthAudio: unsupported speech vendor ${vendor}`);
@@ -183,7 +186,14 @@ async function synthAudio(client, logger, stats, { account_sid,
 
 const synthPolly = async(logger, {credentials, stats, language, voice, engine, text}) => {
   try {
-    const polly = new PollyClient(credentials);
+    const {region, accessKeyId, secretAccessKey} = credentials;
+    const polly = new PollyClient({
+      region,
+      credentials: {
+        accessKeyId,
+        secretAccessKey
+      }
+    });
     const opts = {
       Engine: engine,
       OutputFormat: 'mp3',
@@ -363,10 +373,16 @@ const synthWellSaid = async(logger, {credentials, stats, language, voice, gender
 };
 
 const synthNuance = async(client, logger, {credentials, stats, voice, model, text}) => {
-  /* get a nuance access token */
-  const {client_id, secret} = credentials;
-  const {access_token} = await getNuanceAccessToken(client, logger, client_id, secret, 'tts');
-  const nuanceClient = await createNuanceClient(access_token);
+  let nuanceClient;
+  const {client_id, secret, nuance_tts_uri} = credentials;
+  if (nuance_tts_uri) {
+    nuanceClient = await createKryptonClient(nuance_tts_uri);
+  }
+  else {
+    /* get a nuance access token */
+    const {access_token} = await getNuanceAccessToken(client, logger, client_id, secret, 'tts');
+    nuanceClient = await createNuanceClient(access_token);
+  }
 
   const v = new Voice();
   const p = new AudioParameters();
@@ -428,7 +444,6 @@ const synthNvidia = async(client, logger, {credentials, stats, language,  voice,
   request.setText(text);
 
   return new Promise((resolve, reject) => {
-    console.log(`language ${language} voice ${voice} model ${model} text ${text}`);
     rivaClient.synthesize(request, (err, response) => {
       if (err) {
         console.error(err);
@@ -440,30 +455,56 @@ const synthNvidia = async(client, logger, {credentials, stats, language,  voice,
 };
 
 
-// CustomVendor accept only mp3
-const synthCustomVendor = async(logger, {credentials, stats, language, voice, text}) => {
+const synthCustomVendor = async(logger, {credentials, stats, language, voice, text, filePath}) => {
   const {vendor, auth_token, custom_tts_url} = credentials;
 
   try {
-    const post = bent('POST', 'buffer', {
+    const post = bent('POST', {
       'Authorization': `Bearer ${auth_token}`,
-      'Accept': 'audio/mpeg',
       'Content-Type': 'application/json'
     });
 
-    const mp3 = await post(custom_tts_url, {
+    const response = await post(custom_tts_url, {
       language,
-      format: 'audio/mpeg',
       voice,
       type: text.startsWith('<speak>') ? 'ssml' : 'text',
       text
     });
 
-    return mp3;
+    const regex = /\.[^\.]*$/g;
+    const mime = response.headers['content-type'];
+    const buffer = await response.arrayBuffer();
+    return {
+      audioBuffer: buffer,
+      filePath: filePath.replace(regex, getFileExtFromMime(mime))
+    };
   } catch (err) {
     logger.info({err}, `Vendor ${vendor} returned error`);
     throw err;
   }
 };
 
+const getFileExtFromMime = (mime) => {
+  switch (mime) {
+    case 'audio/wav':
+    case 'audio/x-wav':
+      return '.wav';
+    case /audio\/l16.*rate=8000/.test(mime) ? mime : 'cant match value':
+      return '.r8';
+    case /audio\/l16.*rate=16000/.test(mime) ? mime : 'cant match value':
+      return '.r16';
+    case /audio\/l16.*rate=24000/.test(mime) ? mime : 'cant match value':
+      return '.r24';
+    case /audio\/l16.*rate=32000/.test(mime) ? mime : 'cant match value':
+      return '.r32';
+    case /audio\/l16.*rate=48000/.test(mime) ? mime : 'cant match value':
+      return '.r48';
+    case 'audio/mpeg':
+    case 'audio/mp3':
+      return '.mp3';
+    default:
+      return '.wav';
+  }
+};
+
 module.exports = synthAudio;

lib/utils.js

@@ -77,6 +77,11 @@ const getNuanceAccessToken = async(clientId, secret, scope = 'asr tts') => {
   return json.access_token;
 };
 
