chemistrykit 1.2.1 → 1.3.0

data/.travis.yml

@@ -0,0 +1,9 @@
+language: ruby
+rvm:
+  - 1.9.3
+branches:
+  only:
+    - master
+    - develop
+notifications:
+  hipchat: 2a07784c1aeed1b6efa332242d3fc3@Regulation Station

data/CHANGELOG.md

@@ -1,5 +1,78 @@
-v0.1.1
-======
+#1.3.0 (2013-06-22)
+- Added explicit recursive file loading process for formulas
+- Cleaned up documentation
+- Updated tests
+- Cleaned up logging
+
+#1.2.1 (2013-06-21)
+- Bumping version number and adding Jason Fox as an author
+- Making it so symbol values as tags in beakers will default to true if no value is set to them
+- Pulled out the log value setter from shared_context and rolled this up into selenium-connect instead. Bumped to new version of selenium-connect to get this functionality
+- Reworked the status check and exit logic order and implemented it as a ternary (effectively regressing back to what was there way back when). A better band-aid fix for now.
+- added small tweak to catch the {} returned by calling ckit, which was giving an hash to integer conversion error
+- Bumped version and removed the un-used spec dir
+- Fixes #59. Will need to revisit the approach since it is a workaround given difficulties in getting Thor's self.exit_on_failure?; true; end to work properly Re: https://github.com/erikhuda/thor/issues/244
+- Added a test to capture the exit status issue
+- Updated the spec and moved it up a dir
+- removed symlinking of historical log files and cleaned up the logging system to output the junit xml and the server log to the evidence directory. Added a bit of test for that.
+- updated git ignore for build directory and changed the new feature to a different name to prevent conflicts
+- removed non running spec
+- added a few rake tasks and switched to standard build directory
+#1.1.1 (2013-06-09)
+- Bumping selenium-connect version to account for sauce gem breaking changes
+
+#1.1.0 (2013-06-09)
+- Added the ability to pass in environment variables with ckit brew --params=THING1:value THING2:value
+
+#1.0.0 (2013-06-05)
+- Removed a hard-coded exit code checker from the ckit binary, also adjusted the requires in it to use the top level chemistrykit file in lib. Had to rework that file as well. Also updated the readme
+- Looks like my previous commit didn't include the cleanup of old files. Here you go! Also, updated the readme slightly
+- Wired up selenium-connect and gladly gutted a heap of code that it replaced. Also, updated the readme and updated the gemspec.
+- Readded the new commmand since someone got a little code clippy. Wired up the assertion in the "new" test so it actually works. And made the tmp directory listing explicit.
+- Adding tmp directory to gitignore
+- Updating gitignore for rvmrc
+- Added string interpolation for doc strings with erb and consolidated the brew feature into a single scenario outline with an examples table
+- Simplified ckit generate with a proper subcommand
+- Tested on local branch, tests passing. Removed puts.
+- Add the requires back to the method, not happy about it, but necessary for now
+- Didn't comment out all of one scenario: fixed
+- Disabled sauce with chrome scenario, without a way to check sauce to see if chrome is being used, it's the same as sauce with firefox
+- All four scenarios working. Sacuelabs credentials needed in order for tests to run.
+- Wait only 15 seconds intead of 30 seconds
+- Passing selenium-webdriver and local selenium server scenarios
+- Disabled new.feature, updated brew.feature with 4 types of scenarios
+- Removed usage title, it already has a good title
+- Brew cmd not default, updated documentation to reflect change
+- Removed a duplicate if statement, added a lot of puts to see what's happening
+- Added before(:all) hook to fix #48, refactored more logic out of before(:each)
+- Condensed to methods for driver into one
+- Updated the generated beaker with a formula to use the let incantation.
+- Required formula in generator
+- changelog updated to v0.1.1
+- Added rake back to the gemspec becuase we are using it
+- remove more self's
+- You don't need to reference self so much, it will figure it out
+- You don't need to initialize instance methods unless you want to set their value
+- Passing variables from module Sauce to Ckit in a safer way, fixed bug where executor was being called twice per test
+- Extracted sauce actions into a module
+- Refactored logic out of after:each into methods
+- Case statement for symlinking, added notes about they whys
+- seperated rspec configuration from tagging logic
+- log_timestamp was being called twice and creating two different values - fixed
+- Removal of if statement, using ln_sf forces the creation of a symlink if one already exists
+- Refactored logic for brew command into protected methods
+- Added code climate badge
+- Corrected version #
+- Adding a tag for testing the testing of aruba
+- Added a passing brew commit command and modified support.rb to env.rb like the docs say
+- More updates to make it less overhead
+- Simplified the new cli feature file, I think I'm starting to get aruba
+- Told git to ignore the files that aruba is creating and updated new feature
+- Cucumber spec for ckit new command
+- Removed rake because we're not using it, added aruba and cucumber because living docs for the Cli is cool
+- Please load aruba when you run feature files
+
+#0.1.1 (2013-02-02)
 
