feat(rule): Scrollable region focusable

[jeeyyy]

Rule which checks if scrollable region has keyboard access.

Closes issue: #96

Reviewer checks

Required fields, to be filled out by PR reviewer(s)

• Follows the commit message policy, appropriate for next version
• Has documentation updated, a DU ticket, or requires no documentation change
• Includes new tests, or was unnecessary
• Code is reviewed for security by: Steve

Updated

[straker]

LGTM

[WilcoFiers]

I'm concerned about false positives in this one. Even a single pixel over overflow causes this rule to match the scrollable region. We're not actually looking to see if any of content is hidden. I think that's important. Thinking about this a little more, I'm wondering if we shouldn't exclude elements with overflow:hidden, and elements with the overflow area is small, maybe anything less than 1em or something. Edges get clipped all the time. Please come up with a proposal for this problem.

[dylanb]

@JKODU who has the ball on this PR?

[jeeyyy]

@dylanb it has been reviewed by @WilcoFiers. I am picking up with the changes requested. Should be ready for another round soon.

Updated

[WilcoFiers]

FYI, I didn't review the test cases yet. But I'm done for today, so I figured I'd share what I've got.

Updated

[kensgists]

This new rule is causing standard accessible text clipping techniques to fail. For example, this CSS:

.accessible-text { position: absolute !important;
clip: rect(0.0625rem 0.0625rem 0.0625rem 0.0625rem);
clip: rect(0.0625rem, 0.0625rem, 0.0625rem, 0.0625rem);
padding: 0 !important;
border: 0 !important;
height: 0.0625rem !important;
width: 0.0625rem !important;
overflow: hidden;
}`

is marked as a violation in axe-coconut.

@WilcoFiers pointed out above that elements with overflow:hidden might be exempted. I think this is a good idea.

We use axe-coconut for testing because of some of the useful experimental rules. But now we are getting vast numbers of false positives due to this new rule.