[#73104] Add Sprint column to work package table by EinLama · Pull Request #22706 · opf/openproject

EinLama · 2026-04-09T13:17:36Z

Ticket

https://community.openproject.org/wp/73104

What are you trying to accomplish?

Allows you to select the sprint column within the work package table and sort and group it.

Requires the view_sprints permission in a project to be available.

Screenshots

What approach did you choose and why?

There are some notable points here:

Sorting uses the sprint name (as this is what the user sees in the work packages table). After that, sprints of the same name are sorted via start and finish dates.
Added the ability for models to provide an API resource link name (commit 5014766). A resource link is necessary to make grouping work. The previous implementation added a manual mapping for some classes. Agile::Sprint would be converted to agile_sprint, but sprint is correct. So we have to add a mapping. Since the core should not be required to know about the backlogs module, I added a way for models to provide their API resource link name. The given name will be used by the mapping if available.
To make the GROUP BY statement consider the view_sprints permission, a custom join was necessary - I think. The solution is similar to Implementation/62982 project phase column on work package list #18851, although this might be a case of "if I have a hammer, every problem looks like a nail". On the upside, this solution can be easily extended to work with further visibility checks in the future 👍

Merge checklist

Added/updated tests
Added/updated documentation in Lookbook (patterns, previews, etc)
Tested major browsers (Chrome, Firefox, Edge, ...)
Tested that the PDF export still works with sprints listed (and grouped by sprints)

EinLama · 2026-04-16T10:38:42Z

I think we should not merge this until #22740 is in, so that we don't cause any conflicts for the big clean up PR. I will have to remove some lines about the feature flag, then. But that will be easy.

ulferts

@EinLama as discussed, the preliminary review before checking it out using the application.

ulferts · 2026-04-17T08:51:50Z

+      # Since the property is called `sprint`, but the model_name is `agile_sprint`, there would
+      # otherwise be a mismatch. Avoid that:
+      :sprint
+    end


I personally dislike having code I would consider representation (in the API) within the model. I am aware that this is a preference, not a strict rule.

But maybe backlogs could register the mapping on the ResourceLinkGenerator in its engine.rb

Hmm, yes it is indeed a bit odd, especially since our other models don't do it like this. I moved the name mapping into a patch like suggested.

ulferts · 2026-04-17T09:03:01Z

+    create(:user,
+           member_with_permissions: {
+             project => project_permissions,
+             another_project => another_project_permissions


You should add another shared sprint where the user lacks permissions in the sharing project.

Added some test cases for this. Doing so uncovered an issue with the previous query. Fixed that 👍

ulferts · 2026-04-17T09:08:48Z

+
+    def projects_with_view_sprints_permissions
+      Project.allowed_to(User.current, :view_sprints).select(:id)
+    end


Currently, sprints shared with projects are not all visible. Only if the user has the :view_sprints permission in the sharing project will the sprint be visible.

There is already a visible scope for sprints. This can be used and can more or less replace the hand written subselect (at the cost of a way more complicated SQL the scope hides).

The check in the JOIN condition can then be simplified to visible_sprints.id = work_packages.sprint_id I believe.

This is necessary to make the group_by work

We delegate the resource name to the object itself, that way we can override it in our record and no knowledge about the backlogs module is necessary in the core.

EinLama · 2026-04-20T09:58:48Z

#22740 has been merged, feature flag references have been removed from this PR. Comments have been addressed 🤞

ulferts

Looks good @EinLama .

There is only one thing that I would like fixed before merging which is the unnecessary joining of the projects table.

ulferts · 2026-04-20T11:15:41Z

+    def determine_path_method(record)
+      return :sprint if record.is_a?(Agile::Sprint)
+
+      super


I didn't mean to suggest writing a patch, @EinLama. What I had in mind was something like

class API::V3::Utilities::ResourceLinkGenerator def self.register_mapping(klass, symbol) ... end end class OpenProject::Backlogs::Engine config.to_prepare do API::V3::Utilities::ResourceLinkGenerator.register_mapping(Agile::Sprint, :sprint) end end

But this here will also do the trick.

ulferts · 2026-04-20T11:20:41Z

+          #{visible_sprints.to_sql}
+        ) AS visible_sprints
+        ON visible_sprints.id = work_packages.sprint_id
+        AND "projects"."id" IN (#{projects_with_view_sprints.select(:id).to_sql})


The joining of projects is not necessary here. You can simply check against project_id:

AND "work_packages"."project_id" IN (#{projects_with_view_sprints.select(:id).to_sql})

You could also join the projects_with_view_sprints which might make it a bit easier to read:

LEFT OUTER JOIN (#{projects_with_view_sprints.to_sql}) "view_sprints_projects" ON "view_sprints_projects"."id" = "work_packages"."project_id" LEFT OUTER JOIN ( #{visible_sprints.to_sql} ) AS visible_sprints ON visible_sprints.id = work_packages.sprint_id AND view_sprints_projects.id IS NOT NULL

Then again, having written it, maybe it is not easier to read.

