Quantcast
Channel: Issues for Drupal core
Viewing all 296535 articles
Browse latest View live

SubProcess migrate process plugin does not play nicely with XML-based source data

$
0
0

Problem/Motivation

Currently the SubProcess migrate process plugin has some issues, when dealing with XMLs as source data (in combination with migrate_plus).

Example XML structure:

<?xml version="1.0" encoding="UTF-8" ?>
<article>
  <elements>
    <id>element.text.001</id>
  </elements>
  <elements>
    <id>element.text.002</id>
  </elements>
</article>

The corresponding migration fetches all elements nodes via the following XPath statement: /article/elements.

As a result, I get an array of SimpleXMLElement objects - one for each elements XML node.

I am currently trying to pass this array of SimpleXMLElement objects to the SubProcess plugin. In its transform method, there is the following line of code that creates the new data sub-row (where $new_value is the currently subprocessed SimpleXMLElement of the passed in array):

$new_row = new Row($new_value + $source);

This code unfortunately errors out, due to the fact that you can't merge a SimpleXMLElement object with an array like this.

Proposed resolution

  • Ensure that the $new_value variable is an array before trying to merge it with $source

Remaining tasks

  • Create a patch

User interface changes

n/a

API changes

n/a

Data model changes

n/a


Remove QuickEdit from the Umami demo profile

$
0
0

Problem/Motivation

Same as #3227036: Remove QuickEdit from the Standard profile, but I filed this separately because we need the Umami maintainers to let us know if they would prefer to have QuickEdit turned off by default in the Umami Demo in 9.3, or leave that for when we actually remove it.

Proposed resolution

Discuss whether to remove QuickEdit from the Umami demo profile.

Remaining tasks

TBD

User interface changes

N/A aside from title.

API changes

TBD

Data model changes

TBD

Release notes snippet

TBD

[Investigation] Try removing quickedit from core

$
0
0

Problem/Motivation

#3222947: [policy, no patch] Decide whether to move Quick Edit to contrib Is leaning towards "yes, let's remove it."

This is an issue to do a trial run of removing it, and see what breaks.

Steps to reproduce

n/a

Proposed resolution

Try to remove quickedit and see what happens.

The following API changes need to be made:

  1. QuickEdit annotations on core widgets need to be added via an alter hook.
  2. QuickEdit stylesheets for core themes need to be added via an alter hook.

Remaining tasks

  1. Address tests listed in #6 - see #3227161: Stop using quickedit data attributes as selectors in tests that are not testing quickedit for most of these
  2. Address needed API changes (listed above).
  3. Address any other test failures.
  4. Address other references in #7.

[plan] Move QuickEdit into contrib and remove it from core

$
0
0

Problem/Motivation

In #3222947: [policy, no patch] Decide whether to move Quick Edit to contrib Dries, the committer team, and other stakeholders have signed off on removing the QuickEdit module from core in Drupal 10.

This meta issue is filed as critical because it also unblocks a required Drupal 10 issue for removing our dependency on Backbone.

Proposed resolution

Do the work necessary to move the module into contrib, deprecate it in 9.3, and remove it from core in 10.0.

Remaining tasks

Done

User interface changes

No more QuickEdit.

API changes

TBD, possibly in child issues

Data model changes

TBD, possible in child issues

Release notes snippet

Needed.

Build out + style the user profile page

$
0
0

When first logging in, this is what you get; basically a username and a "member since" date.

Sparse user page

It'd be great if instead there were a few fields here, even if they were auto-generated, and esp. a user picture (even if a placeholder one) to give people some orientation as to where they are and what's possible.

Remove usage of spaceless tag in Twig templates

$
0
0

Problem/Motivation

spaceless tag has been removed in Twig 3. It is not used in Drupal 9 anymore.

Proposed resolution

Use spaceless filter instead in Drupal 8 too to help new sites become compatible with Twig 3 / Drupal 10.
https://twig.symfony.com/doc/3.x/filters/spaceless.html

Before:

{% spaceless %}

{% endspaceless %}

After:

{% apply spaceless %}

{% endapply %}

Remaining tasks

Create a patch.

User interface changes

None.

