Review main Merged

91285[FEATURE] Introduce itemsProcessors over itemsProcFunc

Benni Mack CR +2 V +2 Expert +1972 −126 updated 11/09/25, 16:27

Commit message

Items processing for check, radio and select-type fields is
centralized and handled more safely with strictly-typed code.

As part of this, a new TCA option "itemsProcessors" is introduced,
as a future replacement for good old "itemsProcFunc", but no
deprecation on the horizon for now.

TCA "itemsProcessors" allow for multiple processors to
be chained, whereas extension authors struggled in the
past that "itemsProcFunc" was only available for the last
extension in loading order to set this option.

Beyond a cool new feature, this patch is a building block
for introducing record list filters.

Review

  • Code review CR +2 Done
    • Oli Bartsch +2
    • Garvin Hicking +1
    • Nikita Hovratov +1
  • Verified V +2 Done
    • Oli Bartsch +2
    • Garvin Hicking +1
    • Nikita Hovratov +1
    • core-ci +1 CI

Issues 1

Files 39

Show 19 more files

Review log 116151

Show 96126 earlier messages
  1. Benni Mack uploaded patch set 1 10/27/25, 18:20

  2. core-ci voted on patch set 1 V -1 10/27/25, 18:35 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87712

  3. Francois Suter uploaded patch set 2 V -1 10/28/25, 14:55

  4. core-ci voted on patch set 2 V -1 10/28/25, 15:12 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87843

  5. Francois Suter uploaded patch set 3 V -1 10/28/25, 15:40

  6. core-ci voted on patch set 3 V -1 10/28/25, 15:48 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87853

  7. Francois Suter uploaded patch set 4 V -1 10/28/25, 16:27

  8. core-ci voted on patch set 4 V -1 10/28/25, 16:35 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87863

  9. Stefan Bürk commented on patch set 4 10/28/25, 20:21

    ItemProcessingService.php:79

    I love the idea of this. For some of our extensions we defined itemsProcFunc classes dispatching a event so instances/extending extension can manipulate items in a easier way (which would be still working for now in the current state). Now, it is possible to configure **multiple** items processors for a tca table column which are called in a chain, which is also a good thing. Currently this is considered as "simple" value array which requires to use array methods to eventually rearrange a sorting - not sure if a `identifier` based along with `before/after` would be reasonable or something like that ... and some kind of ordering .. not sure. Would make that easier to use for the developer/integrator on top of the stricter type gain. Just some notes/ideas ... crazy for sure. But maybe not that wrong.
  10. Josef Glatz updated the commit message in patch set 5 V -1 10/28/25, 21:15

  11. core-ci voted on patch set 5 V -1 10/28/25, 21:24 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87891

  12. Francois Suter commented on patch set 5 10/29/25, 08:25

    ItemProcessingService.php:79 Reply

    I thought about that too. I'll discuss it here at the code sprint, but I feel like a full-fledged before/after registration is a bit overkill. But we could use simple numerical keys and do a ksort() before calling the processors.
  13. Francois Suter uploaded patch set 6 V -1 10/29/25, 09:54

  14. Francois Suter commented on patch set 6 10/29/25, 09:55

    ItemProcessingService.php:79 Reply

    Added a simple ksort and adapted RST file accordingly
  15. Francois Suter commented on patch set 6 10/29/25, 09:56

    Still to do: - add TSconfig option for itemsProcessors (same as for itemsProcFunc) - add tests for itemsProcessors
  16. core-ci voted on patch set 6 V +1 10/29/25, 10:03 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87927

  17. Francois Suter uploaded patch set 7 V +1 10/29/25, 11:16

  18. Francois Suter commented on patch set 7 10/29/25, 11:18

    TSconfig handled
  19. core-ci voted on patch set 7 V +1 10/29/25, 11:25 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87948

  20. Francois Suter uploaded patch set 8 V +1 10/29/25, 12:29

  21. Francois Suter updated the commit message in patch set 9 10/29/25, 12:31

  22. core-ci voted on patch set 8 V -1 10/29/25, 12:39 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87962

  23. core-ci voted on patch set 9 V -1 10/29/25, 12:40 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87963

  24. Francois Suter uploaded patch set 10 V -1 10/29/25, 12:44

  25. core-ci voted on patch set 10 V +1 10/29/25, 12:52 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87964

