- Doing 64-bit atomic operations on 32-bit machines is a bit tricky by
golang, as they can only be done under certain set of
conditions(https://pkg.go.dev/sync/atomic#pkg-note-BUG).
- This PR fixes such case whereby the conditions weren't met, it moves
the int64 to the first field of the struct, which will 64-bit operations
happening on this property on 32-bit machines.
- Resolves #19518
Within doArchive there is a service goroutine that performs the
archiving function. This goroutine reports its error using a `chan
error` called `done`. Prior to this PR this channel had 0 capacity
meaning that the goroutine would block until the `done` channel was
cleared - however there are a couple of ways in which this channel might
not be read.
The simplest solution is to add a single space of capacity to the
goroutine which will mean that the goroutine will always complete and
even if the `done` channel is not read it will be simply garbage
collected away.
(The PR also contains two other places when setting up the indexers
which do not leak but where the blocking of the sending goroutine is
also unnecessary and so we should just add a small amount of capacity
and let the sending goroutine complete as soon as it can.)
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: 6543 <6543@obermui.de>
- Apparently `<input>` elements differ from other elements have a size
attribute that act as a `min-width` CSS property, this causes a overflow
on mobile. By setting this size to `1` it doesn't try to force a
min-width and nicely shrink down.
This gets the necessary data to the issuelist for it to support a clickable commit status icon which pops up the full list of commit statuses related to the commit. It accomplishes this without any additional queries or fetching as the existing codepath was already doing the necessary work but only returning the "last" status. All methods were wrapped to call the least-filtered version of each function in order to maximize code reuse.
Note that I originally left `getLastCommitStatus()` in `pull.go` which called to the new function, but `make lint` complained that it was unused, so I removed it. I would have preferred to keep it, but alas.
The only thing I'd still like to do here is force these popups to happen to the right by default instead of the left. I see that the only other place this is popping up right is on view_list.tmpl, but I can't figure out how/why right now.
Fixes #18810
* Improve dashboard's repo list performance
- Avoid a lot of database lookups for all the repo's, by adding a
undocumented "minimal" mode for this specific task, which returns the
data that's only needed by this list which doesn't require any database
lookups.
- Makes fetching these list faster.
- Less CPU overhead when a user visits home page.
* Refactor javascript code + fix Fork icon
- Use async in the function so we can use `await`.
- Remove `archivedFilter` check for count, as it doesn't make sense to
show the count of repos when you can't even see them(as they are
filited away).
* Add `count_only`
* Remove uncessary code
* Improve comment
Co-authored-by: delvh <dev.lh@web.de>
* Update web_src/js/components/DashboardRepoList.js
Co-authored-by: delvh <dev.lh@web.de>
* Update web_src/js/components/DashboardRepoList.js
Co-authored-by: delvh <dev.lh@web.de>
* By default apply minimal mode
* Remove `minimal` paramater
* Refactor count header
* Simplify init
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: zeripath <art27@cantab.net>
* By default force vertical tabs on mobile
- While experimenting with using vertical tabs instead of horizontal
tabs on gitea for a better mobile experience, I made a recent
PR(https://github.com/go-gitea/gitea/pull/19468) in order to see if
there was any objections to this new behavior for the repo headers(one
of the most annoying horizontal tabs). This PR had no objections and
even a user commenting that this change is brilliant.
- This PR now improves upon the previous PR by making this the de-facto
behavior for all menu's on mobile. The only exemption is the navbar
which also uses the menu but caught some layout errors with the changes.
* Fix organisation
* Fix repo/wiki buttons
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
TestAPIGitTag (and likely others) will fail if the running environment contains
GIT_AUTHOR_NAME and other env variables like it.
This PR simply unsets these when running the integration tests.
Fix #14247
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Allow commit status popup on /pulls page
The /pulls page doesn't contain a "repository" element, so the early-out here was preventing the commit status popup hook from working. However, the only thing the .repository element is being used for here is determining whether the popup should be on the right or on the left, so we don't actually need the element to exist for the hook to work.
Pull request #19375 allows the statuses on /pulls pages to appear clickable, but this commit is required to make the popup actually work there.
* Move commit statuses popup hook to dedicated func
* Add missing import
* Use horizontal tabs for repo header on mobile
- The current behavior of the repo header on mobile is to display them
vertically column-by-column. I've only experience annoyance due to this
while trying to visit gitea instanced on mobile. This commit changes
this behavior to use horizontal tabs, it uses less tabs and doesn't
bloat 60% of your mobile screen with the repo headers.
- A small fix added in this commit is to give some space around the repo
buttons, current behavior is that they are too "close" to the repo
title.
* Fix lint
If an `os/exec.Command` is passed non `*os.File` as an input/output, go
will create `os.Pipe`s and wait for their closure in `cmd.Wait()`. If
the code following this is responsible for closing `io.Pipe`s or other
handlers then on process death from context cancellation the `Wait` can
hang.
There are two possible solutions:
1. use `os.Pipe` as the input/output as `cmd.Wait` does not wait for these.
2. create a goroutine waiting on the context cancellation that will close the inputs.
This PR provides the second option - which is a simpler change that can
be more easily backported.
Closes #19448
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Set correct PR status on 3way on conflict checking
- When 3-way merge is enabled for conflict checking, it has a new
interesting behavior that it doesn't return any error when it found a
conflict, so we change the condition to not check for the error, but
instead check if conflictedfiles is populated, this fixes a issue
whereby PR status wasn't correctly on conflicted PR's.
- Refactor the mergeable property(which was incorrectly set and lead me this
bug) to be more maintainable.
- Add a dedicated test for conflicting checking, so it should prevent
future issues with this.
* Fix linter
- Don't panic on `ErrEmailInvalid`, this was caused due that we were
trying to force `ErrEmailCharIsNotSupported` interface, which panics.
- Resolves #19397
* When dumping trim the standard suffices instead of a random suffix
Instead of using the `path.Ext()` to trim the last "extension" suffix, just iterate
through the supported suffices and trim those.
Fix #19424
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fix enum with to have correct supported types only
Co-authored-by: 6543 <6543@obermui.de>
* Don't allow merging PR's which are being conflict checked
- When a PR is still being conflict checked, don't allow the PR to be
merged(the merge button could already be visible before e.g. a new
commit was pushed to the PR).
- Relevant(should prevent such issue from happening) #19352
Co-authored-by: delvh <dev.lh@web.de>
