gitlab-org--gitlab-foss/spec/support/test_env.rb

284 lines
8.8 KiB
Ruby
Raw Normal View History

2013-04-11 20:54:20 +00:00
require 'rspec/mocks'
module TestEnv
extend self
2014-09-29 14:50:20 +00:00
# When developing the seed repository, comment out the branch you will modify.
BRANCH_SHA = {
'signed-commits' => '5d4a1cb',
'not-merged-branch' => 'b83d6e3',
'branch-merged' => '498214d',
2016-08-05 11:15:06 +00:00
'empty-branch' => '7efb185',
'ends-with.json' => '98b0d8b',
2016-08-05 11:15:06 +00:00
'flatten-dir' => 'e56497b',
'feature' => '0b4bc9a',
'feature_conflict' => 'bb5206f',
'fix' => '48f0be4',
'improve/awesome' => '5937ac0',
'markdown' => '0ed8c6c',
'lfs' => 'be93687',
'master' => 'b83d6e3',
'merge-test' => '5937ac0',
2016-08-05 11:15:06 +00:00
"'test'" => 'e56497b',
'orphaned-branch' => '45127a9',
'binary-encoding' => '7b1cf43',
'gitattributes' => '5a62481',
'expand-collapse-diffs' => '4842455',
'symlink-expand-diff' => '81e6355',
2016-08-05 11:15:06 +00:00
'expand-collapse-files' => '025db92',
'expand-collapse-lines' => '238e82d',
'video' => '8879059',
2017-04-11 12:06:45 +00:00
'add-balsamiq-file' => 'b89b56d',
2016-08-05 11:15:06 +00:00
'crlf-diff' => '5938907',
2016-09-01 15:20:29 +00:00
'conflict-start' => '824be60',
2016-08-05 11:15:06 +00:00
'conflict-resolvable' => '1450cd6',
'conflict-binary-file' => '259a6fb',
2016-09-01 15:20:29 +00:00
'conflict-contains-conflict-markers' => '78a3086',
2016-08-05 11:15:06 +00:00
'conflict-missing-side' => 'eb227b3',
'conflict-non-utf8' => 'd0a293c',
2016-08-05 11:15:06 +00:00
'conflict-too-large' => '39fa04f',
'deleted-image-test' => '6c17798',
'wip' => 'b9238ee',
'csv' => '3dd0896',
2017-04-11 20:25:10 +00:00
'v1.1.0' => 'b83d6e3',
2017-05-30 15:00:36 +00:00
'add-ipython-files' => '93ee732',
'add-pdf-file' => 'e774ebd',
'add-pdf-text-binary' => '79faa7b'
2017-02-21 23:32:18 +00:00
}.freeze
# gitlab-test-fork is a fork of gitlab-fork, but we don't necessarily
# need to keep all the branches in sync.
# We currently only need a subset of the branches
FORKED_BRANCH_SHA = {
'add-submodule-version-bump' => '3f547c0',
'master' => '5937ac0',
'remove-submodule' => '2a33e0c',
'conflict-resolvable-fork' => '404fa3f'
2017-02-21 23:32:18 +00:00
}.freeze
TMP_TEST_PATH = Rails.root.join('tmp', 'tests', '**')
# Test environment
#
# See gitlab.yml.example test section for paths
#
def init(opts = {})
2013-04-11 08:50:58 +00:00
# Disable mailer for spinach tests
disable_mailer if opts[:mailer] == false
clean_test_path
# Setup GitLab shell for test instance
setup_gitlab_shell
setup_gitaly
# Create repository for FactoryGirl.create(:project)
setup_factory_repo
# Create repository for FactoryGirl.create(:forked_project_with_submodules)
setup_forked_repo
Merge Request on forked projects The good: - You can do a merge request for a forked commit and it will merge properly (i.e. it does work). - Push events take into account merge requests on forked projects - Tests around merge_actions now present, spinach, and other rspec tests - Satellites now clean themselves up rather then recreate The questionable: - Events only know about target projects - Project's merge requests only hold on to MR's where they are the target - All operations performed in the satellite The bad: - Duplication between project's repositories and satellites (e.g. commits_between) (for reference: http://feedback.gitlab.com/forums/176466-general/suggestions/3456722-merge-requests-between-projects-repos) Fixes: Make test repos/satellites only create when needed -Spinach/Rspec now only initialize test directory, and setup stubs (things that are relatively cheap) -project_with_code, source_project_with_code, and target_project_with_code now create/destroy their repos individually -fixed remote removal -How to merge renders properly -Update emails to show project/branches -Edit MR doesn't set target branch -Fix some failures on editing/creating merge requests, added a test -Added back a test around merge request observer -Clean up project_transfer_spec, Remove duplicate enable/disable observers -Ensure satellite lock files are cleaned up, Attempted to add some testing around these as well -Signifant speed ups for tests -Update formatting ordering in notes_on_merge_requests -Remove wiki schema update Fixes for search/search results -Search results was using by_project for a list of projects, updated this to use in_projects -updated search results to reference the correct (target) project -udpated search results to print both sides of the merge request Change-Id: I19407990a0950945cc95d62089cbcc6262dab1a8
