Lunny Xiao [Fri, 30 Dec 2022 00:08:16 +0000 (08:08 +0800)]
Remove ReverseProxy authentication from the API (#22219) (#22252)
backport #22219
Since we changed the /api/v1/ routes to disallow session authentication
we also removed their reliance on CSRF. However, we left the
ReverseProxy authentication here - but this means that POSTs to the API
are no longer protected by CSRF.
Now, ReverseProxy authentication is a kind of session authentication,
and is therefore inconsistent with the removal of session from the API.
This PR proposes that we simply remove the ReverseProxy authentication
from the API and therefore users of the API must explicitly use tokens
or basic authentication.
Replace #22077
Close #22221
Close #22077
Signed-off-by: Andrew Thornton <art27@cantab.net> Signed-off-by: Andrew Thornton <art27@cantab.net> Co-authored-by: zeripath <art27@cantab.net>
Gusted [Wed, 28 Dec 2022 20:16:18 +0000 (21:16 +0100)]
Fix key signature error page (#22229) (#22231)
- Backport of #22229
- When the GPG key contains an error, such as an invalid signature or an
email address that does not match the user.A page will be shown that
says you must provide a signature for the token.
- This page had two errors: one had the wrong translation key and the
other tried to use an undefined variable
[`.PaddedKeyID`](https://github.com/go-gitea/gitea/blob/e81ccc406bf723a5a58d685e7782f281736affd4/models/asymkey/gpg_key.go#L65-L72),
which is a function implemented on the `GPGKey` struct, given that we
don't have that, we use
[`KeyID`](https://github.com/go-gitea/gitea/blob/e81ccc406bf723a5a58d685e7782f281736affd4/routers/web/user/setting/keys.go#L102)
which is [the fingerprint of the
publickey](https://pkg.go.dev/golang.org/x/crypto/openpgp/packet#PublicKey.KeyIdString)
and is a valid way for opengpg to refer to a key.
<!--
Please check the following:
1. Make sure you are targeting the `main` branch, pull requests on
release branches are only allowed for bug fixes.
2. Read contributing guidelines:
https://github.com/go-gitea/gitea/blob/main/CONTRIBUTING.md
3. Describe what your pull request does and which issue you're targeting
(if any)
Gusted [Tue, 20 Dec 2022 02:07:41 +0000 (03:07 +0100)]
Check for zero time instant in TimeStamp.IsZero() (#22171) (#22173)
- Backport of #22171
- Currently, the 'IsZero' function for 'TimeStamp' just checks if the
unix time is zero, which is not the behavior of 'Time.IsZero()', but
Gitea is using this method in accordance with the behavior of
'Time.IsZero()'.
- Adds a new condition to check for the zero time instant.
- Fixes a bug where non-expiring GPG keys where shown as they expired on
Jan 01, 0001.
- Related https://codeberg.org/Codeberg/Community/issues/791
Do not list active repositories as unadopted (#22034) (#22167)
Backport #22034
This fixes a bug where, when searching unadopted repositories, active
repositories will be listed as well. This is because the size of the
array of repository names to check is larger by one than the
`IterateBufferSize`.
For an `IterateBufferSize` of 50, the original code will pass 51
repository names but set the query to `LIMIT 50`. If all repositories in
the query are active (i.e. not unadopted) one of them will be omitted
from the result. Due to the `ORDER BY` clause it will be the oldest (or
least recently modified) one.
Co-authored-by: Christian Ullrich <christian.ullrich@traditionsa.lu>
KN4CK3R [Wed, 14 Dec 2022 17:49:44 +0000 (18:49 +0100)]
Fix condition for is_internal (#22095) (#22131)
Backport of #22095
I changed it to a static condition because it needs a new version of
xorm which is only available in 1.19. This change is valid because
`SearchLatestVersions` is never called to list internal versions and
there will no change to this behaviour in <1.19.
Jason Song [Tue, 6 Dec 2022 14:15:38 +0000 (22:15 +0800)]
Fix issue/PR numbers (#22037) (#22045)
Backport #22037.
When deleting a closed issue, we should update both `NumIssues`and
`NumClosedIssues`, or `NumOpenIssues`(`= NumIssues -NumClosedIssues`)
will be wrong. It's the same for pull requests.
Releated to #21557.
Alse fixed two harmless problems:
- The SQL to check issue/PR total numbers is wrong, that means it will
update the numbers even if they are correct.
