Commit Graph
1312 Commits
Author SHA1 Message Date
Shivansh Gargandsilverwind bf5c39b310 feat(org): add DEFAULT_TEAM_VISIBILITY instance setting (#40151)
Adds an instance-level `DEFAULT_TEAM_VISIBILITY` setting
(`private`/`limited`/`public`, default `private`) so administrators can
configure the default visibility for newly created teams when it isn't
explicitly chosen in the web form or sent via the API. Also fixes a
template bug where the "Private" radio was always force-checked on the
new-team page regardless of the actual default.

Fixes: https://github.com/go-gitea/gitea/issues/40128

## Verification

Tested all three config values (absent/`private`, `public`, `limited`)
against both the web form default and the API (`POST /orgs/{org}/teams`
without `visibility`) — in each case the correct default was applied.

## Test 1: Key absent in app.ini -> falls back to "private"

<img width="3178" height="1312" alt="image"
src="https://github.com/user-attachments/assets/7ce87372-eb3e-4f03-84da-fc21ace86eca"
/>

## Test 2: DEFAULT_TEAM_VISIBILITY = public
<img width="2802" height="1348" alt="image"
src="https://github.com/user-attachments/assets/aa1af297-9557-434d-82c6-1af253eec2ab"
/>


## Test 3: DEFAULT_TEAM_VISIBILITY = limited


<img width="2778" height="1198" alt="image"
src="https://github.com/user-attachments/assets/1e66565b-c275-454a-a7d1-66d11ef3b9ee"
/>

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-17 08:15:45 +00:00
silverwind 1424a2b7c2 fix(packages): replace blakesmith/ar with an in-tree reader (#40401)
https://github.com/blakesmith/ar is unmaintained since 2019 and panics
on a negative member size (https://pkg.go.dev/vuln/GO-2026-6905), with
no fixed version. Debian packages only need a small subset of the `ar`
format, so this reads it in-tree.

- Malformed or truncated archives are rejected with a 400 instead of a
500 or a panic
- Parsing stops after the first `control.tar` member, deb(5) allows only
one

Assisted-by: Claude Code:claude-opus-5-5
2026-11-17 03:14:40 +00:00
silverwind 055105d42f fix(web): match settings page titles to the UI (#40336)
Many settings pages had a generic or missing title, so browser tabs and
history named the wrong page. Titles now use the active sidebar label,
or the page header on sub-pages, and are set once where several handlers
render the same page. The user and org webhook list headers print the
title, so they now read "Webhooks" instead of "Settings".

| Page | Before | After |
| --- | --- | --- |
| Runners, secrets, Actions general, admin job queue | Actions |
Runners, Secrets, General, Job queue |
| User and org webhooks, org options, 2FA enroll, account and keys after
a form error | Settings | Webhooks, Organization, Enroll in Two-Factor
Authentication, Account, SSH / GPG Keys |
| Admin self check, OAuth2 app edit, repo public access, repo LFS
pointer files | (none), owner/repo | Self Check, Edit OAuth2
Application, Public Access, Find pointer files |
| Admin trace, new branch protection rule, package cleanup rules |
Monitoring, Branch Protection -, Packages | Trace, Add New Rule,
Add/Edit Cleanup Rule, Cleanup Rule Preview |
| Admin webhook add and edit | Add/Update Webhook | Add/Update System
Webhook, Add/Update Default Webhook |
| Sidebar on admin queue, admin webhook edit, user OAuth2 app edit | no
highlight | Queues, Webhooks, Applications |

Assisted-by: Claude Code:claude-opus-5-5
2026-11-17 02:41:49 +00:00
8b48b45def fix: use button to replace anchor elements without href (#40338)
Elements `<a>` without `href` are not focus-able, screen readers didn't
expose it as a button.

Use `<button type="button">` for these elements, and fix what the
conversion exposed:

- Disabled reactions and the copy-content button use `aria-disabled`,
because native `disabled` suppresses their tooltips
- The random label color button gets a label
- Hover, disabled and alignment styles that only matched `<a>` apply to
the buttons
- Unpinning an issue with the keyboard works, and removing a
non-permitted "blocked by" dependency sends the right type

Fixes: https://github.com/go-gitea/gitea/issues/40337

Assisted-by: Claude Code:claude-opus-5-5

---------

Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: silverwind <me@silverwind.io>
2026-11-16 17:48:41 +00:00
bircniandsilverwind 3f1b738f09 enhance(audit): record remaining security-relevant events (#40323)
Adds audit actions for security-relevant changes that were not logged at
all, and records a few remaining events on paths that skipped them:

- Account login prohibition (`user:status:prohibitlogin`) and the
organization setting that lets repository admins change team access
- Actions runner registration tokens issued, runners removed or
enabled/disabled, and Actions variables, for every scope
- Instance config changes, including maintenance mode
(`system:config:update`)
- Deploy token regeneration and pull mirror address changes
- Review deletion and dismissal, package removal, and release and tag
removal
- Repository restores via `gitea admin restore-repo`, OAuth2 secret
rotation on API update, and SSH keys added or removed by authentication
source sync

The authentication source SSH key sync moved from the asymkey model to
the asymkey service so it can record events.

