Merge branch 'blob-tests' into 'master'

Add unit tests for Blob from gitlab-ce + setup code

See merge request gitlab-org/gitaly!920

10 files changed, 643 insertions, 33 deletions

diff --git a/Makefile b/Makefile
index e1346cc77..ffb51d0f8 100644
--- a/Makefile
+++ b/Makefile
@@ -7,6 +7,7 @@ BIN_BUILD_DIR := $(TARGET_DIR)/bin
 PKG_BUILD_DIR := $(TARGET_DIR)/src/$(PKG)
 export TEST_REPO_STORAGE_PATH := $(BUILD_DIR)/internal/testhelper/testdata/data
 TEST_REPO := $(TEST_REPO_STORAGE_PATH)/gitlab-test.git
+GIT_TEST_REPO := $(TEST_REPO_STORAGE_PATH)/gitlab-git-test.git
 INSTALL_DEST_DIR := $(DESTDIR)$(PREFIX)/bin/
 COVERAGE_DIR := $(TARGET_DIR)/cover
 ASSEMBLY_ROOT := $(TARGET_DIR)/assembly
@@ -119,8 +120,11 @@ $(TEST_REPO):
 	# Git notes aren't fetched by default with git clone
 	git -C $@ fetch origin refs/notes/*:refs/notes/*
 
+$(GIT_TEST_REPO):
+	git clone --bare https://gitlab.com/gitlab-org/gitlab-git-test.git $@
+
 .PHONY: prepare-tests
-prepare-tests: $(TARGET_SETUP) $(TEST_REPO) .ruby-bundle
+prepare-tests: $(TARGET_SETUP) $(TEST_REPO) $(GIT_TEST_REPO) .ruby-bundle
 
 .PHONY: test
 test: test-go rspec
@@ -181,7 +185,7 @@ codeclimate-report:
 
 .PHONY: clean
 clean:
-	rm -rf $(TARGET_DIR) $(TEST_REPO) $(TEST_REPO_STORAGE_PATH) ./internal/service/ssh/gitaly-*-pack .ruby-bundle
+	rm -rf $(TARGET_DIR) $(TEST_REPO_STORAGE_PATH) ./ruby/tmp/repositories ./internal/service/ssh/gitaly-*-pack .ruby-bundle
 
 .PHONY: cover
 cover: prepare-tests $(GOCOVMERGE)
diff --git a/ruby/.rubocop_todo.yml b/ruby/.rubocop_todo.yml
index 856c4a165..b2e050bec 100644
--- a/ruby/.rubocop_todo.yml
+++ b/ruby/.rubocop_todo.yml
@@ -68,7 +68,7 @@ Metrics/AbcSize:
 # Offense count: 26
 # Configuration parameters: CountComments, ExcludedMethods.
 Metrics/BlockLength:
-  Max: 203
+  Max: 250
 
 # Offense count: 11
 # Configuration parameters: CountComments.
@@ -292,4 +292,4 @@ Style/TrivialAccessors:
 # Configuration parameters: AllowHeredoc, AllowURI, URISchemes, IgnoreCopDirectives, IgnoredPatterns.
 # URISchemes: http, https
 Metrics/LineLength:
-  Max: 156
+  Max: 200
diff --git a/ruby/lib/gitlab/git/blob.rb b/ruby/lib/gitlab/git/blob.rb
index 9152d79ff..70a835927 100644
--- a/ruby/lib/gitlab/git/blob.rb
+++ b/ruby/lib/gitlab/git/blob.rb
@@ -6,8 +6,7 @@ module Gitlab
 
       # This number is the maximum amount of data that we want to display to
       # the user. We load as much as we can for encoding detection
-      # (Linguist) and LFS pointer parsing. All other cases where we need full
-      # blob data should use load_all_data!.
+      # (Linguist) and LFS pointer parsing.
       MAX_DATA_DISPLAY_SIZE = 10.megabytes
 
       # These limits are used as a heuristic to ignore files which can't be LFS
@@ -160,22 +159,6 @@ module Gitlab
         encode! @data
       end
 
-      # Load all blob data (not just the first MAX_DATA_DISPLAY_SIZE bytes) into
-      # memory as a Ruby string.
-      def load_all_data!(repository)
-        return if @data == '' # don't mess with submodule blobs
-
-        # Even if we return early, recalculate wether this blob is binary in
-        # case a blob was initialized as text but the full data isn't
-        @binary = nil
-
-        return if @loaded_all_data
-
-        @data = repository.gitaly_blob_client.get_blob(oid: id, limit: -1).data
-        @loaded_all_data = true
-        @loaded_size = @data.bytesize
-      end
-
       def name
         encode! @name
       end
diff --git a/ruby/lib/gitlab/git/operation_service.rb b/ruby/lib/gitlab/git/operation_service.rb
index 52a6b617d..c7a24958f 100644
--- a/ruby/lib/gitlab/git/operation_service.rb
+++ b/ruby/lib/gitlab/git/operation_service.rb
@@ -88,9 +88,7 @@ module Gitlab
 
         start_branch_name = nil if start_repository.empty?
 
