Redefine a method for unit tests - ruby

I'm not so string in ruby syntax, but if I understand something when learning it, you can do everything, no restrictions, lets say I got a class method, and I want to redefine it in a unit test to isolate the stuff I'm testing, how do i do that? lets say class name is foo and instance method is bar.

Test::Redef is a gem designed to do exactly this.
test "some method on Foo should be called because under normal circumstances it something useful like send network traffic" do
Test::Redef.rd(
'Foo#bar' => proc {|args| puts 'new behavior goes here!'},
'Foo#baz' => :empty, # same as => proc { }
'Foo#quux' => :wiretap, # no behavior change but you get invocation tracking
'Foo.blat' => :empty, # class method
) do |rd|
# code under test
assert_equal [['invocation 1 arg 1', 'invocation 1 arg 2'],
['invocation 2 arg 1'],
], rd[:bar].args, 'Foo#bar was called with the right arguments'
end
# now all methods are back to normal
end

Related

Uniqueness of symbols in object methods

Given that symbols are unique, how is it possible that the two instances of:test below, which refer to different methods (in different classes), have the same object ID?
class Dope
def test
end
end
class Green
def test
end
end
green = Green.new
dope = Dope.new
green.methods.include?(:test) # => true
dope.methods.include?(:test) # => true
green.test.object_id # => 8
dope.test.object_id # => 8
How can you have two symbols with the same name/object ID that refer to different methods?
Both of your test methods return nil. So the object id that you get is the object id of nil:
nil.object_id
# => 8
A symbol is stored differently than a string. Any existence of :test is always going to equal :test because symbols are immutable, unlike strings. There are more descriptions out there here or here.
The :test used to determine the presence of that method will be the same symbol :test that you store in a different class for params[:test] = "something.
Another thing to note that when you call a method on a ruby object, you are technically sending a message to that class. So you're example shows that they both respond to the message :test. Which will call the method on the class. This is accurate.

How can I mock something that "does not implement" a particular method?

The Background:
I'm trying to use cucumber to do some test-driven (or behavior-driven) development around an interface to AWS, in ruby.
So, I have a step definition that looks like this:
Then(/^the mock object should have had :(.*?) called, setting "(.*?)" to "(.*?)"$/) do |method, param, value|
expect(#mock).to receive(method.to_sym).with(hash_including(param, value))
end
Where #mock was previously set using:
#mock = instance_double(AWS::AutoScaling::Client)
And where I invoke this step definition with a feature line like:
And the mock object should have had :update_auto_scaling_group called, setting "auto_scaling_group_name" to "Some-test-value"
When that step gets run, it gets the following error (leaving out the full error, as I believe this is the most relevant part):
AWS::AutoScaling::Client does not implement: update_auto_scaling_group (RSpec::Mocks::MockExpectationError)
I see that indeed, the checks that RSpec runs (as traced back from where the RSpec::Mocks::MockExpectationError gets thrown) are at least correctly reporting the information that they get from the class:
[1] pry(main)> require 'aws-sdk'
=> true
[2] pry(main)> klass = AWS::AutoScaling::Client
=> AWS::AutoScaling::Client
[3] pry(main)> klass.public_method_defined? "update_auto_scaling_group"
=> false
[4] pry(main)> klass.private_method_defined? "update_auto_scaling_group"
=> false
[5] pry(main)> klass.protected_method_defined? "update_auto_scaling_group"
=> false
And yet, if we ask an actual instance, it lets us know that this is a method it would respond to:
[6] pry(main)> x = klass.new
=> #<AWS::AutoScaling::Client::V20110101>
[7] pry(main)> x.respond_to? "update_auto_scaling_group"
=> true
Even while it doesn't say that about just anything:
[8] pry(main)> x.respond_to? "bogus"
=> false
First questions:
So... is this a bug in the AWS::AutoScaling::Client code (or really, probably here), for not defining the methods in a way that the extant checks ({public,private,protected}_method_defined?) would come back true?
Or perhaps a bug in RSpec's "doubles", for not doing all the checking it could do to try to find out that this is indeed a method that's callable in an instance of that class?
Or perhaps it's simply something that I'm doing wrong here? Other?
More generally:
How can I write tests for the code I'm writing, to ensure that it's making calls to what will be an AWS::AutoScaling::Client instance, with the correct parameters (as defined in several checks that I have)? Are there alternate ways I can write my step definitions that would make this work? Alternative ways to create my mock objects? Other?
I've found a way to dynamically mix in the methods I needed to mock
You could do this with empty methods and then stub them, or just include the stubs in the mixin
require 'rails_helper'
RSpec.describe "users/sessions/new.html.erb", :type => :view do
it "displays login form" do
module DeviseUserBits
def resource
#_DeviseUserBitsUser ||= User.new
end
def resource_name
:user
end
def devise_mapping
Devise.mappings[:user]
end
end
view.class.include DeviseUserBits
render
expect(rendered).to match /form/
end
end
It just adds methods on/after instantiating. It's pretty legal, all ruby classes/objects are open.
Proper answer - you do not want to test what you are trying to test in duck-typed language with open classes and objects. It just does not make sense.
The version 1 AWS SDK for Ruby uses #method_missing as a delegate for building and sending requests. The methods a client responds to are defined in an API definition. This eliminates boiler-plate code, but causes problems if you are trying to reflect the available methods at runtime.
Option A: Use a regular double and apply your assertions on the test double.
Option B: Use the mocking feature of the SDK via AWS.stub! When stubbing is enabled, all clients constructed will respond to their regular methods, but will return dummy responses (empty hashes and arrays). This approach provides the useful ability to specify the data to return from a stub. You can even create a stub response for the express purpose of returning from an assertion.
Going with Option B:
# use `:stub_requests` or call Aws.stub!
as = AWS::AutoScaling::Client.new(:stub_requests: true)
# validates parameters as normal, but returns empty response data
as.update_auto_scaling_group(auto_scaling_group_name: 'name')
#=> {}
# You can access the stub response for any operation by name:
stub = as.stub_for(:describe_auto_scaling_groups)
stub.data[:auto_scaling_group_names] = ["Group1", "Group2"]
# Now calling that operation will return the stubbed data
resp = as.describe_auto_scaling_groups
resp.auto_scaling_group_names
#=> ['Group1', 'Group2']
If you need to assert a method is called against the client, you can do so normally, returning the stubbed response:
expect(#client).to receive(:describe_auto_scaling_groups).
with(hash_including(param, value)).
and_return(#client.stub_for(:describe_auto_scaling_groups))

