Node.js SDK

Prerequisites

Before you use the SDK, read the API reference.

Download and install

1. Download and install the SDK.

   Run the following command to download and install the SDK.

   npm install alibabacloud-nls

2. Import the SDK.

   Use `require` or `import` in your code to import the SDK.

   const Nls = require('alibabacloud-nls')
   // Nls includes SpeechRecognition, SpeechTranscription, and SpeechSynthesizer.
   // The following code shows how to import the SDK using import.
   //import { SpeechRecognition } from "alibabacloud-nls"
   //import { SpeechTranscription } from "alibabacloud-nls"
   //import { SpeechSynthesizer } from "alibabacloud-nls"

Key interfaces and parameters

To implement speech synthesis in Node.js, use the SpeechSynthesizer class and follow the steps below. Steps 2 and 3 are interchangeable.

1. Create a SpeechSynthesizer instance and pass the speech synthesis endpoint and authentication information.

2. Set properties for speech synthesis, such as the voice, sample rate, and audio format. You can do this by creating a new property object or by modifying the default property object that is returned by the defaultStartParams method of the SpeechSynthesizer instance.

3. Set up the on callback function for the SpeechSynthesizer instance. This function uses the observer pattern. The server uses the on callback function to notify the client when events occur, such as a successful connection, synthesized speech, or an exception.

4. Call the start function of the SpeechSynthesizer instance to start speech synthesis.

let tts = new Nls.SpeechSynthesizer({
    url: URL,
    appkey: APPKEY,
    token: TOKEN
})

{
  voice: voice, // The value of voice is the parameter passed when the function is called.
  format: "wav",
  sample_rate: 16000,
  volume: 50,
  speech_rate: 0,
  pitch_rate: 0,
  enable_subtitle: false
}

let param = tts.defaultStartParams();
// Text to be synthesized
param.text = "I raise my head to the bright moon, and lower it to think of home.";
// Voice
param.voice = "aixia";
// Pitch. Range: -500 to 500. Optional. Default value: 0.
// param.pitch_rate = 100;
// Speech rate. Range: -500 to 500. Default value: 0.
// param.speech_rate = 100;
// Set the encoding format of the returned audio.
// param.format = "wav";
// Set the sample rate of the returned audio.
// param.sample_rate = 16000;
// Specifies whether to enable word-level timestamps.
// param.enable_subtitle = true;

let param = {};
// Text to be synthesized
param.text = line;
// Voice
param.voice = "aixia";
// Pitch. Range: -500 to 500. Optional. Default value: 0.
// param.pitch_rate = 100;
// Speech rate. Range: -500 to 500. Default value: 0.
// param.speech_rate = 100;
// Set the encoding format of the returned audio.
// param.format = "wav";
// Set the sample rate of the returned audio.
// param.sample_rate = 16000;
// Specifies whether to enable word-level timestamps.
// param.enable_subtitle = true;

on(which, handler) {
  this._event.on(which, handler)
}

let tts = new Nls.SpeechSynthesizer({
  url: URL,
  appkey:APPKEY,
  token:TOKEN
})

tts.on("meta", (msg)=>{
  console.log("Client received metainfo:", msg)
})

tts.on("data", (msg)=>{
  console.log(`Received size: ${msg.length}`)
  console.log(dumpFile.write(msg, "binary"))
})

tts.on("completed", (msg)=>{
  console.log("Client received completed:", msg)
})

tts.on("closed", () => {
  console.log("Client received closed")
})

tts.on("failed", (msg)=>{
  console.log("Client received failed:", msg)
})

async start(param, enablePing, pingInterval)

Code example

"use strict"

require('log-timestamp')(`${process.pid}`)
const fs = require("fs")
const Nls = require("alibabacloud-nls")
const sleep = (waitTimeInMs) => new Promise(resolve => setTimeout(resolve, waitTimeInMs))
const util = require("util")
const readline = require("readline")
const args = process.argv.slice(2)
//const Memwatch = require("node-memwatch-new")

const URL = "wss://nls-gateway-cn-shanghai.aliyuncs.com/ws/v1"
const APPKEY = "Your Appkey"      // To obtain an AppKey, go to the console: https://nls-portal.console.aliyun.com/applist
const TOKEN = "Your Token"      // For information about how to obtain a token, see https://help.aliyun.com/document_detail/450514.html

let b1 = []
let loadIndex = 0
//let hd = new Memwatch.HeapDiff()
let needDump = true

async function runOnce(line) {
  console.log(`speak: ${line}`)
  loadIndex++

  //let dumpFile = fs.createWriteStream(`${process.pid}.wav`, {flags:"w"})
  let tts = new Nls.SpeechSynthesizer({
    url: URL,
    appkey:APPKEY,
    token:TOKEN
  })

  tts.on("meta", (msg)=>{
    console.log("Client received metainfo:", msg)
  })

  tts.on("data", (msg)=>{
    console.log(`Received size: ${msg.length}`)
    //console.log(dumpFile.write(msg, "binary"))
  })

  tts.on("completed", (msg)=>{
    console.log("Client received completed:", msg)
  })

  tts.on("closed", () => {
    console.log("Client received closed")
  })

  tts.on("failed", (msg)=>{
    console.log("Client received failed:", msg)
  })

  let param = tts.defaultStartParams()
  // Text to be synthesized
  param.text = line
  // Voice
  param.voice = "aixia"
  // Pitch. Range: -500 to 500. Optional. Default value: 0.
  // param.pitch_rate = 100
  // Speech rate. Range: -500 to 500. Default value: 0.
  // param.speech_rate = 100
  // Set the encoding format of the returned audio.
  // param.format = "wav"
  // Set the sample rate of the returned audio.
  // param.sample_rate = 16000
  // Specifies whether to enable word-level timestamps.
  // param.enable_subtitle = true
  try {
    await tts.start(param, true, 6000)
  } catch(error) {
    console.log("error on start:", error)
    return
  } finally {
    //dumpFile.end()
  }
  console.log("synthesis done")
  await sleep(2000)
}

async function test() {
  console.log("Load test case:", args[0])
  const fileStream = fs.createReadStream(args[0])
  const rl = readline.createInterface({
    input: fileStream,
    crlfDelay: Infinity
  })

  for await (const line of rl) {
    b1.push(line)
  }

  while (true) {
    for (let text of b1) {
      await runOnce(text)
    }
  }
}

test()