0.2.11

Feat/893 azure ssml

lib/add-file-to-cache.js

@@ -25,7 +25,7 @@ function getExtensionAndSampleRate(path) {
 }
 
 async function addFileToCache(client, logger, path,
-  {account_sid, vendor, language, voice, deploymentId, engine, model, text}) {
+  {account_sid, vendor, language, voice, deploymentId, engine, model, text, instructions}) {
   let key;
   logger = logger || noopLogger;
 
@@ -38,6 +38,7 @@ async function addFileToCache(client, logger, path,
       engine,
       model,
       text,
+      instructions
     });
     const [extension, sampleRate] = getExtensionAndSampleRate(path);
     const audioBuffer = await fs.readFile(path);

lib/config.js

@@ -2,6 +2,7 @@ const JAMBONES_TTS_TRIM_SILENCE = process.env.JAMBONES_TTS_TRIM_SILENCE;
 const JAMBONES_DISABLE_TTS_STREAMING = process.env.JAMBONES_DISABLE_TTS_STREAMING;
 const JAMBONES_DISABLE_AZURE_TTS_STREAMING = process.env.JAMBONES_DISABLE_AZURE_TTS_STREAMING;
 const JAMBONES_EAGERLY_PRE_CACHE_AUDIO = process.env.JAMBONES_EAGERLY_PRE_CACHE_AUDIO;
+const JAMBONES_AZURE_ENABLE_SSML = process.env.JAMBONES_AZURE_ENABLE_SSML;
 
 const JAMBONES_HTTP_PROXY_IP = process.env.JAMBONES_HTTP_PROXY_IP;
 const JAMBONES_HTTP_PROXY_PORT = process.env.JAMBONES_HTTP_PROXY_PORT;
@@ -21,5 +22,6 @@ module.exports = {
   JAMBONES_TTS_CACHE_DURATION_MINS,
   JAMBONES_EAGERLY_PRE_CACHE_AUDIO,
   TMP_FOLDER,
-  HTTP_TIMEOUT
+  HTTP_TIMEOUT,
+  JAMBONES_AZURE_ENABLE_SSML
 };

lib/purge-tts-cache.js

@@ -12,7 +12,7 @@ const debug = require('debug')('jambonz:realtimedb-helpers');
  * @returns {object} result - {error, purgedCount}
  */
 async function purgeTtsCache(client, logger, {all, account_sid, vendor,
-  language, voice, deploymentId, engine, model, text} = {all: true}) {
+  language, voice, deploymentId, engine, model, text, instructions} = {all: true}) {
   logger = logger || noopLogger;
 
   let purgedCount = 0, error;
@@ -35,6 +35,7 @@ async function purgeTtsCache(client, logger, {all, account_sid, vendor,
         engine,
         model,
         text,
+        instructions
       });
       purgedCount = await client.del(key);
       if (purgedCount === 0) error = 'Specified item not found';

lib/synth-audio.js

@@ -47,6 +47,7 @@ const {
   JAMBONES_HTTP_PROXY_PORT,
   JAMBONES_TTS_CACHE_DURATION_MINS,
   JAMBONES_TTS_TRIM_SILENCE,
+  JAMBONES_AZURE_ENABLE_SSML
 } = require('./config');
 const EXPIRES = JAMBONES_TTS_CACHE_DURATION_MINS;
 const OpenAI = require('openai');
@@ -89,7 +90,7 @@ const trimTrailingSilence = (buffer) => {
  */
 async function synthAudio(client, createHash, retrieveHash, logger, stats, { account_sid,
   vendor, language, voice, gender, text, engine, salt, model, credentials, deploymentId,
-  disableTtsCache, renderForCaching = false, disableTtsStreaming, options
+  disableTtsCache, renderForCaching = false, disableTtsStreaming, options, instructions
 }) {
   let audioData;
   let servedFromCache = false;
@@ -171,7 +172,8 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
     engine,
     // model or model_id is used to identify the tts cache.
     model: model || credentials.model_id,
-    text
+    text,
+    instructions
   });
 
