Fibrelite

Seamlessly use Web Workers to offload work that could block user interaction and page rendering.

With Fibrelite you can turn any async function into a Web Worker. Fibrelite has three core strategies to executing work:

• Execute: run an async function as a Web Worker
• Debounce: Only run the an async function on last set of arguments in a batch of incoming operations
• Prioritise: Run all incoming operations but kill off workers as new operations come in

Demo

Check out Fibrelite in action here. This is a hosted version of the index.html and example.js inside this repository.

Usage

Install

For npm:

npm install fibrelite

Or for yarn:

yarn add fibrelite

CDN

Alternatively you can use the unpkg CDN like so:

<script src="https://unpkg.com/[email protected]/dist/fibrelite.js"></script>

Using Fibrelite

Fibrelite takes three arguments, the second two are optional:

const worker = new fibrelite(asyncHello, numberOfWorkersInPool, debounceInMilliseconds);

Changing numberOfWorkersInPool will make use of a pool of workers in for running you async function (defaults to 1). If you want to change the debounce period for worker.debounce you can set debounceInMilliseconds (defaults to 333 milliseconds).

You can use Fibrelite like this:

(async() => { 

    const asyncHello = async (input) => {
        return "Hello " + input;
    };
    const worker = new fibrelite(asyncHello);
    const response = await worker.execute("World!");
    // const response = await worker.debounce("World!");
    // const response = await worker.prioritise("World!");
    console.log(response); // logs 'Hello World!'

})(); // await calls must be wrapped in an async function

Development

To build Fibrelite you can run:

npm run build

To watch for file changes and build:

npm run watch

To run tests you can do:

npm run test

For convenience you can run the example using live-server using:

npm run serve

Acknowledgements

A massive thanks to Jason Miller for the Greenlet library on which this is heavily based and inspired.

License

MIT