Rxjs-optimizer NPM

RxJS Optimizers

A collection of operators and methods to help optimize RxJS based development.

Get Started

npm install --save rxjs-optimizers
yarn add rxjs-optimizers

List of optimizers

Filter Unchanged

Filter unchanged values to avoid re-processing for the same input

Usage:

import { filterUnchangable } from 'rxjs-optimizer';

observable.pipe(filterUnchanged());
observable.pipe(filterUnchanged((a,b) => a !== b)); // custom change check

Parameters and Options:

import { filterUnchangable } from 'rxjs-optimizer';

filterUnchangable(
    /**
     * [optional] Method to determine if the value has changed
     * @type: (previous: T, current: T) => boolean
     * @default: (a, b) => (a !== b)
     */
    hasChanged
)

Cache Pipe

Given a list of operators (eg. map(), filter(), switchMap(), ...), it will cache the final output value given the the initial inputs, so that if the input goes back to a previous input, it will bypass the operators, and send the cached value before running the operators.

If no key function option is passed, then it will convert these types to string to use as a cache key: (string, number, boolean, symbol). If it's of a different type, then no caching will be done for that input. This behavior can be customized by passing the key option.

Usage:

import { cachePipe } from 'rxjs-optimizer';

observable.pipe(cachePipe(...operators));

// custom options
let smallCache = cachePipe.options({cacheSize: 2});
observable.pipe(smallCache(...operators));

Parameters and Options:

import { cachePipe } from 'rxjs-optimizer';

cachePipe.options({
    /** cacheSize
     * number of input/output pair to keep in the cache
     * @type: number
     * @default: 10
     */
    cacheSize: 10,

    /** key
     * function to get the key given the object received from the source observable
     * @type: (a:any) => string
     * @default: (see description above)
     */
    key: (a) => a.id
})

Single Subscriber

Normally, the observable will call its internal subscribe function with every new subscription. This operator will just subscribe once to the source observable, optionally cache the last value, and send the value to all the subscribers of the resulting observable.

By default, if the subscriber count arrives at zero, then the wrapper will unsubscribe from the wrapped observable, and will reset the cache, to avoid extra use of memory or processing. This behavior can be customized using the cache option

Usage:

import { singleSubscriber } from 'rxjs-optimizer';

observable.pipe(singleSubscriber());

observable.pipe(singleSubscriber({cache: 'none'}));
observable.pipe(singleSubscriber({cache: 'subscribed'})); // default
observable.pipe(singleSubscriber({cache: 'keep'}));

Parameters and Options:

import { singleSubscriber } from 'rxjs-optimizer';

singleSubscriber({
    /**
     * [optional] Enum to determine how the value should be cached:
     *      - 'none'       : Subscribers only receive values emitted while subscribed
     *      - 'subscribed' : If the observable already had a subscriber, then new subscribers
     *                       will receive the last emitted value when they begin subscribing.
     *                       but once all subscribers unsubscribe, the cache is cleared.
     *      - 'keep'       : Caches the last value emitted, and emits it for new subscribers,
     *                       the value is kept in the cache even if all subscribers unsubscribe.
     * @type: 'none' | 'subscribed' | 'keep'
     * @default: 'subscribed'
     */
    cache: 'subscribed'
})

Other optimizer ideas

Preload

Similar to cache, but given the preloading inputs, once the first output goes through, starts loading the other possible inputs, and caches the outputs