Further implicit role updates needed

[scottaohara]

This is a follow-on issue to #392.

The following map to generic

• a w/out href
• area w/out href
• b
• data
• i
• pre
• q
• samp
• small
• span
• u

Repurposing this issue to be scoped only to the roles which need to map to generic. All other roles will be handled in their own issue(s) as some mappings require implementation updates before they can land here, and some roles may even need to be created still.

Related to HTML AAM issue: w3c/html-aam#373

[JAWS-test]

I think most of the elements mentioned should not map to generic, because generic means in the APIs: ROLE_SYSTEM_GROUPING, IA2_ROLE_SECTION, Group, ROLE_SECTION, AXGroup - or will that be changed at https://w3c.github.io/core-aam/ in the future?

If I understand correctly, all these roles are about a group, i.e. an independent page section (as in HTML for the elements div, section, article, fieldset), but not about individual sections within a text or even within a word (as in HTML elements a, b, i, s, span, u, etc.).

For example, I would expect a generic role in the screen reader to cause a reading pause to occur. However, that would be wrong for inline elements.

Currently, JAWS outputs generic elements contiguously, but NVDA separates them with spaces, so that words containing bold or italic letters, for example, are no longer output as a word

[scottaohara]

@JAWS-test core aam needs adjusting, I've brought up the same topic with the ARIA wg a few times now, and the consensus is to use generic and further specify the differences between when a generic would behave like a div vs a span, for example.

[scottaohara]

seems this issue was completed but i failed to close it. so doing that now.