Inline the web app manifest as a `data:` URL instead of serving it from
`/assets/site-manifest.json`. That endpoint has been linked through
`STATIC_URL_PREFIX` since https://github.com/go-gitea/gitea/pull/37405
and fails to load when:
1. a reverse proxy checks auth on every request, because browsers fetch
manifests without cookies
2. `STATIC_URL_PREFIX` serves the `public/` directory from disk, which
has no such file
3. `STATIC_URL_PREFIX` is cross-origin, because the endpoint sends no
CORS headers
The manifest is built per request so `start_url` follows
`PUBLIC_URL_DETECTION`. A reverse proxy that adds a stricter CSP must
allow `manifest-src data:`.
Fixes https://github.com/go-gitea/gitea/issues/40197
---------
Co-authored-by: silverwind <me@silverwind.io>
The job emitter resolves a dependency chain in several rounds in memory,
but a job's `if:` reads its needs' results from the database, where a
need skipped or failed earlier in the same pass was still blocked. A job
with `if: always() && needs.build.result == 'skipped'` was therefore
skipped instead of started.
Finished statuses are now persisted at the end of each round, which also
replaces the separate handling of skipped and failed reusable workflow
callers.
When a member of a CODEOWNERS team approves or rejects a pull request,
`SubmitReview` removes the team's review request. On the next push,
`PullRequestCodeOwnersReview` finds no request for that team and adds it
again, so every team member except the author gets a new "Requested to
review" mail. This happens even when the push doesn't change the PR's
diff (for example, merging the base branch into the PR branch) and the
approval is still valid.
Individual code owners don't have this problem: they are skipped when
they already have a review. This PR applies the same rule to teams. A
team is not requested again while one of its members has an approve or
reject review on the PR that hasn't been dismissed. If that review is
dismissed (for example by "Dismiss stale approvals"), the team is
requested again as before.
Testing: `TestPullView_CodeOwnerTeamAnsweredByMember` creates an org
repo with `README.md @org3/team1`, opens a PR, and checks three things:
the team is requested; after a member approves, a code owner pass (what
a push runs) doesn't request the team again; after the approval is
dismissed, the team is requested again. The test fails without the
change. `make lint-backend` and the `TestPull*` integration tests pass
locally on SQLite.
AI disclosure: I used Claude to help trace the cause and draft the
change, test and description. I reviewed and tested it myself and can
explain every line.
Fixes#34884.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Transferring a repository fired no `repository` webhook. It now sends a
`transferred` action shaped like GitHub's, with the previous owner in
`changes.owner.from.user` or `changes.owner.from.organization`. Like
GitHub, it goes to the new owner's webhooks, plus the repository's own
and system webhooks.
- Chat webhooks render "Repository transferred from <owner>".
- Actions workflows are unaffected, no trigger matches `repository`
events.
Fixes https://github.com/go-gitea/gitea/issues/33584
---------
Co-authored-by: silverwind <me@silverwind.io>
Reuse the request-scoped base git repository in `ToAPIPullRequest`,
`ToPayloadPullRequest` and `ToAPIPullRequests` instead of reopening it
for each conversion, so conversions share the request's `cat-file`
process.
- Same-repo PRs reuse the base repository for the head instead of
opening it a second time.
- Drop a redundant second open of the base repository.
- Fork head repositories stay opened per call. A request-cached fork
keeps its `cat-file` running until the request ends, which blocks
deleting the fork on Windows and keeps one process per fork alive in
batch status changes.
---------
Co-authored-by: silverwind <me@silverwind.io>
Fixes https://github.com/go-gitea/gitea/issues/34317
Removing a label from a pull request still started workflows with
`types: [labeled]` and never matched `types: [unlabeled]`.
https://github.com/go-gitea/gitea/pull/34316 derived
`labeled`/`unlabeled` from the added and removed labels for issues only,
while pull requests kept mapping every `label_updated` to `labeled`.