Follow-up to https://github.com/go-gitea/gitea/pull/40317 and
https://github.com/go-gitea/gitea/pull/40322.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-16 00:10:54 +00:00
bircniandsilverwind c144668e59 fix(audit): record missing user and authentication events (#40317)
Several entry points change users, credentials or 2FA without writing an
audit event, while the equivalent web or service path does. This records
the existing actions on those paths:

- `user:create` for self sign-up, OAuth2 auto-registration, link-account
registration and `gitea admin user create`
- `user:accesstoken:add` for `gitea admin user generate-access-token`
and `user create --access-token`; `user:accesstoken:remove` for `DELETE
/api/v1/token`
- `user:twofactor:disable` for the admin "reset 2FA" option and `gitea
admin user disable-2fa`
- `user:key:ssh:add`, `user:key:gpg:add/remove` for the API
- `user:email:primary`, `user:email:activate`, `user:status:active` for
settings, the activation page/links and OAuth2 source sync deactivation
- `user:oauth2application:grant` when a grant's scope changes,
`user:externallogin:add` on registration via an external login
- `user:authentication:fail:twofactor` for wrong recovery codes, failed
WebAuthn assertions and 2FA during password reset

Should be backported to v28.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-15 20:42:12 +00:00
e5b1ac76f9 fix(audit): record missing repository and content events (#40322)
Several API and background paths change repositories without writing the
audit event the web UI writes for the same change. This records the
existing actions on those paths:

- Webhook add/update/remove via the API (repository, organization, user
and system hooks)
- Branch and tag protection add/update/remove via the API
- Deploy keys and deploy tokens via the API, and deploy tokens generated
in the web UI
- Push mirror add, archive/unarchive and default branch changes via the
API
- Project create/update/delete via the API, and closing/reopening
projects in the web UI
- `repository:create` for migrations, `pr:merge` for scheduled
auto-merges, `pr:comment:create` for reviews (with `review_id` and
`review_type`) and review code comments
- `repository:delete` for repositories removed by the missing-repository
cleanup or an owner purge

Webhook, secret and OAuth2 application events now also keep the
impersonator.

Assisted-by: Claude Code:claude-fable-5-2

---------

Signed-off-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: delvh <dev.lh@web.de>
2026-11-15 13:49:58 +00:00
bircni 9bc19f2823 fix(api): do not allow renaming the Owners team (#40381)
The API accepted a new name for an organization's Owners team. The owner
team is looked up by its name, so the organization had no owner team
afterwards and its owners lost access to it.

The API now rejects the rename, as the web UI already does.

Fixes #40380
2026-11-15 12:50:48 +00:00
6579749a02 fix(repo): do not check out existing repo content during auto-init (#40376)
Plain repository creation marked a new repository as ready before its
git data, hooks and initial commit existed, and auto-init cloned the new
repository to build its initial commit on top of whatever it already
contained.

- Auto-init builds the initial commit in a freshly initialized temp
repository instead of a clone
- New repositories are inserted as being created and marked ready once
initialized, as fork, template generation, adoption and migrations
already do

---------

Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: silverwind <me@silverwind.io>
2026-11-15 11:31:40 +00:00
bircniandwxiaoguang 0a66642da0 fix(repo): limit the commit status fields returned by repo search (#40377)
The web repo search response embedded the whole commit status model
although the dashboard only uses its state and target URL. It now
returns just those two fields.

---------

Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
2026-11-15 11:27:05 +00:00
silverwindandZettat123 3317a31c22 fix(actions): resolve $/ steps against the job's workflow source (#40075)
GitHub resolves `uses: $/{path}` against the repository and commit of
the job's workflow file, not the triggering ones. When those differ, as
with reusable workflows from elsewhere or owner and global workflows:

- such steps are pinned to `self:{repo}/{path}@{sha}` in the runner
payload
- `$/` steps with expressions in their path fail the run as an invalid
workflow, as they can't be pinned safely and GitHub rejects them too

Same-source jobs keep `$/`, so older runners are unaffected.

Assisted-by: Claude Code:claude-opus-5-5

---------

Co-authored-by: Zettat123 <zettat123@gmail.com>
2026-11-14 07:15:46 +00:00
Basavaraj S mandsilverwind eaa4fe918d fix(api): hide org roles from non-members (#40042)
`GET /users/{username}/orgs/{org}/permissions` returned a user's team
roles to anyone who could see the org, revealing concealed members and
their roles. Now only the user, unrestricted org members and site admins
get the real answer, everyone else gets the non-member response:

- Outsiders get no roles, for public members too.
- Restricted members only get their own roles.
- Public-only tokens get none, and site admins need an `all`-scoped
token.

