rspec 0.5.3 → 0.5.4

data/CHANGES

@@ -1,30 +1,55 @@
 = RSpec Changelog
 
+== Version 0.5.4 (in SVN)
+Release notes go here
+
+* Support for running a single spec with --spec
+* Exitcode is now 1 unless all specs pass, in which case it's 0.
+* -v, --version now both mean the same thing
+* For what was verbose output (-v), use --format specdoc or -f s
+* --format rdoc always runs in dry-run mode
+* Removed --doc and added --format and --dry-run
+* Refactored towards more pluggable formatters
+* Use webgen's execute tag when generating website (more accurate)
+* Fixed incorrect quoting of spec_opts in SpecTask
+* Added patch to enable underscored shoulds like 1.should_equal(1) - patch from Rich Kilmer
+* Removed most inherited instance method from Mock, making more methods mockable.
+* Made the RCovVerify task part of the standard toolset.
+* Documented Rake task and how to use it with Rcov
+* Implemented <ruby></ruby> tags for website (hooking into ERB, RedCloth and syntax)
+* RSpec Rake task now takes spec_opts and out params so it can be used for doc generation
+* RCov integration for RSpec Rake task (#4058)
+* Group all results instead of printing them several times (#4057)
+* Mocks can now yield
+* Various improvements to error reporting (including #4191)
+* backtrace excludes rspec code - use -b to include it
+* split examples into examples (passing) and failing_examples
+
 == Version 0.5.3
 The "hurry up, CoR is in two days" release.
 
-* Don't run rcov by default [aslak_hellesoy]
-* Make separate task for running tests with RCov [aslak_hellesoy]
-* Added Rake task to fail build if coverage drops below a certain threshold [aslak_hellesoy]
-* Even more failure output cleanup (simplification) [dchelimsky]
-* Added duck_type constraint for mocks [dchelimsky]
+* Don't run rcov by default
+* Make separate task for running tests with RCov
+* Added Rake task to fail build if coverage drops below a certain threshold
+* Even more failure output cleanup (simplification)
+* Added duck_type constraint for mocks
 
 == Version 0.5.2
 This release has minor improvements to the commandline and fixes some gem warnings
 
-* Readded README to avoid RDoc warnings [aslak_hellesoy]
-* Added --version switch to commandline [aslak_hellesoy]
-* More changes to the mock API [dastels], [dchelimsky]
+* Readded README to avoid RDoc warnings
+* Added --version switch to commandline
+* More changes to the mock API
 
 == Version 0.5.1
 This release is the first release of RSpec with a new website. It will look better soon.
 
-* Added initial documentation for API [dastels]
-* Added website based on webgen [aslak_hellesoy]
-* Modified test task to use rcov [aslak_hellesoy]
-* Deleted unused code (thanks, rcov!) [aslak_hellesoy]
-* Various changes to the mock API [dastels], [dchelimsky]
-* Various improvements to failure reporting [dchelimsky]
+* Added initial documentation for API
+* Added website based on webgen
+* Modified test task to use rcov
+* Deleted unused code (thanks, rcov!)
+* Various changes to the mock API,
+* Various improvements to failure reporting
 
 == Version 0.5.0
 This release introduces a new API and obsolesces previous versions.
@@ -74,23 +99,23 @@ This release improves installation and documentation, mock integration and error
 * Comparison errors now print the class name too.
 * Mocks now take an optional +options+ parameter to specify behaviour.
 * Removed __expects in favour of should_receive
-* Added line number reporting in mock error messages for unreceived message expectations (AH).
-* Added should_match and should_not_match (AH).
-* Added a +mock+ method to Spec::Context which will create mocks that autoverify (no need to call __verify) (AH).
-* Mocks now require names in the constructor to ensure sensible error messages (AH).
-* Made 'spec' executable and updated usage instructions in README accordingly (AH).
-* Made more parts of the Spec::Context API private to avoid accidental usage (AH).
+* Added line number reporting in mock error messages for unreceived message expectations.
+* Added should_match and should_not_match.
+* Added a +mock+ method to Spec::Context which will create mocks that autoverify (no need to call __verify).
+* Mocks now require names in the constructor to ensure sensible error messages.
+* Made 'spec' executable and updated usage instructions in README accordingly.
+* Made more parts of the Spec::Context API private to avoid accidental usage.
 * Added more RDoc to Spec::Context.
 
