@jambonz/speech-utils 0.0.50 → 0.1.0

package/.github/workflows/ci.yml

@@ -8,8 +8,8 @@ jobs:
   build:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-node@v3
+      - uses: actions/checkout@v4
+      - uses: actions/setup-node@v4
         with:
           node-version: lts/*
       - run: npm install
@@ -32,5 +32,7 @@ jobs:
           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

package/index.js

@@ -13,10 +13,10 @@ module.exports = (opts, logger) => {
     getTtsSize: require('./lib/get-tts-size').bind(null, client, logger),
     purgeTtsCache: require('./lib/purge-tts-cache').bind(null, client, logger),
     addFileToCache: require('./lib/add-file-to-cache').bind(null, client, logger),
-    synthAudio: require('./lib/synth-audio').bind(null, client, logger),
+    synthAudio: require('./lib/synth-audio').bind(null, client, createHash, retrieveHash, logger),
     getNuanceAccessToken: require('./lib/get-nuance-access-token').bind(null, client, logger),
     getIbmAccessToken: require('./lib/get-ibm-access-token').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, logger),
+    getTtsVoices: require('./lib/get-tts-voices').bind(null, client, createHash, retrieveHash, logger),
   };
 };

package/lib/get-aws-sts-token.js

@@ -1,32 +1,41 @@
-const { STSClient, GetSessionTokenCommand } = require('@aws-sdk/client-sts');
+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) {
+  logger, createHash, retrieveHash,
+  awsAccessKeyId, awsSecretAccessKey, awsRegion, roleArn = null) {
   logger = logger || noopLogger;
   try {
-    const key = makeAwsKey(awsAccessKeyId);
+    const key = makeAwsKey(roleArn || awsAccessKeyId);
     const obj = await retrieveHash(key);
     if (obj) return {...obj, servedFromCache: true};
 
-    /* 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});
-    const data = await stsClient.send(command);
+    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
     };
 

package/lib/get-tts-voices.js

@@ -6,6 +6,7 @@ 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 getIbmVoices = async(client, logger, credentials) => {
   const {tts_region, tts_api_key} = credentials;
@@ -87,16 +88,26 @@ 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 client = new PollyClient({
-      region,
-      credentials: {
-        accessKeyId,
-        secretAccessKey
-      }
-    });
+    const {region, accessKeyId, secretAccessKey, roleArn} = credentials;
+    let client = null;
+    if (accessKeyId && secretAccessKey) {
+      client = new PollyClient({
+        region,
+        credentials: {
+          accessKeyId,
+          secretAccessKey
+        }
+      });
+    } 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;
@@ -122,7 +133,7 @@ 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),
@@ -137,7 +148,7 @@ 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);
     default:
       break;
   }

package/lib/synth-audio.js

@@ -39,6 +39,7 @@ 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 TMP_FOLDER = '/tmp';
 const OpenAI = require('openai');
+const getAwsAuthToken = require('./get-aws-sts-token');
 
 
 const trimTrailingSilence = (buffer) => {
@@ -75,7 +76,7 @@ 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, renderForCaching, disableTtsStreaming, options
 }) {
@@ -84,10 +85,10 @@ async function synthAudio(client, logger, stats, { account_sid,
   let rtt;
   logger = logger || noopLogger;
 
-  assert.ok(['google', 'aws', 'polly', 'microsoft',
-    'wellsaid', 'nuance', 'nvidia', 'ibm', 'elevenlabs', 'whisper', 'deepgram', 'playht'].includes(vendor) ||
+  assert.ok(['google', 'aws', 'polly', 'microsoft', 'wellsaid', 'nuance', 'nvidia', 'ibm', 'elevenlabs',
+    'whisper', 'deepgram', 'playht', 'rimelabs'].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 ${vendor}`);
   if ('google' === vendor) {
     assert.ok(language, 'synthAudio requires language when google is used');
   }
@@ -123,11 +124,15 @@ async function synthAudio(client, logger, stats, { account_sid,
     assert.ok(voice, 'synthAudio requires voice when elevenlabs is used');
     assert.ok(credentials.api_key, 'synthAudio requires api_key when elevenlabs is used');
     assert.ok(credentials.model_id, 'synthAudio requires model_id when elevenlabs is used');
-  }  else if ('playht' === vendor) {
+  } else if ('playht' === vendor) {
     assert.ok(voice, 'synthAudio requires voice when playht is used');
     assert.ok(credentials.api_key, 'synthAudio requires api_key when playht is used');
     assert.ok(credentials.user_id, 'synthAudio requires user_id when playht is used');
     assert.ok(credentials.voice_engine, 'synthAudio requires voice_engine when playht is used');
+  } else if ('rimelabs' === vendor) {
+    assert.ok(voice, 'synthAudio requires voice when rimelabs is used');
+    assert.ok(credentials.api_key, 'synthAudio requires api_key when rimelabs is used');
+    assert.ok(credentials.model_id, 'synthAudio requires model_id when rimelabs is used');
   } else if ('whisper' === vendor) {
     assert.ok(voice, 'synthAudio requires voice when whisper is used');
     assert.ok(credentials.model_id, 'synthAudio requires model when whisper is used');
@@ -151,7 +156,7 @@ async function synthAudio(client, logger, stats, { account_sid,
     ) ||
     (
       !process.env.JAMBONES_DISABLE_TTS_STREAMING &&
-      ['elevenlabs', 'deepgram'].includes(vendor)
+      ['elevenlabs', 'deepgram', 'rimelabs'].includes(vendor)
     )
   ) {
     filePath = `${TMP_FOLDER}/${key.replace('tts:', `tts-${salt || ''}`)}.r8`;
@@ -183,7 +188,8 @@ 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':
@@ -217,6 +223,12 @@ async function synthAudio(client, logger, stats, { account_sid,
         });
         if (audioBuffer?.filePath) return audioBuffer;
         break;
+      case 'rimelabs':
+        audioBuffer = await synthRimelabs(logger, {
+          credentials, options, stats, language, voice, text, renderForCaching, disableTtsStreaming, filePath
+        });
+        if (audioBuffer?.filePath) return audioBuffer;
+        break;
       case 'whisper':
         audioBuffer = await synthWhisper(logger, {
           credentials, stats, voice, text, renderForCaching, disableTtsStreaming});
@@ -253,16 +265,28 @@ 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 polly = new PollyClient({
-      region,
-      credentials: {
-        accessKeyId,
-        secretAccessKey
-      }
-    });
+    const {region, accessKeyId, secretAccessKey, roleArn} = credentials;
+    let polly;
+    if (accessKeyId && secretAccessKey) {
+      polly = new PollyClient({
+        region,
+        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',
@@ -748,6 +772,52 @@ const synthPlayHT = async(logger, {
   }
 };
 
+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 (!process.env.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 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 */

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.0.50",
+  "version": "0.1.0",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",

package/test/synth.js

@@ -162,6 +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);
@@ -581,14 +608,48 @@ test('PlayHT speech synth tests', async(t) => {
       text,
       renderForCaching: true
     });
-    t.ok(!opts.servedFromCache, `successfully synthesized eleven audio to ${opts.filePath}`);
+    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('..');