Commit Graph
4770 Commits
Author SHA1 Message Date
Mathieu Fenniak 0eca229d15 feat: implement fine-grained access tokens in /teams/{id}/repos
**Breaking*: /teams/{id}/repos previously allowed read access to private
repositories even if a "public-only" access token was in-use.  This has
been restricted to only return public repositories in this case.
2026-03-01 10:54:43 -07:00
Mathieu Fenniak 4bd81d3363 feat: add authorization reducer support to SearchRepoOptions 2026-03-01 10:54:43 -07:00
Mathieu FenniakandMathieu Fenniak c89504d573 feat: implement fine-grained access tokens in /user/subscriptions & /users/{username}/subscriptions 2026-03-01 17:05:53 +01:00
Mathieu FenniakandMathieu Fenniak 9c748e87e1 feat: implement fine-grained access tokens in /user/starred & /users/{username}/starred 2026-03-01 17:05:53 +01:00
Mathieu FenniakandMathieu Fenniak 5582faa0c7 refactor: split AuthorizationReducer into a base RepositoryAuthorizationReducer interface 2026-03-01 17:05:53 +01:00
Mathieu FenniakandMathieu Fenniak 2f19237a14 feat: add GetUserRepoPermissionWithReducer 2026-02-27 17:17:29 +01:00
Mathieu FenniakandMathieu Fenniak a1eff6f0dc feat: backend DB model for fine-grained repo access tokens 2026-02-27 17:17:29 +01:00
Myers CarpenterandMathieu Fenniak 2d4a3e5658 fix(models): deduplicate project sorting values and add unique constraints (#11334)
Project columns and cards use a sorting field for ordering, but nothing prevents duplicate values from being inserted. This causes unpredictable ordering and makes swap-based reordering unsafe. Additionally, the same issue can be added to a project multiple times, which shouldn't be possible.

This PR adds a migration to clean up existing data and enforce three unique constraints:
- (project_id, sorting) on project_board — one sorting value per column per project
- (project_id, issue_id) on project_issue — one card per issue per project
- (project_board_id, sorting) on project_issue — one sorting value per card per column

The migration deduplicates existing rows and reassigns sequential sorting values before adding the constraints.

Changes

- Migration: fix duplicate sorting values in project_board and project_issue, remove duplicate (project_id, issue_id) rows, add three unique constraints
- MoveColumnsOnProject: two-phase swap (negate then set) to avoid constraint collisions
- MoveIssuesOnProjectColumn: three-phase approach with duplicate validation and sorted lock ordering
- UpdateColumn: always persist sorting field (allows setting to 0)
- GetDefaultColumn: query max sorting before auto-creating
- createDefaultColumnsForProject: explicit sequential sorting
- changeProjectStatus: only set ClosedDateUnix when closing

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [x] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11334
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Co-authored-by: Myers Carpenter <myers@maski.org>
Co-committed-by: Myers Carpenter <myers@maski.org>
2026-02-27 16:25:20 +01:00
Mathieu FenniakandMathieu Fenniak 0ae6235386 fix: allow Actions runner to recover tasks lost during fetching from intermittent errors (#11401)
Probably fixes (or improves, at least) https://code.forgejo.org/forgejo/runner/issues/1391, paired with the runner implementation https://code.forgejo.org/forgejo/runner/pulls/1393.

When the FetchTask() API is invoked to create a task, unpreventable environmental errors may occur; for example, network disconnects and timeouts. It's possible that these errors occur after the server-side has assigned a task to the runner during the API call, in which case the error would cause that task to be lost between the two systems -- the server will think it's assigned to the runner, and the runner never received it.  This can cause jobs to appear stuck at "Set up job".

The solution implemented here is idempotency in the FetchTask() API call, which means that the "same" FetchTask() API call is expected to return the same values. Specifically, the runner creates a unique identifier which is transmitted to the server as a header `x-runner-request-key` with each FetchTask() invocation which defines the sameness of the call, and the runner retains the value until the API call receives a successful response. The server implementation returns the same tasks back if a second (or Nth) call is received with the same `x-runner-request-key` header.  In order to accomplish this is records the `x-runner-request-key` value that is used with each request that assigns tasks.

As a complication, the Forgejo server is unable to return the same `${{ secrets.forgejo_token }}` for the task because the server stores that value in a one-way hash in the database.  To resolve this, the server regenerates the token when retrieving tasks for a second time.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests for Go changes

(can be removed for JavaScript changes)

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [x] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I ran...
  - [x] `make pr-go` before pushing

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11401
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-02-22 23:24:38 +01:00
Antonin DelpeuchandMathieu Fenniak 45db3c98a3 feat: link CI job to its defining workflow file (#11216)
Fixes #11036.

This adds a link from a CI run to the file that its workflow was taken from.

| Before | After |
|---------|---------|
| ![screenshot](/attachments/49741492-c98b-4a9a-b8bf-f6628698e008) | ![image](/attachments/8ec7dd76-d4ba-4f58-a63a-dd7886e16aae) |

Before:
* the `test.yml` link points to the list of other runs (`/org123/repo2/actions?workflow=test.yml`)

After:
* the `test.yml` link points to the workflow definition (`/org123/repo2/src/commit/55b048363c8cfa7d9e8b5cade5c75681bd0c7328/.forgejo/workflows/test.yml`)
* the `all runs` link points to the list of other runs (`/org123/repo2/actions?workflow=test.yml`)

I have tried to retain the existing link to the list of workflow runs (moving it to a separate link), but I am not sure if this link should be retained at all and if so how.

