rspec 1.3.0 → 2.0.0

data/Upgrade.rdoc

@@ -1,260 +0,0 @@
-= Upgrade to 1.3.0
-
-== What's changed
-
-=== +be_true+ and +be_false+
-
-These methods now align with Ruby's conditional semantics:
-
-* +be_false+ passes for +false+ and +nil+
-* +be_true+ passes for everything else
-
-This is a change from the previous behaviour (which wasn't working correctly
-anyway), which was supposed to treat +be_true+ as <tt>equal(true)</tt>, and +be_false+ as
-<tt>equal(false)</tt>. 
-
-If the actual values +true+ and +false+ are meaningful to your examples, you'll
-want to change the to use <tt>equal(true)</tt> and <tt>equal(false)</tt> (or
-<tt>==(true)</tt> and <tt>==(false)</tt>).
-
-=== +raise_exception+
-
-We changed the +raise_error+ matcher to +raise_exception+, and aliased it with
-+raise_error+. This maintains backward compatibility, and also gives you the
-option of being more precise in specs about what is expected.
-
-=== Matcher DSL
-
-==== +match+ rescues from +ExpectationNotMetErrror+ by default
-
-This allows you to wrap other expectations in the +match+ method. Consider
-this matcher:
-
-    Spec::Matchers.define :teach do |subject|
-      match do |teacher|
-        teacher.subjects.should include(subject)
-      end
-    end
-
-The block passed to +match+ is called internally by Rspec, and is expected to
-return a boolean value. In this case, if <tt>should include(subject)</tt>
-fails, it raises an +ExpectationNotMetError+.
-
-Beginning with rspec-1.3.0, when the match block raises an
-+ExpectationNotMetError+, it is captured and the block returns +false+.
-Otherwise it returns +true+, so it works like any other matcher.
-
-==== match_unless_raises
-
-The new +match_unless_raises+ method allows you to wrap <tt>Test::Unit</tt>
-assertions by capturing +AssertionFailedError+ and returning false, just as the
-+match+ method now does with +ExpectationNotMetError+.
-
-    Spec::Matchers.define :teach do |subject|
-      match_unless_raises Test::Unit::AssertionFailedError do |teacher|
-        assert teacher.subjects.include?(subject)
-      end
-    end
-
-= Upgrade to rspec-1.2.9
-
-== What's new
-
-=== spec/spec.opts
-
-If you have a spec/spec.opts file, the spec command will now use that
-automatically as long as you don't include any options on the command line.
-
-=== let()
-
-Writing specs tends to follow a regular pattern of using local variables,
-discovering duplication, and then having to convert to local variables to
-instance variables by adding an "@" symbol. The let() method assigns the result
-of a lazy eval'd block as the return value of an instance method using the same
-name. This way you can go from this:
-
-    describe Subscription do
-      it "does something" do
-        subscription = Subscription.create :limit => 1
-        subscription...
-      end
-
-      it "does something else" do
-        subscription = Subscription.create :limit => 1
-        subscription...
-      end
-    end
-
-to this:
-
-    describe Subscription do
-      let(:subscription) { Subscription.create :limit => 1 }
-
-      it "does something" do
-        subscription...
-      end
-
-      it "does something else" do
-        subscription...
-      end
-    end
-
-=== its()
-
-If you're in the habit of writing one-liners using implicit subject, this new
-its() feature is for you. Here's the basic idea:
-
-    describe Array do
-      its(:length) { should == 0 }
-    end
-
-= Upgrade to rspec-1.2.3-1.2.7
-
-== What's Changed
-
-=== Matcher DSL
-
-Use Spec::Matchers.define instead of Spec::Matchers.create (which is now
-deprecated).
-
-=== Explicit Predicate Matchers are deprecated
-
-With the addition of the new Matcher DSL the old, confusing, and
-almost-nobody-uses-it explicit predicate matcher functionality's days are now
-numbered.
-
-If you're not familiar with this feature, don't worry about it. If you have anything
-that looks like this:
-
-    predicate_matchers[:swim] = :can_swim?
-
-Or this
-
-    config.predicate_matchers[:swim] = :can_swim?
-
-Change it to this:
-
-    Spec::Matchers.define :swim do
-      match do |potential_swimmer|