  26. Francois Suter uploaded patch set 11 V +1 10/29/25, 14:51

  27. Oli Bartsch commented on patch set 10 10/29/25, 14:51

    I actually dreamed about such patch also tackling the fact that one can still not rely on itemsProcFunc being evaluated in DataHandler for select fields.
  28. Nikita Hovratov commented on patch set 11 10/29/25, 15:07

    Thinking about the numeric ordering approach. Maybe we should take the opportunity to also introduce a dependency ordering system? Here is why I think this might be helpful: third-party extensions define itemsProcFunc for certain fields like `CType` or `colPos`. It wasn't possible to define multiple processor of course, so they did stuff like this: https://github.com/Gernott/mask/blob/2c83042840c83cf8b1dc13ce1b86e50762ffa31e/Classes/ItemsProcFuncs/CTypeList.php#L74 Storing the current itemsProcFunc and then calling it manually. With dependency ordering ala before/after, those extensions could define dependencies to each other to ensure processing order. Otherwise, everyone would need to agree to arbitrary numeric indexes.
  29. Nikita Hovratov commented on patch set 11 10/29/25, 15:11

    DefaultTcaSchema.php:711

    Let's do `($fieldTypeConfiguration['itemsProcessors'] ?? []) !== []` instead.
  30. Francois Suter commented on patch set 11 10/29/25, 15:20

    Reply

    We discussed this here at the code sprint and felt that it was overkill to introduce a full dependency handling (also thinking that this is called a lot in the core and we shouldn't burden it too much). After all it is pretty easy to check existing indices and resulting call stack.
  31. Francois Suter commented on patch set 11 10/29/25, 15:20

    Reply

    Acknowledged
  32. Francois Suter commented on patch set 11 10/29/25, 15:23

    DefaultTcaSchema.php:711 Reply

    Fine with me. Coming up in next version.
  33. Francois Suter uploaded patch set 12 10/29/25, 15:23

  34. Nikita Hovratov commented on patch set 11 10/29/25, 15:24

    With itemsProcFunc it was possib

    Reply

    Generally speaking, if we do not limit ourselves with a [int => class-string] type of array, then we open up this feature for future enhancement. I could also imagine defining custom config for a processor, which is then available as a parameter. Example from Core: We added the feature "itemGroups", which is an array like [string => string]. The key is the group name and the value the Label. In v13 we needed to define an ordering for those groups, but it was not possible without a breaking change.
  35. Benni Mack commented on patch set 12 10/29/25, 15:31

    Reply

    The main benefit for this change is to have multiple processors. Typical use-cases: - gridelements + container in one installation - I want to add further items, but EXT:news is already doing this. - I want to reorder existing items that already have an itemsProcFunc We've been thinking about using a PSR-14 event but it's too unspecific but it would allow sorting and blocking. The idea of sorting in a fashion that says "I need to hook into itemsProcFunc AFTER container has made their changes" is legit, but rather uncommon, so the sorting (= predictable) by key is a easy way, instead of doing DependencyOrderingService. Further implementation details are that one can now use the same parameters for all places the same way. Do you have further use-cases?
  36. Nikita Hovratov commented on patch set 12 10/29/25, 15:42

    Reply

    Could we at least make it an associative array and not do the same mistake again like for `items`, or `itemGroups` ? ``` 100 => ['class' => MyClass'], 200 => ['class' => MyOtherClass'], ```
  37. Francois Suter commented on patch set 12 10/29/25, 16:03

    Reply

    That could indeed be useful (and I had thought about it at some point). Coming up!
  38. Francois Suter uploaded patch set 13 10/29/25, 16:29

  39. Nikita Hovratov commented on patch set 13 10/29/25, 16:57

    Reply

    Thanks 😊
  40. core-ci voted on patch set 11 V -1 10/29/25, 17:42 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/87994

  41. core-ci voted on patch set 12 V -1 10/29/25, 17:49 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88000

  42. core-ci voted on patch set 13 V -1 10/29/25, 18:01 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88009

  43. Francois Suter uploaded patch set 14 V -1 10/30/25, 09:36

  44. core-ci voted on patch set 14 V +1 10/30/25, 09:46 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88097

  45. Benni Mack updated the commit message in patch set 15 V +1 10/30/25, 10:07

  46. core-ci voted on patch set 15 V +1 10/30/25, 10:16 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88104