Assisted-by: Claude Code:claude-opus-5-5

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-13 19:37:56 +00:00
1f614953a2 enhance: allow admins to force user activity visibility (#40002)
Adds `[service] FORCE_USER_ACTIVITY_VISIBILITY` so admins can force the
activity of all users to be `public` or `private`, e.g. for company
instances that require public activity.

- When the value changes, the next start overwrites the "Hide Activity"
setting of all users. New users get the forced value too.
- The "Hide Activity" checkbox is disabled with an explanatory tooltip
while forced, and `PATCH /api/v1/user/settings` rejects a conflicting
`hide_activity` with 422.

| Before | After |
| --- | --- |
| <img width="122" height="17" alt="before"
src="https://github.com/user-attachments/assets/d138acec-5d40-4a0a-881c-0097e5197a8f"
/> | <img width="258" height="46" alt="after"
src="https://github.com/user-attachments/assets/7e8f7d1b-3303-4473-8d3b-aa35ef8fb20d"
/> |

Fixes: https://github.com/go-gitea/gitea/issues/37961

Assisted-by: Claude Code:claude-opus-5-5

---------

Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
2026-11-13 17:32:29 +00:00
silverwindandClaude 53aafd7ee1 fix(actions): include previous values in edited payloads (#40179)
Actions `edited` payloads omitted `changes`, so workflows could not see
previous values. Issue/PR title and body edits now carry
`changes.title`/`changes.body`, base branch changes add
`changes.base.ref`/`changes.base.sha` (and `changes.ref` now holds the
old branch, not the new one), release edits carry
`changes.name`/`changes.body`.

_Written by Claude (Opus 5.5)._

---------

Co-authored-by: Claude (Opus 5.5) <noreply@anthropic.com>
2026-11-13 08:43:13 +00:00
Rafail Giavrimisandsilverwind 38389f00be feat: unify code search, add search API (#39403)
Adds GitHub-style code search endpoints and makes code search behave the
same in the web UI and the API, whether it runs on git grep, bleve or
Elasticsearch.

- `GET /repos/{owner}/{repo}/code/search` searches any branch, tag or
commit with git grep, so it works without the code indexer.
- `GET /search/code` searches the default branches of all accessible
repositories with the code indexer.
- One query syntax everywhere: every term must match a file's content or
path, case-insensitively. It supports `"phrases"`, `/regexp/` (git grep
only) and the `path:`, `language:`, `repo:`, `user:` and `org:`
qualifiers. Invalid queries return an error.
- All engines apply the same include, exclude, vendored and size rules,
and return every match. The web UI shows them as multiple snippets per
file with the matched text highlighted.

<img width="359" height="295" alt="image"
src="https://github.com/user-attachments/assets/59b7aa66-b1dc-4938-9ef9-66668ae08d7c"
/>

Fixes: https://github.com/go-gitea/gitea/issues/10706
Fixed: https://github.com/go-gitea/gitea/issues/15841
Fixes: https://github.com/go-gitea/gitea/issues/31375
Related: https://gitea.com/gitea/gitea-mcp/issues/280
Docs: https://gitea.com/gitea/docs/pulls/615

AI usage disclosure: Yes, I used AI to write code, review changes and
verify the factuality of the overview I provided. I take full
responsibility for the code in this PR.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-13 07:24:37 +00:00
dc9a2aefdc fix(actions): keep runner_name after ephemeral runner removal (#39725)
Jobs that ran on ephemeral runners lose `runner_name` in the
`workflow_job` completed webhook and the jobs API, because the runner
row is deleted when the task finishes. GitHub keeps both `runner_id` and
`runner_name`, and GARM ignores completed jobs without a name, holding
the runner slot until its periodic cleanup.

The name is now stored on the task when claimed, and a migration
backfills tasks whose runner row still exists.

---------

Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
2026-11-13 01:18:57 +00:00
silverwind 9f93e8d808 test: add e2e tests for comments, file tree, actions and branch protection (#40313)
Add e2e tests for flows that had no browser coverage: editing and
quote-replying an issue with an uploaded image, file tree and go-to-file
navigation, the Actions run view and workflow dispatch, and branch
protection rules with their priority.

The tests exposed bugs that are fixed here:

- Quote reply into an empty comment box left the Comment button disabled
- Removing an uploaded attachment did not update the comment buttons
- Going back from a file opened via the file tree to the initially
loaded page kept showing the other file

Assisted-by: Claude Code:claude-opus-5-5
2026-11-12 07:12:55 +00:00
Shivansh Gargandwxiaoguang aa073fb1be fix(web): avoid compare output route collisions (#40307)
Allow compare refs named diff or patch to use the normal compare page
while preserving explicit .diff and .patch downloads.

Compare refs ending in diff or patch were incorrectly matched as raw
.diff/.patch routes. This caused normal compare pages such as
main...diff and main...patch to return an empty or invalid response.

The route matching is now constrained so valid raw diff/patch downloads
continue working while branches named diff or patch use the normal
compare page.

