`TreeCommon` only holds `ID` and is embedded solely by `Tree`, with no
other users. Inline the field into `Tree`; `t.ID` access is unchanged.
Assisted-by: Codet:codet
When pushing the merge commit to the base repo fails (e.g.
non-fast-forward), the merge intent recorded by
https://github.com/go-gitea/gitea/pull/39442 is left behind.
`merged_commit_id` then points to a commit that does not exist in the
base repo, and `git branch --contains` fails with `no such commit`, so
the PR can never be checked or merged again.
- Clear the recorded merge intent if the push fails and the commit is
not on the base branch
- `IsCommitInBranch` now returns false for a missing commit, so
already-stuck PRs recover on the next check
Fixes https://github.com/go-gitea/gitea/issues/39824
_Written with Codet assistance._
---------
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: Giteabot <teabot@gitea.io>
getPullRequestInfo and getIssuesInfo indexed into the last element of an
empty Assignees slice when Action was HookIssueAssigned, causing a panic
under a race where the assignee list is empty by the time the webhook
payload is built.
Fixes https://github.com/go-gitea/gitea/issues/39826
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: silverwind <me@silverwind.io>
Links to the new issue and new pull request pages can preselect sidebar
values with the same query parameters as the issue list filters:
`labels`, `milestone`, `project` and `assignee`, each taking
comma-separated IDs, e.g. `compare/main...branch?labels=1,2&assignee=3`.
Previously, pull requests supported none of them and issues only
`milestone` and `project`. On the compare page they also expand the new
pull request form, like `title` and `body`. Query values take precedence
over issue template labels and assignees.
* Fix https://github.com/go-gitea/gitea/issues/27341
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Bound the memory used when parsing uploaded packages, so instances with
low memory limits survive non-standard packages.
1. Metadata files are read through a size-limited reader, archives
themselves stay unbounded so large payloads keep working
2. zstd decoding caps the window at 128 MiB, the zstd CLI default
3. RPM headers go through a first-party reader with rpmlib's size and
array count limits that allocates only what the upload contains, until
https://github.com/sassoftware/go-rpmutils/pull/44 is released. Signing
still verifies the payload digest, and its signature header matches
rpmsign's byte for byte
4. Swift and NuGet read the zip entry they size-checked, a duplicate
name could stand in for it before
5. Terraform states check `LIMIT_SIZE_TERRAFORM_STATE` before parsing,
NuGet symbol PDBs share a 256 MiB budget
6. Oversized container configs are rejected instead of stored without
metadata, invalid Vagrant boxes answer 400 instead of 500
Signed-off-by: techknowlogick <git@tklk.dev>
Co-authored-by: techknowlogick <git@tklk.dev>
Co-authored-by: silverwind <me@silverwind.io>
Co-authored-by: techknowlogick <techknowlogick@gitea.com>
Folded (`>`) scalars in workflows gained blank lines and had spaces
turned into newlines every time Gitea re-serialized a job. This bumps
go-yaml to its main branch and drops our workarounds. The runner gets
the same change in https://gitea.com/gitea/runner/pulls/1321.
- go-yaml main fixes the accumulating blank lines
(https://github.com/yaml/go-yaml/pull/366) and the block scalar
indentation indicators (https://github.com/yaml/go-yaml/pull/414), so
`BlockSafeString` and the folded scalar requoting go away
- The job encoder no longer wraps lines, because go-yaml still wraps
more-indented lines of folded scalars and turns a space into a newline
- The fix applies to folded scalars anywhere in a workflow, including
workflow-level `env`
Fixes https://github.com/go-gitea/gitea/issues/39775
Co-authored-by: silverwind <me@silverwind.io>
28.0.0 bumped `github.com/go-webauthn/webauthn` from v0.17.4 to
v0.18.x. v0.18.0 made the ceremony fail whenever the client returns a
client extension output the relying party did not request, unless
`Config.ExtensionsUnsolicitedOutputPolicy` says otherwise; the strict
behavior is the zero value, so not configuring a policy opts into it
(upstream MIGRATION.md §2.4).
Every security-key and passkey login from Safari on iOS fails with 403:
Client returned the "appid" extension output which was not requested
To fix, set `ExtensionsUnsolicitedOutputPolicy:
protocol.UnsolicitedOutputPolicyIgnore` in the webauthn config
Fixes#39790
---------
Signed-off-by: choksi2212 <choksi2212@users.noreply.github.com>
Signed-off-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: choksi2212 <choksi2212@users.noreply.github.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The [Gitea VS Code extension](https://gitea.com/gitea/gitea-vscode)
currently signs in through the builtin `tea` OAuth2 application. Gitea
keeps one grant per user and application, so anyone who has already
signed in with the tea CLI gets `a grant exists with different scope`
when the extension requests its own scopes, and sign-in fails
(gitea/gitea-vscode#24).
