RGAA mapping may have some issues #4926

Product

axe-core

Product Version

4.11

Latest Version

• I have tested the issue with the latest version of the product

Issue Description

Version 4.11 of axe-core now support a mapping with RGAA tests, according to #4850

Here are some issues that we have detected in this mapping:

• aria-hidden-body and aria-hidden-focus are linked to test 7.1.1. Those rules are usually used in test 10.8.1, which is about hidden content.
• valid-lang is linked to 8.7.1. For us this axe rule should be linked to 8.8.1 instead of 8.7.1, as the goal of 8.8.1 is to evaluate if the language code is valid (8.4.1 does the same for the language attribute on the html element, it could also possibly be linked to this rule).
• aria-command-name is mapped to 11.9.1 only, but the rule is more generic and could be linked also to 7.1.1 for menuitems and to 6.2.1 for links
• aria-toggle-field-name is linked to 7.1.1, could it be linked to 11.1.1 instead, as it is about a form field?
• label-content-name-mismatch is linked to 6.1.5 but this rule is broader and supports a lot of roles. It could apply also to 11.9.2 for buttons, 11.2.5 for form fields

About the best practices currently linked to RGAA:

• focus-order-semantics (experimental) has been linked to test 12.8.1. We don't think, that this is part of RGAA, as in 12.8.1 we only check the consistency of the navigation sequence.
• region has been linked to test 9.2.1. We do not think, that this is part of RGAA, in particular 9.2.1 does not require that all elements of a page are in a landmark, but requires some landmarks to be present.
• table-duplicate-name has been linked to test 5.2.1. We do not think that this is part of RGAA, in particular 5.2.1 does not require that all tables have a different name.