Assisted-by: Agent:GPT-5.6 Luna

Fixes https://github.com/go-gitea/gitea/issues/40303

---------

Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
2026-11-11 17:53:39 +00:00
wxiaoguangandsilverwind b3422c0920 refactor: clean up error system (#40328)
1. `ctx.ServerError("LogMessage", err)` is very messy and useless. When
the log is printed, it already contains the caller's file and line
number
2. `ctx.NotFound(nil)` and `ctx.NotFound(err)` were abused a lot, now
only `ctx.NotFound()`

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-11 17:07:57 +00:00
silverwind 03fc2bdcc8 fix(repo): require the unit in AccessibleRepositoryCondition (#40059)
`AccessibleRepositoryCondition(user, unitType)` now requires the repo to
have the requested unit enabled and not globally disabled, consistent
with the per-repo permission checks. This affects code search, Actions
run listings, LFS object reuse, recently pushed branches and org project
milestones.

Written by Claude.
2026-11-11 12:55:04 +00:00
silverwind d399f844e6 test: fix the login tab order e2e test in macOS WebKit (#40316)
The login tab order test fails in WebKit on macOS, so `make test-e2e`
fails locally on every Mac. CI on Linux is unaffected.

- macOS WebKit moves Tab focus only between text fields, plus buttons
and checkboxes when the system keyboard navigation setting is on, and
never to links, like Safari. Option+Tab reaches every element there and
acts as Tab in Linux WebKit, so the test presses Alt+Tab in WebKit.
- Headless WebKit applies the username `autofocus` late, which pulled
focus back from the password field. The test now waits for the autofocus
first, which also covers it.

Follow-up to https://github.com/go-gitea/gitea/pull/40081

Assisted-by: Claude Code:claude-opus-5-5
2026-11-10 20:07:28 +00:00
Gaurav Dubeyandsilverwind d5ce1cad7f fix(indexer): reindex issues when their projects change (#40299)
With the bleve, elasticsearch or meilisearch issue indexer, a keyword
search filtered by project matched issues by the projects they had when
they were last indexed, so newly added issues were missing and removed
ones still showed up.

- Reindex an issue when its projects change, using `IssueChangeProjects`
from https://github.com/go-gitea/gitea/pull/30205.
- Reindex the issues of a deleted project, so the "No project" filter
finds them again.

Entries that are already stale are corrected on the issue's next change
or an index rebuild.

Fixes: https://github.com/go-gitea/gitea/issues/33451

Assisted-by: Claude Code:claude-opus-5-5

---
🤖 Opened with [Superhuman](https://github.com/gaurav0107/superhuman), an
open-source contribution agent.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-10 19:26:12 +00:00
Yonghun Shin 2ef5b7623b fix(packages): generic package deletion ignores terraform state lock and latest version (#40302)
`IsLocked` and `IsLatest` in `services/packages/terraform/state.go`
return early when the package **is** a terraform state (`==` instead of
`!=`), so the `OnBeforeRemovePackageAll` /
`OnBeforeRemovePackageVersion` guards never fire. As a result, deleting
through the generic package API (`DELETE
/api/v1/packages/{owner}/terraform/{name}[/{version}]`) or the web UI
ignores the state lock and allows removing the latest state version,
while the terraform state endpoints reject both.

Changes:

- Flip the type check in `IsLocked` and `IsLatest`.
- `DeleteState` (terraform state `DELETE` endpoint) relied on the bug:
after checking the lock it removed every version one by one with
`RemovePackageVersion`, including the latest. With the guard fixed, it
failed with 500 partway through and left a half-deleted state (the
existing `TestPackageTerraform` subtests catch this). It now calls
`RemovePackage`, which checks the lock and deletes everything in one
transaction.
- The generic package delete API used `APIErrorInternal` for every
error, so the guard errors became 500. It now uses `APIErrorAuto`, which
returns 422 for these.
- Added a `GenericPackageAPI` subtest to `TestPackageTerraform`:
deleting an older version works, deleting the latest version or a locked
state returns 422, and deleting an unlocked state works. It fails on
current `main`.

Tested locally with `go test -run '^TestPackage' ./tests/integration`
(SQLite): all pass.

AI disclosure: I used an AI assistant (Claude) to find the bug, trace
the code paths, and draft the change and this description. I reviewed
the change and ran the tests myself.

Signed-off-by: choam2426 <choam2426@gmail.com>
2026-11-10 17:10:59 +00:00
Gaurav Dubeyandsilverwind 1fcb6b4cd6 fix(packages): rebuild repository index after removing a package (#40306)
Deleting a package or a package version from the web UI, the admin panel
or `DELETE /api/v1/packages/...` removed the database rows but never
rebuilt the Debian, Alpine, RPM or Arch repository index, so clients
kept seeing the deleted version until something else triggered a
rebuild. Only the registries' own delete endpoints and the cleanup task
rebuilt it.

https://github.com/go-gitea/gitea/pull/22810 was closed in favor of
`OnBeforeRemovePackageVersion`, but that hook runs before the version is
gone, so it can't refresh the index. This adds an `OnAfterRemovePackage`
hook and rebuilds the owner's repository files there, the same way the
cleanup task does.

