Bug 313649 – REGRESSION(312211@main) [GTK] Broke copy/pasting relative URLs

• Component: New Bugs
• Status: RESOLVED FIXED

[Vitaly Dyackhov]

Broken tests:
editing/pasteboard/block-wrappers-necessary.html
editing/pasteboard/copy-picture-and-text.html
editing/pasteboard/copy-picture.html
editing/pasteboard/copy-resolves-urls.html
editing/pasteboard/copy-text-with-wrapped-tag.html
editing/pasteboard/paste-4039777-fix.html
editing/pasteboard/paste-noscript.html
editing/pasteboard/reveal-selection-after-pasting-images.html

Diff:
```
--- /sdk/webkit/layout-test-results/editing/pasteboard/copy-resolves-urls-expected.txt
+++ /sdk/webkit/layout-test-results/editing/pasteboard/copy-resolves-urls-actual.txt
@@ -2,4 +2,4 @@
 
 link
 link
-<a href="../relative/path/foo.html">link</a><img src="resources/abe.png">
+<a href="file:///sdk/webkit/LayoutTests/editing/relative/path/foo.html">link</a><img src="file:///sdk/webkit/LayoutTests/editing/pasteboard/resources/abe.png">
```

[Vitaly Dyackhov]

Test gardening pull request: https://github.com/WebKit/WebKit/pull/63893

[EWS]

Test gardening commit 312277@main (bb013742ee86): <https://commits.webkit.org/312277@main>

Reviewed commits have been landed. Closing PR #63893 and removing active labels.

[Diego Pino]

Pull request: https://github.com/webkit/webkit/pull/65316

[EWS]

Committed 313644@main (f5376181a643): <https://commits.webkit.org/313644@main>

Reviewed commits have been landed. Closing PR #65316 and removing active labels.

[Radar WebKit Bug Importer]

<rdar://problem/177591039>

[Diego Pino]

In reply to this comment from Vitaly in the associated PR (https://github.com/WebKit/WebKit/pull/65316#issuecomment-4506061952)

> @dpino I wonder how it works on Apple ports.

In the PR that introduced the regression for WebKitGTK (https://github.com/WebKit/WebKit/pull/63451), the EWS WebKitGTK test bot shows several tests were failing. The PR was merged anyway, which is not great.

That PR was a follow-up to this other commit in the remote branch 'origin/safari-7624-branch':

```
31eb1554d216 2026-01-15 [e-mail]    safari-web-extension url masking bypass. https://webkit.org/b/304155 rdar://problem/166499973
```

Unfortunately, I don't have access to that bugzilla ticket.

It seems to me this was an issue related to Safari and Web Extensions. I think the behavior for WebKitGTK should have been preserved (considering the change made several layout tests fail for WebKitGTK).

BTW, WPE is also affected but right now WPE doesn't run 'editing/pasteboard' tests as they are marked as Skipped in WPE TestExpectations:

```
webkit.org/b/213131 editing/pasteboard [ Skip ]
```