#92937Translation wizard of tt_content having l10n_source = 0 shows too many items
Description
Using the translation wizard in the backend shows too many content elements in selection view
if some translations on the page have l10n_source set to 0 whereas the l18n_parent is set to a correct value.
I do not know, if having different values in l18n_parent and l10n_source should be allowed or not. Researching points to the LocalizationRepository having a wrong while condition.
Reviews 2
Related issues 3
- Bug #94397Translation Button is shown in connected mode if l10n_source differs from l18n_parent
- Bug #97094Transl.Orig field for content element creates confusing translation status
- Bug #111423LocalizationRepository::getRecordsToCopyDatabaseResult() stops at the first translation with l10n_source = 0
Files 6
History 1130
Show 25 earlier messages
-
Gerrit Code Review pushed patch set 1 for master to change 66905 status
NewUnder Review <span data-processor="relativetime" data-value="2020-11-26T08:33:10+00:00" > 11/26/20, 08:33 </span> -
Gerrit Code Review pushed patch set 2 for master to change 66905 <span data-processor="relativetime" data-value="2020-12-07T08:40:45+00:00" > 12/07/20, 08:40 </span>
-
Christian Kuhn commented <span data-processor="relativetime" data-value="2021-03-25T14:06:50+00:00" > 03/25/21, 14:06 </span>
-
Riccardo De Contardi changed Related to #94397 <span data-processor="relativetime" data-value="2021-08-08T08:05:41+00:00" > 08/08/21, 08:05 </span>
-
Gerrit Code Review pushed patch set 3 for master to change 66905 <span data-processor="relativetime" data-value="2021-11-03T17:07:40+00:00" > 11/03/21, 17:07 </span>
-
Gerrit Code Review pushed patch set 4 for master to change 66905 <span data-processor="relativetime" data-value="2021-11-12T16:21:15+00:00" > 11/12/21, 16:21 </span>
-
Gerrit Code Review pushed patch set 5 for master to change 66905 <span data-processor="relativetime" data-value="2021-11-18T10:47:11+00:00" > 11/18/21, 10:47 </span>
-
Gerrit Code Review pushed patch set 6 for main to change 66905 <span data-processor="relativetime" data-value="2021-11-29T14:08:10+00:00" > 11/29/21, 14:08 </span>
-
Gerrit Code Review pushed patch set 7 for main to change 66905 <span data-processor="relativetime" data-value="2022-06-25T21:36:18+00:00" > 06/25/22, 21:36 </span>
-
Gerrit Code Review pushed patch set 8 for main to change 66905 <span data-processor="relativetime" data-value="2022-06-25T22:02:53+00:00" > 06/25/22, 22:02 </span>
-
Gerrit Code Review pushed patch set 9 for main to change 66905 <span data-processor="relativetime" data-value="2022-07-05T11:46:40+00:00" > 07/05/22, 11:46 </span>
-
Gerrit Code Review pushed patch set 10 for main to change 66905 <span data-processor="relativetime" data-value="2022-07-05T11:47:42+00:00" > 07/05/22, 11:47 </span>
-
Andreas Kießling commented <span data-processor="relativetime" data-value="2022-10-12T15:09:46+00:00" > 10/12/22, 15:09 </span>
> l18n_parent = id AND l10n_source = other-id -> The element is in connected mode but has been derived from another element that is a localization of the default language element.
I seem to have that in a project: a page translation is in connected mode, but when i hit Translate again after new elements were added to default, i see a lot of elements that are already translated. When you just keep pressing next in the dialogue, you end up with elements that now have two translations.
-
Gerrit Code Review pushed patch set 11 for main to change 66905 <span data-processor="relativetime" data-value="2023-03-23T11:05:54+00:00" > 03/23/23, 11:05 </span>
-
Gerrit Code Review pushed patch set 12 for main to change 66905 <span data-processor="relativetime" data-value="2023-03-23T11:07:13+00:00" > 03/23/23, 11:07 </span>
-
Gerrit Code Review pushed patch set 13 for main to change 66905 <span data-processor="relativetime" data-value="2023-04-05T07:09:41+00:00" > 04/05/23, 07:09 </span>
-
Gerrit Code Review pushed patch set 14 for main to change 66905 <span data-processor="relativetime" data-value="2023-05-03T09:42:40+00:00" > 05/03/23, 09:42 </span>
-
Gerrit Code Review pushed patch set 15 for main to change 66905 <span data-processor="relativetime" data-value="2023-07-04T20:09:34+00:00" > 07/04/23, 20:09 </span>
-
Markus Klein commented <span data-processor="relativetime" data-value="2023-11-24T08:02:20+00:00" > 11/24/23, 08:02 </span>
> l18n_parent = id AND l10n_source = 0 -> shouldn’t happen / db inconsistency.
I disagree on that one. An editor might change the parent via normal BE editing, but is not able to change the source. This is a valid action in the editor’s POV.
-
Jo Hasenau commented <span data-processor="relativetime" data-value="2024-02-07T08:56:54+00:00" > 02/07/24, 08:56 </span>
Markus Klein wrote in #note-19:
> > l18n_parent = id AND l10n_source = 0 -> shouldn’t happen / db inconsistency.