-        if start_branch_name && !start_repository.branch_exists?(start_branch_name)
-          raise ArgumentError, "Cannot find branch #{start_branch_name} in #{start_repository.relative_path}"
-        end
+        raise ArgumentError, "Cannot find branch #{start_branch_name} in #{start_repository.relative_path}" if start_branch_name && !start_repository.branch_exists?(start_branch_name)
 
         update_branch_with_hooks(branch_name) do
           repository.with_repo_branch_commit(
diff --git a/ruby/lib/gitlab/git/repository.rb b/ruby/lib/gitlab/git/repository.rb
index 383c2a1ac..f3e6505c9 100644
--- a/ruby/lib/gitlab/git/repository.rb
+++ b/ruby/lib/gitlab/git/repository.rb
@@ -296,9 +296,7 @@ module Gitlab
               cat_stdout.close
               cat_stderr.close
 
-              unless [cat_wait_thr, wait_thr].all? { |waiter| waiter.value&.success? }
-                raise ::Gitlab::Git::Repository::GitError, "Unabled to obtain changes between #{old_rev} and #{new_rev}"
-              end
+              raise ::Gitlab::Git::Repository::GitError, "Unabled to obtain changes between #{old_rev} and #{new_rev}" unless [cat_wait_thr, wait_thr].all? { |waiter| waiter.value&.success? }
             end
 