This adds a dedicated builtin application for the extension, following
the same mechanism as the existing CLI entries:
Fixes https://github.com/go-gitea/gitea/issues/39770
`\(x^2\)is` and `\(x^2\)123` rendered as plain text because `\(...\)`
went through the same check on surrounding characters as `$...$`. That
check is there to skip things like `US$1`, which can't happen with
explicit `\(` `\)` delimiters, so it's now skipped for them, the same as
for "$`...`$". `$x$a` still isn't parsed.
`parseGitDiffTreeLine` (`services/gitdiff/git_diff_tree.go`) read
`BasePath`/`HeadPath` straight from `git diff-tree --raw` output without
undoing git's C-quoting, unlike its sibling `parseLsTreeLine`
(`modules/git/parse.go`), which already does this with
`strconv.Unquote`.
This fix applies the same unquoting to both paths in the diff-tree
parser.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
This adds a Branch Deletion setting with three modes: disabled by
default, anyone with push access, or a restricted user/team/deploy-key
allowlist. The policy applies to Git pushes, API/UI deletion, and
post-merge cleanup; the default branch and the pull request target
branch remain undeletable.
Closes https://github.com/go-gitea/gitea/issues/33305
---------
Co-authored-by: Enoque Duarte <12515551+enoquefcd@users.noreply.github.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
1. use correct links for srcset attribute of picture source
2. wrap the picture element with a link correctly
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Gitea changed the `dist.tarball` it advertises for already published
versions twice:
| Gitea | `dist.tarball` path of `@scope/name@1.0.0` |
|---|---|
| 1.27 | `%40scope%2Fname/-/1.0.0/name-1.0.0.tgz` |
| 28.0, https://github.com/go-gitea/gitea/pull/39061 |
`@scope%2Fname/-/1.0.0/name-1.0.0.tgz` |
| 28.1, https://github.com/go-gitea/gitea/pull/39434 |
`@scope/name/-/name-1.0.0.tgz` |
pnpm 11.5.2+ verifies the tarball URLs pinned in its lockfile against
the registry metadata, so frozen installs of lockfiles written against
1.27 fail with `ERR_PNPM_TARBALL_URL_MISMATCH`.
- Versions uploaded before this PR get their 1.27 URL back.
- New versions store their canonical tarball path at upload, so later
URL changes can't reach published versions. The host still follows the
request.
Lockfiles written against 28.0 for scoped packages, or by pnpm 12 with
`@scope:registry` against 28.1, need those pinned `tarball:` entries
removed, as 1.27 lockfiles of the same versions need the other URL.
Signed-off-by: silverwind <me@silverwind.io>
Fixes a few Actions bugs where user input or nondeterminism caused
failures:
- The job log view indexed task steps with the client-supplied cursor
step without a bounds check, so an out-of-range value panicked and
returned a 500.
- Workflow activity type filters (`types:`) were compiled with
`glob.MustCompile`, so an invalid pattern in a workflow file panicked
during event detection. Invalid patterns now simply don't match.
- Listing workflow runs with an unknown `actor` returned a 500; it now
returns 404.
- The v4 artifact `ListArtifacts` response was built by ranging over a
map, so its order changed between calls. It now follows database order.
- The error from marshalling a workflow job when inserting run jobs was
silently ignored.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Since https://github.com/go-gitea/gitea/pull/39703, Gitea fails to start
when its gitconfig holds a `gc.reflogExpire` other than `90` (e.g.
`30.days`), because `git config --unset-all` exits 5 when no value
matches. Value patterns were also unanchored regexps, so `90` also
removed `90.days`.
- Match values exactly in `configUnsetAll` and `configAddNonExist`
- Treat "nothing to unset" as success
- Use `--replace-all` in `configSet` so a key with multiple values no
longer fails startup
Collection of small, behaviour-preserving cleanups in the Actions code.
- Fix log and error prefixes that named non-existent or wrong functions,
and wrap errors with `%w` so `errors.Is` keeps working.
- Remove unused code.
- Rely on `ctx.ServerError` / `ctx.APIErrorAuto`, which already map
`util.ErrNotExist` and friends, instead of equivalent hand-written
switches.
