rspec users - Feb 2012 - Can someone clarify this RSpec block?

I''m new to RSpec, Rails, and Ruby. I''m following the Rails
Tutorial. I was
going over some RSpec code and go wanted some clarification:

describe "User pages" do  
  subject { page }

  describe "sign up" do
   describe "with valid information" do
    before do
     fill_in "Name",         with: "Mickey Mouse"
     fill_in "Email",        with: "mickey at disney.com"
     fill_in "Password",     with: "m1ck3y"
     fill_in "Confirmation", with: "m1ck3y"
    end

    it "should create a user" do # block 1
     expect { click_button "Sign up" }.to change(User, :count).by(1)
    end

    describe "after saving the user" do #block 2
     before { click_button ''Sign up'' }
     let(:user) { User.find_by_email(''mickey at disney.com'') }
     it { should have_selector(''title'', text: user.name) }
     it { should have_link(''Sign out'') }
    end
  end

Notice the "it ''should create a user''" block. Why
doesn''t the user we
create in that block carry over to the next block? ("after saving the 
user")? Why is it necessary to write

before { click_button ''Sign up'' } in order to test for the
user? Shouldn''t
the user already exist as it was created in block 2?

I''m looking to understand the flow of events in RSpec.

Thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://rubyforge.org/pipermail/rspec-users/attachments/20120223/32adfa90/attachment.html>

On Feb 23, 2012, at 4:38 PM, Mohamad El-Husseini wrote:
> I''m new to RSpec, Rails, and Ruby. I''m following the
Rails Tutorial. I was going over some RSpec code and go wanted some
clarification:
> 
> describe "User pages" do  
>   subject { page }
> 
>   describe "sign up" do
>   	describe "with valid information" do
>   	  before do
>   	  	fill_in "Name",         with: "Mickey Mouse"
>   	  	fill_in "Email",        with: "mickey at
disney.com"
>   	  	fill_in "Password",     with: "m1ck3y"
>   	  	fill_in "Confirmation", with: "m1ck3y"
>   	  end
> 
>   	  it "should create a user" do # block 1
>   	  	expect { click_button "Sign up" }.to change(User,
:count).by(1)
>   	  end
> 
>   	  describe "after saving the user" do #block 2
>   	  	before { click_button ''Sign up'' }
>   	  	let(:user) { User.find_by_email(''mickey at
disney.com'') }
>   	  	it { should have_selector(''title'', text: user.name)
}
>   	  	it { should have_link(''Sign out'') }
>   	  end
>   	end
> 
> Notice the "it ''should create a user''" block.
Why doesn''t the user we create in that block carry over to the next
block? ("after saving the user")? Why is it necessary to write
> 
> before { click_button ''Sign up'' } in order to test for
the user? Shouldn''t the user already exist as it was created in block
2?
> 
> I''m looking to understand the flow of events in RSpec.
> 
> Thanks!
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users
Each example (`it`) is run in isolation from the other examples -- there is no
shared state. This keeps things clean and easy to understand. In your example,
just add a method for "signing up":

describe "User pages" do  
  subject { page }

  describe "sign up" do
    describe "with valid information" do
      before do
        fill_in "Name",         with: "Mickey Mouse"
        fill_in "Email",        with: "mickey at disney.com"
        fill_in "Password",     with: "m1ck3y"
        fill_in "Confirmation", with: "m1ck3y"
      end

      def click_sign_up_button
        click_button "Sign up" 
      end

      it "should create a user" do # block 1
        expect { click_sign_up_button }.to change(User, :count).by(1)
      end

      describe "after saving the user" do #block 2
        before { click_sign_up_button }
        let(:user) { User.find_by_email(''mickey at
disney.com'') }
        it { should have_selector(''title'', text: user.name) }
        it { should have_link(''Sign out'') }
      end
    end

Each "it" runs in isolation. The flow is, all the "before"
blocks run in order from outer to inner, before each "it" block.
You can think of each "it" expectation as sending you to the start, to
test another isolated expectation about the process.
This is good, because a failure in one part of the system won''t
necessarily break tests for another part.

 -- matt

On Feb 23, 2012, at 3:38 PM, Mohamad El-Husseini <husseini.mel at
gmail.com> wrote:
> I''m new to RSpec, Rails, and Ruby. I''m following the
Rails Tutorial. I was going over some RSpec code and go wanted some
clarification:
> 
> describe "User pages" do  
>   subject { page }
> 
>   describe "sign up" do
>   	describe "with valid information" do
>   	  before do
>   	  	fill_in "Name",         with: "Mickey Mouse"
>   	  	fill_in "Email",        with: "mickey at
disney.com"
>   	  	fill_in "Password",     with: "m1ck3y"
>   	  	fill_in "Confirmation", with: "m1ck3y"
>   	  end
> 
>   	  it "should create a user" do # block 1
>   	  	expect { click_button "Sign up" }.to change(User,
:count).by(1)
>   	  end
> 
>   	  describe "after saving the user" do #block 2
>   	  	before { click_button ''Sign up'' }
>   	  	let(:user) { User.find_by_email(''mickey at
disney.com'') }
>   	  	it { should have_selector(''title'', text: user.name)
}
>   	  	it { should have_link(''Sign out'') }
>   	  end
>   	end
> 
> Notice the "it ''should create a user''" block.
Why doesn''t the user we create in that block carry over to the next
block? ("after saving the user")? Why is it necessary to write
> 
> before { click_button ''Sign up'' } in order to test for
the user? Shouldn''t the user already exist as it was created in block
2?
> 
> I''m looking to understand the flow of events in RSpec.
> 
> Thanks!
> _______________________________________________
> rspec-users mailing list
> rspec-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rspec-users