[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 #29105
`ctx.Error` only displays the text but `ctx.ServerError` renders the
usual error page.
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
(cherry picked from commit a1c0b3a02e54e5fe879dabccb71fba9498b64051)
Backport #29347 by @carlosfelgueiras
Fixes#27188.
Introduces a check on the installation that tries to parse the FROM
address. If it fails, shows a new error message to the user.
Co-authored-by: Carlos Felgueiras <carlosfelgueiras@tecnico.ulisboa.pt>
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
(cherry picked from commit 35db5a373babd9af157fd63eeb20d6da53320b73)
Backport #29342 by @Zettat123
Fix#29249
~~Use the `/repos/{owner}/{repo}/archive/{archive}` API to download.~~
Apply #26430 to archive download URLs.
Co-authored-by: Zettat123 <zettat123@gmail.com>
(cherry picked from commit 829b807a91f9895e3f4b262f688a8d0d9a44caf6)
Backport #29295 by @lunny
Fix#28843
This PR will bypass the pushUpdateTag to database failure when
syncAllTags. An error log will be recorded.
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
(cherry picked from commit b78f5fc60f510a58d58535af77c5b424a8b5a660)
Fixes#28945
Backport #28948
Setting the avatar is wrong and creating a random password is equal to
leave it empty.
Co-authored-by: KN4CK3R <admin@oldschoolhack.me>
(cherry picked from commit 7ea2ffaf166780b7786291f7ff022e3f5b49e8c2)
- Backport of #2518
- The content history table contains the content history of issues and
comments. For issues they are saved with an comment id of zero.
- If you want to check if the issue has an content history, it should
take into account that SQL has `comment_id = 0`, as it otherwise could
return incorrect results when for example the issue already has an
comment that has an content history.
- Fix the code of `HasIssueContentHistory` to take this into account, it
relied on XORM to generate the SQL from the non-default values of the
struct, this wouldn't generate the `comment_id = 0` SQL as `0` is the
default value of an integer.
- Remove an unncessary log (it's not the responsibility of `models`
code to do logging).
- Adds unit test.
- Resolves#2513
(cherry picked from commit 331fa44956)
- Backport of #2507
- The CODEOWNER feature relies on the changed files to determine which
reviewers should be added according to the `CODEOWNER` file.
- The current approach was to 'diff' between the base and head branch,
which seems logical but fail in practice when the pull request is out of
date with the base branch. Therefore it should instead diff between the
head branch and the merge base of the head and base branch, so only the
actual affected files by the pull requests are used, the same approach
is used by the diff of an unmerged pull request.
- Add integration testing (for the feature as well).
- Resolves#2458
(cherry picked from commit fb2795b5bb)
- Backport of #2489
- If the user is searching repositories with an specific topic, adding
any other filter option, such as showing unrelevant repositories or
using another sort Forgejo should remember that 'topic only' was set.
- Adds integration test.
- Resolves#2461
(cherry picked from commit b4360d504c)
It's possible for reviews to not be assiocated with users, when they
were migrated from another forge instance. In the migration code,
there's no sanitization check for author names, so they could contain
HTML tags and thus needs to be properely escaped.
(cherry picked from commit ca798e4cc2)
- It's possible for reviews to not be assiocated with users, when they
were migrated from another forge instance. In the migration code,
there's no sanitization check for author names, so they could contain
HTML tags and thus needs to be properely escaped.
- Pass `$reviewerName` trough `Escape`.
On the wiki and revisions page, information is shown about the last
commit that modified that wiki page. This includes the time it was last
edited and by whom. Verify it is sanitized.
(cherry picked from commit 565e331238)
- On the wiki and revisions page, information is shown about the last
commit that modified that wiki page. This includes the time it was last
edited and by whom. That whole string is not being sanitized (passed
trough `Safe` in the templates), because the last edited bit is
formatted as an HTML element and thus shouldn't be sanitized. The
problem with this is that now `.Author.Name` is not being sanitized.
- This can be exploited, the names of authors and commiters on a Git
commit is user controlled, they can be any value and thus also include
HTML. It's not easy to actually exploit this, as you cannot use the
official git binary to do use, as they actually strip `<` and `>` from
user names (trivia: this behaviour was introduced in the initial commit
of Git). In the integration testing, go-git actually has to generate
this commit as they don't have such restrictions.
- Pass `.Author.Name` trough `Escape` in order to be sanitized.
- Backport of #2416
- The hook regeneration code relies on `git.SupportProcReceive` being
set to determine if the `proc-receive` hook should be written, this
variable is set when the git module is initialized.
- Resolves#2414
(cherry picked from commit 815abad84c)
Backport #28551
RequestReview get deleted on review.
So we don't have to try to load them on comments.
(cherry picked from commit 0ac3186267b717bce7076ef44f883df7720d7a2d)
