npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

pixl-cache

v1.0.6

Published

A simple LRU cache module.

Downloads

3,186

Readme

Overview

pixl-cache is a very simple LRU (Least Recently Used) cache module for Node.js. It works like a hash map with set() and get() methods, but when it detects an overflow (configurable by total keys or total bytes) it automatically expunges the least recently accessed objects from the cache. It is fast, stable, and has no dependencies.

Internally the cache is implemented as a combination of a hash and a double-linked list. When items are accessed (added, replaced or fetched) they are promoted to the front of the linked list. When the max size (keys or bytes) is exceeded, items are dropped from the back of the list. Items can also have an optional max age (i.e. expiration date).

Features

  • Simple and straightforward API
  • Fast and stable
  • Low memory overhead
  • Predictable results on overflow
  • Can expire based on key count or byte count
  • Optional expiration date for items
  • Event listener for ejecting expired items
  • Can store custom metadata along with cache objects
  • No dependencies

Benchmarks

Note that while pixl-cache may be the fastest and use the least memory among the packages tested, it also has the least amount of features. The other packages listed here are all awesome, and you should consider using them. Most have been battle-tested in production for years, whereas pixl-cache has not.

Read and write 1,000,000 keys (shorter times are better):

| Package | Version | Write Time | Read Time | Max Memory Usage | |---------|---------|------------|-----------|------------------| | stale-lru-cache | v5.1.1 | 5531 ms | 216 ms | 147.9 MB | | fast-lru | v3.0.1 | 5816 ms | 149 ms | 142.1 MB | | lru-cache | v5.1.1 | 5263 ms | 62 ms | 140.7 MB | | node-cache | v3.1.0 | 6097 ms | 713 ms | 221.5 MB | | pixl-cache | v1.0.4 | 4253 ms | 37 ms | 128.1 MB |

Tests performed on a MacBook Pro 2016 (2.9 GHz Intel Core i7) with Node v10.14.1.

Usage

Use npm to install the module:

npm install pixl-cache

Here is a simple usage example, enforcing 5 max items:

const LRU = require('pixl-cache');
var cache = new LRU({ maxItems: 5 });

cache.set( 'key1', "Simple String" );
cache.set( 'key2', Buffer.alloc(10) );
cache.set( 'key3', { complex: { item: 1234 } } );
cache.set( 'key4', true );
cache.set( 'key5', "Cache is full now" );
cache.set( 'key6', "Oops, key1 is gone now");

var value = cache.get( 'key1' );
// value === undefined (key1 got expunged)

cache.delete( 'key2' ); // manual delete

if (cache.has('key3')) {
	// check if key exists (true)
}

cache.clear(); // wipe entire cache

Instead of setting a maximum item count, you can set a total byte count:

var cache = new LRU({ maxBytes: 1024 * 1024 }); // 1 MB
cache.set( 'key1', Buffer.alloc(1024 * 512) ); // 50% full here
cache.set( 'key2', Buffer.alloc(1024 * 256) ); // 75% full here

Or expire objects by their age in the cache:

var cache = new LRU({ maxAge: 86400 }); // 24 hours
cache.set( 'key1', "Expires tomorrow!" );

Overflow

pixl-cache can handle overflow in two different ways: by enforcing a maximum number of items, and/or a maximum number of bytes. When either of these limits are exceeded, the least recently used object(s) will be expunged. These limits can be passed to the class constructor like so:

var cache = new LRU({ 
	maxItems: 1000, 
	maxBytes: 1048576
});

This would allow up to 1,000 items or 1 MB of total value length, whichever is reached first. When using maxBytes the cache needs to calculate the size of your objects. The byte size includes both the key and the value. The process is automatic if you use simple primitive value types such as strings, buffers, numbers or booleans. Examples:

cache.set( 'key1', "ABCDEFGHIJ" ); // 20 bytes + 8 for key
cache.set( 'key2', Buffer.alloc(10) ); // 10 bytes + 8 for key
cache.set( 'key3', 12345 ); // 8 bytes + 8 for key
cache.set( 'key4', true ) ; // 4 bytes + 8 for key

Note that string length does not equal byte length. This is because strings are internally represented as 16-bits (2 bytes) per character in Node.js, so they're basically double size. That is why the 10-character string ABCDEFGHIJ is actually 20 bytes in memory.

