0.1.2

Improve handling of TTS cache by adding the file extension to the cac…

.github/workflows/ci.yml

@@ -8,12 +8,15 @@ jobs:
   build:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
-      - uses: actions/setup-node@v3
+      - uses: actions/setup-node@v4
         with:
           node-version: lts/*
       - run: npm install
       - run: npm run jslint
+      - run: sudo apt update && sudo apt install -y squid
+      - run: sudo cp test/squid.conf /etc/squid/squid.conf
+      - run: sudo systemctl start squid
       - run: npm test
         env:
           AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
@@ -24,4 +27,12 @@ jobs:
           IBM_TTS_API_KEY: ${{ secrets.IBM_TTS_API_KEY }}
           IBM_TTS_REGION: ${{ secrets.IBM_TTS_REGION }}
           MICROSOFT_API_KEY: ${{ secrets.MICROSOFT_API_KEY }}
-          MICROSOFT_REGION: ${{ secrets.MICROSOFT_REGION }}
+          MICROSOFT_REGION: ${{ secrets.MICROSOFT_REGION }}
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+          ELEVENLABS_API_KEY: ${{ secrets.ELEVENLABS_API_KEY }}
+          ELEVENLABS_VOICE_ID: ${{ secrets.ELEVENLABS_VOICE_ID }}
+          ELEVENLABS_MODEL_ID: ${{ secrets.ELEVENLABS_MODEL_ID }}
+          PLAYHT_USER_ID: ${{ secrets.PLAYHT_USER_ID }}
+          PLAYHT_API_KEY: ${{ secrets.PLAYHT_API_KEY }}
+          JAMBONES_HTTP_PROXY_IP: 127.0.0.1
+          JAMBONES_HTTP_PROXY_PORT: 3128

.gitignore

@@ -39,3 +39,5 @@ node_modules
 examples/*
 
 .vscode
+
+.env

README.md

@@ -1,2 +1,3 @@
-# speech-utils
+# speech-utils ![CI](https://github.com/jambonz/speech-utils/workflows/CI/badge.svg)
-TTS-related speech utilities for jambonz
+TTS-related speech utilities for jambonz.
+

index.js

@@ -1,41 +1,23 @@
 const {noopLogger} = require('./lib/utils');
-const Redis = require('ioredis');
 
 module.exports = (opts, logger) => {
   logger = logger || noopLogger;
-  const connectionOpts = {...opts};
+  const {
-  // Support legacy app
+    client,
-  if (process.env.JAMBONES_REDIS_USERNAME && process.env.JAMBONES_REDIS_PASSWORD) {
+    createHash,
-    if (Array.isArray(connectionOpts)) {
+    retrieveHash
-      for (const o of opts) {
+  } = require('@jambonz/realtimedb-helpers')(opts, logger);
-        o.username = process.env.JAMBONES_REDIS_USERNAME;
-        o.password = process.env.JAMBONES_REDIS_PASSWORD;
-      }
-    } else {
-      connectionOpts.username = process.env.JAMBONES_REDIS_USERNAME;
-      connectionOpts.password = process.env.JAMBONES_REDIS_PASSWORD;
-    }
-  }
-
-  const client = new Redis(connectionOpts);
-  ['ready', 'connect', 'reconnecting', 'error', 'end', 'warning']
-    .forEach((event) => {
-      client.on(event, (...args) => {
-        if ('error' === event) {
-          if (process.env.NODE_ENV === 'test' && args[0]?.code === 'ECONNREFUSED') return;
-          logger.error({...args}, '@jambonz/realtimedb-helpers - redis error');
-        }
-        else logger.debug({args}, `redis event ${event}`);
-      });
-    });
 
   return {
     client,
     getTtsSize: require('./lib/get-tts-size').bind(null, client, logger),
     purgeTtsCache: require('./lib/purge-tts-cache').bind(null, client, logger),
-    synthAudio: require('./lib/synth-audio').bind(null, client, logger),
+    addFileToCache: require('./lib/add-file-to-cache').bind(null, client, logger),
+    synthAudio: require('./lib/synth-audio').bind(null, client, createHash, retrieveHash, logger),
+    getVerbioAccessToken: require('./lib/get-verbio-token').bind(null, client, logger),
     getNuanceAccessToken: require('./lib/get-nuance-access-token').bind(null, client, logger),
     getIbmAccessToken: require('./lib/get-ibm-access-token').bind(null, client, logger),
-    getTtsVoices: require('./lib/get-tts-voices').bind(null, client, logger),
+    getAwsAuthToken: require('./lib/get-aws-sts-token').bind(null, logger, createHash, retrieveHash),
+    getTtsVoices: require('./lib/get-tts-voices').bind(null, client, createHash, retrieveHash, logger),
   };
 };

lib/add-file-to-cache.js

@@ -0,0 +1,31 @@
+const fs = require('fs/promises');
+const {noopLogger, makeSynthKey} = require('./utils');
+const {JAMBONES_TTS_CACHE_DURATION_MINS} = require('./config');
+const EXPIRES = JAMBONES_TTS_CACHE_DURATION_MINS;
+
+async function addFileToCache(client, logger, path,
+  {account_sid, vendor, language, voice, deploymentId, engine, text}) {
+  let key;
+  logger = logger || noopLogger;
+
+  try {
+    key = makeSynthKey({
+      account_sid,
+      vendor,
+      language: language || '',
+      voice: voice || deploymentId,
+      engine,
+      text,
+    });
+    const audioBuffer = await fs.readFile(path);
+    await client.setex(key, EXPIRES, audioBuffer.toString('base64'));
+  } catch (err) {
+    logger.error(err, 'addFileToCache: Error');
+    return;
+  }
+
+  logger.debug(`addFileToCache: added ${path} to cache with key ${key}`);
+  return key;
+}
+
+module.exports = addFileToCache;

lib/config.js

@@ -0,0 +1,20 @@
+const JAMBONES_TTS_TRIM_SILENCE = process.env.JAMBONES_TTS_TRIM_SILENCE;
+const JAMBONES_DISABLE_TTS_STREAMING = process.env.JAMBONES_DISABLE_TTS_STREAMING;
+
+const JAMBONES_HTTP_PROXY_IP = process.env.JAMBONES_HTTP_PROXY_IP;
+const JAMBONES_HTTP_PROXY_PORT = process.env.JAMBONES_HTTP_PROXY_PORT;
+const JAMBONES_TTS_CACHE_DURATION_MINS = (parseInt(process.env.JAMBONES_TTS_CACHE_DURATION_MINS) || 4 * 60) * 60; // cache tts for 4 hours
+
+const TMP_FOLDER = '/tmp';
+
+const HTTP_TIMEOUT = 5000;
+
+module.exports = {
+  JAMBONES_TTS_TRIM_SILENCE,
+  JAMBONES_DISABLE_TTS_STREAMING,
+  JAMBONES_HTTP_PROXY_IP,
+  JAMBONES_HTTP_PROXY_PORT,
+  JAMBONES_TTS_CACHE_DURATION_MINS,
+  TMP_FOLDER,
+  HTTP_TIMEOUT
+};

lib/get-aws-sts-token.js

@@ -0,0 +1,54 @@
+const { STSClient, GetSessionTokenCommand, AssumeRoleCommand } = require('@aws-sdk/client-sts');
+const {makeAwsKey, noopLogger} = require('./utils');
+const debug = require('debug')('jambonz:speech-utils');
+const EXPIRY = 3600;
+
+async function getAwsAuthToken(
+  logger, createHash, retrieveHash,
+  awsAccessKeyId, awsSecretAccessKey, awsRegion, roleArn = null) {
+  logger = logger || noopLogger;
+  try {
+    const key = makeAwsKey(roleArn || awsAccessKeyId);
+    const obj = await retrieveHash(key);
+    if (obj) return {...obj, servedFromCache: true};
+
+    let data;
+    if (roleArn) {
+      const stsClient = new STSClient({ region: awsRegion});
+      const roleToAssume = { RoleArn: roleArn, RoleSessionName: 'Jambonz_Speech', DurationSeconds: EXPIRY};
+      const command = new AssumeRoleCommand(roleToAssume);
+
+      data = await stsClient.send(command);
+    } else {
+      /* access token not found in cache, so generate it using STS */
+      const stsClient = new STSClient({
+        region: awsRegion,
+        credentials: {
+          accessKeyId: awsAccessKeyId,
+          secretAccessKey: awsSecretAccessKey,
+        }
+      });
+      const command = new GetSessionTokenCommand({DurationSeconds: EXPIRY});
+      data = await stsClient.send(command);
+    }
+
+    const credentials = {
+      accessKeyId: data.Credentials.AccessKeyId,
+      secretAccessKey: data.Credentials.SecretAccessKey,
+      sessionToken: data.Credentials.SessionToken,
+      securityToken: data.Credentials.SessionToken
+    };
+
+    /* expire 10 minutes before the hour, so we don't lose the use of it during a call */
+    createHash(key, credentials, EXPIRY - 600)
+      .catch((err) => logger.error(err, `Error saving hash for key ${key}`));
+
+    return {...credentials, servedFromCache: false};
+  } catch (err) {
+    debug(err, 'getAwsAuthToken: Error retrieving AWS auth token');
+    logger.error(err, 'getAwsAuthToken: Error retrieving AWS auth token');
+    throw err;
+  }
+}
+
+module.exports = getAwsAuthToken;

