ResqueSpec

Build
Status

A test double of Resque for RSpec and Cucumber. The code was originally based on http://github.com/justinweiss/resque_unit.

ResqueSpec will also fire Resque hooks if you are using them. See below.

Install

Update your Gemfile to include resque_spec only in the test group (Not using bundler? Do the necessary thing for your app's gem management and use bundler. resque_spec monkey patches resque it should only be used with your tests!)

group :test do
  gem 'resque_spec'
end

Cucumber

By default, the above will add the ResqueSpec module and make it available in Cucumber. If you want the with_resque and without_resque helpers, manually require the resque_spec/cucumber module:

require 'resque_spec/cucumber'

This can be done in features/support/env.rb or in a specific support file such as features/support/resque.rb.

What is ResqueSpec?

ResqueSpec implements the stable API for Resque 1.19+ (which is enqueue, enqueue_to, dequeue, peek, reserve, size, the Resque hooks, and because of the way resque_scheduler works Job.create and Job.destroy).

It does not have a test double for Redis, so this may lead to some interesting and puzzling behaviour if you use some of the popular Resque plugins (such as resque_lock).

Resque with Specs

Given this scenario

Given a person
When I recalculate
Then the person has calculate queued

And I write this spec using the resque_spec matcher

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the Person queue" do
    person.recalculate
    expect(Person).to have_queued(person.id, :calculate)
  end
end

And I see that the have_queued assertion is asserting that the Person queue has a job with arguments person.id and :calculate

And I take note of the before block that is calling reset! for every spec

And I might use the in statement to specify the queue:

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the Person queue" do
    person.recalculate
    expect(Person).to have_queued(person.id, :calculate).in(:people)
  end
end

And I might write this as a Cucumber step

Then /the (\w?) has (\w?) queued/ do |thing, method|
  thing_obj = instance_variable_get("@#{thing}")
  expect(thing_obj.class).to have_queued(thing_obj.id, method.to_sym)
end

Then I write some code to make it pass:

class Person
  @queue = :people

  def recalculate
    Resque.enqueue(Person, id, :calculate)
  end
end

You can check the size of the queue in your specs too.

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds an entry to the Person queue" do
    person.recalculate
    expect(Person).to have_queue_size_of(1)
  end
end

Turning off ResqueSpec and calling directly to Resque

Occasionally, you want to run your specs directly against Resque instead of ResqueSpec. For one at a time use, pass a block to the without_resque_spec helper:

describe "#recalculate" do
  it "recalculates the persons score" do
    without_resque_spec do
      person.recalculate
    end
    ... assert recalculation after job done
  end
end

Or you can manage when ResqueSpec is disabled by flipping the ResqueSpec.disable_ext flag:

# disable ResqueSpec
ResqueSpec.disable_ext = true

You will most likely (but not always, see the Resque docs) need to ensure that you have redis running.

ResqueMailer with Specs

To use with ResqueMailer you should have an initializer that does not exclude the test (or cucumber) environment. Your initializer will probably end up looking like:

# config/initializers/resque_mailer.rb
Resque::Mailer.excluded_environments = []

If you have a mailer like this:

class ExampleMailer < ActionMailer::Base
  include Resque::Mailer

  def welcome_email(user_id)
  end
end

You can write a spec like this:

describe "#welcome_email" do
  before do
    ResqueSpec.reset!
    Examplemailer.welcome_email(user.id).deliver
  end

  subject { described_class }
  it { should have_queue_size_of(1) }
  it { should have_queued(:welcome_email, [user.id]) }
end

resque-scheduler with Specs

Update the Gemfile to enable the resque-schedular matchers:

group :test do
  gem 'resque_spec', require: 'resque_spec/scheduler'
end

Given this scenario

Given a person
When I schedule a recalculate
Then the person has calculate scheduled

And I write this spec using the resque_spec matcher

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the Person queue" do
    person.recalculate
    expect(Person).to have_scheduled(person.id, :calculate)
  end
end

And I might use the at statement to specify the time:

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the Person queue" do
    person.recalculate

    # Is it scheduled to be executed at 2010-02-14 06:00:00 ?
    expect(Person).to have_scheduled(person.id, :calculate).at(Time.mktime(2010,2,14,6,0,0))
  end
end

And I might use the in statement to specify time interval (in seconds):

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the Person queue" do
    person.recalculate

    # Is it scheduled to be executed in 5 minutes?
    expect(Person).to have_scheduled(person.id, :calculate).in(5 * 60)
  end
end

You can also check the size of the schedule:

describe "#recalculate" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the Person queue" do
    person.recalculate

    expect(Person).to have_schedule_size_of(1)
  end
end

(And I take note of the before block that is calling reset! for every spec)

You can explicitly specify the queue when using enqueue_at_with_queue and enqueue_in_with_queue:

describe "#recalculate_in_future" do
  before do
    ResqueSpec.reset!
  end

  it "adds person.calculate to the :future queue" do
    person.recalculate_in_future

    Person.should have_schedule_size_of(1).queue(:future)
  end
end

And I might write this as a Cucumber step

Then /the (\w?) has (\w?) scheduled/ do |thing, method|
  thing_obj = instance_variable_get("@#{thing}")
  expect(thing_obj.class).to have_scheduled(thing_obj.id, method.to_sym)
end

Then I write some code to make it pass:

class Person
  @queue = :people

  def recalculate
    Resque.enqueue_at(Time.now + 3600, Person, id, :calculate)
  end

  def recalculate_in_future
    Resque.enqueue_at_with_queue(:future, Time.now + 3600, Person, id, :calculate)
  end
end

Performing Jobs in Specs

Normally, ResqueSpec does not perform queued jobs within tests. You may want to make assertions based on the result of your jobs. ResqueSpec can process jobs immediately as they are queued or under your control.

