[git-annex](https://git-annex.branchable.com/) is a more complicated cousin to
git-lfs, storing large files in an optional-download side content. Unlike lfs,
it allows mixing and matching storage remotes, so the content remote(s) doesn't
need to be on the same server as the git remote, making it feasible to scatter
a collection across cloud storage, old harddrives, or anywhere else storage can
be scavenged. Since this can get complicated, fast, it has a content-tracking
database (`git annex whereis`) to help find everything later.
The use-case we imagine for including it in Gitea is just the simple case, where
we're primarily emulating git-lfs: each repo has its large content at the same URL.
Our motivation is so we can self-host https://www.datalad.org/ datasets, which
currently are only hostable by fragilely scrounging together cloud storage --
and having to manage all the credentials associated with all the pieces -- or at
https://openneuro.org which is fragile in its own ways.
Supporting git-annex also allows multiple Gitea instance to be annex remotes for
each other, mirroring the content or otherwise collaborating the split up the
hosting costs.
Enabling
--------
TODO
HTTP
----
TODO
Permission Checking
-------------------
This tweaks the API in routers/private/serv.go to expose the calling user's
computed permission, instead of just returning HTTP 403.
This doesn't fit in super well. It's the opposite from how the git-lfs support is
done, where there's a complete list of possible subcommands and their matching
permission levels, and then the API compares the requested with the actual level
and returns HTTP 403 if the check fails.
But it's necessary. The main git-annex verbs, 'git-annex-shell configlist' and
'git-annex-shell p2pstdio' are both either read-only or read-write operations,
depending on the state on disk on either end of the connection and what the user
asked it to ask for, with no way to know before git-annex examines the situation.
So tell the level via GIT_ANNEX_READONLY and trust it to handle itself.
In the older Gogs version, the permission was directly read in cmd/serv.go:
```
mode, err = db.UserAccessMode(user.ID, repo)
```
- 966e925cf3/internal/cmd/serv.go (L334)
but in Gitea permission enforcement has been centralized in the API layer.
(perhaps so the cmd layer can avoid making direct DB connections?)
Deletion
--------
git-annex has this "lockdown" feature where it tries
really quite very hard to prevent you deleting its
data, to the point that even an rm -rf won't do it:
each file in annex/objects/ is nested inside a
folder with read-only permissions.
The recommended workaround is to run chmod -R +w when
you're sure you actually want to delete a repo. See
https://git-annex.branchable.com/internals/lockdown
So we edit util.RemoveAll() to do just that, so now
it's `chmod -R +w && rm -rf` instead of just `rm -rf`.
Backport #28576 by wxiaoguang
Regression of #28454 . Now the string is escaped HTML, so it doesn't
need `| Safe`.
Fix#28575
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit acc8100d47bab3b13311bcf4c330b19e2153d4ca)
Backport #28555 by @fuxiaohei
Related to https://github.com/go-gitea/gitea/issues/28279
When merging artifact chunks, it lists chunks from storage. When storage
is minio, chunk's path contains `MINIO_BASE_PATH` that makes merging
break.
<del>So trim the `MINIO_BASE_PATH` when handle chunks.</del>
Update the chunk file's basename to retain necessary information. It
ensures that the directory in the chunk's path remains unaffected.
Co-authored-by: FuXiaoHei <fuxiaohei@vip.qq.com>
(cherry picked from commit 8ca32dc8733c0dad147cc7f734097dbe1113e9a9)
Backport #28552 by @6543
can we please PLEAS PLEASE only use raw SQL statements if it is relay
needed!!!
source is https://github.com/go-gitea/gitea/pull/28544 (before
refactoring)
Co-authored-by: 6543 <m.huber@kithara.com>
(cherry picked from commit 16263af9715afcfa249cf43157b2ae9b4d4765f4)
Backport #28520 by @framitdavid
There is an accessibility issue in the interface when attempting to
delete a repository. When I click on "Delete repository," a dialog box
appears, requiring confirmation to proceed with the repository deletion.
