We're looking for a user researcher to understand the needs of developers and designers. Is this you or someone you know? Check out the post: https://mzl.la/2IGzdXS

この記事はまだボランティアによって 日本語 に翻訳されていません。ぜひ MDN に参加して翻訳を手伝ってください!
この記事を English (US) で読むこともできます。

This feature is non-standard and is not on a standards track. Do not use it on production sites facing the Web: it will not work for every user. There may also be large incompatibilities between implementations and the behavior may change in the future.


The -moz-image-rect value for CSS background-image lets you use a portion of a larger image as a background. This allows you to, for example, use different parts of one larger image as backgrounds in different parts of your content.

This works very similarly to the -moz-image-region property, which is used with the list-style-image property to use parts of an image as the bullets in lists. However, this can be used for any CSS background.

The syntax for the rectangle is similar to the rect() function generating a <<shape>()> CSS type. All four values are relative to the upper left corner of the image.


-moz-image-rect(<uri>, top, right, bottom, left);


The URI of the image from which to take the sub-image.
The top edge, specified as a <length>, of the sub-image within the specified image.
The right edge, specified as a <length>, of the sub-image within the specified image.
The bottom edge, specified as a <length>, of the sub-image within the specified image.
The left edge, specified as a <length>, of the sub-image within the specified image.


This example loads an image and uses it in four segments to draw the Firefox logo in four <div> blocks. Clicking on their container causes the four segments to rotate around by swapping the background-image property values among the four <div> blocks.


The CSS defines one container style, then the styles for the four boxes that comprise the complete image.

The container looks like this:

#container {
  text-shadow:white 0px 0px 6px;

Then the four boxes defining the segments of the image are defined. Let's look at them one at a time.

#box1 {
  background-image: -moz-image-rect(url(https://mdn.mozillademos.org/files/12053/firefox.png), 0%, 50%, 50%, 0%);

This is the top-left corner of the image. It defines a rectangle containing the top-left quarter of the image in the file firefox.jpg.

#box2 {
  background-image: -moz-image-rect(url(https://mdn.mozillademos.org/files/12053/firefox.png), 0%, 100%, 50%, 50%);

This defines the top-right corner of the image.

The other corners follow a similar pattern:

#box3 {
  background-image: -moz-image-rect(url(https://mdn.mozillademos.org/files/12053/firefox.png), 50%, 50%, 100%, 0%);
#box4 {
  background-image: -moz-image-rect(url(https://mdn.mozillademos.org/files/12053/firefox.png), 50%, 100%, 100%, 50%);


The HTML is quite simple:

<div id="container" onclick="rotate()">
  <div id="box1" style="left:0px;top:0px;">Top left</div>
  <div id="box2" style="left:133px;top:0px;">Top right</div>
  <div id="box3" style="left:0px;top:136px;">Bottom left</div>
  <div id="box4" style="left:133px;top:136px;">Bottom right</div>

This places the four segments of our image in a two-by-two box grid. These four segments are all contained within a larger <div> block whose primary purpose is to receive click events and dispatch them to our JavaScript code.

The JavaScript code

This code handles the click event when the container receives a mouse click.

function rotate() {
  var prevStyle = window.getComputedStyle(document.getElementById("box4"), null).getPropertyValue("background-image");
  // Now that we've saved the last one, start rotating
  for (var i=1; i<=4; i++) {
    var curId = "box" + i;
    // Shift the background images
    var curStyle = window.getComputedStyle(document.getElementById(curId), null).getPropertyValue("background-image");
    document.getElementById(curId).style.backgroundImage = prevStyle;
    prevStyle = curStyle;

This uses window.getComputedStyle() to fetch the style of each element, shifting it to the following element. Notice that before it begins doing so it saves a copy of the last box's style since it will be overwritten by the third element's style. By simply copying the values of the background-image property from one element to the next with each mouse click, we achieve the desired effect.

What it looks like

Browser compatibility

FeatureChromeEdgeFirefoxInternet ExplorerOperaSafari
Basic support No No4 No No No1
FeatureAndroid webviewChrome for AndroidEdge mobileFirefox for AndroidOpera AndroidiOS SafariSamsung Internet
Basic support No No No4 No No1 No

1. See WebKit bug 32177


最終更新者: ExE-Boss,