Changes look good. However, there is still an open thread in the merge request that is possibly outdated?
Ran into this issue as well after updating to beta3. The issue was fixed in the -dev version, but the change to does apply cleanly as a patch
#3538150-18: 502 Bad Gateway on Frontend After Update to beta 3 (from 2) → mentions
And if everything is ok, I'll publish a 2.0.0-beta4 release.
A new beta version should prevent other sites from running into this issue, so a new release tag would be very welcome
📌 Convert to OOP Hooks Active was committed.
📌 Convert to OOP Hooks Active was committed.
The unit tests is renamed so it is detected on Drupal.org infrastructure, see https://git.drupalcode.org/issue/search_api_attachments-3510970/-/pipeli...
Berdir mentions in [3396165-25]:
I realized that we forgot about another generic plugin feature beside deriver, the provider. It isn't used a lot in core for obvious reasons, but it's a valid use case to implement a plugin if another module exists.
Was this ever implemented for plugin attributes? Running into this now at the Linkit conversion: 📌 Support attribute based plugin discovery Active
The merge request fixes phpstan/phpcs/stylelint pipeline job findings
The autocomplete icon in the CKEditor widget cannot be updated with this approach without duplicating css, so the approach in 🐛 Improve autocomplete UI Active is preferable
Icon looks much better now, thanks!
Merge conflict resolved after 📌 Use latest rules for postcss-nesting Active was committed.
The WebformBreadcrumbBuilderTest @dataProvider is now a static method
This feature is available through contrib, see https://www.drupal.org/project/composer_deploy →
https://git.drupalcode.org/project/varnish_purge/-/merge_requests/12/dif...
In the commit message the dependencies in the info.yml file are assumed to resolve automatically. This assumption is incorrect for this module.
The rename of the info.yml file is out of scope for this issue
In the DisplayFeedTest.php the description is only checked for html, but not the CDATA. Can you update the assertion to include the CDATA, since this allows the inner text to be html ?
The proposed resolution is to add a composer.json. The MR does not match the proposed resolution.
Tests are failing on Drupal 11.2 regardless, see 📌 Fix tests on 11.2 Active
The merge request:
- fixes tests on 11.2
- includes the deprecation for invalidateAll(), see https://www.drupal.org/node/3500622 →
I added a screencast on a new Drupal standard installation on 11.2.x with steps to reproduce
The redundant lines were removed.
Updated the issue summary so it matches the current implementation
Code change works as expected, thanks!
I filed a follow-up issue to fix the unrelated issue in the cspell pipeline job, see 📌 cspell issues reported in pipeline Active .
idebr → changed the visibility of the branch 3185024-change-feature-policy to hidden.
Not sure why this was postponed in #5? Seems like a useful feature
To do: find out how to migrate the provider attribute. See #3396165-25: [meta] Convert all core plugin types to attribute discovery →
phpcs finding were fixed in 📌 Fix phpcs pipeline job findings Active
phpcs finding were fixed in 📌 Fix phpcs pipeline job findings Active
Test coverage shows the issue is fixed. Contrib tests in 🐛 Fix tests (next minor) Active are fixed as well with this change. Thanks!
I filed an issue for Drupal Core: 🐛 A reverted revision is not listed on the version history Active
\Drupal\node\Form\NodeRevisionRevertForm does not call \Drupal\Core\Entity\ContentEntityStorageBase::createRevision that sets the revision_translation_affected.
See \Drupal\Core\Entity\Form\RevisionRevertForm::prepareRevision for reference
These test failures now affect the regular 'phpunit' pipeline step
Closed as a duplicate of 🐛 Fix tests (next minor) Active
Patch file for 8.x-1.x for projects running the final version of 8.x-1.x
The merge request removes the outdated \Drupal\Tests\diff\Functional\CoreVersionUiTestTrait
Closed as a duplicate of 🐛 Fix tests (next minor) Active
The merge request updates 'assertEquals()' arguments in tests to expected, actual
The merge request adds support for Datetime Range field type and adds test coverage for the daterange field type
8.x-2.4 was released 7 April 2025
@rang501 can you create a merge request instead of uploading a patch? Working with merge requests is the preferred workflow
Let's close this issue as a duplicate of ✨ Missing schema tags for courses Active , so we can focus our efforts in the related issue.
Settings to 'Needs work' to fix the test failures
BC break reported in Editor Advanced Link: 🐛 D11.2: Uncaught CKEditorError: Cannot read properties of undefined (reading 'viewUid') Active
The merge request now include a custom Content Lock phpcs ruleset that excludes Drupal.Arrays.Array.LongLineDeclaration
Changes for Drupal.Arrays.Array.LongLineDeclaration were reverted.
Gitlab CI will pick up a custom phpcs configuration, as do IDEs like PHPStorm. Is Drupal.Arrays.Array.LongLineDeclaration the only rule that should not be enforced?
See webform for reference: https://git.drupalcode.org/project/webform/-/blob/6.3.x/phpcs.xml?ref_ty...
The merge request applies gitlab CI defaults for phpcs
Thanks!
Domain source uses the base_url option for its output. There is no need to add the Country path prefix anymore, as it is already available in the base url.
The 'external' option is not a supported key for outbound processing, so that can be removed.
BC break reported in Linkit: 🐛 Future compatibility fix for v45 of CKEditor5 Active
✨ Linkit for Link field Fixed introduced the link-widget-title--auto class that indicates the title was set manually and should not be overwritten
The regression was introduced in 10.4.x, but the fix is only available in 11.x. Any chance of a backport to 10.x?
The merge requests updates the Domain Source OutboundPathProcessor to not rewrite the outbound URL if the source domain is the active domain
Nice to fix an issue by removing code, thanks!
alexpott → credited idebr → .
I think it was a prerequisite for ✨ Release lock when browsing away from locked form RTBC and it has been enabled ever since
Oh I see.
The merge request works as expected, thanks!
Nice to see the CSRF token in place
What is the use for the destination query parameter in content_lock_js_url? This URL is only called in the background and not something a user actually visits?
I expected a CSRF token instead of a custom hash. Can you elaborate why the custom hash is preferred in this case?
The phpstan pipeline job findings are fixed in the merge request