Add latest changes from gitlab-org/gitlab@master

6 files changed, 355 insertions, 36 deletions

diff --git a/qa/qa/resource/issue.rb b/qa/qa/resource/issue.rb
index ffffa0eecda..2144e619c70 100644
--- a/qa/qa/resource/issue.rb
+++ b/qa/qa/resource/issue.rb
@@ -14,11 +14,11 @@ module QA
         end
       end
 
-      attribute :id
-      attribute :iid
-      attribute :assignee_ids
-      attribute :labels
-      attribute :title
+      attributes :id,
+                 :iid,
+                 :assignee_ids,
+                 :labels,
+                 :title
 
       def initialize
         @assignee_ids = []
@@ -41,13 +41,21 @@ module QA
       end
 
       def api_get_path
-        "/projects/#{project.id}/issues/#{id}"
+        "/projects/#{project.id}/issues/#{iid}"
       end
 
       def api_post_path
         "/projects/#{project.id}/issues"
       end
 
+      def api_put_path
+        "/projects/#{project.id}/issues/#{iid}"
+      end
+
+      def api_comments_path
+        "#{api_get_path}/notes"
+      end
+
       def api_post_body
         {
           assignee_ids: assignee_ids,
@@ -59,20 +67,28 @@ module QA
         end
       end
 
-      def api_put_path
-        "/projects/#{project.id}/issues/#{iid}"
-      end
-
       def set_issue_assignees(assignee_ids:)
         put_body = { assignee_ids: assignee_ids }
         response = put Runtime::API::Request.new(api_client, api_put_path).url, put_body
 
         unless response.code == HTTP_STATUS_OK
-          raise ResourceUpdateFailedError, "Could not update issue assignees to #{assignee_ids}. Request returned (#{response.code}): `#{response}`."
+          raise(
+            ResourceUpdateFailedError,
+            "Could not update issue assignees to #{assignee_ids}. Request returned (#{response.code}): `#{response}`."
+          )
         end
 
         QA::Runtime::Logger.debug("Successfully updated issue assignees to #{assignee_ids}")
       end
+
+      # Get issue comments
+      #
+      # @return [Array]
+      def comments(auto_paginate: false)
+        return parse_body(api_get_from(api_comments_path)) unless auto_paginate
+
+        auto_paginated_response(Runtime::API::Request.new(api_client, api_comments_path, per_page: '100').url)
+      end
     end
   end
 end
diff --git a/qa/qa/resource/merge_request.rb b/qa/qa/resource/merge_request.rb
index 8d9de0ea718..419893f0b11 100644
--- a/qa/qa/resource/merge_request.rb
+++ b/qa/qa/resource/merge_request.rb
@@ -160,9 +160,10 @@ module QA
       # Get MR comments
       #
       # @return [Array]
-      def comments
-        response = get(Runtime::API::Request.new(api_client, api_comments_path).url)
-        parse_body(response)
+      def comments(auto_paginate: false)
+        return parse_body(api_get_from(api_comments_path)) unless auto_paginate
+
+        auto_paginated_response(Runtime::API::Request.new(api_client, api_comments_path, per_page: '100').url)
       end
 
       private
diff --git a/qa/qa/resource/project.rb b/qa/qa/resource/project.rb
index 70cec0cf747..12ae644cab8 100644
--- a/qa/qa/resource/project.rb
+++ b/qa/qa/resource/project.rb
@@ -32,7 +32,7 @@ module QA
       end
 
       attribute :path_with_namespace do
-        "#{sandbox_path}#{group.path}/#{name}" if group
+        "#{group.full_path}/#{name}"
       end
 
       alias_method :full_path, :path_with_namespace
@@ -268,14 +268,16 @@ module QA
         result[:import_status]
       end
 
-      def commits
-        response = get(request_url(api_commits_path))
-        parse_body(response)
+      def commits(auto_paginate: false)
+        return parse_body(api_get_from(api_commits_path)) unless auto_paginate
+
+        auto_paginated_response(request_url(api_commits_path, per_page: '100'))
       end
 
-      def merge_requests
-        response = get(request_url(api_merge_requests_path))
-        parse_body(response)
+      def merge_requests(auto_paginate: false)
+        return parse_body(api_get_from(api_merge_requests_path)) unless auto_paginate
+
+        auto_paginated_response(request_url(api_merge_requests_path, per_page: '100'))
       end
 
       def merge_request_with_title(title)
@@ -299,9 +301,10 @@ module QA
         parse_body(response)
       end
 