How to use yield self within Rspec

This is a description of how to create a helper method in Rspec taken from the Rspec book (page 149). This example assumes that there is a method called 'set_status' which is triggered when the 'Thing' object is created.
Both sets of code create a new 'Thing' object, set the status, then do 'fancy_stuff'. The first set of code is perfect clear to me. One of the 'it' statements it triggered, which then calls the 'create_thing' method with options. A new 'Thing' object is created and the 'set_status' method is called with the 'options' attribute as the parameter.
The second set of code is similar. One of the 'it' statements is triggered, which then calls the 'given_thing_with' method while passing ':status' hash assignment as a parameter. Within the 'given_thing_with' method the 'yield' is triggered taking the 'Thing.new' as a parameter. This is where I am having trouble. When I try to run this code I get an error of "block given to yield". I understand that whatever attributes that are passed by yield will be returned to the 'thing' in pipe brace from the 'it' statement that called the 'given_thing_with' method. I can get the new
What I don't understand is why the code block is not called in the 'given_thing_with' method after the 'yield' command. In other words, I can't code in that block to run.
Thanks in advance for your help.
The remainder of this question is quoted directly from the Rspec book:
describe Thing do
def create_thing(options)
thing = Thing.new
thing.set_status(options[:status])
thing
end
it "should do something when ok" do
thing = create_thing(:status => 'ok')
thing.do_fancy_stuff(1, true, :move => 'left', :obstacles => nil)
...
end
it "should do something else when not so good" do
thing = create_thing(:status => 'not so good')
thing.do_fancy_stuff(1, true, :move => 'left', :obstacles => nil)
...
end
end
One idiom you can apply to clean this up even more is to yield self from initializers in your objects. Assuming that Thing's initialize() method does this and set_status() does as well, you can write the previous like this:
describe Thing do
def given_thing_with(options)
yield Thing.new do |thing|
thing.set_status(options[:status])
end
end
it "should do something when ok" do
given_thing_with(:status => 'ok') do |thing|
thing.do_fancy_stuff(1, true, :move => 'left', :obstacles => nil)
...
end
end
it "should do something else when not so good" do
given_thing_with(:status => 'not so good') do |thing|
thing.do_fancy_stuff(1, true, :move => 'left', :obstacles => nil)
...
end
end
end
The example in the book is a bit confusing because the implementation of Thing is not shown. To make this work you need to write Thing like so:
class Thing
def initialize
yield self
end
end
When given_thing_with is called it yields a new Thing, which will yield itself when it is constructed. This means that when the inner code block (the one containing thing.set_status) is executed it will have a reference to he newly built Thing.
There are 2 issues with the code from book.
1. Setting up the initializer to yield itself
When the Thing object is created, it needs an initializer and need yield itself.
class Thing
def initialize
yield self
end
end
However, this alone will still causes an error, at least on my system, which is Ruby 1.9.3. Specifically, the error is 'block given to yield (SyntaxError)'. This doesn't make much sense, since that is what we want it to do. Regarless, that is the error I get.
2. Fixing the 'block given to yield' error
This is not as obvious and has something to do with either Ruby or the 'yield' statement, but creating a block using 'do...end' as was written in the book and is shown below causes the error.
yield Thing.new do |thing|
thing.set_status(options[:status])
end
Fixing this error is simlpy a matter of creating the block using braces, '{...}', as is shown below.
yield Thing.new { |thing|
thing.set_status(options[:status])
}
This is not good form for multiline Ruby code, but it works.
Extra. How the series of yields works to set the parameters of the 'Thing' object
The problem is already fixed, but this explains how it works.
the "caller block" calls 'given_thing_with' method with a parameter
that method yields back to the "caller block" a new "Thing" and a block (I'll call it the "yield block")
to execute the "yield block", the Thing class needs the initialization and 'yield self', otherwise the 'set_status' method will never be run because the block will be ignored
the new "Thing" is already in the "caller block" and has it's status set and now the relevant method is executed