   debug(`synth key is ${key}`);
@@ -215,7 +217,8 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
         audioData = await synthNuance(client, logger, {credentials, stats, voice, model, text});
         break;
       case 'nvidia':
-        audioData = await synthNvidia(client, logger, {credentials, stats, language, voice, model, text});
+        audioData = await synthNvidia(client, logger, {credentials, stats, language, voice, model, text,
+          renderForCaching, disableTtsStreaming});
         break;
       case 'ibm':
         audioData = await synthIbm(logger, {credentials, stats, voice, text});
@@ -241,7 +244,7 @@ async function synthAudio(client, createHash, retrieveHash, logger, stats, { acc
         break;
       case 'whisper':
         audioData = await synthWhisper(logger, {
-          credentials, stats, voice, text, renderForCaching, disableTtsStreaming});
+          credentials, stats, voice, text, instructions, renderForCaching, disableTtsStreaming});
         break;
       case 'verbio':
         audioData = await synthVerbio(client, logger, {
@@ -464,7 +467,6 @@ async function _synthOnPremMicrosoft(logger, {
 }) {
   const {use_custom_tts, custom_tts_endpoint_url, api_key} = credentials;
   let content = text;
-
   if (use_custom_tts && !content.startsWith('<speak')) {
     /**
      * Note: it seems that to use custom voice ssml is required with the voice attribute
@@ -477,9 +479,13 @@ async function _synthOnPremMicrosoft(logger, {
     /* microsoft enforces some properties and uses voice xml element so if the user did not supply do it for them */
     const words = content.slice(7, -8).trim().replace(/(\r\n|\n|\r)/gm, ' ');
     // eslint-disable-next-line max-len
-    content = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="${language}"><voice name="${voice}">${words}</voice></speak>`;
+    content = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="${language}"><lang xml:lang="${language}"><voice name="${voice}">${words}</voice></lang></speak>`;
     logger.info({content}, 'synthMicrosoft');
   }
+  else if (JAMBONES_AZURE_ENABLE_SSML) {
+    // 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}"><lang xml:lang="${language}">${text}</lang></voice></speak>`;
+  }
 
   try {
     const trimSilence = JAMBONES_TTS_TRIM_SILENCE;
@@ -516,19 +522,23 @@ const synthMicrosoft = async(logger, {
     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
+     * 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"
+     * 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 */
+    /* 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}">${words}</voice></speak>`;
+      content = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="${language}"><lang xml:lang="${language}"><voice name="${voice}">${words}</voice></lang></speak>`;
       logger.info({content}, 'synthMicrosoft');
     }
+    else if (JAMBONES_AZURE_ENABLE_SSML) {
+      // eslint-disable-next-line max-len
+      content = `<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xmlns:mstts="https://www.w3.org/2001/mstts" xml:lang="${language}"><voice name="${voice}"><lang xml:lang="${language}">${text}</lang></voice></speak>`;
+    }
     if (!JAMBONES_DISABLE_TTS_STREAMING && !JAMBONES_DISABLE_AZURE_TTS_STREAMING &&
       !renderForCaching && !disableTtsStreaming) {
       let params = '';
@@ -708,8 +718,24 @@ const synthNuance = async(client, logger, {credentials, stats, voice, model, tex
   });
 };
 
-const synthNvidia = async(client, logger, {credentials, stats, language,  voice, model, text}) => {
+const synthNvidia = async(client, logger, {
+  credentials, stats, language,  voice, model, text, renderForCaching, disableTtsStreaming
+}) => {
   const {riva_server_uri} = credentials;
+  if (!JAMBONES_DISABLE_TTS_STREAMING && !renderForCaching && !disableTtsStreaming) {
+    let params = '';
+    params += `{riva_server_uri=${riva_server_uri}`;
+    params += `,voice=${voice}`;
+    params += `,language=${language}`;
+    params += ',write_cache_file=1';
+    params += '}';
+
+    return {
+      filePath: `say:${params}${text.replace(/\n/g, ' ')}`,
+      servedFromCache: false,
+      rtt: 0
+    };
+  }
   let rivaClient, request;
   const sampleRate = 8000;
   try {
@@ -837,11 +863,10 @@ const synthElevenlabs = async(logger, {
 const synthPlayHT = async(client, logger, {
   credentials, options, stats, voice, language, text, renderForCaching, disableTtsStreaming
 }) => {
-  const {api_key, user_id, voice_engine, options: credOpts} = credentials;
+  const {api_key, user_id, voice_engine, playht_tts_uri, options: credOpts} = credentials;
   const opts = !!options && Object.keys(options).length !== 0 ? options : JSON.parse(credOpts || '{}');
 
-  let synthesizeUrl = 'https://api.play.ht/api/v2/tts/stream';
+  let synthesizeUrl = playht_tts_uri ? `${playht_tts_uri}/api/v2/tts/stream` : 'https://api.play.ht/api/v2/tts/stream';
-
   // If model is play3.0, the synthesizeUrl is got from authentication endpoint
   if (voice_engine === 'Play3.0') {
     try {
@@ -946,6 +971,11 @@ const synthRimelabs = async(logger, {
     params += ',write_cache_file=1';
     if (opts.speedAlpha) params += `,speed_alpha=${opts.speedAlpha}`;
     if (opts.reduceLatency) params += `,reduce_latency=${opts.reduceLatency}`;
+    // Arcana model parameters
+    if (opts.temperature) params += `,temperature=${opts.temperature}`;
+    if (opts.repetition_penalty) params += `,repetition_penalty=${opts.repetition_penalty}`;
+    if (opts.top_p) params += `,top_p=${opts.top_p}`;
+    if (opts.max_tokens) params += `,max_tokens=${opts.max_tokens}`;
     params += '}';
 
     return {
@@ -1026,7 +1056,8 @@ const synthVerbio = async(client, logger, {credentials, stats, voice, text, rend
   }
 };
 
