This commit forces the resource owner (user) to always approve OAuth 2.0
authorization requests if the client is public (e.g. native
applications).
As detailed in [RFC 6749 Section 10.2](https://www.rfc-editor.org/rfc/rfc6749.html#section-10.2),
> The authorization server SHOULD NOT process repeated authorization
requests automatically (without active resource owner interaction)
without authenticating the client or relying on other measures to ensure
that the repeated request comes from the original client and not an
impersonator.
With the implementation prior to this patch, attackers with access to
the redirect URI (e.g., the loopback interface for
`git-credential-oauth`) can get access to the user account without any
user interaction if they can redirect the user to the
`/login/oauth/authorize` endpoint somehow (e.g., with `xdg-open` on
Linux).
Fixes #25061.
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit 5c542ca94caa3587329167cfe9e949357ca15cf1)
Before, we would just throw 500 if a user passes an attachment that is
not an allowed type. This commit catches this error and throws a 422
instead since this should be considered a validation error.
(cherry picked from commit 872caa17c0a30d95f85ab75c068d606e07bd10b3)
Conflicts:
tests/integration/api_comment_attachment_test.go
tests/integration/api_issue_attachment_test.go
trivial context conflict because of 'allow setting the update date on issues and comments'
Makes it easier to use because you see which square is currently
hovered:
<img width="314" alt="Screenshot 2024-05-02 at 15 38 20"
src="https://github.com/go-gitea/gitea/assets/115237/3a15dad1-2259-4f28-9fae-5cf6ad3d8798">
I did try a `scoped` style for this, but that did not work for some
reason.
(cherry picked from commit 6f89d5e3a0886d02ead732005f593ae003f78f78)
The test had a dependency on `https://api.pwnedpasswords.com` which
caused many failures on CI recently:
```
--- FAIL: TestPassword (2.37s)
pwn_test.go:41: Get "https://api.pwnedpasswords.com/range/e6b6a": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
FAIL
coverage: 82.9% of statements
```
(cherry picked from commit 9235442ba58524c8d12ae54865d583acfa1f439d)
Co-authored-by: silverwind <me@silverwind.io>
(cherry picked from commit be112c1fc30f87a248b30f48e891d1c8c18e8280)
Conflicts:
routers/web/web.go
trivial conflict because of https://codeberg.org/forgejo/forgejo/pulls/1533
This is because it doesn't exist as an adapter. The `redis` adapter
already handles Redis cluster configurations.
Fixes #30534.
(cherry picked from commit f135cb7c9457f7b9bdc43601f44757834573950f)
Conflicts:
docs/content/administration/config-cheat-sheet.en-us.md
does not exist in Forgejo
This is a follow-up for 5e1bd8af5f, which
was my first commit to Gitea. It is also a follow up for the
Gitea PR #29300 (https://github.com/go-gitea/gitea/pull/23900) created
by myself, which turned stale.
This change partially restores the behavior of Gitea PR #23747
(https://github.com/go-gitea/gitea/pull/23747) by wxiaoguang, but
maintains the lock.
The original idea was to differentiate things from GitHub and GitLab a
little bit, and show the email address on the profile. The profile is
not only a place where the user chooses to show how they present
themselves on an instance, it is also a place where they can assess
their relationship *with* the instance, as it provides features such
as the Public Activity feed that can be only shown to the user, in
private.
It's, in some way, a dashboard. The email was shown there to remind
the user that this is the primary email that will be used by a supposed
administrator to contact them. There were other motivations behind that
change as well, but, long story short, the idea did not work very well,
as some people (e.g. people livestreaming on the Internet, or 'normal'
users sharing their screens) do not want to put their email address
out there when showing their screen to other people.
Other alternatives, such as blurring the text or only showing the real
email address, were explored, but were rejected because of
browser compatibility and simplicity reasons. The padlock icon that
is shown when showing the email address to other people has been kept.
One viable alternative could be displaying the placeholder email
instead, but that requires some more thought.
Fixes https://codeberg.org/forgejo/forgejo/issues/1950.
Backport #30770
If an user is deactivated, it should not be in the list of users who are
suggested to be assigned or review-requested.
old assignees or reviewers are not affected.
---
*Sponsored by Kithara Software GmbH*
(cherry picked from commit f2d8ccc5bb2df25557cc0d4d23f2cdd029358274)
Conflicts:
models/repo/user_repo_test.go
because there is one less fixture user compared to Gitea
Resolve all cases for `unused parameter` and `unnecessary type
arguments`
Related: #30729
---------
Co-authored-by: Giteabot <teabot@gitea.io>
(cherry picked from commit e80466f7349164ce4cf3c07bdac30d736d20f035)
Conflicts:
modules/markup/markdown/transform_codespan.go
modules/setting/incoming_email.go
routers/api/v1/admin/user_badge.go
routers/private/hook_pre_receive.go
tests/integration/repo_search_test.go
resolved by discarding the change, this is linting only and
for the sake of avoiding future conflicts
When you cross-compile Gitea and you specify one of the envrionment
variables related to C flags, cgo will fail to build the generator
programs (e.g. generate-bindata) because GOOS and GOARCH are unset, but
those additional flags variables are not unset together with those.
