@jlmurphysa/rate-limiter v0.0.6

rate-limiter

A powerful rate-limiter with annotation and channel support written for TypeScript.

Features

• Annotation/decorator support with @throttle
• Implements the token-bucket algorithm for flexible rate options.
• Support for independent configurable Channels that are isolated from each other.

Installation

$ npm install @jlmurphysa/rate-limiter

Usage

Simple default channel

import { throttle, Channels } from '@jlmurphysa/rate-limiter'

class API {
    @throttle()
    static async request() {
        return 'OK';
    }
}
Channels.create(10, 'second')

Simple named channels

import { Channels, throttle } from '@jlmurphysa/rate-limiter'

class API {
    @throttle('google-places')
    static async requestToPlaces() {
        return 'OK';
    }

    @throttle('google-translate')
    static async requestToAnother() {

    }
}
Channels.create('google-places', 10, 'second')
Channels.create('google-translate', 20, 'hour')

Advanced

import {
    Channel,
    Channels,
    throttle
} from '@jlmurphysa/rate-limiter'

Channels.add('google-places', new Channel({
    interval: 1000,         // Token refresh interval
    bucketSize: 25,         // Max tokens the bucket will hold
    tokensPerInterval: 10,  // Number of tokens we receive per interval
    tokens: 25              // Initial tokens
})

const throttleOpts = {
    channel: 'google-places',
    cost: 5,    // The token cost of each call
    ttl: 5000   // Throw if we'll have to wait more than 5s.
}

class API {
    @throttle(throttleOpts)
    static async requestToPlaces() {
        return 'OK'
    }

    @throttle({...throttleOpts, cost: 8})
    static async requestToPlacesRich() {
        return 'OK!'
    }
}

Future Work

• 'priority' option - not exactly trivial in the current implementation.