 == Version 0.1.6
 
 More should methods.
 
-* Added should_match and should_not_match (AH).
+* Added should_match and should_not_match.
 
 == Version 0.1.5
 
-Included examples and tests in gem (AH).
+Included examples and tests in gem.
 
 == Version 0.1.4
 
@@ -106,23 +131,23 @@ Improved mocking:
 
 This release adds some improvements to the mock API and minor syntax improvements
 
-* Added Mock.should_expect for a more consistent DSL (AH).
-* Added MockExpectation.and_returns for a better DSL (AH).
-* Made Mock behave as a null object after a call to Mock.ignore_missing (AH)
-* Internal syntax improvements (AH).
-* Improved exception trace by adding exception class name to error message (AH).
-* Renamed some tests for better consistency (AH).
+* Added Mock.should_expect for a more consistent DSL.
+* Added MockExpectation.and_returns for a better DSL.
+* Made Mock behave as a null object after a call to Mock.ignore_missing
+* Internal syntax improvements.
+* Improved exception trace by adding exception class name to error message.
+* Renamed some tests for better consistency.
 
 == Version 0.1.1
 
 This release adds some shoulds and improves error reporting
 
-* Added should_be_same_as and should_not_be_same_as (AH).
-* Improved error reporting for comparison expectations (AH).
+* Added should_be_same_as and should_not_be_same_as.
+* Improved error reporting for comparison expectations.
 
 == Version 0.1.0
 
 This is the first preview release of RSpec, a Behaviour-Driven Development library for Ruby 
 
-* Added Rake script with tasks for gems, rdoc etc (AH).
-* Added an XForge task to make release go easier (AH).
+* Added Rake script with tasks for gems, rdoc etc.
+* Added an XForge task to make release go easier.

data/Rakefile

@@ -5,10 +5,16 @@ require 'rake/contrib/rubyforgepublisher'
 require 'rake/clean'
 require 'rake/testtask'
 require 'rake/rdoctask'
-require 'lib/spec/version'
-require 'lib/spec/rake/spectask'
+require 'spec/version'
+require 'spec/rake/spectask'
+require 'spec/rake/rcov_verify'
 require 'test/rcov/rcov_testtask'
-require 'test/rcov/rcov_verify'
+
+# Some of the tasks are in separate files since they are also part of the website documentation
+load File.dirname(__FILE__) + '/test/tasks/examples.rake'
+load File.dirname(__FILE__) + '/test/tasks/examples_specdoc.rake'
+load File.dirname(__FILE__) + '/test/tasks/examples_with_rcov.rake'
+load File.dirname(__FILE__) + '/test/tasks/rcov_verify.rake'
 
 PKG_NAME = "rspec"
 # Versioning scheme: MAJOR.MINOR.PATCH
@@ -27,11 +33,11 @@ PKG_FILES = FileList[
   'doc/**/*'
 ]
 
-task :default => [:test] #, :test_text_runner]
+task :default => [:test]
 
-Spec::Rake::SpecTask.new do |t|
-  t.spec_files = FileList['examples/**/*_spec.rb']
-  t.verbose = true
+desc "Run all failing examples"
+Spec::Rake::SpecTask.new('failing_examples') do |t|
+  t.spec_files = FileList['failing_examples/**/*_spec.rb']
 end
 
