[Accessibility] SVG examples to check results and monitoring how browsers treats SVG elements

[giacomo-petri]

Closes: web-platform-tests/interop-accessibility#173

Description:
These WPT tests are designed to evaluate how browsers expose roles for different SVG elements.

Note:
I haven't set any expectations. While the specs require all SVGs to be treated equally as graphics-document role, I believe there should be (and likely already is) some heuristic to communicate SVGs based on their intended purpose for users.

[cookiecrook]

Related to (possibly blocked by?)

• Define computedrole for the SVG roles table w3c/svg-aam#24 (comment)
• Should concrete shape/object roles like circle have a more specific role description? w3c/svg-aam#30

[cookiecrook]

FYI @cyns re: blocking issues since she's resurrecting SVG-AAM now...

[cyns]

We're just getting started, and it seems likely some of these mappings will change. What is the impact of having these tests merged? Is it likely to push implementations, before we're ready?

[cookiecrook]

Mainly to illustrate the differences here... (as a tentative test file, that's okay, as long as the test expectation isn't confused for the real implementation, hence the ALL CAPS SENTENCE)

the expectation "should always be graphics document" is a bit wordy and possibly misleading... Perhaps it should be "UNDEFINED in SVG-AAM, but possibly graphics-document." Giacomo's expectation is from the base graphics-aria listings of computed role, means that Firefox's implementation is closest to spec, but the existing open questions prevent us from adding the concrete expectation. Even in a tentative test.

• Define computedrole for the SVG roles table w3c/svg-aam#24 (comment)
• Should concrete shape/object roles like circle have a more specific role description? w3c/svg-aam#30

[cookiecrook]

r+ with update diffs