Both events now share one conversion, so they cannot drift apart again.
Two other label changes also triggered the wrong workflows:
- Clearing all labels started no `unlabeled` workflow, because the
Actions notifier ignored it
- Replacing labels reported every old label as removed, so a replace
that kept a label also matched `unlabeled`
AI disclosure: written with Claude Code (claude-opus-5-5), session
https://claude.ai/code/session_01Kw9ySGNd4ynmyX9GcVdUfw
---------
Co-authored-by: silverwind <me@silverwind.io>
Remove the `UnmarshalHandleDoubleEncode` runtime workaround for an old
xorm bug, and normalize legacy double-encoded JSON in `repo_unit.config`
and `login_source.cfg` with a migration instead.
- Config `FromDB` no longer fails on corrupted values. A broken auth
source is deactivated, so admins can still fix it in the admin panel
- Remove `ProjectsConfig.GetProjectsMode`, `FromDB` already defaults
`ProjectsMode`
Actually it doesn't need so many changes, just because AI thinks it
should handle "corrupted database value" for auth sources.
---------
Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: silverwind <me@silverwind.io>
Clearing "Full Name" in the profile settings shows success but keeps the
old value: since #31959 it is only updated when non-empty, because the
input is disabled (not submitted) when `change_full_name` is disabled.
Use `*string` for the field, so a missing field keeps the value and an
empty one clears it. With `change_full_name` disabled, any submitted
value (including empty) is now rejected. Not `optional.Option[string]`,
because `MaxSize` would count slice elements instead of characters.
Fixes#40185
---
AI disclosure: written with Claude Code (claude-opus-5-5) and tested
locally.
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: silverwind <me@silverwind.io>
A failed contributor stats generation never released its lock, so after
the cached error expired, the contributors and code frequency pages kept
saying stats are being generated until a restart. A generation
outlasting the 5 second wait also leaked its goroutine on an unread
channel.
The lock is now claimed atomically and released together with the
completion signal when generation ends, and a cache miss after the wait
counts as still generating.
Fixes https://github.com/go-gitea/gitea/issues/40037
`DELETE /repos/{owner}/{repo}/push_mirrors/{name}` only deleted the
`push_mirror` row and left the `remote_mirror_*` git remotes, which
embed the mirror credentials, in the repository and wiki config. Each
create/delete cycle through the API left more credentials behind.
- Remove the git remotes before deleting the row and record it in the
audit log, like the web settings handler.
- Return 404 for an unknown name, as documented, instead of 204.
- Ignore an already missing code remote, so the wiki remote is still
removed and such mirrors can be deleted through both the API and the web
settings.
---
AI disclosure: I used an AI coding assistant to help write this change
and the PR description. I reviewed the code and ran the tests myself.
---------
Signed-off-by: sarankumar1325 <120080287+sarankumar1325@users.noreply.github.com>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Pages showing commit statuses loaded the repository of every status
separately, so a PR with ~75 statuses ran ~75 identical repository
queries per merge box render (about 3.4s in a production trace).
`statusRepoCanReadActions` only loaded the repository inside the
permission cache loader, which never runs once `RepoAssignment` has
published the permission.
- `CommitStatus.loadRepository` resolves through the request context
cache
- `statusRepoCanReadActions` loads the repository before the permission
lookup
- `RepoAssignment` publishes its repository to the context cache next to
the permission
- `GetCommitActionsStatusMap` drops its own repository map
Queries per request with 50 statuses (45 Actions, 5 external) on the
head commit, SQLite:
| Page | Repository queries | Total queries |
|---|---|---|
| PR view | 61 → 11 | 157 → 107 |
| Commit | 6 → 5 | 24 → 23 |
| Branches | 6 → 5 | 33 → 32 |
| Commits | 6 → 5 | 23 → 22 |
| Repo home | 8 → 7 | 35 → 34 |
| `/pulls` | 15 → 15 | 37 → 37 |
---------
Co-authored-by: silverwind <me@silverwind.io>
Default-branch pushes and Actions unit saves recreated the repository's
schedules with `Next` recomputed from that moment, so a push between a
cron's due minute and the scheduler's pass skipped the occurrence, and
`@every` timers restarted.