 Rake::TestTask.new do |t|
@@ -46,7 +52,7 @@ end
 
 desc 'Generate HTML documentation'
 task :doc do
-  sh %{pushd doc; webgen; popd }
+  sh %{pushd doc; webgen; popd}
 end
 
 desc 'Generate RDoc'
@@ -54,8 +60,7 @@ rd = Rake::RDocTask.new("rdoc") do |rdoc|
   rdoc.rdoc_dir = 'doc/output/rdoc'
   rdoc.title    = "RSpec"
   rdoc.options << '--line-numbers' << '--inline-source' << '--main' << 'README'
-  rdoc.rdoc_files.include('README', 'CHANGES')
-  rdoc.rdoc_files.include('lib/**/*.rb', 'doc/**/*.rdoc')
+  rdoc.rdoc_files.include('README', 'CHANGES', 'EXAMPLES.rd', 'lib/**/*.rb')
 end
 
 spec = Gem::Specification.new do |s|
@@ -120,15 +125,10 @@ task :clobber do
   rm_rf 'doc/output'
 end
 
-task :release => [:clobber, :verify_user, :verify_password, :test, :upload_releases, :publish_website, :publish_news]
+task :release => [:clobber, :verify_user, :verify_password, :test, :publish_packages, :publish_website, :publish_news]
 
 desc "Build the website with rdoc and rcov, but do not publish it"
-task :website => [:clobber, :copy_rcov_report, :doc, :rdoc]
-
-RCov::VerifyTask.new do |t|
-  t.threshold = 99.1 # Don't make it lower unless you have a damn good reason.
-  t.index_html = 'coverage/index.html'
-end
+task :website => [:clobber, :copy_rcov_report, :doc, :examples_specdoc, :rdoc]
 
 task :copy_rcov_report => [:test_with_rcov, :rcov_verify] do
   rm_rf 'doc/output/coverage'
@@ -138,6 +138,7 @@ end
 
 task :verify_user do
   raise "RUBYFORGE_USER environment variable not set!" unless ENV['RUBYFORGE_USER']
+  raise "BEHAVIOURDRIVEN_USER environment variable not set!" unless ENV['BEHAVIOURDRIVEN_USER']
 end
 
 task :verify_password do
@@ -147,16 +148,16 @@ end
 desc "Upload Website to RubyForge"
 task :publish_website => [:verify_user, :website] do
   publisher = Rake::SshDirPublisher.new(
-    "#{ENV['RUBYFORGE_USER']}@rubyforge.org",
-    "/var/www/gforge-projects/#{PKG_NAME}",
+    "#{ENV['BEHAVIOURDRIVEN_USER']}@www.behaviourdriven.org",
+    "/var/www/behaviourdriven.org/htdocs/#{PKG_NAME}",
     "doc/output"
   )
 
   publisher.upload
 end
 
-desc "Release gem+tgz+zip to RubyForge. You must make sure lib/version.rb is aligned with the CHANGELOG file"
-task :upload_releases => [:verify_user, :verify_password, :package] do
+desc "Publish gem+tgz+zip on RubyForge. You must make sure lib/version.rb is aligned with the CHANGELOG file"
+task :publish_packages => [:verify_user, :verify_password, :package] do
   require 'meta_project'
   require 'rake/contrib/xforge'
   release_files = FileList[

data/bin/spec

@@ -1,18 +1,16 @@
 #!/usr/bin/env ruby
 
-require 'find'
-require File.dirname(__FILE__) + "/../lib/spec"
+require File.expand_path(File.dirname(__FILE__) + "/../lib/spec") # better stack traces this way
 
-context_runner = Spec::Runner::ContextRunner.new(ARGV)
-Spec::Runner::Context.context_runner = context_runner
+$context_runner = ::Spec::Runner::OptionParser.create_context_runner(ARGV, false, STDERR, STDOUT)
 
-ARGV.each do |file|
-  Find.find(file) do |path|
-    if FileTest.file?(path) and File.extname(path) == '.rb'
-      require path
-    end
-  end
+ARGV.each do |file_or_dir|
+	if File.lstat(file_or_dir).directory? then
+		(Dir.open(file_or_dir).select {|f| f =~ /.*\.rb/}).each {|file| require "#{file_or_dir}/#{file}"}
+	else
+		require file_or_dir
+	end
 end
 