## Checklist

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [x] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [x] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

<!--start release-notes-assistant-->

## Release notes
<!--URL:https://codeberg.org/forgejo/forgejo-->
- Features
  - [PR](https://codeberg.org/forgejo/forgejo/pulls/11216): <!--number 11216 --><!--line 0 --><!--description bGluayBDSSBqb2IgdG8gaXRzIGRlZmluaW5nIHdvcmtmbG93IGZpbGU=-->link CI job to its defining workflow file<!--description-->
<!--end release-notes-assistant-->

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11216
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Co-authored-by: Antonin Delpeuch <antonin@delpeuch.eu>
Co-committed-by: Antonin Delpeuch <antonin@delpeuch.eu>
2026-02-20 03:11:29 +01:00
limiting-factorandGusted d4de6a4e5c fix: UserTypeRemoteUser is an eligible organization member (#11326)
A [safeguard][0] was added to ensure a team member is not an organization. A [discussion][1] concluded that remote users should not yet be allowed to be team members, at least until it is implemented (by F3 or ActivityPub).

As [gof3 v3.11.25][2] has support for [teams][3], remote users need to be added as an eligible team member, otherwise compliance tests for the Forgejo internal driver [implementation][4] fail.

The IsUserByID function is renamed to IsAnEligibleTeamMemberByID to clarify its purpose. The TestIsUserConsistency is removed because the IsUser function is used in [an entirely different context][5] and there is no reason to ensure both are consistent. A unit test is added for IsAnEligibleTeamMemberByID as a replacement and also verifies organizations are not considered eligible.

[0]: https://codeberg.org/forgejo/forgejo/pulls/9757
[1]: https://codeberg.org/forgejo/forgejo/pulls/9757#issuecomment-7802255
[2]: https://code.forgejo.org/f3/gof3/releases/tag/v3.11.25
[3]: https://code.forgejo.org/f3/gof3/pulls/484/files
[4]: https://code.forgejo.org/forgefriends/forgefriends/commit/9c565a9b37ec88d1d6591289784ac003056e5b4d#diff-241521b6ebc5a1da660421daea23da1dd1086158
[5]: https://codeberg.org/forgejo/forgejo/src/commit/b66a76686d64d5b0b2062a9a8cc65c7bc68eac7e/routers/api/v1/api.go#L299-L308

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11326
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: limiting-factor <limiting-factor@posteo.com>
Co-committed-by: limiting-factor <limiting-factor@posteo.com>
2026-02-17 02:39:46 +01:00
Manuel GanterandMathieu Fenniak 5b6bbabd74 feat: implement ephemeral runners (#9962)
As described in [this comment](https://gitea.com/gitea/act_runner/issues/19#issuecomment-739221) one-job runners are not secure when running in host mode. We implemented a routine preventing runner tokens from receiving a second job in order to render a potentially compromised token useless. Also we implemented a routine that removes finished runners as soon as possible.

Big thanks to [ChristopherHX](https://github.com/ChristopherHX) who did all the work for gitea!

Rel: #9407

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [x] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [ ] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [ ] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/9962
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Co-authored-by: Manuel Ganter <manuel.ganter@think-ahead.tech>
Co-committed-by: Manuel Ganter <manuel.ganter@think-ahead.tech>
2026-02-16 18:56:56 +01:00
5589182c54 fix: honor org/user project in new issue (#9906)
Previously, when a project id was passed into the new issue form as a
query parameter, it wouldn’t be selected if the project belonged to the
user or organization instead of directly to the repository.

Resolves forgejo/forgejo#8489

Signed-off-by: Nils Philippsen <nils@redhat.com>
Co-authored-by: Gusted <postmaster@gusted.xyz>
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/9906
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Reviewed-by: Robert Wolff <mahlzahn@posteo.de>
Co-authored-by: Nils Philippsen <nilsph@noreply.codeberg.org>
Co-committed-by: Nils Philippsen <nilsph@noreply.codeberg.org>
2026-02-16 05:52:47 +01:00
Robert WolffandMathieu Fenniak 54d952ee66 feat(ui): support additional job status selection in dropdown menu on Actions tab (#11156)
### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [x] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

### Disclaimer

Generative AI (ChatGPT) was used to debug the e2e test, with copied code lines below threshold of originality.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11156
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Robert Wolff <mahlzahn@posteo.de>
Co-committed-by: Robert Wolff <mahlzahn@posteo.de>
2026-02-11 03:22:53 +01:00
Mathieu FenniakandMathieu Fenniak 51d0188533 refactor: replace Value() from Option[T] with Get() & ValueOrZeroValue() (#11218)
`Option[T]` currently exposes a method `Value()` which is permitted to be called on an option that has a value, and an option that doesn't have a value.  This API is awkward because the behaviour if the option doesn't have a value isn't clear to the caller, and, because almost all accesses end up being `.Has()?` then `OK, use .Value()`.

`Get() (bool, T)` is added as a better replacement, which both returns whether the option has a value, and the value if present.  Most call-sites are rewritten to this form.

`ValueOrZeroValue()` is a direct replacement that has the same behaviour that `Value()` had, but describes the behaviour if the value is missing.