Deleting the last RPM of a group also removed the repodata of all other
groups of the owner. Now only the emptied group's files are removed.

Fixes: https://github.com/go-gitea/gitea/issues/32830
Fixes: https://github.com/go-gitea/gitea/issues/40304
Fixes: https://github.com/go-gitea/gitea/issues/40305

---
🤖 Opened with [Superhuman](https://github.com/gaurav0107/superhuman), an
open-source contribution agent.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-10 16:29:39 +00:00
wxiaoguangandsilverwind d1d702bea9 refactor: user error system (#40308)
Introduce "FlashErrorAuto" for "XxxErrorAuto" family and simplify the
error handling logic.

* Use ErrorUnwrapForHTML for web error messages 
* Use ErrorUnwrapForUser for api error messages
* Make Flash message can take up to about 3000 bytes

Next step: completely refactor "ctx.ServerError"

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-10 16:29:36 +00:00
Shivansh Gargandwxiaoguang 640cd615c4 feat: restrict pull/issue creation on access basis (#40069)
Adds a repository-level **"Restrict to collaborators"** option under the
Issues and Pull Request unit settings. When enabled, only users with
collaborator access to the base repository (owner, team members, or
explicit collaborators) can open issues or pull requests against it.

Public repositories often need to be visible/cloneable by everyone, but
owners may want to limit who can open pull requests (e.g. restrict to an
internal team while still keeping the code public)

* Fixes: https://github.com/go-gitea/gitea/issues/40015
* Fixes: https://github.com/go-gitea/gitea/issues/31890
* Fixes: https://github.com/go-gitea/gitea/issues/28021
* Fixes: https://github.com/go-gitea/gitea/issues/14801

---------

Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
2026-11-10 00:53:49 +00:00
silverwindandLunny Xiao 3c41728017 test(e2e): assert server-rendered biography after profile update (#40176)
Since Playwright 1.56, a WebKit click on a native submit button can
return before the form navigation starts, so `toHaveValue` passed on the
old page with the value `fill()` just set, and the next step raced the
reload.

Assert the textarea's text instead, its server-rendered default, which
only the reloaded page has.

---------

Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
2026-11-09 21:50:27 +00:00
Lunny Xiaoandsilverwind 3561478109 fix(projects): move only the dragged card on project boards (#40182)
The board reported every card of the target column on each drag, and the
server wrote all of them into that column. A stale board in another
browser therefore pulled cards moved by someone else back to their old
column.

The client now sends only the dragged issue and its predecessor, and the
server computes the position in one transaction:

- Only the dragged issue changes column, the others are just renumbered.
- A predecessor that is no longer in the column puts the card at the
end.

Added an e2e test.

Fixes https://github.com/go-gitea/gitea/issues/21243
Fixes https://github.com/go-gitea/gitea/issues/32070
Fixes https://github.com/go-gitea/gitea/issues/13498

---
Assisted-by: Codet:claude

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-09 21:39:22 +00:00
Lunny Xiao bccc90d654 fix(api): lock and unlock issues as the signed-in user, not the repository owner (#40291)
`PUT/DELETE /repos/{owner}/{repo}/issues/{index}/lock` passed
`ctx.ContextUser` as the doer of `LockIssue` / `UnlockIssue`. In the
repo routes that is the owner of the repository, so on a repository of
an organization the lock and unlock timeline comments were created by
the organization instead of the user who did it. The web handler already
uses the signed-in user; this uses `ctx.Doer` as well.

Adds a subtest `DoerOnOrganizationRepo` to the existing
`TestAPILockIssue`: a member of an organization locks and unlocks an
issue of an organization repository, and the comments must be posted by
that member.

Found while testing audit logging of the issue lock events downstream.

_Authored by Codet (Claude)._
2026-11-09 18:00:06 +00:00
wxiaoguang 772f365254 refactor: use RunWithStderr for git commands (#40295)
If the stderr output is not handled, then it needs to use RunWithStderr
to make sure the git error message is included in the returned error.
Otherwise, when the git command fails, the error log won't show useful
information.

Ref:

* #39795 : no useful information in log messages
*
https://github.com/go-gitea/gitea/pull/40160#pullrequestreview-5752720866
2026-11-09 17:59:52 +00:00
bircniandsilverwind e0b2f3425e fix(actions): run pull_request workflows on the merge commit (#40076)
`pull_request` workflow runs checked out the PR head, so a PR could pass
CI and still break the base branch once merged.

Gitea now maintains `refs/pull/<index>/merge`, a test merge of the head
into the base branch, for open pull requests. It is updated on creation
and mergeability checks and removed on close or merge.

- `pull_request` runs read their workflows from the test merge and
expose it as `github.ref`, `github.sha` and
`github.event.pull_request.merge_commit_sha`. Commit statuses, the API's
`head_sha`, `workflow_run` payloads and `paths` filters stay on the
head.
- Runs fall back to the head when the merge conflicts, the pull request
is closed or git is older than 2.40.
- Hidden refs keep each run's test merge (`refs/gitea/run/<id>`) and
merged commits (`refs/gitea/pull/<index>/merged`) from `git gc`, for
reruns and after the base branch is deleted or force-pushed.