Olivero: Wide search form should close on blur

$
0
0

Similar to #3191692: Have secondary menus close on blur, the search form dropdown should close on when focus is removed from the search form.

This can problem reduce or eliminate issues such as #3209871: Olivero's: The header is showing odd behavior when the search bar is opened.

Testing instructions

  1. Load the patch or visit the tugboat link: https://3226785-wide-search-blur-ndgukawgpielewnzaznw3cxtmoia97gi.tugboa...
  2. Load the page at wide widths
  3. Open the search form (can be either click or via keyboard)
  4. Tab outside of the search form - verify the form closes after you tab away.
  5. Shift-tab to before the search form. Verify the form will close after either the form or the form's disclosure button loses focus.

NodeController::revisionOverview is uncacheable

$
0
0

Problem/Motivation

NodeController calls $this->renderer->addCacheableDependency with an object that doesn't implement CacheableDependencyInterface, this results in an uncacheable page for DPC.

Steps to reproduce

Visit the node overview page
Notice that the page is uncacheable in dynamic page cache

Proposed resolution

username is a render array, use the correct API for merging cacheability metadata

Remaining tasks

User interface changes

API changes

Data model changes

Release notes snippet


Improve StringItem::generateSampleValue()

$
0
0

Problem/Motivation

Sample entity value generation generates long strings for text fields. This can cause the layout to break, as noted in #3016507: Break long text strings in layout edit

Proposed resolution

Rather than use \Drupal\Component\Utility\Random::word, which often generates very long words, use \Drupal\Component\Utility\Random::paragraphs, which uses words from a selection of shorter words.

Remaining tasks

Write a patch.

User interface changes

None-ish.

API changes

None

Data model changes

None

Auto fix Illegal choice n in Weight element.

$
0
0

Problem/Motivation

Less errors in watchdog

Steps to reproduce

Deletion of an element

This most likely isn't the correct place but I'm putting it in here so our fix is documented.

Order of terms with same weight messed up after saving

$
0
0

path: #overlay=admin/structure/taxonomy/tagname
- click 'Show row weights' after creating some term ( all weights are '0' )
- set the weight of one term
- click save

now the items are not sort as they are supposed to.

Using the normal 'drag to re-order' works fine.

tested in firefox 3.6.10 and chrome 8.0.552.0 dev

issue occurs only when the weights of multiple terms are '0'

Taxonomy term ordering on node view & edit not consistent with taxonomy weight

$
0
0

Taxonomy is listed according to field delta and not the weight assigned in taxonomy admin screens. This is inconsistent and (I believe) is wrong for a number of reasons. These are:

  • Allowing taxonomy admins to order tags in the administration theme implies this will have an effect. For nodes with vocab fields using the autocomplete term widget it never does – for the other form elements available (eg. select list) it does but only after the node is re-saved.
  • The order in which terms are listed on edit and view is inconsistent – if the autocomplete term is used terms are listed in the order entered by the node creator/editor – for the other options they are list in order of weight at the time of content creation.
  • Other CMS's free tagging systems obey alphabetical order – I believe this is natural for free tagging and we are now making this impossible to achieve with vanilla Drupal where as it was out-of-the-box with D6.

In order to recreate this issue with a vanilla drupal standard install:

  1. Create an article type node with "mongoose, zebra, aardvark" in the free tagging field and save
  2. View the node... the order is "aardvark, zebra, mongoose" (i.e. the reverse order to which the terms are entered)
  3. Go to the taxonomy list screen for the default vocab you will see that the terms are in alphabetical order
  4. Change it to any other order (except the order entered in the field) eg. Zebra, Mongoose, Aardvark
  5. Go back to the created article - you will see it has no effect

UI missing for translatable data type unless form_element_class is set

$
0
0

Problem/Motivation

For some data types such as string, boolean and uri, there is no UI to translate them. Even when they are marked as translatable, for example:

url:
  type: uri
  label: Link URL
  translatable: true

The config_translation module exposes a UI to translate configuration of various data types by assigning the form_element_class that best represents a data type. The module file contains a comment describing a fallback behavior if this property is not set:

Enhance the text and date type definitions with classes to generate proper form elements in ConfigTranslationFormBase. Other translatable types will appear as a one line textfield.
-- in config_translation_config_schema_info_alter

However, when this class name is not defined, no UI is offered for config marked as translatable, due to the isset($definition['form_element_class'] check in Drupal\config_translation\Form\ConfigTranslationFormBase::createFormElement(). Thus the value cannot be translated at all, which is at odds with the comment in the module file as well as the expectation of the user.