In addition to the current API being awkward, the core reason for this change is that `Value()` conflicts with the `Value()` function from the `driver.Valuer` interface.  If this interface was implemented, it would allow `Option[T]` to be used to represent a nullable field in an xorm bean struct (requires: https://code.forgejo.org/xorm/xorm/pulls/66).

_Note:_ changes are extensive in this PR, but are almost all changes are easy, mechanical transitions from `.Has()` to `.Get()`.  All of this work was performed by hand.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [x] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11218
Reviewed-by: Otto <otto@codeberg.org>
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-02-10 16:41:21 +01:00
Andreas AhlenstorfandGusted f7873ba393 fix: normalize secrets consistently, display accurate help (#11052)
Forgejo's UI claims that whitespace is removed from the beginning and the end of the values of Forgejo Actions variables and secrets. However, that is not correct. The entered values are stored as-is. Only CRLF is replaced with LF, which is also the desired behaviour.

This PR changes the incorrect text which is also no longer displayed as placeholder but as a proper help text below the input fields. Furthermore, tests were added to verify the behaviour.

While adding tests, I discovered and fixed another inconsistency. Depending on whether secrets were managed using the UI or the HTTP API, they were treated differently. CRLF in secrets entered in the UI was correctly replaced with LF while secrets created using the HTTP API kept CRLF.

Fixes #11003.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11052
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Andreas Ahlenstorf <andreas@ahlenstorf.ch>
Co-committed-by: Andreas Ahlenstorf <andreas@ahlenstorf.ch>
2026-02-09 17:02:18 +01:00
LuisandGusted 3f7859f52d feat: improve label filtering exclusion (#10702)
Adds a new button on the right side of the label's filter menu items to explicitly exclude labels.

The new button is reachable with the keyboard by using the vertical arrow keys to reach the label you want to exclude and then the horizontal arrow keys to select the exclusion button.

The new button will only be visible when hovering the menu item or reaching it with the keyboard.

Adjusted the alignment of labels when at least one label is selected so that users can clearly discern which labels are selected or not.

Resolves #3302

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10702
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Luis <luis@adame.dev>
Co-committed-by: Luis <luis@adame.dev>
2026-02-08 00:31:31 +01:00
Mathieu FenniakandMathieu Fenniak cc47a4057f ci: introduce semgrep to prevent using xorm.Sync() incorrectly in new migrations (#11142)
Adds a CI check which detects any usage of xorm's `Sync` method that doesn't include `IgnoreDropIndices: true`, and causes an error.

`semgrep` is a semantic grep tool that allows for the relatively easy authoring of linting tools that are customized to a project's specific needs, rather than generic like `golangci` and related tools.  Although `semgrep` offers a suite of out-of-the-box rules (and a paid set of rules), neither of those are used here -- only one Forgejo-specific rule is added in `.semgrep/xorm.yaml`.

My intent with this change is to introduce the idea and infrastructure of `semgrep` with a single minimal rule.  Once in-place, this will become a tool that we can use when we recognize bad coding patterns and wish to correct them permanently, rather than relying on human code review.  While generic linting tools do this well for general patterns, this will allow Forgejo to apply domain-specific checks.  For example, in #11112, an error indicates that it might be appropriate for us to always use `.StorageEngine("InnoDB")` when using an xorm engine -- if we made that determination, it could be cemented in-place with a `semgrep` rule relatively easily.

This specific rule looks for any access for xorm's `Sync` or `SyncWithOptions` methods on the `*xorm.Engine` or `*xorm.Session`.  They are then considered errors if they don't include `IgnoreDropIndices: true`.  This is *typically* correct and safe, but can also be ignored when specifically needed.  In the `.semgrep/tests` folder, test code is added which validates that the `semgrep` rule matches the expected patterns; this self-test is run before `semgrep` runs on the PR in CI.

As a demonstration, when `IgnoreDropIndices` is removed from a migration, here's an error: https://codeberg.org/forgejo/forgejo/actions/runs/135750/jobs/12/attempt/1

```
    models/forgejo_migrations/v14b_add-action_run-preexecutionerrorcode.go
   ❯❯❱ semgrep.xorm-sync-missing-ignore-drop-indices
          xorm Sync operation may drop indices if used on an incomplete bean definition for an existing table.
          Use SyncWithOptions with IgnoreDropIndices: true instead.

           22┆ _, err := x.SyncWithOptions(xorm.SyncOptions{}, new(ActionRun))
```

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [x] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11142
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-02-07 21:52:43 +01:00
Mathieu FenniakandMathieu Fenniak f24a97f719 fix: newly expanded dynamic matrix jobs can become stuck in a 'blocked' state (#11184)
The end-to-end tests are currently failing on v15: https://code.forgejo.org/forgejo/end-to-end/actions/runs/4900/jobs/8/attempt/1#jobstep-4-356   This is a regression from #11164.

