* Fix ParsePatch to work properly with quoted diff --git string
Currently ParsePatch fails when a diff contains a quoted diff line like:
diff --git "a/file" "b/file"
This patch makes it properly parse the line when that happens.
Fixes#6309
* Add test for regular case while here
* Simplify string modification
In #6211, we started creating repo_redirects for ownership transfers,
however that opens an edge case where a user might perform the
following sequence:
rename org1/repo1 -> org1/repo2 (creates org1/repo1 redirect)
transfer org2/repo1 -> org1/repo1 (org1/repo1 redirect continues to exist)
rename org1/repo1 -> org1/repo3 (fails due to existing org1/repo1 redirect)
This change ensures that each time we rename or transfer a repo,
we delete any existing redirects at the target location. This
already happens when a new repo is created. By doing this we ensure
that we'll never have both a repo and a redirect at the same location.
Signed-off-by: James E. Blair <jeblair@redhat.com>
* Show email if the authenticated user owns the profile page being
requested for.
Also removed `setting.UI.ShowUserEmail` as it's documentation says it
only controls the email setting on the explore page
* fix current user check... This prevents a panic as a user must be signed in before ctx.User is called
* fix panic in tests
* try to fix tests
* Update year
* Test CI fail
* Revert change
* User 3 is not allowed to authorize
* Set user2 email to be private
* Change to user4 in explore page as user2 now has private email option set
* Use git plumbing for upload: #5621 repo_editor.go: UploadRepoFile
* Use git plumbing for upload: #5621 repo_editor.go: GetDiffPreview
* Use git plumbing for upload: #5621 repo_editor.go: DeleteRepoFile
* Use git plumbing for upload: #5621 repo_editor.go: UploadRepoFiles
* Move branch checkout functions out of repo_editor.go as they are no longer used there
* BUGFIX: The default permissions should be 100644
This is a change from the previous code but is more in keeping
with the default behaviour of git.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Standardise cleanUploadFilename to more closely match git
See verify_path in: 7f4e641693/read-cache.c (L951)
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Redirect on bad paths
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Refactor to move the uploading functions out to a module
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add LFS support
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Update upload.go attribution header
Upload.go is essentially the remnants of repo_editor.go. The remaining code is essentially unchanged from the Gogs code, hence the Gogs attribution.
* Delete upload files after session committed
* Ensure that GIT_AUTHOR_NAME etc. are valid for git
see #5774
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add in test cases per @lafriks comment
* Add space between gitea and github imports
Signed-off-by: Andrew Thornton <art27@cantab.net>
* more examples in TestCleanUploadName
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fix formatting
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Set the SSH_ORIGINAL_COMMAND to ensure hooks are run
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Switch off SSH_ORIGINAL_COMMAND
Signed-off-by: Andrew Thornton <art27@cantab.net>
* fixes#5957
* add tests to make sure config option is respected
* use already defined struct
* - use migration to make the flag repo wide not for the entire gitea instance
Also note that the config value can still be set so as to be able to control the value for new repositories that are to be created
- fix copy/paste error in copyright header year and rearrange import
- use repo config instead of server config value to determine if a commit should close an issue
- update testsuite
* use global config only when creating a new repository
* allow repo admin toggle feature via UI
* fix typo and improve testcase
* fix fixtures
* add DEFAULT prefix to config value
* fix test
* Fix serving of raw wiki files other than .md
Closes#4690.
Closes#4395.
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Simplify code at routers/repo/wiki.go
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Add more files to user2/repo1.wiki for testing
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Update macaron to v1.3.2
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Add tests for WikiRaw
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Fix NewResponseWriter usage due to macaron update
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Add raw to reserved wiki names
Signed-off-by: Gabriel Silva Simões <simoes.sgabriel@gmail.com>
* Don't display buttons if there are no notices
* clear stopwatch on merging a PR
* remove redundant gt check
* use ctx.Flash as per @bkcsoft comment
* stop timer on closing issues/PRs too
* updated translation as per review
* redirect to login page after successfully activating account
* remove unrelated changes
* stop timer for issues that are closed via commits too..Not just the 'close' UI button
1. A key can either be an ssh user key or a deploy key. It cannot be both.
2. If a key is a user key - it can only be associated with one user.
3. If a key is a deploy key - it can be used in multiple repositories and the permissions it has on those repositories can be different.