  47. Francois Suter uploaded patch set 16 V +1 10/30/25, 10:38

  48. Oli Bartsch commented on patch set 15 10/30/25, 10:40

    Reply

    Sorry I should have been more clear. I miss the itemsProcFunc evaluation in the `checkValueForGroupFolderSelect.`
  49. core-ci voted on patch set 16 V +1 10/30/25, 10:47 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88107

  50. Francois Suter commented on patch set 16 10/30/25, 10:55

    Reply

    Now I understand. I would consider that a separate patch (a bugfix actually), since "itemsProcFunc" is not currently evaluated at all. What do you think?
  51. Oli Bartsch commented on patch set 16 10/30/25, 11:28

    Reply

    I would like to do this in a whole now, as this was always the idea, when introducing a new API to cover all parts especially in DataHandler.
  52. Francois Suter commented on patch set 16 10/30/25, 11:38

    Reply

    OK, I'll look into that.
  53. Oli Bartsch commented on patch set 16 10/30/25, 11:58

    Reply

    Just for background: we’ve always had to add disclaimers to the itemsProcFunc API, such as “Note that this cannot be properly validated in DataHandler due to missing context.” This has long been a major issue, since proper item validation couldn’t be ensured when using an itemsProcFunc. That’s unfortunate, because it’s actually a solid and widely used API — even in core. Therefore, we should finally address this once and for all.
  54. Garvin Hicking commented on patch set 16 10/30/25, 12:03

    ItemProcessorInterface.php:25

    phpdoc here is mostly redundant, better only keep @param SelectItem[] $items @return SelectItem[] as this is the only non-native declaration.
  55. Garvin Hicking commented on patch set 16 10/30/25, 12:11

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:19

    ```suggestion as an array of arbitrary data. ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:24

    ```suggestion Among the arguments that the processor instance receives, the items are an array of ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:25

    ```suggestion :php:`\TYPO3\CMS\Core\Schema\Struct\SelectItem` instances. The processor must return ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:26

    ```suggestion a similar structure. Added items cannot be untyped arrays anymore. This makes the ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:37

    ```suggestion .. code-block:: php :caption: EXT:my_extension/Configuration/TCA/my_table.php ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:59

    also adjust this syntax to the sub-array above? Or is a non-array syntax also supported?

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:69

    ```suggestion .. code-block:: php :caption: EXT:my_extension/Classes/MyPackage/Processors/SpecialRelationsProcessor.php ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:111

    ```suggestion Note how the numerical key of the processor must be reused. With this setup, ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:112

    ```suggestion the class :php`SpecialRelationsProcessor` would receive the PHP array :php:`['foo' => 'bar']` ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:118

    ```suggestion .. code-block:: xml :caption: EXT:my_extension/Configuration/FlexForms/SomeForm.xml ```

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:138

    Please also describe what happens if one TCA field has both itemsProcFunc and itemsProcessors set - I suppose the latter will be preferred, and the former ignored?
  56. Francois Suter commented on patch set 16 10/30/25, 16:50

    Thanks for all these improvements. Done in upcoming new version.

    ItemProcessorInterface.php:25 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:19 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:24 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:25 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:26 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:37 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:59 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:69 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:111 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:112 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:118 Reply

    Acknowledged

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:138 Reply

    Acknowledged
  57. Francois Suter uploaded patch set 17 V +1 10/30/25, 16:52

  58. Francois Suter commented on patch set 17 10/30/25, 17:01

    Reply

    After some discussion, this would be done in a separate bugfix, as it touches on the RelationHandler and would ideally imply refactoring it beforehand.
  59. core-ci voted on patch set 17 V +1 10/30/25, 17:02 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88172

  60. Garvin Hicking commented on patch set 17 10/31/25, 07:35

    I'm testing this now. Don't be afraid, the comments are actually only addressing two things. ;)

    AbstractItemProvider.php:131

    Currently our convention according to https://docs.typo3.org/m/typo3/reference-coreapi/main/en-us/PhpArchitecture/NamedArguments.html is that we only use named arguments for value objects / data (transfer) objects for their constructor calls in CPP context. I understand why you used it for this large amount of parameters, however we would like to not make a precedent exemption to bend the current rules for this. Could you check the patch for usage of these named parameters and adapt it? Another sidenote: Could you use an injectItemProcessingService like the injections above to prevent utilizing makeInstance?

