### What / why
`TimeEstimateParse` (used by the issue time-estimate form) only checked
that the first token starts at the beginning of the string and the last
token ends at its end, but never checked the gaps between consecutive
tokens. Non-whitespace garbage embedded between two valid units was
silently dropped and the string accepted with a wrong value instead of
being reported as invalid.
Examples that were wrongly accepted before this change:
- `1h 2x 3m` → 3780s (parsed as 1h3m)
- `1h_2m` → 3720s
- `1h,1m` → 3660s
All three now return an "invalid time string" error, while valid inputs
such as `1h 1m 1s` and `1h1m1s` keep working.
### How
Reject any non-whitespace content between two matched units.
---------
Signed-off-by: TowyTowy <towy@airreps.link>
Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
### Description
This PR implements an optimized, adaptive client-side auto-refresh
mechanism for the Gitea Actions workflow runs list page. It allows users
to see workflow progress updates dynamically without having to reload
the page.
Fixes https://github.com/go-gitea/gitea/issues/37457
---------
Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Fixes#38217
## Problem
Turnstile CAPTCHA verification uses `http.DefaultClient`, so the request
to `challenges.cloudflare.com` bypasses Gitea's configured HTTP proxy —
unlike other outbound HTTP clients such as the update checker
(`modules/updatechecker/update_checker.go`) and migrations. In
deployments where egress is only permitted through the configured proxy,
verification fails.
## Fix
Build the client with `proxy.Proxy()` as the transport proxy, mirroring
the update checker:
```go
func httpClient() *http.Client {
return &http.Client{
Transport: &http.Transport{
Proxy: proxy.Proxy(),
},
}
}
```
The client is built per call (rather than a package-level var) because
`proxy.Proxy()` reads `setting.Proxy` when invoked; building it at
request time ensures it reflects the loaded settings. When no proxy is
configured, behavior is unchanged (`proxy.Proxy()` returns a no-op /
`http.ProxyFromEnvironment`).
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Fixes#35472.
`Comment.LoadAssigneeUserAndTeam` already has a Ghost user fallback
for a deleted assignee user, but the parallel branch for a deleted
assignee team just swallowed the not-found error and left
`AssigneeTeam` as `nil`. This is inconsistent (the reporter's example
shows `assignee` becoming a Ghost user while `assignee_team` becomes
`null`), and it's also a latent nil pointer bug: other code that
assumes `AssigneeTeam` is set once this function returns without
error will panic.
Added `organization.NewGhostTeam()` / `Team.IsGhost()`, mirroring the
existing `user_model.NewGhostUser()` / `User.IsGhost()` pattern, and
used it in the same fallback branch.
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The flex-list refactor (#37505) raised the shared `.item-trailing`
selector's specificity, so its `flex-wrap: wrap` started overriding the
run list's intended `flex-wrap: nowrap` — a long branch name pushed the
trailing content past its fixed 280px width and wrapped the kebab menu
onto its own line.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The `official` flag of a pull request review is computed against the
target branch's protection rules at submit time and stored on the review
record. `ChangeTargetBranch` updated the base branch but never
re-evaluated it, so an `official` approval obtained against an
unprotected branch could be retargeted onto a protected branch and
satisfy its required approvals, bypassing the branch protection.
This re-evaluates the `official` flag of the latest approve/reject
reviews against the new base branch whenever the target branch changes.
The contribution calendar manually constructed localized date strings
instead of using the browser's locale-aware date formatting. Replace
this with `toLocaleDateString()` to correctly format dates for all
locales and calendar systems, including non-Gregorian calendars.
Fixes https://github.com/go-gitea/gitea/issues/38375
- **Locale DoS:** the `Locale` middleware passed the raw
`Accept-Language` header to `ParseAcceptLanguage`, whose guard only
counts `-` while the scanner aliases `_` to `-` — a large `_`-separated
header on an unauthenticated request burned CPU. The header is now
length-bounded before parsing.
- **Public-only token scope:** `GET /teams/{id}/repos`,
`.../repos/{org}/{repo}`, `/teams/{id}/activities/feeds`, and
`/users/{username}/orgs/{org}/permissions` still returned private
repo/activity/permission data to a public-only token. They now filter
via `TokenCanAccessRepo` / `ApplyPublicOnly` and reject non-public org
permissions.