             result
diff --git a/ruby/spec/lib/gitlab/git/blob_spec.rb b/ruby/spec/lib/gitlab/git/blob_spec.rb
new file mode 100644
index 000000000..cfae0ec1d
--- /dev/null
+++ b/ruby/spec/lib/gitlab/git/blob_spec.rb
@@ -0,0 +1,296 @@
+require "spec_helper"
+
+describe Gitlab::Git::Blob, :seed_helper do
+  include TestRepo
+
+  let(:repository) { gitlab_git_from_gitaly(git_test_repo_read_only) }
+  let(:rugged) do
+    Rugged::Repository.new(GIT_TEST_REPO_PATH)
+  end
+
+  describe 'initialize' do
+    let(:blob) { Gitlab::Git::Blob.new(name: 'test') }
+
+    it 'handles nil data' do
+      expect(blob.name).to eq('test')
+      expect(blob.size).to eq(nil)
+      expect(blob.loaded_size).to eq(nil)
+    end
+  end
+
+  describe '.find' do
+    context 'nil path' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, nil) }
+
+      it { expect(blob).to eq(nil) }
+    end
+
+    context 'utf-8 branch' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, 'Ääh-test-utf-8', "files/ruby/popen.rb") }
+
+      it { expect(blob.id).to eq(SeedRepo::RubyBlob::ID) }
+    end
+
+    context 'blank path' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, '') }
+
+      it { expect(blob).to eq(nil) }
+    end
+
+    context 'file in subdir' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, "files/ruby/popen.rb") }
+
+      it { expect(blob.id).to eq(SeedRepo::RubyBlob::ID) }
+      it { expect(blob.name).to eq(SeedRepo::RubyBlob::NAME) }
+      it { expect(blob.path).to eq("files/ruby/popen.rb") }
+      it { expect(blob.commit_id).to eq(SeedRepo::Commit::ID) }
+      it { expect(blob.data[0..10]).to eq(SeedRepo::RubyBlob::CONTENT[0..10]) }
+      it { expect(blob.size).to eq(669) }
+      it { expect(blob.mode).to eq("100644") }
+    end
+
+    context 'file in root' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, ".gitignore") }
+
+      it { expect(blob.id).to eq("dfaa3f97ca337e20154a98ac9d0be76ddd1fcc82") }
+      it { expect(blob.name).to eq(".gitignore") }
+      it { expect(blob.path).to eq(".gitignore") }
+      it { expect(blob.commit_id).to eq(SeedRepo::Commit::ID) }
+      it { expect(blob.data[0..10]).to eq("*.rbc\n*.sas") }
+      it { expect(blob.size).to eq(241) }
+      it { expect(blob.mode).to eq("100644") }
+      it { expect(blob).not_to be_binary }
+    end
+
+    context 'file in root with leading slash' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, "/.gitignore") }
+
+      it { expect(blob.id).to eq("dfaa3f97ca337e20154a98ac9d0be76ddd1fcc82") }
+      it { expect(blob.name).to eq(".gitignore") }
+      it { expect(blob.path).to eq(".gitignore") }
+      it { expect(blob.commit_id).to eq(SeedRepo::Commit::ID) }
+      it { expect(blob.data[0..10]).to eq("*.rbc\n*.sas") }
+      it { expect(blob.size).to eq(241) }
+      it { expect(blob.mode).to eq("100644") }
+    end
+
+    context 'non-exist file' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, "missing.rb") }
+
+      it { expect(blob).to be_nil }
+    end
+
+    context 'six submodule' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, 'six') }
+
+      it { expect(blob.id).to eq('409f37c4f05865e4fb208c771485f211a22c4c2d') }
+      it { expect(blob.data).to eq('') }
+
+      it 'does not mark the blob as binary' do
+        expect(blob).not_to be_binary
+      end
+    end
+
+    context 'large file' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, 'files/images/6049019_460s.jpg') }
+      let(:blob_size) { 111_803 }
+      let(:stub_limit) { 1000 }
+
+      before do
+        stub_const('Gitlab::Git::Blob::MAX_DATA_DISPLAY_SIZE', stub_limit)
+      end
+
+      it { expect(blob.size).to eq(blob_size) }
+      it { expect(blob.data.length).to eq(stub_limit) }
+
+      it 'check that this test is sane' do
+        # It only makes sense to test limiting if the blob is larger than the limit.
+        expect(blob.size).to be > Gitlab::Git::Blob::MAX_DATA_DISPLAY_SIZE
+      end
+
+      it 'marks the blob as binary' do
+        expect(Gitlab::Git::Blob).to receive(:new)
+          .with(hash_including(binary: true))
+          .and_call_original
+
+        expect(blob).to be_binary
+      end
+    end
+  end
+
+  describe 'encoding' do
+    context 'file with russian text' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, "encoding/russian.rb") }
+
+      it { expect(blob.name).to eq("russian.rb") }
+      it { expect(blob.data.lines.first).to eq("Хороший файл") }
+      it { expect(blob.size).to eq(23) }
+      it { expect(blob.truncated?).to be_falsey }
+      # Run it twice since data is encoded after the first run
+      it { expect(blob.truncated?).to be_falsey }
+      it { expect(blob.mode).to eq("100755") }
+    end
+
+    context 'file with Chinese text' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::Commit::ID, "encoding/テスト.txt") }
+
+      it { expect(blob.name).to eq("テスト.txt") }
+      it { expect(blob.data).to include("これはテスト") }
+      it { expect(blob.size).to eq(340) }
+      it { expect(blob.mode).to eq("100755") }
+      it { expect(blob.truncated?).to be_falsey }
+    end
+
+    context 'file with ISO-8859 text' do
+      let(:blob) { Gitlab::Git::Blob.find(repository, SeedRepo::LastCommit::ID, "encoding/iso8859.txt") }
+
+      it { expect(blob.name).to eq("iso8859.txt") }
+      it { expect(blob.loaded_size).to eq(4) }
+      it { expect(blob.size).to eq(4) }