-        potential_swimmer.can_swim?
-      end
-    end
-
-== Custom Formatters
-
-If you have an custom formatter, the <tt>add_example_group</tt> method has
-been changed to <tt>example_group_started</tt>, and kept as an alias so your
-formatters will still work. Though not yet, <tt>add_example_group</tt> will be
-deprecated in a future minor release, and removed in a future major release,
-so we recommend you make this change now.
-
-= Upgrade to rspec-1.2.2
-
-== What's Changed
-
-=== <tt>require 'rubygems' unless ENV['NO_RUBYGEMS']</tt>
-
-After minor public outcry and confusion, we restored necessary references to
-rubygems in rspec. If you use a different mechanism for managing gems, just
-set a <tt>NO_RUBYGEMS</tt> environment variable (to any non-nil value).
-
-=== Proxies and locations
-
-This is probably only interesting to you if you use custom formatters.
-
-Formatters now receive Spec::Example::ExampleGroupProxy and
-Spec::Example::ExampleGroup objects with cohesive APIs for reporting. See the
-RDoc for those classes and Spec::Runner::Formatter::BaseFormatter for more
-information.
-
-== What's new
-
-=== The new matcher DSL works with test/unit (without the rest of rspec)
-
-We'll be separating this out to its own gem for rspec 2.0, but for now, just install
-rspec >= 1.2.1 and add the following to your <tt>test_helper</tt> file:
-
-    require 'spec/expectations'
-    class Test::Unit::TestCase
-      include Spec::Matchers
-    end
-
-This will add <tt>should()</tt> and <tt>should_not()</tt> to your objects, make all of
-rspec's built-in matchers available to your tests, INCLUDING rspec's DSL for
-creating matchers (see below, under Upgrade to rspec-1.2.0)
-
-=== debugger
-
-If you have ruby-debug installed, you can set a breakpoint by adding <tt>debugger()</tt>
-in your code:
-
-    # some code .....
-    debugger
-    # some more code ....
-
-... and using the <tt>--debugger</tt> or <tt>-u</tt> command line option.
-
-    spec path/to/file.rb --debugger
-
-= Upgrade to rspec-1.2.0
-
-== What's Changed
-
-=== WARNINGS
-
-* If you use the ruby command to run specs instead of the spec command, you'll
-  need to require 'spec/autorun' or they won't run. This won't affect you if
-  you use the spec command or the Spec::Rake::SpecTask that ships with RSpec.
-
-* require 'spec/test/unit' to invoke test/unit interop if you're using
-  RSpec's core (this is handled implicitly with spec-rails)
-
-* setup and teardown are gone - use before and after instead
-
-  * you can still use setup and teardown if you're using
-    Test::Unit::TestCase as the base ExampleGroup class (which is implicit
-    in rspec-rails)
-
-* The matcher protocol has been improved. The old protocol is still supported,
-  but we added support for two new methods that speak a bit more clearly:
-
-    failure_message          => failure_message_for_should
-    negative_failure_message => failure_message_for_should_not
-
-* All references to rubygems have been removed from within rspec's code.
-
-  * See http://gist.github.com/54177 for rationale and suggestions on
-    alternative approaches to loading rubygems
-
-== What's New
-
-=== Ruby 1.9
-
-RSpec now works with Ruby 1.9.1. See http://wiki.github.com/dchelimsky/rspec/ruby-191
-for useful information.
-
-=== Improved heckle integration
-
-RSpec works with heckle again! Gotta use heckle >= 1.4.2 for this to work
-though, and it only works with ruby-1.8.6 and 1.8.7 (heckle doesn't support
-1.9.1 yet).
-
-    [sudo] gem install heckle --version ">=1.4.2"
-    spec spec/game/mastermind.rb --heckle Game::Mastermind
-
-=== New Matcher DSL
-
-We've added a new DSL for generating custom matchers very simply and cleanly.
-We'll still support the simple_matcher method, so never fear if you're using
-that, but we recommend that you start developing your new matchers with this
-new syntax.
-
-    Spec::Matchers.create :be_a_multiple_of do |smaller|
-      match do |bigger|
-        bigger % smaller == 0
-      end
-    end
-
-    9.should be_a_multiple_of(3)
-
-See <tt>features/matchers/define_matcher.feature</tt> for more examples