4. If a repository is deleted, its deploy keys must be deleted too.
We currently don't enforce any of this and multiple repositories access with different permissions doesn't work at all. This PR enforces the following constraints:
- [x] You should not be able to add the same user key as another user
- [x] You should not be able to add a ssh user key which is being used as a deploy key
- [x] You should not be able to add a ssh deploy key which is being used as a user key
- [x] If you add an ssh deploy key to another repository you should be able to use it in different modes without losing the ability to use it in the other mode.
- [x] If you delete a repository you must delete all its deploy keys.
Fix#1357
Previous code could potentially dereference nil - this PR ensures
that the poster is loaded before dereferencing it.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Add a new config option to enable / disable the automatic watching of
repos for new repositories and if a user is added to a team.
Fixes#653
Signed-off-by: Sebastian Sauer <sauer.sebastian@gmail.com>
* Ensure valid git author names passed in signatures
Fix#5772 - Git author names are not allowed to include `\n` `<` or `>` and
must not be empty. Ensure that the name passed in a signature is valid.
* Account for pathologically named external users
LDAP and the like usernames are not checked in the same way that users who signup are.
Therefore just ensure that user names are also git safe and if totally pathological -
Set them to "user-$UID"
* Add Tests and adjust test users
Make our testcases a little more pathological so that we be sure that integration
tests have a chance to spot these cases.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Optimize pulls merging
By utilizing `git clone -s --no-checkout` rather than cloning the whole
repo.
* Use sparse-checkout to speedup pulls merge
* Use bytes.Buffer instead of strings.Builder for backward compatibility
* Fix empty diff-tree output for repos with only the initial commit
* Fix missing argument for the format string
* Rework diff-tree-list generation
* Remove logging code
* File list for sparse-checkout must be prefix with /
Otherwise, they would match all files with the same name under
subdirectories.
* Update onto the rebased head
* Use referecen repo to avoid fetching objects
* Pooled and buffered gzip implementation
* Add test for gzip
* Add integration test
* Ensure lfs check within transaction
The previous code made it possible for a race condition to occur whereby a LFSMetaObject could be checked into the database twice. We should check if the LFSMetaObject is within the database and insert it if not in one transaction.
* Try to avoid primary key problem in postgres
The integration tests are being affected by
https://github.com/go-testfixtures/testfixtures/issues/39 if we set the
primary key high enough, keep a count of this and remove at the end of
each test we shouldn't be affected by this.
Unfortunately the last fix didn't completely fix the migration to v79 of the db
due to bug with schema locking during Sync2. This should fix this issue.
Fix#5759
Signed-off-by: Andrew Thornton <art27@cantab.net>
* api: Add an endpoint to list a particular member of team.
* models: Rename `GetUserTeams()` to `GetUserOrgTeams()` in `org_team` model.
`GetUserTeams()` sounds a bit misnomer since it actually returns
the teams that user belongs to in a given organization rather than
all the teams across all the organization that the user has joined.
* models: Add `GetUserTeams()`.
Returns all the teams that a user belongs to.
* api: Add an endpoint for GET '/user/teams'.
A GET request to this endpoint lists all the teams that a user
belongs to.
* Fixed deadlock in CreateComment
* Fix possible deadlock in UpdateIssueDeadline from createDeadlineComment
* Ensure that calls to IsTimeTracker enabled are called within session
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Ensure that calls to reactionList are also called within session
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Ensure all calls in NewPullRequest with the session are called within the session
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Deal with potential deadlocks in repo
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Ensure that isStaring is checked within our transaction
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Fix mistake in isOrganizationMember
Sorry.