    BackendUtility.php:1228

    ```suggestion ($columnConfig['itemsProcessors'] ?? []) !== [] ``` Just a minor thing, but we usually do this "empty array" comparison in the core to prevent `count()` usage.

    BackendUtility.php:1231

    (named arguments)

    BackendUtility.php:1315

    (also adjust `count()`)

    BackendUtility.php:1319

    (named arguments)

    DataHandler.php:2031

    (`count()` replacement)

    DataHandler.php:2035

    (named arguments)

    DataHandler.php:2126

    (`count()` replacement)

    DataHandler.php:2130

    (named arguments)

    ItemProcessingService.php:97

    (named arguments)

    ItemProcessingService.php:111

    (not sure if adding a phpstan-ignore here is too intrusive or if we could work around the php-doc inferred types, but I suppose it's better than to add this to the baseline. Marking as resolved and just a comment if someone else stumbles accross it)

    DefaultTcaSchema.php:711

    (note: here the `count()` is already avoided)
  61. Garvin Hicking commented on patch set 17 10/31/25, 08:43

    ItemProcessingService.php:65

    I wonder if we should take the extra time to convert $items from an array to a DO instead, ensure better type handling and having something like ``` $itemGroup->add(new SelectItem(...)) ``` and a return type of "ItemGroup" then. This would eradicate some instanceof and item checks down the way. And we could auto-convert/migrate the $items array by the procFunc into that structure for the legacy options. If we don't do it now, it'll probably be too annoying to change afterwards (and would be a BC). Thoughts?
  62. Garvin Hicking commented on patch set 17 10/31/25, 08:53

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:66

    ```suggestion :php:`SpecialRelationsProcessor2` will be called before :php:`SpecialRelationsProcessor`. ```
  63. Garvin Hicking commented on patch set 17 10/31/25, 09:09

    Can you check out `typo3/sysext/backend/Classes/Form/FormDataProvider/TcaSiteLanguage.php` at line 68 - probably this check also needs to take `itemsProcessors` into consideration?
  64. Garvin Hicking commented on patch set 17 V +1 10/31/25, 09:16

    OK so this functionally works fine for me at this point, many thanks! Adding a "verified" vote for this state to show my appreciation :-) Some further things though: - Please add `itemsProcessors` to `Build/Sources/TypeScript/backend/code-editor/stream-parser/typoscript.ts` as well, to make it a recognized parsable TypoScript keyword - Maybe we should include one core TCA itemProcFunc and convert it to an itemsProcessor in this patch to showcase conversion? Like maybe the TCA of typo3/sysext/indexed_search/Configuration/TCA/index_config.php - that could be with little impact. - Please add at one showcase/example also in EXT:styleguide
  65. Francois Suter uploaded patch set 18 V +1 11/02/25, 18:52

  66. core-ci voted on patch set 18 V -1 11/02/25, 18:59 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88323

  67. Francois Suter uploaded patch set 19 V -1 11/02/25, 19:36

  68. Francois Suter uploaded patch set 20 11/02/25, 19:40

  69. Francois Suter commented on patch set 20 11/02/25, 19:43

    Reply

    I added "itemsProcessors" to Build/Sources/TypeScript/backend/code-editor/stream-parser/typoscript.ts. I hope I did it correctly, because it's the first time I ever see this file. About showcasing a conversion, I think we should rather not. I discussed with Benni about migrating the usages of "itemsProcFunc" in the core and it would actually be a breaking change, because people who used to override an existing "itemsProcFunc" would not be overriding it anymore. So it would have to be a separate patch (and the deadline to 14.0 being close, it feels like other areas are more important; this migration can be done for the next version since "itemsProcFunc" is not deprecated anyway). But if someone wants to do it, I'm fine. I'm keen on keeping my limited time to work on other things that Benni showed me. I'll add an example to styleguide, though, in the next patch version.

    Reply

    This belongs to site settings, which use a pseudo-TCA based in YAML to then feed to the FormEngine. It follows a different path and is thus not covered by this patch.
    @garvin@hick.ing If I haven't missed anything, I have addressed all your remarks, mostly by changing things, except for a couple of items, for which I have explained the justification. I would be glad if you could in particular check the `SelectItemCollection` class which I introduce, because I'm not sure if that was the right way to do it. I took the `CategoryCollection` class as an example.

    AbstractItemProvider.php:131 Reply

    Done. Injection is missing, coming the next patch version.

