96306[BUGFIX] Validation of TCA type "check"
Commit message
Test this change
Run a command in your clone of the TYPO3 Core. Then try what the change does and vote on Verified in Gerrit. How to review a patch
- Check out patch set 1
-
git fetch https://review.typo3.org/Packages/TYPO3.CMS refs/changes/06/96306/1 && git checkout FETCH_HEAD - Cherry-pick onto main
-
git fetch https://review.typo3.org/Packages/TYPO3.CMS refs/changes/06/96306/1 && git cherry-pick FETCH_HEAD
Review
-
Code review CR -1 Needs a vote
- Oli Bartsch -1
- Benni Mack +1
-
Verified V +1 Needs a vote
- Benni Mack +1
- core-ci +1 CI
Merge conflicts 16
- main 97304[WIP][BUGFIX] Render the record form from unsaved values
- main 97560[TASK] Find subpages with one query when copying and deleting pages
- main 97603[TASK] Introduce RecordTypeResolver and deprecate getTCAtypeValue()
- main 93819[WIP][BUGFIX] Guard translated tt_content free-mode moves
- main 94753[FEATURE] New TCA type "coordinate"
- main 96543[WIP][POC][TASK] Write and read records by the same rules and rights
- main 74781[FEATURE] Introduce new slug builders for records and string
- main 92859[WIP][BUGFIX] Make MM tables language-aware and workspace-aware
- main 95145[TASK] Migrate LanguageService->sL() labels to new domain syntax
- main 95343[FEATURE] Add `adminOnly` for TCA columns
- main 95638[TASK] Extend FieldValidationService to more TCA field types
- main 95639[TASK] Extract a read-only UniquenessChecker from DataHandler
- main 93407[TASK] Modernize `RecordHistoryStore`
- main 92214[TASK] Introduce generic permission sets for accessing data
- main 85574[WIP][FEATURE] Language-aware slugs and slug normalizer API
- main 87719[WIP][BUGFIX] Mitigate missing column in cached schema information
Issues 1
Files 3
Review log 45
-
Daniel Windloff uploaded patch set 1 10/02/26, 11:21
-
core-ci voted on patch set 1 V +1 10/02/26, 11:29 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/109921
-
Benni Mack voted on patch set 1 CR +1 V +1 10/02/26, 13:26
-
Benni Mack wrote 10/16/26, 07:45
Topic set to DataHandler -
Oli Bartsch commented on patch set 1 CR -1 10/30/26, 22:08
Unresolved
This documents the new behaviour as the old bug, which IMO misleads anyone (including me) reading the commit. let's say that pid 0 was treated as "no pid filter" and the global result overwrote the per-pid result. Now pid 0 is a real pid and both limits apply. Please also make the subject imperative, e.g. `[BUGFIX] Enforce both maximumRecordsChecked limits for type check`.DataHandler.php:2118UnresolvedI don't quite get this comment. Is this really related to the check? Anyways, I wonder if we really need to evaluate both TCA options, in case the first already fails. I think of moving this check directly into line 2112 and only move on with `maximumRecordsChecked` check if passed / not set. So skipping the global check when the per-pid check already reset the value.DataHandler.php:2762UnresolvedThe scalar types and the changed `$pageId` default are more than the fix needs, and only the `null` sentinel is required. This is an `@internal` method, but third-party callers passing a string uid or an explicit `0` could hit a `TypeError` or changed behaviour on a LTS branch. Please keep the parameter types out of the patch (or at least the backport) and change only the default and the `!== null` check.CheckboxValidationTest.php:83Unresolved`ForPidZero` and `ForPidNonZero` pass on the base code, and `ForPidNonZero` duplicates the existing InPid test. They add runtime but no regression protection. Please drop them or fold them into `ForPidNonZeroAndZero`. A data provider would also be nicer than five copy-pasted methods.CheckboxValidationTest.php:211UnresolvedThe third record is created on `$pid2`, which already holds a checked record, so `maximumRecordsCheckedInPid=1` rejects it. `$pid3` is created but never used. The test therefore still passes with the `maximumRecordsChecked` check disabled and doesn't guard the global limit. Please create the third record on `$pid3`.tx_testdatahandler_element_check.php:8UnresolvedThe fixture has copy leftovers: ctrl label `title` has no column, `showitem` lists the non-existent `check_maximum_records`, the palette `1` is empty, and the versioningWS and language fields are unused. Let's trim it to the two columns and `rootLevel` -1 that the tests need.