2013-04-25 14:15:33 +00:00
end
def cleanup
stop_gitaly
end
def disable_mailer
2017-06-21 13:48:12 +00:00
allow_any_instance_of(NotificationService).to receive(:mailer)
.and_return(double.as_null_object)
Merge Request on forked projects The good: - You can do a merge request for a forked commit and it will merge properly (i.e. it does work). - Push events take into account merge requests on forked projects - Tests around merge_actions now present, spinach, and other rspec tests - Satellites now clean themselves up rather then recreate The questionable: - Events only know about target projects - Project's merge requests only hold on to MR's where they are the target - All operations performed in the satellite The bad: - Duplication between project's repositories and satellites (e.g. commits_between) (for reference: http://feedback.gitlab.com/forums/176466-general/suggestions/3456722-merge-requests-between-projects-repos) Fixes: Make test repos/satellites only create when needed -Spinach/Rspec now only initialize test directory, and setup stubs (things that are relatively cheap) -project_with_code, source_project_with_code, and target_project_with_code now create/destroy their repos individually -fixed remote removal -How to merge renders properly -Update emails to show project/branches -Edit MR doesn't set target branch -Fix some failures on editing/creating merge requests, added a test -Added back a test around merge request observer -Clean up project_transfer_spec, Remove duplicate enable/disable observers -Ensure satellite lock files are cleaned up, Attempted to add some testing around these as well -Signifant speed ups for tests -Update formatting ordering in notes_on_merge_requests -Remove wiki schema update Fixes for search/search results -Search results was using by_project for a list of projects, updated this to use in_projects -updated search results to reference the correct (target) project -udpated search results to print both sides of the merge request Change-Id: I19407990a0950945cc95d62089cbcc6262dab1a8
2013-04-25 14:15:33 +00:00
end
def enable_mailer
2017-06-21 13:48:12 +00:00
allow_any_instance_of(NotificationService).to receive(:mailer)
.and_call_original
Merge Request on forked projects The good: - You can do a merge request for a forked commit and it will merge properly (i.e. it does work). - Push events take into account merge requests on forked projects - Tests around merge_actions now present, spinach, and other rspec tests - Satellites now clean themselves up rather then recreate The questionable: - Events only know about target projects - Project's merge requests only hold on to MR's where they are the target - All operations performed in the satellite The bad: - Duplication between project's repositories and satellites (e.g. commits_between) (for reference: http://feedback.gitlab.com/forums/176466-general/suggestions/3456722-merge-requests-between-projects-repos) Fixes: Make test repos/satellites only create when needed -Spinach/Rspec now only initialize test directory, and setup stubs (things that are relatively cheap) -project_with_code, source_project_with_code, and target_project_with_code now create/destroy their repos individually -fixed remote removal -How to merge renders properly -Update emails to show project/branches -Edit MR doesn't set target branch -Fix some failures on editing/creating merge requests, added a test -Added back a test around merge request observer -Clean up project_transfer_spec, Remove duplicate enable/disable observers -Ensure satellite lock files are cleaned up, Attempted to add some testing around these as well -Signifant speed ups for tests -Update formatting ordering in notes_on_merge_requests -Remove wiki schema update Fixes for search/search results -Search results was using by_project for a list of projects, updated this to use in_projects -updated search results to reference the correct (target) project -udpated search results to print both sides of the merge request Change-Id: I19407990a0950945cc95d62089cbcc6262dab1a8