However, when I press the "Repo name" label, the wrong input field gains
focus. The focused field is located behind the dialog and is intended
for renaming the repository.
I am submitting these pull requests to ensure that the correct input
field is focused when the user clicks on the label. This change will
also facilitate the writing of tests using Playwright or Testing Library
to retrieve elements based on roles. This PR will also improve
acessibility of this area.
Co-authored-by: David Øvrelid <46874830+framitdavid@users.noreply.github.com>
(cherry picked from commit 932e282e153ca30f6afaac738469137b761cf904)
Backport #28491 by @appleboy
- Modify the `Password` field in `CreateUserOption` struct to remove the
`Required` tag
- Update the `v1_json.tmpl` template to include the `email` field and
remove the `password` field
Signed-off-by: Bo-Yi Wu <appleboy.tw@gmail.com>
Co-authored-by: Bo-Yi Wu <appleboy.tw@gmail.com>
(cherry picked from commit 411310d698e86bd639b31f2f5a8b856365b4590f)
Backport #28454 (the only conflict is caused by some comments)
* Close#24483
* Close#28123
* Close#23682
* Close#23149
(cherry picked from commit a3f403f438e7f5b5dca3a5042fae8e68a896b1e7)
Conflicts:
modules/setting/ui.go
trivial context conflict
Backport #28487 by @earl-warren
- When a repository is orphaned and has objects stored in any of the
storages such as repository avatar or attachments the delete function
would error, because the storage module wasn't initalized.
- Add code to initialize the storage module.
Refs: https://codeberg.org/forgejo/forgejo/pulls/1954
Co-authored-by: Earl Warren <109468362+earl-warren@users.noreply.github.com>
Co-authored-by: Gusted <postmaster@gusted.xyz>
(cherry picked from commit 8ee1ed877b0207d9d8733ac32270325c54659909)
Backport #28421 by wxiaoguang
Refactor the code and add tests, keep the old logic.
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit 6cbb6f303a602e566f80b0e82ad73a333614efd5)
Backport #28441 by wxiaoguang
Fix#28319
It only polyfills if there is no "SubmitEvent" class, so it has no side
effect for most users.
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit 6af698fb81b9928c5d88c6bbca4b57229db1c2f3)
Backport #28392 by @nekrondev
Windows-based shells will add a CRLF when piping the token into
ssh-keygen command resulting in
verification error. This resolves#21527.
Co-authored-by: nekrondev <heiko@noordsee.de>
Co-authored-by: Heiko Besemann <heiko.besemann@qbeyond.de>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit b47482d58e86c636006f7b17b4d91786e6ed4d37)
Backport #28457 by @KN4CK3R
Recently Docker started to use the optional `POST /v2/token` endpoint
which should respond with a `404 Not Found` status code instead of the
current `405 Method Not Allowed`.
> Note: Not all token servers implement oauth2. If the request to the
endpoint returns 404 using the HTTP POST method, refer to Token
Documentation for using the HTTP GET method supported by all token
servers.
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
(cherry picked from commit 74ab7980339da50e8a42db00677b457d5f340a5a)
Backport #28428 by @KN4CK3R
There could be a nil pointer exception if the file is not found because
that specific error is suppressed but not handled.
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
(cherry picked from commit 97a0bf151a1fed2119b0ecb3c94f68462979c699)
Since a scheduled action is only run from the default branch, it
cannot be anything else.
Refs: https://codeberg.org/forgejo/forgejo/issues/1926
(cherry picked from commit eff0822856fd727915f6e6493a80844cffd7b02a)
- The transaction in combination with Git push was causing deadlocks if
you had the `push_update` queue set to `immediate`. This was the root
cause of slow integration tests in CI.
- Remove the sync branch code as this is already being done in the Git
post-receive hook.
- Add tests to proof the branch models are in sync even with this code
removed.