-context_runner.run
+$context_runner.run(true)
 

data/doc/README

@@ -1,5 +1,3 @@
 The website can be generated with
   webgen
-(Install the following gems: webgen, redcloth, bluecloth)
-
-TODO: write rake task that will generate website, put rdoc underneath it and upload everything.
+(Install the following gems: webgen, redcloth, syntax)

data/doc/plugin/syntax.rb

@@ -1,6 +1,29 @@
+# gem install syntax
+require 'syntax/convertors/html'
+RUBY2HTML = Syntax::Convertors::HTML.for_syntax "ruby"
+
+class ERB
+  class Compiler
+    alias old_compile compile
+  
+    def compile(s)
+      s.gsub! /<ruby>/n, "<notextile><%= ruby <<-EOR"
+      s.gsub! /<\/ruby>/n, "EOR\n%></notextile>"
+      s.gsub! /<ruby\s+file="(.*)"\s*\/>/, "{ruby_inline: {filename: \\1}}"
+      old_compile(s)
+    end
+  end
+end
+
+module FileHandlers
+  class PageFileHandler < DefaultFileHandler
+    def ruby(code)
+      RUBY2HTML.convert(code)
+    end
+  end
+end
+
 module Tags
-  # gem install syntax
-  require 'syntax/convertors/html'
 
   # prints out nicely formatted ruby code in html
   class RubyInliner < DefaultTag
@@ -9,11 +32,10 @@ module Tags
     add_param 'filename', nil, 'The File to insert'
     set_mandatory 'filename', true
     
-    CONVERTOR = Syntax::Convertors::HTML.for_syntax "ruby"
     
     def initialize
       super
-      register_tag( 'ruby_inline')
+      register_tag('ruby_inline')
     end
 
     def process_tag(tag, node, ref_node)
@@ -32,7 +54,7 @@ module Tags
       rescue
         self.logger.error { "Given file <#{filename}> does not exist (tag specified in <#{ref_node.recursive_value( 'src' )}>" }
       end
-      CONVERTOR.convert(content)
+      RUBY2HTML.convert(content)
     end
   end
 end

data/doc/src/core_team.page

@@ -0,0 +1,22 @@
+---
+title: The Core Team
+inMenu: true
+---
+
+h2. The Core Team
+
+* Dave Astels
+* Steve Baker
+* David Chelimsky
+* Aslak Hellesoy
+
+<table>
+<tr>
+<td><img src="images/David_and_Aslak.jpg"></td>
+<td><img src="images/Whats_That_Dude.jpg"></td>
+</tr>
+<tr>
+<td align="center">David and Aslak getting ready for 1.0</td>
+<td align="center">What's that, Dude?</td>
+</tr>
+</table>

data/doc/src/default.css

@@ -21,7 +21,7 @@ acronym {
 }
 #container {
 	font-size:1.2em;
-	width:550px;
+	width:94%;
 	margin:0 auto;
 	border-right:3px double #000;
 }
@@ -42,7 +42,7 @@ acronym {
 	color:#c00;
 }
 #header h2 {
-	margin:140px 0 0 0;
+	margin:30px 0 0 0;
 	font-size:1em;
 	font-weight:normal;
 	color:#999;
@@ -52,10 +52,7 @@ acronym {
 }
 #navigation {
 	text-align:right;
-	margin:10px 10px 10px auto;
-}
-#navigation {
-	margin:0 10px 0 0;
+	margin:0 18px 10px auto;
 	padding:0;
 }
 #navigation li {
@@ -76,6 +73,9 @@ acronym {
 	color:#C00;
 	border-bottom:1px solid #ddd;
 }
+#navigation li.webgen-menu-item-selected a, #naviation li.webgen-menu-item-selected a:hover {
+    color: #000;
+}
 #content {
 	margin:20px 0 20px 10px;
 	padding:0;
@@ -98,7 +98,7 @@ acronym {
 }
 #content h2 {
 	display:block;
-	margin:20px 0 0 0px;
+	margin:20px 20px 0 0px;
 	padding:0;
 	font-family:"Trebuchet MS",arial,sans-serif;
 	font-size:1.1em;
@@ -137,9 +137,9 @@ acronym {
 	text-indent:0;
 }
 #content code {
-    font-family:"courier new";
-    color:#444;
-    font-weight:bold;
+    font-family:monospace;
+    color:#222;
+    font-size:1.3em;
 }
 #content pre {
 	padding:10px;
@@ -160,7 +160,7 @@ acronym {
 #footer {
 	font-size:0.9em;
 	font-family:arial,sans-serif;
-	margin:90px 0 20px 0;
+	margin:60px 18px 20px 0;
 	padding:5px;
 	border-top:1px solid #c00;
 }