- Load runs referenced by commit statuses with one query instead of one
per run (used on every PR view); the result order is now stable.
- Encode an event's payload once instead of up to three times.
- Share the duplicated runner-token "get or create" logic and the
"notify and emit" call pair.
- Simplify `ToActionArtifact` / `ToActionRunner` signatures.
---------
Signed-off-by: bircni <bircni@icloud.com>
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Gitea's default `gc.reflogExpire = 90` is parsed by git as
`1990-<month>-<day>`, currently `1990-10-09`, so reachable reflog
entries never expired.
Since git 2.54, auto maintenance runs its reflog-expire task in the
foreground of every push, and its trigger ignores reachability, so busy
repos rerun `git reflog expire --all` on every push without pruning
anything, stalling large repos for over a minute.
- Stop setting `gc.reflogExpire` so git's own defaults apply, and remove
the `90` written by earlier versions
- Treat the legacy `[git.reflog] EXPIRATION` as days, as documented
The next push to each repo prunes the accumulated entries once.
Fixes: https://github.com/go-gitea/gitea/issues/39693
Previously, recognition of escaped alert markers such as `\[!NOTE\]`
accidentally depended on enabling the optional `inline-parentheses` math
setting, resulting in such escaped alerts not rendering properly.
Goldmark splits the escaped marker differently based on the setting
being enabled or disabled, and only one form was handled. This commit
handles both forms.
For example, in a README or issue:
```
> \[!NOTE\]
> text
```
Before this commit, we'd get an ordinary blockquote displaying literal
`[!NOTE]` followed by `text`. With this commit, we get the styled
admonition.
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Go rejects scp-style addresses like `git@host:owner/repo.git` in the
go-import meta tag because they have no URL scheme, so
GO_GET_CLONE_URL_PROTOCOL=ssh did not work with the default SSH port
unless USE_COMPAT_SSH_URI was set. Always use the ssh:// form for
go-get.
When DISABLE_HTTP_GIT is enabled, the https clone URL can never work, so
GO_GET_CLONE_URL_PROTOCOL now defaults to ssh in that case.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Creating a repository with an initial commit, via `auto_init` or a
template with git content, pushed that commit through receive-pack and
three server-side hooks that do nothing for this internal push except
refresh server info. Fetching it into the bare repository instead makes
such repository creation about 3.7x faster locally (median 790ms to
214ms). E2E tests that create initialized repositories get about
0.2-0.3s faster each.
1. Run `update-server-info` after template content generation, which the
push's `post-receive` hook did before.
2. Remove the now unused `InternalPushingEnvironment` and
`PushOptions.LocalRefName`.
Introduce second list of addresses which are classified as dialable if
explicitly allowed when in Lax mode.
Restricted pool now includes: link-local, site local, private (including
ULA), CGNAT, discard, dummy, documentation and test addreses.
Reserved pool shrinks to: this network, wireserver embedding/translation
ranges and multicasts
Rationale for the choice is that while items in restricted pool can be
dangerous to allow they could be a legitimate target in some
deployments. Ranges left in reserved list are ranges which make no sense
to dial, are public (wireserver) or are 6to4 embedding which cannot be
reasonably verified to be safe. To unlock those a proxy should be used
instead
fixes: https://github.com/go-gitea/gitea/issues/39557
---------
Signed-off-by: TheFox0x7 <thefox0x7@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Rendering a CITATION.cff could use memory far out of proportion to the
file, as every YAML alias copies its target into the formatted citation
and the parser copies `%TAG` prefixes into every node. Files past these
limits show no citation, like unparseable ones do today.
- Skip files over 256 KiB, largest real-world file found is 80 KiB
- Skip files with `%TAG` directives
- Skip files whose aliases add more than 64 Ki nodes and value bytes
- Skip self-referencing anchors, except a sequence listing itself
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Raise the minimum git version to 2.34, the version in Ubuntu 22.04,
Debian 12 and RHEL 8 ship newer, and remove the fallbacks it makes
obsolete.
- Always enable AGit
- Use `diff --skip-to` and `apply -3` unconditionally
- Set the default branch of new repos and wikis via `git init
--initial-branch`
- Detect rebase conflicts via `REBASE_HEAD`
Replace citation-js with a Go port of ruby-cff, which GitHub uses for
"Cite this repository", rendering APA and BibTeX server-side and
dropping about 770KB of JS. Output matches GitHub on 1568 of 1571
real-world files, the rest are improvements over GitHub.