+const createKryptonClient = async(uri) => {
+  const client = new SynthesizerClient(uri, grpc.credentials.createInsecure());
+  return client;
+};
+
 const createNuanceClient = async(access_token) => {
 
   //if (nuanceClientMap.has(access_token)) return nuanceClientMap.get(access_token);
@@ -108,6 +113,7 @@ module.exports = {
   makeIbmKey,
   getNuanceAccessToken,
   createNuanceClient,
+  createKryptonClient,
   createRivaClient,
   makeBasicAuthHeader,
   NUANCE_AUTH_ENDPOINT,

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.2",
+  "version": "0.0.9",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",
@@ -24,18 +24,18 @@
   },
   "homepage": "https://github.com/jambonz/speech-utils#readme",
   "dependencies": {
-    "@aws-sdk/client-polly": "^3.276.0",
-    "@google-cloud/text-to-speech": "^4.2.0",
-    "@grpc/grpc-js": "^1.8.7",
-    "@jambonz/realtimedb-helpers": "^0.6.3",
-    "aws-sdk": "^2.1310.0",
+    "@aws-sdk/client-polly": "^3.303.0",
+    "@google-cloud/text-to-speech": "^4.2.1",
+    "@grpc/grpc-js": "^1.8.13",
+    "@jambonz/promisify-redis": "^0.0.6",
     "bent": "^7.3.12",
     "debug": "^4.3.4",
-    "google-protobuf": "^3.21.2",
-    "ibm-watson": "^7.1.2",
     "form-urlencoded": "^6.1.0",
-    "microsoft-cognitiveservices-speech-sdk": "^1.25.0",
-    "undici": "^5.19.1"
+    "google-protobuf": "^3.21.2",
+    "ibm-watson": "^8.0.0",
+    "microsoft-cognitiveservices-speech-sdk": "^1.26.0",
+    "redis": "^3.1.2",
+    "undici": "^5.21.0"
   },
   "devDependencies": {
     "config": "^3.3.9",

test/index.js

@@ -1,5 +1,4 @@
 require('./docker_start');
 require('./synth');
-require('./nuance');
-require('./ibm');
+require('./list-voices');
 require('./docker_stop');