 * `ckit new` generates a chemistrykit project
 * Execute tests in SauceLabs Ondemand

data/CONTRIBUTORS.md

@@ -1,5 +1,5 @@
 In alphabetical order:
 
+* Jason Fox (@jasonrobertfox)
 * Adam Goucher (@adamgoucher)
 * Dave Haeffner (@tourdedave)
-* Jason Wieringa (@jwieringa)

data/README.md

@@ -1,6 +1,10 @@
 [![Gem Version](https://badge.fury.io/rb/chemistrykit.png)](http://badge.fury.io/rb/chemistrykit)
 [![Code Climate](https://codeclimate.com/github/arrgyle/chemistrykit.png)](https://codeclimate.com/github/arrgyle/chemistrykit)
 
+Master branch: [![Build Status](https://travis-ci.org/arrgyle/chemistrykit.png?branch=master)](https://travis-ci.org/jrobertfox/chef-broiler-platter)
+
+Develop branch: [![Build Status](https://travis-ci.org/arrgyle/chemistrykit.png?branch=develop)](https://travis-ci.org/jrobertfox/chef-broiler-platter)
+
 ChemistryKit
 ============================================================
 
@@ -28,8 +32,18 @@ This will run ckit and execute your beakers. By default it will run the tests lo
 
 
 ## Contributing
+This project conforms to the [neverstopbuilding/craftsmanship](https://github.com/neverstopbuilding/craftsmanship) guidelines. Please see them for details on:
+- Branching theory
+- Documentation expectations
+- Release process
+
+###It's simple
+
+1. Create a feature branch from develop: `git checkout -b feature/myfeature develop` or `git flow feature start myfeature`
+2. Do something awesome.
+3. Submit a pull request.
 
-This project conforms to the [neverstopbuilding/craftsmanship](https://github.com/neverstopbuilding/craftsmanship) guidelines. Please see them for details.
+All issues and questions related to this project should be logged using the [github issues](https://github.com/arrgyle/chemistrykit/issues) feature.
 
 ### Install Dependencies
 
@@ -41,4 +55,4 @@ This project conforms to the [neverstopbuilding/craftsmanship](https://github.co
 
 ### Run the local version of the executable:
 
-    bin/ckit
+    ckit

data/Rakefile

@@ -3,14 +3,21 @@ require "cucumber"
 require "cucumber/rake/task"
 require 'rspec/core/rake_task'
 
-task :default => [:build]
+task :default => [:build_ci]
 
 desc 'Runs standard build activities.'
-task :build => [:clean, :spec, :cucumber]
+task :build => [:clean, :prepare, :spec, :cucumber]
+
+desc 'Runs standard build activities for ci server.'
+task :build_ci => [:clean, :prepare, :spec]
 
 desc 'Removes the build directory.'
 task :clean do
-  FileUtils.rm_rf('build');
+  FileUtils.rm_rf('build')
+end
+desc 'Adds the build tmp directory for test kit creation.'
+task :prepare do
+  FileUtils.mkdir_p('build/tmp')
 end
 