Re-registration now updates schedule rows in place and keeps the spec
rows of unchanged crons, so their cursor survives and an in-flight
scheduler pass still finds its schedule. Re-detection also clears
schedules when no scheduled workflow remains.
Fixes https://github.com/go-gitea/gitea/issues/40062
Written with AI assistance.
---------
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Claude (Opus 5.5) <noreply@anthropic.com>
Large GitHub migrations failed once the API budget ran out. A rejected
request never updated the downloader's rate state, so go-github refused
every retry locally.
- Record the rate limit from error responses too, so the next request
waits for the reset
- Retry repository-wide comment pages like the other downloader calls
- Stop `waitAndPickClient` spinning when the rate refresh fails
- Skip reaction requests when the reaction rollup reports none
- Log a failed patch download instead of failing the migration
---------
Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Giteabot <teabot@gitea.io>
The post-receive hook wrote each push's new commit into the branch
table. Each push runs its own hook after the ref is updated, so with
pushes in close succession, like CI pushing back or a quick force-push,
a slower first hook can finish last. The table then keeps the older
commit, or marks a re-created branch as deleted.
- The sync reads each pushed branch's state from git under a repo-scoped
lock
- Deletions go through the same sync, so a re-created branch stays
active
- A branch that fails to sync is logged and skipped instead of aborting
the batch
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Giteabot <teabot@gitea.io>
The repository issue and pull request lists filter `access` by
`repo_id`, which no index covers, so every load scans the whole table.
Issue counts on larger repositories merge three single-column indices,
and the API pull request list loads the latest reviews with `id IN
(SELECT max(id) ... GROUP BY ...)`, which MySQL runs as a scan of every
review, about 1.1s on
https://gitea.com/api/v1/repos/gitea/gitea-mcp/pulls?limit=1.
- Index `access.repo_id`
- Index `team_repo.repo_id`
- Replace the single `repo_id`, `is_closed` and `is_pull` indices with
one `issue` index on `(repo_id, is_closed, is_pull)`
- Join the latest review ids instead of `id IN (subquery)`
Median times in ms on MySQL 8.4 with tables sized like gitea.com (1M
access rows, 500k issues, 200k reviews), for a copy of
https://gitea.com/gitea/gitea-mcp and a repository with 9.7k issues:
| Page | Before | After |
|---|---|---|
| `/{owner}/{repo}/issues` | 107 | 40 |
| `/{owner}/{repo}/pulls` | 106 | 32 |
| `/{owner}/{repo}/pulls?state=closed` | 153 | 35 |
| `/{owner}/{repo}/issues/{index}` | 141 | 60 |
| `/{owner}/{repo}/issues`, 9.7k issues | 303 | 53 |
| `/api/v1/repos/{owner}/{repo}/pulls?limit=50` | 180 | 46 |
| `/api/v1/repos/{owner}/{repo}/pulls?limit=1` | 137 | 16 |
| `/api/v1/repos/{owner}/{repo}/issues`, 9.7k issues | 248 | 43 |
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Giteabot <teabot@gitea.io>
Project board cards list the pull requests that reference their issue,
but loaded them without checking the viewer's access. A private pull
request mentioning a public issue showed its title, number and state on
that issue's board to anyone who can view the board, anonymous visitors
included.
Repository and organization boards now skip linked pull requests in
repositories where the viewer cannot read pull requests, matching how
the issue page already hides such references.
---------
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: Giteabot <teabot@gitea.io>
Adds migration from bitbucket.org as the Cloud variant of the Bitbucket
entry, in the web UI and via `POST /repos/migrate` with
`service=bitbucket`. It migrates the repository and pull requests with
their comments and reviews. Bitbucket Cloud has no releases, and
Atlassian removed its issues and wikis on 2026-08-20
(https://community.atlassian.com/forums/Bitbucket-articles/Announcing-sunset-of-Bitbucket-Issues-and-Wikis/ba-p/3193882).