Fixes: https://github.com/go-gitea/gitea/issues/27039
Fixes: https://github.com/go-gitea/gitea/issues/13134
Fixes: https://github.com/go-gitea/gitea/issues/30972
Docs: https://gitea.com/gitea/docs/pulls/617

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-09 12:42:40 +00:00
silverwind d715a5ae75 test(api): cover null milestone due_on (#40172)
Removing a milestone deadline with `"due_on": null` was untested, only
`""` was. On release/v28 the null case returned HTTP 500
(https://github.com/go-gitea/gitea/pull/40171). main already handles it,
this adds the case to the existing test so it stays that way.
2026-11-08 20:53:59 +00:00
silverwind 8cb6fb2001 fix: add missing checks to more handlers and services (#40131)
Several code paths skipped checks that their siblings already perform.

- Code comment replies are rejected when they target a review of another
pull request or another user's pending review
- OAuth2 refuses tokens for users who can't sign in, and introspection
reports their tokens and already used refresh tokens as inactive
- AGit proc-receive runs the same pull request permission and wiki
checks as pre-receive
- Repository transfer drops the repository's owner-level scoped workflow
source registrations, and registrations whose source the owner no longer
owns are ignored
- The web editor's base branch copy into a fork and the fork's upstream
divergence info require read access to the base repository
- LFS downloads require sign-in under strict `REQUIRE_SIGNIN_VIEW`, like
git smart HTTP
- Unknown AGit target branches are reported as an unexpected ref again
2026-11-08 09:02:51 +00:00
silverwind cd92c6a008 fix: paste copied code as text, not as a markdown table (#40053)
Copying lines from a diff, file view, blame, code search or code preview
and pasting them into a comment on Gitea or GitHub produced a markdown
table with escaped pipes and stripped indentation. Both use
`@github/paste-markdown`, which converts copied HTML tables unless they
carry `data-paste-markdown-skip`, as GitHub's own code tables do.
Gitea's code tables now set it too.

The test is a bit more complicated because of a upstream bug which
causes `paste-markdown` to insert pasted content twice, once
https://github.com/github/paste-markdown/pull/128 is merged and
released, the negative assertion could be removed.

---------

Signed-off-by: silverwind <me@silverwind.io>
2026-11-08 09:01:50 +00:00
c54ec52f0f fix(auth): limit failed TOTP attempts per account (#39993)
Gitea accepted unlimited wrong TOTP passcodes, so anyone holding an
account's password could enumerate the 6-digit code in minutes. This
allows 5 failed attempts per account, then refuses TOTP until 15 minutes
after the last one.

- The count lives in the cache, and the global lock spans the check,
validation and counting, so new sessions and concurrent requests cannot
widen it. With Redis, both are shared across instances
- Applies to every TOTP check: the web 2FA page, account recovery and
Basic auth with `X-Gitea-OTP` (API, git, packages)
- Only wrong or replayed passcodes count, server errors do not
- Scratch codes and WebAuthn stay usable as the recovery path while
locked
- Failures log `Failed authentication attempt for <user> from <ip>`,
which the fail2ban filter already matches

---------

Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
2026-11-07 23:01:49 +00:00
Lazizbek Ergashevandsilverwind 884884308d fix(pull): don't add a timeline entry for pending API reviews (#40260)
Pending reviews created through the API went through `SubmitReview`,
which added a timeline entry and sent notifications before the review
was submitted, so submitting it later showed the review twice.

- Pending reviews skip the timeline entry and notifications until
submitted, like on GitHub
- `CreatePullReview` rejects comments without a `path` or a positive
line with 422, like GitHub
- Already duplicated reviews stay as they are

Fixes https://github.com/go-gitea/gitea/issues/40259

Assisted-by: Claude Code:claude-opus-5-5

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-07 14:19:50 +00:00
Renoxandwxiaoguang 462ab3801a fix(web): don't block README images with REQUIRE_SIGNIN_VIEW = expensive (#40264)
Relax expensive endpoint list, then README can render its images.

---------

Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
2026-11-07 13:34:01 +00:00
Mitrahsoftandsilverwind 3aca9c8fb7 enhance(actions): create commit status for workflow_dispatch runs (#39961)
Runs triggered by `workflow_dispatch` never created a commit status, so
the commit showed no sign that a workflow ran on it.

- Post the status on the run's commit, e.g. `ci / e2e
(workflow_dispatch)`. `schedule` runs stay unchanged.
- These statuses don't count for pull requests. They don't show in a
PR's checks and can't satisfy or fail its required checks, even through
broad globs like `ci / e2e*`.

Closes https://github.com/go-gitea/gitea/issues/39639

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-07 12:55:19 +00:00
Trung Heheandsilverwind 6bc92f464a fix(org): enforce unique team names in the database (#40228)
Concurrent requests could create teams with the same name in one
organization, because only the service checked the name. Editing such a
team, or renaming a team to a taken name, then failed with a 500.

