Using the built-in Ruby Minitest framework, is there a way to run some code once before the entire suite runs, or even once before an entire TestClass runs? I see in the answer to this question that Test::Unit::after_tests can be used to run code after all tests have been run; is there a similar method to run code before all tests have run?
I would like to use this functionality to initialize a test database before the tests run and tear it down after they have all run.
Thanks!
This is modified from the MiniTest docs (under Customizable Test Runner Types).
class Burger
def initialize
puts "YOU CREATED A BURGER"
end
def has_cheese?
true
end
def has_pickle?
false
end
end
gem 'minitest'
require 'minitest/unit'
MiniTest::Unit.autorun
class MyMiniTest
class Unit < MiniTest::Unit
def before_suites
# code to run before the first test
p "Before everything"
end
def after_suites
# code to run after the last test
p "After everything"
end
def _run_suites(suites, type)
begin
before_suites
super(suites, type)
ensure
after_suites
end
end
def _run_suite(suite, type)
begin
suite.before_suite if suite.respond_to?(:before_suite)
super(suite, type)
ensure
suite.after_suite if suite.respond_to?(:after_suite)
end
end
end
end
MiniTest::Unit.runner = MyMiniTest::Unit.new
class BurgerTest < MiniTest::Unit::TestCase
def self.before_suite
p "hi"
end
def self.after_suite
p "bye"
end
def setup
@burger = Burger.new
end
def test_has_cheese
assert_equal true, @burger.has_cheese?
end
def test_has_pickle
assert_equal false, @burger.has_pickle?
end
end
Note that you I included gem 'minitest'
to use the gem instead of the bundled version which didn't have the MiniTest::Unit.runner
method. Here's the output.
Run options: --seed 49053
# Running tests:
"Before everything"
"hi"
YOU CREATED A BURGER
.YOU CREATED A BURGER
."bye"
"After everything"
Finished tests in 0.000662s, 3021.1480 tests/s, 3021.1480 assertions/s.
2 tests, 2 assertions, 0 failures, 0 errors, 0 skips
So it calls #setup
twice, but .before_suite
and .after_suite
only once, which is what you are looking for I think.
An alternative way to get a handle on the time before and after all the tests in a MiniTest suite have run is to put if
blocks in the setup
& teardown
methods to control that these blocks only get called once.
In this way, you can load a browser and other dependencies such as page objects just once at the beginning of your test suite and then close the browser at the end when all tests have completed.
Here's an example of this using MiniTest 5.5.1 and Watir:
class CoolTests < Minitest::Test
@@setupComplete = false # tracks whether 1-time setup has completed, so we only instantiate a browser and dependent pages/modules one time per suite run
@@testsRun = 0 # tracks how many tests have run so we can close the browser when all tests complete
def setup # Minitest#setup runs before every #test method
@@testsRun+=1 # increment tetsRun indicating that a test has run
if (!@@setupComplete) # we load the browser and necessary page objects here one-time if we haven't already
@@driver = Watir::Browser.new :chrome # instantiate new chrome browser
@@driver.window.maximize # maximize the browser window so we expect to test against Desktop UI/UX rather than Mobile UI/UX
@@setupComplete = true # setupComplete is now true as we've loaded up everything we need for our tests
end
end
def teardown # Minitest#teardown runs after every #test method
if (@@testsRun == CoolTests.runnable_methods.length) # if we've run all the tests in the suite we are finished and can then close the browser
@@driver.quit
end
end
#Tests
def test_one
p __method__
@@driver.goto('www.google.com')
assert_equal 'Google', @@driver.title, 'browser should be at google.com'
end
def test_two
p __method__
@@driver.goto('www.bing.com')
assert_equal 'Bing', @@driver.title, 'browser should be at bing.com'
end