Why is the stubbed method not called?

It seems I understood something wrong. I have a class
module Spree
class OmnikassaPaymentResponse
#...
# Finds a payment with provided parameters trough ActiveRecord.
def payment(state = :processing)
Spree::Payment.find(:first, :conditions => { :amount => #amount, :order_id => #order_id, :state => state } ) || raise(ActiveRecord::RecordNotFound)
end
end
end
Which is specced in Rspec:
describe "#payment" do
it 'should try to find a Spree::Payment' do
Spree::Payment.any_instance.stub(:find).and_return(Spree::Payment.new)
Spree::Payment.any_instance.should_receive(:find)
Spree::OmnikassaPaymentResponse.new(#seal, #data).payment
end
end
This, however, always throws ActiveRecord::RecordNotFound. I expected any_instance.stub(:find).and_return() to make sure that whenever, wherever I call a #find on whatever instance I happen to have of Spree::Payment, it returns something.
In other words: I would expect the stub.and_return would avoid getting to || raise(ActiveRecord::RecordNotFound). But it does not.
Is my assumption wrong, my code? Something else?
In your case find is not an instance method, but a class method of Spree::Payment. That means you should stub it directly without any_instance like that:
Spree::Payment.stub(:find).and_return(Spree::Payment.new)

Is it possible to ask Factory Girl what associations a given factory has?

Factory Girl is incredibly useful for functional testing, but has one annoying property that makes it slightly harder to use in unit tests, where I don't want to rely on the test database. I often use Factory.build to create a factory that I can then pass around or assign to an ActiveRecord.find call using flexmock:
require 'test_helper'
require 'flexmock'
class SomeMixinTest < ActiveSupport::TestCase
include FlexMock::TestCase
def setup
#foo = Factory.build(:foo, :id => 123,
:bar => Factory.build(:bar, :id => 456,
:baz => Factory.build(:baz, :id => 789)
)
)
flexmock Foo, :find => #foo
end
def test_find_by_reverse_id
assert_equal #foo, Foo.find_by_reverse_id(321)
end
end
This pattern is very nice, since it cares not about the presence of the database, and runs much faster than if the objects had to actually be persisted. However, it is a bit annoying to have to build the associated objects manually. If you don't, the associated objects are actually created in the database by the build call, as if you had used create instead.
assert_equal [], Foo.all
foo = Factory.build :foo # build my associations too, please
assert_equal [], Foo.all # look Ma, no mocks!
assert_equal [], Bar.all # <=== ASSERTION FAILED
assert_equal [], Baz.all
This is non-intuitive to say the least, and causes an actual problem when I'm testing a few classes that need to play nicely with a mixin. I want to be able to do this:
KLASSES_UNDER_TEST = [Foo, Bar, Baz]
def test_find_by_reverse_id
KLASSES_UNDER_TEST.each do |klass|
objects = (123..456).map do |id|
Factory.build klass.to_s.downcase.to_sym, :id => id
end
flexmock klass, :all => objects
objects.each do |object|
assert_equal object, klass.find_by_reverse_id(object.id.to_s.reverse), "#{klass} #{object.id}"
end
end
But this has the nasty side effect of creating 333 Bars and 666 Bazes ("Baz" does sound kind of like a demon's nickname, so maybe that's fitting) in the database, making this test slower than molasses flowing uphill in the winter.
I'd like to create a helper method like this:
def setup_mocks(klass)
klass_sym = klass.to_s.downcase.to_sym
objects = (123..456).map{|id|
associated_objects = Hash[
Factory.associations(klass_sym).map do |association|
[ association, setup_mocks(association, 1) ]
end
]
Factory.build klass_sym, associated_objects
end
flexmock klass, :all => objects
objects
end
So, does anything like Factory.associations exist?
I've not tested this, but looking at the source it seems that something like this should work:
FactoryGirl.find(:factory_name).associations

Resources