I replaced the condition to use work_packages.project_id 👍 Is that what you meant?

We cannot get rid of the additional join on the projects table at the top of the SQL statement due to similar reasons like the project phase definition. If that line/join is removed, we get the same error we got back then:

PG::UndefinedTable: ERROR: missing FROM-clause entry for table "projects" LINE 1: ...status_id" WHERE ((statuses.is_closed=FALSE) AND (projects.i... ^

ulferts · 2026-04-20T11:21:42Z

+    #    directly. Without this second condition, a sprint that is shared *to*
+    #    project_receiving could leak into the sort/group for work packages that live in
+    #    the sharer project itself, because `sprint_source_for` transitively includes the
+    #    sharer when the user only has permission in the receiver.


Good catch here. Took me a while to understand but makes sense.

EinLama force-pushed the feature/73104-sprint-column-sort-and-group-for-work-packages-table branch 9 times, most recently from e4ff513 to ce40bf4 Compare April 14, 2026 05:46

EinLama marked this pull request as ready for review April 14, 2026 06:08

EinLama added the DO NOT MERGE label Apr 16, 2026

ulferts requested changes Apr 17, 2026

View reviewed changes

EinLama force-pushed the feature/73104-sprint-column-sort-and-group-for-work-packages-table branch from 90275fd to efa5a0a Compare April 17, 2026 14:12

EinLama requested a review from ulferts April 20, 2026 08:09

EinLama added 15 commits April 20, 2026 11:56

[#73104] Add Sprint column to work package table

ef5daf2

[#73104] Check permission for Sprint filter

08e620e

[#73104] Allow resource linking for sprints

c952fad

This is necessary to make the group_by work

[#73104] Remove backlogs module pollution from core

6c770ff

We delegate the resource name to the object itself, that way we can override it in our record and no knowledge about the backlogs module is necessary in the core.

[#73104] Add explanatory comment to #api_resource_link_name

d620d79

[#73104] Make GROUP BY work with permission checking

3e08fb7

[#73104] Add test case for user not being a project member

8e4c484

[#73104] Specs for WorkPackageSprintSelect

7a49bdf

[#73104] Small refactor

2a0e691

[#73104] Consider sprint sharing by using the visible scope

9181ebf

[#73104] First specs for sprint sharing

071d1c2

[#73104] More specs for sprint sharing and query fix

74e5fa2

[#73104] Move api resource link out of the model

c1e93e1

[#73104] Use correct expectation for empty, but settable field

b5fef86

[#73104] Remove feature flag checks

457affb

EinLama force-pushed the feature/73104-sprint-column-sort-and-group-for-work-packages-table branch from 596b5d6 to 457affb Compare April 20, 2026 09:58

EinLama removed the DO NOT MERGE label Apr 20, 2026

[#73104] Remove unnecessary require

a5bc0ad

ulferts approved these changes Apr 20, 2026

View reviewed changes

[#73104] projects.id -> work_packages.project_id

3391ab7

EinLama mentioned this pull request Apr 20, 2026

Remove legacy Sprint model #22836

Merged

11 tasks

EinLama merged commit e66e682 into dev Apr 21, 2026
17 of 18 checks passed

EinLama deleted the feature/73104-sprint-column-sort-and-group-for-work-packages-table branch April 21, 2026 11:55

github-actions Bot locked and limited conversation to collaborators Apr 21, 2026

Conversation

EinLama commented Apr 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Ticket

What are you trying to accomplish?

Screenshots

What approach did you choose and why?

Merge checklist

Uh oh!

EinLama commented Apr 16, 2026

Uh oh!

ulferts left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

EinLama commented Apr 20, 2026

Uh oh!

ulferts left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

2 participants

EinLama commented Apr 9, 2026 •

edited

Loading