lib/get-ibm-access-token.js

@@ -2,8 +2,8 @@ const formurlencoded = require('form-urlencoded');
 const {Pool} = require('undici');
 const pool = new Pool('https://iam.cloud.ibm.com');
 const {makeIbmKey, noopLogger} = require('./utils');
+const { HTTP_TIMEOUT } = require('./config');
 const debug = require('debug')('jambonz:realtimedb-helpers');
-const HTTP_TIMEOUT = 5000;
 
 async function getIbmAccessToken(client, logger, apiKey) {
   logger = logger || noopLogger;

lib/get-nuance-access-token.js

@@ -2,8 +2,8 @@ const formurlencoded = require('form-urlencoded');
 const {Pool} = require('undici');
 const pool = new Pool('https://auth.crt.nuance.com');
 const {makeNuanceKey, makeBasicAuthHeader, noopLogger} = require('./utils');
+const { HTTP_TIMEOUT } = require('./config');
 const debug = require('debug')('jambonz:realtimedb-helpers');
-const HTTP_TIMEOUT = 5000;
 
 async function getNuanceAccessToken(client, logger, clientId, secret, scope) {
   logger = logger || noopLogger;

lib/get-tts-voices.js

@@ -1,11 +1,16 @@
 const assert = require('assert');
 const {noopLogger, createNuanceClient, createKryptonClient} = require('./utils');
 const getNuanceAccessToken = require('./get-nuance-access-token');
+const getVerbioAccessToken = require('./get-verbio-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 getAwsAuthToken = require('./get-aws-sts-token');
+const {Pool} = require('undici');
+const { HTTP_TIMEOUT } = require('./config');
+const verbioVoicePool = new Pool('https://us.rest.speechcenter.verbio.com');
 
 const getIbmVoices = async(client, logger, credentials) => {
   const {tts_region, tts_api_key} = credentials;
@@ -87,17 +92,27 @@ const getGoogleVoices = async(_client, logger, credentials) => {
   return await client.listVoices();
 };
 
-const getAwsVoices = async(_client, logger, credentials) => {
+const getAwsVoices = async(_client, createHash, retrieveHash, logger, credentials) => {
   try {
-    const {region, accessKeyId, secretAccessKey} = credentials;
+    const {region, accessKeyId, secretAccessKey, roleArn} = credentials;
-    const client = new PollyClient({
+    let client = null;
-      region,
+    if (accessKeyId && secretAccessKey) {
-      credentials: {
+      client = new PollyClient({
-        accessKeyId,
+        region,
-        secretAccessKey
+        credentials: {
-      }
+          accessKeyId,
-    });
+          secretAccessKey
-    const command = new DescribeVoicesCommand({LanguageCode: 'en-US'});
+        }
+      });
+    } else if (roleArn) {
+      client = new PollyClient({
+        region,
+        credentials: await getAwsAuthToken(logger, createHash, retrieveHash, null, null, region, roleArn),
+      });
+    } else {
+      client = new PollyClient({region});
+    }
+    const command = new DescribeVoicesCommand({});
     const response = await client.send(command);
     return response;
   } catch (err) {
@@ -106,6 +121,26 @@ const getAwsVoices = async(_client, logger, credentials) => {
   }
 };
 
+const getVerbioVoices = async(client, logger, credentials) => {
+  try {
+    const access_token = await getVerbioAccessToken(client, logger, credentials);
+    const { body} =  await verbioVoicePool.request({
+      path: '/api/v1/voices',
+      method: 'GET',
+      headers: {
+        'Authorization': `Bearer ${access_token.access_token}`,
+        'User-Agent': 'jambonz'
+      },
+      timeout: HTTP_TIMEOUT,
+      followRedirects: false
+    });
+    return await body.json();
+  } catch (err) {
+    logger.info({err}, 'getVerbioVoices - failed to list voices for Verbio');
+    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
@@ -122,10 +157,10 @@ const getAwsVoices = async(_client, logger, credentials) => {
  * @returns object containing filepath to an mp3 file in the /tmp folder containing
  * the synthesized audio, and a variable indicating whether it was served from cache
  */
-async function getTtsVoices(client, logger, {vendor, credentials}) {
+async function getTtsVoices(client, createHash, retrieveHash, logger, {vendor, credentials}) {
   logger = logger || noopLogger;
 
-  assert.ok(['nuance', 'ibm', 'google', 'aws', 'polly'].includes(vendor),
+  assert.ok(['nuance', 'ibm', 'google', 'aws', 'polly', 'verbio'].includes(vendor),
     `getTtsVoices not supported for vendor ${vendor}`);
 
   switch (vendor) {
@@ -137,7 +172,9 @@ async function getTtsVoices(client, logger, {vendor, credentials}) {
       return getGoogleVoices(client, logger, credentials);
     case 'aws':
     case 'polly':
-      return getAwsVoices(client, logger, credentials);
+      return getAwsVoices(client, createHash, retrieveHash, logger, credentials);
+    case 'verbio':
+      return getVerbioVoices(client, logger, credentials);
     default:
       break;
   }

lib/get-verbio-token.js

@@ -0,0 +1,51 @@
+const {Pool} = require('undici');
+const { noopLogger, makeVerbioKey } = require('./utils');
+const { HTTP_TIMEOUT } = require('./config');
+const pool = new Pool('https://auth.speechcenter.verbio.com:444');
+const debug = require('debug')('jambonz:realtimedb-helpers');
+
+async function getVerbioAccessToken(client, logger, credentials) {
+  logger = logger || noopLogger;
+  const { client_id, client_secret } = credentials;
+  try {
+    const key = makeVerbioKey(client_id);
+    const access_token = await client.get(key);
+    if (access_token) {
+      return {access_token, servedFromCache: true};
+    }
+
+    const payload = {
+      client_id,
+      client_secret
+    };
+
+    const {statusCode, headers, body} =  await pool.request({
+      path: '/api/v1/token',
+      method: 'POST',
+      headers: {
+        'Content-Type': 'application/json',
+        'User-Agent': 'jambonz'
+      },
+      body: JSON.stringify(payload),
+      timeout: HTTP_TIMEOUT,
+      followRedirects: false
+    });
+
+    if (200 !== statusCode) {
+      logger.debug({statusCode, headers, body: await body.text()}, 'error fetching access token from Verbio');
+      const err = new Error();
+      err.statusCode = statusCode;
+      throw err;
+    }
+    const json = await body.json();
+    const expiry =  Math.floor(json.expiration_time - Date.now() / 1000 - 30);
+    await client.set(key, json.access_token, 'EX', expiry);
+    return {...json, servedFromCache: false};
+  } catch (err) {
+    debug(err, `getVerbioAccessToken: Error retrieving Verbio access token for client_id ${client_id}`);
+    logger.error(err, `getVerbioAccessToken: Error retrieving Verbio access token for client_id ${client_id}`);
+    throw err;
+  }
+}
+
+module.exports = getVerbioAccessToken;

