unlazy

Lazy loading library for placeholder images

unlazy library

unlazy

npm version

Universal lazy loading library leveraging native browser APIs. It's intended to be used with the loading="lazy" attribute alongside (blurry) placeholder images and with a BlurHash or ThumbHash string.

It's the modernized and now framework-agnostic rewrite of my Loadeer.js library, which uses an Intersection Observer to detect when an image enters the viewport.

Features

  • 🎀 Native: Utilizes the loading="lazy" attribute
  • 🎛️ Framework-agnostic: Works with any framework or no framework at all
  • 🌊 BlurHash & ThumbHash support: SSR & client-side BlurHash and ThumbHash decoding
  • 🪄 Sizing: Automatically calculates the sizes attribute
  • 🔍 SEO-friendly: Detects search engine bots and preloads all images
  • 🎟 <picture>: Supports multiple image tags
  • 🏎 Auto-initialize: Usable without a build step

Note

Although the loading="lazy" attribute is supported in all major browsers, it is only available in Safari 16.4 (released March 2023) and later versions by default. It is important to consider this limitation when using unlazy for your project, as it might impact the user experience for visitors using older Safari versions or other unsupported browsers.

Setup

📖 Read the documentation

# pnpm
pnpm add unlazy

# npm
npm i unlazy

Basic Usage

📖 Read the documentation

To apply lazy loading to all images with the loading="lazy" attribute, import the lazyLoad function and call it without any arguments:

import { lazyLoad } from 'unlazy'

// Apply lazy loading for all images by the selector `img[loading="lazy"]`
lazyLoad()

You can target specific images by passing a CSS selector, a DOM element, a list of DOM elements, or an array of DOM elements to lazy-load to lazyLoad.

💻 Development

  1. Clone this repository
  2. Enable Corepack using corepack enable
  3. Install dependencies using pnpm install
  4. Run pnpm run dev:prepare
  5. Start development server using pnpm run dev inside the one of the packages directories

License

MIT License © 2023-present Johann Schopplich