- Replace legacy `num_issues = num_issues + 1` operations with
`UpdateRepoIssueNumbers`.
Gusted [Fri, 18 Nov 2022 14:24:49 +0000 (15:24 +0100)]
Prevent dangling user redirects (#21856) (#21859)
- Backport #21856
- It's possible that the `user_redirect` table contains a user id that
no longer exists.
- Delete a user redirect upon deleting the user.
- Add a check for these dangling user redirects to check-db-consistency.
zeripath [Mon, 14 Nov 2022 07:58:11 +0000 (07:58 +0000)]
Fix enabling partial clones on 1.17 (#21809)
When backporting #20902 in #21058 there was a slight misbackport. It was
missed that we needed to remove the global command option before setting
the settings.
zeripath [Sun, 13 Nov 2022 22:43:40 +0000 (22:43 +0000)]
Prevent panic in doctor command when running default checks (#21791) (#21808)
Backport #21791
There was a bug introduced in #21352 due to a change of behaviour caused
by #19280. This causes a panic on running the default doctor checks
because the panic introduced by #19280 assumes that the only way
opts.StdOut and opts.Stderr can be set in RunOpts is deliberately.
Unfortunately, when running a git.Command the provided RunOpts can be
set, therefore if you share a common set of RunOpts these two values can
be set by the previous commands.
This PR stops using common RunOpts for the commands in that doctor check
but secondly stops RunCommand variants from changing the provided
RunOpts.
Xinyu Zhou [Fri, 11 Nov 2022 06:38:52 +0000 (14:38 +0800)]
Add HEAD fix to gitea doctor (#21352) (#21751)
Backport #21352
Due to a bug in presumably an older version of Gitea, multiple of my
repositories still have their HEADs pointing to a `master` branch while
the default branch on the UI is listed as `main`. This adds a `gitea
doctor` command that will fix all of the HEAD references for repos when
they're not synchronized with the default branch in the DB.
This will help with cloning to ensure that git automatically checks out
the right branch, instead of a nonexistent one.
Note: I'm not sure if I actually need to do more other than add a file
here. Will try testing this out on my server soon.
Wayne Starr [Wed, 9 Nov 2022 15:02:21 +0000 (09:02 -0600)]
Remove semver compatible flag and change pypi to an array of test cases (#21708) (#21729)
Backport (#21708)
This addresses #21707 and adds a second package test case for a
non-semver compatible version (this might be overkill though since you
could also edit the old package version to have an epoch in front and
see the error, this just seemed more flexible for the future).
Wayne Starr [Wed, 9 Nov 2022 06:00:09 +0000 (00:00 -0600)]
Allow for resolution of NPM registry paths that match upstream (#21568) (#21723)
Backport (#21568)
This PR fixes issue #21567 allowing for package tarball URLs to match
the upstream registry (and GitLab/JFrog Artifactory URLs). It uses a
regex to parse the filename (which contains the NPM version) and does a
fuzzy search to pull it out. The regex was built/expanded from
http://json.schemastore.org/package,
https://github.com/Masterminds/semver, and
https://docs.npmjs.com/cli/v6/using-npm/semver and is testable here:
https://regex101.com/r/OydBJq/5
zeripath [Tue, 1 Nov 2022 19:24:37 +0000 (19:24 +0000)]
Fix repository adoption on Windows (#21646) (#21651)
Backport #21646
A bug was introduced in #17865 where filepath.Join is used to join
putative unadopted repository owner and names together. This is
incorrect as these names are then used as repository names - which shoud
have the '/' separator. This means that adoption will not work on
Windows servers.
Ashley Nelson [Wed, 26 Oct 2022 07:44:05 +0000 (02:44 -0500)]
Update milestone counters when issue is deleted (#21459) (#21586)
Backports #21459
When actions besides "delete" are performed on issues, the milestone
counter is updated. However, since deleting issues goes through a
different code path, the associated milestone's count wasn't being
updated, resulting in inaccurate counts until another issue in the same
milestone had a non-delete action performed on it.