lib/synth-audio.js

@@ -19,9 +19,11 @@ const {
   createNuanceClient,
   createKryptonClient,
   createRivaClient,
-  noopLogger
+  noopLogger,
+  makeFilePath
 } = require('./utils');
 const getNuanceAccessToken = require('./get-nuance-access-token');
+const getVerbioAccessToken = require('./get-verbio-token');
 const {
   SynthesisRequest,
   Voice,
@@ -36,9 +38,15 @@ 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 || 4 * 60) * 60; // cache tts for 4 hours
+const {
-const TMP_FOLDER = '/tmp';
+  JAMBONES_DISABLE_TTS_STREAMING,
+  JAMBONES_HTTP_PROXY_IP,
+  JAMBONES_HTTP_PROXY_PORT,
+  JAMBONES_TTS_CACHE_DURATION_MINS,
+} = require('./config');
+const EXPIRES = JAMBONES_TTS_CACHE_DURATION_MINS;
 const OpenAI = require('openai');
+const getAwsAuthToken = require('./get-aws-sts-token');
 
 
 const trimTrailingSilence = (buffer) => {
@@ -75,18 +83,19 @@ const trimTrailingSilence = (buffer) => {
  * @returns object containing filepath to an mp3 file in the /tmp folder containing
  * the synthesized audio, and a variable indicating whether it was served from cache
  */
-async function synthAudio(client, logger, stats, { account_sid,
+async function synthAudio(client, createHash, retrieveHash, logger, stats, { account_sid,
-  vendor, language, voice, gender, text, engine, salt, model, credentials, deploymentId, disableTtsCache
+  vendor, language, voice, gender, text, engine, salt, model, credentials, deploymentId,
+  disableTtsCache, renderForCaching, disableTtsStreaming, options
 }) {
   let audioBuffer;
   let servedFromCache = false;
   let rtt;
   logger = logger || noopLogger;
 
-  assert.ok(['google', 'aws', 'polly', 'microsoft',
+  assert.ok(['google', 'aws', 'polly', 'microsoft', 'wellsaid', 'nuance', 'nvidia', 'ibm', 'elevenlabs',
-    'wellsaid', 'nuance', 'nvidia', 'ibm', 'elevenlabs', 'whisper'].includes(vendor) ||
+    'whisper', 'deepgram', 'playht', 'rimelabs', 'verbio'].includes(vendor) ||
   vendor.startsWith('custom'),
-  `synthAudio supported vendors are google, aws, microsoft, nuance, nvidia and wellsaid, not ${vendor}`);
+  `synthAudio supported vendors are google, aws, microsoft, nuance, nvidia and wellsaid ..etc, not ${vendor}`);
   if ('google' === vendor) {
     assert.ok(language, 'synthAudio requires language when google is used');
   }
@@ -122,12 +131,25 @@ async function synthAudio(client, logger, stats, { account_sid,
     assert.ok(voice, 'synthAudio requires voice when elevenlabs is used');
     assert.ok(credentials.api_key, 'synthAudio requires api_key when elevenlabs is used');
     assert.ok(credentials.model_id, 'synthAudio requires model_id when elevenlabs is used');
+  } else if ('playht' === vendor) {
+    assert.ok(voice, 'synthAudio requires voice when playht is used');
+    assert.ok(credentials.api_key, 'synthAudio requires api_key when playht is used');
+    assert.ok(credentials.user_id, 'synthAudio requires user_id when playht is used');
+    assert.ok(credentials.voice_engine, 'synthAudio requires voice_engine when playht is used');
+  } else if ('rimelabs' === vendor) {
+    assert.ok(voice, 'synthAudio requires voice when rimelabs is used');
+    assert.ok(credentials.api_key, 'synthAudio requires api_key when rimelabs is used');
+    assert.ok(credentials.model_id, 'synthAudio requires model_id when rimelabs is used');
   } else if ('whisper' === vendor) {
     assert.ok(voice, 'synthAudio requires voice when whisper is used');
     assert.ok(credentials.model_id, 'synthAudio requires model when whisper is used');
     assert.ok(credentials.api_key, 'synthAudio requires api_key when whisper is used');
   } else  if (vendor.startsWith('custom')) {
     assert.ok(credentials.custom_tts_url, `synthAudio requires custom_tts_url in credentials when ${vendor} is used`);
+  } else if ('verbio' === vendor) {
+    assert.ok(voice, 'synthAudio requires voice when verbio is used');
+    assert.ok(credentials.client_id, 'synthAudio requires client_id when verbio is used');
+    assert.ok(credentials.client_secret, 'synthAudio requires client_secret when verbio is used');
   }
   const key = makeSynthKey({
     account_sid,
@@ -138,15 +160,7 @@ async function synthAudio(client, logger, stats, { account_sid,
     text
   });
   let filePath;
-  if (['nuance', 'nvidia'].includes(vendor) ||
+  filePath = makeFilePath(vendor, key, salt);
-    (
-      process.env.JAMBONES_TTS_TRIM_SILENCE &&
-      ['microsoft', 'azure'].includes(vendor)
-    )
-  ) {
-    filePath = `${TMP_FOLDER}/${key.replace('tts:', `tts-${salt || ''}`)}.r8`;
-  }
-  else filePath = `${TMP_FOLDER}/${key.replace('tts:', `tts-${salt || ''}`)}.mp3`;
   debug(`synth key is ${key}`);
   let cached;
   if (!disableTtsCache) {
@@ -173,12 +187,14 @@ async function synthAudio(client, logger, stats, { account_sid,
       case 'aws':
       case 'polly':
         vendorLabel = 'aws';
-        audioBuffer = await synthPolly(logger, {credentials, stats, language, voice, text, engine});
+        audioBuffer = await synthPolly(createHash, retrieveHash, logger,
+          {credentials, stats, language, voice, text, engine});
         break;
       case 'azure':
       case 'microsoft':
         vendorLabel = 'microsoft';
-        audioBuffer = await synthMicrosoft(logger, {credentials, stats, language, voice, text, deploymentId, filePath});
+        audioBuffer = await synthMicrosoft(logger, {credentials, stats, language, voice, text, deploymentId,
+          filePath, renderForCaching, disableTtsStreaming});        
         break;
       case 'nuance':
         model = model || 'enhanced';
@@ -194,10 +210,32 @@ async function synthAudio(client, logger, stats, { account_sid,
         audioBuffer = await synthWellSaid(logger, {credentials, stats, language, voice, text, filePath});
         break;
       case 'elevenlabs':
-        audioBuffer = await synthElevenlabs(logger, {credentials, stats, language, voice, text, filePath});
+        audioBuffer = await synthElevenlabs(logger, {
+          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
+        });        
+        break;
+      case 'playht':
+        audioBuffer = await synthPlayHT(logger, {
+          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
+        });        
+        break;
+      case 'rimelabs':
+        audioBuffer = await synthRimelabs(logger, {
+          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
+        });        
         break;
       case 'whisper':
-        audioBuffer = await synthWhisper(logger, {credentials, stats, voice, text});
+        audioBuffer = await synthWhisper(logger, {
+          credentials, stats, voice, text, renderForCaching, disableTtsStreaming});        
+        break;
+      case 'verbio':
+        audioBuffer = await synthVerbio(client, logger, {
+          credentials, stats, voice, text, renderForCaching, disableTtsStreaming});
+        if (audioBuffer?.filePath) return audioBuffer;
+        break;
+      case 'deepgram':
+        audioBuffer = await synthDeepgram(logger, {credentials, stats, model, text,
+          renderForCaching, disableTtsStreaming});        
         break;
       case vendor.startsWith('custom') ? vendor : 'cant_match_value':
         ({ audioBuffer, filePath } = await synthCustomVendor(logger,
@@ -206,6 +244,7 @@ async function synthAudio(client, logger, stats, { account_sid,
       default:
         assert(`synthAudio: unsupported speech vendor ${vendor}`);
     }
+    if('filePath' in audioBuffer) return audioBuffer;
     const diff = process.hrtime(startAt);
     const time = diff[0] * 1e3 + diff[1] * 1e-6;
     rtt = time.toFixed(0);
@@ -225,16 +264,28 @@ async function synthAudio(client, logger, stats, { account_sid,
   });
 }
 
