nestworker

A simple library that provides an abstraction for the Node.js worker_threads module. You can run your function in a dedicated thread by working with Promises.

Example

import { executeInThread } from 'nestworker';

async function calculate(): Promise<void> {
  const values = await Promise.all([
    executeInThread(() => 2 ** 10), // this doesn't block the main thread
    executeInThread(() => 3 ** 10),
  ]);

  console.log(values); // [1024, 59049]
}

calculate();

This example demonstrates the optimization of two resource-intensive calculations through parallel execution in distinct threads. By distributing the tasks across separate threads, significant time savings are achieved.

Nestworker's takes a task function as its parameter, orchestrates its execution in a new thread, and subsequently delivers a Promise.

Surprisingly simple, isn't it?

Installation

$ npm i nestworker

All examples:

• Basic example
• Parameters for the thread task
• Async function inside the thread
• Error handling
• Use modules inside the thread

API

executeInThread(task, { args: any[] }

Runs the specified function in a separate thread.

Parameters

• Task (Function): The function to be executed in a thread.
  • This can also be a async function (promise).
• ...params (Any): Additional arguments to be passed to the Task function.
  • Parameter cann't be a function.

const task = function(a: number, b: object, c: boolean) { ... };
executeInThread(task, { args: [1, {}, true] })

The executeInThread function allows you to execute a given task function in a dedicated thread, similar to the behavior of setTimeout or setInterval. You provide the main function to be executed, along with any additional arguments (...args) that should be passed to the given function.

Returns

Promise<any>: A Promise that resolves with the return value of the callback.

Inside the provided function, you have the flexibility to return any value, including a Promise. The returned value, whether it's a standard value or a Promise, will be passed back to you as the resolved result of the Promise returned by the executeInThread function.

const number = await executeInThread(() => 123); // 123
const name = await executeInThread(() => Promise.resolve('John')); // John

Important (limitation)

Access to data outside of the task function is restricted. If you require the use of a module, it should be required within the task function. The sole method for accessing data within a task function from external sources is through the utilization of the parameters. Closures do not function in this context.

In this example, we're reading a file in a separate thread and returning the data in string format. We start by defining a task function that will run within the thread, and then we prepare the necessary parameters to be passed as inputs to that function.

import { executeInThread } from 'nestworker';

async function task(filename: string) {
// Closure doesn't work here
  const { readFile } = await import('fs/promises');
  const content = await readFile(filename);
  return content.toString();
}

async function read() {
  const content = await executeInThread(task, { args: [filename] });
  console.log(content);
}

read();

There is also another option if you don't want to use import inside the function.

import { executeInThread } from 'nestworker';

// this will be executed in a dedicated thread
async function task(modules: { 'fs/promises': typeof import('fs/promises') }) {
  // Closure doesn't work here
  const { readFile } = modules['fs/promises'];

  const content = await readFile(__filename);

  return content.toString();
}

async function read() {
  const content = await executeInThread(task, {
    threadModules: ['fs/promises'],
  });

  console.log(content);
}

read();

The threadModules parameter is an array of strings that represent the modules you want to use in the thread. The modules will be imported and passed to the task function first argument as an object.

Contributing

See the contributing guide for detailed instructions on how to get started with our project.

Author

Vahe Hakobyan: Telegram

License

Licensed under MIT.