- `CITATION.cff` wins over `CITATION.bib`, matched case-insensitively
and through symlinks
- `CITATION.bib` is offered as-is, without APA
Signed-off-by: silverwind <me@silverwind.io>
The `gitea.com/go-chi/session` package only exists for Gitea, so it
moves into `modules/session` to fix its bugs directly. Fixes the flake
in
https://github.com/go-gitea/gitea/actions/runs/36726154500/job/109923538400.
- Sessions are only written back when changed, so a read-only request
can't revert a concurrent change or restore a logged-out session, like
https://github.com/go-macaron/session/commit/ae808a4a4660c802965c834299ab08f167effd12
- The session cookie is only set once a session holds data
- Every backend refreshes the expiry on load and file sessions are
written atomically
- Also fix https://github.com/go-gitea/gitea/issues/36176
## ⚠️ BREAKING ⚠️
* the `mysql`, `postgres`, `couchbase` and `memcache` session providers
are removed, use `file`, `db` or `redis` instead
* login-related cookies are renamed to `gitea_session` and
`gitea_remember`, if you'd like to use the old names, set `COOKIE_NAME`
and `COOKIE_REMEMBER_NAME` in app.ini
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The `gitea.com/go-chi/cache` package only exists for Gitea, so it moves
into `modules/cache`.
- `ITEM_TTL = -1` disables caching as documented
- Sub-second TTLs round up instead of never expiring
- The Redis adapter no longer grows a `MacaronCache` hash
- Use two-queue cache for in-memory cache
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The `gitea.com/go-chi/captcha` package only exists for Gitea,
so it moves into `modules/imagecaptcha`.
- Reloading an expired challenge shows a new image instead of a broken one
- Every answer is consumed on its first check
- OpenID registration stops after a failed captcha
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
The `gitea.com/go-chi/binding` package only exists for Gitea, so it
moves into `modules/web/binding` to fix its bugs directly. Split out of
https://github.com/go-gitea/gitea/pull/39504.
- GET and HEAD always bind the query
- JSON `null` slice elements and nested `TrimSpace` fields bind
correctly
- Integer fields reject out-of-range values instead of wrapping
- An empty JSON body binds nothing and an unknown binding rule is an
error
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: bircni <bircni@icloud.com>
Replace time.After with a stoppable time.Timer in BatchChecker.CheckPath.
Avoid accumulating up to 6 pending 5-second timers per file on the normal path.
Preserve the existing per-attribute timeout behavior.
---------
Signed-off-by: Calvin Tjoaquinn <calvintjoa23@gmail.com>
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Speeds up two slow git paths. Results match `main` apart
from the `log.follow` fix.
- Activity top authors skip `--numstat` and an unused `rev-list
--count`, they only need names and emails
- Subdirectory listings pass only the directory as pathspec, which
already matches all its entries
- Directory listings pass `--no-follow` to `git log`, a configured
`log.follow` disabled parent rewriting and gave wrong last commits
| Benchmark | main | PR | Change |
|---|--:|--:|--:|
| Top authors, gitea, 1 month | 632 ms | 21 ms | -97% |
| Top authors, gitea, 1 year | 3216 ms | 68 ms | -98% |
| Top authors, tea | 62 ms | 12 ms | -80% |
| Listing, git `Documentation/technical` | 189 ms | 91 ms | -52% |
| Listing, gitea `options/license` | 329 ms | 208 ms | -37% |
| Listing, gitea `templates/repo` | 294 ms | 223 ms | -24% |
Tested with unit and sqlite integration tests in default and `gogit`
builds on git 2.25 and 2.56, and by comparing listing results and cache
writes with `main` on randomized histories and the gitea, tea and git
repos. Benchmarks are medians of 8 interleaved macOS runs.
Aligns the npm registry with what npm, pnpm and yarn expect:
1. Raise the publish body cap from
https://github.com/go-gitea/gitea/pull/37890 to 256 MiB like npmjs,
larger bodies get 413
2. Pick the tarball attachment by name, `npm publish --provenance`
failed at random
3. Store and serve `libc`, so mismatched glibc/musl optional binaries
are skipped
4. Treat root `*.gyp` files as an install script, like npm does
5. Always serve a `latest` dist-tag, yarn and pnpm fail without it
6. Take top-level metadata from `latest` and drop the per-version readme
7. Serve tarballs at the npmjs path `/<name>/-/<file>`, former URLs keep
working
8. Add ETag revalidation for metadata, `npm ping` and `npm whoami`
Tested with npm 12.1, pnpm 12.4, yarn 1.22 and yarn 4.18.
---------
Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>
Lunny Xiao