>
> I disagree on that one. An editor might change the parent via normal BE editing, but is not able to change the source. This is a valid action in the editor’s POV.And this has to be changed, since it would be a DB inconsistency. Changing the value of l18n_parent must push that value to l10n_source too, if l10n_source was empty before. That’s exactly one of those issues we have to tackle with the upcoming TransFusion extension.
-
Jo Hasenau commented <span data-processor="relativetime" data-value="2024-02-07T09:39:19+00:00" > 02/07/24, 09:39 </span>
Andreas Kiessling wrote in #note-13:
> > l18n_parent = id AND l10n_source = other-id -> The element is in connected mode but has been derived from another element that is a localization of the default language element.
>
> I seem to have that in a project: a page translation is in connected mode, but when i hit Translate again after new elements were added to default, i see a lot of elements that are already translated. When you just keep pressing next in the dialogue, you end up with elements that now have two translations.You should not be able to hit “translate” again, but it seems that related methods don’t properly check the values of those fields.
i.e. BackendUtility::getLocalizedRecord() checks for$tcaCtrl['translationSource'] ?? $tcaCtrl['transOrigPointerField']which is the wrong order, since transOrigPointerField must have a higher priority.
-
Gerrit Code Review pushed patch set 16 for main to change 66905 <span data-processor="relativetime" data-value="2024-10-18T14:08:35+00:00" > 10/18/24, 14:08 </span>
-
André Buchmann commented and changed attached clipboard-202503302226-py7go.png attached clipboard-202503302226-rnrmm.png attached clipboard-202503302227-aa1z9.png attached clipboard-202503302227-laifg.png attached clipboard-202503302257-z7r3k.png attached clipboard-202503302259-2auor.png <span data-processor="relativetime" data-value="2025-03-30T21:14:02+00:00" > 03/30/25, 21:14 </span>
>> l18n_parent = id AND l10n_source = 0 -> shouldn’t happen / db inconsistency.
>
> I disagree on that one. An editor might change the parent via normal BE editing, but is not able to change the source. This is a valid action in the editor’s POV.I totally agree with @liayn . I don’t see it as a database inconsistency, as long as we support unconnected content elements (free mode) :-)
Behaviour still reproducable without patch in dev-main (14.0). I’ve created two elements in the default language and one element in a translation (free mode). Then I’ve translated one element of the default language via the “New”-select in the content-element editor.
Initial setup in List!
Initial setup in “Page”. The translation button shows up with both elements:
clipboard-202503302226-rnrmm.pngThe reconnected element with empty l10n_source is translated again:
clipboard-202503302227-aa1z9.pngResult:
clipboard-202503302227-laifg.png—-
With the patch, the translation button still shows up, but the wizard is empty:
clipboard-202503302257-z7r3k.pngClick on next, it crashes -> Fix needs to be improved:
clipboard-202503302259-2auor.png -
Gerrit Code Review pushed patch set 17 for main to change 66905 <span data-processor="relativetime" data-value="2025-04-12T12:25:13+00:00" > 04/12/25, 12:25 </span>
-
Gerrit Code Review pushed patch set 18 for main to change 66905 <span data-processor="relativetime" data-value="2026-04-22T06:50:45+00:00" > 04/22/26, 06:50 </span>
-
Alexander Stehlik changed Duplicated by #105343 <span data-processor="relativetime" data-value="2026-06-18T22:59:49+00:00" > 06/18/26, 22:59 </span>
-
Gerrit Code Review pushed patch set 19 for main to change 66905 <span data-processor="relativetime" data-value="2026-06-18T23:21:00+00:00" > 06/18/26, 23:21 </span>
-
Peter Kraume commented and changed TYPO3 Version
1114 <span data-processor="relativetime" data-value="2026-10-16T08:45:16+00:00" > 10/16/26, 08:45 </span>I recently had this issue in TYPO3 v14 where the wizard suggested records for translation which were already translated. In the next step of the wizard, this caused a TYPO3 error.
Patch set 19 worked well to fix the problem. -
Benni Mack changed Related to #97094 <span data-processor="relativetime" data-value="2026-11-11T10:59:11+00:00" > 11/11/26, 10:59 </span>
-
Oliver Bartsch changed Related to #111423 <span data-processor="relativetime" data-value="2026-11-16T16:28:22+00:00" > 11/16/26, 16:28 </span>
Hey. I did not look at the pending patch in detail (yet), but I can elaborate on l18n_parent vs l10n_source a bit:
- l10n_source (tca ctrl translationSource): This is an indicator from which other content element this localization has been derived from.
- It is 0 for sys_language_uid=0 content elements.
- It is 0 for elements newly created in a non default language in free mode.
- It is the id of a default-language content element if the element has been derived from that default language element. both in free and connected mode.
- It is the id of a non-default-language record if this content element locolization has been derived from another non-default-language content element (localization-chains: de is a localization of default-language en, and fr is a localization of de)
This allows the following combinations for any given non-default sys_language_uid > 0 element:Note there are some additional cases / combinations in workspaces …