Huge News!Announcing our $40M Series B led by Abstract Ventures.Learn More
Socket
Sign inDemoInstall
Socket

fs-store

Package Overview
Dependencies
Maintainers
1
Versions
6
Alerts
File Explorer

Advanced tools

Socket logo

Install Socket

Detect and block malicious and high-risk dependencies

Install

fs-store

File system backed store for node with a trivial interface, multiple backups, robust failure recovery, single async load, in-memory cache (effectively synchronous API), background flushing.

  • 0.3.3
  • latest
  • Source
  • npm
  • Socket score

Version published
Maintainers
1
Created
Source

node-fs-store

Trivially simple interface for a robust file-system backed store for node. The goal of this module is to provide a simple to use module that will persist data locally in a robust manner (recovers if your process exits while writing, etc), for simple applications that do not wish to spend the time to set up a full-fledged database or something similar.

Features:

  • multiple backups
  • robust failure recovery
  • single sync load
  • in-memory cache (effectively synchronous API)
  • rate-limited background (async) flushing

Getting Started

Install the module with: npm install fs-store

Non-NPM install: Save index.js as fs-store.js and instead use require('./fs-store.js')

var FileStore = require('fs-store').FileStore;

// Create a store
var my_store = new FileStore('example_data.json');

// Get a value, providing a default
var number_of_runs = my_store.get('number_of_runs', 0);

// Store a value (will be written to disk asynchronously)
my_store.set('number_of_runs', number_of_runs + 1);

// Get the value back (immediately reflected in the store,
// even if it is not yet on disk)
console.log('This example has run ' + my_store.get('number_of_runs') + ' time(s)');

Advanced Configuration

var FileStore = require('fs-store').FileStore;

var my_store = new FileStore({
  // File name to use to back the store, required parameter
  filename: 'example_data.json',

  // Providing a default object to be used if the file does
  // not exist (or all backups are unrecoverable)
  // Default value = {}
  default_object: { foo: 'bar' },

  // Minimum milliseconds between two consecutive saves
  // Default value = 1000 (1 second)
  min_save_interval: 2500,

  // Maximum number of additional backups to keep.  Must be >= 1
  // Default value = 3
  max_backups: 3,

  // Set read_only to true to disable writes to disk (in-memory
  // changes to the store will still work as usual).  May be useful
  // when (cluster.isWorker && cluster.worker.id !== 1) and all cluster
  // members are doing the exact same logic on the store.
  // Default value = false;
  read_only: true,
});

API Reference

new FileStore(options)

Creates a FileStore and synchronously loads the store from disk or a backup if the primary file is unable to be loaded for any reason. See "Advanced Configuration" above for details on the options parameter.

store.set(key, value)

Sets a value in the store, queues up an asynchronous save to disk.

store.get(key, [default_value])

Returns a value from the store, or the provided default_value if no value has previously been set with that key.

store.getStore()

This should almost never be needed. Returns that storage object for the store for direct modifications or iteration. If any changes are made, you must manually call store.save when you are done. In theory this could be used to very efficiently make a bunch of changes, but, if you need that much efficiency, you probably should not be using a stupid-simple on-disk store like this.

store.save()

Manually trigger a save of the store (after making modifications through store.getStore()).

store.onFlush(cb)

Calls cb after any changes have been flushed to disk. Callback will always be called asynchronously even if nothing is waiting to be flushed to disk. This has very little general use, but is invaluable in writing deterministic unit tests.

Notes

All saves happen asynchronously, so if your program is entirely synchronous, no data will be saved until the main part of the program exits (at which point it will execute a single write).

The asynchronous saves, by default, happen at most once per second, and batch all modifications smartly so if you make three .set() calls in a block of code, all of the modifications will be saved at once (e.g. it does not start saving after the first set() call, waiting before saving the second and third, but instead starts saving once your code execution has yielded and saves all three values).

Keywords

FAQs

Package last updated on 25 Apr 2024

Did you know?

Socket

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Install

Related posts

SocketSocket SOC 2 Logo

Product

  • Package Alerts
  • Integrations
  • Docs
  • Pricing
  • FAQ
  • Roadmap
  • Changelog

Packages

npm

Stay in touch

Get open source security insights delivered straight into your inbox.


  • Terms
  • Privacy
  • Security

Made with ⚡️ by Socket Inc