Performing jobs immediately

To perform jobs immediately, you can pass a block to the with_resque helper:

Given this scenario

Given a game
When I score
Then the game has a score

I might write this as a Cucumber step

When /I score/ do
  with_resque do
    visit game_path
    click_link 'Score!'
  end
end

Or I write this spec using the with_resque helper

describe "#score!" do
  before do
    ResqueSpec.reset!
  end

  it "increases the score" do
    with_resque do
      game.score!
    end
    expect(game.score).to == 10
  end
end

You can turn this behavior on by setting ResqueSpec.inline = true.

Performing jobs at your discretion

You can perform the first job on a queue at a time, or perform all the jobs on a queue. Use ResqueSpec#perform_next(queue_name) or ResqueSpec#perform_all(queue_name)

Given this scenario:

Given a game
When I score
And the score queue runs
Then the game has a score

I might write this as a Cucumber step

When /the (\w+) queue runs/ do |queue_name|
  ResqueSpec.perform_all(queue_name)
end

Hooks

Resque provides hooks at different points of the queueing lifecylce. ResqueSpec fires these hooks when appropriate.

The before and after enqueue hooks are always called when you use Resque#enqueue. If your before_enqueue hook returns false, the job will not be queued and after_enqueue will not be called.

The perform hooks: before, around, after, and on failure are fired by ResqueSpec if you are using the with_resque helper or set ResqueSpec.inline = true.

Important! If you are using resque-scheduler, Resque#enqueue_at/enqueue_in does not fire the after enqueue hook (the job has not been queued yet!), but will fire the perform hooks if you are using inline mode.

Note on Patches/Pull Requests

  • Fork the project.
  • Make your feature addition or bug fix.
  • Add tests for it. This is important so I don't break it in a future version unintentionally.
  • Commit, do not mess with rakefile, version, or history. (if you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull)
  • Send me a pull request. Bonus points for topic branches.

Author

I made resque_spec because resque is awesome and should be easy to spec. Follow me on Github and Twitter.

Contributors

  • Kenneth Kalmer (@kennethkalmer) : rspec dependency fix
  • Brian Cardarella (@bcardarella) : fix mutation bug
  • Joshua Davey (@joshdavey) : with_resque helper
  • Lar Van Der Jagt (@supaspoida) : with_resque helper
  • Evan Sagge (@evansagge) : Hook in via Job.create, have_queued.in
  • Jon Larkowski (@l4rk) : inline perform
  • James Conroy-Finn (@jcf) : spec fix
  • Dennis Walters (@ess) : enqueue_in support
  • (@RipTheJacker) : remove_delayed support
  • Kurt Werle (@kwerle) : explicit require spec for v020
  • (@dwilkie) : initial before_enqueue support
  • Marcin Balinski (@marcinb) : have_schedule_size_of matcher, schedule matcher at, in
  • (@alexeits) : fix matcher in bug with RSpec 2.8.0
  • (@ToadJamb) : encode/decode of Resque job arguments
  • Mateusz Konikowski (@mkonikowski) : support for anything matcher
  • Mathieu Ravaux (@mathieuravaux) : without_resque_spec support
  • Arjan van der Gaag (@avdgaag) : peek support
  • (@dtsiknis) : Updated removed_delayed
  • Li Ellis Gallardo (@lellisga) : fix inline/disable_ext bug
  • Jeff Deville (@jeffdeville) : Resque.size
  • Frank Wambutt (@opinel) : Fix DST problem in have_scheduled
  • Luke Melia (@lukemelia) : Add times chained matcher
  • Pablo Fernandez (@heelhook) : Add have_queue_size_of_at_least and have_schedule_size_of_at_least matchers
  • (@k1w1) : Add support for enqueue_at_with_queue/enqueue_in_with_queue
  • Ozéias Sant'ana (@ozeias) : Update specs to RSpec 2.10
  • Yuya Kitajima (@yuyak) : Add ResqueMailer examples to README
  • Andrés Bravo (@andresbravog) : Replace rspec dependency with explicit dependencies
  • Ben Woosley (@Empact) : Loosen rubygems version constraint
  • Jeff Dickey (@dickeyxxx) : Remove 2.0 warnings, added Travis
  • Earle Clubb (@eclubb) : be_queued matcher
  • Erkki Eilonen (@erkki) : RSpec 3 support
  • Gavin Heavyside (@gavinheavyside) : RSpec three warnings
  • Pavel Khrulev (@PaulSchuher) : Resque 2 and RSpec 3 support
  • Ilya Katz (@ilyakatz) : Cleanup README.md for RSpec 3
  • (@addbrick) : Compare times as integers in have_scheduled matcher
  • Serious Haircut (@serioushaircut) : Fix ArgumentListMatcher to make it work with any_args
  • Harry Lascelles (@hlascelles) : Fix error when resque-spec is disabled

Copyright

Copyright (c) 2010-2015 Les Hill. See LICENSE for details.