The code worked under the assumption that anything under `refs/tags` are tag objects, and all the rest would be peelable to a commit. As it is completely valid to have tags to blobs under a non `refs/tags` ref, this would cause failures when trying to peel a tag to a commit. Fix the broken filtering by switching to `git_revwalk_push_glob`, which already handles this case.
A newly added test uses the `git_repository_new` function without the corresponding header file being included. While this works due to the compiler deducing the correct function signature, we should obviously just include the function's declaration file.
Given a wholly in-memory repository, ensure that we can create an anonymous remote and perform actions on it.
An anonymous remote is not configured and cannot therefore have configured refspecs. Remove the parameter which adds this from the constructor.
A couple of tests use the wrong remote to push to. We did not notice up to now because the local push would copy individual objects, and those already existed, so it became a no-op. Once we made local push create the packfile, it became noticeable that there was a new packfile where it didn't belong.
The base refspecs changing can be a cause of confusion as to what is the
current base refspec set and complicate saving the remote's
configuration.
Change `git_remote_add_{fetch,push}()` to update the configuration
instead of an instance.
This finally makes `git_remote_save()` a no-op, it will be removed in a
later commit.
This is another option which we should not be keeping in the remote, but is specific to each particular operation.