data/bin/autospec

@@ -1,4 +0,0 @@
-#!/usr/bin/env ruby
-ENV['RSPEC'] = 'true'     # allows autotest to discover rspec
-ENV['AUTOTEST'] = 'true'  # allows autotest to run w/ color on linux
-system (RUBY_PLATFORM =~ /mswin|mingw/ ? 'autotest.bat' : 'autotest'), *ARGV

data/bin/spec

@@ -1,5 +0,0 @@
-#!/usr/bin/env ruby
-rspec_dir = File.expand_path(File.join(File.dirname(__FILE__), '..', 'lib'))
-$LOAD_PATH.unshift(rspec_dir) unless $LOAD_PATH.include?(rspec_dir)
-require 'spec/autorun'
-exit ::Spec::Runner::CommandLine.run

data/cucumber.yml

@@ -1,5 +0,0 @@
-default: --require features features -t ~@wip
-pending: --require features features-pending
-no_heckle: --require features features --exclude heckle
-focus: --require features features -t focus
-wip: --require features features -t @wip

data/examples/failing/README.txt

@@ -1,11 +0,0 @@
-"Why have failing examples?", you might ask.
-
-They allow us to see failure messages. RSpec wants to provide meaningful and
-helpful failure messages. The failures in this directory not only provide you
-a way of seeing the failure messages, but they provide RSpec's own specs a way
-of describing what they should look like and ensuring they stay correct.
-
-To see the types of messages you can expect, stand in the root directory and
-run:
-
-bin/spec examples/failing/*.rb

data/examples/failing/diffing_spec.rb

@@ -1,36 +0,0 @@
-describe "Running specs with --diff" do
-  it "should print diff of different strings" do
-    uk = <<-EOF
-RSpec is a
-behaviour driven development
-framework for Ruby
-EOF
-    usa = <<-EOF
-RSpec is a
-behavior driven development
-framework for Ruby
-EOF
-    usa.should == uk
-  end
-
-  class Animal
-    def initialize(name,species)
-      @name,@species = name,species
-    end
-
-    def inspect
-      <<-EOA
-<Animal
-name=#{@name},
-species=#{@species}
->
-      EOA
-    end
-  end
-
-  it "should print diff of different objects' pretty representation" do
-    expected = Animal.new "bob", "giraffe"
-    actual   = Animal.new "bob", "tortoise"
-    expected.should eql(actual)
-  end
-end

data/examples/failing/failing_implicit_docstrings_example.rb

@@ -1,17 +0,0 @@
-# Run spec w/ -fs to see the output of this file
-
-describe "Failing examples with no descriptions" do
-
-  # description is auto-generated as "should equal(5)" based on the last #should
-  it do
-    3.should equal(2)
-    5.should equal(5)
-  end
-
-  it { 3.should be > 5 }
-
-  it { ["a"].should include("b") }
-
-  it { [1,2,3].should_not respond_to(:size) }
-
-end

data/examples/failing/failure_in_after.rb

@@ -1,10 +0,0 @@
-describe "This example" do
-  
-  it "should be listed as failing in after" do
-  end
-  
-  after(:each) do
-    NonExistentClass.new
-  end
-  
-end

data/examples/failing/failure_in_before.rb

@@ -1,10 +0,0 @@
-describe "This example" do
-  
-  before(:each) do
-    NonExistentClass.new
-  end
-  
-  it "should be listed as failing in each" do
-  end
-  
-end

data/examples/failing/mocking_example.rb

@@ -1,38 +0,0 @@
-describe "Mocker" do
-
-  it "should be able to call mock()" do
-    mock = mock("poke me")
-    mock.should_receive(:poke)
-    mock.poke
-  end
-
-  it "should fail when expected message not received" do
-    mock = mock("poke me")
-    mock.should_receive(:poke)
-  end
-  
-  it "should fail when messages are received out of order" do
-    mock = mock("one two three")
-    mock.should_receive(:one).ordered
-    mock.should_receive(:two).ordered
-    mock.should_receive(:three).ordered
-    mock.one
-    mock.three
-    mock.two
-  end
-
-  it "should get yelled at when sending unexpected messages" do
-    mock = mock("don't talk to me")
-    mock.should_not_receive(:any_message_at_all)
-    mock.any_message_at_all
-  end
-
-  it "has a bug we need to fix" do
-    pending "here is the bug" do
-      # Actually, no. It's fixed. This will fail because it passes :-)
-      mock = mock("Bug")
-      mock.should_receive(:hello)
-      mock.hello
-    end
-  end
-end