-const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCaching, disableTtsStreaming}) => {
+const synthWhisper = async(logger, {credentials, stats, voice, text, instructions,
+  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) {
@@ -1037,6 +1068,8 @@ const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCa
     params += `,voice=${voice}`;
     params += ',write_cache_file=1';
     if (speed) params += `,speed=${speed}`;
+    // comma is used to separated parameters in freeswitch tts module
+    if (instructions) params += `,instructions=${instructions.replace(/\n/g, ' ').replace(/,/g, ';')}`;
     params += '}';
 
     return {
@@ -1056,6 +1089,7 @@ const synthWhisper = async(logger, {credentials, stats, voice, text, renderForCa
       model: model_id,
       voice,
       input: text,
+      ...(instructions && {instructions}),
       response_format: 'mp3'
     });
     return {

lib/utils.js

@@ -23,10 +23,11 @@ function makeSynthKey({
   voice,
   engine = '',
   model = '',
-  text
+  text,
+  instructions = '',
 }) {
   const hash = crypto.createHash('sha1');
-  hash.update(`${language}:${vendor}:${voice}:${engine}:${model}:${text}`);
+  hash.update(`${language}:${vendor}:${voice}:${engine}:${model}:${text}:${instructions}`);
   const hexHashKey = hash.digest('hex');
   const accountKey = account_sid ? `:${account_sid}` : '';
   const key = `tts${accountKey}:${hexHashKey}`;