-const synthPolly = async(logger, {credentials, stats, language, voice, engine, text}) => {
+const synthPolly = async(createHash, retrieveHash, logger,
+  {credentials, stats, language, voice, engine, text}) => {
   try {
-    const {region, accessKeyId, secretAccessKey} = credentials;
+    const {region, accessKeyId, secretAccessKey, roleArn} = credentials;
-    const polly = new PollyClient({
+    let polly;
-      region,
+    if (accessKeyId && secretAccessKey) {
-      credentials: {
+      polly = new PollyClient({
-        accessKeyId,
+        region,
-        secretAccessKey
+        credentials: {
-      }
+          accessKeyId,
-    });
+          secretAccessKey
+        }
+      });
+    } else if (roleArn) {
+      polly = new PollyClient({
+        region,
+        credentials: await getAwsAuthToken(logger, createHash, retrieveHash, null, null, region, roleArn),
+      });
+    } else {
+      // AWS RoleArn assigned to Instance profile
+      polly = new PollyClient({region});
+    }
     const opts = {
       Engine: engine,
       OutputFormat: 'mp3',
@@ -328,7 +379,7 @@ async function _synthOnPremMicrosoft(logger, {
   text,
   filePath
 }) {
-  const {use_custom_tts, custom_tts_endpoint_url} = credentials;
+  const {use_custom_tts, custom_tts_endpoint_url, api_key} = credentials;
   let content = text;
 
   if (use_custom_tts && !content.startsWith('<speak')) {
@@ -352,7 +403,8 @@ async function _synthOnPremMicrosoft(logger, {
     const post = bent('POST', 'buffer', {
       'X-Microsoft-OutputFormat': trimSilence ? 'raw-8khz-16bit-mono-pcm' : 'audio-16khz-32kbitrate-mono-mp3',
       'Content-Type': 'application/ssml+xml',
-      'User-Agent': 'Jambonz'
+      'User-Agent': 'Jambonz',
+      ...(api_key && {'Ocp-Apim-Subscription-Key': api_key})
     });
     const mp3 = await post(custom_tts_endpoint_url, content);
     return mp3;
@@ -368,10 +420,48 @@ const synthMicrosoft = async(logger, {
   language,
   voice,
   text,
-  filePath
+  filePath,
+  renderForCaching,
+  disableTtsStreaming
 }) => {
   try {
     const {api_key: apiKey, region, use_custom_tts, custom_tts_endpoint, custom_tts_endpoint_url} = credentials;
+    // let clean up the text
+    let content = text;
+    if (use_custom_tts && !content.startsWith('<speak')) {
+      /**
+       * Note: it seems that to use custom voice ssml is required with the voice attribute
+       * Otherwise sending plain text we get "Voice does not match"
+       */
+      content = `<speak>${text}</speak>`;
+    }
+
+    if (content.startsWith('<speak>')) {
+      /* microsoft enforces some properties and uses voice xml element so if the user did not supply do it for them */
+      const words = content.slice(7, -8).trim().replace(/(\r\n|\n|\r)/gm, ' ');
+      // eslint-disable-next-line max-len
+      content = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="${language}"><voice name="${voice}">${words}</voice></speak>`;
+      logger.info({content}, 'synthMicrosoft');
+    }
+    if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+      let params = '';
+      params += `{api_key=${apiKey}`;
+      params += `,language=${language}`;
+      params += ',vendor=microsoft';
+      params += `,voice=${voice}`;
+      params += ',write_cache_file=1';
+      if (region) params += `,region=${region}`;
+      if (custom_tts_endpoint) params += `,endpointId=${custom_tts_endpoint}`;
+      if (custom_tts_endpoint_url) params += `,endpoint=${custom_tts_endpoint_url}`;
+      if (JAMBONES_HTTP_PROXY_IP) params += `,http_proxy_ip=${JAMBONES_HTTP_PROXY_IP}`;
+      if (JAMBONES_HTTP_PROXY_PORT) params += `,http_proxy_port=${JAMBONES_HTTP_PROXY_PORT}`;
+      params += '}';
+      return {
+        filePath: `say:${params}${content.replace(/\n/g, ' ')}`,
+        servedFromCache: false,
+        rtt: 0
+      };
+    }
     if (use_custom_tts && custom_tts_endpoint_url) {
       return await _synthOnPremMicrosoft(logger, {
         credentials,
@@ -383,32 +473,22 @@ const synthMicrosoft = async(logger, {
       });
     }
     const trimSilence = filePath.endsWith('.r8');
-    let content = text;
     const speechConfig = SpeechConfig.fromSubscription(apiKey, region);
     speechConfig.speechSynthesisLanguage = language;
     speechConfig.speechSynthesisVoiceName = voice;
     if (use_custom_tts && custom_tts_endpoint) {
       speechConfig.endpointId = custom_tts_endpoint;
     }
-    if (use_custom_tts && !content.startsWith('<speak')) {
-      /**
-       * Note: it seems that to use custom voice ssml is required with the voice attribute
-       * Otherwise sending plain text we get "Voice does not match"
-       */
-      content = `<speak>${text}</speak>`;
-    }
     speechConfig.speechSynthesisOutputFormat = trimSilence ?
       SpeechSynthesisOutputFormat.Raw8Khz16BitMonoPcm :
       SpeechSynthesisOutputFormat.Audio16Khz32KBitRateMonoMp3;
-    const synthesizer = new SpeechSynthesizer(speechConfig);
 
-    if (content.startsWith('<speak>')) {
+    if (JAMBONES_HTTP_PROXY_IP && JAMBONES_HTTP_PROXY_PORT) {
-      /* microsoft enforces some properties and uses voice xml element so if the user did not supply do it for them */
+      logger.debug(
-      const words = content.slice(7, -8).trim().replace(/(\r\n|\n|\r)/gm, ' ');
+        `synthMicrosoft: using proxy ${JAMBONES_HTTP_PROXY_IP}:${JAMBONES_HTTP_PROXY_PORT}`);
-      // eslint-disable-next-line max-len
+      speechConfig.setProxy(JAMBONES_HTTP_PROXY_IP, JAMBONES_HTTP_PROXY_PORT);
-      content = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="${language}"><voice name="${voice}">${words}</voice></speak>`;
-      logger.info({content}, 'synthMicrosoft');
     }
+    const synthesizer = new SpeechSynthesizer(speechConfig);
 
     return new Promise((resolve, reject) => {
       const speakAsync = content.startsWith('<speak') ?
@@ -585,21 +665,50 @@ const synthCustomVendor = async(logger, {credentials, stats, language, voice, te
   }
 };
 
-const synthElevenlabs = async(logger, {credentials, stats, language, voice, text}) => {
+const synthElevenlabs = async(logger, {
-  const {api_key, model_id} = credentials;
+  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+}) => {
+  const {api_key, model_id, options: credOpts} = credentials;
+  const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
+
+  /* default to using the streaming interface, unless disabled by env var OR we want just a cache file */
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{api_key=${api_key}`;
+    params += ',vendor=elevenlabs';
+    params += `,voice=${voice}`;
+    params += `,model_id=${model_id}`;
+    params += `,optimize_streaming_latency=${opts.optimize_streaming_latency || 2}`;
+    params += ',write_cache_file=1';
+    if (opts.voice_settings?.similarity_boost) params += `,similarity_boost=${opts.voice_settings.similarity_boost}`;
+    if (opts.voice_settings?.stability) params += `,stability=${opts.voice_settings.stability}`;
+    if (opts.voice_settings?.style) params += `,style=${opts.voice_settings.style}`;
+    if (opts.voice_settings?.use_speaker_boost === false) params += ',use_speaker_boost=false';
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ').replace(/\r/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+
+  const optimize_streaming_latency = opts.optimize_streaming_latency ?
+    `?optimize_streaming_latency=${opts.optimize_streaming_latency}` : '';
   try {
     const post = bent('https://api.elevenlabs.io', 'POST', 'buffer', {
       'xi-api-key': api_key,
       'Accept': 'audio/mpeg',
       'Content-Type': 'application/json'
     });
-    const mp3 = await post(`/v1/text-to-speech/${voice}`, {
+    const mp3 = await post(`/v1/text-to-speech/${voice}${optimize_streaming_latency}`, {
       text,
       model_id,
       voice_settings: {
         stability: 0.5,
         similarity_boost: 0.5
-      }
+      },
+      ...opts
     });
     return mp3;
   } catch (err) {
@@ -609,8 +718,166 @@ const synthElevenlabs = async(logger, {credentials, stats, language, voice, text
   }
 };
 