data/doc/src/default.template

@@ -15,7 +15,6 @@
 	</div>
 
 	<div id="navigation">
-	    <!-- Not sure how to get this in the menu -->
 	    {menu: {menuStyle: horizontal}}
 	</div>
 

data/doc/src/documentation/index.page

@@ -1,13 +1,188 @@
 ---
-title: Writing Specifications
+title: Core API
 inMenu: true
+ordering: 5
 ---
-h2. Writing Specifications
+h2. Core API
 
-An RSpec specification is a Ruby file with a context and one or more specifications. Example:
+When RSpec executes specifications, it defines a method <code>should</code> on every object in the system. This <code>should</code> method is your entry to the magic of RSpec.
 
-{ruby_inline: {filename: ../examples/empty_stack_spec.rb}}
-  
-A context represents a collection of specifications and must be defined with a name and a block (do-end).
-A specification is defined with the *specify* keyword and a sentence describing what's being specified.
-It is highly recommended to include
+Almost all expectation forms have a corresponding negated form. It is listed when it is supported and, unless otherwise stated, is met when ever the non-negated form would be violated.
+
+h3. General
+
+h4. Arbitrary Block
+
+<ruby>
+target.should.satisfy {|arg| ...}
+target.should.not.satisfy {|arg| ...}
+</ruby>
+
+The supplied block is evaluated, passing <code>target</code> as the sole argument. If the block evaluates to <code>false</code>, <code>ExpectationNotMetError</code> is raised.
+
+<ruby>
+target.should.satisfy {|arg| arg > 0}
+</ruby>
+
+h4. Equality
+
+<ruby>
+target.should.equal <value>
+target.should.not.equal <value>
+</ruby>
+
+The target object is compared to <code>value</code> using ==. If the result is <code>false</code>, <code>ExpectationNotMetError</code> is raised.
+
+h4. Floating Point Comparison
+
+<ruby>
+target.should.be.close <value>, <tolerance>
+target.should.not.be.close <value>, <tolerance>
+</ruby>
+
+The target object is compared to <code>value</code>. If they differ by more that <code>tolerance</code>, <code>ExpectationNotMetError</code> is raised. In the negated case, <code>ExpectationNotMetError</code> is raised if they differ by less than <code>tolerance</code>.
+
+<ruby>
+target.should.be.close 27.35, 0.05
+</ruby>
+
+h4. Identity
+
+<ruby>
+target.should.be <value>
+target.should.not.be <value>
+</ruby>
+
+The target object is compared to <code>value</code> using <code>equal?</code>. If the result is <code>false</code>, <code>ExpectationNotMetError</code> is raised.
+
+h4. Arbitrary Predicate
+
+<ruby>
+target.should.predicate [optional args]
+target.should.be.predicate [optional args]
+target.should.not.predicate [optional args]
+target.should.not.be.predicate [optional args]
+</ruby>
+
+The message <code>predicate?</code> is sent to <code>target</code> with any supplied arguments. If the result is <code>false</code>, <code>ExpectationNotMetError</code> is raised.
+
+For example:
+
+<ruby>
+container.should.include('a') => container.include?('a')
+container.should.be.empty => container.empty?
+</ruby>
+
+h4. Pattern Matching
+
+<ruby>
+target.should.match <regex>
+target.should.not.match <regex>
+</ruby>
+
+The <code>target</code> is matched against <code>regex</code>. An <code>ExpectationNotMetError</code> is raised if the match fails.
+
+h3. Class/Type
+