package-lock.json

@@ -1,14 +1,15 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.2.6",
+  "version": "0.2.11",
   "lockfileVersion": 2,
   "requires": true,
   "packages": {
     "": {
       "name": "@jambonz/speech-utils",
-      "version": "0.2.6",
+      "version": "0.2.11",
       "license": "MIT",
       "dependencies": {
+        "23": "^0.0.0",
         "@aws-sdk/client-polly": "^3.496.0",
         "@aws-sdk/client-sts": "^3.496.0",
         "@cartesia/cartesia-js": "^2.1.0",
@@ -21,7 +22,7 @@
         "google-protobuf": "^3.21.2",
         "ibm-watson": "^8.0.0",
         "microsoft-cognitiveservices-speech-sdk": "1.38.0",
-        "openai": "^4.25.0",
+        "openai": "^4.98.0",
         "undici": "^7.5.0"
       },
       "devDependencies": {
@@ -1107,27 +1108,6 @@
         "node": "^12.22.0 || ^14.17.0 || >=16.0.0"
       }
     },
-    "node_modules/@cartesia/cartesia-js/node_modules/ws": {
-      "version": "8.18.0",
-      "resolved": "https://registry.npmjs.org/ws/-/ws-8.18.0.tgz",
-      "integrity": "sha512-8VbfWfHLbbwu3+N6OKsOMpBdT4kXPDDB9cJk2bJ6mh9ucxdlnNvH1e+roYkKmN9Nxw2yjz7VzeO9oOz2zJ04Pw==",
-      "license": "MIT",
-      "engines": {
-        "node": ">=10.0.0"
-      },
-      "peerDependencies": {
-        "bufferutil": "^4.0.1",
-        "utf-8-validate": ">=5.0.2"
-      },
-      "peerDependenciesMeta": {
-        "bufferutil": {
-          "optional": true
-        },
-        "utf-8-validate": {
-          "optional": true
-        }
-      }
-    },
     "node_modules/@eslint-community/eslint-utils": {
       "version": "4.4.0",
       "resolved": "https://registry.npmjs.org/@eslint-community/eslint-utils/-/eslint-utils-4.4.0.tgz",
@@ -2299,6 +2279,12 @@
       "integrity": "sha512-I4q9QU9MQv4oEOz4tAHJtNz1cwuLxn2F3xcc2iV5WdqLPpUnj30aUuxt1mAxYTG+oe8CZMV/+6rU4S4gRDzqtQ==",
       "license": "MIT"
     },
+    "node_modules/23": {
+      "version": "0.0.0",
+      "resolved": "https://registry.npmjs.org/23/-/23-0.0.0.tgz",
+      "integrity": "sha512-uAETf9Okr72trtp1pNXYKhFCTTI1EKGcYMA8gw3jLGhlbaDX+grrNToEWrpt8luxRAvrZWBTvyB5wk3PpNjGQQ==",
+      "license": "ISC"
+    },
     "node_modules/abort-controller": {
       "version": "3.0.0",
       "resolved": "https://registry.npmjs.org/abort-controller/-/abort-controller-3.0.0.tgz",
@@ -2541,11 +2527,6 @@
       "integrity": "sha512-3oSeUO0TMV67hN1AmbXsK4yaqU7tjiHlbxRDZOpH0KW9+CeX4bRAaX0Anxt0tx2MrpRpWwQaPwIlISEJhYU5Pw==",
       "dev": true
     },
-    "node_modules/base-64": {
-      "version": "0.1.0",
-      "resolved": "https://registry.npmjs.org/base-64/-/base-64-0.1.0.tgz",
-      "integrity": "sha512-Y5gU45svrR5tI2Vt/X9GPd3L0HNIKzGu202EjxrXMpuc2V2CiKgemAbUUsqYmZJvPtCXoUKjNZwBJzsNScUbXA=="
-    },
     "node_modules/base64-js": {
       "version": "1.5.1",
       "resolved": "https://registry.npmjs.org/base64-js/-/base64-js-1.5.1.tgz",
@@ -2781,14 +2762,6 @@
         "url": "https://github.com/chalk/chalk?sponsor=1"
       }
     },
-    "node_modules/charenc": {
-      "version": "0.0.2",
-      "resolved": "https://registry.npmjs.org/charenc/-/charenc-0.0.2.tgz",
-      "integrity": "sha512-yrLQ/yVUFXkzg7EDQsPieE/53+0RlaWTs+wBrvW36cyilJ2SaDWfl4Yj7MtLTXleV9uEKefbAGUPv2/iWSooRA==",
-      "engines": {
-        "node": "*"
-      }
-    },
     "node_modules/clean-stack": {
       "version": "2.2.0",
       "resolved": "https://registry.npmjs.org/clean-stack/-/clean-stack-2.2.0.tgz",
@@ -2891,14 +2864,6 @@
         "node": ">= 8"
       }
     },
-    "node_modules/crypt": {
-      "version": "0.0.2",
-      "resolved": "https://registry.npmjs.org/crypt/-/crypt-0.0.2.tgz",
-      "integrity": "sha512-mCxBlsHFYh9C+HVpiEacem8FEBnMXgU9gy4zmNC+SXAZNB/1idgp/aulFJ4FgCi7GPEVbfyng092GqL2k2rmow==",
-      "engines": {
-        "node": "*"
-      }
-    },
     "node_modules/d": {
       "version": "1.0.1",
       "resolved": "https://registry.npmjs.org/d/-/d-1.0.1.tgz",
@@ -3053,15 +3018,6 @@