I verified this change fixes the inaccurate counts using a local docker
build.
unfortunately their api documentation leaves this out:
https://github.com/npm/registry/blob/master/docs/responses/package-metadata.md#abbreviated-version-objectdoc
which is likely to be the reason this was left out in gitea's initial
implementation
this response is critical for npm to install the binary in the .bin
folder so as to be included on the users default bin path, resulting in
immediate access to any binaries provided by the package
i have tested upload and installing through npm and can confirm the npm
registry now responds with bin in the version metadata and results in
the binary being available after install.
this fixes https://github.com/go-gitea/gitea/issues/21303
delvh [Thu, 20 Oct 2022 15:25:54 +0000 (17:25 +0200)]
Ignore error when retrieving changed PR review files (#21487) (#21524)
When a PR reviewer reviewed a file on a commit that was later gc'ed,
they would always get a `500` response from then on when loading the PR.
This PR simply ignores that error and instead marks all files as
unchanged.
This approach was chosen as the only feasible option without diving into
**a lot** of error handling.
silverwind [Wed, 19 Oct 2022 20:12:37 +0000 (22:12 +0200)]
Enable Monaco automaticLayout (#21516)
Enable
[`automaticLayout`](https://microsoft.github.io/monaco-editor/api/interfaces/monaco.editor.IDiffEditorOptions.html#automaticLayout)
for monaco so it can reflow itself.
zeripath [Wed, 12 Oct 2022 07:52:21 +0000 (08:52 +0100)]
Do DB update after merge in hammer context (#21401) (#21416)
Backport #21401
When merge was changed to run in the background context, the db updates
were still running in request context. This means that the merge could
be successful but the db not be updated.
This PR changes both these to run in the hammer context, this is not
complete rollback protection but it's much better.
This fixes error "unauthorized_client: invalid client secret" when
client includes secret in Authorization header rather than request body.
OAuth spec permits both:
https://www.rfc-editor.org/rfc/rfc6749#section-2.3.1
Clients in possession of a client password MAY use the HTTP Basic
authentication scheme ... Alternatively, the authorization server MAY
support including the client credentials in the request-body
Sanity validation that client id and client secret in request are
consistent with Authorization header.
Improve error descriptions. Error codes remain the same.
Respect `REQUIRE_SIGNIN_VIEW` for packages (#20873) (#21232)
Backport of #20873
When REQUIRE_SIGNIN_VIEW = true, even with public repositories, you can only see them after you login. The packages should not be accessed without login.
Prevent invalid behavior for file reviewing when loading more files (#21230) (#21234)
Backport of #21230
The problem was that many PR review components loaded by `Show more`
received the same ID as previous batches, which confuses browsers (when
clicked). All such occurrences should now be fixed.
Additionally improved the background of the `viewed` checkbox.
Fix hard-coded timeout and error panic in API archive download endpoint (#20925) (#21051)
Backport #20925
This commit updates the `GET /api/v1/repos/{owner}/{repo}/archive/{archive}`
endpoint which prior to this PR had a couple of issues.
1. The endpoint had a hard-coded 20s timeout for the archiver to complete after
which a 500 (Internal Server Error) was returned to client. For a scripted
API client there was no clear way of telling that the operation timed out and
that it should retry.
2. Whenever the timeout _did occur_, the code used to panic. This was caused by
the API endpoint "delegating" to the same call path as the web, which uses a
slightly different way of reporting errors (HTML rather than JSON for
example).
More specifically, `api/v1/repo/file.go#GetArchive` just called through to
`web/repo/repo.go#Download`, which expects the `Context` to have a `Render`
field set, but which is `nil` for API calls. Hence, a `nil` pointer error.
The code addresses (1) by dropping the hard-coded timeout. Instead, any
timeout/cancelation on the incoming `Context` is used.
The code addresses (2) by updating the API endpoint to use a separate call path
for the API-triggered archive download. This avoids producing HTML-errors on
errors (it now produces JSON errors).
Signed-off-by: Peter Gardfjäll <peter.gardfjall.work@gmail.com> Signed-off-by: Peter Gardfjäll <peter.gardfjall.work@gmail.com> Signed-off-by: Andrew Thornton <art27@cantab.net> Co-authored-by: Peter Gardfjäll <peter.gardfjall.work@gmail.com> Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Set uploadpack.allowFilter etc on gitea serv to enable partial clones with ssh (#20902) (#21058)
Backport #20902
When setting.Git.DisablePartialClone is set to false then the web server will add filter support to web http. It does this by using`-c` command arguments but this will not work on gitea serv as the upload-pack and receive-pack commands do not support this.
Instead we move these options into the .gitconfig instead.
Fix #20400
Signed-off-by: Andrew Thornton <art27@cantab.net> Signed-off-by: Andrew Thornton <art27@cantab.net>
projects / gitea.git / log