josebaseba / hyperlru

Tiny & Fast LRU Implementation as possible.

hyperlru

Tiny & Fast LRU Implementation as possible.

Features

• Fast: Ultra-high performance (See benchmark).
• Simple: The whole project is ~60 lines of code.
• Lightweight: The package weighs less than a megabyte, with zero dependencies.

hyperlru claim to be fast and simple as possible. There are a lot of LRU implementations, but most of them have a poor perfomance and they are hard to understand.

It's Inspired by the hashlru algorithm.

Install

$ npm install hyperlru --save

Usage

const hyperlru = require('hyperlru')
const cache = hyperlru({max: 1000})

Benchmark

name	size	gzip	set	get1	update	get2	evict
lru-cache	19.1 kB	6.23 kB	922 ops	3030 ops	2105 ops	4082 ops	917 ops
modern-lru	2.27 kB	907 B	813 ops	1299 ops	752 ops	1515 ops	629 ops
mkc	10.5 kB	3.61 kB	784 ops	766 ops	810 ops	1149 ops	697 ops
lru_cache	2.19 kB	756 B	6452 ops	18182 ops	14286 ops	9524 ops	3509 ops
hashlru	628 B	332 B	5263 ops	6897 ops	4545 ops	7407 ops	3922 ops
tiny-lru	4 kB	1.64 kB	3509 ops	11765 ops	15385 ops	11765 ops	2222 ops
quick-lru	1.23 kB	489 B	3279 ops	2198 ops	3226 ops	2105 ops	2778 ops
lru	6.07 kB	1.86 kB	2500 ops	3846 ops	3175 ops	3226 ops	976 ops
hyperlru	889 B	395 B	2299 ops	25000 ops	2410 ops	22222 ops	2439 ops
lru-fast	2.34 kB	793 B	1869 ops	6897 ops	4167 ops	16667 ops	3077 ops
simple-lru-cache	1.43 kB	565 B	1709 ops	13333 ops	5128 ops	22222 ops	3922 ops
secondary-cache	22.6 kB	6.54 kB	1695 ops	2632 ops	2667 ops	5556 ops	1408 ops

See more at bench-lru.

API

hyperlru([options])

options

max

Type: number
Default: 1000

Max of element to keep into the cache.

.set(key, value)

Set the value of the key and mark the key as most recently used.

It returns the value.

.get(key)

Query the value of the key and mark the key as most recently used.

It returns the value of key if found; undefined otherwise.

.keys()

It retrieves all the keys currently in the cache.

.clear()

Clear all the elements in the cache.

License

MIT © Kiko Beats.