DEV Community

Cover image for Lazy Loading Image - [2/2]
Francesco Di Donato
Francesco Di Donato

Posted on

Lazy Loading Image - [2/2]

Recap

In the previous post, I built a React component Image that receives two sources, one for a low-resolution version of the image and one for the high-resolution one. Shows the former, which is promptly replaced by the latter as soon as its download is complete.

Repo 📑
Previous post 😺

Abstract

A further performance improvement is to start the download of the high-resolution image only when the component is in view.
Still, with a view to modern React, I build a custom hook which, having received a ref associated with an HTML element, uses the IntersectionObserver API to evaluate if the element is in view

Process

I add the hook in the appropriate folder built previously

touch src/hooks/useIntersectionObserver.js
Enter fullscreen mode Exit fullscreen mode

The IntersectionObserver must be instantiated in a useEffect whose execution depends on the elementRef that the hook receives as an argument. This is necessary for the functionality of the hook to be responsive if a different ref is conditionally provided during use

One way to proceed is to bind the IntersectionObserver to a ref declared in the hook itself. In this way, at the unmount of the component using the hook, React will take care of the clean up of the aforementioned ref

In the IntersectionObserver callback it is sufficient to set the entry that is observed. This makes it easy to find outside the useEffect

useIntersectionObserver.js
import { useRef, useEffect, useState } from 'react'

const useIntersectionObserver = elementRef => {
  const observer = useRef()
  const [entry, setEntry] = useState()

  const options = {
    threshold: 0.1,
    root: null,
    rootMargin: '0%',
  }

  const updateEntry = entries => {
    setEntry(entries[0])
  }

  useEffect(() => {
    const node = elementRef?.current
    if (!node) return

    if (observer.current) observer.current.disconnect()

    observer.current = new IntersectionObserver(updateEntry, options)

    const { current: currentObserver } = observer

    currentObserver.observe(node)

    return () => currentObserver.disconnect()
  }, [elementRef])

  return { isVisible: !!entry?.isIntersecting, entry }
}

export default useIntersectionObserver
Enter fullscreen mode Exit fullscreen mode

A boolean is returned indicating the presence or absence of the component in the view

There are two observer.current.disconnect(). The first is executed only if the observer was already active but under observation on a different elementRef. In the second case, the disconnection occurs in the cleanup phase of the useEffect which, by extension, uniquely corresponds to the moment in which the component that makes use of the hook is removed from the DOM

For the purposes of this demo the hook always refers to the whole view. However, it is not difficult to take a second options argument and pass it into the IntersectionObserver instance (remember to add it to the useEffect dependencies)


The use in the <Image> component (the same as in the previous post) is immediate. I declare a ref (imageRef) and bind it to the root element of the component (div.wrapper). The same ref is supplied to the useIntersectionObserver hook which returns isVisible

Conditionally showing the second <img> tag, that is the one associated with the high-resolution image, you will get that the feature implemented in the previous post is used only when the element enters view. In the meantime, the user is shown the low-resolution image

Image.js (* to indicate the changes from the previous one)
import { useRef } from 'react'
import useImageOnLoad from '../hooks/useImageOnLoad'
import useIntersectionObserver from '../hooks/useIntersectionObserver'

const Image = ({ width = '100%', height = '100%', lowResSrc, highResSrc }) => {
  const { handleImageOnLoad, transitionStyles } = useImageOnLoad()

  const imageRef = useRef() // *
  const { isVisible } = useIntersectionObserver(imageRef) // *

  const styles = {
    wrapper: {
      position: 'relative',
      width,
      height,
    },
    image: {
      position: 'absolute',
      width: '100%',
      height: '100%',
      objectPosition: 'center center',
      objectFit: 'cover',
    },
  }

  const lowResStyle = {
    ...styles.image,
    ...transitionStyles.lowRes,
  }
  const hightResStyle = {
    ...styles.image,
    ...transitionStyles.highRes,
  }

  return (
    <div style={styles.wrapper} ref={imageRef}>
      <img src={lowResSrc} style={lowResStyle} />
      {isVisible && ( // *
        <img
          src={highResSrc}
          style={hightResStyle}
          onLoad={handleImageOnLoad}
        />
      )}
    </div>
  )
}

export default Image
Enter fullscreen mode Exit fullscreen mode

The simplest way to check if the desired effect is present is to move the image outside the screen

App.js (detail)
<div style={{ position: 'relative', height: '200vh' }}>
  <div style={{ position: 'absolute', bottom: 0 }}>
    <ImageIO
      width={600}
      height={400}
      lowResSrc={srcTuple[0]}
      highResSrc={srcTuple[1]}
    />
  </div>
</div>
Enter fullscreen mode Exit fullscreen mode

From the Network tab of the Developer Tools, you can see how the low-resolution image download is performed as soon as possible. On the other hand, that of the high-resolution image is started only when the component is in view

intersection observer demo


Thanks for reading, continue to the next post (work in progress)

Repo 📑

If you like it, let's get in touch
🐙, 🐦 and 💼

Top comments (0)