The cause of this regression is:
- When the job emitter emits new jobs, it *now* sets their `Needs` field correctly, fixed in #11164.
- If a job has `needs` set to a non-empty array, it will start as status **blocked**: https://codeberg.org/forgejo/forgejo/src/commit/db037fca507178feb2bd05e0bf59df7c7534116d/models/actions/run.go#L369-L370

This got past manual testing and the end-to-end test run in #11164 because it is intermittent.  If the runner invokes `UpdateTask` multiple times once the status of the job is settled, then the API will invoke the job emitter multiple times -- a second run would unblock the newly blocked jobs.  Podman-based runners do this often due to a long cleanup time, and Docker-based runners (like the end-to-end test) can do it randomly depending on a race condition.

https://codeberg.org/forgejo/forgejo/src/commit/db037fca507178feb2bd05e0bf59df7c7534116d/routers/api/actions/runner/runner.go#L241-L244

The fix for this is to reinvoke the job emitter's logic whenever new jobs are created, so that their new blocked state can be reevaluated to see if it is correct or not.  This is treated cautiously by examining if new jobs are present.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [x] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.
    - This will be a fix to an unreleased regression within the milestone, and so won't be user-facing.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11184
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-02-07 14:36:49 +01:00
Mathieu FenniakandMathieu Fenniak 462ad7bb33 fix: don't abandon Action jobs waiting for approval (#11145)
On an open PR that is waiting for job approval, if jobs haven't been approved by the time the abandon timeout occurs they get marked as cancelled.  This doesn't match the expectations of abandoned jobs in my opinion, which is that they were never able to be dispatched to a runner (no matching labels), but these jobs never got a chance.  They should remain valid and blocked until approved.

Discovered while testing #11125, but unrelated to the behaviour fixed there.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11145
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-02-04 16:00:18 +01:00
Panagiotis "Ivory" VasilopoulosandGusted 81601eab85 feat(activitypub): use structure @PreferredUsername@host.tld:port for actors (#9254)
This modifies usernames of ActivityPub accounts to use the @example@example.tld
format with an additional optional port component (e.g. @user@example.tld:42).
This allows accounts from ActivityPub servers with more relaxed username
requirements than those of Forgejo's to interact with Forgejo. Forgejo would
also follow a "de facto" standard of ActivityPub implementations.

By separating different information using @'s, we also gain future
opportunities to store more information about ActivityPub accounts internally,
so that we won't have to rely on e.g. the amount of dashes in a username as
my migration currently does.

Continuation of Aravinth's work: https://codeberg.org/forgejo/forgejo/pulls/4778

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/9254
Reviewed-by: jerger <jerger@noreply.codeberg.org>
Reviewed-by: Ellen Εμιλία Άννα Zscheile <fogti@noreply.codeberg.org>
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Panagiotis "Ivory" Vasilopoulos <git@n0toose.net>
Co-committed-by: Panagiotis "Ivory" Vasilopoulos <git@n0toose.net>
2026-01-30 23:45:11 +01:00
Mathieu FenniakandMathieu Fenniak c198082975 fix: empty dynamic matrix can leave action run hanging incomplete (#11063)
Fixes #11030.

When a `strategy.matrix` needs to be evaluated on the output of another job, it can become evaluated into an empty set of jobs.  In this case, and assuming no other jobs in the run are active, the run should reach a settled state.  The logic to check the other jobs in the run and determine if this state has been hit needs to be explicitly added to the job emitter.

To accomplish this change, this action run state logic was extracted out of `UpdateRunJobWithoutNotification` where it could be reused.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11063
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-27 17:10:59 +01:00
christopher-beschandGusted d9545c503e fix: decrease watch count when blocking user (#10882)
Fixes #10881

Call the proper function for each repository the user watches, so adjusting the watch count can be done properly.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10882
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: christopher-besch <mail@chris-besch.com>
Co-committed-by: christopher-besch <mail@chris-besch.com>
2026-01-26 22:58:22 +01:00
Christoph MewesandGusted 023a894677 chore: fix typos throughout the codebase (#10753)
This PR fixes a number of typos throughout the entire repository. Running https://github.com/crate-ci/typos and then changing all occurrences that I naively deemed "safe enough".

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10753
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Christoph Mewes <christoph@kubermatic.com>
Co-committed-by: Christoph Mewes <christoph@kubermatic.com>
2026-01-26 22:57:33 +01:00
0ko 3cafb7fa6c chore(ui): change /devtest to /-/demo (#11019)
It has always been largely used for showcasing UI elements but that name didn't work too well for it.

Testing:
Some of existing tests depend on these pages, making it redundant to create extra tests.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11019
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
2026-01-26 13:12:25 +01:00
christopher-beschandGusted 7408bf9b1e chore: remove dead code in WatchIfAuto and general model documentation (#10880)
I've been reading through the watch model code and found that the `isWrite` flag is dead code — it isn't used anywhere. To be honest I don't see why we should need that anyways. Therefore, I deleted it.

Also I've added some comments for the `WatchMode`. It took me quite some time to figure out why there are four options.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10880
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: christopher-besch <mail@chris-besch.com>
Co-committed-by: christopher-besch <mail@chris-besch.com>
2026-01-26 12:03:08 +01:00
Andreas AhlenstorfandMathieu Fenniak d996dfb476 feat: filter action runs by Git reference (#11013)
Make it possible to filter action runs returned by `/api/v1/repos/andreas/test/actions/runs` by Git reference.

Resolves https://codeberg.org/forgejo/forgejo/issues/10874.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [x] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11013
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Co-authored-by: Andreas Ahlenstorf <andreas@ahlenstorf.ch>
Co-committed-by: Andreas Ahlenstorf <andreas@ahlenstorf.ch>
2026-01-24 21:25:31 +01:00
Nils GorollandMathieu Fenniak 2d55ad5585 fix NewMockWebServer(): Headers never reached the http client (#11007)
Found while working on https://codeberg.org/forgejo/forgejo/pulls/10798#issuecomment-10083846: The symptom was that the go-github client never returned a `resp.After`, so I tracked down the root cause, which was that, with the mocked http server ...

