Related to https://codeberg.org/Codeberg/Community/issues/1944
* Allowed the githubdownloaderv3 to know whether issues and, or PRs are requested to migrate
* Used this information to decide to filter for "/pulls/" or "/issues"
* Or not to filter at all if issues == true && prs == true
* Added isolated test for the downloader and for the uploader
* Created a new test_repo in github.com/forgejo and set it up properly together with @Gusted
* Updated github_downloader_test with the new URLs and test data from the repo
* Recorded the API calls for local testing
* Added a minimal gitbucket test (which uses the github downloader under the hood)
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8892
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-authored-by: patdyn <patdyn@noreply.codeberg.org>
Co-committed-by: patdyn <patdyn@noreply.codeberg.org>
- gopkg.in/yaml.v3 is archived and unmaintained
- go.yaml.in/yaml/v3 is a compatible fork under the umbrella
of https://yaml.org/
### Tests
There is no need for more tests than already provided: it is like an upgrade to a minor version, only from a fork. I browsed the changes and there are some bug fixes. They all seem reasonably minimal. It is not one of those forks that went crazy with breaking changes 😁 And there is a non zero chance that [a bug that matters to Forgejo Actions](https://github.com/yaml/go-yaml/issues/76) is fixed there. It is rare and can wait but it did happen on Codeberg.
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8956
Reviewed-by: oliverpool <oliverpool@noreply.codeberg.org>
Co-authored-by: Earl Warren <contact@earl-warren.org>
Co-committed-by: Earl Warren <contact@earl-warren.org>
Add support for migrating Pagure repositories (including metadata) to Forgejo via the migration tool.
One 'discrepancy' with migration from other forges is how privates issues are handled, they are migrated when a API token is set for the migration, in that case users are advised to set their repositories visibility to private to avoid leaking such private issues.
Co-authored-by: Akashdeep Dhar <akashdeep.dhar@gmail.com>
Co-authored-by: Ryan Lerch <rlerch@redhat.com>
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8513
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Co-committed-by: Akashdeep Dhar <akashdeep.dhar@gmail.com>
- It seems that mid-May (according to some system notices in Codeberg) Github started (or converted all) release asssets to be made available under `https://release-assets.githubusercontent.com/`.
- Update the migration code to allow this baseURL for Github release assets.
- ResolvesCodeberg/Community#2061
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8771
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: Gusted <postmaster@gusted.xyz>
Co-committed-by: Gusted <postmaster@gusted.xyz>
This PR contains the following updates:
| Package | Type | Update | Change |
|---|---|---|---|
| [gitlab.com/gitlab-org/api/client-go](https://gitlab.com/gitlab-org/api/client-go) | require | minor | `v0.129.0` -> `v0.130.1` |
---
### Release Notes
<details>
<summary>gitlab-org/api/client-go (gitlab.com/gitlab-org/api/client-go)</summary>
### [`v0.130.1`](https://gitlab.com/gitlab-org/api/client-go/tags/v0.130.1)
[Compare Source](https://gitlab.com/gitlab-org/api/client-go/compare/v0.130.0...v0.130.1)
#### [0.130.1](https://gitlab.com/gitlab-org/api/client-go/compare/v0.130.0...v0.130.1) (2025-06-11)
##### Bug Fixes
* add missing nil check on create group with avatar ([3298a05](3298a058f3))
### [`v0.130.0`](https://gitlab.com/gitlab-org/api/client-go/tags/v0.130.0)
[Compare Source](https://gitlab.com/gitlab-org/api/client-go/compare/v0.129.0...v0.130.0)
### [0.130.0](https://gitlab.com/gitlab-org/api/client-go/compare/v0.129.0...v0.130.0) (2025-06-11)
##### Bug Fixes
* **workflow:** the `release.config.mjs` file mustn't be hidden ([5d423a5](5d423a55d5))
##### Features
* add "emoji_events" support to group hooks ([c6b770f](c6b770f350))
* Add `active` to ListProjects ([7818155](78181558db))
* add generated_file field for MergeRequestDiff ([4b95dac](4b95dac3ef))
* add support for `administrator` to Group `project_creation_level` enums ([664bbd7](664bbd7e3c))
* add the `WithTokenSource` client option ([6ccfcf8](6ccfcf857a))
* add url field to MergeCommentEvent.merge_request ([bd639d8](bd639d811c))
* implement a specialized `TokenSource` interface ([83c2e06](83c2e06cbe))
* **projects:** add ci_push_repository_for_job_token_allowed parameter ([3d539f6](3d539f66fd))
* **terraform-states:** add Terraform States API ([082b81c](082b81cd45))
</details>
---
### Configuration
📅 **Schedule**: Branch creation - Between 12:00 AM and 03:59 AM ( * 0-3 * * * ) (UTC), Automerge - Between 12:00 AM and 03:59 AM ( * 0-3 * * * ) (UTC).