- Add a unique index on `team (org_id, lower_name)`. Migration 368
renames every duplicate except the oldest to a free `<name>-<n>` and
logs each rename. IDs, members, repositories and permissions are kept,
but lookups by the old name (CODEOWNERS, group-to-team sync, mentions,
URLs) now reach only the kept team.
- Team writes report a name rejected by the index as
`ErrTeamAlreadyExist`, so duplicates, concurrent ones included, return
409 from `POST /orgs/{org}/teams` (previously 422) and `PATCH
/teams/{id}`, and a form error in the web UI.

Fixes https://github.com/go-gitea/gitea/issues/39632

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-07 11:16:11 +00:00
5d9bfa1e14 feat(packages): add private Terraform module registry (#38040)
Adds a Terraform module registry package type implementing the [module
registry
protocol](https://developer.hashicorp.com/terraform/internals/module-registry-protocol),
so `terraform init` and Terragrunt `tfr://` can resolve modules from
Gitea:

```hcl
module "vpc" {
  source  = "gitea.example.com/owner/vpc/aws"
  version = "~> 1.0"
}
```

- Modules are uploaded with `PUT
/api/packages/-/terraform/modules/{owner}/{name}/{provider}/{version}`
as a `.tar.gz` with the module at its root, stored verbatim and
immutable.
- `X-Terraform-Get` is a short-lived signed URL because Terraform sends
no credentials when downloading the archive.
- The package page shows install snippets, README, the root module's
inputs, outputs and providers, and `modules/*` submodules.
- Metadata is read by a small in-tree HCL scanner instead of
`hashicorp/hcl/v2`, which would add 5 modules and ~1.4 MB to the binary.

Docs: https://gitea.com/gitea/docs/pulls/433

---------

Signed-off-by: Michael Lopez <mickael.lopez@gmail.com>
Co-authored-by: TheFox0x7 <thefox0x7@gmail.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: silverwind <me@silverwind.io>
2026-11-07 10:45:05 +00:00
b23c70d439 fix: add missing checks to several handlers and services (#40120)
Several code paths skipped checks their siblings perform, or cleaned up
more than they validated.

- Official team review requests now count, and removing one no longer
marks an unrelated review official
- Release edits only delete files of validated attachments
- Rejected container uploads no longer delete an existing blob
- Attachment removal only removes the uploader's own pending uploads
- OpenID sign-in checks the allow and block lists against the normalized
identifier
- HTTP git access is rejected while a repository is being created, like
SSH

---------

Signed-off-by: silverwind <me@silverwind.io>
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
2026-11-07 09:43:39 +00:00
Mitrahsoftandsilverwind b6c291f87a fix(api): honor branch protection force push and allowlist options on create and edit (#39990)
The branch protection API handled push and force push options
inconsistently. Create dropped force push unless push was enabled, and
edit ignored allowlist fields unless their parent was set to true in the
same request.

- Create treats force push independently of push, like edit and the web
UI. Enforcement is unchanged.
- Edit applies `enable_push_whitelist`, `enable_force_push_allowlist`
and both deploy key flags against the rule's resulting state, like the
deletion fields. `{"enable_push_whitelist": false}` now disables the
whitelist instead of being ignored.
- Documents that `enable_force_push_allowlist` requires
`enable_force_push` on create.

Fixes https://github.com/go-gitea/gitea/issues/35121

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-07 09:34:21 +00:00
Zettat123andsilverwind c74f2f9216 fix(repo): mark repository as non-empty before git push returns (#40102)
The post-receive hook syncs pushed branches to the database, but
`is_empty` and the default branch of an empty repository were only
updated later by the `push_update` queue. Until then the API treated the
repository as empty (branch list `null`, `GET /branches/{branch}` 500),
and the queue overwrote the default branch chosen by the hook or set via
the API.

`SyncBranchesToDB` now initializes an empty repository and repairs a
default branch missing in git in its locked transaction, so the state is
final when `git push` returns. The push queue and web file edits retry
it while the repository is still empty.

- The default branch is the configured one if it exists, else `main`,
else `master`, else the first branch by name, like GitLab. HEAD follows
it.
- A first push of several branches no longer depends on push order. The
push that initializes a repository still prints no pull request hints,
like GitHub.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-07 03:20:25 +00:00
silverwind ede9ec3e10 fix: stop discarding file close and test setup errors (#39987)
- Fix
https://github.com/go-gitea/gitea/security/quality/rules/go%2Funhandled-writable-file-close,
ignored close errors hide failed flushes, which could lose rotated logs
and truncate authorized_keys
- Fix `TestProjects` fixture reset, from
https://github.com/go-gitea/gitea/security/quality/ai-findings,
`TruncateBeans` rejected the table names and the ignored error let the
test run against fixture data
2026-11-06 20:45:17 +00:00
silverwind 24700dca2a test(e2e): retry clipboard reads rejected by Firefox (#40155)
The Firefox mermaid e2e test fails intermittently with `Clipboard read
operation is not allowed.`, for example in
https://github.com/go-gitea/gitea/actions/runs/41281998298/job/124776072700.