data/examples/failing/mocking_with_flexmock.rb

@@ -1,26 +0,0 @@
-# stub frameworks like to gum up Object, so this is deliberately
-# set NOT to run so that you don't accidentally run it when you
-# run this dir.
-
-# To run it, stand in this directory and say:
-#
-#   RUN_FLEXMOCK_EXAMPLE=true ruby ../bin/spec mocking_with_flexmock.rb
-
-if ENV['RUN_FLEXMOCK_EXAMPLE']
-  Spec::Runner.configure do |config|
-    config.mock_with :flexmock
-  end
-
-  describe "Flexmocks" do
-    it "should fail when the expected message is received with wrong arguments" do
-      m = flexmock("now flex!")
-      m.should_receive(:msg).with("arg").once
-      m.msg("other arg")
-    end
-
-    it "should fail when the expected message is not received at all" do
-      m = flexmock("now flex!")
-      m.should_receive(:msg).with("arg").once
-    end
-  end
-end

data/examples/failing/mocking_with_mocha.rb

@@ -1,25 +0,0 @@
-# stub frameworks like to gum up Object, so this is deliberately
-# set NOT to run so that you don't accidentally run it when you
-# run this dir.
-
-# To run it, stand in this directory and say:
-#
-#   RUN_MOCHA_EXAMPLE=true ruby ../bin/spec mocking_with_mocha.rb
-
-if ENV['RUN_MOCHA_EXAMPLE']
-  Spec::Runner.configure do |config|
-    config.mock_with :mocha
-  end
-  describe "Mocha framework" do
-    it "should should be made available by saying config.mock_with :mocha" do
-      m = mock()
-      m.expects(:msg).with("arg")
-      m.msg
-    end
-    it "should should be made available by saying config.mock_with :mocha" do
-      o = Object.new
-      o.expects(:msg).with("arg")
-      o.msg
-    end
-  end
-end

data/examples/failing/mocking_with_rr.rb

@@ -1,27 +0,0 @@
-# stub frameworks like to gum up Object, so this is deliberately
-# set NOT to run so that you don't accidentally run it when you
-# run this dir.
-
-# To run it, stand in this directory and say:
-#
-#   RUN_RR_EXAMPLE=true ruby ../bin/spec mocking_with_rr.rb
-
-if ENV['RUN_RR_EXAMPLE']
-  Spec::Runner.configure do |config|
-    config.mock_with :rr
-  end
-  describe "RR framework" do
-    it "should should be made available by saying config.mock_with :rr" do
-      o = Object.new
-      mock(o).msg("arg")
-      o.msg
-    end
-    it "should should be made available by saying config.mock_with :rr" do
-      o = Object.new
-      mock(o) do |m|
-        m.msg("arg")
-      end
-      o.msg
-    end
-  end
-end

data/examples/failing/partial_mock_example.rb

@@ -1,18 +0,0 @@
-class MockableClass
-  def self.find id
-    return :original_return
-  end
-end
-
-describe "A partial mock" do
-
-  it "should work at the class level (but fail here due to the type mismatch)" do
-    MockableClass.should_receive(:find).with(1).and_return {:stub_return}
-    MockableClass.find("1").should equal(:stub_return)
-  end
-
-  it "should revert to the original after each spec" do
-    MockableClass.find(1).should equal(:original_return)
-  end
-
-end

data/examples/failing/pending_example.rb

@@ -1,7 +0,0 @@
-describe "pending example (which is fixed)" do
-  it %Q|reports "FIXED ... Expected ... to fail.  No Error was raised."| do
-    pending("for some reason") do
-      # success
-    end
-  end
-end