🚦 **Automerge**: Disabled by config. Please merge this manually once you are satisfied.
♻ **Rebasing**: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.
🔕 **Ignore**: Close this PR and you won't be reminded about this update again.
---
- [ ] <!-- rebase-check -->If you want to rebase/retry this PR, check this box
---
This PR has been generated by [Renovate Bot](https://github.com/renovatebot/renovate).
<!--renovate-debug:eyJjcmVhdGVkSW5WZXIiOiI0MC40OC40IiwidXBkYXRlZEluVmVyIjoiNDAuNDguNCIsInRhcmdldEJyYW5jaCI6ImZvcmdlam8iLCJsYWJlbHMiOlsiZGVwZW5kZW5jeS11cGdyYWRlIiwidGVzdC9ub3QtbmVlZGVkIl19-->
Co-authored-by: Earl Warren <contact@earl-warren.org>
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8163
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: Renovate Bot <forgejo-renovate-action@forgejo.org>
Co-committed-by: Renovate Bot <forgejo-renovate-action@forgejo.org>
urfave/cli v2 will eventually become unmaintained, switch over to v3 which is the latest supported version.
Note: the `docs` command would be a lot of work to restore with v3 ([the package is still in alpha](https://github.com/urfave/cli-docs)) An alternative to avoid a breaking change would be to not upgrade from v2 to v3 for that reason alone.
Note: these commits were cherry-picked from https://code.forgejo.org/forgefriends/forgefriends
Note: it is best reviewed side by side with no display of whitespace changes (there are a lot of those when converting vars to func).
- a few functional changes were necessary and are noted in context in the file changes tab
- https://cli.urfave.org/migrate-v2-to-v3/ upgrade instructions were followed in the most minimal way possible
- upgrade gof3 to v3.10.8 which includes and upgrade from urfave/cli v2 to urfave/cli v3
- upgrade gitlab.com/gitlab-org/api/client-go v0.129.0 because it is an indirect dependency of gof3 and requires a change because of a deprecated field that otherwise triggers a lint error but nothing else otherwise
- verified that the [script](https://codeberg.org/forgejo/docs/src/branch/next/scripts/cli-docs.sh) that generates the [CLI documentation](https://codeberg.org/forgejo/docs/src/branch/next/scripts/cli-docs.sh) still works. There are cosmetic differences and the **help** subcommand is no longer advertised (although it is still supported) but the `--help` option is advertised as expected so it is fine.
- end-to-end tests [passed](https://code.forgejo.org/forgejo/end-to-end/pulls/667) (they use the Forgejo CLI to some extent)
## Checklist
The [contributor guide](https://forgejo.org/docs/next/contributor/) contains information that will be helpful to first time contributors. There also are a few [conditions for merging Pull Requests in Forgejo repositories](https://codeberg.org/forgejo/governance/src/branch/main/PullRequestsAgreement.md). You are also welcome to join the [Forgejo development chatroom](https://matrix.to/#/#forgejo-development:matrix.org).
### Tests
- I added test coverage for Go changes...
- [ ] in their respective `*_test.go` for unit tests.
- [ ] in the `tests/integration` directory if it involves interactions with a live Forgejo server.
- I added test coverage for JavaScript changes...
- [ ] in `web_src/js/*.test.js` if it can be unit tested.
- [ ] in `tests/e2e/*.test.e2e.js` if it requires interactions with a live Forgejo server (see also the [developer guide for JavaScript testing](https://codeberg.org/forgejo/forgejo/src/branch/forgejo/tests/e2e/README.md#end-to-end-tests)).
### Documentation
- [ ] I created a pull request [to the documentation](https://codeberg.org/forgejo/docs) to explain to Forgejo users how to use this change.
- [x] I did not document these changes and I do not expect someone else to do it.
### Release notes
- [ ] I do not want this change to show in the release notes.
- [ ] I want the title to show in the release notes with a link to this pull request.
- [x] I want the content of the `release-notes/<pull request number>.md` to be be used for the release notes instead of the title.