 RSpec::Core::RakeTask.new(:spec)

data/chemistrykit.gemspec

@@ -1,12 +1,12 @@
 Gem::Specification.new do |s|
   s.name          = "chemistrykit"
-  s.version       = "1.2.1"
+  s.version       = "1.3.0"
   s.platform      = Gem::Platform::RUBY
   s.authors       = ["Dave Haeffner", "Jason Fox"]
   s.email         = ["dave@arrgyle.com", "jason@arrgyle.com"]
   s.homepage      = "https://github.com/arrgyle/chemistrykit"
   s.summary       = "A simple and opinionated web testing framework for Selenium that follows convention over configuration."
-  s.description   = "Fixing the exit codes better, and addressing a bug with logging."
+  s.description   = "Updated file loading, documentation and tests."
   s.license       = 'MIT'
 
   s.files         = `git ls-files`.split($/)

data/features/brew.feature

@@ -37,6 +37,7 @@ Feature: Brewing a ChemistryKit project
   Scenario Outline: Run All Configurations
     When I overwrite _config.yaml with:
       """
+      log: 'evidence'
       host: '<%= <hostname> %>'
       sauce_username: 'testing_arrgyle'
       sauce_api_key:  'ab7a6e17-16df-42d2-9ef6-c8d2539cc38a'
@@ -51,3 +52,5 @@ Feature: Brewing a ChemistryKit project
     | hostname    |
     | "localhost" |
 #    | "saucelabs" |
+
+

data/features/load_page_objects.feature

@@ -0,0 +1,47 @@
+Feature: Load Page Objects
+Formulas should be loaded in the correct order with thier dependencies
+
+  Scenario: Load the libs first
+    Given I run `ckit new big-project`
+    And I cd to "big-project"
+    And a file named "formulas/big.rb" with:
+      """
+      module Formulas
+        class BigProject < Formula
+
+          def open(url)
+            helper_open(url)
+          end
+
+        end
+      end
+      """
+
+    And a file named "formulas/lib/formula.rb" with:
+      """
+      module Formulas
+        class Formula
+          def initialize(driver)
+            @driver = driver
+          end
+
+          def helper_open(url)
+            @driver.get url
+          end
+        end
+      end
+      """
+
+    And a file named "beaker/big_beaker.rb" with:
+      """
+      describe "Big", :depth => 'shallow' do
+        let(:book) { Formulas::BigProject.new(@driver) }
+
+        it "loads an external web page" do
+          book.open "http://www.google.com"
+        end
+      end
+      """
+
+    When I run `ckit brew`
+    Then the stdout should contain "1 example, 0 failures"

data/lib/chemistrykit/cli/cli.rb

@@ -4,6 +4,7 @@ require 'ci/reporter/rake/rspec_loader'
 require 'chemistrykit/cli/new'
 require 'chemistrykit/cli/formula'
 require 'chemistrykit/cli/beaker'
+require 'chemistrykit/cli/helpers/formula_loader'
 
 module ChemistryKit
   module CLI
@@ -45,7 +46,8 @@ module ChemistryKit
       end
 
       def load_page_objects
-        Dir["#{Dir.getwd}/formulas/*.rb"].each {|file| require file }
+        loader = ChemistryKit::CLI::Helpers::FormulaLoader.new
+        loader.get_formulas(File.join(Dir.getwd, 'formulas')).each {|file| require file }
       end
 
       def set_logs_dir

data/lib/chemistrykit/cli/helpers/formula_loader.rb

@@ -0,0 +1,46 @@
+module ChemistryKit
+  module CLI
+    module Helpers
+      class FormulaLoader
+        #this loader returns the formulas found in a path according to these rules:
+        #- directories loaded in alpha order
+        #- children directories loaded before parents
+        #- files loaded in alpha order
+        #- lib directories are loaded before any other
+        #- rules stack
+
+        def initialize
+          @formulas = Array.new
+        end
+
+        def get_formulas(path)
+          gather(path)
+          @formulas
+        end
+
+        private
+          def gather(path)
+            #get all the directories for path
+            directories = Dir.entries(path).select {|entry| File.directory? File.join(path,entry) and !(entry =='.' || entry == '..') }
+            if directories.empty?
+              #if there are no directories is empty just get all the files and add it to the instance variable
+              @formulas.concat Dir.glob(File.join(path, "*.rb")).sort
+            else
+              #otherwise for each of them recurse into it
+              directories.sort!
+              if directories.include? 'lib'
+                #if there is a lib directory, put that at the front
+                directories.delete('lib')
+                directories.unshift('lib')
+              end
+              directories.each do |directory|
+                gather(File.join(path, directory))
+              end
+              #and then after add the files in the original parent
+              @formulas.concat Dir.glob(File.join(path, "*.rb")).sort
+            end
+          end
+      end
+    end
+  end
+end