    BackendUtility.php:1228 Reply

    Done

    BackendUtility.php:1231 Reply

    Done

    BackendUtility.php:1315 Reply

    Done

    BackendUtility.php:1319 Reply

    Done

    DataHandler.php:2031 Reply

    Done

    DataHandler.php:2035 Reply

    Done

    DataHandler.php:2126 Reply

    Done

    DataHandler.php:2130 Reply

    Done

    ItemProcessingService.php:65 Reply

    I introduced a SelectItemCollection class.

    ItemProcessingService.php:97 Reply

    Done

    DefaultTcaSchema.php:711 Reply

    Done

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:66 Reply

    Done
  70. core-ci voted on patch set 19 V -1 11/02/25, 19:44 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88324

  71. core-ci voted on patch set 20 V -1 11/02/25, 19:48 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88325

  72. Garvin Hicking commented on patch set 20 11/02/25, 19:56

    Reply

    Understood, thanks for checking!

    Reply

    Totally fine, thanks for elaborating. I understand and agree with this in mind.

    Reply

    I will check this out, thank you (unresolving so that I remember tomorrow). Thanks so much for getting on this, and I'm really sorry my notes were "intimidating" - I wanted to value your contribution and review thoroughly.
  73. Francois Suter commented on patch set 20 11/02/25, 20:27

    @garvin@hick.ing Could you also check with Core CI why it isn't happy? It seems touching the ".ts" file triggered a rebuild of the CSS and something fails, but that's an area about which I don't know anything. Thanks in advance for your help.
  74. Garvin Hicking commented on patch set 20 11/02/25, 20:41

    Reply

    Yes, tried to ping you via slack - please run `Build/Scripts/runTests -s build`. This runs the TypeScript parsers which turns the .ts files into .js files, and once those are in sync, the CI will no longer complain. You commit-amend the changed .js file to your patchset as well.
  75. Francois Suter uploaded patch set 21 V -1 11/03/25, 07:34

  76. core-ci voted on patch set 21 V +1 11/03/25, 07:42 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88348

  77. Garvin Hicking commented on patch set 21 CR +1 V +1 11/03/25, 08:10

    Reply

    Done
    I do think the "overhead" of the SelectItemCollection is useful to have for safe parameter handling, and is a way forward to have all items handled like that (as noted in the todo), and uses the chance we have now when introducing this to not be a breaking change in the future. Current version works fine for me! Thanks a lot!
  78. Simon Schaufelberger voted on patch set 21 CR +1 V +1 11/04/25, 18:21

  79. Chris Müller commented on patch set 21 11/04/25, 18:26

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:29

    ```suggestion A new Page TSconfig option is also available, mirroring the existing option for ```
  80. Garvin Hicking commented on patch set 22 CR +1 V +1 11/04/25, 18:34

    Published edit on patch set 21.
  81. Garvin Hicking commented on patch set 22 CR +1 V +1 11/04/25, 18:34

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:29 Reply

    Fix applied.
  82. core-ci voted on patch set 22 V +1 11/04/25, 18:41 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88477

  83. Nikita Hovratov commented on patch set 22 11/04/25, 19:57

    TcaCheckboxItems.php:53

    `$config` is not a reference. I think this needs to be changed back to the full `$result` path.
  84. Nikita Hovratov uploaded patch set 23 CR +1 V +1 11/04/25, 20:42

  85. Nikita Hovratov commented on patch set 23 11/04/25, 20:42

    TcaCheckboxItems.php:53 Reply

    Done
  86. core-ci voted on patch set 23 V -1 11/04/25, 20:55 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88480

  87. Simon Schaufelberger voted on patch set 23 CR +1 V +1 11/04/25, 20:55

  88. Simon Schaufelberger voted on patch set 23 CR 0 V 0 11/04/25, 20:55

  89. Nikita Hovratov uploaded patch set 24 V -1 11/04/25, 21:06

  90. core-ci voted on patch set 24 V -1 11/04/25, 21:18 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88482

  91. Nikita Hovratov uploaded patch set 25 V -1 11/04/25, 21:19

  92. core-ci voted on patch set 25 V +1 11/04/25, 21:26 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88483