Mocked headers never reached the calling client, because w.WriteHeader()
was called before the headers were set in the response.

Fix by moving w.WriteHeader() to the right place just before w.Write(),
which writes the body.

Test added which fails without the fix and succeeds with it.

### Tests

- I added test coverage for Go changes...
  - [X] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [X] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [X] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/11007
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Co-authored-by: Nils Goroll <nils.goroll@uplex.de>
Co-committed-by: Nils Goroll <nils.goroll@uplex.de>
2026-01-24 04:51:18 +01:00
Mathieu FenniakandMathieu Fenniak 374b3d9f4d fix: remove infinite loop in UpdateRunJobWithoutNotification when run in transaction (#10945)
#10893 introduced a retry loop to manage concurrent updates when updating the state of `action_run` in the function `UpdateRunJobWithoutNotification`.  However, when `UpdateRunJobWithoutNotification` is called from within a transaction, the retry loop continues to read the same data from the DB (due to repeatable read isolation) and loops infinitely.

As #10893 was later identified to not be required to fix the target problem (https://code.forgejo.org/forgejo/runner/issues/1302), this PR reverts the change.  The only retained change is that the error `ErrActionRunOutOfDate` is a constant rather than `errors.New("run has changed")`.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
      - [x] Reverted the test added for 10893 after confirming that it is the cause of the problem.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10945
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Reviewed-by: Ellen Εμιλία Άννα Zscheile <fogti@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-20 17:34:59 +01:00
0ko 10aaef5c7b feat(perf): remove unused size url parameter for local avatars (#10932)
Avatars storage handler ignores the size= url parameter, and we don't have any code for creating multiple sizes of the same avatar.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10932
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
2026-01-20 04:59:15 +01:00
Mathieu FenniakandMathieu Fenniak 9f8dda6325 fix: retry ActionRun updates when optimistic-concurrency-control indicates record changed (#10893)
When concurrent updates occur to the `action_run` table, fetching a task via `FetchTask` can result in an error:
```
time="2026-01-16T16:02:30Z" level=error msg="failed to fetch task" error="internal: pick task: CreateTaskForRunner: update run 2358339: run has changed"
```

This is an error with no known harm.  However, it is more efficient to recover from doing 90% of the work to fetch a task, and complete that work, rather than error out and rollback.  Discovered while investigating https://code.forgejo.org/forgejo/runner/issues/1302#issuecomment-73859, although conclusions do not indicate that it is a source of this bug.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [ ] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10893
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-17 23:09:49 +01:00
Mathieu FenniakandMathieu Fenniak 009e3c670b fix: use ALTER TABLE in SQLite DropTableColumns(), allowing unexpected database sources to work better in migrations (#10888)
The existing implementation of `DropTableColumns()` came from before SQLite had the ability to `ALTER TABLE ... DROP COLUMN ...`.  It works by parsing the table definition and rewriting it without the columns that are to be dropped, but it will fail to do this correctly if the table definition is not in the exact expected format.  In #10887, a database that had probably come through some migration tool was not exactly formatted the way Forgejo expected, resulting in a migration failure.

This replaces `DropTableColumns()`'s hacky SQLite implementation with a more straightforward implementation.  Affected indexes touching the target fields are dropped, then the field is dropped.

DROP COLUMN is supported on SQLite since [3.35.0, 2021-03-12](https://sqlite.org/releaselog/3_35_0.html).

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- Existing `test-sqlite-migration` coverage is relied upon for this change.  During development it was proven to exercise the affected code -- in other words, multiple iterations of changes were required due to it failing as I worked on it.
- No coverage is added for "database with unexpected schema definition format" as the trigger issue for this change though, a point that can be raised if someone believes it is worthwhile.
- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] This change will be noticed by a Forgejo user or admin (feature, bug fix, performance, etc.). I suggest to include a release note for this change.
- [ ] This change is not visible to a Forgejo user or admin (refactor, dependency upgrade, etc.). I think there is no need to add a release note for this change.

*The decision if the pull request will be shown in the release notes is up to the mergers / release team.*

The content of the `release-notes/<pull request number>.md` file will serve as the basis for the release notes. If the file does not exist, the title of the pull request will be used instead.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10888
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-17 21:28:35 +01:00
Mathieu FenniakandMichael Kriese 1aca323acb fix: make concurrency group job cancellation effect runs that are failed (#10863)
When an action's job fails, it marks the entire run as failed.  Concurrency group cancellation was only looking for runs that are in a pending state, and therefore after a single job failed in the run, none of the other jobs in the run could be cancelled by a matching cancel-in-progress job.