The byte length of object values is not automatically calculated. Meaning, if you pass an object as a value, its byte count is read from a length property if it exists (i.e. for buffers), or if not found it defaults to 0. So, you may want to specify your own custom lengths in certain cases, especially if you are storing objects in the cache. To do this, pass a metadata object to set() as the 3rd argument, and include an explicit length property therein:

cache.set( 'key1', { "name": "Joe" }, { length: 200 } );

This would record the length of the object value as 200 bytes.

Expiration

In addition to expiring the least recently used objects when the cache is full, objects can also expire based on age. You can specify a maxAge configuration property when constructing your cache:

var cache = new LRU({ maxAge: 86400 }); // 24 hours

And in this case all items added to the cache will be expired after 24 hours. Replacing existing items resets their age clock. Each item has its own internal expiration date, and if that date comes to pass, fetching the item will immediately cause it to be deleted, and return undefined. Note that items are not "actively deleted" based on an interval timer, but rather expired items delete themselves on fetch (or may be expunged for other reasons, i.e. maxItems and/or maxBytes).

You can specify a custom expiration date for your items individually, by passing a metadata object to set() as the 3rd argument, and including an explicit expires property therein:

var someFutureDate = ( Date.now() / 1000 ) + 86400; // 24 hours from now
cache.set( 'key1', "ABCDEFGHIJ", { expires: someFutureDate } );

The expires property needs to be an Epoch timestamp, as shown above. You do not need to enable maxAge in order to use the custom expires property. It works with any combination of configuration options, and overrides maxAge if it is also set. If you set an expires date in the past, the key is immediately deleted upon the next fetch.

API

constructor

const LRU = require('pixl-cache');
var cache = new LRU();

The constructor creates a cache instance. You can optionally pass an object containing any of these properties:

| Property | Default | Description | |----------|---------|-------------| | maxItems | 0 | Specifies the maximum number of objects allowed in the cache, before it starts expunging the least recently used. | | maxBytes | 0 | Specifies the maximum value bytes allowed in the cache, before it starts expunging the least recently used. | | maxAge | 0 | Specifies the default maximum age in seconds for all keys added to the cache, before they are deleted on fetch. |

The default of 0 means infinite. If multiple configuration properties are specified, whichever one kicks in first will expire keys. Here is an example with all properties set:

var cache = new LRU({
	maxItems: 1000,
	maxBytes: 1048576, // 1 MB
	maxAge: 86400 // 24 hours
});

set

cache.set( 'key1', "Value 1" );

The set() method adds or replaces a single object in the cache, given a key and value. The key is assumed to be a string, but the value can be anything. Anytime an object is added or replaced, it becomes the most recently used.

Adding new keys may cause the least recently used object(s) to be expired from the cache.

You can optionally pass an object containing arbitrary metadata as the 3rd argument to set(). This is stored in the internal cache database, and does not add to the total byte count. Example:

cache.set( 'key1', "Value 1", { mytag: "frog" } );

Please make sure your metadata object does not include the following four keys: key, value, next or prev. Those are for internal cache use. You can, however, include a length key in the metadata object, which overrides the default length calculation for the object, and/or an expires key, which sets the expiration date for the key (Epoch timestamp).

Subsequent calls to set() replacing a key with differing metadata is shallow-merged. If a subsequent call omits metadata entirely, the original data is preserved.

You can use getMeta() to retrieve cache objects including your metadata.

Note that if you want to store either null or undefined as cache values, you must specify a non-zero length in the metadata object. Otherwise pixl-cache will throw an exception trying to compute the length.

get

var value = cache.get( 'key1' );

The get() method fetches a value given a key. If the key is not found in the cache (or it exists but is expired) the return value will be undefined. Note that when fetching any key, that object becomes the most recently used.

