There are a few places in FlushQueueWithContext which make an incorrect
assumption about how `select` on multiple channels works.
The problem is best expressed by looking at the following example:
```go
package main
import "fmt"
func main() {
closedChan := make(chan struct{})
close(closedChan)
toClose := make(chan struct{})
count := 0
for {
select {
case <-closedChan:
count++
fmt.Println(count)
if count == 2 {
close(toClose)
}
case <-toClose:
return
}
}
}
```
This PR double-checks that the contexts are closed outside of checking
if there is data in the dataChan. It also rationalises the WorkerPool
FlushWithContext because the previous implementation failed to handle
pausing correctly. This will probably fix the underlying problem in
#22145
Fix #22145
Signed-off-by: Andrew Thornton <art27@cantab.net>
Signed-off-by: Andrew Thornton <art27@cantab.net>
This PR changed the Auth interface signature from
`Verify(http *http.Request, w http.ResponseWriter, store DataStore, sess
SessionStore) *user_model.User`
to
`Verify(http *http.Request, w http.ResponseWriter, store DataStore, sess
SessionStore) (*user_model.User, error)`.
There is a new return argument `error` which means the verification
condition matched but verify process failed, we should stop the auth
process.
Before this PR, when return a `nil` user, we don't know the reason why
it returned `nil`. If the match condition is not satisfied or it
verified failure? For these two different results, we should have
different handler. If the match condition is not satisfied, we should
try next auth method and if there is no more auth method, it's an
anonymous user. If the condition matched but verify failed, the auth
process should be stop and return immediately.
This will fix #20563
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
Co-authored-by: Jason Song <i@wolfogre.com>
If user has reached the maximum limit of repositories:
- Before
- disallow create
- allow fork without limit
- This patch:
- disallow create
- disallow fork
- Add option `ALLOW_FORK_WITHOUT_MAXIMUM_LIMIT` (Default **true**) :
enable this allow user fork repositories without maximum number limit
fixed https://github.com/go-gitea/gitea/issues/21847
Signed-off-by: Xinyu Zhou <i@sourcehut.net>
Since #20218 introduced internal watching template, template watching
should be removed from `air`. This will prevent restart the whole server
once the template files changed to speed up developing when using `make
watch`.
To ensure `make watch` will reuse template watching, this PR introduced
a new ENV `GITEA_RUN_MODE` to make sure `make watch` will always run in
a dev mode of Gitea so that template watching will open.
This PR also added more exclude testdata directories.
- Remove code that isn't being used.
Found this is my stash from a few weeks ago, not sure how I found this
in the first place.
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Fixes #22178
After this change upload versions with different semver metadata are
treated as the same version and trigger a duplicated version error.
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
For a long time Gitea has tested PR patches using a git apply --check
method, and in fact prior to the introduction of a read-tree assisted
three-way merge in #18004, this was the only way of checking patches.
Since #18004, the git apply --check method has been a fallback method,
only used when the read-tree three-way merge method has detected a
conflict. The read-tree assisted three-way merge method is much faster
and less resource intensive method of detecting conflicts. #18004 kept
the git apply method around because it was thought possible that this
fallback might be able to rectify conflicts that the read-tree three-way
merge detected. I am not certain if this could ever be the case.
Given the uncertainty here and the now relative stability of the
read-tree method - this PR makes using this fallback optional and
disables it by default. The hope is that users will not notice any
significant difference in conflict detection and we will be able to
remove the git apply fallback in future, and/or improve the read-tree
three-way merge method to catch any conflicts that git apply method
might have been able to fix.
An additional benefit is that patch checking should be significantly
less resource intensive and much quicker.
(See
https://github.com/go-gitea/gitea/issues/22083\#issuecomment-1347961737)
Ref #22083
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
The PR #21198 introduced a probable security vulnerability which
resulted in making all storage files be marked as executable.
This PR ensures that these are forcibly marked as non-executable.
Fix #22161
Signed-off-by: Andrew Thornton <art27@cantab.net>
As recognised in #21841 the rendering of plain text files is somewhat
incorrect when there are ambiguous characters as the html code is double
escaped. In fact there are several more problems here.
We have a residual isRenderedHTML which is actually simply escaping the
file - not rendering it. This is badly named and gives the wrong
impression.
There is also unusual behaviour whether the file is called a Readme or
not and there is no way to get to the source code if the file is called
README.
In reality what should happen is different depending on whether the file
is being rendered a README at the bottom of the directory view or not.