  93. Nikita Hovratov voted on patch set 25 CR +1 V +1 11/04/25, 21:27

  94. Garvin Hicking voted on patch set 25 CR +1 V +1 11/04/25, 21:33

  95. Simon Schaufelberger voted on patch set 25 CR +1 V +1 11/04/25, 21:59

  96. Stefan Bürk commented on patch set 25 CR +1 V +1 11/05/25, 01:30

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:18

    Nitpick, special due to the votings on the change already. I guess it would be good to state the sorting order (`ascending` meaning from lower to upper (-PHP_INT_MAX -> +PHP_INT_MAX) ... or what does the others/documentation verifier think ? Don't make it a blocking ... and maybe we can enhance that afterwards.
  97. Oli Bartsch commented on patch set 25 V -1 11/05/25, 07:41

    Reviewing.
  98. Oli Bartsch commented on patch set 25 11/05/25, 07:43

    Feature-107889-IntroduceTCAOptionItemsProcessors.rst:18 Reply

    I’d personally not add that level of detail — the fact that arrays with numeric keys are iterated in ascending order is basic PHP behavior and quite fundamental programming knowledge. Adding it here would bloat the documentation and set a precedent to explain such language-level concepts elsewhere.
  99. Oli Bartsch uploaded patch set 26 CR +1 V +1 11/05/25, 10:05

  100. core-ci voted on patch set 26 V +1 11/05/25, 10:12 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88500

  101. Garvin Hicking commented on patch set 26 11/05/25, 10:22

    TcaSelectTreeItems.php:128

    Can we throw an exception if people accidentally use "itemsProcessors" instead of "itemProcessors"? I believe (untested) using an unknown key just silently fails at this time. Doing a search+replace here for custom code might easily lead people into thinking to replace "itemsProcFunc" with "itemsProcessors". I think that was the initial reasoning when creating the new naming.
  102. Oli Bartsch uploaded patch set 27 V +1 11/05/25, 10:24

  103. Benni Mack commented on patch set 27 11/05/25, 10:26

    TcaSelectTreeItems.php:128 Reply

    Originally, it should be an "Item Processor"
  104. Oli Bartsch commented on patch set 27 11/05/25, 10:28

    TcaSelectTreeItems.php:128 Reply

    I don't think this is necessary. The key was never introduced or documented. If someone uses it by mistake, nothing will happen, which should naturally lead them to check the documentation again. I’m against adding exceptions for arbitrary or misspelled keys that were never part of the API in the first place.
  105. Oli Bartsch voted on patch set 27 CR +1 V +1 11/05/25, 10:29

  106. core-ci voted on patch set 27 V +1 11/05/25, 10:32 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88502

  107. Nikita Hovratov commented on patch set 27 11/05/25, 10:56

    phpstan-baseline.neon:1198

    Can we avoid this?
  108. Oli Bartsch commented on patch set 27 11/05/25, 11:13

    phpstan-baseline.neon:1198 Reply

    The interface does not allow typed parameters. Therefore we can only add them as doc types. This therefore needs those entries. See phpstan.neon for a description about this.
  109. Nikita Hovratov commented on patch set 27 11/05/25, 11:24

    phpstan-baseline.neon:1198 Reply

    ok, thanks.
  110. Lina Wolf commented on patch set 27 11/05/25, 11:30

    TcaSelectTreeItems.php:128 Reply

    I agress with Oli, we never do that for unexpected keys in TCA. Also it is possible for third party extensions to introduce and interpret additional keys so throwing an exception for an unexpected key could then break code of third party extensions.
  111. Oli Bartsch commented on patch set 27 11/05/25, 11:40

    TcaSelectTreeItems.php:128 Reply

    Thanks Lina, I even forgot about this option. So I think we can resolve this.
  112. Lina Wolf commented on patch set 27 CR +1 V +1 11/05/25, 11:45

    Tested with the styleguide extension and an additional field with select mutltiple side by side. Code looks good to me
  113. Nikita Hovratov commented on patch set 27 11/05/25, 12:23

    Reply

    Done
  114. Nikita Hovratov commented on patch set 27 11/05/25, 12:34

    AbstractItemProvider.php:140

    The inner array_map can be removed here, as it is already handled in SelectItemCollection.
  115. Nikita Hovratov uploaded patch set 28 CR +1 V +1 11/05/25, 12:40

  116. Lina Wolf voted on patch set 28 CR +1 V +1 11/05/25, 12:43

  117. core-ci voted on patch set 28 V -1 11/05/25, 12:48 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88516