test/list-voices.js

@@ -0,0 +1,205 @@
+const test = require('tape').test ;
+const config = require('config');
+const opts = config.get('redis');
+const fs = require('fs');
+const logger = require('pino')({level: 'error'});
+process.on('unhandledRejection', (reason, p) => {
+  console.log('Unhandled Rejection at: Promise', p, 'reason:', reason);
+});
+
+const stats = {
+  increment: () => {},
+  histogram: () => {}
+};
+
+test('IBM - create access key', async(t) => {
+  const fn = require('..');
+  const {client, getIbmAccessToken} = fn(opts, logger);
+
+  if (!process.env.IBM_API_KEY ) {
+      t.pass('skipping IBM test since no IBM api_key provided');
+      t.end();
+      client.quit();
+      return;
+  }
+  try {
+    let obj = await getIbmAccessToken(process.env.IBM_API_KEY);
+    //console.log({obj}, 'received access token from IBM');
+    t.ok(obj.access_token && !obj.servedFromCache, 'successfull received access token from IBM');
+
+    obj = await getIbmAccessToken(process.env.IBM_API_KEY);
+    //console.log({obj}, 'received access token from IBM - second request');
+    t.ok(obj.access_token && obj.servedFromCache, 'successfully received access token from cache');
+ 
+    await client.flushallAsync();
+    t.end();
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('IBM - retrieve tts voices test', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices} = fn(opts, logger);
+
+  if (!process.env.IBM_TTS_API_KEY || !process.env.IBM_TTS_REGION) {
+      t.pass('skipping IBM test since no IBM api_key and/or region provided');
+      t.end();
+      client.quit();
+      return;
+  }
+  try {
+    const opts = {
+      vendor: 'ibm',
+      credentials: {
+        tts_api_key: process.env.IBM_TTS_API_KEY,
+        tts_region: process.env.IBM_TTS_REGION
+      }
+    };
+    const obj = await getTtsVoices(opts);
+    const {voices} = obj.result;
+    //console.log(JSON.stringify(voices));
+    t.ok(voices.length > 0 && voices[0].language, 
+      `GetVoices: successfully retrieved ${voices.length} voices from IBM`);
+ 
+    await client.flushallAsync();
+
+    t.end();
+
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('Nuance hosted tests', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices} = fn(opts, logger);
+
+  if (!process.env.NUANCE_CLIENT_ID || !process.env.NUANCE_SECRET ) {
+      t.pass('skipping Nuance hosted test since no Nuance client_id and secret provided');
+      t.end();
+      client.quit();
+      return;
+  }
+  try {
+    const opts = {
+      vendor: 'nuance',
+      credentials: {
+        client_id: process.env.NUANCE_CLIENT_ID,
+        secret: process.env.NUANCE_SECRET
+      }
+    };
+    let voices = await getTtsVoices(opts);
+    t.ok(voices.length > 0 && voices[0].language, 
+      `GetVoices: successfully retrieved ${voices.length} voices from Nuance`);
+
+    await client.flushallAsync();
+
+    t.end();
+
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('Nuance on-prem tests', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices} = fn(opts, logger);
+
+  if (!process.env.NUANCE_TTS_URI ) {
+      t.pass('skipping Nuance on-prem test since no Nuance uri provided');
+      t.end();
+      client.quit();
+      return;
+  }
+  try {
+    const opts = {
+      vendor: 'nuance',
+      credentials: {
+        nuance_tts_uri: process.env.NUANCE_TTS_URI
+      }
+    };
+    let voices = await getTtsVoices(opts);
+    t.ok(voices.length > 0 && voices[0].language, 
+      `GetVoices: successfully retrieved ${voices.length} voices from Nuance`);
+
+    await client.flushallAsync();
+
+    t.end();
+
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('Google tests', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices} = fn(opts, logger);
+
+  if (!process.env.GCP_FILE && !process.env.GCP_JSON_KEY) {
+    t.pass('skipping google speech synth tests since neither GCP_FILE nor GCP_JSON_KEY provided');
+    return t.end();
+  }
+  try {
+    const str = process.env.GCP_JSON_KEY || fs.readFileSync(process.env.GCP_FILE);
+    const credentials = JSON.parse(str);
+    const opts = {
+      vendor: 'google',
+      credentials
+    };
+    let result = await getTtsVoices(opts);
+    t.ok(result[0].voices.length > 0, `GetVoices: successfully retrieved ${result[0].voices.length} voices from Google`);
+
+    await client.flushallAsync();
+
+    t.end();
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('AWS tests', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices} = fn(opts, logger);
+
+  if (!process.env.AWS_ACCESS_KEY_ID || !process.env.AWS_SECRET_ACCESS_KEY || !process.env.AWS_REGION) {
+    t.pass('skipping AWS speech synth tests since AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, or AWS_REGION not provided');
+    return t.end();
+  }
+  try {
+    const opts = {
+      vendor: 'aws',
+      credentials: {
+        accessKeyId: process.env.AWS_ACCESS_KEY_ID,
+        secretAccessKey: process.env.AWS_SECRET_ACCESS_KEY,
+        region: process.env.AWS_REGION,
+      }
+    };
+    let result = await getTtsVoices(opts);
+    t.ok(result?.Voices?.length > 0, `GetVoices: successfully retrieved ${result.Voices.length} voices from AWS`);
+
+    await client.flushallAsync();
+
+    t.end();
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});

test/nuance.js

@@ -12,12 +12,12 @@ const stats = {
   histogram: () => {}
 };
 