-const synthWhisper = async(logger, {credentials, stats, voice, text}) => {
+const synthPlayHT = async(logger, {
-  const {api_key, model_id, baseURL, timeout} = credentials;
+  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+}) => {
+  const {api_key, user_id, voice_engine, options: credOpts} = credentials;
+  const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
+
+  /* default to using the streaming interface, unless disabled by env var OR we want just a cache file */
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{api_key=${api_key}`;
+    params += `,user_id=${user_id}`;
+    params += ',vendor=playht';
+    params += `,voice=${voice}`;
+    params += `,voice_engine=${voice_engine}`;
+    params += ',write_cache_file=1';
+    if (opts.quality) params += `,quality=${opts.quality}`;
+    if (opts.speed) params += `,speed=${opts.speed}`;
+    if (opts.seed) params += `,style=${opts.seed}`;
+    if (opts.temperature) params += `,temperature=${opts.temperature}`;
+    if (opts.emotion) params += `,emotion=${opts.emotion}`;
+    if (opts.voice_guidance) params += `,voice_guidance=${opts.voice_guidance}`;
+    if (opts.style_guidance) params += `,style_guidance=${opts.style_guidance}`;
+    if (opts.text_guidance) params += `,text_guidance=${opts.text_guidance}`;
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ').replace(/\r/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+
+  try {
+    const post = bent('https://api.play.ht', 'POST', 'buffer', {
+      'AUTHORIZATION': api_key,
+      'X-USER-ID': user_id,
+      'Accept': 'audio/mpeg',
+      'Content-Type': 'application/json'
+    });
+    const mp3 = await post('/api/v2/tts/stream', {
+      text,
+      voice,
+      voice_engine,
+      output_format: 'mp3',
+      sample_rate: 8000,
+      ...opts
+    });
+    return mp3;
+  } catch (err) {
+    logger.info({err}, 'synth PlayHT returned error');
+    stats.increment('tts.count', ['vendor:playht', 'accepted:no']);
+    throw err;
+  }
+};
+
+const synthRimelabs = async(logger, {
+  credentials, options, stats, voice, text, renderForCaching, disableTtsStreaming
+}) => {
+  const {api_key, model_id, options: credOpts} = credentials;
+  const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
+
+  /* default to using the streaming interface, unless disabled by env var OR we want just a cache file */
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{api_key=${api_key}`;
+    params += `,model_id=${model_id}`;
+    params += ',vendor=rimelabs';
+    params += `,voice=${voice}`;
+    params += ',write_cache_file=1';
+    if (opts.speedAlpha) params += `,speed_alpha=${opts.speedAlpha}`;
+    if (opts.reduceLatency) params += `,reduce_latency=${opts.reduceLatency}`;
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ').replace(/\r/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+
+  try {
+    const post = bent('https://users.rime.ai', 'POST', 'buffer', {
+      'Authorization': `Bearer ${api_key}`,
+      'Accept': 'audio/mp3',
+      'Content-Type': 'application/json'
+    });
+    const mp3 = await post('/v1/rime-tts', {
+      speaker: voice,
+      text,
+      modelId: model_id,
+      samplingRate: 8000,
+      ...opts
+    });
+    return mp3;
+  } catch (err) {
+    logger.info({err}, 'synth rimelabs returned error');
+    stats.increment('tts.count', ['vendor:rimelabs', 'accepted:no']);
+    throw err;
+  }
+};
+const synthVerbio = async(client, logger, {credentials, stats, voice, text, renderForCaching, disableTtsStreaming}) => {
+  //https://doc.speechcenter.verbio.com/#tag/Text-To-Speech-REST-API
+  if (text.length > 2000) {
+    throw new Error('Verbio cannot synthesize for the text length larger than 2000 characters');
+  }
+  const token = await getVerbioAccessToken(client, logger, credentials);
+  if (!process.env.JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{access_token=${token.access_token}`;
+    params += ',vendor=verbio';
+    params += `,voice=${voice}`;
+    params += ',write_cache_file=1';
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+
+  try {
+    const post = bent('https://us.rest.speechcenter.verbio.com', 'POST', 'buffer', {
+      'Authorization': `Bearer ${token.access_token}`,
+      'User-Agent': 'jambonz',
+      'Content-Type': 'application/json'
+    });
+    const r8 = await post('/api/v1/synthesize', {
+      voice_id: voice,
+      output_sample_rate: '8k',
+      output_encoding: 'pcm16',
+      text
+    });
+    return r8;
+  } catch (err) {
+    logger.info({err}, 'synth Verbio returned error');
+    stats.increment('tts.count', ['vendor:verbio', 'accepted:no']);
+    throw err;
+  }
+};
+
+const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCaching, disableTtsStreaming}) => {
+  const {api_key, model_id, baseURL, timeout, speed} = credentials;
+  /* if the env is set to stream then bag out, unless we are specifically rendering to generate a cache file */
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{api_key=${api_key}`;
+    params += `,model_id=${model_id}`;
+    params += ',vendor=whisper';
+    params += `,voice=${voice}`;
+    params += ',write_cache_file=1';
+    if (speed) params += `,speed=${speed}`;
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
   try {
     const openai = new OpenAI.OpenAI({
       apiKey: api_key,
@@ -630,8 +897,41 @@ const synthWhisper = async(logger, {credentials, stats, voice, text}) => {
     stats.increment('tts.count', ['vendor:openai', 'accepted:no']);
     throw err;
   }
-}
+};
-;
+
+const synthDeepgram = async(logger, {credentials, stats, model, text, renderForCaching, disableTtsStreaming}) => {
+  const {api_key} = credentials;
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{api_key=${api_key}`;
+    params += ',vendor=deepgram';
+    params += `,voice=${model}`;
+    params += ',write_cache_file=1';
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
+  try {
+    const post = bent('https://api.deepgram.com', 'POST', 'buffer', {
+      'Authorization': `Token ${api_key}`,
+      'Accept': 'audio/mpeg',
+      'Content-Type': 'application/json'
+    });
+    const mp3 = await post(`/v1/speak?model=${model}`, {
+      text
+    });
+    return mp3;
+  } catch (err) {
+    logger.info({err}, 'synth Deepgram returned error');
+    stats.increment('tts.count', ['vendor:deepgram', 'accepted:no']);
+    throw err;
+  }
+};
+
 const getFileExtFromMime = (mime) => {
   switch (mime) {
     case 'audio/wav':

lib/utils.js

@@ -3,10 +3,10 @@ const {SynthesizerClient} = require('../stubs/nuance/synthesizer_grpc_pb');
 const {RivaSpeechSynthesisClient} = require('../stubs/riva/proto/riva_tts_grpc_pb');
 const {Pool} = require('undici');
 const pool = new Pool('https://auth.crt.nuance.com');
-const HTTP_TIMEOUT = 5000;
 const NUANCE_AUTH_ENDPOINT = 'tts.api.nuance.com:443';
 const grpc = require('@grpc/grpc-js');
 const formurlencoded = require('form-urlencoded');
+const { JAMBONES_DISABLE_TTS_STREAMING, JAMBONES_TTS_TRIM_SILENCE, TMP_FOLDER, HTTP_TIMEOUT } = require('./config');
 
 const debug = require('debug')('jambonz:realtimedb-helpers');
 /**
@@ -19,7 +19,44 @@ const debug = require('debug')('jambonz:realtimedb-helpers');
 function makeSynthKey({account_sid = '', vendor, language, voice, engine = '', text}) {
   const hash = crypto.createHash('sha1');
   hash.update(`${language}:${vendor}:${voice}:${engine}:${text}`);
-  return `tts${account_sid ? (':' + account_sid) : ''}:${hash.digest('hex')}`;
+  const hexHashKey = hash.digest('hex');
+  const accountKey = account_sid ? `:${account_sid}` : '';
+  const namespace = vendor.startsWith('custom') ? vendor : getFileExtension(vendor);
+  const key = `tts${accountKey}:${namespace}:${hexHashKey}`;
+  return key;
+}
+
+function makeFilePath(vendor, key, salt = '') {
+  const extension = getFileExtension(vendor);
+  return `${TMP_FOLDER}/${key.replace('tts:', `tts-${salt}`)}.${extension}`;
+}
+
+function getFileExtension(vendor) {
+  const mp3Extension = 'mp3';
+  const r8Extension = 'r8';
+
+  switch (vendor) {
+    case 'azure':
+    case 'microsoft':
+      if (!JAMBONES_DISABLE_TTS_STREAMING || JAMBONES_TTS_TRIM_SILENCE) {
+        return r8Extension;
+      } else {
+        return mp3Extension;
+      }
+    case 'deepgram':
+    case 'elevenlabs':
+    case 'rimlabs':
+      if (!JAMBONES_DISABLE_TTS_STREAMING) {
+        return r8Extension;
+      } else {
+        return mp3Extension;
+      }
+    case 'nuance':
+    case 'nvidia':
+      return r8Extension;
+    default:
+      return mp3Extension;
+  }
 }
 
 const noopLogger = {
@@ -43,6 +80,18 @@ function makeIbmKey(apiKey) {
   return `ibm:${hash.digest('hex')}`;
 }
 