<!--start release-notes-assistant-->
## Release notes
<!--URL:https://codeberg.org/forgejo/forgejo-->
- Breaking features
- [PR](https://codeberg.org/forgejo/forgejo/pulls/8035): <!--number 8035 --><!--line 0 --><!--description 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-->The `forgejo docs` command is deprecated and CLI errors are now displayed on stderr instead of stdout. These breaking changes happened because the package used to parse the command line arguments was [upgraded from v2 to v3](https://cli.urfave.org/migrate-v2-to-v3/). A [separate project was initiated](https://github.com/urfave/cli-docs) to re-implement the `docs` command, but it is not yet production ready.<!--description-->
<!--end release-notes-assistant-->
Co-authored-by: limiting-factor <limiting-factor@posteo.com>
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/8035
Reviewed-by: Gusted <gusted@noreply.codeberg.org>
Git authorization was not taking into account multiple token feature,
leading to auth failures
Closes: https://github.com/go-gitea/gitea/issues/34141
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit 8a6df00c532becd4d10efb70827ccf80b2bf74e2)
- When migrating a pull requests from a Gitea or Forgejo instance, check if the pull request was created via the AGit flow and transfer that bit of information to the migrated pull request.
- Expose this bit of information as the `flow` field for the pull request.
- We have to do a horrible Go hack with Gitea's [go-sdk](gitea.com/gitea/go-sdk) to list all pull requests while being able to decode it to a struct that contains the new `Flow` field. The library does not allow you to do this out of the box, so we have to use `go:linkname` to access the private method that allows us to do this. This in turn means we have to do some boilerplate code that the library otherwise would do for us. The better option would be forking, but that would be a hassle of keeping the library in sync.
- Resolvesforgejo/forgejo#5848
- Unit test added.
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/7421
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: Gusted <postmaster@gusted.xyz>
Co-committed-by: Gusted <postmaster@gusted.xyz>
- In the case that a deleted user modified the state of an issue or pull request, the user field in the API response for that state event will be `null`. Handle this by falling back to Forgejo's internal Ghost user.
- No testing, this bug was hit on Codeberg with a instance that is only IPv6-accessible and otherwise might be phased out. So I will do some mental gymnastics and argue, migration feature will someday be replaced by F3 and considering the logic that was added its not worth the tradeoff to add testing for this by trying to recreate the same scenario on another Gitlab instance and then use that as a testing vector. To still give some confidence in this patch, it was confirmed that this exact fix worked on Codeberg.
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/7210
Reviewed-by: Otto <otto@codeberg.org>
Co-authored-by: Gusted <postmaster@gusted.xyz>
Co-committed-by: Gusted <postmaster@gusted.xyz>
- Add a new function `IsPushMirrorURLAllowed` that will allow `ssh://` url and make the existing `IsMigrateURLAllowed` not allow such URLs anymore.
- Resolvesforgejo/forgejo#6960
- Existing integration tests make sure that SSH urls are still allowed for the push mirror feature and added unit test to ensure that `IsMigrateURLAllowed` no longer allows SSH urls.
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/7004
Reviewed-by: 0ko <0ko@noreply.codeberg.org>
Co-authored-by: Gusted <postmaster@gusted.xyz>
Co-committed-by: Gusted <postmaster@gusted.xyz>
* cleanup: remove not used properties
* feat: implement migration of website field from gogs
* feat: implement dumping and restoring website field
* feat: implement migration of website field from gitea
* feat: implement migration of homepage/website field from github
* feat: implement website properties for repository migration
Gogs migration is untested for now.
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/6474
Reviewed-by: Otto <otto@codeberg.org>
Co-authored-by: ThomasBoom89 <thomasboom89@noreply.codeberg.org>
Co-committed-by: ThomasBoom89 <thomasboom89@noreply.codeberg.org>
- Continuation of https://github.com/go-gitea/gitea/pull/18835 (by
@Gusted, so it's fine to change copyright holder to Forgejo).
- Add the option to use SSH for push mirrors, this would allow for the
deploy keys feature to be used and not require tokens to be used which
cannot be limited to a specific repository. The private key is stored
encrypted (via the `keying` module) on the database and NEVER given to
the user, to avoid accidental exposure and misuse.
- CAVEAT: This does require the `ssh` binary to be present, which may
not be available in containerized environments, this could be solved by
adding a SSH client into forgejo itself and use the forgejo binary as
SSH command, but should be done in another PR.
- CAVEAT: Mirroring of LFS content is not supported, this would require
the previous stated problem to be solved due to LFS authentication (an
attempt was made at forgejo/forgejo#2544).
- Integration test added.
- Resolves#4416
Uses `gopls check <files>` as a linter. Tested locally and brings up 149
errors currently for me. I don't think I want to fix them in this PR,
but I would like at least to get this analysis running on CI.