+      it { expect(blob.mode).to eq("100644") }
+      it { expect(blob.truncated?).to be_falsey }
+    end
+  end
+
+  describe 'mode' do
+    context 'file regular' do
+      let(:blob) do
+        Gitlab::Git::Blob.find(
+          repository,
+          'fa1b1e6c004a68b7d8763b86455da9e6b23e36d6',
+          'files/ruby/regex.rb'
+        )
+      end
+
+      it { expect(blob.name).to eq('regex.rb') }
+      it { expect(blob.path).to eq('files/ruby/regex.rb') }
+      it { expect(blob.size).to eq(1200) }
+      it { expect(blob.mode).to eq("100644") }
+    end
+
+    context 'file binary' do
+      let(:blob) do
+        Gitlab::Git::Blob.find(
+          repository,
+          'fa1b1e6c004a68b7d8763b86455da9e6b23e36d6',
+          'files/executables/ls'
+        )
+      end
+
+      it { expect(blob.name).to eq('ls') }
+      it { expect(blob.path).to eq('files/executables/ls') }
+      it { expect(blob.size).to eq(110_080) }
+      it { expect(blob.mode).to eq("100755") }
+    end
+
+    context 'file symlink to regular' do
+      let(:blob) do
+        Gitlab::Git::Blob.find(
+          repository,
+          'fa1b1e6c004a68b7d8763b86455da9e6b23e36d6',
+          'files/links/ruby-style-guide.md'
+        )
+      end
+
+      it { expect(blob.name).to eq('ruby-style-guide.md') }
+      it { expect(blob.path).to eq('files/links/ruby-style-guide.md') }
+      it { expect(blob.size).to eq(31) }
+      it { expect(blob.mode).to eq("120000") }
+    end
+
+    context 'file symlink to binary' do
+      let(:blob) do
+        Gitlab::Git::Blob.find(
+          repository,
+          'fa1b1e6c004a68b7d8763b86455da9e6b23e36d6',
+          'files/links/touch'
+        )
+      end
+
+      it { expect(blob.name).to eq('touch') }
+      it { expect(blob.path).to eq('files/links/touch') }
+      it { expect(blob.size).to eq(20) }
+      it { expect(blob.mode).to eq("120000") }
+    end
+  end
+
+  describe 'lfs_pointers' do
+    context 'file a valid lfs pointer' do
+      let(:blob) do
+        Gitlab::Git::Blob.find(
+          repository,
+          '33bcff41c232a11727ac6d660bd4b0c2ba86d63d',
+          'files/lfs/image.jpg'
+        )
+      end
+
+      it { expect(blob.lfs_pointer?).to eq(true) }
+      it { expect(blob.lfs_oid).to eq("4206f951d2691c78aac4c0ce9f2b23580b2c92cdcc4336e1028742c0274938e0") }
+      it { expect(blob.lfs_size).to eq(19548) }
+      it { expect(blob.id).to eq("f4d76af13003d1106be7ac8c5a2a3d37ddf32c2a") }
+      it { expect(blob.name).to eq("image.jpg") }
+      it { expect(blob.path).to eq("files/lfs/image.jpg") }
+      it { expect(blob.size).to eq(130) }
+      it { expect(blob.mode).to eq("100644") }
+    end
+
+    describe 'file an invalid lfs pointer' do
+      context 'with correct version header but incorrect size and oid' do
+        let(:blob) do
+          Gitlab::Git::Blob.find(
+            repository,
+            '33bcff41c232a11727ac6d660bd4b0c2ba86d63d',
+            'files/lfs/archive-invalid.tar'
+          )
+        end
+
+        it { expect(blob.lfs_pointer?).to eq(false) }
+        it { expect(blob.lfs_oid).to eq(nil) }
+        it { expect(blob.lfs_size).to eq(nil) }
+        it { expect(blob.id).to eq("f8a898db217a5a85ed8b3d25b34c1df1d1094c46") }
+        it { expect(blob.name).to eq("archive-invalid.tar") }
+        it { expect(blob.path).to eq("files/lfs/archive-invalid.tar") }
+        it { expect(blob.size).to eq(43) }
+        it { expect(blob.mode).to eq("100644") }
+      end
+
+      context 'with correct version header and size but incorrect size and oid' do
+        let(:blob) do
+          Gitlab::Git::Blob.find(
+            repository,
+            '33bcff41c232a11727ac6d660bd4b0c2ba86d63d',
+            'files/lfs/picture-invalid.png'
+          )
+        end
+
+        it { expect(blob.lfs_pointer?).to eq(false) }
+        it { expect(blob.lfs_oid).to eq(nil) }
+        it { expect(blob.lfs_size).to eq(1_575_078) }
+        it { expect(blob.id).to eq("5ae35296e1f95c1ef9feda1241477ed29a448572") }
+        it { expect(blob.name).to eq("picture-invalid.png") }
+        it { expect(blob.path).to eq("files/lfs/picture-invalid.png") }
+        it { expect(blob.size).to eq(57) }
+        it { expect(blob.mode).to eq("100644") }
+      end
+
+      context 'with correct version header and size but invalid size and oid' do
+        let(:blob) do
+          Gitlab::Git::Blob.find(
+            repository,
+            '33bcff41c232a11727ac6d660bd4b0c2ba86d63d',
+            'files/lfs/file-invalid.zip'
+          )
+        end
+
+        it { expect(blob.lfs_pointer?).to eq(false) }
+        it { expect(blob.lfs_oid).to eq(nil) }
+        it { expect(blob.lfs_size).to eq(nil) }
+        it { expect(blob.id).to eq("d831981bd876732b85a1bcc6cc01210c9f36248f") }
+        it { expect(blob.name).to eq("file-invalid.zip") }
+        it { expect(blob.path).to eq("files/lfs/file-invalid.zip") }
+        it { expect(blob.size).to eq(60) }
+        it { expect(blob.mode).to eq("100644") }
+      end
+    end
+  end
+end
diff --git a/ruby/spec/spec_helper.rb b/ruby/spec/spec_helper.rb
index 75652b01a..54aa9038d 100644
--- a/ruby/spec/spec_helper.rb
+++ b/ruby/spec/spec_helper.rb
@@ -3,4 +3,6 @@ require_relative '../lib/gitlab/git.rb'
 require_relative 'support/sentry.rb'
 require 'test_repo_helper'
 
