english2009-02-07 jQuery Accessible Tabs - How to make tabs REALLY accessible
The one Problem even the better ones have is the missing Feedback of the content change. Most of the tabs scripts change the style of the tabs and the visibility of their corresponding content but leave the user just where he was - on the tab-link - not knowing that something happened (just like clicking on a, ever so popular, inactive link (a href="#")).
This is also the big difference this Script makes. While the user experience for users without visual impairment is exactly the same, there's a lot happening under the hood. But let's start from the beginning.
jQuery Accessible Tabs uses a very simple and flexible HTML markup as it's base to build upon. All it needs it's a wrapper containing Headlines and content elements one after the other. This provides a nice non-javascript fallback that does not look like something is missing or broken.
<div class="tabs">
<h2>some dummy headline</h2>
<div class="tabbody">
<p>Lorem ipsum dolor sit amet, consectetuer adipiscing elit.</p>
<h3>Lorem ipsum</h3>
<p>Nullam malesuada suscipit pede. Nullam ipsum lacus, varius</p>
</div>
<h2>another dummy headline</h2>
<div class="tabbody">
<p>Integer tincidunt. Cras dapibus. Vivamus elementum nisi.</p>
<p>Quisque rutrum. Aenean imperdiet. Etiam ultricies nisi vel.</p>
</div>
<h2>anything else</h2>
<div class="tabbody">
<p>Here could be your content</p>
</div>
</div>
A simple jQuery call then transforms this to the accessible Tabs:
$(document).ready(function(){
$('.tabs').accessibleTabs();
});
To not only be accessible but also to be as flexible as possible the script allows configurations to be passed to.
$('.tabs').accessibleTabs({
// Classname to apply to the div that is wrapped around
// the original Markup
wrapperClass: 'content',
// Classname to apply to the LI of the selected Tab
currentClass: 'current',
// Tag or valid Query Selector of the Elements to Transform the
// Tabs-Navigation from (originals are removed)
tabhead: 'h4',
// Tag or valid Query Selector of the Elements to be treated as
// the Tab Body
tabbody: '.tabbody',
// can be 'fadeIn', 'slideDown', 'show'
fx:'show',
// speed (String|Number): 'slow', 'normal', or 'fast') or the number of
// milliseconds to run the animation
fxspeed: 'normal',
// text to indicate for screenreaders which tab is the current one
currentInfoText: 'current tab: ',
// Definition where to insert the Info Text.
// Can be either 'prepend' or 'append'
currentInfoPosition: 'prepend',
// Class to apply to the span wrapping the CurrentInfoText
currentInfoClass: 'current-info'
});
So, why is this Tabs-Script more accessible than others?
The main thing this Script does better than the others is providing a click feedback for Screenreader Users. When the user clicks on one of the tabs there's actually a navigation going on on the page.
Every link in the tabs is pointing to an actually focusable content. To allow this, we create a named anchor inside of the tabs content. This anchor, as it has no href property, is not focusable by default though. To make it focusable we gave it a tabindex="0". Why 0? Because a tabindex of 0 does not change the natural tab flow of the page but it makes an element focusable which is not able to receive focus by default.
<h4>
<a id="accessibletabscontent" name="accessibletabscontent" class="accessibletabsanchor" tabindex="0">
some dummy headline
</a>
</h4>
Inside of the named anchor we add the same text as the tab that was just clicked and wrapped it into a headline element. This provides a very nice user experience for screenreader users, as it nicely confirms that you actually jumped into a new content of the page.
The other thing that adds benefit for screenreader users is a little text that indicates the currently selected tab. It defaults to "current tab: " followed by the default text of the tab. For internationalisation reasons this text as well as the text position (either before or after) can easily be configured.
Problem: Focussing hidden or recently hidden Elements
During the testing of the tabs we discovered a very interesting, yet very annoying, behaviour of the Screenreader Jaws (and probably others too). We wanted to focus the headline in the content of the current tab after clicking the tab which made the content visible. Everytime we tested, the Screenreader was unable to find the element to focus on and simply jumped to the end of the page or randomly somewhere else. We discovered that this is only happening when the focus target was recently hidden. This phenomenon is caused by the screenreaders virtual buffer, an internal copy of the actual DOM in which the Screenreader user navigates. The virtual buffer does regularly update itself from the Browser DOM but there's always a delay.
To work around this issue we used a different approach. The named anchor we're linking to is not actually in the hidden tab content but before it in the DOM of all of the contents and hidden offscreen. Also actually every link in the tabs is pointing to the same anchor. We're simply dynamically rewriting the headline text on the fly and show and hide the contents accordingly. The user experience is just the same. The screenreader user clicks on the tab, follows the link and finds the linked content starting with a corresponding headline.
You can see the script in action in a simple version here and in a Yaml environment here. This plugin will also be part of the The CSS Framework Yaml in its upcoming version 3.1.1.
You can download the script including the demos here.