-      def repository_branches
-        response = get(request_url(api_repository_branches_path))
-        parse_body(response)
+      def repository_branches(auto_paginate: false)
+        return parse_body(api_get_from(api_repository_branches_path)) unless auto_paginate
+
+        auto_paginated_response(request_url(api_repository_branches_path, per_page: '100'))
       end
 
       def repository_tags
@@ -324,19 +327,22 @@ module QA
         parse_body(response)
       end
 
-      def issues
-        response = get(request_url(api_issues_path))
-        parse_body(response)
+      def issues(auto_paginate: false)
+        return parse_body(api_get_from(api_issues_path)) unless auto_paginate
+
+        auto_paginated_response(request_url(api_issues_path, per_page: '100'))
       end
 
-      def labels
-        response = get(request_url(api_labels_path))
-        parse_body(response)
+      def labels(auto_paginate: false)
+        return parse_body(api_get_from(api_labels_path)) unless auto_paginate
+
+        auto_paginated_response(request_url(api_labels_path, per_page: '100'))
       end
 
-      def milestones
-        response = get(request_url(api_milestones_path))
-        parse_body(response)
+      def milestones(auto_paginate: false)
+        return parse_body(api_get_from(api_milestones_path)) unless auto_paginate
+
+        auto_paginated_response(request_url(api_milestones_path, per_page: '100'))
       end
 
