Microtask Collector

Creates a collection of values from the current microtask

Example

import { Collector } from "microtask-collector"

const collector = new Collector({
    map: values => Object.freeze(values)
})

async function watch() {
    for await (const values of collector) {
        console.log({ values })
    }
}

async function producer() {
    collector.add(1)
    collector.add(2)
    await wait()
    collector.add(3)
    await wait()
    collector.add(4)
    collector.add(5)
    collector.close()
}

async function wait(ms = 0) {
    return new Promise(resolve => setTimeout(resolve, ms))
}

await Promise.all([watch(), producer()])

The above example will log:

{ values: [1, 2] }
{ values: [3] }
{ values: [4, 5] }

options.queueMicrotask

By default queueMicrotask will be used to determine microtasks, however the queueMicrotask option can be provided as a function which accepts a single callback.

For example:

import { Collector } from "microtask-collector"

const collector = new Collector({
    map: values => Object.freeze(values),
    queueMicrotask: callback => setTimeout(callback, 0)
})

If queueMicrotask is not available from the global scope then setImmediate will be used, or setTimeout with 0 as the timeout argument.

options.map

The map option is required so that any transformations to the value can happen once for all code observing, an array of the collected values will be provided as an argument, this array won't be changed by the collector once map has been invoked.

Collector.prototype.add(value)

Invoked with a value associated to the current microtask

Collector.prototype.close()

This should be invoked when the collector is no longer going to produce new values. This allows associated async iterators to finalise.