Fix#32795
If a job uses a matrix, multiple `ActionRunJobs` may have the same
`JobID`. We need to merge the outputs of these jobs to make them
available to the jobs that need them.
Resolve#31492
The response time for the Pull Requests API has improved significantly,
dropping from over `2000ms` to about `350ms` on my local machine. It's
about `6` times faster.
A key area for further optimization lies in batch-fetching data for
`apiPullRequest.ChangedFiles, apiPullRequest.Additions, and
apiPullRequest.Deletions`.
Tests `TestAPIViewPulls` does exist and new tests added.
- This PR also fixes some bugs in `GetDiff` functions.
- This PR also fixes data inconsistent in test data. For a pull request,
the head branch's reference should be equal to the reference in
`pull/xxx/head`.
When running e2e tests on flaky networks, gravatar can cause a timeout
and test failures. Turn off, and populate avatars on e2e test suite run
to make them reliable.
We have some actions that leverage the Gitea API that began receiving
401 errors, with a message that the user was not found. These actions
use the `ACTIONS_RUNTIME_TOKEN` env var in the actions job to
authenticate with the Gitea API. The format of this env var in actions
jobs changed with go-gitea/gitea/pull/28885 to be a JWT (with a
corresponding update to `act_runner`) Since it was a JWT, the OAuth
parsing logic attempted to parse it as an OAuth token, and would return
user not found, instead of falling back to look up the running task and
assigning it to the actions user.
Make ACTIONS_RUNTIME_TOKEN in action runners could be used,
attempting to parse Oauth JWTs. The code to parse potential old
`ACTION_RUNTIME_TOKEN` was kept in case someone is running an older
version of act_runner that doesn't support the Actions JWT.
In profiling integration tests, I found a couple places where per-test
overhead could be reduced:
* Avoiding disk IO by synchronizing instead of deleting & copying test
Git repository data. This saves ~100ms per test on my machine
* When flushing queues in `PrintCurrentTest`, invoke `FlushWithContext`
in a parallel.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
currently restricted users can only see the repos of teams in orgs they
are part at.
they also should only see the users that are also part at the same team.
---
*Sponsored by Kithara Software GmbH*
Gitea instance keeps reporting a lot of errors like "LFS SSH transfer connection denied, pure SSH protocol is disabled". When starting debugging the problem, there are more problems found. Try to address most of them:
* avoid unnecessary server side error logs (change `fail()` to not log them)
* figure out the broken tests/user2/lfs.git (added comments)
* avoid `migratePushMirrors` failure when a repository doesn't exist (ignore them)
* avoid "Authorization" (internal&lfs) header conflicts, remove the tricky "swapAuth" and use "X-Gitea-Internal-Auth"
* make internal token comparing constant time (it wasn't a serous problem because in a real world it's nearly impossible to timing-attack the token, but good to fix and backport)
* avoid duplicate routers (introduce AddOwnerRepoGitLFSRoutes)
* avoid "internal (private)" routes using session/web context (they should use private context)
* fix incorrect "path" usages (use "filepath")
* fix incorrect mocked route point handling (need to check func nil correctly)
* split some tests from "git general tests" to "git misc tests" (to keep "git_general_test.go" simple)
Still no correct result for Git LFS SSH tests. So the code is kept there
(`tests/integration/git_lfs_ssh_test.go`) and a FIXME explains the details.
Closes https://github.com/go-gitea/gitea/issues/30296
- Adds a DB fixture for actions artifacts
- Adds artifacts test files
- Clears artifacts test files between each run
- Note: I initially initialized the artifacts only for artifacts tests,
but because the files are small it only takes ~8ms, so I changed it to
always run in test setup for simplicity
- Fix some otherwise flaky tests by making them not depend on previous
tests
This is a large and complex PR, so let me explain in detail its changes.
First, I had to create new index mappings for Bleve and ElasticSerach as
the current ones do not support search by filename. This requires Gitea
to recreate the code search indexes (I do not know if this is a breaking
change, but I feel it deserves a heads-up).
I've used [this
approach](https://www.elastic.co/guide/en/elasticsearch/reference/7.17/analysis-pathhierarchy-tokenizer.html)
to model the filename index. It allows us to efficiently search for both
the full path and the name of a file. Bleve, however, does not support
this out-of-box, so I had to code a brand new [token
filter](https://blevesearch.com/docs/Token-Filters/) to generate the
search terms.
I also did an overhaul in the `indexer_test.go` file. It now asserts the
order of the expected results (this is important since matches based on
the name of a file are more relevant than those based on its content).
I've added new test scenarios that deal with searching by filename. They
use a new repo included in the Gitea fixture.
The screenshot below depicts how Gitea shows the search results. It
shows results based on content in the same way as the current version
does. In matches based on the filename, the first seven lines of the
file contents are shown (BTW, this is how GitHub does it).