List of errors:
```
modules/indexer/code/indexer.go:181:11: impossible condition: nil != nil
routers/private/hook_post_receive.go:120:15: tautological condition: nil == nil
services/auth/source/oauth2/providers.go:185:9: tautological condition: nil == nil
services/convert/issue.go:216:11: tautological condition: non-nil != nil
tests/integration/git_test.go:332:9: impossible condition: nil != nil
services/migrations/migrate.go:179:24-43: unused parameter: ctx
services/repository/transfer.go:288:48-69: unused parameter: doer
tests/integration/api_repo_tags_test.go:75:41-61: unused parameter: session
tests/integration/git_test.go:696:64-74: unused parameter: baseBranch
tests/integration/gpg_git_test.go:265:27-39: unused parameter: t
tests/integration/gpg_git_test.go:284:23-29: unused parameter: tmpDir
tests/integration/gpg_git_test.go:284:31-35: unused parameter: name
tests/integration/gpg_git_test.go:284:37-42: unused parameter: email
```
(cherry picked from commit 816222243af523316041692622be6f48ef068693)
Conflicts:
Makefile
trivial context conflict and also ask renovate to watch over it
do not include it in lint-backend because the errors are not fixed
When performing migrations, and need to remap external users to local
ones, when no local mapping is possible, map the external user to Ghost,
rather than the user who initiated the migration.
Mapping the external user to the migration initiator has the potential
of breaking assumptions elsewhere, like only having one review per pull
request per user. Mapping these migrated, locally unavailable users to
Ghost makes sure these - often hidden - assumptions do not break.
Fixes#3860.
Signed-off-by: Gergely Nagy <forgejo@gergo.csillger.hu>
When creating a repo, the "FORCE_PRIVATE" config option should be
respected, `readonly` doesn't work for checkbox, so it should use
`disabled` attribute.
(cherry picked from commit edbf74c418061b013a5855f604dd6be6baf34132)
Conflicts:
templates/repo/create.tmpl
templates/repo/migrate/codebase.tmpl
templates/repo/migrate/git.tmpl
templates/repo/migrate/gitbucket.tmpl
templates/repo/migrate/gitea.tmpl
templates/repo/migrate/github.tmpl
templates/repo/migrate/gitlab.tmpl
templates/repo/migrate/gogs.tmpl
templates/repo/migrate/onedev.tmpl
already in forgejo fc0c5e80da Fix and improve repo visibility checkbox when FORCE_PRIVATE is on (#3786)
enforcing FORCE_PRIVATE on repo settings was manually tested
with a repository of an unprivileged user after setting
FORCE_PRIVATE = true
More about codespell: https://github.com/codespell-project/codespell .
I personally introduced it to dozens if not hundreds of projects already and so far only positive feedback.
```
❯ grep lint-spell Makefile
@echo " - lint-spell lint spelling"
@echo " - lint-spell-fix lint spelling and fix issues"
lint: lint-frontend lint-backend lint-spell
lint-fix: lint-frontend-fix lint-backend-fix lint-spell-fix
.PHONY: lint-spell
lint-spell: lint-codespell
.PHONY: lint-spell-fix
lint-spell-fix: lint-codespell-fix
❯ git grep lint- -- .forgejo/
.forgejo/workflows/testing.yml: - run: make --always-make -j$(nproc) lint-backend checks-backend # ensure the "go-licenses" make target runs
.forgejo/workflows/testing.yml: - run: make lint-frontend
```
so how would you like me to invoke `lint-codespell` on CI? (without that would be IMHO very suboptimal and let typos sneak in)
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/3270
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: Yaroslav Halchenko <debian@onerussian.com>
Co-committed-by: Yaroslav Halchenko <debian@onerussian.com>
Resolve all cases for `unused parameter` and `unnecessary type
arguments`
Related: #30729
---------
Co-authored-by: Giteabot <teabot@gitea.io>
(cherry picked from commit e80466f7349164ce4cf3c07bdac30d736d20f035)
Conflicts:
modules/markup/markdown/transform_codespan.go
modules/setting/incoming_email.go
routers/api/v1/admin/user_badge.go
routers/private/hook_pre_receive.go
tests/integration/repo_search_test.go
resolved by discarding the change, this is linting only and
for the sake of avoiding future conflicts
Also add a test for GogsDownloaderFactory.New() to make sure
that the URL of the source repository is parsed correctly.
When the source gogs instance is hosted at a subpath like `https://git.example.com/gogs/<username>/<reponame>` the migration fails.
This PR fixes that.
