This is a headless polyfill for the CSS object-fit
property which defines the sizing mode for content images (similar to background-size for CSS background sources).
Please note that as Microsoft Edge does not support a good way to get matching CSS rules, and there are a couple of other issues that aren’t likely to be fixed, this polyfill is of limited benefit for most projects as all other browser vendors support object-fit
natively now.
object-fit-images
and fitie
can be used instead of this polyfill.
The specification for object-fit
is to be found at W3C CSS3 Images. The property scales the image to fit in a certain way into a defined area, e.g:
img {
width: 100%; // dimensions are mandatory
height: 35em; // dimensions are mandatory
object-fit: cover;
overflow: hidden; // Cuts off the parts of the image poking out
}
Normally, the image would be stretched to the specified dimensions but due to the usage of the CSS property object-fit: cover;
the image now is scaled proportionally, until every pixel of the defined area is covered by parts of it. In the case of cover this means that parts of the image will overlap the given area.
The following are the possible values and their implications:
fill
streches the image exactly to the defined dimensions which results in a distorted image. Comparable tobackground-size: 100% 100%
. That's the default value.none
leaves the image at its natural size and centers it inside within the defined area. If the image's natural dimensions are larger than the defined area parts of the image will poke out of it unless you also setoverflow: hidden
on it. Comparable tobackground-size: auto auto; background-position: center center
.contain
scales the image up or down until all of it fits into the defined area. This mode respects the image's natural aspect-ratio. It's also called "letterbox view". Comparable tobackground-size: contain
.cover
scales the image up or down until every pixel of the defined area is covered with parts of the image. Sort of "pan and scan view". This means that parts of the image will poke out of the defined area unless you also setoverflow: hidden
on it. This mode respects the image's natural aspect-ratio. Comparable tobackground-size: cover
.
The polyfill uses a feature detection method to see if object-fit is supported. If it's not it will active itself.
This polyfill works in all major browsers as well as in IE9+. Find out which browsers support object-fit
natively.
Browser | polyfill? | natively? |
---|---|---|
Google Chrome | yes | v31+ |
Opera | yes | v24+ |
Firefox | 4+ (#13) | v36+ |
Internet Explorer | 9+ | "under consideration" |
This polyfill is available as Bower component or via npm. Use it right away from bower:
$ bower install --save object-fit
or set up via npm
$ npm install --save object-fit
The --save
flag is used to store the package dependency in the package.json so it can be automatically fetched next time using npm install
. Use --save-dev
to use it only as development dependency (but only do if you are sure you know what you do).
Or set up manually by grabbing the download from GitHub.
Then include the CSS file polyfill.object-fit.css
in your HTML <head>
, the JavaScript file polyfill.object-fit.min.js
at the bottom of your HTML <body>
. Right behind the JavaScript file reference you now need to call the polyfill:
<script>
objectFit.polyfill({
selector: 'img', // this can be any CSS selector
fittype: 'cover', // either contain, cover, fill or none
disableCrossDomain: 'true' // either 'true' or 'false' to not parse external CSS files.
});
</script>
You can find sample implementations in our test directory.
Due to CSP restrictions and our CSS parser there’s no way to test this polyfill from a filesystem. You need to set up a local server that serves from root directory. Calling http://localhost:8000/tests/index-cover.html
should work then. php -S localhost:8000
for example would start a local PHP server on your current directory.
In browsers greater IE8 the polyfill uses DOM Mutation Events or Mutation Observers (depending on what's available) to detect the injection of further images matching the defined selector. This means that it will also apply itself to any images that you append to the DOM at any later point. And it will detach itself from images that you remove from the DOM. Since this feature is sort of complicated to craft in a rock solid way, you might look out for unexpected behaviors.
If you use any third party CSS or mixed content on your website (Webfonts from a service, a CDN, or similar), the polyfill might not work as expected. You can still use the polyfill but then need to set some options regarding CSP:
For example you need to set the header to:
'Access-Control-Allow-Origin: *'
This should fix the issue. If you also need to support credentials, you can’t use *
but need the server reply with two headers (server needs also to reply with Access-Control-Allow-Credentials: true
), one of which includes the origin in question.
It is recommended to add the attribute crossorigin=""
to your CSS link
element that is called from the external resource to indicate what type of CORS the server should reply with.
In case you can’t alter the CSP / CORS settings of the server in question, you can disable parsing external CSS files in the config of the call:
<script>
objectFit.polyfill({
selector: 'img',
fittype: 'cover',
disableCrossDomain: 'true'
});
</script>
This polyfill is written by Anselm Hannemann and Christian "Schepp" Schaefer. Follow them on Twitter via @helloanselm and @derSchepp or check their GitHub profiles via anselmh and Schepp for more information.
This project is under the MIT Open Source License. See the LICENSE file for more information.