         "node": "^10.13.0 || ^12.13.0 || ^14.15.0 || >=15.0.0"
       }
     },
-    "node_modules/digest-fetch": {
-      "version": "1.3.0",
-      "resolved": "https://registry.npmjs.org/digest-fetch/-/digest-fetch-1.3.0.tgz",
-      "integrity": "sha512-CGJuv6iKNM7QyZlM2T3sPAdZWd/p9zQiRNS9G+9COUCwzWFTs0Xp8NF5iePx7wtvhDykReiRRrSeNb4oMmB8lA==",
-      "dependencies": {
-        "base-64": "^0.1.0",
-        "md5": "^2.3.0"
-      }
-    },
     "node_modules/dotenv": {
       "version": "16.4.7",
       "resolved": "https://registry.npmjs.org/dotenv/-/dotenv-16.4.7.tgz",
@@ -4531,11 +4487,6 @@
         "url": "https://github.com/sponsors/ljharb"
       }
     },
-    "node_modules/is-buffer": {
-      "version": "1.1.6",
-      "resolved": "https://registry.npmjs.org/is-buffer/-/is-buffer-1.1.6.tgz",
-      "integrity": "sha512-NcdALwpXkTm5Zvvbk7owOUSvVvBKDgKP5/ewfXEznmQFfs4ZRmanOeKBTjRVjka3QFoN6XJ+9F3USqfHqTaU5w=="
-    },
     "node_modules/is-callable": {
       "version": "1.2.7",
       "resolved": "https://registry.npmjs.org/is-callable/-/is-callable-1.2.7.tgz",
@@ -5259,16 +5210,6 @@
         "semver": "bin/semver.js"
       }
     },
-    "node_modules/md5": {
-      "version": "2.3.0",
-      "resolved": "https://registry.npmjs.org/md5/-/md5-2.3.0.tgz",
-      "integrity": "sha512-T1GITYmFaKuO91vxyoQMFETst+O71VUPEU3ze5GNzDm0OWdP8v1ziTaAEPUr/3kLsY3Sftgz242A1SetQiDL7g==",
-      "dependencies": {
-        "charenc": "0.0.2",
-        "crypt": "0.0.2",
-        "is-buffer": "~1.1.6"
-      }
-    },
     "node_modules/micromatch": {
       "version": "4.0.8",
       "resolved": "https://registry.npmjs.org/micromatch/-/micromatch-4.0.8.tgz",
@@ -5328,6 +5269,27 @@
         "uuid": "dist/bin/uuid"
       }
     },
+    "node_modules/microsoft-cognitiveservices-speech-sdk/node_modules/ws": {
+      "version": "7.5.10",
+      "resolved": "https://registry.npmjs.org/ws/-/ws-7.5.10.tgz",
+      "integrity": "sha512-+dbF1tHwZpXcbOJdVOkzLDxZP1ailvSxM6ZweXTegylPny803bFhA+vqBYw4s31NSAk4S2Qz+AKXK9a4wkdjcQ==",
+      "license": "MIT",
+      "engines": {
+        "node": ">=8.3.0"
+      },
+      "peerDependencies": {
+        "bufferutil": "^4.0.1",
+        "utf-8-validate": "^5.0.2"
+      },
+      "peerDependenciesMeta": {
+        "bufferutil": {
+          "optional": true
+        },
+        "utf-8-validate": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/mime-db": {
       "version": "1.52.0",
       "resolved": "https://registry.npmjs.org/mime-db/-/mime-db-1.52.0.tgz",
@@ -5723,22 +5685,33 @@
       }
     },
     "node_modules/openai": {
-      "version": "4.25.0",
+      "version": "4.98.0",
-      "resolved": "https://registry.npmjs.org/openai/-/openai-4.25.0.tgz",
+      "resolved": "https://registry.npmjs.org/openai/-/openai-4.98.0.tgz",
-      "integrity": "sha512-qLMFOizjxKuDfQkBrczZPYo6XVL4bdcuz9MR11Q+M91kGcs8dQw+O90nRcC+qWuhaGphQkfXQJMn4cd7Yew3Kg==",
+      "integrity": "sha512-TmDKur1WjxxMPQAtLG5sgBSCJmX7ynTsGmewKzoDwl1fRxtbLOsiR0FA/AOAAtYUmP6azal+MYQuOENfdU+7yg==",
+      "license": "Apache-2.0",
       "dependencies": {
         "@types/node": "^18.11.18",
         "@types/node-fetch": "^2.6.4",
         "abort-controller": "^3.0.0",
         "agentkeepalive": "^4.2.1",
-        "digest-fetch": "^1.3.0",
         "form-data-encoder": "1.7.2",
         "formdata-node": "^4.3.2",
-        "node-fetch": "^2.6.7",
+        "node-fetch": "^2.6.7"
-        "web-streams-polyfill": "^3.2.1"
       },
       "bin": {
         "openai": "bin/cli"
+      },
+      "peerDependencies": {
+        "ws": "^8.18.0",
+        "zod": "^3.23.8"
+      },
+      "peerDependenciesMeta": {
+        "ws": {
+          "optional": true
+        },
+        "zod": {
+          "optional": true
+        }
       }
     },
     "node_modules/openai/node_modules/@types/node": {
@@ -7162,14 +7135,6 @@
         "uuid": "dist/bin/uuid"
       }
     },
-    "node_modules/web-streams-polyfill": {
-      "version": "3.2.1",