+h4. Direct Instance
+
+<ruby>
+target.should.be.an.instance.of <class>
+target.should.not.be.an.instance.of <class>
+</ruby>
+
+An <code>ExpectationNotMetError</code> is raised if <code>target</code> is not or is, respectively, an direct instance of <code>class</code>. As expected this correlates to <code>target.instance_of?  class</code>.
+
+h4. Ancestor Class
+
+<ruby>
+target.should.be.a.kind.of <class>
+target.should.not.be.a.kind.of <class>
+</ruby>
+
+As above, but uses <code>target.kind_of? class</code>: checking whether <code>class</code> is the direct class of <code>target</code>, or an ancestor of <code>target</code>'s direct class.
+
+h4. Type
+
+<ruby>
+target.should.respond.to <symbol>
+target.should.not.respond.to <symbol>
+</ruby>
+
+Uses <code>target.respond_to?(symbol)</code> to check whether <code>symbol</code> is the name of a message that <code>target</code> understands.
+
+h3. Procs
+
+h4. Raising
+
+<ruby>
+proc.should.raise <exception>
+proc.should.not.raise <exception>
+</ruby>
+
+Checks that <code>proc</code> causes the named exception to be raised or not. The latter is actually one of two cases: some other exception is raised, or no exception is raised. Typically the <code>proc</code> is created in place using <code>lambda</code>. For example:
+
+<ruby>
+lambda { 3 / 0 }.should.raise ZeroDivisionError
+</ruby>
+
+There is a more general form as well.
+
+<ruby>
+proc.should.raise
+proc.should.not.raise
+</ruby>
+
+These forms don't worry about what exception is raised (or not). All they are concerned with is that some except was raised, or that no exception was.
+
+h4. Throwing
+
+<ruby>
+proc.should.throw <symbol>
+proc.should.not.throw <symbol>
+</ruby>
+
+Similar to the above, but checks that <code>symbol</code> is thrown from within <code>proc</code>, or not. The latter is actually one of two cases: some other symbol is thrown, or no symbol is thrown.
+
+<ruby>
+proc.should.not.throw
+</ruby>
+
+This form is more specific.  It checks that no symbol is thrown from within <code>proc</code>.
+
+h3. Collections
+
+h4. Containment
+
+<ruby>
+target.should.include <object>
+target.should.not.include <object>
+</ruby>
+
+This is simply a specific case of the arbitrary predicate form. It uses <code>target.include?(object)</code> and raises an <code>ExpectationNotMetError</code> if that returns false. 
+
+The remaining collection forms are a little more involved. They rely on two things: 1) <code>target</code> responds to the message <code>things</code> by returning an object that 2) responds to either <code>length</code> or <code>size</code>, which return a number that is a measure of size. Currently <code>length</code> is used if is appropriate, otherwise <code>size</code> is attempted.
+
+h4. Exact Size
+
+<ruby>
+target.should.have(<number>).things
+</ruby>
+
+The <code>things</code> of <code>target</code> has a length/size of exactly <code>number</code>. 
+
+h4. Lower Bound
+
+<ruby>
+target.should.have.at.least(<number>).things
+</ruby>
+
+The <code>things</code> of <code>target</code> has a length/size of no less than <code>number</code>.
+
+h4. Upper Bound
+
+<ruby>
+target.should.have.at.most(<number>).things
+</ruby>
+
+The <code>things</code> of <code>target</code> has a length/size of no more than <code>number</code>.