This repo is inspired by paralleljs.
Make easier when handle parallel computing in javascript.
Unlike paralleljs, which stores the function to process any array data.
npm i worker-parallel
Use this constructor to create instance Parallel to store function need use to execute parallel process.
Example
const hardwork = (n: number): number => {
let out = 0
for(let i = 0; i < n; i++) {
out += i
}
return out
}
const pInstance = new Parallel(hardwork) To binding data need to execute parallel with function when construct.
calculateData return self instance to keep handle other logic.
Example
let data = [1_000_000, 1_500_000, 2_000_000, 2_500_000, 3_000_000]
pInstance.calculateData(data)When function contain context this like that:
const A = {
a: 1,
hardwork: function(n: number): number {
let out = 0
for(let i = 0; i < n; i++) {
out += i
}
return out + this.a
}
}When routing processing to Workers. The context of this is not guaranteed to be kept correct. So we need specific context this when handle parallel in Workers.
Example
const pInstance = new Parallel(A.hardwork)
pInstance.bindingParallel(A)After binding data and bindingInstance success we can execute parallel and receive Promise response to get value after execute done or error.
Example
const A = {
a: 1,
hardwork: function(n: number): number {
let out = 0
for(let i = 0; i < n; i++) {
out += i
}
return out + this.a
}
}
let data = [1_000_000, 1_500_000, 2_000_000, 2_500_000, 3_000_000]
const startTime = performance.now()
const pInstance = new Parallel(A.hardwork)
pInstance.calculateData(data)
.bindingParallel(A)
.execute<number[]>()
.then(result => {
console.log(result)
console.log(performance.now() - startTime)
})Any issues or information can be left at the repo or contact me.