- Private repositories need an API or access token, which also raises
the rate limit.
- Rate-limited requests back off and retry, Bitbucket sends no
`Retry-After`.
- Tests replay recorded responses, set `BITBUCKET_READ_TOKEN` to
re-record.
Co-authored-by: silverwind <me@silverwind.io>
Repository-level and instance-level runners still picked up waiting jobs
of repositories with Actions disabled.
https://github.com/go-gitea/gitea/pull/30331 only covered owner-level
runners, by filtering them out in SQL.
Pickup now checks the Actions unit of each job it is about to claim, for
every runner scope, and cancels the job instead when Actions is
disabled, also honoring `DISABLED_REPO_UNITS`. Disabling Actions
therefore cancels its queued jobs once a matching runner polls, instead
of leaving them waiting until `ABANDONED_JOB_TIMEOUT`.
Runs triggered by pull request reviews took their ref and commit from
the review's `commit_id` instead of the pull request. They now use the
pull request head like other pull request events, and their commit
status is set on the commit that ran.
Deleting an Actions run (UI, API or `RUN_RETENTION_DAYS`) left its
commit statuses linking to the deleted run, so "Details" gave a 404.
`DeleteRun` now clears the `target_url` of the run's statuses and of the
commit's status summary on the run's commit, in the same transaction.
States are kept, so required status checks and merging are unaffected.
This differs from GitHub, which deletes the checks together with the
run.
Fixes https://github.com/go-gitea/gitea/issues/40079
Co-authored-by: silverwind <me@silverwind.io>
Package files carry type specific properties like `debian.distribution`,
`debian.component`, `rpm.group` or `alpine.branch` that decide where a
file lives in a registry, but the API never returned them. A client
listing `GET /packages/{owner}/{type}/{name}/{version}/files` could not
tell which distribution or component a file belongs to, or tell apart
same-named files in different ones. Tools copying packages between
instances, like https://gitea.com/gitea/importer, need them to re-upload
a file to the same place.
This adds a `properties` map to `PackageFile`, limited to the properties
that locate a file. Internal ones like the rpm, alpine, arch and conda
metadata or the Debian control file stay hidden, as they can be many MiB
per file.
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Keep notification counts and stopwatches in sync when tabs open or
reconnect, closing the gap behind [this
flake](https://github.com/go-gitea/gitea/actions/runs/31312525323/job/93242429394).
1. Send current notification counts and stopwatch state on each
WebSocket connection, replacing the separate state fetches tabs made on
every connect.
2. Tabs joining an already open connection fetch current state once,
since its snapshot went out before they attached.
3. Send events queued during the handshake before the snapshot, so older
events can't roll it back.
4. Wait for the Redis SUBSCRIBE acknowledgement, so the snapshot read
after subscribing can't miss an event.
Several handlers and parsers skipped checks that their siblings already
perform.
- Reject issue, comment, label, milestone and release mutations in
archived repositories via the API with one guard per route group,
matching the web UI
- Hide timeline references and dependencies pointing at issues the doer
cannot read, in the web UI and the API, including pull requests of the
same repository without pull request access
- Reject workflow runs with more than `MaxJobNumPerRun` jobs, including
inline reusable workflow jobs, and show rejected pushes as invalid
workflow runs
- Bound how much of a Helm chart archive is inflated while looking for
`Chart.yaml`
---------
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
OneDev 16.8 moved the server version endpoint from
`/~api/version/server` to `/~api/server/version`. Both are tried, the
new one first, and both errors are reported when neither works.
---------
Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: silverwind <me@silverwind.io>
The file view, blame and web editor read `.editorconfig` from the
default branch, so branch-specific settings like tab width were ignored.