+Dir[File.join(__dir__, 'support/helpers/*.rb')].each { |f| require f }
+
 ENV['GITALY_RUBY_GIT_BIN_PATH'] ||= 'git'
diff --git a/ruby/spec/support/generate-seed-repo-rb b/ruby/spec/support/generate-seed-repo-rb
new file mode 100755
index 000000000..b82297f53
--- /dev/null
+++ b/ruby/spec/support/generate-seed-repo-rb
@@ -0,0 +1,162 @@
+#!/usr/bin/env ruby
+#
+# # generate-seed-repo-rb
+#
+# This script generates the seed_repo.rb file used by lib/gitlab/git
+# tests. The seed_repo.rb file needs to be updated anytime there is a
+# Git push to https://gitlab.com/gitlab-org/gitlab-git-test.
+#
+# Usage:
+#
+#   ./spec/support/generate-seed-repo-rb > spec/support/helpers/seed_repo.rb
+#
+#
+
+require 'erb'
+require 'tempfile'
+
+SOURCE = File.expand_path('../../../internal/testhelper/testdata/data/gitlab-git-test.git', __dir__).freeze
+SCRIPT_NAME = 'generate-seed-repo-rb'.freeze
+REPO_NAME = 'gitlab-git-test.git'.freeze
+
+def main
+  Dir.mktmpdir do |dir|
+    abort "git clone failed" unless system("git", "clone", "--bare", SOURCE.to_s, REPO_NAME.to_s, chdir: dir)
+
+    repo = File.join(dir, REPO_NAME)
+    erb = ERB.new(DATA.read)
+    erb.run(binding)
+  end
+end
+
+def capture!(cmd, dir)
+  output = IO.popen(cmd, 'r', chdir: dir, &:read)
+  raise "command failed with #{$?}: #{cmd.join(' ')}" unless $?.success?
+
+  output.chomp
+end
+
+main
+
+__END__
+# This file is generated by <%= SCRIPT_NAME %>. Do not edit this file manually.
+#
+# Seed repo:
+<%= capture!(%w{git log --format=#\ %H\ %s}, repo) %>
+
+module SeedRepo
+  module BigCommit
+    ID               = "913c66a37b4a45b9769037c55c2d238bd0942d2e".freeze
+    PARENT_ID        = "cfe32cf61b73a0d5e9f13e774abde7ff789b1660".freeze
+    MESSAGE          = "Files, encoding and much more".freeze
+    AUTHOR_FULL_NAME = "Dmitriy Zaporozhets".freeze
+    FILES_COUNT      = 2
+  end
+
+  module Commit
+    ID               = "570e7b2abdd848b95f2f578043fc23bd6f6fd24d".freeze
+    PARENT_ID        = "6f6d7e7ed97bb5f0054f2b1df789b39ca89b6ff9".freeze
+    MESSAGE          = "Change some files\n\nSigned-off-by: Dmitriy Zaporozhets <dmitriy.zaporozhets@gmail.com>\n".freeze
+    AUTHOR_FULL_NAME = "Dmitriy Zaporozhets".freeze
+    FILES            = ["files/ruby/popen.rb", "files/ruby/regex.rb"].freeze
+    FILES_COUNT      = 2
+    C_FILE_PATH      = "files/ruby".freeze
+    C_FILES          = ["popen.rb", "regex.rb", "version_info.rb"].freeze
+    BLOB_FILE        = %{%h3= @key.title\n%hr\n%pre= @key.key\n.actions\n  = link_to 'Remove', @key, :confirm => 'Are you sure?', :method => :delete, :class => \"btn danger delete-key\"\n\n\n}.freeze
+    BLOB_FILE_PATH   = "app/views/keys/show.html.haml".freeze
+  end
+
+  module EmptyCommit
+    ID               = "b0e52af38d7ea43cf41d8a6f2471351ac036d6c9".freeze
+    PARENT_ID        = "40f4a7a617393735a95a0bb67b08385bc1e7c66d".freeze
+    MESSAGE          = "Empty commit".freeze
+    AUTHOR_FULL_NAME = "Rémy Coutable".freeze
+    FILES            = [].freeze
+    FILES_COUNT      = FILES.count
+  end
+
+  module EncodingCommit
+    ID               = "40f4a7a617393735a95a0bb67b08385bc1e7c66d".freeze
+    PARENT_ID        = "66028349a123e695b589e09a36634d976edcc5e8".freeze
+    MESSAGE          = "Add ISO-8859-encoded file".freeze
+    AUTHOR_FULL_NAME = "Stan Hu".freeze
+    FILES            = ["encoding/iso8859.txt"].freeze
+    FILES_COUNT      = FILES.count
+  end
+
+  module FirstCommit
+    ID               = "1a0b36b3cdad1d2ee32457c102a8c0b7056fa863".freeze
+    PARENT_ID        = nil
+    MESSAGE          = "Initial commit".freeze
+    AUTHOR_FULL_NAME = "Dmitriy Zaporozhets".freeze
+    FILES            = ["LICENSE", ".gitignore", "README.md"].freeze
+    FILES_COUNT      = 3
+  end
+
+  module LastCommit
+    ID               = <%= capture!(%w[git show -s --format=%H HEAD], repo).inspect %>.freeze
+    PARENT_ID        = <%= capture!(%w[git show -s --format=%P HEAD], repo).split.last.inspect %>.freeze
+    MESSAGE          = <%= capture!(%w[git show -s --format=%s HEAD], repo).inspect %>.freeze
+    AUTHOR_FULL_NAME = <%= capture!(%w[git show -s --format=%an HEAD], repo).inspect %>.freeze
