Javascript errors in entity-form.js when retranslate checkbox does not exist

Don't think this needs a test case as it's more reverting a previous change.

For record the JS use to be before 📌 Refactor (if feasible) use of jquery is function to use vanillaJS Fixed

         translate = $checkbox.is(':checked')
              ? Drupal.t('Flag other translations as outdated')
              : Drupal.t('Do not flag other translations as outdated');

So thinking adding the change here should be fine.

I think that argument (no tests needed) doesn't quite hold up, we did regress on this and didn't notice. But it was caught in contrib tests and is likely not trivial to reproduce in a test, the failing paragraph tests are complex (paragraphs library + multilingual + content moderation + entity browser) and I'm not sure which part exactly is even causing this.

Somehow I missed this when I created 🐛 Recent removal of jQuery.is() results in undefined element errors in entity-embed.js Closed: duplicate . Closed that as a dupe as there is more activity here, but the MR there additionally changes block.js and layout-buider.js as they seem to be at similar risk of attempting to access a checked property on an undefined object.

Reproducing the way contrib tests caught the problem could be a pain, but the error could be brute forced by enabling content translation and using a form alter on the node form to make sure there aren't .js-form-item-translation-translate or .js-form-item-translation-retranslate classes present. While it doesn't recreate the exact circumstances that surfaced this error, form_altering these elements off the page seems like a valid enough use case to create the test coverage for. Core JS should be able to gracefully handle them not being present. I would commit a test with this approach.

It turns out Drupal.behaviors.entityContentDetailsSummaries doesn't work for the translation details element anyway. The javascript is looking for js-form-item-translation-translate but the actual class is js-form-item-content-translation-retranslate. I've tested this in 11.x, 10.0.x, 9.5.x and 9.3.x. So yeah, this probably needs test coverage.

This prevents green tests for paragraphs module on 10.2. If it turns out those test are a bit complicated and the implementation requires more changes, could we do a follow-up to properly fix and test it?

This prevents green tests for paragraphs module on 10.2. If it turns out those test are a bit complicated and the implementation requires more changes, could we do a follow-up to properly fix and test it?

+1 on this. I prefer keeping the scope to fixing the regression that impacts Paragraphs

It looks like the JS looking for a not-present js-form-item-translation-translate was due to a change made in early 2013. Since the underlying issue has been around 10+ years, I think it's safe to offload the (valid!) concern in #10 to a separate issue, and keep this issue's scope lean so Paragraphs maintainers can get their tests working again.

So do we need to update the MR to fix block.js and layout-buider.js as per #7 or should they be separate as well? Opened 🐛 Translation details summary does not update when selecting the (re) translate checkbox Active for fixing the selector and adding test coverage.

So do we need to update the MR to fix block.js and layout-buider.js as per #7 or should they be separate as well?

I think it's fine for that to happen here or in followups, provided those followups are created. I'd commit either approach (though in this case I worked on it a bit so probably best if I'm not the one committing)

Based on #14 seems next step is to RTBC. And seems it's been through a few reviews already.

lauriii → closed merge request !4995

Committed a071abe and pushed to 11.x. Also cherry-picked to 10.2.x. Thanks!

I re-opened 🐛 Recent removal of jQuery.is() results in undefined element errors in entity-embed.js Closed: duplicate for the remaining items in #7.

Thank you @mstrelan! Was supposed to do that myself but I guess I forgot 🤦‍♂️

Automatically closed - issue fixed for 2 weeks with no activity.