The file view's middleware ran before the ref was resolved, and blame
never loaded it.
- `GetEditorconfig` reads the current commit, falling back to the
default branch only where no ref is resolved
- `Home` and `RefBlame` load the editorconfig after ref resolution
Fixes: https://github.com/go-gitea/gitea/issues/27413
Signed-off-by: silverwind <me@silverwind.io>
Co-authored-by: silverwind <me@silverwind.io>
Only the plain git migration can clone from a local server path. Other
services (Gitea, GitLab, Gogs, OneDev, GitBucket, Codebase, CodeCommit,
Bitbucket Server) are accessed through their remote API, so a local path
never works, yet their migrate forms advertised "or a local server
path".
- Remove the local path hint from those templates.
- Add `IsMigrateURLAllowedForService` and use it in the web and API
migrate handlers and in `MigrateRepository`, rejecting local paths for
any service other than plain git.
- Add a unit test.
Raised in review of #39913.
_Authored with Codet (AI assistant) on behalf of @lunny._
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
When opening a pull request that renamed a CSV file:
the compare page fails with a 500.
Fix: load the left blob for every file that was not added, and the right blob
(from `Name`) for every file that was not deleted. That matches what the
templates did before #38517.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Remove the workaround for chi router, do as much as possible in
middlewares
---------
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Giteabot <teabot@gitea.io>
Webhook delivery history (request/response, status, timestamp) was only
reachable through the repo settings UI, which shows the last ~10
deliveries with no filtering. There was no way to fetch this data
programmatically for debugging integrations or auditing webhook
activity.
This adds two read endpoints at the repository level:
- `GET /repos/{owner}/{repo}/hooks/{id}/deliveries` — paginated list of
a hook's deliveries, most recent first
- `GET /repos/{owner}/{repo}/hooks/{id}/deliveries/{uuid}` — a single
delivery, including request/response headers, body, and status
Both reuse the existing `hook_task` storage and retention policy (the
`CleanupHookTaskTable` cron job) that already back the "Recent
Deliveries" web UI, and are gated by the same `reqAdmin()` +
`reqWebhooksEnabled()` guard as the existing hook endpoints.
Org, user, and admin/system-level webhooks are intentionally out of
scope for this PR to keep it reviewable; they can follow as separate
PRs using the same pattern.
## Usage
```bash
curl -H "Authorization: token $TOKEN" \
https://gitea.example.com/api/v1/repos/{owner}/{repo}/hooks/{id}/deliveries
curl -H "Authorization: token $TOKEN" \
https://gitea.example.com/api/v1/repos/{owner}/{repo}/hooks/{id}/deliveries/{uuid}
```
## Testing
- Unit tests for the new model-layer query functions
(`ListHookTasks`, `GetHookTaskByUUID`)
- Integration tests covering: list + get, an undelivered task's response
is `null`, a hook/delivery UUID from another repository 404s instead
of leaking across repos, and pagination (`page`/`limit`, `Link` and
`X-Total-Count` headers, correct ordering/no overlap across pages)
- `make lint-go`, `make generate-swagger`
Fixes: https://github.com/go-gitea/gitea/issues/38463.
Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>
Add migration from Bitbucket Server (Data Center). It has no issues,
wiki, milestones or releases, so those are not offered.
Migrated: git data with tags and LFS, and pull requests with state,
draft flag, merge commit/base, comments, approvals and inline code
comments including resolved threads. Multi-line ranges become a `Lines
a-b:` text prefix, as Gitea anchors code comments to one line. Auth is
an HTTP access token (personal, project or repository) or
username/password; 429 responses are retried.
Changes outside the downloader:
- `ReviewComment.Resolved`, applied by the uploader.
- The uploader creates a fork PR's head branch from its head commit when
already in the repository, instead of fetching the fork (Bitbucket keeps
it in `refs/pull-requests/<id>/from`; private forks need credentials).