+    FILES            = <%=
+      parents = capture!(%w[git show -s --format=%P HEAD], repo).split
+      merge_base = parents.size > 1 ? capture!(%w[git merge-base] + parents, repo) : parents.first
+      capture!( %W[git diff --name-only #{merge_base}..HEAD --], repo).split("\n").inspect
+    %>.freeze
+    FILES_COUNT      = FILES.count
+  end
+
+  module Repo
+    HEAD = "master".freeze
+    BRANCHES = %w[
+<%=   capture!(%W[git for-each-ref --format=#{'  ' * 3}%(refname:strip=2) refs/heads/], repo) %>
+    ].freeze
+    TAGS = %w[
+<%=   capture!(%W[git for-each-ref --format=#{'  ' * 3}%(refname:strip=2) refs/tags/], repo) %>
+    ].freeze
+  end
+
+  module RubyBlob
+    ID = "7e3e39ebb9b2bf433b4ad17313770fbe4051649c".freeze
+    NAME = "popen.rb".freeze
+    CONTENT = <<-BLOB_CONTENT.freeze
+require 'fileutils'
+require 'open3'
+
+module Popen
+  extend self
+
+  def popen(cmd, path=nil)
+    unless cmd.is_a?(Array)
+      raise RuntimeError, "System commands must be given as an array of strings"
+    end
+
+    path ||= Dir.pwd
+
+    vars = {
+      "PWD" => path
+    }
+
+    options = {
+      chdir: path
+    }
+
+    unless File.directory?(path)
+      FileUtils.mkdir_p(path)
+    end
+
+    @cmd_output = ""
+    @cmd_status = 0
+
+    Open3.popen3(vars, *cmd, options) do |stdin, stdout, stderr, wait_thr|
+      @cmd_output << stdout.read
+      @cmd_output << stderr.read
+      @cmd_status = wait_thr.value.exitstatus
+    end
+
+    return @cmd_output, @cmd_status
+  end
+end
+    BLOB_CONTENT
+  end
+end
diff --git a/ruby/spec/support/helpers/seed_repo.rb b/ruby/spec/support/helpers/seed_repo.rb
new file mode 100644
index 000000000..19a9b2de6
--- /dev/null
+++ b/ruby/spec/support/helpers/seed_repo.rb
@@ -0,0 +1,154 @@
+# This file is generated by generate-seed-repo-rb. Do not edit this file manually.
+#
+# Seed repo:
+# 4b4918a572fa86f9771e5ba40fbd48e1eb03e2c6 Merge branch 'master' into 'master'
+# 0e1b353b348f8477bdbec1ef47087171c5032cd9 adds an executable with different permissions
+# 0e50ec4d3c7ce42ab74dda1d422cb2cbffe1e326 Merge branch 'lfs_pointers' into 'master'
+# 33bcff41c232a11727ac6d660bd4b0c2ba86d63d Add valid and invalid lfs pointers
+# 732401c65e924df81435deb12891ef570167d2e2 Update year in license file
+# b0e52af38d7ea43cf41d8a6f2471351ac036d6c9 Empty commit
+# 40f4a7a617393735a95a0bb67b08385bc1e7c66d Add ISO-8859-encoded file
+# 66028349a123e695b589e09a36634d976edcc5e8 Merge branch 'add-comments-to-gitmodules' into 'master'
+# de5714f34c4e34f1d50b9a61a2e6c9132fe2b5fd Add comments to the end of .gitmodules to test parsing
+# fa1b1e6c004a68b7d8763b86455da9e6b23e36d6 Merge branch 'add-files' into 'master'
+# eb49186cfa5c4338011f5f590fac11bd66c5c631 Add submodules nested deeper than the root
+# 18d9c205d0d22fdf62bc2f899443b83aafbf941f Add executables and links files
+# 5937ac0a7beb003549fc5fd26fc247adbce4a52e Add submodule from gitlab.com
+# 570e7b2abdd848b95f2f578043fc23bd6f6fd24d Change some files
+# 6f6d7e7ed97bb5f0054f2b1df789b39ca89b6ff9 More submodules
+# d14d6c0abdd253381df51a723d58691b2ee1ab08 Remove ds_store files
+# c1acaa58bbcbc3eafe538cb8274ba387047b69f8 Ignore DS files
+# ae73cb07c9eeaf35924a10f713b364d32b2dd34f Binary file added
+# 874797c3a73b60d2187ed6e2fcabd289ff75171e Ruby files modified
+# 2f63565e7aac07bcdadb654e253078b727143ec4 Modified image
+# 33f3729a45c02fc67d00adb1b8bca394b0e761d9 Image added
+# 913c66a37b4a45b9769037c55c2d238bd0942d2e Files, encoding and much more
+# cfe32cf61b73a0d5e9f13e774abde7ff789b1660 Add submodule
+# 6d394385cf567f80a8fd85055db1ab4c5295806f Added contributing guide
+# 1a0b36b3cdad1d2ee32457c102a8c0b7056fa863 Initial commit
+
+module SeedRepo
+  module BigCommit
+    ID               = "913c66a37b4a45b9769037c55c2d238bd0942d2e".freeze
+    PARENT_ID        = "cfe32cf61b73a0d5e9f13e774abde7ff789b1660".freeze
+    MESSAGE          = "Files, encoding and much more".freeze
+    AUTHOR_FULL_NAME = "Dmitriy Zaporozhets".freeze
+    FILES_COUNT      = 2
+  end
+
+  module Commit
+    ID               = "570e7b2abdd848b95f2f578043fc23bd6f6fd24d".freeze
+    PARENT_ID        = "6f6d7e7ed97bb5f0054f2b1df789b39ca89b6ff9".freeze