Raised in https://codeberg.org/Codeberg/Community/issues/2315.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10863
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-16 10:54:01 +01:00
Mathieu FenniakandMichael Kriese c78a4198b1 fix: recreate-tables doesn't work on PostgreSQL with multiple Forgejo schemas (#10854)
Discovered while trying to reproduce #10848 -- when using `forgejo doctor recreate-tables` against a PostgreSQL database with multiple Forgejo schemas in it, it fails.  The reason is that when querying for index and sequence information, it begins to get information from the other schemas.

```
2026/01/15 15:19:15 ...3.6.1/command_run.go:288:run() [I] PING DATABASE postgresschema
2026/01/15 15:19:15 ...igrations/base/db.go:51:func2() [I] Creating temp table: tmp_recreate__external_login_user for Bean: ExternalLoginUser
2026/01/15 15:19:15 ...igrations/base/db.go:108:func2() [I] Copying table external_login_user to temp table tmp_recreate__external_login_user
2026/01/15 15:19:15 ...igrations/base/db.go:120:func2() [I] Dropping existing table external_login_user, and renaming temp table tmp_recreate__external_login_user in its place
2026/01/15 15:19:15 cmd/doctor.go:216:func1() [E] [Error SQL Query] ALTER INDEX "external_login_user_pkey" RENAME TO "external_login_user_pkey" [] - ERROR: relation "external_login_user_pkey" does not exist (SQLSTATE 42P01)
2026/01/15 15:19:15 ...igrations/base/db.go:404:renameTable() [E] Unable to rename external_login_user_pkey to external_login_user_pkey. Error: ERROR: relation "external_login_user_pkey" does not exist (SQLSTATE 42P01)
Command error: migrate: ERROR: relation "external_login_user_pkey" does not exist (SQLSTATE 42P01)
```

This is a very niche use-case that is likely to only affect a developer using PostgreSQL and popping back to older releases often enough to keep them around in different DB schemas.  I don't think it's worth an automated test, which would require creating a secondary DB schema in a specific migration test.  Manually tested on my dev environment.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10854
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-16 10:53:19 +01:00
Mario MinardiandMathieu Fenniak c84cbd56a1 feat: add OIDC workload identity federation support (#10481)
Add support for OIDC workload identity federation.

Add ID_TOKEN_SIGNING_ALGORITHM, ID_TOKEN_SIGNING_PRIVATE_KEY_FILE, and
ID_TOKEN_EXPIRATION_TIME settings to settings.actions to allow for admin
configuration of this functionality.

Add OIDC endpoints (/.well-known/openid-configuration and /.well-known/keys)
underneath the "/api/actions" route.

Add a token generation endpoint (/_apis/pipelines/workflows/{run_id}/idtoken)
underneath the "/api/actions" route.

Depends on: https://code.forgejo.org/forgejo/runner/pulls/1232
Docs PR: https://codeberg.org/forgejo/docs/pulls/1667

Signed-off-by: Mario Minardi <mminardi@shaw.ca>

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [x] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [x] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [ ] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [ ] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10481
Reviewed-by: Mathieu Fenniak <mfenniak@noreply.codeberg.org>
Co-authored-by: Mario Minardi <mminardi@shaw.ca>
Co-committed-by: Mario Minardi <mminardi@shaw.ca>
2026-01-15 03:39:00 +01:00
limiting-factorandGusted a71392c6aa chore: update gof3/v3 v3.11.15 (#10673)
Update the Forgejo driver for gof3 with modifications for non-backward compatible changes. The changes are isolated behind the f3.Enable flag and not yet functional. The purpose of this upgrade is to not drift from the gof3 implementation while the work continues.

The `fix: include remote users when counting users` commit is a functional change to Forgejo itself but does not change the behavior because the remote users are only created in fixtures or by F3.

https://codeberg.org/forgejo/forgejo/src/commit/59c721d26b1a01e6e6ca812b63066a33e7e5e4da/models/user/user.go#L65-L66

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10673
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: limiting-factor <limiting-factor@posteo.com>
Co-committed-by: limiting-factor <limiting-factor@posteo.com>
2026-01-13 16:59:56 +01:00
Mathieu FenniakandMathieu Fenniak 0eb0179c1a feat: add foreign keys to the action_runner_token table (#10756)
In support of adding foreign keys to the `action_runner_token` table, this PR also had to:
- Add detection and error if a table with "soft delete" is used with a foreign key, because it causes a tricky to track down foreign key violation
- Remove unused xorm "soft delete" capability on the `action_runner_token` table
- Change the `RepoID` and `OwnerID` fields to use the value `NULL` to indicate that this scope wasn't valid for the token, rather than the value `0`

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10756
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-12 21:59:40 +01:00
Έλλεν Εμίλια Άννα ZscheileandGusted 4fd275116d feat: teach lint-locale-usage about ObjectVerification.Reason (#10755)
Add special parsing to handle the keys found in the `Reason` field of `ObjectVerification` structs.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10755
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Έλλεν Εμίλια Άννα Zscheile <fogti+devel@ytrizja.de>
Co-committed-by: Έλλεν Εμίλια Άννα Zscheile <fogti+devel@ytrizja.de>
2026-01-11 09:40:49 +01:00
Mathieu FenniakandGusted 270317a3ad fix: add forgejo doctor cleanup-commit-status command (#10686)
```
NAME:
   forgejo doctor cleanup-commit-status - Cleanup extra records in commit_status table

USAGE:
   forgejo doctor cleanup-commit-status

DESCRIPTION:
   Forgejo suffered from a bug which caused the creation of more entries in the
   "commit_status" table than necessary. This operation removes the redundant
   data caused by the bug. Removing this data is almost always safe.
   These reundant records can be accessed by users through the API, making it
   possible, but unlikely, that removing it could have an impact to
   integrating services (API: /repos/{owner}/{repo}/commits/{ref}/statuses).

   It is safe to run while Forgejo is online.

   On very large Forgejo instances, the performance of operation will improve
   if the buffer-size option is used with large values. Approximately 130 MB of
   memory is required for every 100,000 records in the buffer.

   Bug reference: https://codeberg.org/forgejo/forgejo/issues/10671

OPTIONS:
   --help, -h                       show help
   --custom-path string, -C string  Set custom path (defaults to '{WorkPath}/custom')
   --config string, -c string       Set custom config file (defaults to '{WorkPath}/custom/conf/app.ini')
   --work-path string, -w string    Set Forgejo's working path (defaults to the directory of the Forgejo binary)
   --verbose, -V                    Show process details
   --dry-run                        Report statistics from the operation but do not modify the database
   --buffer-size int                Record count per query while iterating records; larger values are typically faster but use more memory (default: 100000)
   --delete-chunk-size int          Number of records to delete per DELETE query (default: 1000)
```

