So, the behaviour you're getting from VoiceOver is not necessarily standard behaviour. It may help if you can post some of your code.Screen readers are all over the map with respect to interpretting aria-label on non-interactive elements.Īccording to David MacDonald's tests, VoiceOver and TalkBack will read the aria-label text on static elements, but JAWS and NVDA will not unless the static content (like a ) has the role="navigation", role="search", role="main", or role="img".īy design, aria-label is meant to override the existing text of an element because it is what most screen readers will interpret as the element's label. If your buttons are not really part of a group but you're still hearing group info, then you have a bug in your code. It also depends how you navigate to the button, whether via TAB or the up/down arrow keys to walk the DOM.īut the lesson here is that if you have a group of things and it's properly coded, then you want the screen reader to announce the group info. Some will say "1 of 2", some might just say it's a group (without the number). Note that different screen readers may announce groups of things differently. Instead of using a real, I could use role="list" and role="listitem" and get the same type of "1 of 2" messages. You don't want to take that feature away unless you are misusing a grouping feature and the button isn't really part of a group.Įxamples of different ways to get the group position announced. Sighted users get that information with their eyes. Hearing "1 of 2" is a fantastic feature of screen readers to give the user an idea of how big a set is and which item they're on in the set. You can also get "1 of 2" announced if you're using ARIA attributes such as aria-setsize and aria-posinset, but those attributes aren't heavily used so I'm guessing it's not that. So if you have your buttons or links inside of a list ( or ) or if the buttons are natively part of a group such as a radio button group ( ), then it's absolutely correct for NVDA to announce the item number within the set. What it does know about is semantic HTML elements.
NVDA doesn't know anything about arrays or the language you're using to code in.
0 kommentar(er)