+    MESSAGE          = "Change some files\n\nSigned-off-by: Dmitriy Zaporozhets <dmitriy.zaporozhets@gmail.com>\n".freeze
+    AUTHOR_FULL_NAME = "Dmitriy Zaporozhets".freeze
+    FILES            = ["files/ruby/popen.rb", "files/ruby/regex.rb"].freeze
+    FILES_COUNT      = 2
+    C_FILE_PATH      = "files/ruby".freeze
+    C_FILES          = ["popen.rb", "regex.rb", "version_info.rb"].freeze
+    BLOB_FILE        = %(%h3= @key.title\n%hr\n%pre= @key.key\n.actions\n  = link_to 'Remove', @key, :confirm => 'Are you sure?', :method => :delete, :class => \"btn danger delete-key\"\n\n\n).freeze
+    BLOB_FILE_PATH   = "app/views/keys/show.html.haml".freeze
+  end
+
+  module EmptyCommit
+    ID               = "b0e52af38d7ea43cf41d8a6f2471351ac036d6c9".freeze
+    PARENT_ID        = "40f4a7a617393735a95a0bb67b08385bc1e7c66d".freeze
+    MESSAGE          = "Empty commit".freeze
+    AUTHOR_FULL_NAME = "Rémy Coutable".freeze
+    FILES            = [].freeze
+    FILES_COUNT      = FILES.count
+  end
+
+  module EncodingCommit
+    ID               = "40f4a7a617393735a95a0bb67b08385bc1e7c66d".freeze
+    PARENT_ID        = "66028349a123e695b589e09a36634d976edcc5e8".freeze
+    MESSAGE          = "Add ISO-8859-encoded file".freeze
+    AUTHOR_FULL_NAME = "Stan Hu".freeze
+    FILES            = ["encoding/iso8859.txt"].freeze
+    FILES_COUNT      = FILES.count
+  end
+
+  module FirstCommit
+    ID               = "1a0b36b3cdad1d2ee32457c102a8c0b7056fa863".freeze
+    PARENT_ID        = nil
+    MESSAGE          = "Initial commit".freeze
+    AUTHOR_FULL_NAME = "Dmitriy Zaporozhets".freeze
+    FILES            = ["LICENSE", ".gitignore", "README.md"].freeze
+    FILES_COUNT      = 3
+  end
+
+  module LastCommit
+    ID               = "4b4918a572fa86f9771e5ba40fbd48e1eb03e2c6".freeze
+    PARENT_ID        = "0e1b353b348f8477bdbec1ef47087171c5032cd9".freeze
+    MESSAGE          = "Merge branch 'master' into 'master'".freeze
+    AUTHOR_FULL_NAME = "Stan Hu".freeze
+    FILES            = ["bin/executable"].freeze
+    FILES_COUNT      = FILES.count
+  end
+
+  module Repo
+    HEAD = "master".freeze
+    BRANCHES = %w[
+      feature
+      fix
+      fix-blob-path
+      fix-existing-submodule-dir
+      fix-mode
+      gitattributes
+      gitattributes-updated
+      master
+      merge-test
+      rd-add-file-larger-than-1-mb
+      Ääh-test-utf-8
+    ].freeze
+    TAGS = %w[
+      v1.0.0
+      v1.1.0
+      v1.2.0
+      v1.2.1
+    ].freeze
+  end
+
+  module RubyBlob
+    ID = "7e3e39ebb9b2bf433b4ad17313770fbe4051649c".freeze
+    NAME = "popen.rb".freeze
+    CONTENT = <<~BLOB_CONTENT.freeze
+      require 'fileutils'
+      require 'open3'
+
+      module Popen
+        extend self
+
+        def popen(cmd, path=nil)
+          unless cmd.is_a?(Array)
+            raise RuntimeError, "System commands must be given as an array of strings"
+          end
+
+          path ||= Dir.pwd
+
+          vars = {
+            "PWD" => path
+          }
+
+          options = {
+            chdir: path
+          }
+
+          unless File.directory?(path)
+            FileUtils.mkdir_p(path)
+          end
+
+          @cmd_output = ""
+          @cmd_status = 0
+
+          Open3.popen3(vars, *cmd, options) do |stdin, stdout, stderr, wait_thr|
+            @cmd_output << stdout.read
+            @cmd_output << stderr.read
+            @cmd_status = wait_thr.value.exitstatus
+          end
+
+          return @cmd_output, @cmd_status
+        end
+      end
+    BLOB_CONTENT
+  end
+end
diff --git a/ruby/spec/test_repo_helper.rb b/ruby/spec/test_repo_helper.rb
index a9ecb7dc5..9d8f39a9d 100644
--- a/ruby/spec/test_repo_helper.rb
+++ b/ruby/spec/test_repo_helper.rb
@@ -7,14 +7,27 @@ DEFAULT_STORAGE_DIR = File.expand_path('../tmp/repositories', __dir__)
 DEFAULT_STORAGE_NAME = 'default'.freeze
 TEST_REPO_PATH = File.join(DEFAULT_STORAGE_DIR, 'gitlab-test.git')
 TEST_REPO_ORIGIN = '../internal/testhelper/testdata/data/gitlab-test.git'.freeze
+GIT_TEST_REPO_PATH = File.join(DEFAULT_STORAGE_DIR, 'gitlab-git-test.git')
+GIT_TEST_REPO_ORIGIN = '../internal/testhelper/testdata/data/gitlab-git-test.git'.freeze
 