-      "resolved": "https://registry.npmjs.org/web-streams-polyfill/-/web-streams-polyfill-3.2.1.tgz",
-      "integrity": "sha512-e0MO3wdXWKrLbL0DgGnUV7WHVuw9OUvL4hjgnPkIeEvESk74gAITi5G606JtZPp39cd8HA9VQzCIvA49LpPN5Q==",
-      "engines": {
-        "node": ">= 8"
-      }
-    },
     "node_modules/webidl-conversions": {
       "version": "3.0.1",
       "resolved": "https://registry.npmjs.org/webidl-conversions/-/webidl-conversions-3.0.1.tgz",
@@ -7318,16 +7283,16 @@
       }
     },
     "node_modules/ws": {
-      "version": "7.5.10",
+      "version": "8.18.2",
-      "resolved": "https://registry.npmjs.org/ws/-/ws-7.5.10.tgz",
+      "resolved": "https://registry.npmjs.org/ws/-/ws-8.18.2.tgz",
-      "integrity": "sha512-+dbF1tHwZpXcbOJdVOkzLDxZP1ailvSxM6ZweXTegylPny803bFhA+vqBYw4s31NSAk4S2Qz+AKXK9a4wkdjcQ==",
+      "integrity": "sha512-DMricUmwGZUVr++AEAe2uiVM7UoO9MAVZMDu05UQOaUII0lp+zOzLLU4Xqh/JvTqklB1T4uELaaPBKyjE1r4fQ==",
       "license": "MIT",
       "engines": {
-        "node": ">=8.3.0"
+        "node": ">=10.0.0"
       },
       "peerDependencies": {
         "bufferutil": "^4.0.1",
-        "utf-8-validate": "^5.0.2"
+        "utf-8-validate": ">=5.0.2"
       },
       "peerDependenciesMeta": {
         "bufferutil": {
@@ -7398,6 +7363,11 @@
     }
   },
   "dependencies": {
+    "23": {
+      "version": "0.0.0",
+      "resolved": "https://registry.npmjs.org/23/-/23-0.0.0.tgz",
+      "integrity": "sha512-uAETf9Okr72trtp1pNXYKhFCTTI1EKGcYMA8gw3jLGhlbaDX+grrNToEWrpt8luxRAvrZWBTvyB5wk3PpNjGQQ=="
+    },
     "@aashutoshrathi/word-wrap": {
       "version": "1.2.6",
       "resolved": "https://registry.npmjs.org/@aashutoshrathi/word-wrap/-/word-wrap-1.2.6.tgz",
@@ -8232,12 +8202,6 @@
             "process": "^0.11.10",
             "string_decoder": "^1.3.0"
           }
-        },
-        "ws": {
-          "version": "8.18.0",
-          "resolved": "https://registry.npmjs.org/ws/-/ws-8.18.0.tgz",
-          "integrity": "sha512-8VbfWfHLbbwu3+N6OKsOMpBdT4kXPDDB9cJk2bJ6mh9ucxdlnNvH1e+roYkKmN9Nxw2yjz7VzeO9oOz2zJ04Pw==",
-          "requires": {}
         }
       }
     },
@@ -9328,11 +9292,6 @@
       "integrity": "sha512-3oSeUO0TMV67hN1AmbXsK4yaqU7tjiHlbxRDZOpH0KW9+CeX4bRAaX0Anxt0tx2MrpRpWwQaPwIlISEJhYU5Pw==",
       "dev": true
     },
-    "base-64": {
-      "version": "0.1.0",
-      "resolved": "https://registry.npmjs.org/base-64/-/base-64-0.1.0.tgz",
-      "integrity": "sha512-Y5gU45svrR5tI2Vt/X9GPd3L0HNIKzGu202EjxrXMpuc2V2CiKgemAbUUsqYmZJvPtCXoUKjNZwBJzsNScUbXA=="
-    },
     "base64-js": {
       "version": "1.5.1",
       "resolved": "https://registry.npmjs.org/base64-js/-/base64-js-1.5.1.tgz",
@@ -9470,11 +9429,6 @@
         "supports-color": "^7.1.0"
       }
     },
-    "charenc": {
-      "version": "0.0.2",
-      "resolved": "https://registry.npmjs.org/charenc/-/charenc-0.0.2.tgz",
-      "integrity": "sha512-yrLQ/yVUFXkzg7EDQsPieE/53+0RlaWTs+wBrvW36cyilJ2SaDWfl4Yj7MtLTXleV9uEKefbAGUPv2/iWSooRA=="
-    },
     "clean-stack": {
       "version": "2.2.0",
       "resolved": "https://registry.npmjs.org/clean-stack/-/clean-stack-2.2.0.tgz",
@@ -9555,11 +9509,6 @@
         "which": "^2.0.1"
       }
     },
-    "crypt": {
-      "version": "0.0.2",
-      "resolved": "https://registry.npmjs.org/crypt/-/crypt-0.0.2.tgz",
-      "integrity": "sha512-mCxBlsHFYh9C+HVpiEacem8FEBnMXgU9gy4zmNC+SXAZNB/1idgp/aulFJ4FgCi7GPEVbfyng092GqL2k2rmow=="
-    },
     "d": {
       "version": "1.0.1",
       "resolved": "https://registry.npmjs.org/d/-/d-1.0.1.tgz",
@@ -9666,15 +9615,6 @@
       "resolved": "https://registry.npmjs.org/diff-sequences/-/diff-sequences-27.5.1.tgz",
       "integrity": "sha512-k1gCAXAsNgLwEL+Y8Wvl+M6oEFj5bgazfZULpS5CneoPPXRaCCW7dm+q21Ky2VEE5X+VeRDBVg1Pcvvsr4TtNQ=="
     },
-    "digest-fetch": {
-      "version": "1.3.0",
-      "resolved": "https://registry.npmjs.org/digest-fetch/-/digest-fetch-1.3.0.tgz",
