Count first 1000 deployments or return '999+'.

Changelog: added

Add new graphql field that returns yaml
error messages for a pipeline.

Changelog: added

This MR sets up validation for semantic versioning when releasing a new
vesion of the catalog resource.

* Add migration for adding columns to support parsing of SemVer
* Updates the Versions::CreateService to use `release.tag` for
  versioning
* Updates specs that were not using semantic version, thus testing the
  semantic version validation further
* Quarantines an out of date feature spec

Changelog: added

Changelog: added

Here we have changed the base class of Organizations::GroupsResolver and
Organizations::ProjectsResolver from BaseResolver to GroupsResolver and
ProjectsResolver respectively to use the existing functionalities of
both the resolvers.

We also need a `search` argument in the Organizations::ProjectsResolver
and changing its base class added this argument as well.

The `GroupsResolver` was missing the sort argument which has been added
as well.

Changelog: added

Changelog: removed

Changelog: changed

Instance setting `math_rendering_limits_enabled` can now
be queried and set on groups as well. It is a cascading
setting, so it can be locked at any level.

Changelog: added

Add design widget to
widget_definition table and add
a design widget wrapper
graphql query for designCollection

Issue: https://gitlab.com/gitlab-org/gitlab/-/issues/387552
Changelog: added

Changelog: added

Changelog: other

Add name argument to workItemTypes GQL field to be able to query a specific type.
Also remove the existing (experimental) taskable argument because it becomes redundant.

Changelog: changed

Changelog: removed

- Improve spec coverage
- Update WorkItemType model used in DB migration

projectPath was the very first implementation of the component resource resolver.
Since then, we have moved away from using the namespace for the component search
so the projectPath is no longer needed.

Changelog: changed

Following comments in front-end MR:
https://gitlab.com/gitlab-org/gitlab/-/merge_requests/135959

Split the GraphQL updates into a new merge request.

This only exposes the ci_pipeline_schedules project
plan limits

Changelog: added

This change adds centralized service classes for pipeline trigger token
CRUD in order to standardize behavior and reduce inconsistent behaviors
across different endpoints and uses the new CreateService.

Changelog: other

While working on [MR 142396][0], a bot mentioned that the
[`namespace_projects_resolver_spec.rb`][1]
makes heavy use of `let` for groups and projects.

I have refactored the spec to make use of `let_it_be`

Why 2 less examples
======================================================================

Two of my changes resulted in 2 less examples because they were
exercising the same code and covered by other examples.

1. `it 'finds all projects not aimed for deletion'` and `it 'finds all
   projects not aimed for deletion including the subgroups'` was
   collapsed into 1 example

2. The original `context 'with an user namespace'.it 'finds all
   projects'` and `context 'with an user namespace'.it 'finds all
   projects including the subgroups'` were actually the same example and
   not actually testing distinct situations. I collapsed them into a
   single example

Stats
======================================================================

Summary
-----------------------------------

When running the spec locally with

```sh
bundle exec rspec spec/graphql/resolvers/namespace_projects_resolver_spec.rb
```

I see the following timing differences in the output:

- before: `Finished in 26.14 seconds (files took 11.58 seconds to load)`
- after: `Finished in 6.63 seconds (files took 11.43 seconds to load)`

Before
-----------------------------------
`RD_PROF=1 bundle exec rspec spec/graphql/resolvers/namespace_projects_resolver_spec.rb`

```sh
[TEST PROF INFO] RSpecDissect enabled

Test environment set up in 0.926308698 seconds
...............
[TEST PROF INFO] RSpecDissect report

Total time: 00:23.597

Total `let` time: 00:17.428
Total `before(:each)` time: 00:23.036

Finished in 26.92 seconds (files took 11.62 seconds to load)
15 examples, 0 failures

[TEST PROF INFO] Time spent in factories: 00:17.444 (61.5% of total time)
```