Proposed resolution

Do as described in hook_config_schema_info_alter and provide single line text fields for translatable elements, even if form_element_class is not set.

Remaining tasks

Write tests & review.

User interface changes

Single line text fields are offered for any configuration type definition that are marked as translatable: true but without a dedicated form_element_class.

Data model changes

Existing configuration schema definitions should not require changes. However, if the translatable flag has been set on inappropriate elements (e.g. config_object containers), this will lead to errors due to there not being a value.

[META] Taxonomy admin usability Improvements

$
0
0

Posted by xjm

There are a number of existing issues on ways to improve the D8 taxonomy UI/UX, many of which are related, overlapping, or contradictory. Here's a list:

Term properties and fields

Vocabulary properties

Terminology

Help text

...Okay, wow, that was more than I thought.

I think we should probably get some general direction on this: decide which of the above should be won't fix, which can have a shared solution, etc.

drupal/core is implicitly allowed by scaffold

$
0
0

Problem/Motivation

The README for the Composer Scaffold plugin uses drupal/core as an example allowed-package. This is a poor example, because drupal/core is implicitly allowed and explicitly listing it has no effect. Using drupal/core as an example allowed-package misleads people into thinking that it has to be explicitly listed.

Proposed resolution

Either clearly state in the examples that drupal/core is implicitly allowed, or choose a different example package


Fix copy-paste mistake in code comment

$
0
0

Problem/Motivation

FormSubscriber calls itself FormValidator in describing its constructor statement. (And when you lie to grep, you confuse developers who are desperately looking for a way to override a tiny piece of FormValidator!)

Olivero doesn't have inline links css classes which are available in Bartik

$
0
0

My module quote don't work with Olivero because Olivero don't have inline links css classes (bartik have, see screenshots)

Empty breadcrumb at node/add and node/add/{content_type} when frontpage view is enabled

$
0
0

Problem/Motivation

The fix in #2716019: View titles in breadcrumb and metatag title don't get properly translated introduced a small bug that creates an empty breadcrumb menu item because the frontpage view has no title.

The result:
breadcrumb error showing Home, an empty space, Add content

Steps to reproduce

1. Create a Basic Page with the title "Home" and path /home
2. Set the default front page to /home
3. Visit node/add/article

Proposed resolution

Current proposal

Add a breadcrumb builder that applies to node/add and node/add/{content_type} that outputs the respective breadcrumbs:

  • For node/add: Home - Administration - Content
  • For node/add/{content type}: Home - Administration - Content - Add content

The breadcrumb builder will be marked as deprecated for removal in 10.x, with a follow-up issue being filed against 10.x to move relevant node/add paths to sit under admin/content.

Original proposal

Add a default title to the view, suggestion 'Promoted content' as in essence that's what it is
Add a test for this displaying in breadcrumbs per the steps to reproduce
A UX team review decided that this is an unacceptable workaround, as it has unintended consequences.

Workarounds

If this is impacting your site, there are at least two ways a site administrator can work around the problem:

  • Set a title for the Frontpage view.
  • Disable the Frontpage view.

Remaining tasks

User interface changes

API changes

Data model changes

Release notes snippet

Meeting of the Bug Smash Initiative 2021-08-03

$
0
0

Agenda items

Standing items

  • Who is here
  • Any other suggested topics
  • Wins / Thanks / Blockers
  • Statistics and functionality update. X new bugs created in last two weeks. Y closed by the initiative.
  • The bug I would most like to see fixed
  • Update on target items from last meeting
  • What should be the initiative focus for the next 2 weeks?

Statistics

Statistics graph

ToDos

anyone from US/NZL - send a reminder about Thu/Friday triage session this week.

0️⃣ Who is here today? Comment in the thread to introduce yourself. We’ll keep the meeting open for 24 hours to allow for all timezones.