Tested with unit tests, responses recorded from Bitbucket 10.5.1, and an
end-to-end migration. Independent of
https://github.com/go-gitea/gitea/pull/38607 (Bitbucket Cloud); the
service type enum value needs adjusting for whichever lands second.
---------
Co-authored-by: bircni <bircni@icloud.com>
Adds a batched GraphQL path for GitHub migrations that fetches each
issue or pull request together with its comments, reviews, reactions and
timeline events instead of the REST N+1 pattern, sharply reducing
rate-limit pressure on large repositories.
- GraphQL is used for authenticated migrations with issues or pull
requests and falls back to REST automatically when it fails
- timeline events (close/reopen, lock, rename, label, milestone, merge,
branch deletion, pin, auto-merge) are imported as typed comments showing
their GitHub author
- GitBucket issue migrations no longer stop after the first page,
GitBucket sends no Link header
Fixes: https://github.com/go-gitea/gitea/issues/13243
Fixes: https://github.com/go-gitea/gitea/issues/31856
Signed-off-by: David E. Lovas <david@rexovas.com>
Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Giteabot <teabot@gitea.io>
This PR introduces project workflows, allowing users to configure
event-based automations directly from the UI.
Users can define filters and actions for each workflow event, and
duplicate existing events when needed.
### Concepts
Every workflow is composed of `event`, `filters`, and `actions`.
- **Event** - An Event defines when a workflow should be triggered. It
represents specific activities such as an issue being added to the
project, a pull request being merged, or a project column being changed.
- **Filter** - A Filter refines when an event should take effect. It
allows users to specify conditions like issue type, labels, or project
columns to ensure actions only run in targeted scenarios.
- **Action** - An Action defines what happens when the event and filters
match. It performs operations such as adding labels, moving issues
between columns, or changing the issue or pull request state.
### Available Events
Currently, the following eight events are supported:
- On issues/pull requests closed
- On issues/pull requests reopened
- On issues/pull requests was added to the project, including when
created in it
- On issues/pull requests was removed from the project
- On issues/pull requests' column was changed
- On code changes requested
- On code review approved
- On pull request merged
Each event type supports its own set of filters and actions.
### Available Filters
There are four types of filters that can be configured for an event:
- **Apply to** - Specify the target type: `issues only`, `pull requests
only`, or `issues and pull requests`.
- **Labels** - Trigger the event only if the issue or pull request
contains one or more specified labels.
- **Move from column** - Filter events based on the original project
column.
- **Move to column** - Filter events based on the destination project
column.
### Available Actions
There are four types of actions that can be executed when an event is
triggered:
- **Add labels** - Automatically add one or more labels to the issue or
pull request.
- **Remove labels** - Remove one or more specified labels.
- **Move to column** - Move the issue or pull request to a designated
project column.
- **Issue state** - Close or reopen the issue. Pull requests can only be
closed.
### Enable/Disable an event workflow
Each event workflow can be individually enabled or disabled. When
disabled, the workflow will be ignored even if the event occurs. This
feature is useful when you want to temporarily stop certain automated
actions without deleting the workflow configuration. Workflows of closed
projects don't run either.
### Clone an event workflow
When you need multiple workflows for the same event, you can clone an
existing event to create a new one with different filters and actions.
This allows you to customize how the same event behaves under different
conditions, making workflows more flexible and powerful.
### Permissions
Permissions are checked when workflows are managed, not when they run:
- Viewing a project's workflows needs read access to the project
- Managing them needs write access to the project plus write access to
both issues and pull requests of a repository project's repository,
being an owner of an organization project's organization or a site
administrator, or being the user of a user project. Tokens limited to
public repositories can't manage owner-level projects' workflows
- A workflow runs when its event happens to an item of an open project
that can hold the item's repository, unless that repository is archived,
no matter who caused the event, and applies its actions without further
permission checks. Pull requests are never reopened
- Changes show as made by "Project workflow" on behalf of the user who
triggered the event and don't trigger other workflows
- Only users who can read an item can put it on a board or move its card
The management rule is also documented on `CanManageWorkflows` in
`services/projects/workflow.go`.