* add migration and basic UI for changing a user's theme
* update user themem
* use right text on button
* load theme based on users' selection
* load theme based on users' selection in pwa too
* update sample config
* delete older theme loading
* implement AfterLoad to set users' theme properly
* set up default theme when creating a user. This uses the installation wide theme
* use flash messages for error
* set default theme when creating a user from the cli
* fix @lunny review
* Fix sqlite deadlock when assigning to a PR
Fix 5639
Signed-off-by: Andrew Thornton <art27@cantab.net>
* More possible deadlocks found and fixed
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Added test environment for m$sql
* Added template for test environment for m$sql
* Fix password
* Fix password (again)
* Fix password (again again)
* Fix db
* Ci trigger (Looking at you drone....)
* Ci trigger (Looking at you drone....)
* Ci trigger (Looking at you drone....)
* Ci trigger (Looking at you drone....)
* Create master database for mssql integration tests
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Create database only if master do not exist
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Fix mssql integration tests by using custom database "gitea"
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Moved defer
* bump xorm
* updated xorm
* Fixed build
* Add branch protection for approvals
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Add required approvals
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Add missing comments and fmt
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Add type = approval and group by reviewer_id to review
* Prevent users from adding negative review limits
* Add migration for approval whitelists
Signed-off-by: Jonas Franz <info@jonasfranz.software>
MSSQL is using the wrong type here which results in a strconv.ParseInt: parsing "2018-12-07T00:00:00Z": invalid syntax error.
The added datediff(SECOND, '19700101', x) results in the unix timestamp to be returned.
Signed-off-by: Christopher Dziomba <christopher.dziomba@gmail.com>
right now the `closed_at` field for json responses is not filled during
the `APIIssue` creation for api responses.
For a closed issue you get a result like:
```json
"state":"open","comments":0,"created_at":"2018-11-29T16:39:24+01:00",
"updated_at":"2018-11-30T10:49:19+01:00","closed_at":null,
"due_date":null,"pull_request":null}
```
which has no information about the closing date. (which exists in the
db and ui)
with this PR the result changes to this:
```json
:null,"assignee":null,"assignees":null,
"state":"closed",
"comments":0,"created_at":"2018-11-29T16:43:05+01:00",
"updated_at":"2018-12-02T19:17:05+01:00",
"closed_at":"2018-12-02T19:17:05+01:00",
"due_date":null,"pull_request":null}
```
fixes: https://github.com/go-gitea/gitea/issues/5446
Signed-off-by: Roman <romaaan.git@gmail.com>
* add milestone issues and pulls page instead of redirecting issues page
* add milestone when creating issue from milestone page
* refactor to merge similiar codes as a new function issues
* remove milestone info on milestone issues list
* fix missing params
* show only opened milestones on issues page milestone filter
* update Godoc
* update Godoc everywhere
* update swagger
* use false instead of 0
* Add seccond ordering by ID for milestones where no deadline is set
* Remove field from migration to support upgrades from older version
That will ensure the field does not get queried in the Select if it does
not exist yet:
```
[I] [SQL] SELECT "id", "repo_id", "index", "poster_id", "name", "content", "milestone_id", "priority", "assignee_id", "is_closed", "is_pull", "num_comments", "ref", "deadline_unix", "created_unix", "updated_unix
[...itea/routers/init.go:60 GlobalInit()] [E] Failed to initialize ORM engine: migrate: do migrate: pq: column "ref" does not exist
```
see #5318
* Skip remove stale watcher migration if not required
Otherwise the migration will fail if executed from a older database
version without multiple IssueWatch feature.