 module TestRepo
   def self.prepare_test_repository
     FileUtils.rm_rf(Dir["#{DEFAULT_STORAGE_DIR}/mutable-*"])
-    return if File.directory?(TEST_REPO_PATH)
 
     FileUtils.mkdir_p(DEFAULT_STORAGE_DIR)
-    clone_new_repo!(TEST_REPO_PATH)
+
+    {
+      TEST_REPO_ORIGIN => TEST_REPO_PATH,
+      GIT_TEST_REPO_ORIGIN => GIT_TEST_REPO_PATH
+    }.each do |origin, path|
+      next if File.directory?(path)
+
+      clone_new_repo!(origin, path)
+    end
+  end
+
+  def git_test_repo_read_only
+    Gitaly::Repository.new(storage_name: DEFAULT_STORAGE_NAME, relative_path: File.basename(GIT_TEST_REPO_PATH))
   end
 
   def test_repo_read_only
@@ -23,7 +36,7 @@ module TestRepo
 
   def new_mutable_test_repo
     relative_path = "mutable-#{SecureRandom.hex(6)}.git"
-    TestRepo.clone_new_repo!(File.join(DEFAULT_STORAGE_DIR, relative_path))
+    TestRepo.clone_new_repo!(TEST_REPO_ORIGIN, File.join(DEFAULT_STORAGE_DIR, relative_path))
     Gitaly::Repository.new(storage_name: DEFAULT_STORAGE_NAME, relative_path: relative_path)
   end
 
@@ -54,8 +67,8 @@ module TestRepo
     )
   end
 
-  def self.clone_new_repo!(destination)
-    return if system("git", "clone", "--quiet", "--bare", TEST_REPO_ORIGIN.to_s, destination.to_s)
+  def self.clone_new_repo!(origin, destination)
+    return if system("git", "clone", "--quiet", "--bare", origin.to_s, destination.to_s)
 
     abort "Failed to clone test repo. Try running 'make prepare-tests' and try again."
   end