Wpt Versions Save

[html/canvas] tests for TextMetrics.advances should not be in WPT as this attribute is not in the spec (#46231)

Replace 2d.text.measure.advances test with a test that asserts advances does not exist, as it is not in the spec.

Add missing ARIAMixin reflection tests

Exclude interpolation test from overflow shorthand test list (#46227)

Many of the subtests depend on transition-behavior via interpolation-testcommon.js.

CSS highlight pseudos: parse ::search-text and ::search-text:current

This patch adds a new runtime enabled feature for ::search-text (SearchTextHighlightPseudo), new PseudoId and PseudoType and devtools protocol variants for ::search-text and :current, and extends the selector parser to accept ::search-text and ::search-text:current when the feature is enabled.

:current has an alternate functional form [1] that, at least in a WebVTT context, allows you to match ancestors of the current element. Parsing this will be left to a separate patch.

[1] https://drafts.csswg.org/selectors/#the-current-pseudo

Bug: 339298411 Change-Id: I2f12acd9580ae03dbdd816a03cbb68017864d136 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5515107 Commit-Queue: Delan Azabani [email protected] Reviewed-by: Anders Hartvoll Ruud [email protected] Reviewed-by: Dmitry Gozman [email protected] Cr-Commit-Position: refs/heads/main@{#1299975}

WebKit export: Re-sync color conversion constants with latest in spec https://bugs.webkit.org/show_bug.cgi?id=272952

Test removed canvas features are removed

See https://github.com/whatwg/html/pull/9979 & https://github.com/whatwg/html/pull/8229 for context.

Fetch: fix variable typo in redirect-referrer-override.any.js

Revert "[css-anchor-position-1] Fix flicker with try-option with fixed-position anchor"

This reverts commit 94af1f503694c3a7ffece9c910516b3bbe9d50c2.

Reason for revert: Causing crashes in GC, see bug.

Bug: 339967265, 338929745

Original change's description:

[css-anchor-position-1] Fix flicker with try-option with fixed-position anchor

Previously we supposed the default anchor wouldn't change in try-options, but the latest draft spec allows that.

When the currently tried option will use a default anchor that is different from the currently used default anchor, checking overflow with the snapshotted AnchorPositionScrollData will be incorrect, and the incorrect results caused oscillation between try options.

To resolve that, when the total offset is requested for a different default anchor, we need to calculate it on the fly instead of getting it from the snapshotted result.

Also allow one more layout if ValidateSnapshot() (called after the first layout in a document lifecycle update) finds that the snapshotted scroll data needs update, to ensure correct layout and paint with the changed anchor and scroll adjustment in most cases.

In rare cases when we need more than one additional layout to get the stable layout, ShouldScheduleNextService() will schedule another lifecycle.

Fixed: 338929745 Change-Id: I295fbc605ac8b7f0298eef7aeff4434c6bd54628 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5529048 Reviewed-by: Ian Kilpatrick [email protected] Reviewed-by: Philip Rogers [email protected] Commit-Queue: Xianzhu Wang [email protected] Cr-Commit-Position: refs/heads/main@{#1299608}

Change-Id: I222274aa6af5da55d5d6dc0867d3cda684efff49 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/5526668 Bot-Commit: Rubber Stamper [email protected] Commit-Queue: Michael Lippautz [email protected] Cr-Commit-Position: refs/heads/main@{#1299965}

WebKit export of https://bugs.webkit.org/show_bug.cgi?id=273751

[WebDriver BiDi] Remove context as the interception of Realms does not include it (#45910)