- **Push-option visibility:** `repo.private` / `repo.template` push
options were applied to any existing repo, letting an owner/admin
silently flip visibility bypassing audit, webhooks, and notifications.
They are now honored only on push-to-create.
Committer can also be co-author, it should only not be included in the
co-author list if it is not in the "Co-author-by" list.
* Author & Co-author: they changed the code (attribution)
* Committer: they submitted the commit but didn't change the code (e.g.:
maintainer signed a commit)
Fix#38384
The `/user/starred` and `/user/subscriptions` endpoints returned private
repositories a user had starred/watched even after their access to those
repositories was revoked, still exposing the repository name,
description and visibility (including later metadata changes).
Private repositories in the starred/watched queries are now gated on the
actor's current access via `AccessibleRepositoryCondition`, so users who
no longer have access no longer receive the metadata. Public
repositories and public-only tokens are unaffected.
The LFS batch and upload handlers linked an object that already existed
in the content store but was not linked to the current repo whenever the
token's user could access it in another repo. Deploy-key tokens carry
the repo owner's identity, so a single-repo write deploy key could link
and then download objects from any repo the owner can see.
This drops the cross-repo access check: the batch handler now makes the
client upload (hash-verified) any object not yet linked to the repo, and
the upload handler skips proof of possession only when the object is
already linked to the current repo.
The commit page built its "co-authored by" list from
`AllParticipantIdentities()[1:]`, which only drops the author and leaves
the committer in the list even though the committer is already shown
separately as "committed by". This caused the committer to be wrongly
rendered as a co-author (issue #38384): a commit authored by
`silverwind`, committed by `bircni`, with a `Co-authored-by: silverwind`
trailer displayed "co-authored by bircni" instead of the actual trailer
identity. This adds a `Commit.CoAuthorIdentities()` method that excludes
both the author and the committer, uses it on the commit page, and
covers the fix with a regression test.
Closes#38384
Follow #38237#38237 posts "skipped" commit statuses for every workflow that is not
triggered due to a filter (e.g. `paths` or `branches`) mismatch.
However, for non-required workflows, creating "skipped" commit statuses
for them would generate a lot of noise.
To address this issue, this PR adds a check before creating commit
status:
- For the context that matches any required status check patterns, a
"skipped" commit status will be created. The `Required` label can inform
users that this status check is required, but has been skipped because
of a filter mismatch.
- For a non-required context, nothing will be created.
NOTE: Reducing noise is a best-effort approach and isn't entirely
accurate. When creating commit statuses, it is impossible to predict
which branch protection rule will take effect. Therefore, we have to
compare the commit status context against the required patterns from all
rules. If any rule matches, the context is considered "required".
https://github.com/go-gitea/gitea/pull/37594 widened the author column
from `three wide` to `four wide`, leaving a large empty gap around the
SHA column in the common single-author case. The old author cell was
capped at 180px, so the wider column only adds whitespace: avatar-stack
names ellipsize at 240px each, and wider multi-author rows still expand
naturally in the auto-layout table. Restore the pre-existing
`three`/`eight` widths.
Co-authored-by: bircni <bircni@icloud.com>
`data-render-name` is set before the plugin's async render runs, so
measuring the container height right after the attribute appears can
observe the pre-render 48px height when the `pdfobject` chunk loads
slowly (flaked in CI). Poll for the height instead, like the asciicast
test in the same file does.
Before, the logic is already there for "pull merge box".
After, the logic is extracted into a general class ActivePageTimer and
will help more pages (including #38329)
3 reductions in the DB load generated by many runners polling
`FetchTask`:
**1. Debounce runner heartbeat writes**
Every poll wrote `last_online`, and every `UpdateTask`/`UpdateLog` wrote
`last_active` — while a runner streams logs that is many writes per
second per runner. These are now persisted only when stale enough to
actually affect the active/offline status (`ShouldPersistLastOnline` /
`ShouldPersistLastActive`), using the existing columns.
**2. Throttle concurrent task picks**
A new in-process semaphore (`MAX_CONCURRENT_TASK_PICKS`) bounds how many
runners run the task-assignment transaction at once, so a fleet polling
together cannot stampede the query. Throttled polls retry on their next
poll without advancing the runner's tasks version.
