add support for nuance tts on-prem

add support for Nuance TTS on-prem (vs hosted)

lib/get-tts-voices.js

@@ -1,5 +1,5 @@
 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');
@@ -21,15 +21,20 @@ 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);

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();
@@ -440,30 +456,56 @@ 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-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.2",
+  "version": "0.0.7",
   "lockfileVersion": 2,
   "requires": true,
   "packages": {
     "": {
       "name": "@jambonz/speech-utils",
-      "version": "0.0.2",
+      "version": "0.0.7",
       "license": "MIT",
       "dependencies": {
         "@aws-sdk/client-polly": "^3.276.0",

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.2",
+  "version": "0.0.7",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",

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');