Add a 'machine_name' widget for string field types with a UniqueField constraint

Patch #112 (a re-roll of #106) omitted the new files that were being added in #106. (And the interdiff for it, posted in #113, therefore showed those files removed.)

Here's a replacement patch that includes the new files. This is a re-roll of #106, including the suggestion from #111 to use the new callable syntax. I include an interdiff of this vs #106. Drupal 9.5 sites may not support the callable syntax new to PHP 8.1, so I include a patch for D9.5.x too (without interdiff - I've just retained the previous uasort($components, [SortArray::class, 'sortByWeightElement']); line).

I've also included some screenshots of the new widget settings (collapsed+expanded), and the widget itself in action (before+after clicking the 'edit' link). I trust those might be enough screenshots? Since this is new functionality, I presume we don't need screenshots of how it looked before the change? More can be made if needed of course!

The last submitted patch, 114: drupal-10-1x-machine-name-widget-2685749-114.patch, failed testing. View results →

Fixed the test failure which was due to case sensitivity. (To my surprise!)

I also spotted that the widget's isApplicable() relies on the options having been set when adding the constraint. I don't see anywhere documented that requires that; so I've replaced an isset() for an array_key_exists() so that calls like ->addConstraint('UniqueField') will work rather than requiring an empty second parameter ->addConstraint('UniqueField', []). (There are instances in core of other constraints being added without setting the options parameter, such as in \Drupal\Core\Entity\EntityType::__construct().)

The last submitted patch, 116: drupal-10-1x-machine-name-widget-2685749-116.patch, failed testing. View results →

Previous test failure was for a hopefully unrelated Ckeditor5 test. Tests all passed on a retest!

According to the issue summary #2905594: Missing entity validation constraint: don't allow new entities when there is an existing one with the same ID → is still remaining. So my question is should this ticket be postponed on that one? Or is #2905594: Missing entity validation constraint: don't allow new entities when there is an existing one with the same ID → not needed for this to move forward?

That issue talks about using a "duplicate ID" - if that's talking about a field with a unique value constraint, it could actually be solved by the work that has now been done here. All that would be needed (on top of this work) is for the constraint to be added to that field, I believe.

Or it's talking about using the entity ID field itself, rather than a field using a machine name constraint/field/widget. Which would be related, but I don't think blocks this issue at all. It might even still be able to make use of it in the same way - by simply adding the constraint to the appropriate entity ID field.

So I could be wrong, but I'm updating the IS based on that belief that it's a related issue, not a blocker to this one. Correct me if I am indeed wrong!

Thanks for updating that! This one may need more eyes but noticed the change record was written for D8 and this probably will only make it in 10.2. So if that could be updated.

Ok; updated the change record to now target 10.2.0.

10.2.x-dev isn’t an option for the issue metadata here though. I guess we just wait to get it into that or main or whatever becomes available?

So tested by enabling entity_test module and going to entity_test_string_id/structure/entity_test_string_id/form-display
Verifying the ID field can switch between textfield and machine_name.

Verified that regular Text(only) field doesn't get machine_widget.

Not sure how other to test this one. But lets get into committer eyes. May be easier to get into 10.2 early

Thanks to everyone here for moving this 7 year old issue along!

I triaged this and updated the remaining tasks accordingly. Here are the details.

The issue summary states that there are no user interface changes but this is adding a new widget. And in #107, screenshots were asked for. I eventually found them in #144. @james.williams, thanks for making them. It really helps reviewers/committers if the screenshots are in the IS or a link to the comment so they can be found. Also, with a note saying which patch/MR version they were made with. (I just added it)

There were many questions raised in the reviews, that is good to see and read. However, #80 has not been answered. I have added that to the remaining tasks.

#124 says "Not sure how other to test this one. But lets get into committer eyes." Instead, before setting to RTBC ask in #contribute for first. An issue that is Reviewed & tested by the community ("RTBC") → should not have any questions unanswered, it should be ready for commit.

I read the CR and made changes to use plain English. There should be no changes to the content but someone should check that. I am adding a review of the CR to the remaining tasks. The CR mentions the benefits to site builders but site builder' is not selected in the 'Impacts'. Should it be? The CR should also include a screenshot. These two look the most informative,

https://www.drupal.org/files/issues/2023-05-16/widget-settings-collapsed... →
https://www.drupal.org/files/issues/2023-05-16/widget-settings-expanded.png →

The latest patch is failing tests and is on 10.1x. This needs to be on 11.x with passing tests. Setting to NW.

I am updating credit, always a challenge on issues with > 100 comments.

A reminder to all that when making a patch, an interdiff or diff is needed. There are instructions for creating an interdiff → . If you think a diff is not needed, add a comment stating why.

Hello.

You can try using this module https://www.drupal.org/project/machine_name_widget →

The patch in #2685749-116: Add a 'machine_name' widget for string field types with a UniqueField constraint → for Drupal 9.5 is broken.
It adds two unnecessary files
core/b/core/tests/Drupal/FunctionalJavascriptTests/Core/Field/Widget/MachineNameWidgetTest.php
core/b/core/lib/Drupal/Core/Field/Plugin/Field/FieldWidget/MachineNameWidget.php

They do not differ from the ones at the correct path, so I just removed them.

It should look like the one attached.

Re-roll of the patch from 116 for 11.x. Fixing the testPublishWorkspace as well.

Fix failure of custom commands

Thanks @a.dmitriiev! :) Still 1 failing test left.

Could you perhaps use a MR to make reviews easier?

Here is the patch the same as MR. I figured out the failing test: after changing machine name to client side only, the trailing replace char is also removed. I hope now the tests will pass.

Still the same test fails. However, the expected string is exactly the same as widget produces in admin UI when checking the same thing manually.

Looks like the machine name is wrong. When I add a new field anywhere in Drupal, naming it Test value !0-9@, the machine name gets auto-created as field_test_value_0_9_.

Which would be strange, though, as the machine name test says also says Test value !0-9@ translates to test_value_0_9: https://git.drupalcode.org/project/drupal/-/blob/11.x/core/tests/Drupal/... 🧐

leymannx → changed the visibility of the branch 11.x to hidden.

Okay, I reverted the commit. That wasn't it.

Okay, I found a way to output the machine name value from the field and it's a completely different value than the one that's expected: jvfavt8z == test_value_0_9 wich of course is false. Same can be seen in the test artefacts.

It's also the same what I experienced after I put $settings['extension_discovery_scan_tests'] = TRUE; into my settings.php and enabled the entity_test module. I opened URL /entity_test_string_id/add and it always contained some random name in the name field and the machine name field was always pre-filled with this machine name. When I changed the name, the machine name didn't change accordingly. It always stayed the same.

(It also looks really different from all other machine name elements I know in Drupal.)

By the name of Dries I couldn't find where this arbitrary name value was coming from. On every page load it was a different one. I also don't understand why all other asserts succeed and only this one doesn't.

Does anybody here have an idea why when accessing /entity_test_string_id/add the name is always pre-filled with a random value? Where does it come from?

Re-rolled patch for Drupal 10.3, still tests are not fixed.