If the item is fetched on or after its expiration date (i.e. when using maxAge or setting an explicit date), it will be deleted, and the call will return undefined.

getMeta

var item = cache.getMeta( 'key1' );
var mytag = item.mytag; // custom metadata

The getMeta() method fetches the internal cache object wrapper given its key, which includes any metadata you may have specified when you called set(). The object will always have the following properties, along with your metadata merged in:

| Key | Description | |-----|-------------| | key | The raw cache object key, as passed to set(). | | value | The raw cache object value, as passed to set() and returned by get(). | | next | A pointer to the next cache object in the linked list. Please do not touch. | | prev | A pointer to the previous cache object in the linked list. Please do not touch. |

The metadata object may also have one or both of these additional properties:

| Key | Description | |-----|-------------| | length | The length of the object's value, if it was provided explicitly when set() was called, or calculated automatically. | | expires | The expiration date of the object as an Epoch timestamp, if the maxAge configuration property is set, or a custom expires was provided when set() was called. |

has

if (cache.has('key1')) console.log("key1 is present!");

The has() method checks if a specified key exists in the cache, and returns true or false. This does not cause the object to get promoted to the most recently used. It is more of a "peek".

Note that if a cache object is technically still present but expired, the return value of this call will be false. This is because the item is "effectively" deleted at this point (i.e. it can no longer be fetched), so we properly represent what would happen if you tried to call get() after calling has().

delete

cache.delete( 'key1' );

The delete() method deletes the specified key from the cache. If successful, this returns true. If the key was not found it returns false. This will return true (success) for deleting expired keys.

clear

cache.clear();

The clear() method clears the entire cache, deleting all objects. It does not reset any configuration properties, however.

getStats

var stats = cache.getStats();

The getStats() method returns an object containing some basic statistics about the cache, including the current total keys, total bytes, fullness percentage estimate, and the 10 hottest (most used) keys. Example response, formatted as JSON:

{
	"count": 158,
	"bytes": 15098,
	"full": "15.8%",
	"hotKeys": [
		"index/myapp/data/2665",
		"index/myapp/data/2664",
		"index/myapp/data/2663",
		"index/myapp/data/2662",
		"index/myapp/data/2661",
		"index/myapp/data/2660",
		"index/myapp/data/2659",
		"index/myapp/data/2658",
		"index/myapp/data/2657",
		"index/myapp/data/2656"
	]
}

on

cache.on( 'expire', function(item, reason) {
	console.log(`Cache expired ${item.key} because of ${reason}.`);
});

The pixl-cache class inherits from Node's EventEmitter, so it has all those methods available including on() and off().

Events

The following events are emitted:

expire

The expire event is emitted when the cache is about to expire the least recently used object, either due to total key count, byte count or age. The event listener will be passed two arguments, the item being expired, and a reason:

cache.on( 'expire', function(item, reason) {
	console.log(`Cache expired ${item.key} because of ${reason}.`);
});

The item is an object containing the original key and value as originally passed to set(), along with any custom metadata if applicable. It is the same object that getMeta() returns. You can use the expire hook to persist data to disk, for example.

The reason will be either count (expired due to total key count), bytes (expired due to total byte count), or age (expired due to age). It is always a string.

Properties

The following properties are available on pixl-cache class instances:

items

The items property is a hash containing all the objects currently in the cache, keyed by their actual keys. Please do not directly manipulate this object, as it will become out of sync with the internal linked list. However, you are free to read from it.

count

The count property contains the current total key count in the cache.

bytes

The bytes property contains the current total byte count in the cache.

Development

To install pixl-cache for development, run these commands:

git clone https://github.com/jhuckaby/pixl-cache.git
cd pixl-cache
npm install

The npm install is only needed for running unit tests.

Unit Tests

When installing locally for development, pixl-unit will be installed as a dev dependency. Then, to run the unit tests, issue this command:

npm test

License

The MIT License (MIT)

Copyright (c) 2019 Joseph Huckaby.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.