After
-----------------------------------
`RD_PROF=1 bundle exec rspec spec/graphql/resolvers/namespace_projects_resolver_spec.rb`

```sh
[TEST PROF INFO] RSpecDissect enabled

Test environment set up in 0.878414488 seconds
.............
[TEST PROF INFO] RSpecDissect report

Total time: 00:01.030

Total `let` time: 00:00.420
Total `before(:each)` time: 00:00.475

Finished in 6.63 seconds (files took 11.43 seconds to load)
13 examples, 0 failures

[TEST PROF INFO] Time spent in factories: 00:02.126 (26.4% of total time)
```

resolves: https://gitlab.com/gitlab-org/gitlab/-/issues/438946

[0]:https://gitlab.com/gitlab-org/gitlab/-/merge_requests/142396
[1]:https://gitlab.com/gitlab-org/gitlab/-/blob/05b9a56233fa157c853e78a24b874f08c0730420/spec/graphql/resolvers/namespace_projects_resolver_spec.rb

We can't just call super because that just raises an exception. Since we
already have the accepts logic here, we just need to return the type if
it is part of the accepted types.

clientMutationId is now being added as an argument of the mutation class
instead of just the mutation input object. See
https://github.com/rmosolgo/graphql-ruby/pull/4581

Changelog: added

Closes https://gitlab.com/gitlab-org/gitlab/-/issues/437143

This addition to the project query will feed the dropdown for
creating and editing branch rules.

The query returns a list of branches that can be protected. This does
not take into account wildcard branch protections.

Changelog: added

Adds a new resolver that allows searching through the versions of a
machine learning model, along with model changes necessary to support
the searching and ordering.

Changelog: added

Will be used for organization project lists

Changelog: added

It will return the maximum access level that the current user has inside
the project or group.

Changelog: added

This Commit
-----------------------

This commit exposes the new `include_archived` filter param in the
`groupProjects` graphql resolver that was added in [MR 141989][2]

As the existing behavior was to include archived projects, the default
for the new param is `include_archived: true` to keep any other
usages of this graphql resolver unaffected.

Context
-----------------------

For the group-level view of the Vulnerability Report, we exclude
archived projects.

However, the project filter list includes archived projects as
selections:

![image](/uploads/451fd85e4848f1e7ad361eaee4299aad/image.png)

We want to exclude archived projects from that list.

The list is populated by a graphql
resolver ([`security_dashboard/graphql/queries/group_projects.query.graphql`][1])

That resolver ultimately ends up calling the finder which was updated
in [MR 141989][2] to accept a new filter param (`include_archived`)
([`src`][0]).

[0]:https://gitlab.com/gitlab-org/gitlab/-/blob/b1f2a2d85e2f920b1e0a98f71d1d29a9aed3fc5c/app/graphql/resolvers/namespace_projects_resolver.rb
[1]:https://gitlab.com/gitlab-org/gitlab/-/blob/b1f2a2d85e2f920b1e0a98f71d1d29a9aed3fc5c/ee/app/assets/javascripts/security_dashboard/graphql/queries/group_projects.query.graphql#L4
[2]:https://gitlab.com/gitlab-org/gitlab/-/merge_requests/141989

resolves: https://gitlab.com/gitlab-org/gitlab/-/work_items/438814
related to: https://gitlab.com/gitlab-org/gitlab/-/issues/437004

Changelog: changed

This mutation was introduced as alpha so
it doesn't need to follow the deprecation process.
The mutation is no longer needed so we are removing it.

Changelog: removed

Changelog: added

Changelog: other

When anonymous users would access a query on mlModels, they would crash
even if they did have access to the endpoint. This fixes by using
Ability.allowed? instead of user.can?

- Allows users to add a description to instance-level CI variables.
- Include description in query results for inherited CI variables.
- Exposes the `description` field so that the frontend can pass the
values.

Changelog: added

The `path` field on `ci_components` currently holds the file path of the
component within the component project repository. However, we aren't
currently using that data. Instead, we need to provide the include path
for the component so we can display it on the catalog resource details
  page. This commit stores the include path instead of the file path.

Existing component records will still have the file path. We'll decide
later whether it is necessary to update them with a data migration.

Changelog: changed

Changelog: added

Changelog: removed
MR: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/140970

This MR adds the 'name' argument to the Ci catalog
VersionsResolver and adds a 'readme_html' field
to VersionType. This allows FE to query the versioned
README.md content for the catalog resource.

Changelog: added