  118. Nikita Hovratov uploaded patch set 29 CR +1 V +1 11/05/25, 12:56

  119. Nikita Hovratov uploaded patch set 30 11/05/25, 13:01

  120. Nikita Hovratov commented on patch set 30 11/05/25, 13:02

    AbstractItemProvider.php:140 Reply

    Done
  121. core-ci voted on patch set 29 V -1 11/05/25, 13:04 Core CI is not happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88517

  122. core-ci voted on patch set 30 V +1 11/05/25, 13:09 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88518

  123. Nikita Hovratov voted on patch set 30 CR +1 V +1 11/05/25, 13:09

  124. Nikita Hovratov commented on patch set 30 11/05/25, 13:13

    SelectItemCollection.php:41

    Maybe we should remove the elseif and expect the $data array to be either SelectItem or array. Otherwise throw an exception. Right now an invalid item would just be ignored. This is less strict than it was before.
  125. Oli Bartsch commented on patch set 30 11/05/25, 14:42

    SelectItemCollection.php:41 Reply

    But then we would still need to check for array to throw the exception in an else block?
  126. Nikita Hovratov commented on patch set 30 11/05/25, 14:54

    SelectItemCollection.php:41 Reply

    yep, so 2 ifs with early return and otherwise exception. No else/elseifs.
  1. Oli Bartsch commented on patch set 30 11/05/25, 14:55

    SelectItemCollection.php:41 Reply

    Oke for me. Can you update the patch?
  2. Simon Schaufelberger wrote 11/05/25, 15:11

    Removed reviewer Simon Schaufelberger.
  3. Garvin Hicking voted on patch set 30 CR +1 V +1 11/05/25, 17:27

  4. Nikita Hovratov uploaded patch set 31 CR +1 V +1 11/06/25, 08:25

  5. Nikita Hovratov commented on patch set 31 11/06/25, 08:25

    SelectItemCollection.php:41 Reply

    Done
  6. core-ci voted on patch set 31 V +1 11/06/25, 08:34 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88555

  7. Nikita Hovratov voted on patch set 31 CR +1 V +1 11/06/25, 08:35

  8. Oli Bartsch commented on patch set 31 V -1 11/06/25, 10:19

    Talked to Garvin yesterday. I think we should rework the naming to acutally be consistent now. We are introducing Processor*s* for tca item*s*. Therefore, we should name everything this way. The option should be `ìtemsProcessors`, the interface `ItemsProcessorIntrface`, and so on..
  9. Oli Bartsch uploaded patch set 32 CR +1 V +1 11/07/25, 07:56

  10. core-ci voted on patch set 32 V +1 11/07/25, 08:04 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88624

  11. Oli Bartsch commented on patch set 32 CR +1 V +1 11/07/25, 08:05

    Reply

    Done
  12. Lina Wolf commented on patch set 32 CR +1 V +1 11/07/25, 12:27

    Tested with the styleguide extension
  13. Oli Bartsch uploaded patch set 33 CR +1 V +1 11/07/25, 16:50

  14. Oli Bartsch uploaded patch set 34 11/07/25, 16:52

  15. core-ci voted on patch set 33 V +1 11/07/25, 16:59 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88657

  16. core-ci voted on patch set 34 V +1 11/07/25, 17:00 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88658

  17. Oli Bartsch voted on patch set 34 CR +1 V +1 11/07/25, 17:27

  18. Garvin Hicking commented on patch set 34 CR -1 11/07/25, 17:34

    reviewing, some more changes needed
  19. Garvin Hicking uploaded patch set 35 CR +1 V +1 11/07/25, 17:47

  20. core-ci voted on patch set 35 V +1 11/07/25, 17:57 Core CI is happy: https://git.typo3.org/typo3/CI/cms/-/pipelines/88665

  21. Garvin Hicking commented on patch set 35 CR +1 V +1 11/07/25, 18:07

    One drinks for "naming things"
  22. Nikita Hovratov voted on patch set 35 CR +1 V +1 11/07/25, 23:23

  23. Oli Bartsch voted on patch set 35 CR +2 V +2 11/09/25, 16:11

  24. Oli Bartsch merged patch set 36 11/09/25, 16:11

  25. Francois Suter commented on patch set 36 11/09/25, 16:27

    Wow, what a bumpy road we traveled with this patch! Thanks for all the care you have given it.