```
2018/11/11 23:51:14 [I] [SQL] SELECT DISTINCT "issue_watch"."user_id", "issue"."repo_id" FROM "issue_watch" INNER JOIN issue ON issue_watch.issue_id = issue.id WHERE (issue_watch.is_watching = $1) LIMIT 50 []int
[...itea/routers/init.go:60 GlobalInit()] [E] Failed to initialize ORM engine: migrate: do migrate: pq: relation "issue_watch" does not exist
```
see #5318
* Add private information to the deploy keys api
This commit adds more information to the deploy keys to allow for back
reference in to the main keys list. It also adds information about the
repository that the key is referring to.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add private information to the user keys API
This adjusts the keys API to give out private information to user keys if
the current user is the owner or an admin.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add ability to search keys by fingerprint
This commit adds the functionality to search ssh-keys by fingerprint of
the ssh-key. Deploy keys per repository can also be searched. There is
no current clear API point to allow search of all deploy keys by
fingerprint or keyID.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Add integration test
* Fix issue where ecdsa and other key types are not synced from LDAP authentication provider fixes#5092
* integrations/auth_ldap_test.go: Add Hermes Conrad new ecdsa-sha2-nistp256 publickey fingerprint
* integrations/auth_ldap_test.go: Use ssh-keygen -lf <filename> -E sha256
* Added basic heatmap data
* Added extra case for sqlite
* Built basic heatmap into user profile
* Get contribution data from api & styling
* Fixed lint & added extra group by statements for all database types
* generated swagger spec
* generated swagger spec
* generated swagger spec
* fixed swagger spec
* fmt
* Added tests
* Added setting to enable/disable user heatmap
* Added locale for loading text
* Removed UseTiDB
* Updated librejs & moment.js
* Fixed import order
* Fixed heatmap in postgresql
* Update docs/content/doc/advanced/config-cheat-sheet.en-us.md
Co-Authored-By: kolaente <konrad@kola-entertainments.de>
* Added copyright header
* Fixed a bug to show the heatmap for the actual user instead of the currently logged in
* Added integration test for heatmaps
* Added a heatmap on the dashboard
* Fixed timestamp parsing
* Hide heatmap on mobile
* optimized postgresql group by query
* Improved sqlite group by statement
`show` is keyword in MySQL and has to be quoted to reference a column name. Use grave accents (ASCII code 96) for quoting to match rest of the source code. It's non-standard SQL, but it's supported by SQLite and MySQL.
Signed-off-by: Filip Navara <navara@emclient.com>
* add notification interface and refactor UI notifications
* add missing methods on notification interface and notifiy only issue status really changed
* implement NotifyPullRequestReview for ui notification
* Hide outdated comments in file view
Signed-off-by: Jonas Franz <info@jonasfranz.software>
* Add unit test by adding "invalidated" comment to fixtures
Signed-off-by: Jonas Franz <info@jonasfranz.software>
opening issues generates a webhook to discord that contains
a url to the gitea api. the message title in discord is therefore
referencing to the api instead of the issue itself.
* Code review UI improvements
* More fixes to dark theme
* Style fix
* Fix to allow add code review comments only on review files tab
* More readability dark style fixes
* Fix commenting on deleted files. Fixes#4752
* Fix line blame getting for multiple corner cases
* redirect to login page after successfully activating account
* force users to change password if account was created by an admin
* force users to change password if account was created by an admin
* fixed build
* fixed build
* fix pending issues with translation and wrong routes
* make sure path check is safe
* remove unneccessary newline
* make sure users that don't have to view the form get redirected
* move route to use /settings prefix so as to make sure unauthenticated users can't view the page
* update as per @lafriks review
* add necessary comment
* remove unrelated changes
* support redirecting to location the user actually want to go to before being forced to change his/her password
* run make fmt
* added tests
* improve assertions
* add assertion
* fix copyright year
Signed-off-by: Lanre Adelowo <yo@lanre.wtf>
* Add whitespace handling to PR-comparsion
In a PR we have to keep an eye on a lot of different things. But sometimes the
bare code is the key-thing we want to care about and just don't want to care
about fixed indention on some places. Especially if we follow the pathfinder
rule we face a lot of these situations because these changes don't break the
code in many languages but improve the readability a lot.
So this change introduce a fine graned button to adjust the way how the
reviewer want to see whitespace-changes within the code.
The possibilities reflect the possibilities from git itself except of the
`--ignore-blank-lines` flag because that one is also handled by `-b` and is
really rare.
Signed-off-by: Felix Nehrke <felix@nehrke.info>