2013-04-25 14:15:33 +00:00
end
def disable_pre_receive
allow_any_instance_of(Gitlab::Git::Hook).to receive(:trigger).and_return([true, nil])
end
# Clean /tmp/tests
#
# Keeps gitlab-shell and gitlab-test
def clean_test_path
Dir[TMP_TEST_PATH].each do |entry|
unless File.basename(entry) =~ /\A(gitaly|gitlab-(shell|test|test_bare|test-fork|test-fork_bare))\z/
FileUtils.rm_rf(entry)
end
end
FileUtils.mkdir_p(repos_path)
FileUtils.mkdir_p(backup_path)
FileUtils.mkdir_p(pages_path)
end
def clean_gitlab_test_path
Dir[TMP_TEST_PATH].each do |entry|
if File.basename(entry) =~ /\A(gitlab-(test|test_bare|test-fork|test-fork_bare))\z/
FileUtils.rm_rf(entry)
end
end
end
def setup_gitlab_shell
shell_needs_update = component_needs_update?(Gitlab.config.gitlab_shell.path,
Gitlab::Shell.version_required)
unless !shell_needs_update || system('rake', 'gitlab:shell:install')
raise 'Can`t clone gitlab-shell'
end
Merge Request on forked projects The good: - You can do a merge request for a forked commit and it will merge properly (i.e. it does work). - Push events take into account merge requests on forked projects - Tests around merge_actions now present, spinach, and other rspec tests - Satellites now clean themselves up rather then recreate The questionable: - Events only know about target projects - Project's merge requests only hold on to MR's where they are the target - All operations performed in the satellite The bad: - Duplication between project's repositories and satellites (e.g. commits_between) (for reference: http://feedback.gitlab.com/forums/176466-general/suggestions/3456722-merge-requests-between-projects-repos) Fixes: Make test repos/satellites only create when needed -Spinach/Rspec now only initialize test directory, and setup stubs (things that are relatively cheap) -project_with_code, source_project_with_code, and target_project_with_code now create/destroy their repos individually -fixed remote removal -How to merge renders properly -Update emails to show project/branches -Edit MR doesn't set target branch -Fix some failures on editing/creating merge requests, added a test -Added back a test around merge request observer -Clean up project_transfer_spec, Remove duplicate enable/disable observers -Ensure satellite lock files are cleaned up, Attempted to add some testing around these as well -Signifant speed ups for tests -Update formatting ordering in notes_on_merge_requests -Remove wiki schema update Fixes for search/search results -Search results was using by_project for a list of projects, updated this to use in_projects -updated search results to reference the correct (target) project -udpated search results to print both sides of the merge request Change-Id: I19407990a0950945cc95d62089cbcc6262dab1a8
2013-04-25 14:15:33 +00:00
end
def setup_gitaly
socket_path = Gitlab::GitalyClient.address('default').sub(/\Aunix:/, '')
gitaly_dir = File.dirname(socket_path)
gitaly_needs_update = component_needs_update?(gitaly_dir,
Gitlab::GitalyClient.expected_server_version)
unless !gitaly_needs_update || system('rake', "gitlab:gitaly:install[#{gitaly_dir}]")
raise "Can't clone gitaly"
end
start_gitaly(gitaly_dir)
end
def start_gitaly(gitaly_dir)
2017-07-31 13:17:14 +00:00
if ENV['CI'].present?
# Gitaly has been spawned outside this process already
return
end
spawn_script = Rails.root.join('scripts/gitaly-test-spawn').to_s
@gitaly_pid = Bundler.with_original_env { IO.popen([spawn_script], &:read).to_i }
end
def stop_gitaly
return unless @gitaly_pid
Process.kill('KILL', @gitaly_pid)
end
def setup_factory_repo
setup_repo(factory_repo_path, factory_repo_path_bare, factory_repo_name,
BRANCH_SHA)