data/spec/chemistrykit/cli/helpers/formula_loader_spec.rb

@@ -0,0 +1,87 @@
+require 'spec_helper'
+#might be able to use something like this: https://github.com/alexeypetrushin/class_loader
+
+describe ChemistryKit::CLI::Helpers::FormulaLoader do
+
+  before(:each) do
+    @loader = ChemistryKit::CLI::Helpers::FormulaLoader.new
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula'))
+  end
+
+  it 'should return a new instance of the loader' do
+    @loader.should be_an_instance_of ChemistryKit::CLI::Helpers::FormulaLoader
+  end
+
+  it 'should return ruby files in alphabetical order' do
+    #create some test files
+    d = File.join(TEST_TMP_PATH, 'formula', 'd.rb')
+    a = File.join(TEST_TMP_PATH, 'formula', 'a.rb')
+    b = File.join(TEST_TMP_PATH, 'formula', 'b.rb')
+    FileUtils.touch(d)
+    FileUtils.touch(a)
+    FileUtils.touch(b)
+
+    #ensure the result
+    @loader.get_formulas(File.join(TEST_TMP_PATH, 'formula')).should eq [a, b, d]
+
+  end
+
+  it 'should order files in child directories before parent directories' do
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'child'))
+    d = File.join(TEST_TMP_PATH, 'formula', 'child' ,'d.rb')
+    b = File.join(TEST_TMP_PATH, 'formula', 'child' ,'b.rb')
+    a = File.join(TEST_TMP_PATH, 'formula', 'a.rb')
+    c = File.join(TEST_TMP_PATH, 'formula', 'c.rb')
+    FileUtils.touch(d)
+    FileUtils.touch(b)
+    FileUtils.touch(a)
+    FileUtils.touch(c)
+
+    @loader.get_formulas(File.join(TEST_TMP_PATH, 'formula')).should eq [b, d, a, c]
+  end
+
+  it 'should load directories in alphabetical order' do
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'abby'))
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'road'))
+    d = File.join(TEST_TMP_PATH, 'formula', 'abby' ,'d.rb')
+    b = File.join(TEST_TMP_PATH, 'formula', 'road' ,'b.rb')
+    FileUtils.touch(d)
+    FileUtils.touch(b)
+    @loader.get_formulas(File.join(TEST_TMP_PATH, 'formula')).should eq [d, b]
+  end
+
+  it 'should load any lib directory before any other' do
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'lib'))
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'abby'))
+    d = File.join(TEST_TMP_PATH, 'formula', 'lib' ,'d.rb')
+    b = File.join(TEST_TMP_PATH, 'formula', 'abby' ,'b.rb')
+    FileUtils.touch(d)
+    FileUtils.touch(b)
+    @loader.get_formulas(File.join(TEST_TMP_PATH, 'formula')).should eq [d, b]
+  end
+
+  it 'should apply stack these rules deeply' do
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'lib'))
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'abby'))
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'abby', 'lib'))
+    Dir.mkdir(File.join(TEST_TMP_PATH, 'formula', 'abby', 'road'))
+
+    a = File.join(TEST_TMP_PATH, 'formula', 'lib' ,'a.rb')
+    c = File.join(TEST_TMP_PATH, 'formula', 'abby' ,'c.rb')
+    b = File.join(TEST_TMP_PATH, 'formula', 'b.rb')
+    d = File.join(TEST_TMP_PATH, 'formula', 'abby' , 'lib', 'd.rb')
+    e = File.join(TEST_TMP_PATH, 'formula', 'abby' , 'road', 'e.rb')
+
+    FileUtils.touch(a)
+    FileUtils.touch(b)
+    FileUtils.touch(c)
+    FileUtils.touch(d)
+    FileUtils.touch(e)
+
+    @loader.get_formulas(File.join(TEST_TMP_PATH, 'formula')).should eq [a, d, e, c, b]
+  end
+
+  after(:each) do
+    FileUtils.rm_rf(File.join(TEST_TMP_PATH, 'formula'))
+  end
+end

