Mnemonist

The LRUMap, LRU standing for least recently used, can be seen as a a fixed-capacity key-value store that will evict infrequent items when full and setting new keys.

For instance, if one creates a LRUMap with a capacity of 1000 and one inserts a thousand-and-first key, the cache will forget its least recently used key-value pair in order not to overflow the allocated memory.

This structure is very useful to cache the result of costly operations when one cannot afford to keep every result in memory and only want to keep the most frequent ones.

For more information, you can check this Wikipedia page.

For a detailed explanation of how this cache is implemented, you can also read this blog post.

This implementation has been designed to work with an ES6 Map object. You can alternatively find an implementation relying on a javascript raw object here. Depending on the precise use case (string keys, integer keys etc.), one or the other might be faster depending on js engine magic.

const LRUMap = require('mnemonist/lru-map');
// If you need deletions
const LRUMapWithDelete = require('mnemonist/lru-map-with-delete');

Constructor

The LRUMap takes a single argument: the desired capacity.

const cache = new LRUMap(1000);

Optionally, you can type the used keys & values in order to be more memory-efficient:

// First argument will be instantiated for keys, second one for values
const cache = new LRUMap(Uint32Array, Float32Array, 1000);

Static #.from

Alternatively, one can build a LRUMap from an arbitrary JavaScript iterable likewise:

// Attempting the guess the given iterable's length/size
const cache = LRUMap.from({one: 1, two: 2});

// Providing the desired capacity
const cache = LRUMap.from({one: 1, two: 2}, 10);

// Typing the cache
const cache = LRUMap.from({one: 1, two: 2}, Array, Uint8Array, 10);

Members

• #.capacity
• #.size

Methods

Mutation

• #.set
• #.setpop
• #.delete (only on LRUMapWithDelete)
• #.remove (only on LRUMapWithDelete)
• #.clear

Read

• #.get
• #.peek
• #.has

Iteration

• #.forEach
• #.keys
• #.values
• #.entries
• Iterable

#.capacity

Maximum number of items the cache is able to store.

const cache = new LRUMap(10);
cache.capacity
>>> 10

#.size

Number of items actually in the cache.

const cache = new LRUMap(10);
cache.size
>>> 0
cache.set('one', 1);
cache.size
>>> 1

#.set

Sets a value for the given key in the cache. If the cache is already full, the least recently used key will be dropped from the cache.

O(1)

const cache = new LRUMap(10);
cache.set('one', 1);
cache.has('one');
>>> true

#.setpop

Sets a value for the given key in the cache. If the cache is already full, the least recently used key will be dropped from the cache, and an object containing the dropped key, dropped value and evicted = true will be returned. If the key already exists, an object containing the key, previous value and evicted = false will be returned. If no eviction or overwrite occurs, null is returned.

O(1)

const cache = new LRUMap(1);
cache.setpop('one', 1);
>>> null
cache.setpop('one', 10);
>>> {key: 'one', value: 1, evicted: false}
cache.setpop('two', 2);
>>> {key: 'one', value: 10, evicted: true}

#.delete

Deletes the given key and returns whether its was actually in the cache before deletion.

Beware, for performance reasons this method is only available on LRUMapWithDelete.

O(1)

const cache = new LRUMapWithDelete(1);
cache.set('one', 1)

cache.delete('one');
>>> true

cache.delete('one');
>>> false

#.remove

Deletes the given key and returns the associated value if the key was actually in the cache or a default value if it was not.

Beware, for performance reasons this method is only available on LRUMapWithDelete.

O(1)

const cache = new LRUMapWithDelete(1);
cache.set('one', 1)

cache.remove('one');
>>> 1

cache.remove('one');
>>> undefined

cache.remove('one', 'not-found');
>>> 'not-found'

#.clear

Completely clears the cache.

O(1)

const cache = new LRUMap(10);
cache.set('one', 1);
cache.clear();

cache.has('one');
>>> false
cache.size
>>> 0

#.get

Retrieves the value associated to the given key in the cache or undefined if the key is not found.

If the key is found, the key is moved to the front of the underlying list to be the most recently used item.

O(1)

const cache = new LRUMap(10);
cache.set('one', 1);
cache.get('one');
>>> 1

#.peek

Retrieves the value associated to the given key in the cache or undefined if the key is not found.

Unlike #.get, it does not modify the underlying list.

O(1)

const cache = new LRUMap(10);
cache.set('one', 1);
cache.peek('one');
>>> 1

#.has

Retrieves whether the given key exists in the cache.

O(1)

const cache = new LRUMap(10);
cache.set('one', 1);
cache.has('one');
>>> true

cache.has('two');
>>> false

#.forEach

Iterates over the cache from the most to the least recently used key-value pair.

const cache = new LRUMap(10);

cache.set('one', 1);
cache.set('two', 2);

cache.forEach((value, key, cache) => {
  console.log(key, value);
});

#.keys

Returns an iterator over the cache’s keys from the most to the least recently used key.

const cache = new LRUMap(10);

cache.set('one', 1);
cache.set('two', 2);

const iterator = cache.keys();

iterator.next().value
>>> 'two'

#.values

Returns an iterator over the cache’s values from the most to the least recently used value.

const cache = new LRUMap(10);

cache.set('one', 1);
cache.set('two', 2);

const iterator = cache.values();

iterator.next().value
>>> 2

#.entries

Returns an iterator over the cache’s entries from the most to the least recently used entry.

const cache = new LRUMap(10);

cache.set('one', 1);
cache.set('two', 2);

const iterator = cache.entries();

iterator.next().value
>>> ['two', 2]

Iterable

Alternatively, you can iterate over a cache’s entries using ES2015 for...of protocol:

const cache = new LRUMap(10);

cache.set('one', 1);
cache.set('two', 2);

for (const [key, value] of cache) {
  console.log(key, value);
}