cleanup WPT css-transitions timing-function parsing tests

[chromium-wpt-export-bot]

Move all such tests into parsing.
Eliminate duplicate tests and manual versions of automated tests.

Change-Id: I0ef507a1832c9ce113ed1a855653af7fd22bb0b0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1809508
Commit-Queue: George Steel gtsteel@chromium.org
Auto-Submit: George Steel gtsteel@chromium.org
Reviewed-by: Stephen McGruer smcgruer@chromium.org
Cr-Commit-Position: refs/heads/master@{#699042}

[wpt-pr-bot]

Already reviewed downstream.

[george-steel]

@foolip There was a final change upstream fixing a nit that never got pushed to the pull request (blink-wpt-bot desynchronized). What do we do to fix this?

[foolip]

@george-steel that is pretty serious, I'm not sure if I've seen that happen before. I'll check what was landed in Chromium and make sure WPT matches. There may be a revert of changes by importing WPT in the interim, but eventually it will stabilize.

[foolip]

To figure out what happened here I ran the steps for manually exporting a change in https://github.com/foolip/ecosystem-infra-rotation/blob/master/docs/export.md.

Applying the changes at the same point where this was landed (7b9a8b1) leads to conflicts in css/css-transitions/parsing/transition-timing-function-invalid.html and I presume that is why @Hexcles merged in master to resolve the conflicts.

Here's the conflict markers:

<<<<<<< HEAD
test_invalid_value("transition-timing-function", "initial, cubic-bezier(0, -2, 1, 3)");
test_invalid_value("transition-timing-function", "cubic-bezier(0, -2, 1, 3), initial");
=======
test_invalid_value("transition-timing-function", "steps(foobar)");
test_invalid_value("transition-timing-function", "steps(3.3, end)");
test_invalid_value("transition-timing-function", "steps(3, top)");
test_invalid_value("transition-timing-function", "steps(-3, top)");
test_invalid_value("transition-timing-function", "steps(0, jump-start)");
test_invalid_value("transition-timing-function", "steps(0, jump-end)");
test_invalid_value("transition-timing-function", "steps(0, jump-both)");
test_invalid_value("transition-timing-function", "steps(1, jump-none)");

>>>>>>> cleanup WPT css-transitions timing-function parsing tests

The first two lines come from #19195 (landed before this) and the rest are from https://chromium-review.googlesource.com/c/chromium/src/+/1809508.

Both should be kept. @Hexcles resolved the conflict by doing that, just in the other order compared to the conflict markers.

The remaining difference then is in css/css-transitions/parsing/transition-timing-function-computed.html, exactly as reported in #19124 (comment).

That change was reverted by import in https://chromium.googlesource.com/chromium/src/+/dd5d2254250ea85d458f6e6d581f137022ca89d5.

I'll send a Chromium CL to rectify this, which will then be exported too.

[foolip]

I've sent https://chromium-review.googlesource.com/c/chromium/src/+/1824266/ to restore the change and filed https://bugs.chromium.org/p/chromium/issues/detail?id=1007920 about how it could happen.

For anything seeing this and thinking that conflicts must be a major burden on 2-way sync, conflicts are actually very rare and not the main thing that requires human intervention. There are things that could be done to make conflicts less likely but so far that hasn't seemed like the right tradeoff.

[Hexcles]

Oops, sorry I totally missed the last patchset in the CL when I tried to resolve the merge conflict -- this is exactly why one shouldn't work when on vaca :)

Thanks for fixing this, @foolip .