Backport of https://codeberg.org/forgejo/forgejo/pulls/1962
(cherry picked from commit a064065cb9a6e39597e38c37a405d066cfabf7f7)
This reverts commit 172fdd0d32.
This code was introduced for a test case that is no longer in use. It
should guard against the provider being null but that's not worth the
effort for deadcode. Just remove it.
Refs: https://codeberg.org/forgejo/forgejo/issues/1984
It will determine how anchors are created and will break existing
links otherwise.
Adapted from Revert "Make `user-content-* ` consistent with github (#26388)
It shows warnings although the setting is not set, this will surely be
fixed later but there is no sense in spaming the users right now. This
revert can be discarded when another fix lands in v1.21.
su -c "forgejo admin user generate-access-token -u root --raw --scopes 'all,sudo'" git
2023/12/12 15:54:45 .../setting/security.go:166:loadSecurityFrom() [W] Enabling Query API Auth tokens is not recommended. DISABLE_QUERY_AUTH_TOKEN will default to true in gitea 1.23 and will be removed in gitea 1.24.
This reverts commit 0e3a5abb69.
Conflicts:
routers/api/v1/api.go
To be rebuilt with latest golang version
---------
Co-authored-by: Lauris BH <lauris@nix.lv>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
(cherry picked from commit 96d3fcf179d4eea98f62d380e04f246ff65694d7)
Backport #28390 by @jackHay22
## Changes
- Add deprecation warning to `Token` and `AccessToken` authentication
methods in swagger.
- Add deprecation warning header to API response. Example:
```
HTTP/1.1 200 OK
...
Warning: token and access_token API authentication is deprecated
...
```
- Add setting `DISABLE_QUERY_AUTH_TOKEN` to reject query string auth
tokens entirely. Default is `false`
## Next steps
- `DISABLE_QUERY_AUTH_TOKEN` should be true in a subsequent release and
the methods should be removed in swagger
- `DISABLE_QUERY_AUTH_TOKEN` should be removed and the implementation of
the auth methods in question should be removed
## Open questions
- Should there be further changes to the swagger documentation?
Deprecation is not yet supported for security definitions (coming in
[OpenAPI Spec version
3.2.0](https://github.com/OAI/OpenAPI-Specification/issues/2506))
- Should the API router logger sanitize urls that use `token` or
`access_token`? (This is obviously an insufficient solution on its own)
Co-authored-by: Jack Hay <jack@allspice.io>
Co-authored-by: delvh <dev.lh@web.de>
(cherry picked from commit f144521aea0d7a08b9bd5f17e49bae4021bd7a45)
Backport #28302 by @yp05327
Close#28287
## How to test it in local
convert Makefile L34 into:
```
cd .tmp/upstream-docs && git clean -f && git reset --hard && git fetch origin pull/28302/head:pr28302 && git switch pr28302
```
Co-authored-by: yp05327 <576951401@qq.com>
(cherry picked from commit 40d51188c039f6a674a37e67d0ea5504a7a3e282)
Fix#28056
Backport #28361
This PR will check whether the repo has zero branch when pushing a
branch. If that, it means this repository hasn't been synced.
The reason caused that is after user upgrade from v1.20 -> v1.21, he
just push branches without visit the repository user interface. Because
all repositories routers will check whether a branches sync is necessary
but push has not such check.
For every repository, it has two states, synced or not synced. If there
is zero branch for a repository, then it will be assumed as non-sync
state. Otherwise, it's synced state. So if we think it's synced, we just
need to update branch/insert new branch. Otherwise do a full sync. So
that, for every push, there will be almost no extra load added. It's
high performance than yours.
For the implementation, we in fact will try to update the branch first,
if updated success with affect records > 0, then all are done. Because
that means the branch has been in the database. If no record is
affected, that means the branch does not exist in database. So there are
two possibilities. One is this is a new branch, then we just need to
insert the record. Another is the branches haven't been synced, then we
need to sync all the branches into database.
(cherry picked from commit 87db4a47c8e22b7c2e4f2b9f9efc8df1e3622884)