The cleanup effectively performs `SELECT * FROM commit_status ORDER BY repo_id, sha, context, index, id`, and iterates through the records.  Whenever `index, id` changes without the other fields changing, then it's a useless record that can be deleted.  The major complication is doing that at scale without bringing the entire database table into memory, which is performed through a new iteration method `IterateByKeyset`.

Manually tested against a 455,303 record table in PostgreSQL, MySQL, and SQLite, which was reduced to 10,781 records, dropping 97.5% of the records.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10686
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-10 23:15:21 +01:00
GustedandMathieu Fenniak e559a5fe6c fix: use correct GPG key for export
`GPGKeyToEntity` incorrectly assumed that within a keyring with multiple
keys that the first key is verified and should be exported. Look at all
keys and find the one that matches the verified key ID.
2026-01-06 09:58:20 -07:00
Mathieu Fenniak f7442ac4c7 fix: hide user profile anonymous options on public repo APIs 2026-01-06 09:58:20 -07:00
luisadameandGusted d8501b42fc fix: don't display pending reviews as participants (#10528)
Fixes #10155

When participants are displayed, don't include those that only have made a pending review. Those should not yet be revealed as participants.

Apart from adding automated tests, this is the manual verification process I've followed:
1. Set up three users
2. User 1 creates a repository, then creates a pull request adding a new file
3. User 2 creates a new code comment but doesn't not publish the review, shows as pending.
4. User 3 creates a new code comment and publishes the review.
5. From everyone's perspective the number of participants is: 2. And, the participants displayed in the list are 1 and 3. User 2, which hasn't yet published the review is not displayed.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10528
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: luisadame <luisadame@noreply.codeberg.org>
Co-committed-by: luisadame <luisadame@noreply.codeberg.org>
2026-01-06 10:47:21 +01:00
Mathieu FenniakandMathieu Fenniak 9d6ae1471e fix: re-running an expanded reusable workflow causes duplicate "attempt 1" job (#10666)
Placeholder tasks, which are used to store the outputs of a reusable workflow, were hard coded to always have attempt 1.  If you executed "Re-run all jobs" with a reusable workflow, a second placeholder task would be created with the same attempt, which caused: (a) Forgejo to not know which attempt, and therefore which outputs, were valid, and (b) the UI to be stuck in "You are viewing an out-of-date run of this job..." when viewing the job.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [x] I do not want this change to show in the release notes.
- [ ] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10666
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2026-01-02 15:26:11 +01:00
limiting-factorandGusted 2faaa4c5b4 chore: move all test blank imports in a single package (#10662)
- Create `modules/testimport/import.go` to centralize blank import needed for tests (in order to run the `init` function)  to simplify maintenance.
- Remove the imports that are not needed.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10662
Reviewed-by: Michael Kriese <michael.kriese@gmx.de>
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: limiting-factor <limiting-factor@posteo.com>
Co-committed-by: limiting-factor <limiting-factor@posteo.com>
2026-01-02 05:32:32 +01:00
Mathieu FenniakandMathieu Fenniak 75cb38faa6 feat: support reusable workflow expansion when with or strategy.matrix contains ${{ needs... }} (#10647)
This change allows the `with:` field of a reusable workflow to reference a previous job, such as `with: { some-input: "${{ needs.other-job.outputs.other-output }}" }`.  `strategy.matrix` can also reference `${{ needs... }}`.

When a job is parsed and encounters this situation, the outer job of the workflow is marked with a field `incomplete_with` (or `incomplete_matrix`), indicating to Forgejo that it can't be executed as-is and the other jobs in its `needs` list need to be completed first.  And then in `job_emitter.go` when one job is completed, it checks if other jobs had a `needs` reference to it and unblocks those jobs -- but if they're marked with `incomplete_with` then they can be sent back through the job parser, with the now-available job outputs, to be expanded into the correct definition of the job.

The core functionality for this already exists to allow `runs-on` and `strategy.matrix` to reference the outputs of other jobs, but it is expanded upon here to include `with` for reusable workflows.

There is one known defect in this implementation, but it has a limited scope -- if this code path is used to expand a nested reusable workflow, then the `${{ input.... }}` context will be incorrect.  This will require an update to the jobparser in runner version 12.4.0, and so it is out-of-scope of this PR.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).
- **end-to-end test:** will require the noted "known defect" to be resolved, but tests are authored at https://code.forgejo.org/forgejo/end-to-end/compare/main...mfenniak:expand-reusable-workflows-needs

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10647
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2025-12-31 19:04:35 +01:00
Mathieu FenniakandMathieu Fenniak 9b2f7c557b feat: support jobs.<job_id>.secrets with reusable workflow expansion (#10627)
Follow-up to #10525; adds support for `jobs.<job_id>.secrets` to expanded reusable workflows (when no `runs-on` is specified in a job that `uses: ...` another workflow).