Indrajith KBHi.. Indrajith From India:wave:
Griffyn HeelsGriff, Aus :wave:
mohit_agheraMohit, India :wave:
Suresh PrabhuSuresh, India
larowlanLee, AU
quietoneVicki, NZ
jibranJibran, AU.
kimb0Kim, AU
SpokjeFrank, NL
mansoor20Mansoor, India :wave:
Paulo Henrique Cota StarlingPaulo, from Belo Horizonte :flag-br:
catchNat, UK
dwwDerek, US

0️⃣ .5️⃣ Last helpful article/video you have read/watched on the internet?

Griffyn HeelsI've been working on a Python Course recently and have found this site invaluable: https://www.programiz.com/python-programming
larowlanhttps://www.youtube.com/watch?v=dsTXcSeAZq8&pp=sAQA
quietonehttps://www.thelancet.com/journals/landia/article/PIIS2213-8587%2821%290...

1️⃣ Do you have suggested topics you are looking to discuss? Post in this thread and we’ll open threads for them as appropriate.

quietoneApprove the previous minutes, checking the credit as well. #3222714: Meeting of the Bug Smash Initiative 2021-07-20
jibranWe have weekly group triage session scheduled and we are are not utilizing those slots. Given that @Kristen Pol (she/her) posted this last week https://drupal.slack.com/archives/C014QES6HSQ/p1627674236002000, how can we be more active during those slots.
Kristen Pol (she/her):slightly_smiling_face:

2️⃣.1️⃣ Wins - what wins have you had since the last meeting? Share your successes!

mohit_aghera#1822440: "Content access" filter should check for node_grants implementations before adding node access grant queries
Indrajith KB#2886816: When you remove the "Breadcrumb" block in Seven, there's no top padding for the main content
kimb0These bugs are resolved as duplicate of an issue I have been working on. :tada:#1048114: file_copy(), $destination cannot be NULL
SpokjeNot (really) related to BugSmash, but I've been chipping away at Coding Standards and making (test-)code more readable with stuff like this: #3221312: ->willReturn(...) would make more sense here Bonus not (really) related to BugSmash: Being back in the working saddle again on actual websites I'm (finally) back at posting patches for Contrib Modules. Stuff learned in here make them whizz through to RBTC/Fixed way more faster compared to times I wasn't involved in BugSmash. So there's a hidden shoot-out with Thank You All! in there somewhere :)

2️⃣.2️⃣ Thanks - a space for callouts to other contributors who’ve helped you

kimb0Happy birthday for tomorrow @jibran :birthday:
jibranThank you! Big 35!
kimb0:dancerwookie:
dwwThanks @kimb0 for working on the big gnarly file system patches!  I hope to have time to get back into the saddle with those again in the coming weeks...
kimb0Legend! :muscle:  Thanks!
dww(although I guess they're not exactly bugs...) :wink:  But still, thanks!

2️⃣.3️⃣ Blockers - what’s getting in the way of your issues / goals

larowlanworking on contrib
quietoneSooo many bugs ..
SpokjeWorking in general
dwwLife is tricky right now.  In the process of moving, which is taking up a lot of time/energy. Family stuff.  Health stuff.  Since I'm already at reduced capacity for $day_job, it's been hard to eek out upstream contribution time... I'm barely staying on top of my "real" commitments as it is.
larowlanknow that feeling
larowlantake care of yourself

3️⃣ Core bug statistics - current state of open bugs

quietoneToday: 2021-08-03

From: 2021-07-20 to 2021-08-03

All bugs
Open: 48, including 12 fixed
Closed: 83
Age Count
00 25
01 08
02 03
03 04
04 04
05 07
06 06
07 09
08 08
09 01
10 04
11 01
12 03

Status Count
closed (duplicate) 25
closed (cannot reproduce) 6
closed (outdated) 19
closed (fixed) 19
closed (works as designed) 14

Approximately 378 yr reduction in total number of years of all open bugs.