       def wikis
diff --git a/qa/qa/specs/features/api/1_manage/import_large_github_repo_spec.rb b/qa/qa/specs/features/api/1_manage/import_large_github_repo_spec.rb
new file mode 100644
index 00000000000..22917d0d146
--- /dev/null
+++ b/qa/qa/specs/features/api/1_manage/import_large_github_repo_spec.rb
@@ -0,0 +1,286 @@
+# frozen_string_literal: true
+
+require 'octokit'
+require 'parallel'
+
+# rubocop:disable Rails/Pluck
+module QA
+  # Only executes in custom job/pipeline
+  RSpec.describe 'Manage', :github, :requires_admin, only: { job: 'large-github-import' } do
+    describe 'Project import' do
+      let(:api_client) { Runtime::API::Client.as_admin }
+      let(:group) do
+        Resource::Group.fabricate_via_api! do |resource|
+          resource.api_client = api_client
+        end
+      end
+
+      let(:user) do
+        Resource::User.fabricate_via_api! do |resource|
+          resource.api_client = api_client
+          resource.hard_delete_on_api_removal = true
+        end
+      end
+
+      let(:differ) { RSpec::Support::Differ.new(color: true) }
+      let(:github_repo) { 'allure-framework/allure-ruby' }
+
+      let(:github_client) do
+        Octokit.middleware = Faraday::RackBuilder.new do |builder|
+          builder.response(:logger, Runtime::Logger.logger, headers: false, bodies: false)
+        end
+
+        Octokit::Client.new(access_token: Runtime::Env.github_access_token, auto_paginate: true)
+      end
+
+      let(:gh_branches) { github_client.branches(github_repo).map(&:name) }
+      let(:gh_commits) { github_client.commits(github_repo).map(&:sha) }
+      let(:gh_repo) { github_client.repository(github_repo) }
+      let(:gh_labels) { github_client.labels(github_repo) }
+      let(:gh_milestones) { github_client.list_milestones(github_repo, state: 'all') }
+
+      let(:gh_all_issues) do
+        github_client.list_issues(github_repo, state: 'all')
+      end
+
+      let(:gh_prs) do
+        gh_all_issues.select(&:pull_request).each_with_object({}) do |pr, hash|
+          hash[pr.title] = {
+            body: pr.body || '',
+            comments: [*gh_pr_comments[pr.html_url], *gh_issue_comments[pr.html_url]].compact.sort
+          }
+        end
+      end
+
+      let(:gh_issues) do
+        gh_all_issues.reject(&:pull_request).each_with_object({}) do |issue, hash|
+          hash[issue.title] = {
+            body: issue.body || '',
+            comments: gh_issue_comments[issue.html_url]
+          }
+        end
+      end
+
+      let(:gh_issue_comments) do
+        github_client.issues_comments(github_repo).each_with_object(Hash.new { |h, k| h[k] = [] }) do |c, hash|
+          hash[c.html_url.gsub(/\#\S+/, "")] << c.body # use base html url as key
+        end
+      end
+
+      let(:gh_pr_comments) do
+        github_client.pull_requests_comments(github_repo).each_with_object(Hash.new { |h, k| h[k] = [] }) do |c, hash|
+          hash[c.html_url.gsub(/\#\S+/, "")] << c.body # use base html url as key
+        end
+      end
+
+      let(:imported_project) do
+        Resource::ProjectImportedFromGithub.fabricate_via_api! do |project|
+          project.add_name_uuid = false
+          project.name = 'imported-project'
+          project.group = group
+          project.github_personal_access_token = Runtime::Env.github_access_token
+          project.github_repository_path = github_repo
+          project.api_client = api_client
+        end
+      end
+
+      before do
+        group.add_member(user, Resource::Members::AccessLevel::MAINTAINER)
+      end
+
+      it 'imports large Github repo via api' do
+        imported_project # import the project
+        fetch_github_objects # fetch all objects right after import has started
+
+        expect { imported_project.reload!.import_status }.to eventually_eq('finished').within(
+          duration: 3600,
+          interval: 30
+        )
+
+        aggregate_failures do
+          verify_repository_import
+          verify_merge_requests_import
+          verify_issues_import
+          verify_labels_import
+          verify_milestones_import
+        end
+      end
+
+      # Persist all objects from repository being imported
+      #
+      # @return [void]
+      def fetch_github_objects
+        Runtime::Logger.debug("Fetching objects for github repo: '#{github_repo}'")
+
+        gh_repo
+        gh_branches
+        gh_commits
+        gh_prs
+        gh_issues
+        gh_labels
+        gh_milestones
+      end
+
+      # Verify repository imported correctly
+      #
+      # @return [void]
+      def verify_repository_import
+        branches = imported_project.repository_branches(auto_paginate: true).map { |b| b[:name] }
+        commits = imported_project.commits(auto_paginate: true).map { |c| c[:id] }
+
+        expect(imported_project.description).to eq(gh_repo.description)
+        # check via include, importer creates more branches
+        # https://gitlab.com/gitlab-org/gitlab/-/issues/332711
+        expect(branches).to include(*gh_branches)
+        expect(commits).to match_array(gh_commits)
+      end
+
+      # Verify imported merge requests and mr issues
+      #
+      # @return [void]
+      def verify_merge_requests_import
+        verify_mrs_or_issues('mrs')
+      end
+
+      # Verify imported issues and issue comments
+      #
+      # @return [void]
+      def verify_issues_import
+        verify_mrs_or_issues('issues')
+      end
+
+      # Verify imported labels
+      #
+      # @return [void]
+      def verify_labels_import
+        labels = imported_project.labels(auto_paginate: true).map { |label| label.slice(:name, :color) }
+        actual_labels = gh_labels.map { |label| { name: label.name, color: "##{label.color}" } }
+
+        expect(labels.length).to eq(actual_labels.length)
+        expect(labels).to match_array(actual_labels)
+      end
+
+      # Verify milestones import
+      #
+      # @return [void]
+      def verify_milestones_import
+        milestones = imported_project.milestones(auto_paginate: true).map { |ms| ms.slice(:title, :description) }
+        actual_milestones = gh_milestones.map { |ms| { title: ms.title, description: ms.description } }
+
+        expect(milestones.length).to eq(actual_milestones.length)
+        expect(milestones).to match_array(actual_milestones)
+      end
+
+      private
+
+      # Verify imported mrs or issues
+      #
+      # @param [String] type verification object, 'mrs' or 'issues'
+      # @return [void]
+      def verify_mrs_or_issues(type)
+        msg = ->(title) { "expected #{type} with title '#{title}' to have" }
+        expected = type == 'mrs' ? mrs : gl_issues
+        actual = type == 'mrs' ? gh_prs : gh_issues
+
+        expect(expected.keys).to match_array(actual.keys)
+        actual.each do |title, actual_item|
+          expected_item = expected[title]
+
+          expect(expected_item).to be_truthy, "#{msg.call(title)} been imported"
+          next unless expected_item
+
+          expect(expected_item[:body]).to(
+            include(actual_item[:body]),
+            "#{msg.call(title)} same description. #{diff(expected_item[:body], actual_item[:body])}"
+          )
+          expect(expected_item[:comments].length).to(
+            eq(actual_item[:comments].length),
+            "#{msg.call(title)} same amount of comments"
+          )
+          expect(expected_item[:comments]).to match_array(actual_item[:comments])
+        end
+      end
+
+      # Imported project merge requests
+      #
+      # @return [Hash]
+      def mrs
+        @mrs ||= begin
+          imported_mrs = imported_project.merge_requests(auto_paginate: true)
+          # fetch comments in parallel since we need to do it for each mr separately
+          mrs_hashes = Parallel.map(imported_mrs, in_processes: 5) do |mr|
+            resource = Resource::MergeRequest.init do |resource|
+              resource.project = imported_project
+              resource.iid = mr[:iid]
+              resource.api_client = api_client
+            end
+
+            {
+              title: mr[:title],
+              body: mr[:description],
+              comments: resource.comments(auto_paginate: true)
+                # remove system notes
+                .reject { |c| c[:system] || c[:body].match?(/^(\*\*Review:\*\*)|(\*Merged by:).*/) }
+                .map { |c| sanitize(c[:body]) }
+            }
+          end
+
+          mrs_hashes.each_with_object({}) do |mr, hash|
+            hash[mr[:title]] = {
+              body: mr[:body],
+              comments: mr[:comments]
+            }
+          end
+        end
+      end
+
+      # Imported project issues
+      #
+      # @return [Hash]
+      def gl_issues
+        @gl_issues ||= begin
+          imported_issues = imported_project.issues(auto_paginate: true)
+          # fetch comments in parallel since we need to do it for each mr separately
+          issue_hashes = Parallel.map(imported_issues, in_processes: 5) do |issue|
+            resource = Resource::Issue.init do |issue_resource|
+              issue_resource.project = imported_project
+              issue_resource.iid = issue[:iid]
+              issue_resource.api_client = api_client
+            end
+
+            {
+              title: issue[:title],
+              body: issue[:description],
+              comments: resource.comments(auto_paginate: true).map { |c| sanitize(c[:body]) }
+            }
+          end
+
+          issue_hashes.each_with_object({}) do |issue, hash|
+            hash[issue[:title]] = {
+              body: issue[:body],
+              comments: issue[:comments]
+            }
+          end
+        end
+      end
+
+      # Remove added prefixes by importer
+      #
+      # @param [String] body
+      # @return [String]
+      def sanitize(body)
+        body.gsub(/\*Created by: \S+\*\n\n/, "")
+      end
+
+      # Diff of 2 objects
+      #
+      # @param [Object] actual
+      # @param [Object] expected
+      # @return [String]
+      def diff(actual, expected)
+        "diff:\n#{differ.diff(actual, expected)}"
+      end
+    end
+  end
+end
+# rubocop:enable Rails/Pluck
diff --git a/qa/qa/support/api.rb b/qa/qa/support/api.rb
index de9da3171b0..1493feeeed7 100644
--- a/qa/qa/support/api.rb
+++ b/qa/qa/support/api.rb
@@ -79,11 +79,20 @@ module QA
         error.response
       end
 