Fixes: #14359Fixes: #26704Fixes: #27990
Replaces: #28745
Documentation PR: https://gitea.com/gitea/docs/pulls/308
Co-authored-by: bircni <bircni@icloud.com>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Giteabot <teabot@gitea.io>
For repos with many issues, the GitHub API only allows cursor-based
pagination instead of page-based pagination after a certain page.
A full migration of such repos failed with a 422 response.
Now, we dynamically switch how we migrate based on the response from the GitHub API.
Fixes#39882
---------
Signed-off-by: Phoenix <44333391+yuhongwei380@users.noreply.github.com>
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: silverwind <me@silverwind.io>
Release webhook notifications never mention attached files, although the
payload carries them.
The summary shared by all webhook providers now appends an asset count
like `(assets: 2)` for created and updated releases.
Closes https://github.com/go-gitea/gitea/issues/39630
Co-authored-by: silverwind <me@silverwind.io>
Code comments next to a diff hunk could show one line too high in the
conversation view vs. in the code view that showed them on the correct
line.
- Treat the hunk end in `CutDiffAroundLine` as exclusive, so the line
after a hunk falls back to the unchanged-line context
- Don't count `\ No newline at end of file` markers as lines when
cutting the hunk
- Share the unchanged-line fallback with migrated review comments
Matching git's stderr text is fragile because it breaks silently when
git rewords a message. Where git offers an exit code or a query with a
defined answer, use that instead, and drop matches that could never
trigger. Follow-up to the discussion in https://github.com/go-gitea/gitea/pull/39857.
- Detect a missing remote by `git remote` exit code 2
- Resolve annotated tags to their tree with cat-file `^{tree}` instead
of spawning `ls-tree` when listing tree entries
- Skip the wiki page lookup when the default branch doesn't exist yet
- Remove unreachable "no merge base" and "unknown revision" checks
Adds a repository setting for the default squash merge commit message
body with GitHub's options `PR_BODY`, `COMMIT_MESSAGES` and `BLANK`,
also settable via `squash_merge_commit_message` in the API. The commit
title stays the pull request title.
1. Breaking: API squash merges without `merge_message_field` now get the
same default message as the web merge box instead of a title-only
commit, which affects clients sending only `Do` like Renovate. An empty
value keeps title-only.
2. Breaking: `[repository.pull-request] DEFAULT_SQUASH_MERGE_MESSAGE`
replaces `POPULATE_SQUASH_COMMENT_WITH_COMMIT_MESSAGES`, which still
works as a deprecated fallback. Repositories store the default when
created or next saved.
3. Deprecated app.ini keys now migrate through one table instead of
per-loader fallbacks, which also fixes the ignored `[mailer] CERT_FILE`
and `KEY_FILE` fallbacks. Deprecation warnings no longer claim a removal
version.
<img width="1891" height="571" alt="grafik"
src="https://github.com/user-attachments/assets/0fdcde13-1f11-4156-b598-d80eb0d04cbe"
/>
---------
Co-authored-by: silverwind <me@silverwind.io>
The token a request authenticated with now lives on the doer, so
permission and visibility checks apply its scope from the doer alone.
Replaces https://github.com/go-gitea/gitea/pull/39167, following
https://github.com/go-gitea/gitea/pull/39167#issuecomment-6063839619.
1. Public-only tokens get 404 instead of 403 for hidden users, orgs and
repos, like GitHub.
2. Org ownership, site admin, repo and org creation, generation and
transfer endpoints reject public-only tokens.
3. API pull request responses set `head.repo` to `null` and
`head.repo_id` to `-1` when the caller cannot see the head repository.
4. Public-only tokens see org memberships like anonymous callers, so
concealed members stay hidden.
5. Site admin access in activity feeds, follower lists, org members,
Actions run lists and packages needs a session or an `all`-scoped token.
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>