Resolves#32096
---------
Signed-off-by: Bruno Sofiato <bruno.sofiato@gmail.com>
Misspell 0.5.0 supports passing a csv file to extend the list of
misspellings, so I added some common ones from the codebase. There is at
least one typo in a API response so we need to decided whether to revert
that and then likely remove the dict entry.
On creation of an empty project (no template) a default board will be
created instead of falling back to the uneditable pseudo-board.
Every project now has to have exactly one default boards. As a
consequence, you cannot unset a board as default, instead you have to
set another board as default. Existing projects will be modified using a
cron job, additionally this check will run every midnight by default.
Deleting the default board is not allowed, you have to set another board
as default to do it.
Fixes#29873Fixes#14679 along the way
Fixes#29853
Co-authored-by: delvh <dev.lh@web.de>
Refactor the webhook logic, to have the type-dependent processing happen
only in one place.
---
## Current webhook flow
1. An event happens
2. It is pre-processed (depending on the webhook type) and its body is
added to a task queue
3. When the task is processed, some more logic (depending on the webhook
type as well) is applied to make an HTTP request
This means that webhook-type dependant logic is needed in step 2 and 3.
This is cumbersome and brittle to maintain.
Updated webhook flow with this PR:
1. An event happens
2. It is stored as-is and added to a task queue
3. When the task is processed, the event is processed (depending on the
webhook type) to make an HTTP request
So the only webhook-type dependent logic happens in one place (step 3)
which should be much more robust.
## Consequences of the refactor
- the raw event must be stored in the hooktask (until now, the
pre-processed body was stored)
- to ensure that previous hooktasks are correctly sent, a
`payload_version` is added (version 1: the body has already been
pre-process / version 2: the body is the raw event)
So future webhook additions will only have to deal with creating an
http.Request based on the raw event (no need to adjust the code in
multiple places, like currently).
Moreover since this processing happens when fetching from the task
queue, it ensures that the queuing of new events (upon a `git push` for
instance) does not get slowed down by a slow webhook.
As a concrete example, the PR #19307 for custom webhooks, should be
substantially smaller:
- no need to change `services/webhook/deliver.go`
- minimal change in `services/webhook/webhook.go` (add the new webhook
to the map)
- no need to change all the individual webhook files (since with this
refactor the `*webhook_model.Webhook` is provided as argument)
Fix for regressions introduced by #28805
Enabled projects on repos created before the PR weren't detected. Also,
the way projects mode was detected in settings didn't match the way it
was detected on permission check, which leads to confusion.
Co-authored-by: Giteabot <teabot@gitea.io>
Part of #23318
Add menu in repo settings to allow for repo admin to decide not just if
projects are enabled or disabled per repo, but also which kind of
projects (repo-level/owner-level) are enabled. If repo projects
disabled, don't show the projects tab.
---------
Co-authored-by: delvh <dev.lh@web.de>
Fix#14459
The following users can add/remove review requests of a PR
- the poster of the PR
- the owner or collaborators of the repository
- members with read permission on the pull requests unit
Adds a new API `/repos/{owner}/{repo}/commits/{sha}/pull` that allows
you to get the merged PR associated to a commit.
---------
Co-authored-by: 6543 <6543@obermui.de>
Fixes#28660
Fixes an admin api bug related to `user.LoginSource`
Fixed `/user/emails` response not identical to GitHub api
This PR unifies the user update methods. The goal is to keep the logic
only at one place (having audit logs in mind). For example, do the
password checks only in one method not everywhere a password is updated.
After that PR is merged, the user creation should be next.
Fixes#22236
---
Error occurring currently while trying to revert commit using read-tree
-m approach:
> 2022/12/26 16:04:43 ...rvices/pull/patch.go:240:AttemptThreeWayMerge()
[E] [63a9c61a] Unable to run read-tree -m! Error: exit status 128 -
fatal: this operation must be run in a work tree
> - fatal: this operation must be run in a work tree
We need to clone a non-bare repository for `git read-tree -m` to work.
bb371aee6e
adds support to create a non-bare cloned temporary upload repository.
After cloning a non-bare temporary upload repository, we [set default
index](https://github.com/go-gitea/gitea/blob/main/services/repository/files/cherry_pick.go#L37)
(`git read-tree HEAD`).
This operation ends up resetting the git index file (see investigation
details below), due to which, we need to call `git update-index
--refresh` afterward.
Here's the diff of the index file before and after we execute
SetDefaultIndex: https://www.diffchecker.com/hyOP3eJy/
Notice the **ctime**, **mtime** are set to 0 after SetDefaultIndex.
You can reproduce the same behavior using these steps:
```bash
$ git clone https://try.gitea.io/me-heer/test.git -s -b main
$ cd test
$ git read-tree HEAD
$ git read-tree -m 1f085d7ed8 1f085d7ed8 9933caed00
error: Entry '1' not uptodate. Cannot merge.