-      "integrity": "sha512-CGJuv6iKNM7QyZlM2T3sPAdZWd/p9zQiRNS9G+9COUCwzWFTs0Xp8NF5iePx7wtvhDykReiRRrSeNb4oMmB8lA==",
-      "requires": {
-        "base-64": "^0.1.0",
-        "md5": "^2.3.0"
-      }
-    },
     "dotenv": {
       "version": "16.4.7",
       "resolved": "https://registry.npmjs.org/dotenv/-/dotenv-16.4.7.tgz",
@@ -10737,11 +10677,6 @@
         "has-tostringtag": "^1.0.0"
       }
     },
-    "is-buffer": {
-      "version": "1.1.6",
-      "resolved": "https://registry.npmjs.org/is-buffer/-/is-buffer-1.1.6.tgz",
-      "integrity": "sha512-NcdALwpXkTm5Zvvbk7owOUSvVvBKDgKP5/ewfXEznmQFfs4ZRmanOeKBTjRVjka3QFoN6XJ+9F3USqfHqTaU5w=="
-    },
     "is-callable": {
       "version": "1.2.7",
       "resolved": "https://registry.npmjs.org/is-callable/-/is-callable-1.2.7.tgz",
@@ -11292,16 +11227,6 @@
         }
       }
     },
-    "md5": {
-      "version": "2.3.0",
-      "resolved": "https://registry.npmjs.org/md5/-/md5-2.3.0.tgz",
-      "integrity": "sha512-T1GITYmFaKuO91vxyoQMFETst+O71VUPEU3ze5GNzDm0OWdP8v1ziTaAEPUr/3kLsY3Sftgz242A1SetQiDL7g==",
-      "requires": {
-        "charenc": "0.0.2",
-        "crypt": "0.0.2",
-        "is-buffer": "~1.1.6"
-      }
-    },
     "micromatch": {
       "version": "4.0.8",
       "resolved": "https://registry.npmjs.org/micromatch/-/micromatch-4.0.8.tgz",
@@ -11344,6 +11269,12 @@
           "version": "9.0.1",
           "resolved": "https://registry.npmjs.org/uuid/-/uuid-9.0.1.tgz",
           "integrity": "sha512-b+1eJOlsR9K8HJpow9Ok3fiWOWSIcIzXodvv0rQjVoOVNpWMpxf1wZNpt4y9h10odCNrqnYp1OBzRktckBe3sA=="
+        },
+        "ws": {
+          "version": "7.5.10",
+          "resolved": "https://registry.npmjs.org/ws/-/ws-7.5.10.tgz",
+          "integrity": "sha512-+dbF1tHwZpXcbOJdVOkzLDxZP1ailvSxM6ZweXTegylPny803bFhA+vqBYw4s31NSAk4S2Qz+AKXK9a4wkdjcQ==",
+          "requires": {}
         }
       }
     },
@@ -11634,19 +11565,17 @@
       }
     },
     "openai": {
-      "version": "4.25.0",
+      "version": "4.98.0",
-      "resolved": "https://registry.npmjs.org/openai/-/openai-4.25.0.tgz",
+      "resolved": "https://registry.npmjs.org/openai/-/openai-4.98.0.tgz",
-      "integrity": "sha512-qLMFOizjxKuDfQkBrczZPYo6XVL4bdcuz9MR11Q+M91kGcs8dQw+O90nRcC+qWuhaGphQkfXQJMn4cd7Yew3Kg==",
+      "integrity": "sha512-TmDKur1WjxxMPQAtLG5sgBSCJmX7ynTsGmewKzoDwl1fRxtbLOsiR0FA/AOAAtYUmP6azal+MYQuOENfdU+7yg==",
       "requires": {
         "@types/node": "^18.11.18",
         "@types/node-fetch": "^2.6.4",
         "abort-controller": "^3.0.0",
         "agentkeepalive": "^4.2.1",
-        "digest-fetch": "^1.3.0",
         "form-data-encoder": "1.7.2",
         "formdata-node": "^4.3.2",
-        "node-fetch": "^2.6.7",
+        "node-fetch": "^2.6.7"
-        "web-streams-polyfill": "^3.2.1"
       },
       "dependencies": {
         "@types/node": {
@@ -12674,11 +12603,6 @@
       "integrity": "sha512-+NYs2QeMWy+GWFOEm9xnn6HCDp0l7QBD7ml8zLUmJ+93Q5NF0NocErnwkTkXVFNiX3/fpC6afS8Dhb/gz7R7eg==",
       "dev": true
     },
-    "web-streams-polyfill": {
-      "version": "3.2.1",
-      "resolved": "https://registry.npmjs.org/web-streams-polyfill/-/web-streams-polyfill-3.2.1.tgz",
-      "integrity": "sha512-e0MO3wdXWKrLbL0DgGnUV7WHVuw9OUvL4hjgnPkIeEvESk74gAITi5G606JtZPp39cd8HA9VQzCIvA49LpPN5Q=="
-    },
     "webidl-conversions": {
       "version": "3.0.1",
       "resolved": "https://registry.npmjs.org/webidl-conversions/-/webidl-conversions-3.0.1.tgz",
@@ -12802,9 +12726,9 @@
       }
     },
     "ws": {
-      "version": "7.5.10",
+      "version": "8.18.2",
-      "resolved": "https://registry.npmjs.org/ws/-/ws-7.5.10.tgz",
+      "resolved": "https://registry.npmjs.org/ws/-/ws-8.18.2.tgz",
-      "integrity": "sha512-+dbF1tHwZpXcbOJdVOkzLDxZP1ailvSxM6ZweXTegylPny803bFhA+vqBYw4s31NSAk4S2Qz+AKXK9a4wkdjcQ==",
+      "integrity": "sha512-DMricUmwGZUVr++AEAe2uiVM7UoO9MAVZMDu05UQOaUII0lp+zOzLLU4Xqh/JvTqklB1T4uELaaPBKyjE1r4fQ==",
       "requires": {}
     },
     "y18n": {