**3. Paginate the task-pick query**
`CreateTaskForRunner` previously loaded every waiting job in the
runner's scope into memory on each poll (no `LIMIT`). Now it pages
through the waiting backlog oldest-first with `LIMIT`, claiming the
first label-matching job.
---------
Co-authored-by: Zettat123 <zettat123@gmail.com>
Pull mirror sync ran `git fetch` / `remote update` / `remote prune`
without disabling HTTP redirects. A mirror remote that later starts
redirecting to an otherwise-blocked or internal address could be used as
an SSRF/exfiltration vector on scheduled syncs, bypassing the
allow/block validation applied at migration time.
This sets `http.followRedirects=false` on all three remote-contacting
commands in the pull mirror path, matching the existing guard already
present on the clone path.
Adds a new `branch-name` value for the `[repository.pull-request]`
`DEFAULT_TITLE_SOURCE` setting that always uses the normalized branch
name as the PR title, regardless of commit count.
Fix#38317
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
When a runner's `FetchTask` request context is cancelled after the job
is claimed but before the task reaches the runner (e.g. request
timeout), the job was left referencing a running task no runner ever
executes, so it stayed unpickable.
Fix: Check the context after assembling the task and release the task so
the job returns to waiting status for another runner.
`TestInitKeys` verified whether a host key file was regenerated by
comparing `os.FileInfo` before and after a `InitDefaultHostKeys` call.
On systems where the OS clock / filesystem timestamp granularity is
coarse, both writes land in the same tick and get an identical mtime.
The resulting `FileInfo` is then byte-for-byte equal even though the key
was actually regenerated, so `assert.NotEqual` fails.
Since a regenerated key always produces different random bytes,
comparing the content can reliably detect regeneration.
This fixes the web release edit flow so renamed release attachments are
validated against `[repository.release] ALLOWED_TYPES`.
Previously, the API attachment edit endpoint already enforced release
attachment type restrictions, but the web release edit form passed
`attachment-edit-*` values into `release_service.UpdateRelease`, which
updated attachment names directly without validating the new filename
against `setting.Repository.Release.AllowedTypes`.
As a result, a user with repository write access could rename an
existing release attachment to a disallowed extension through the web
UI.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The runner-listing API endpoints (`admin`, `org`, `user`, `repo`, and
`shared`) return runners in a non-deterministic order across pages. When
paging through runners, some runners from page 1 could reappear on page
2 (and others get skipped entirely).
The cause is in `FindRunnerOptions.ToOrders()`. Most sort modes order by
a **non-unique** column only:
- default / `online` / `offline` → `last_online`
- `alphabetically` / `reversealphabetically` → `name`
When multiple runners tie on the sort key (e.g. every offline runner
shares `last_online = 0`, or two runners have the same name), the
database is free to return the tied rows in any order between separate
queries. Combined with `LIMIT`/`OFFSET` pagination, this means the same
runner can land on more than one page.
## Fix
Append the unique primary key `id` as a stable tiebreaker to each
non-unique sort order:
The `newest`/`oldest` modes already sort by the unique `id`, so they are
left unchanged.
Draft-release access control was enforced only on the API release
endpoints (`/api/v1/repos/{owner}/{repo}/releases/...`) but not on the
UUID-based web attachment endpoints (`/attachments/{uuid}`,
`/{owner}/{repo}/attachments/{uuid}`,
`/{owner}/{repo}/releases/attachments/{uuid}`).
Anyone who obtained an attachment UUID — including unauthenticated
callers — could download files belonging to a hidden draft release,
since `ServeAttachment` only checked repo-level read permission and
never the release's draft state.
The nightly snap build fails with `snapcraft remote-build`'s
`BadRequest()`: it force-pushes Gitea's full history to a fresh
Launchpad repo each run and creates the recipe before Launchpad has
indexed the `main` ref. The race is unwinnable at Gitea's repo size, and
Canonical does not support `remote-build` in CI.
Build locally on native amd64 + arm64 runners with
`snapcore/action-build` + `snapcore/action-publish` instead — no
Launchpad, no race. Drops the `LAUNCHPAD_CREDENTIALS` secret (publishing
keeps `SNAPCRAFT_STORE_CREDENTIALS`); the `snap/` recipe is unchanged,
so the same snaps ship to `latest/edge`.
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
TowyTowy