-test('Nuance tests', async(t) => {
+test('Nuance hosted tests', async(t) => {
   const fn = require('..');
   const {client, getTtsVoices} = fn(opts, logger);
 
   if (!process.env.NUANCE_CLIENT_ID || !process.env.NUANCE_SECRET ) {
-      t.pass('skipping Nuance test since no Nuance client_id and secret provided');
+      t.pass('skipping Nuance hosted test since no Nuance client_id and secret provided');
       t.end();
       client.quit();
       return;
@@ -31,8 +31,39 @@ test('Nuance tests', async(t) => {
       }
     };
     let voices = await getTtsVoices(opts);
-    //console.log(`received ${voices.length} voices from Nuance`);
-    //console.log(JSON.stringify(voices));
+    t.ok(voices.length > 0 && voices[0].language, 
+      `GetVoices: successfully retrieved ${voices.length} voices from Nuance`);
+
+    await client.flushallAsync();
+
+    t.end();
+
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('Nuance on-prem tests', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices} = fn(opts, logger);
+
+  if (!process.env.NUANCE_TTS_URI ) {
+      t.pass('skipping Nuance on-prem test since no Nuance uri provided');
+      t.end();
+      client.quit();
+      return;
+  }
+  try {
+    const opts = {
+      vendor: 'nuance',
+      credentials: {
+        nuance_tts_uri: process.env.NUANCE_TTS_URI
+      }
+    };
+    let voices = await getTtsVoices(opts);
     t.ok(voices.length > 0 && voices[0].language, 
       `GetVoices: successfully retrieved ${voices.length} voices from Nuance`);
 

test/synth.js

@@ -212,7 +212,7 @@ test('Azure custom voice speech synth tests', async(t) => {
   client.quit();
 });
 
-test('Nuance speech synth tests', async(t) => {
+test('Nuance hosted speech synth tests', async(t) => {
   const fn = require('..');
   const {synthAudio, client} = fn(opts, logger);
 
@@ -251,6 +251,43 @@ test('Nuance speech synth tests', async(t) => {
   client.quit();
 });
 
+test('Nuance on-prem speech synth tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.NUANCE_TTS_URI) {
+    t.pass('skipping Nuance on prem speech synth tests since NUANCE_TTS_URI not provided');
+    return t.end();
+  }
+  try {
+    let opts = await synthAudio(stats, {
+      vendor: 'nuance',
+      credentials: {
+        nuance_tts_uri: process.env.NUANCE_TTS_URI
+      },
+      language: 'en-US',
+      voice: 'Evan',
+      text: 'This is a test of on-prem.  This is only a test',
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized nuance audio to ${opts.filePath}`);
+
+    opts = await synthAudio(stats, {
+      vendor: 'nuance',
+      credentials: {
+        nuance_tts_uri: process.env.NUANCE_TTS_URI
+      },
+      language: 'en-US',
+      voice: 'Evan',
+      text: 'This is a test of on-prem.  This is only a test',
+    });
+    t.ok(opts.servedFromCache, `successfully retrieved nuance audio from cache ${opts.filePath}`);
+  } catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
 test('Nvidia speech synth tests', async(t) => {
   const fn = require('..');
   const {synthAudio, client} = fn(opts, logger);
@@ -348,7 +385,6 @@ test('Custom Vendor speech synth tests', async(t) => {
     let obj = await getJSON(`http://127.0.0.1:3100/lastRequest/somethingnew`);
     t.ok(obj.headers.Authorization == 'Bearer some_jwt_token', 'Custom Vendor Authentication Header is correct');
     t.ok(obj.body.language == 'en-US', 'Custom Vendor Language is correct');
-    t.ok(obj.body.format == 'audio/mpeg', 'Custom Vendor format is correct');
     t.ok(obj.body.voice == 'English-US.Female-1', 'Custom Vendor voice is correct');
     t.ok(obj.body.type == 'text', 'Custom Vendor type is correct');
     t.ok(obj.body.text == 'This is a test.  This is only a test', 'Custom Vendor text is correct');