end
# This repo has a submodule commit that is not present in the main test
# repository.
def setup_forked_repo
setup_repo(forked_repo_path, forked_repo_path_bare, forked_repo_name,
FORKED_BRANCH_SHA)
end
def setup_repo(repo_path, repo_path_bare, repo_name, refs)
clone_url = "https://gitlab.com/gitlab-org/#{repo_name}.git"
Merge Request on forked projects The good: - You can do a merge request for a forked commit and it will merge properly (i.e. it does work). - Push events take into account merge requests on forked projects - Tests around merge_actions now present, spinach, and other rspec tests - Satellites now clean themselves up rather then recreate The questionable: - Events only know about target projects - Project's merge requests only hold on to MR's where they are the target - All operations performed in the satellite The bad: - Duplication between project's repositories and satellites (e.g. commits_between) (for reference: http://feedback.gitlab.com/forums/176466-general/suggestions/3456722-merge-requests-between-projects-repos) Fixes: Make test repos/satellites only create when needed -Spinach/Rspec now only initialize test directory, and setup stubs (things that are relatively cheap) -project_with_code, source_project_with_code, and target_project_with_code now create/destroy their repos individually -fixed remote removal -How to merge renders properly -Update emails to show project/branches -Edit MR doesn't set target branch -Fix some failures on editing/creating merge requests, added a test -Added back a test around merge request observer -Clean up project_transfer_spec, Remove duplicate enable/disable observers -Ensure satellite lock files are cleaned up, Attempted to add some testing around these as well -Signifant speed ups for tests -Update formatting ordering in notes_on_merge_requests -Remove wiki schema update Fixes for search/search results -Search results was using by_project for a list of projects, updated this to use in_projects -updated search results to reference the correct (target) project -udpated search results to print both sides of the merge request Change-Id: I19407990a0950945cc95d62089cbcc6262dab1a8
2013-04-25 14:15:33 +00:00
unless File.directory?(repo_path)
system(*%W(#{Gitlab.config.git.bin_path} clone -q #{clone_url} #{repo_path}))
2014-09-29 14:50:20 +00:00
end
set_repo_refs(repo_path, refs)
2014-09-29 14:50:20 +00:00
unless File.directory?(repo_path_bare)
# We must copy bare repositories because we will push to them.
system(git_env, *%W(#{Gitlab.config.git.bin_path} clone -q --bare #{repo_path} #{repo_path_bare}))
end
2013-11-12 14:11:46 +00:00
end
def copy_repo(project, bare_repo:, refs:)
target_repo_path = File.expand_path(project.repository_storage_path + "/#{project.full_path}.git")
FileUtils.mkdir_p(target_repo_path)
FileUtils.cp_r("#{File.expand_path(bare_repo)}/.", target_repo_path)
FileUtils.chmod_R 0755, target_repo_path
set_repo_refs(target_repo_path, refs)
end
def repos_path
Gitlab.config.repositories.storages.default['path']
end
2014-09-27 07:56:19 +00:00
def backup_path
Gitlab.config.backup.path
end
def pages_path
Gitlab.config.pages.path
end
# When no cached assets exist, manually hit the root path to create them
#
# Otherwise they'd be created by the first test, often timing out and
# causing a transient test failure
def eager_load_driver_server
return unless ENV['CI']
return unless defined?(Capybara)
puts "Starting the Capybara driver server..."
Capybara.current_session.visit '/'
end
def factory_repo_path_bare
"#{factory_repo_path}_bare"
end
def forked_repo_path_bare
"#{forked_repo_path}_bare"
end
2014-09-27 07:56:19 +00:00
private
def factory_repo_path
2014-09-29 14:50:20 +00:00
@factory_repo_path ||= Rails.root.join('tmp', 'tests', factory_repo_name)
end
2014-09-27 07:56:19 +00:00
def factory_repo_name
'gitlab-test'
end
def forked_repo_path
@forked_repo_path ||= Rails.root.join('tmp', 'tests', forked_repo_name)
end
def forked_repo_name
'gitlab-test-fork'
end
# Prevent developer git configurations from being persisted to test
# repositories
def git_env
{ 'GIT_TEMPLATE_DIR' => '' }
end
def set_repo_refs(repo_path, branch_sha)
instructions = branch_sha.map { |branch, sha| "update refs/heads/#{branch}\x00#{sha}\x00" }.join("\x00") << "\x00"
update_refs = %W(#{Gitlab.config.git.bin_path} update-ref --stdin -z)
reset = proc do
Dir.chdir(repo_path) do
IO.popen(update_refs, "w") { |io| io.write(instructions) }
$?.success?
end
end
# Try to reset without fetching to avoid using the network.
unless reset.call
raise 'Could not fetch test seed repository.' unless system(*%W(#{Gitlab.config.git.bin_path} -C #{repo_path} fetch origin))
# Before we used Git clone's --mirror option, bare repos could end up
# with missing refs, clearing them and retrying should fix the issue.
cleanup && clean_gitlab_test_path && init unless reset.call
end
end
def component_needs_update?(component_folder, expected_version)
version = File.read(File.join(component_folder, 'VERSION')).strip
# Notice that this will always yield true when using branch versions
# (`=branch_name`), but that actually makes sure the server is always based
# on the latest branch revision.
version != expected_version
rescue Errno::ENOENT
true
end
end