Only bugs tagged Bug Smash Initiative
Open: 6, including 4 fixed
Closed: 48
Age Count
00 08
01 02
02 02
03 03
04 03
05 07
06 04
07 06
08 05
09 01
10 03
11 01
12 03

Status Count
closed (cannot reproduce) 4
closed (outdated) 15
closed (fixed) 8
closed (duplicate) 13
closed (works as designed) 8

Approximately 278 yr reduction in total number of years of all open bugs.

mohit_agheraThanks @quietoneJust curious, how do you create these reports?
quietoneIt is another script in https://gitlab.com/Lendude/bug-smash-initiative but I haven't pushed it a long while.
larowlanThanks for these, another good fortnight

4️⃣ Nominate your pet issue for a bug-smash target for the next fortnight

quietoneFix the 2038 year problem,#3215062: Update hook_schema for Y2038
Paulo Henrique Cota Starling#3186834: Sorting nodes by title aren't returned in correct order.

5️⃣ Help me - Is there anything you could use a :hand:

dwwIt'd be nice to know if https://www.drupal.org/project/drupal/issues/139015 is viable for a D7 commit, or if I should close it.  That's my oldest open bug report. ;)
jibranMaybe post in #contribute and ping D7 maintainers.
dwwDone, thanks for the suggestion, @jibran!

6️⃣  @quietone Approve the previous minutes, checking the credit as well.

mohit_agheraIssue: #3222714: Meeting of the Bug Smash Initiative 2021-07-20
SpokjeI'm normally the one "moaning" about something but this looks perfect to me
dwwRTBC'ed ;)  Thanks, @quietone!
dwwI was expecting to find the sample meeting agenda at #3153713: Bug Smash Initiative meetings - next: 2021-08-17 but lo, that summary doesn't contain our meeting agenda template.  Do we always just cut/paste from the previous meeting's "Standing items" section?  We should add this as a standing item so we always remember to open a thread for it so @quietone doesn't have to be there to remind us. :smile:

7️⃣ @jibran: How can we effectively use the triage session timeslots.

mohit_agherahttps://drupal.slack.com/archives/C014QES6HSQ/p1627674236002000
larowlangood question
quietoneSome thoughts and ideas that come to mind, not in any order.Advertise before hand a set of related issues that will be discussed. This may help folks to prepare for that topic.I lean towards a video call for these but not necessary.Have one person lead the triage on an issue and usually the same person to write the comment. So that while they are writing the comment someone else can lead the next issue.Then the facilitator shouldn't be leading any triage, they need to keep the meeting flowing.
dwwPoint 1 there would also help by simply letting #bugsmash know there's going to be a triage session at all. ;)  I keep meaning to join one and never know if/when they're happening.  Like we do with the "hey, next meeting starting in 3 hours" sorts of pings, would be lovely to see a "next triage in 3 hours" or whatever.  If possible.  Is the info at https://www.drupal.org/community-initiatives/bug-smash-initiative/workin... accurate?We have weekly group triage sessions on Google Meet, alternating between two times....
larowlanthey're in the core calendar yeah
jibranIf someone from US/NZL will send a reminder about Thu/Friday one this week then I'm happy to send a reminder for Tues one.

Participants:

quietone, jibran, Kristen Pol, mohit_aghera, Indrajith KB, kim.pepper, Spokje, dww, larowlan, Paulo Henrique Cota Starling

Config export key order is not predictable, use config schema to order keys for maps

$
0
0

Problem/Motivation

Follow-up to #2361539: Config export key order is not predictable for sequences, add orderby property to config schema

As shown in #2350821: Sort views displays by display name and #2350537: Enforce order of display of components in config export shows that even though #2256679: Use config schema to determine which config entity properties to export regardless of whether they are public or protected made use of config schema to export the top level keys, for lower levels we still rely on the data provided as-is. That means that if keys are reordered for some reason, there will be unnecessary diffs. We should sorting maps in the same way order that the mapping is defined in the schema.

In the absence of a solution in core, various approaches have sprung up in contrib aimed at normalizing config for purposes of comparison or merging. These include:

Proposed resolution

Use the order defined in a config schema data types mapping to sort the result configuration array. This makes the entire system work the same was as #2256679: Use config schema to determine which config entity properties to export regardless of whether they are public or protected.

Remaining tasks

Agree. Do it.

User interface changes

None.

API changes

TBD

Data model changes

Maybe some configuration re-ordered - tbd.

Viewing all 296535 articles
Browse latest View live


<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>