+      def auto_paginated_response(url)
+        pages = []
+        with_paginated_response_body(url) { |response| pages << response }
+
+        pages.flatten
+      end
+
       def with_paginated_response_body(url)
         loop do
           response = get(url)
+          page, pages = response.headers.values_at(:x_page, :x_total_pages)
+          api_endpoint = url.match(%r{v4/(\S+)\?})[1]
 
-          QA::Runtime::Logger.debug("Fetching page #{response.headers[:x_page]} of #{response.headers[:x_total_pages]}...")
+          QA::Runtime::Logger.debug("Fetching page (#{page}/#{pages}) for '#{api_endpoint}' ...") unless pages.to_i <= 1
 
           yield parse_body(response)
 
@@ -96,7 +105,7 @@ module QA
 
       def pagination_links(response)
         response.headers[:link].split(',').map do |link|
-          match = link.match(/\<(?<url>.*)\>\; rel=\"(?<rel>\w+)\"/)
+          match = link.match(/<(?<url>.*)>; rel="(?<rel>\w+)"/)
           break nil unless match
 
           { url: match[:url], rel: match[:rel] }
diff --git a/qa/spec/support/matchers/eventually_matcher.rb b/qa/spec/support/matchers/eventually_matcher.rb
index 3f0afd6fb54..2de6598be4c 100644
--- a/qa/spec/support/matchers/eventually_matcher.rb
+++ b/qa/spec/support/matchers/eventually_matcher.rb
@@ -24,6 +24,7 @@ module Matchers
       chain(:within) do |options = {}|
         @duration = options[:duration]
         @attempts = options[:attempts]
+        @interval = options[:interval]
       end
 
       def supports_block_expectations?
@@ -55,7 +56,7 @@ module Matchers
         QA::Support::Retrier.retry_until(
           max_attempts: @attempts,
           max_duration: @duration,
-          sleep_interval: 0.5
+          sleep_interval: @interval || 0.5
         ) do
           public_send(expectation_name, actual)
         rescue RSpec::Expectations::ExpectationNotMetError, QA::Resource::ApiFabricator::ResourceNotFoundError