1. If it is rendered as a README on a directory - it should simply be
escaped and rendered as `<pre>` text.
2. If it is rendered as a file then it should be rendered as source
code.
This PR therefore does:
1. Rename IsRenderedHTML to IsPlainText
2. Readme files rendered at the bottom of the directory are rendered
without line numbers
3. Otherwise plain text files are rendered as source code.
Replace #21841
Signed-off-by: Andrew Thornton <art27@cantab.net>
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
The recent PR adding orphaned checks to the LFS storage is not
sufficient to completely GC LFS, as it is possible for LFSMetaObjects to
remain associated with repos but still need to be garbage collected.
Imagine a situation where a branch is uploaded containing LFS files but
that branch is later completely deleted. The LFSMetaObjects will remain
associated with the Repository but the Repository will no longer contain
any pointers to the object.
This PR adds a second doctor command to perform a full GC.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Gitea will attempt to lookup its location using LookPath however, this
fails on cmd.exe if gitea is in the current working directory.
exec.LookPath will return an exec.ErrDot error which we can test for and
then simply using filepath.Abs(os.Args[0]) to absolute gitea against the
current working directory.
Fix #22063
Signed-off-by: Andrew Thornton <art27@cantab.net>
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: techknowlogick <techknowlogick@gitea.io>
Close #14601
Fix #3690
Revive of #14601.
Updated to current code, cleanup and added more read/write checks.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Signed-off-by: Andre Bruch <ab@andrebruch.com>
Co-authored-by: zeripath <art27@cantab.net>
Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: Norwin <git@nroo.de>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Gitea will migrate the database model version automatically, but it
should be able to be disabled and keep Gitea shutdown if the version is
not matched.
Although there are per-locale fallbacks for ambiguity the locale names
for Chinese do not quite match our locales. This PR simply maps zh-CN on
to zh-hans and other zh variants on to zh-hant.
Ref #20999
Signed-off-by: Andrew Thornton <art27@cantab.net>
When getting tracked times out of the db and loading their attributes
handle not exist errors in a nicer way. (Also prevent an NPE.)
Fix #22006
Signed-off-by: Andrew Thornton <art27@cantab.net>
Paths in git are always separated by `/` not `\` - therefore we should
`path` and not `filepath`
Fix #21987
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: Lauris BH <lauris@nix.lv>
Provide a new type to make it easier to parse a ref name.
Actually, it's picked up from #21937, to make the origin PR lighter.
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
A complement to #21985.
I overlooked it because the name of the switch is `StartServer`, not
`Enabled`. I believe the weird name is a legacy, but renaming is out of
scope.
Fixes #21865.
Scheme-based normalization ([RFC 3986, section
6.2.3](https://www.rfc-editor.org/rfc/rfc3986#section-6.2.3)) was
already implemented, but only for `defaultAppURL`.
This PR implements the same for `AppURL`.
Signed-off-by: Saswat Padhi <saswatpadhi@protonmail.com>
Co-authored-by: John Olheiser <john.olheiser@gmail.com>
`hex.EncodeToString` has better performance than `fmt.Sprintf("%x",
[]byte)`, we should use it as much as possible.
I'm not an extreme fan of performance, so I think there are some
exceptions:
- `fmt.Sprintf("%x", func(...)[N]byte())`
- We can't slice the function return value directly, and it's not worth
adding lines.
```diff
func A()[20]byte { ... }
- a := fmt.Sprintf("%x", A())
- a := hex.EncodeToString(A()[:]) // invalid
+ tmp := A()
+ a := hex.EncodeToString(tmp[:])
```
- `fmt.Sprintf("%X", []byte)`
- `strings.ToUpper(hex.EncodeToString(bytes))` has even worse
performance.
Change all license headers to comply with REUSE specification.
Fix #16132
Co-authored-by: flynnnnnnnnnn <flynnnnnnnnnn@github>
Co-authored-by: John Olheiser <john.olheiser@gmail.com>
Unfortunately the fallback configuration code for [mailer] that were
added in #18982 are incorrect. When you read a value from an ini section
that key is added. This leads to a failure of the fallback mechanism.
Further there is also a spelling mistake in the startTLS configuration.
This PR restructures the mailer code to first map the deprecated
settings on to the new ones - and then use ini.MapTo to map those on to
the struct with additional validation as necessary.
