On Tue, 3 Mar 2026 21:57:23 GMT, Andy Goryachev <[email protected]> wrote:
> Can this be tested with the monkey tester? If so, what should be the expected > outcome? Sure. The outcome is that it should not differ from the previous implementation (and it does, thanks for reporting that). > Also, for the bookkeeping sake, do you think it might be worth to create an > umbrella task linking all the similar tickets? There already is - [JDK-8301311](https://bugs.openjdk.org/browse/JDK-8301311) i just forgot to lnk this new bug to it and i already done so. > VoiceOver does not announce "visited" state with this change, the master > branch does. To reproduce, open two monkey testers one from the master > branch, the other from this PR, select Hyperlink page and click between the > two hyperlinks. > > This PR says "java, Monkey Tester - Hyperlink, window, link, short" > > The master branch announces: "java, Monkey Tester - Hyperlink, window, > visited, link, short" > > ("short" is the text in the hyperlink) I will fix that shortly. I haven't found a new function based call that reports visited status so i am going to expose the method that allows access to the accessibility properties in classes where it is required starting from this one. > modules/javafx.graphics/src/main/native-glass/mac/a11y/AccessibleBase.m line > 55: > >> 53: [rolesMap setObject:@"JFXCheckboxAccessibility" >> forKey:@"TOGGLE_BUTTON"]; >> 54: [rolesMap setObject:@"JFXStaticTextAccessibility" forKey:@"TEXT"]; >> 55: [rolesMap setObject:@"JFXLinkAccessibility" forKey:@"HYPERLINK"]; > > Would it make sense to sort the source by `forKey` name (for human > consumption)? > > I take it since it's a map, the declaration order is not important? Sure, i added the new bindings in order of implementation and this was a special case because i initially aimed as binding the hyperlink to the static text accessibility peer and just add the special case reporting accessibility role but then i decided to create a new class and i'm glad i did - looks like it differs not only in a role reported but functionally too. So yes, changing the order based on the requested role is reasonable. ------------- PR Comment: https://git.openjdk.org/jfx/pull/2094#issuecomment-3999800612 PR Comment: https://git.openjdk.org/jfx/pull/2094#issuecomment-3999813693 PR Review Comment: https://git.openjdk.org/jfx/pull/2094#discussion_r2885723266