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@jambonz/speech-utils",
-  "version": "0.2.6",
+  "version": "0.2.11",
   "description": "TTS-related speech utilities for jambonz",
   "main": "index.js",
   "author": "Dave Horton",
@@ -26,6 +26,7 @@
   },
   "homepage": "https://github.com/jambonz/speech-utils#readme",
   "dependencies": {
+    "23": "^0.0.0",
     "@aws-sdk/client-polly": "^3.496.0",
     "@aws-sdk/client-sts": "^3.496.0",
     "@cartesia/cartesia-js": "^2.1.0",
@@ -38,7 +39,7 @@
     "google-protobuf": "^3.21.2",
     "ibm-watson": "^8.0.0",
     "microsoft-cognitiveservices-speech-sdk": "1.38.0",
-    "openai": "^4.25.0",
+    "openai": "^4.98.0",
     "undici": "^7.5.0"
   },
   "devDependencies": {

test/synth.js

@@ -887,7 +887,8 @@ test('TTS Cache tests', async(t) => {
     // save some random tts keys to cache
     const minRecords = 8;
     for (const i in Array(minRecords).fill(0)) {
-      await client.set(makeSynthKey({vendor: i, language: i, voice: i, engine: i, model: i, text: i}), i);
+      await client.set(makeSynthKey({vendor: i, language: i, voice: i, engine: i, model: i, text: i,
+        instructions: i}), i);
     }
     const count = await getTtsSize();
     t.ok(count >= minRecords, 'getTtsSize worked.');
@@ -906,7 +907,7 @@ test('TTS Cache tests', async(t) => {
   try {
     // save some random tts keys to cache
     for (const i in Array(10).fill(0)) {
-      await client.set(makeSynthKey({vendor: i, language: i, voice: i, engine: i, text: i}), i);
+      await client.set(makeSynthKey({vendor: i, language: i, voice: i, engine: i, text: i, instructions: i}), i);
     }
     // save a specific key to tts cache
     const opts = {vendor: 'aws', language: 'en-US', voice: 'MALE', engine: 'Engine', text: 'Hello World!'};
@@ -944,10 +945,10 @@ test('TTS Cache tests', async(t) => {
     const account_sid = "12412512_cabc_5aff"
     const account_sid2 = "22412512_cabc_5aff"
     for (const i in Array(minRecords).fill(0)) {
-      await client.set(makeSynthKey({account_sid, vendor: i, language: i, voice: i, engine: i, text: i}), i);
+      await client.set(makeSynthKey({account_sid, vendor: i, language: i, voice: i, engine: i, text: i, instructions: i}), i);
     }
     for (const i in Array(minRecords).fill(0)) {
-      await client.set(makeSynthKey({account_sid: account_sid2, vendor: i, language: i, voice: i, engine: i, text: i}), i);
+      await client.set(makeSynthKey({account_sid: account_sid2, vendor: i, language: i, voice: i, engine: i, text: i, instructions: i}), i);
     }
     const {purgedCount} = await purgeTtsCache({account_sid});
     t.equal(purgedCount, minRecords, `successfully purged at least ${minRecords} tts records from cache for account_sid:${account_sid}`);