To solve this issue, the simplest way that I've found is to disable cgo
in the `go generate` command as it's not really used there.
For example, I've had this problem with cross-compiling Gitea on FreeBSD
x86_64 to ARMv7 where it's necessary to pass `--target` to `clang` via
`CGO_CFLAGS`:
```
GOOS=freebsd \
GOARCH=arm \
GGOARM=7 \
CGO_ENABLED=1 \
SYSROOT=/usr/local/freebsd-sysroot/armv7 \
CC=clang \
CGO_CFLAGS="--target=armv7-unknown-freebsd13.2-gnueabihf" \
TAGS="bindata sqlite sqlite_unlock_notify" \
make SHELL='sh -x' build
```
```
Running go generate...
# runtime/cgo
In file included from gcc_freebsd_amd64.c:9:
In file included from /usr/include/signal.h:42:
/usr/include/sys/_ucontext.h:44:2: error: unknown type name 'mcontext_t'
modules/migration/schemas_bindata.go:8: running "go": exit status 1
# runtime/cgo
In file included from gcc_freebsd_amd64.c:9:
In file included from /usr/include/signal.h:42:
/usr/include/sys/_ucontext.h:44:2: error: unknown type name 'mcontext_t'
modules/options/options_bindata.go:8: running "go": exit status 1
# runtime/cgo
In file included from gcc_freebsd_amd64.c:9:
In file included from /usr/include/signal.h:42:
/usr/include/sys/_ucontext.h:44:2: error: unknown type name 'mcontext_t'
modules/public/public_bindata.go:8: running "go": exit status 1
# runtime/cgo
In file included from gcc_freebsd_amd64.c:9:
In file included from /usr/include/signal.h:42:
/usr/include/sys/_ucontext.h:44:2: error: unknown type name 'mcontext_t'
modules/templates/templates_bindata.go:8: running "go": exit status 1
gmake[1]: *** [Makefile:781: generate-go] Error 1
*** Error code 2
Stop.
```
But with this fix Gitea compiles successfully.
(cherry picked from commit d11133b83652238023b52576e0d3e57a4f4b21c9)
This allows `nix flake metadata` and nix in general to lock a *branch*
tarball link in a manner that causes it to fetch the correct commit even
if the branch is updated with a newer version.
For further context, Nix flakes are a feature that, among other things,
allows for "inputs" that are "github:someuser/somerepo",
"https://some-tarball-service/some-tarball.tar.gz",
"sourcehut:~meow/nya" or similar. This feature allows our users to fetch
tarballs of git-based inputs to their builds rather than using git to
fetch them, saving significant download time.
There is presently no gitea or forgejo specific fetcher in Nix, and we
don't particularly wish to have one. Ideally (as a developer on a Nix
implementation myself) we could just use the generic tarball fetcher and
not add specific forgejo support, but to do so, we need additional
metadata to know which commit a given *branch* tarball represents, which
is the purpose of the Link header added here.
The result of this patch is that a Nix user can specify `inputs.something.url =
"https://forgejo-host/some/project/archive/main.tar.gz"` in flake.nix
and get a link to some concrete tarball for the actual commit in the
lock file, then when they run `nix flake update` in the future, they
will get the latest commit in that branch.
Example of it working locally:
» nix flake metadata --refresh 'http://localhost:3000/api/v1/repos/jade/cats/archive/main.tar.gz?dir=configs/nix'
Resolved URL: http://localhost:3000/api/v1/repos/jade/cats/archive/main.tar.gz?dir=configs/nix
Locked URL: 804ede182b.tar.gz?dir=configs
/nix&narHash=sha256-yP7KkDVfuixZzs0fsqhSETXFC0y8m6nmPLw2GrAMxKQ%3D
Description: Computers with the nixos
Path: /nix/store/s856c6yqghyan4v0zy6jj19ksv0q22nx-source
Revision: 804ede182b6b66469b23ea4d21eece52766b7a06
Last modified: 2024-05-02 00:48:32
For details on the header value, see:
56763ff918/doc/manual/src/protocols/tarball-fetcher.md
In `repo.RemoveDependency`, use `PostFormValue` instead of
`PostForm.Get`. The latter requires `ParseForm()` to be called prior,
and in this case, has no benefit over `PostFormValue` anyway (which
calls `ParseForm()` if necessary).
While this currently does not cause any issue as far as I can tell, it
feels like a bug lying in wait for the perfect opportunity. Lets squash
it before it can do harm.
Signed-off-by: Gergely Nagy <forgejo@gergo.csillger.hu>
Adjust the `anyHashPattern` to match URL query parameters too, and
adjust `fullHashPatternProcessor` accordingly.
Includes a test case, and an update to an existing one to account for
the new capture group.
Fixes #3548.
Signed-off-by: Gergely Nagy <forgejo@gergo.csillger.hu>
