Don't use rugged in Repository#refs_hash

The refs hash is used to determine what branches and tags have a commit
as head in the network graph. The previous implementation depended on
Rugged#references. The problem with this implementation was that it
depended on rugged, but also that it iterated over all references and
thus loading more data than needed if for example the project uses CI/CD
environments, Pipelines, or Merge Requests.

Given only refs are checked the network cares about the GraphHelper#refs
method has no need to reject those, simplifying the method.

Closes gitlab-org/gitaly#880

1 files changed, 6 insertions, 2 deletions

diff --git a/spec/lib/gitlab/git/repository_spec.rb b/spec/lib/gitlab/git/repository_spec.rb
index ec1c7a96f92..edcf8889c27 100644
--- a/spec/lib/gitlab/git/repository_spec.rb
+++ b/spec/lib/gitlab/git/repository_spec.rb
@@ -600,12 +600,16 @@ describe Gitlab::Git::Repository, seed_helper: true do
   end
 
   describe "#refs_hash" do
-    let(:refs) { repository.refs_hash }
+    subject { repository.refs_hash }
 
     it "should have as many entries as branches and tags" do
       expected_refs = SeedRepo::Repo::BRANCHES + SeedRepo::Repo::TAGS
       # We flatten in case a commit is pointed at by more than one branch and/or tag
-      expect(refs.values.flatten.size).to eq(expected_refs.size)
+      expect(subject.values.flatten.size).to eq(expected_refs.size)
+    end
+
+    it 'has valid commit ids as keys' do
+      expect(subject.keys).to all( match(Commit::COMMIT_SHA_PATTERN) )
     end
   end