data/examples/failing/predicate_example.rb

@@ -1,32 +0,0 @@
-class BddFramework
-  def intuitive?
-    true
-  end
-
-  def adopted_quickly?
-    #this will cause failures because it reallly SHOULD be adopted quickly
-    false
-  end
-end
-
-describe "BDD framework" do
-
-  before(:each) do
-    @bdd_framework = BddFramework.new
-  end
-
-  it "should be adopted quickly" do
-    #this will fail because it reallly SHOULD be adopted quickly
-    @bdd_framework.should be_adopted_quickly
-  end
-
-  it "should be intuitive" do
-    @bdd_framework.should be_intuitive
-  end
-
-  it "should not respond to test" do
-    #this will fail
-    @bdd_framework.test
-  end
-
-end

data/examples/failing/raising_example.rb

@@ -1,47 +0,0 @@
-describe "This example" do
-  
-  it "should show that a NoMethodError is raised but an Exception was expected" do
-    proc { ''.nonexistent_method }.should raise_error
-  end
-  
-  it "should pass" do
-    proc { ''.nonexistent_method }.should raise_error(NoMethodError)
-  end
-  
-  it "should show that a NoMethodError is raised but a SyntaxError was expected" do
-    proc { ''.nonexistent_method }.should raise_error(SyntaxError)
-  end
-  
-  it "should show that nothing is raised when SyntaxError was expected" do
-    proc { }.should raise_error(SyntaxError)
-  end
-
-  it "should show that a NoMethodError is raised but a Exception was expected" do
-    proc { ''.nonexistent_method }.should_not raise_error
-  end
-  
-  it "should show that a NoMethodError is raised" do
-    proc { ''.nonexistent_method }.should_not raise_error(NoMethodError)
-  end
-  
-  it "should also pass" do
-    proc { ''.nonexistent_method }.should_not raise_error(SyntaxError)
-  end
-  
-  it "should show that a NoMethodError is raised when nothing expected" do
-    proc { ''.nonexistent_method }.should_not raise_error(Exception)
-  end
-  
-  it "should show that the wrong message was received" do
-    proc { raise StandardError.new("what is an enterprise?") }.should raise_error(StandardError, "not this")
-  end
-  
-  it "should show that the unexpected error/message was thrown" do
-    proc { raise StandardError.new("abc") }.should_not raise_error(StandardError, "abc")
-  end
-  
-  it "should pass too" do
-    proc { raise StandardError.new("abc") }.should_not raise_error(StandardError, "xyz")
-  end
-  
-end

data/examples/failing/syntax_error_example.rb

@@ -1,7 +0,0 @@
-describe "when passing a block to a matcher" do
-  it "you should use {} instead of do/end" do
-    Object.new.should satisfy do
-      "this block is being passed to #should instead of #satisfy - use {} instead"
-    end
-  end
-end

data/examples/failing/team_spec.rb

@@ -1,41 +0,0 @@
-class Team
-  attr_reader :players
-  def initialize
-    @players = Players.new
-  end
-end
-
-class Players
-  def initialize
-    @players = []
-  end
-  def size
-    @players.size
-  end
-  def include? player
-    raise "player must be a string" unless player.is_a?(String)
-    @players.include? player
-  end
-end
-
-describe "A new team" do
-  
-  before(:each) do
-    @team = Team.new
-  end
-  
-  it "should have 3 players (failing example)" do
-    @team.should have(3).players
-  end
-  
-  it "should include some player (failing example)" do
-    @team.players.should include("Some Player")
-  end
-
-  it "should include 5 (failing example)" do
-    @team.players.should include(5)
-  end
-  
-  it "should have no players"
-  
-end

data/examples/failing/timeout_behaviour.rb

@@ -1,5 +0,0 @@
-describe "Something really slow" do
-  it "should be failed by RSpec when it takes longer than --timeout" do
-    sleep(2)
-  end
-end

data/examples/passing/custom_formatter.rb

@@ -1,11 +0,0 @@
-require 'spec/runner/formatter/progress_bar_formatter'
-
-# Example of a formatter with custom bactrace printing. Run me with:
-# ruby bin/spec examples/failing -r examples/passing/custom_formatter.rb -f CustomFormatter
-class CustomFormatter < Spec::Runner::Formatter::ProgressBarFormatter
-  def backtrace_line(line)
-    line.gsub(/([^:]*\.rb):(\d*)/) do
-      "<a href=\"file://#{File.expand_path($1)}\">#{$1}:#{$2}</a> "
-    end
-  end
-end