Co-authored-by: hecker <tomas.hecker@gmail.com>
Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/3572
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>
Co-authored-by: varp0n <tom@gkstn.de>
Co-committed-by: varp0n <tom@gkstn.de>
Noteable additions:
- `redefines-builtin-id` forbid variable names that shadow go builtins
- `empty-lines` remove unnecessary empty lines that `gofumpt` does not
remove for some reason
- `superfluous-else` eliminate more superfluous `else` branches
Rules are also sorted alphabetically and I cleaned up various parts of
`.golangci.yml`.
(cherry picked from commit 74f0c84fa4245a20ce6fb87dac1faf2aeeded2a2)
Conflicts:
.golangci.yml
apply the linter recommendations to Forgejo code as well
This patch improves the migration from gitbucket to gitea.
The gitbucket uses it's own internal perPage value (= 25) for paging and
ignore per_page arguments in the requested URL. This cause gitea to
migrate only 25 issues and 25 PRs from gitbucket repository. This may
not happens on old gitbucket. But recent gitbucket 4.40 or 4.38.4 has
this problem.
This patch change to use this internally hardcoded perPage of gitbucket
as gitea's maxPerPage numer when migrating from gitbucket. There are
several perPage values in gitbucket like 25 for Isseus/PRs and 10 for
Releases. Some of those API doesn't support paging yet. It sounds
difficult to implement, but using the minimum number among them worked
out very well. So, I use 10 in this patch.
Brief descriptions of problems and this patch are also available in
https://github.com/go-gitea/gitea/issues/30316.
In addition, I'm not sure what kind of test cases are possible to write
here. It's a test for migration, so it requires testing gitbucket server
and gitea server, I guess. Please let me know if it is possible to write
such test cases here. Thanks!
(cherry picked from commit 7af074dbeebc3c863618992b43f84ec9e5ab9657)
See also: [GITEA] Enable mocked HTTP responses for GitLab migration test
make GITLAB_READ_TOKEN=XXXX GO_TEST_PACKAGES='code.gitea.io/gitea/services/migrations/... -test.v' FLAGS= 'test#TestGitlabDownloadRepo'
to update the captured GitLab HTTP requests used for offline testing.
Some specific events on Gitlab issues and merge requests are stored
separately from comments as "resource state events". With this change,
all relevant resource state events are downloaded during issue and merge
request migration, and converted to comments.
This PR also updates the template used to render comments to add support
for migrated comments of these types.
ref: https://docs.gitlab.com/ee/api/resource_state_events.html
(cherry picked from commit 17f170ee3724d8bdf2ddaad4211b12433f78ff0e)
GitLab generates "system notes" whenever an event happens within the
platform. Unlike Gitea, those events are stored and retrieved as text
comments with no semantic details. The only way to tell whether a
comment was generated in this manner is the `system` flag on the note
type.
This PR adds detection for a new specific kind of event: Changing the
target branch of a PR. When detected, it is downloaded using Gitea's
type for this event, and eventually uploaded into Gitea in the expected
format, i.e. with no text content in the comment.
This PR also updates the template used to render comments to add support
for migrated comments of this type.
ref:
11bd6dc826/app/services/system_notes/merge_requests_service.rb (L102)
(cherry picked from commit 6e5966597c2d498d1a8540dad965461d44ff8e57)
Now we can get object format name from git command line or from the
database repository table. Assume the column is right, we don't need to
read from git command line every time.
This also fixed a possible bug that the object format is wrong when
migrating a sha256 repository from external.
<img width="658" alt="image"
src="https://github.com/go-gitea/gitea/assets/81045/6e9a9dcf-13bf-4267-928b-6bf2c2560423">
(cherry picked from commit b79c30435f439af8243ee281310258cdf141e27b)
Conflicts:
routers/web/repo/blame.go
services/agit/agit.go
context
GitLab generates "system notes" whenever an event happens within the
platform. Unlike Gitea, those events are stored and retrieved as text
comments with no semantic details. The only way to tell whether a
comment was generated in this manner is the `system` flag on the note
type.
This PR adds detection for two specific kinds of events: Scheduling and
un-scheduling of automatic merges on a PR. When detected, they are
downloaded using Gitea's type for these events, and eventually uploaded
into Gitea in the expected format, i.e. with no text content in the
comment.
This PR also updates the template used to render comments to add support
for migrated comments of these two types.
ref:
11bd6dc826/app/services/system_notes/merge_requests_service.rb (L6-L17)
---------
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit a70c00b80bcb5de8479e407f1b8f08dcf756019d)
