This reverts commit 6c75bd01, reversing
changes made to 1be7f5aa.

Add `GetArchiveRequest` to git-archive params.

Modifies `Git::Repository#archive_metadata` to append `path`
to `ArchivePrefix` so it'll not hit the cache of repository archive
when it already exists.

This changes the repository type from a binary `wiki?` to a type. So
we can have more than 2 repository types.

Now everywhere we called `.wiki?` and expected a boolean, we check
that type.

This reverts merge request !26183

Signed-off-by: Rémy Coutable <remy@rymai.me>

When an empty project is loaded in the UI, there are 15 separate Gitaly
FindDefaultBranch calls to determine the root_ref. Previously, it was
not cached even within the request. This change caches it within the
request so only a single FindDefaultBranch RPC is needed.

Closes https://gitlab.com/gitlab-org/gitlab-ce/issues/58684

This style change enforces `return if ...` instead of
`return nil if ...` to save maintainers a few minor review points

Adds the ground work for writing into
the merge ref refs/merge-requests/:iid/merge the
merge result between source and target branches of
a MR, without further side-effects such as
mailing, MR updates and target branch changes.

Adds the client call for the gitaly rpc CountDivergingCommits

fixing signature

simplifying commit logic

adding test for max-count

refactoring tests

When hashed storage is in use, it's helpful to have the project
name associated with the request.

Closes https://gitlab.com/gitlab-org/gitaly/issues/1394

This engine was replaced with CommonMarker in 11.4, it was deprecated
since then.

The data migration looks for code owner file and errs
if repository is missing.

This reverts commit e65e1842, reversing
changes made to 646ba241.

This implements Repository#ambiguous_ref? and checks if a ref is
ambiguous before trying to resolve the ref in Project#protected_for?

Reworks Project#resolve_ref to return Gitlab::Git::Branch,
Gitlab::Git::Tag or raise an AmbiguousRef error.

Determined by running the script:

```
included = `git grep --name-only ShellAdapter`.chomp.split("\n")
used = `git grep --name-only gitlab_shell`.chomp.split("\n")
included - used
```

Caches repository.path into Repository#readme_path

Use shelling out to git to write refs instead of rugged, hoping to
avoid creating invalid refs.

To update HEAD we switched to using `git symbolic-ref`.

This new endpoint allow users to update a submodule's reference.

The MR involves adding a new operation RPC operation in gitaly-proto
(see gitlab-org/gitaly-proto!233) and change Gitaly to use this
new version (see gitlab-org/gitaly!936).

See gitlab-org/gitlab-ce!20949

This indirection doesn't provide any value, so remove it

GitLab keeps a cache of the rendered HTML for a repository's README as
stored in the HEAD branch. However, it was not used in all
circumstances. In particular, the new blob viewer framework bypassed
this cache entirely.

This MR ensures a ::ReadmeBlob is returned instead of a ::Blob when
asking a repository for an individual blob, if the commit and path
match the readme for HEAD. This makes the cached HTML available to
consumers, including the blob viewer.

The ReadmeBlob is a simple delegator to the Blob, so should be
compatible in all cases. Adding the rendered_markdown method is the
only additional behaviour it contains.

This integration no longer works and does not appear to be supported.

Closes https://gitlab.com/gitlab-org/gitlab-ce/issues/39697

As we now support getting the merge base for multiple revisions in
gitaly, we can provide this functionality in our API

Implements list_last_commits_for_tree to communicate with the
ListLastCommitsForTree Gitaly RPC

Bumps the Gitaly server version

Bumps the Gitaly-Proto gem version

* Only truthy values are cached in Redis.
* All values are cached in RequestStore and in an instance variable.

And reuse `Gitlab::Utils::StrongMemoize`.

There is a subtle behavior change required to reuse StrongMemoize in
this case. The early fallback check now occurs *before* reading the
memoized value instead of after.

I think this is fine since a memoized value should only exist if
`exists?` is also already memoized as `true`.

We should not synchronize the source repository with
the target root ref while updating a push mirror.

CE mirror of 17de13ada1a98da060802c55889489a512183cd1

CE mirror of 03c6094997023d9c8875ced421a6c9ef39a4af44