+function makeAwsKey(awsAccessKeyId) {
+  const hash = crypto.createHash('sha1');
+  hash.update(awsAccessKeyId);
+  return `aws:${hash.digest('hex')}`;
+}
+
+function makeVerbioKey(client_id) {
+  const hash = crypto.createHash('sha1');
+  hash.update(client_id);
+  return `verbio:${hash.digest('hex')}`;
+}
+
 function makeNuanceKey(clientId, secret, scope) {
   const hash = crypto.createHash('sha1');
   hash.update(`${clientId}:${secret}:${scope}`);
@@ -110,11 +159,14 @@ module.exports = {
   makeSynthKey,
   makeNuanceKey,
   makeIbmKey,
+  makeAwsKey,
+  makeVerbioKey,
   getNuanceAccessToken,
   createNuanceClient,
   createKryptonClient,
   createRivaClient,
   makeBasicAuthHeader,
   NUANCE_AUTH_ENDPOINT,
-  noopLogger
+  noopLogger,
+  makeFilePath
 };

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.25",
+  "version": "0.1.2",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",
@@ -12,6 +12,7 @@
     "test": "NODE_ENV=test node test/ ",
     "coverage": "nyc --reporter html --report-dir ./coverage npm run test",
     "jslint": "eslint index.js lib",
+    "jslint:fix": "eslint --fix '**/*.js'",
     "build": "./build_stubs.sh"
   },
   "repository": {
@@ -24,25 +25,26 @@
   },
   "homepage": "https://github.com/jambonz/speech-utils#readme",
   "dependencies": {
-    "@aws-sdk/client-polly": "^3.359.0",
+    "@aws-sdk/client-polly": "^3.496.0",
-    "@google-cloud/text-to-speech": "^4.2.1",
+    "@aws-sdk/client-sts": "^3.496.0",
-    "@grpc/grpc-js": "^1.8.13",
+    "@google-cloud/text-to-speech": "^5.0.2",
+    "@grpc/grpc-js": "^1.9.14",
+    "@jambonz/realtimedb-helpers": "^0.8.7",
     "bent": "^7.3.12",
     "debug": "^4.3.4",
-    "form-urlencoded": "^6.1.0",
+    "form-urlencoded": "^6.1.4",
     "google-protobuf": "^3.21.2",
     "ibm-watson": "^8.0.0",
-    "ioredis": "^5.3.2",
+    "microsoft-cognitiveservices-speech-sdk": "1.36.0",
-    "microsoft-cognitiveservices-speech-sdk": "^1.31.0",
+    "openai": "^4.25.0",
-    "openai": "^4.16.2",
+    "undici": "^6.4.0"
-    "undici": "^5.21.0"
   },
   "devDependencies": {
-    "config": "^3.3.9",
+    "config": "^3.3.10",
-    "eslint": "^8.33.0",
+    "eslint": "^8.56.0",
     "eslint-plugin-promise": "^6.1.1",
     "nyc": "^15.1.0",
-    "pino": "^7.2.0",
+    "pino": "^8.17.0",
-    "tape": "^5.1.1"
+    "tape": "^5.7.3"
   }
 }

stubs/nuance/synthesizer_grpc_pb.js

@@ -62,9 +62,9 @@ function deserialize_nuance_tts_v1_UnarySynthesisResponse(buffer_arg) {
 
 //
 // The Synthesizer service offers these functionalities:
-// - GetVoices: Queries the list of available voices, with filters to reduce the search space.  
+// - GetVoices: Queries the list of available voices, with filters to reduce the search space.
-// - Synthesize: Synthesizes audio from input text and parameters, and returns an audio stream. 
+// - Synthesize: Synthesizes audio from input text and parameters, and returns an audio stream.
-// - UnarySynthesize: Synthesizes audio from input text and parameters, and returns a single audio response. 
+// - UnarySynthesize: Synthesizes audio from input text and parameters, and returns a single audio response.
 var SynthesizerService = exports.SynthesizerService = {
   getVoices: {
     path: '/nuance.tts.v1.Synthesizer/GetVoices',

stubs/riva/proto/riva_audio_grpc_pb.js

@@ -1 +1 @@
-// GENERATED CODE -- NO SERVICES IN PROTO
+// GENERATED CODE -- NO SERVICES IN PROTO

stubs/riva/proto/riva_tts_grpc_pb.js

@@ -57,7 +57,7 @@ function deserialize_nvidia_riva_tts_SynthesizeSpeechResponse(buffer_arg) {
 var RivaSpeechSynthesisService = exports.RivaSpeechSynthesisService = {
   // Used to request text-to-speech from the service. Submit a request containing the
 // desired text and configuration, and receive audio bytes in the requested format.
-synthesize: {
+  synthesize: {
     path: '/nvidia.riva.tts.RivaSpeechSynthesis/Synthesize',
     requestStream: false,
     responseStream: false,
@@ -69,9 +69,9 @@ synthesize: {
     responseDeserialize: deserialize_nvidia_riva_tts_SynthesizeSpeechResponse,
   },
   // Used to request text-to-speech returned via stream as it becomes available.
-// Submit a SynthesizeSpeechRequest with desired text and configuration,
+  // Submit a SynthesizeSpeechRequest with desired text and configuration,
-// and receive stream of bytes in the requested format.
+  // and receive stream of bytes in the requested format.
-synthesizeOnline: {
+  synthesizeOnline: {
     path: '/nvidia.riva.tts.RivaSpeechSynthesis/SynthesizeOnline',
     requestStream: false,
     responseStream: true,
@@ -83,7 +83,7 @@ synthesizeOnline: {
     responseDeserialize: deserialize_nvidia_riva_tts_SynthesizeSpeechResponse,
   },
   // Enables clients to request the configuration of the current Synthesize service, or a specific model within the service.
-getRivaSynthesisConfig: {
+  getRivaSynthesisConfig: {
     path: '/nvidia.riva.tts.RivaSpeechSynthesis/GetRivaSynthesisConfig',
     requestStream: false,
     responseStream: false,

test/aws.js

@@ -0,0 +1,40 @@
+const test = require('tape').test ;
+const config = require('config');
+const opts = config.get('redis');
+const logger = require('pino')({level: 'error'});
+process.on('unhandledRejection', (reason, p) => {
+  console.log('Unhandled Rejection at: Promise', p, 'reason:', reason);
+});
+
+const sleep = (ms) => new Promise((resolve) => setTimeout(resolve, ms));
+
+test('AWS - create and cache auth token', async(t) => {
+  const fn = require('..');
+  const {client, getAwsAuthToken} = fn(opts, logger);
+
+  if (!process.env.AWS_ACCESS_KEY_ID || !process.env.AWS_SECRET_ACCESS_KEY || !process.env.AWS_REGION) {
+      t.pass('skipping AWS auth token tests since no AWS credentials provided');
+      t.end();
+      client.quit();
+      return;
+  }
+  try {
+    let obj = await getAwsAuthToken(process.env.AWS_ACCESS_KEY_ID, process.env.AWS_SECRET_ACCESS_KEY, process.env.AWS_REGION);
+    //console.log({obj}, 'received auth token from AWS');
+    t.ok(obj.securityToken && !obj.servedFromCache, 'successfullY generated auth token from AWS');
+
+    await sleep(250);
+    obj = await getAwsAuthToken(process.env.AWS_ACCESS_KEY_ID, process.env.AWS_SECRET_ACCESS_KEY, process.env.AWS_REGION);
+    //console.log({obj}, 'received auth token from AWS - second request');
+    t.ok(obj.securityToken && obj.servedFromCache, 'successfully received access token from cache');
+ 
+    await client.flushall();
+    t.end();
+  }
+  catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+

test/index.js

@@ -1,4 +1,7 @@
 require('./docker_start');
 require('./synth');
 require('./list-voices');
+require('./aws');
+require('./ibm');
+require('./nuance');
 require('./docker_stop');

test/list-voices.js

@@ -12,6 +12,32 @@ const stats = {
   histogram: () => {}
 };
 