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).
- **end-to-end testing**: [prepared, PR n](https://code.forgejo.org/forgejo/end-to-end/pulls/1351)

### Documentation

- [x] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
    - [ ] Doc to be created
- [ ] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10627
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2025-12-30 17:33:21 +01:00
Mathieu FenniakandMathieu Fenniak f7d2f51bf7 feat: support workflow inputs on expanded reusable workflows (#10614)
Follow-up to #10525; adds support for `on.workflow_call.inputs` to expanded reusable workflows (when no `runs-on` is specified in a job that `uses: ...` another workflow).

The majority of the work for this is done by the `jobparser` library which evaluates inputs automatically when the job is being parsed and stores those inputs on the expanded jobs as the "default" value in `on.workflow_call.inputs`.  Forgejo's role here is just to to ensure that `forgejo.event_name` is set to `"workflow_call"` when a job is dispatched, which causes the runner to use the inputs that are stored -- https://code.forgejo.org/forgejo/runner/src/commit/34c20aa50f1d65e1f0265b5b5274a0bbe5db128e/act/runner/expression.go#L452-L467.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).
- **end-to-end testing**: https://code.forgejo.org/forgejo/end-to-end/pulls/1323

### Documentation

- [x] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
    - https://codeberg.org/forgejo/docs/pulls/1664
- [ ] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

<!--start release-notes-assistant-->

## Release notes
<!--URL:https://codeberg.org/forgejo/forgejo-->
- Features
  - [PR](https://codeberg.org/forgejo/forgejo/pulls/10614): <!--number 10614 --><!--line 0 --><!--description c3VwcG9ydCB3b3JrZmxvdyBpbnB1dHMgb24gZXhwYW5kZWQgcmV1c2FibGUgd29ya2Zsb3dz-->support workflow inputs on expanded reusable workflows<!--description-->
<!--end release-notes-assistant-->

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10614
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2025-12-29 15:37:44 +01:00
Mathieu FenniakandMathieu Fenniak 81cce58a48 migration: update existing foreign key migrations to automatically fix inconsistencies (#10568)
Changes foreign key database inconsistency handling so that inconsistent records are automatically deleted with an administrator warning during migration.  As noted in discussion: https://codeberg.org/forgejo/discussions/issues/385#issuecomment-9175566

Intended for backport to v14 to eliminate pain during upgrades.

Because these migrations are now deleting data, rather than allowing the administrator to do it, all migrations have been covered with an integration test that verifies expected data is deleted.  This is particularly interesting with nullable fields.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10568
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2025-12-29 02:50:20 +01:00
Mathieu FenniakandMathieu Fenniak fb35abb7e1 feat: support workflow outputs on expanded reusable workflows (#10578)
Follow-up to #10525; adds support for `on.workflow_call.outputs` to expanded reusable workflows (when no `runs-on` is specified on a job that `uses: ...` another workflow).  When all the inner jobs of a workflow call complete, `on.workflow_call.outputs` is evaluated and the related outputs are stored on the outer job's `ActionTask`.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [x] in their respective `*_test.go` for unit tests.
  - [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).
- **end-to-end testing**: https://code.forgejo.org/forgejo/end-to-end/pulls/1322

### Documentation

- [x] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
    - https://codeberg.org/forgejo/docs/pulls/1661
- [ ] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

<!--start release-notes-assistant-->

## Release notes
<!--URL:https://codeberg.org/forgejo/forgejo-->
- Features
  - [PR](https://codeberg.org/forgejo/forgejo/pulls/10578): <!--number 10578 --><!--line 0 --><!--description c3VwcG9ydCB3b3JrZmxvdyBvdXRwdXRzIG9uIGV4cGFuZGVkIHJldXNhYmxlIHdvcmtmbG93cw==-->support workflow outputs on expanded reusable workflows<!--description-->
<!--end release-notes-assistant-->

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10578
Reviewed-by: Andreas Ahlenstorf <aahlenst@noreply.codeberg.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2025-12-28 20:05:13 +01:00
Mathieu FenniakandMathieu Fenniak 663acf102a fix: ListTrackedTimes API has no defined record ordering (#10588)
API call `GET /repos/{owner}/{repo}/issues/{index}/times` has no defined ordering implemented in it, causing PostgreSQL to have intermittent test failures on `TestAPIGetTrackedTimes` which expected records to be returned in ID order.  ID order is reasonable enough, so this PR adds that ordering.

Fixes #10577.

## Checklist

The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).

### Tests

- I added test coverage for Go changes...
  - [ ] in their respective `*_test.go` for unit tests.
  - [x] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
  - [ ] in `web_src/js/*.test.js` if it can be unit tested.
  - [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).

### Documentation

- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.

### Release notes

- [ ] I do not want this change to show in the release notes.
- [x] I want the title to show in the release notes with a link to this pull request.
- [ ] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/10588
Reviewed-by: Cyborus <cyborus@disroot.org>
Co-authored-by: Mathieu Fenniak <mathieu@fenniak.net>
Co-committed-by: Mathieu Fenniak <mathieu@fenniak.net>
2025-12-26 20:03:12 +01:00