Firefox rejects `readText()` when it overlaps the page's still pending
clipboard write, and `expect.poll` does not retry errors thrown by its
callback, so one overlapping read fails the test.
https://github.com/go-gitea/gitea/pull/39838 only removed the paste
prompt, not this rejection.

- A new `assertClipboardText` helper treats a rejected read as a
mismatch so it is polled again
- The bot token copy test had the same race and uses it too

_Written by Claude (Opus 5.5)._
2026-11-06 20:44:00 +00:00
silverwind 7e18e81c7d refactor: remove duplicated and dead Go code (#40244)
Pure refactors with no behavior change, mostly reusing existing helpers
and dropping duplicated or dead code. Net −222 lines.

- Reuse existing helpers instead of inlined copies: `getIssueIndex`,
`NewLabel`, `ToCommitUser`, `GetPackageFileDescriptors`,
`util.SplitTrimSpace`
- Merge copy-pasted blocks in review comment deletion, project deletion
and the pre/post-receive hook templates
- Drop dead code: an error `toStringKeys` never returned, impossible
`bytes.Buffer` write errors, redundant nil-map branches, a `base64`
wrapper
- Use stdlib where it fits: `strings.Cut`, `bytes.IndexAny`,
`cmp.Compare`, `min`/`max`, `slices.Concat`, map lookups instead of key
scans
2026-11-06 11:59:49 +00:00
Shivansh Gargandsilverwind a5201dc427 feat(release): allow setting a release as the latest (#40214)
Gitea picks the latest release by commit date, while GitHub keeps an
explicit pointer, so repositories migrated from GitHub could show a
different latest release than their source. Fixes
https://github.com/go-gitea/gitea/issues/35273.

- Add a "Set as Latest Release" checkbox and GitHub's `make_latest` API
option. Without them, the newest stable release by commit date stays the
latest.
- Migrations from GitHub and Gitea carry over the source's latest
release.
- Migrated releases are dated by their commit like native ones, and
equally dated ones keep the source's order.

<img width="355" height="122" alt="Set as Latest Release checkbox"
src="https://github.com/user-attachments/assets/d35f5324-7ae0-4e9f-b73c-d3c90496fb00"
/>

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-06 08:23:13 +00:00
Lazizbek Ergashevandsilverwind 98c937170f fix(repo): reject unparsable required approvals in branch protection (#40213)
Saving a branch protection rule with a "Required approvals" value that
can't be parsed as an integer (like `99999999999999999999` or `abc`)
showed the success message and stored 0, because the binding error was
never checked. The form now uses `form-fetch-action` with
`GetFetchActionForm`, so invalid input is rejected and nothing is saved.
Dev mode still answers unparsable numbers with a 500 until
`TypeCastError` gets a message, see
https://github.com/go-gitea/gitea/pull/40212.

- Validation errors show on the page and highlight the offending field,
keeping the form input
- Negative approvals are rejected before any rule lookup, and the input
gets `min="0"`

Fixes https://github.com/go-gitea/gitea/issues/40211

Assisted-by: Claude Code:claude-opus-5-5

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-06 08:08:20 +00:00
Basavaraj S mandsilverwind 8d14c9af20 fix(pull): enforce user blocks on pull request reviews (#40194)
Code review never checked user blocks, so a user blocked by the poster
or repository owner could still post reviews, code comments and replies
through the web, API and email.

- `CreateCodeComment` and `SubmitReview` now check blocks, with the repo
admin exemption issue comments have.
- Official reviewers blocked only by the author can still review, so an
author can't keep a maintainer from requesting changes.
- Review requests are handled in
https://github.com/go-gitea/gitea/pull/39296.

---------

Co-authored-by: silverwind <me@silverwind.io>
2026-11-06 07:31:58 +00:00
wxiaoguang fd98ab5472 enhance(codeowners): support section / glob / file view display / skip review request (#40243)
Document: https://gitea.com/gitea/docs/pulls/634 ([doc PR
preview](https://gitea.com/gitea/docs/src/commit/955ae7d75c3377a15fafe465d85e484fc8dc7d2a/docs/usage/repository/code-owners.md))

1. If the pattern starts with `*` or contains `/*`, use our glob package
to parse it as glob to regexp
* Make it easier to use. Otherwise it's really ugly to keep writing a
lot of backslash escapes
2. Add GitLab-style "section" support
   * Fix #28926

```codeowners
[Section] @user
*.js
```

3. Add our annotation support, at the moment only need to support
review-request=true/false
    * Fix #40241

```codeowners
## review-request=false # this is global

*.js $user ## review-request=true # this is line-level

[Section] @user ## review-request=true # this is section-level
*.md
```

4. Add GitHub/GitLab-style "show codeowner when viewing a file"
    * Fix #28137
2026-11-06 07:19:51 +00:00