Fix css-nesting parsing typo

[mdubet]

https://wpt.fyi/results/css/css-nesting/parsing.html?label=experimental&label=master&aligned

[wpt-pr-bot]

There are no reviewers for this pull request. Please reach out on the chat room to get help with this. Thank you!

[tabatkins]

I swear I made that edit already, but in any case, thanks!

[emilio]

@tabatkins are the expectations here correct? At the very least this probably needs:

diff --git a/testing/web-platform/tests/css/css-nesting/parsing.html b/testing/web-platform/tests/css/css-nesting/parsing.html
index b189190f41257..3ef3811b311cb 100644
--- a/testing/web-platform/tests/css/css-nesting/parsing.html
+++ b/testing/web-platform/tests/css/css-nesting/parsing.html
@@ -42,7 +42,7 @@
   testNestedSelector("&.bar");
   testNestedSelector("& .bar");
   testNestedSelector("& > .bar");
-  testNestedSelector("> & .bar");
+  testNestedSelector("> & .bar", {expected: "& > & .bar" });
   testNestedSelector("> .bar", {expected:"& > .bar"});
   testNestedSelector("+ .bar &", {expected:"& + .bar &"});
   testNestedSelector(".test > & .bar");

As otherwise the selector is invalid in non-nested contexts?

But this also expects that ampersands are introduced if not there already for non-relative selectors, in some cases even for non-relative selectors, which is very weird?

E.g, the test expects: .bar, .foo & to get turned into & .bar, .foo &, which doesn't seem necessary? All browsers right now serializes that as .bar, .foo & which seems fine, since it's not an invalid selector list in any context.

Just wanted to double-check the expectations before changing behavior here...

[emilio]

Also testNestedSelector(":is(.bar, .baz)", {expected:"&:is(.bar, .baz)"}); should probably be & :is(.bar, .baz) (note the extra space)?

[emilio]

(If we want to do that, which per my previous comment we might not...)

[tabatkins]

...sigh, the reason I was so confused is that I did fix a lot of these issues in #41050, but that never merged.

[tabatkins]

-  testNestedSelector("> & .bar");
+  testNestedSelector("> & .bar", {expected: "& > & .bar" });

Yup.

But this also expects that ampersands are introduced if not there already for non-relative selectors, in some cases even for non-relative selectors, which is very weird?

Hm, you're right, I just misinterpreted the resolution. Curiously enough, I wrote the spec correctly, matching the resolution.

Both of these are fixed in #41213.

[emilio]

Thanks!