Ref #21744
Signed-off-by: Andrew Thornton <art27@cantab.net>
This PR addresses #19586
I added a mutex to the upload version creation which will prevent the
push errors when two requests try to create these database entries. I'm
not sure if this should be the final solution for this problem.
I added a workaround to allow a reupload of missing blobs. Normally a
reupload is skipped because the database knows the blob is already
present. The workaround checks if the blob exists on the file system.
This should not be needed anymore with the above fix so I marked this
code to be removed with Gitea v1.20.
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Committer avatar rendered by `func AvatarByEmail` are not vertical align
as `func Avatar` does.
- Replace literals `ui avatar` and `ui avatar vm` with the constant
`DefaultAvatarClass`
Although git does expect that author names should be of the form: `NAME
<EMAIL>` some users have been able to create commits with: `<EMAIL>`
Fix #21900
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: Lauris BH <lauris@nix.lv>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Enable this to require captcha validation for user login. You also must
enable `ENABLE_CAPTCHA`.
Summary:
- Consolidate CAPTCHA template
- add CAPTCHA handle and context
- add `REQUIRE_CAPTCHA_FOR_LOGIN` config and docs
- Consolidate CAPTCHA set-up and verification code
Partially resolved #6049
Signed-off-by: Xinyu Zhou <i@sourcehut.net>
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Andrew Thornton <art27@cantab.net>
Fix #20456
At some point during the 1.17 cycle abbreviated refishs to issue
branches started breaking. This is likely due serious inconsistencies in
our management of refs throughout Gitea - which is a bug needing to be
addressed in a different PR. (Likely more than one)
We should try to use non-abbreviated `fullref`s as much as possible.
That is where a user has inputted a abbreviated `refish` we should add
`refs/heads/` if it is `branch` etc. I know people keep writing and
merging PRs that remove prefixes from stored content but it is just
wrong and it keeps causing problems like this. We should only remove the
prefix at the time of
presentation as the prefix is the only way of knowing umambiguously and
permanently if the `ref` is referring to a `branch`, `tag` or `commit` /
`SHA`. We need to make it so that every ref has the appropriate prefix,
and probably also need to come up with some definitely unambiguous way
of storing `SHA`s if they're used in a `ref` or `refish` field. We must
not store a potentially
ambiguous `refish` as a `ref`. (Especially when referring a `tag` -
there is no reason why users cannot create a `branch` with the same
short name as a `tag` and vice versa and any attempt to prevent this
will fail. You can even create a `branch` and a
`tag` that matches the `SHA` pattern.)
To that end in order to fix this bug, when parsing issue templates check
the provided `Ref` (here a `refish` because almost all users do not know
or understand the subtly), if it does not start with `refs/` add the
`BranchPrefix` to it. This allows people to make their templates refer
to a `tag` but not to a `SHA` directly. (I don't think that is
particularly unreasonable but if people disagree I can make the `refish`
be checked to see if it matches the `SHA` pattern.)
Next we need to handle the issue links that are already written. The
links here are created with `git.RefURL`
Here we see there is a bug introduced in #17551 whereby the provided
`ref` argument can be double-escaped so we remove the incorrect external
escape. (The escape added in #17551 is in the right place -
unfortunately I missed that the calling function was doing the wrong
thing.)
Then within `RefURL()` we check if an unprefixed `ref` (therefore
potentially a `refish`) matches the `SHA` pattern before assuming that
is actually a `commit` - otherwise is assumed to be a `branch`. This
will handle most of the problem cases excepting the very unusual cases
where someone has deliberately written a `branch` to look like a `SHA1`.
But please if something is called a `ref` or interpreted as a `ref` make
it a full-ref before storing or using it. By all means if something is a
`branch` assume the prefix is removed but always add it back in if you
are using it as a `ref`. Stop storing abbreviated `branch` names and
`tag` names - which are `refish` as a `ref`. It will keep on causing
problems like this.
Fix #20456
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Lauris BH <lauris@nix.lv>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
This patch provide a mechanism to disable RSS/Atom feed.
Signed-off-by: Xinyu Zhou <i@sourcehut.net>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: 6543 <6543@obermui.de>
This PR adds a context parameter to a bunch of methods. Some helper
`xxxCtx()` methods got replaced with the normal name now.
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
- 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.
* Closes #19997
Adds an `updated_at` time field to the `PullReview` API object to
specify when the pull request review's state changed.
Signed-off-by: Yarden Shoham <hrsi88@gmail.com>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