```
After which, we can fix like this:
```
$ git update-index --refresh
$ git read-tree -m 1f085d7ed8 1f085d7ed8 9933caed00
```
Unfortunately, when a system setting hasn't been stored in the database,
it cannot be cached.
Meanwhile, this PR also uses context cache for push email avatar display
which should avoid to read user table via email address again and again.
According to my local test, this should reduce dashboard elapsed time
from 150ms -> 80ms .
> ### Description
> If a new branch is pushed, and the repository has a rule that would
require signed commits for the new branch, the commit is rejected with a
500 error regardless of whether it's signed.
>
> When pushing a new branch, the "old" commit is the empty ID
(0000000000000000000000000000000000000000). verifyCommits has no
provision for this and passes an invalid commit range to git rev-list.
Prior to 1.19 this wasn't an issue because only pre-existing individual
branches could be protected.
>
> I was able to reproduce with
[try.gitea.io/CraigTest/test](https://try.gitea.io/CraigTest/test),
which is set up with a blanket rule to require commits on all branches.
Fix#25565
Very thanks to @Craig-Holmquist-NTI for reporting the bug and suggesting
an valid solution!
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
## Archived labels
This adds the structure to allow for archived labels.
Archived labels are, just like closed milestones or projects, a medium to hide information without deleting it.
It is especially useful if there are outdated labels that should no longer be used without deleting the label entirely.
## Changes
1. UI and API have been equipped with the support to mark a label as archived
2. The time when a label has been archived will be stored in the DB
## Outsourced for the future
There's no special handling for archived labels at the moment.
This will be done in the future.
## Screenshots
Part of https://github.com/go-gitea/gitea/issues/25237
---------
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
- The permalink and 'Reference in New issue' URL of an renderable file
(those where you can see the source and a rendered version of it, such
as markdown) doesn't contain `?display=source`. This leads the issue
that the URL doesn't have any effect, as by default the rendered version
is shown and thus not the source.
- Add `?display=source` to the permalink URL and to 'Reference in New
Issue' if it's renderable file.
- Add integration testing.
Refs: https://codeberg.org/forgejo/forgejo/pulls/1088
Co-authored-by: Gusted <postmaster@gusted.xyz>
Co-authored-by: Giteabot <teabot@gitea.io>
Fix#25934
Add `ignoreGlobal` parameter to `reqUnitAccess` and only check global
disabled units when `ignoreGlobal` is true. So the org-level projects
and user-level projects won't be affected by global disabled
`repo.projects` unit.
Fixes (?) #25538
Fixes https://codeberg.org/forgejo/forgejo/issues/972
Regression #23879#23879 introduced a change which prevents read access to packages if a
user is not a member of an organization.
That PR also contained a change which disallows package access if the
team unit is configured with "no access" for packages. I don't think
this change makes sense (at the moment). It may be relevant for private
orgs. But for public or limited orgs that's useless because an
unauthorized user would have more access rights than the team member.
This PR restores the old behaviour "If a user has read access for an
owner, they can read packages".
---------
Co-authored-by: Giteabot <teabot@gitea.io>
Fix#25558
Extract from #22743
This PR added a repository's check when creating/deleting branches via
API. Mirror repository and archive repository cannot do that.
Related #14180
Related #25233
Related #22639Close#19786
Related #12763
This PR will change all the branches retrieve method from reading git
data to read database to reduce git read operations.
- [x] Sync git branches information into database when push git data
- [x] Create a new table `Branch`, merge some columns of `DeletedBranch`
into `Branch` table and drop the table `DeletedBranch`.
- [x] Read `Branch` table when visit `code` -> `branch` page
- [x] Read `Branch` table when list branch names in `code` page dropdown
- [x] Read `Branch` table when list git ref compare page
- [x] Provide a button in admin page to manually sync all branches.
- [x] Sync branches if repository is not empty but database branches are
empty when visiting pages with branches list
- [x] Use `commit_time desc` as the default FindBranch order by to keep
consistent as before and deleted branches will be always at the end.
---------
Co-authored-by: Jason Song <i@wolfogre.com>
Enable deduplication of unofficial reviews. When pull requests are
configured to include all approvers, not just official ones, in the
default merge messages it was possible to generate duplicated
Reviewed-by lines for a single person. Add an option to find only
distinct reviews for a given query.
fixes#24795
---------
Signed-off-by: Cory Todd <cory.todd@canonical.com>
Co-authored-by: Giteabot <teabot@gitea.io>
Fixes#24145
To solve the bug, I added a "computed" `TargetBehind` field to the
`Release` model, which indicates the target branch of a release.
This is particularly useful if the target branch was deleted in the
meantime (or is empty).
I also did a micro-optimization in `calReleaseNumCommitsBehind`. Instead
of checking that a branch exists and then call `GetBranchCommit`, I
immediately call `GetBranchCommit` and handle the `git.ErrNotExist`
error.
This optimization is covered by the added unit test.