+test('Verbio - get Access key and voices', async(t) => {
+  const fn = require('..');
+  const {client, getTtsVoices, getVerbioAccessToken} = fn(opts, logger);
+  if (!process.env.VERBIO_CLIENT_ID || !process.env.VERBIO_CLIENT_SECRET) {
+    t.pass('skipping Verbio test since no Verbio Keys provided');
+    t.end();
+    client.quit();
+    return;
+  }
+
+  try {
+    const credentials = {
+      client_id: process.env.VERBIO_CLIENT_ID,
+      client_secret: process.env.VERBIO_CLIENT_SECRET
+    };
+    let obj = await getVerbioAccessToken(credentials);
+    t.ok(obj.access_token , 'successfully received access token not from cache');
+    const voices = await getTtsVoices({vendor: 'verbio', credentials});
+    t.ok(voices && voices.length != 0, 'successfully received verbio voices');
+  } catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
 test('IBM - create access key', async(t) => {
   const fn = require('..');
   const {client, getIbmAccessToken} = fn(opts, logger);

test/squid.conf

@@ -0,0 +1,85 @@
+#
+# Recommended minimum configuration:
+#
+
+# Example rule allowing access from your local networks.
+# Adapt to list your (internal) IP networks from where browsing
+# should be allowed
+acl localnet src 0.0.0.1-0.255.255.255	# RFC 1122 "this" network (LAN)
+acl localnet src 10.0.0.0/8		# RFC 1918 local private network (LAN)
+acl localnet src 100.64.0.0/10		# RFC 6598 shared address space (CGN)
+acl localnet src 169.254.0.0/16 	# RFC 3927 link-local (directly plugged) machines
+acl localnet src 172.16.0.0/12		# RFC 1918 local private network (LAN)
+acl localnet src 192.168.0.0/16		# RFC 1918 local private network (LAN)
+acl localnet src fc00::/7       	# RFC 4193 local private network range
+acl localnet src fe80::/10      	# RFC 4291 link-local (directly plugged) machines
+
+acl SSL_ports port 443
+acl Safe_ports port 80		# http
+acl Safe_ports port 21		# ftp
+acl Safe_ports port 443		# https
+acl Safe_ports port 70		# gopher
+acl Safe_ports port 210		# wais
+acl Safe_ports port 1025-65535	# unregistered ports
+acl Safe_ports port 280		# http-mgmt
+acl Safe_ports port 488		# gss-http
+acl Safe_ports port 591		# filemaker
+acl Safe_ports port 777		# multiling http
+
+#
+# Recommended minimum Access Permission configuration:
+#
+# Deny requests to certain unsafe ports
+http_access deny !Safe_ports
+
+# Deny CONNECT to other than secure SSL ports
+http_access allow CONNECT !SSL_ports
+
+# Only allow cachemgr access from localhost
+http_access allow localhost manager
+http_access deny manager
+
+# This default configuration only allows localhost requests because a more
+# permissive Squid installation could introduce new attack vectors into the
+# network by proxying external TCP connections to unprotected services.
+http_access allow localhost
+
+# The two deny rules below are unnecessary in this default configuration
+# because they are followed by a "deny all" rule. However, they may become
+# critically important when you start allowing external requests below them.
+
+# Protect web applications running on the same server as Squid. They often
+# assume that only local users can access them at "localhost" ports.
+http_access deny to_localhost
+
+# Protect cloud servers that provide local users with sensitive info about
+# their server via certain well-known link-local (a.k.a. APIPA) addresses.
+http_access deny to_linklocal
+
+#
+# INSERT YOUR OWN RULE(S) HERE TO ALLOW ACCESS FROM YOUR CLIENTS
+#
+
+# For example, to allow access from your local networks, you may uncomment the
+# following rule (and/or add rules that match your definition of "local"):
+# http_access allow localnet
+
+# And finally deny all other access to this proxy
+http_access deny all
+
+# Squid normally listens to port 3128
+http_port 3128
+
+# Uncomment and adjust the following to add a disk cache directory.
+#cache_dir ufs /usr/local/var/cache/squid 100 16 256
+
+# Leave coredumps in the first cache dir
+coredump_dir /usr/local/var/cache/squid
+
+#
+# Add any of your own refresh_pattern entries above these.
+#
+refresh_pattern ^ftp:		1440	20%	10080
+refresh_pattern ^gopher:	1440	0%	1440
+refresh_pattern -i (/cgi-bin/|\?) 0	0%	0
+refresh_pattern .		0	20%	4320

test/synth.js

@@ -20,7 +20,7 @@ const stats = {
 
 test('Google speech synth tests', async(t) => {
   const fn = require('..');
-  const {synthAudio, client} = fn(opts, logger);
+  const {synthAudio, addFileToCache, client} = 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');
@@ -58,6 +58,14 @@ test('Google speech synth tests', async(t) => {
     });
     t.ok(opts.servedFromCache, `successfully retrieved cached google audio from ${opts.filePath}`);
 
+    const success = await addFileToCache(opts.filePath, {
+      vendor: 'google',
+      language: 'en-GB',
+      gender: 'FEMALE',
+      text: 'This is a test.  This is only a test'
+    });
+    t.ok(success, `successfully added ${opts.filePath} to cache`);
+
     opts = await synthAudio(stats, {
       vendor: 'google',
       credentials: {
@@ -154,6 +162,33 @@ test('AWS speech synth tests', async(t) => {
   client.quit();
 });
 
+test('AWS speech synth tests by RoleArn', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.AWS_ROLE_ARN || !process.env.AWS_REGION) {
+    t.pass('skipping AWS speech synth tests by RoleArn since AWS_ROLE_ARN or AWS_REGION not provided');
+    return t.end();
+  }
+  try {
+    let opts = await synthAudio(stats, {
+      vendor: 'aws',
+      credentials: {
+        roleArn: process.env.AWS_ROLE_ARN,
+        region: process.env.AWS_REGION,
+      },
+      language: 'en-US',
+      voice: 'Joey',
+      text: 'This is a test.  This is only a test',
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized aws by roleArn audio to ${opts.filePath}`);
+  } catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
 test('Azure speech synth tests', async(t) => {
   const fn = require('..');
   const {synthAudio, client} = fn(opts, logger);
@@ -180,9 +215,12 @@ test('Azure speech synth tests', async(t) => {
       language: 'en-US',
       voice: 'en-US-ChristopherNeural',
       text: longText,
+      renderForCaching: true
     });
     t.ok(!opts.servedFromCache, `successfully synthesized microsoft audio to ${opts.filePath}`);
-
+    if (process.env.JAMBONES_HTTP_PROXY_IP && process.env.JAMBONES_HTTP_PROXY_PORT) {
+      t.pass('successfully used proxy to reach microsoft tts service');
+    }
 
     opts = await synthAudio(stats, {
       vendor: 'microsoft',
@@ -193,6 +231,7 @@ test('Azure speech synth tests', async(t) => {
       language: 'en-US',
       voice: 'en-US-ChristopherNeural',
       text: longText,
+      renderForCaching: true
     });
     t.ok(opts.servedFromCache, `successfully retrieved microsoft audio from cache ${opts.filePath}`);
   } catch (err) {
@@ -202,6 +241,58 @@ test('Azure speech synth tests', async(t) => {
   client.quit();
 });
 
