radio-groupname-case.html still fails one caseless comparison test

[jdm]

https://github.com/servo/servo/pull/7777/files#diff-ed2c588bed7bbd35cf3ebc3cc09ded10R3

This points at an incomplete implementation in https://github.com/SimonSapin/rust-caseless, but we should figure out precisely what the failure case is in the testcase after #7777 merges.

cc @KiChjang in case you're interested in tackling this :)

[KiChjang]

I think I'll dive into this one once I get some of my other stuff done. In the meantime, if anyone is interested in this issue, go ahead ;)

[psdh]

This seems to fail on https://github.com/w3c/web-platform-tests/blob/master/html/semantics/forms/the-input-element/radio-groupname-case.html#L17 (on K to be specific). But at the same time, line number 34 doesn't fail.

[KiChjang]

For reference, K is the Kelvin sign used for measuring temperature.

[psdh]

@KiChjang Yep it is, the same kelvin sign matches correctly to 'k' (006b) in line 34. Not sure why that is happening

[psdh]

As SimonSapin pointed out, html spec says "no language-specific tailorings", and Unicode's CaseFolding.txt says about T "For non-Turkic languages, this mapping is normally not used."

[SimonSapin]

I think this is a test bug: web-platform-tests/wpt#2579

[annevk]

FYI: HTML changed in whatwg/html#1941.

[nox]

This seems to still be relevant.