+test('Azure SSML tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.MICROSOFT_API_KEY || !process.env.MICROSOFT_REGION) {
+    t.pass('skipping Microsoft speech synth tests since MICROSOFT_API_KEY or MICROSOFT_REGION not provided');
+    return t.end();
+  }
+  try {
+    const text = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="en-US">
+    <voice name="en-US-JennyMultilingualNeural">
+    <mstts:express-as style="cheerful" styledegree="2">That'd be just amazing!
+    </mstts:express-as>
+    </voice>
+    </speak>`;
+
+    let opts = await synthAudio(stats, {
+      vendor: 'microsoft',
+      credentials: {
+        api_key: process.env.MICROSOFT_API_KEY,
+        region: process.env.MICROSOFT_REGION,
+      },
+      language: 'en-US',
+      voice: 'en-US-ChristopherNeural',
+      text,
+      renderForCaching: true
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized microsoft audio to ${opts.filePath}`);
+    if (process.env.JAMBONES_HTTP_PROXY_IP && process.env.JAMBONES_HTTP_PROXY_PORT) {
+      t.pass('successfully used proxy to reach microsoft tts service');
+    }
+
+    opts = await synthAudio(stats, {
+      vendor: 'microsoft',
+      credentials: {
+        api_key: process.env.MICROSOFT_API_KEY,
+        region: process.env.MICROSOFT_REGION,
+      },
+      language: 'en-US',
+      voice: 'en-US-ChristopherNeural',
+      text,
+      renderForCaching: true
+    });
+    t.ok(opts.servedFromCache, `successfully retrieved microsoft audio from cache ${opts.filePath}`);
+  } catch (err) {
+    console.error(err);
+    t.end(err);
+  }
+  client.quit();
+});
+
+
 test('Azure custom voice speech synth tests', async(t) => {
   const fn = require('..');
   const {synthAudio, client} = fn(opts, logger);
@@ -223,6 +314,7 @@ test('Azure custom voice speech synth tests', async(t) => {
       language: 'en-US',
       voice: process.env.MICROSOFT_CUSTOM_VOICE,
       text,
+      renderForCaching: true
     });
     t.ok(!opts.servedFromCache, `successfully synthesized microsoft audio to ${opts.filePath}`);
 
@@ -237,6 +329,7 @@ test('Azure custom voice speech synth tests', async(t) => {
       language: 'en-US',
       voice: process.env.MICROSOFT_CUSTOM_VOICE,
       text,
+      renderForCaching: true
     });
     t.ok(opts.servedFromCache, `successfully retrieved microsoft custom voice audio from cache ${opts.filePath}`);
   } catch (err) {
@@ -459,7 +552,16 @@ test('Elevenlabs speech synth tests', async(t) => {
       vendor: 'elevenlabs',
       credentials: {
         api_key: process.env.ELEVENLABS_API_KEY,
-        model_id: process.env.ELEVENLABS_MODEL_ID
+        model_id: process.env.ELEVENLABS_MODEL_ID,
+        options: JSON.stringify({
+          optimize_streaming_latency: 1,
+          voice_settings: {
+            similarity_boost: 1,
+            stability: 0.8,
+            style: 1,
+            use_speaker_boost: true
+          }
+        })
       },
       language: 'en-US',
       voice: process.env.ELEVENLABS_VOICE_ID,
@@ -474,6 +576,81 @@ test('Elevenlabs speech synth tests', async(t) => {
   client.quit();
 })
 
+test('PlayHT speech synth tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.PLAYHT_API_KEY || !process.env.PLAYHT_USER_ID) {
+    t.pass('skipping PlayHT speech synth tests since PLAYHT_API_KEY or PLAYHT_USER_ID is/are not provided');
+    return t.end();
+  }
+  const text = 'Hi there and welcome to jambones!';
+  try {
+    let opts = await synthAudio(stats, {
+      vendor: 'playht',
+      credentials: {
+        api_key: process.env.PLAYHT_API_KEY,
+        user_id: process.env.PLAYHT_USER_ID,
+        voice_engine: 'PlayHT2.0-turbo',
+        options: JSON.stringify({
+          quality: "medium",
+          speed: 1,
+          seed: 1,
+          temperature: 1,
+          emotion: "female_happy",
+          voice_guidance: 3,
+          style_guidance: 20,
+          text_guidance: 1,
+        })
+      },
+      language: 'en-US',
+      voice: 's3://voice-cloning-zero-shot/d9ff78ba-d016-47f6-b0ef-dd630f59414e/female-cs/manifest.json',
+      text,
+      renderForCaching: true
+    });
+    t.ok(!opts.servedFromCache, `successfully playht eleven audio to ${opts.filePath}`);
+
+  } catch (err) {
+    console.error(JSON.stringify(err));
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('rimelabs speech synth tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.RIMELABS_API_KEY) {
+    t.pass('skipping rimelabs speech synth tests since RIMELABS_API_KEY is not provided');
+    return t.end();
+  }
+  const text = 'Hi there and welcome to jambones!';
+  try {
+    let opts = await synthAudio(stats, {
+      vendor: 'rimelabs',
+      credentials: {
+        api_key: process.env.RIMELABS_API_KEY,
+        model_id: 'mist',
+        options: JSON.stringify({
+          speedAlpha: 1.0,
+          reduceLatency: false
+        })
+      },
+      language: 'en-US',
+      voice: 'amber',
+      text,
+      renderForCaching: true
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized rimelabs audio to ${opts.filePath}`);
+
+  } catch (err) {
+    console.error(JSON.stringify(err));
+    t.end(err);
+  }
+  client.quit();
+});
+
 test('whisper speech synth tests', async(t) => {
   const fn = require('..');
   const {synthAudio, client} = fn(opts, logger);
@@ -493,9 +670,71 @@ test('whisper speech synth tests', async(t) => {
       language: 'en-US',
       voice: 'alloy',
       text,
+      renderForCaching: true
     });
     t.ok(!opts.servedFromCache, `successfully synthesized whisper audio to ${opts.filePath}`);
 
+  } catch (err) {
+    console.error(JSON.stringify(err));
+    t.end(err);
+  }
+  client.quit();
+});
+
+test('Verbio speech synth tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.VERBIO_CLIENT_ID || !process.env.VERBIO_CLIENT_SECRET) {
+    t.pass('skipping Verbio Synthesize test since no Verbio Keys provided');
+    t.end();
+    client.quit();
+    return;
+  }
+
+  const text = 'Hi there and welcome to jambones!';
+  try {
+    let opts = await synthAudio(stats, {
+      vendor: 'verbio',
+      credentials: {
+        client_id: process.env.VERBIO_CLIENT_ID,
+        client_secret: process.env.VERBIO_CLIENT_SECRET
+      },
+      language: 'en-US',
+      voice: 'tommy_en-us',
+      text,
+      renderForCaching: true
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized whisper audio to ${opts.filePath}`);
+
+  } catch (err) {
+    console.error(JSON.stringify(err));
+    t.end(err);
+  }
+  client.quit();
+})
+
+test('Deepgram speech synth tests', async(t) => {
+  const fn = require('..');
+  const {synthAudio, client} = fn(opts, logger);
+
+  if (!process.env.DEEPGRAM_API_KEY) {
+    t.pass('skipping Deepgram speech synth tests since DEEPGRAM_API_KEY');
+    return t.end();
+  }
+  const text = 'Hi there and welcome to jambones!';
+  try {
+    let opts = await synthAudio(stats, {
+      vendor: 'deepgram',
+      credentials: {
+        api_key: process.env.DEEPGRAM_API_KEY
+      },
+      model: 'aura-asteria-en',
+      text,
+      renderForCaching: true
+    });
+    t.ok(!opts.servedFromCache, `successfully synthesized deepgram audio to ${opts.filePath}`);
+
   } catch (err) {
     console.error(JSON.stringify(err));
     t.end(err);
@@ -550,8 +789,19 @@ test('TTS Cache tests', async(t) => {
     t.ok(error, `error returned when specified key was not found`);
 
     // make sure other tts keys are still there
-    const cached = (await client.keys('tts:*')).length;
+    const cached = await client.keys('tts:*')
-    t.ok(cached >= 1, `successfully kept all non-specified tts records in cache`);
+    t.ok(cached.length >= 1, `successfully kept all non-specified tts records in cache`);
+
+    // retrieve keys from cache and check the key contains the file extension    
+    let key = cached[0];
+    t.ok(key.includes('mp3'), `tts cache extension shoult be part of the key and equal mp3`);
+
+    process.env.VG_TRIM_TTS_SILENCE = 'true';    
+    await client.set(makeSynthKey({ vendor: 'azure' }), 'value');
+
+    const r8Keys = await client.keys('tts:r8*');
+    key = r8Keys[0];
+    t.ok(key.includes('r8'), `tts cache extension shoult be part of the key and equal r8`);
 
   } catch (err) {
     console.error(JSON.stringify(err));