data/spec/spec_helper.rb

@@ -4,6 +4,15 @@
 # loaded once.
 #
 # See http://rubydoc.info/gems/rspec-core/RSpec/Core/Configuration
+#
+require 'rubygems'
+require 'bundler/setup'
+require 'fileutils'
+#not sure it this is the right way to include all the files under test.
+require_relative '../lib/chemistrykit/cli/helpers/formula_loader'
+
+TEST_TMP_PATH = File.join(Dir.pwd, 'build', 'tmp')
+
 RSpec.configure do |config|
   config.treat_symbols_as_metadata_keys_with_true_values = true
   config.run_all_when_everything_filtered = true

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: chemistrykit
 version: !ruby/object:Gem::Version
-  version: 1.2.1
+  version: 1.3.0
   prerelease: 
 platform: ruby
 authors:
@@ -172,7 +172,7 @@ dependencies:
     - - ~>
       - !ruby/object:Gem::Version
         version: 10.0.3
-description: Fixing the exit codes better, and addressing a bug with logging.
+description: Updated file loading, documentation and tests.
 email:
 - dave@arrgyle.com
 - jason@arrgyle.com
@@ -184,6 +184,7 @@ extra_rdoc_files: []
 files:
 - .gitignore
 - .rspec
+- .travis.yml
 - CHANGELOG.md
 - CONTRIBUTORS.md
 - Gemfile
@@ -196,6 +197,7 @@ files:
 - ext/mkrf_conf.rb
 - features/brew.feature
 - features/exit_status.feature
+- features/load_page_objects.feature
 - features/new.feature
 - features/step_definitions/steps.rb
 - features/support/env.rb
@@ -203,7 +205,7 @@ files:
 - lib/chemistrykit/cli/beaker.rb
 - lib/chemistrykit/cli/cli.rb
 - lib/chemistrykit/cli/formula.rb
-- lib/chemistrykit/cli/generators.rb
+- lib/chemistrykit/cli/helpers/formula_loader.rb
 - lib/chemistrykit/cli/new.rb
 - lib/chemistrykit/shared_context.rb
 - lib/templates/beaker.tt
@@ -214,8 +216,7 @@ files:
 - lib/templates/chemistrykit/evidence/.gitkeep
 - lib/templates/chemistrykit/formulas/.gitkeep
 - lib/templates/formula.tt
-- readme2.md
-- spec/chemistrykit/cli_spec.rb
+- spec/chemistrykit/cli/helpers/formula_loader_spec.rb
 - spec/spec_helper.rb
 homepage: https://github.com/arrgyle/chemistrykit
 licenses:
@@ -238,7 +239,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.24
+rubygems_version: 1.8.25
 signing_key: 
 specification_version: 3
 summary: A simple and opinionated web testing framework for Selenium that follows
@@ -246,8 +247,9 @@ summary: A simple and opinionated web testing framework for Selenium that follow
 test_files:
 - features/brew.feature
 - features/exit_status.feature
+- features/load_page_objects.feature
 - features/new.feature
 - features/step_definitions/steps.rb
 - features/support/env.rb
-- spec/chemistrykit/cli_spec.rb
+- spec/chemistrykit/cli/helpers/formula_loader_spec.rb
 - spec/spec_helper.rb

data/lib/chemistrykit/cli/generators.rb

@@ -1,12 +0,0 @@
-require 'thor'
-require 'chemistrykit/cli/formula'
-require 'chemistrykit/cli/beaker'
-
-module ChemistryKit
-  module CLI
-    class Generate < Thor
-      register(ChemistryKit::CLI::PageObjectGenerator, 'formula', 'formula [NAME]', 'generates a page object')
-      register(ChemistryKit::CLI::BeakerGenerator, 'beaker', 'beaker [NAME]', 'generates a beaker')
-    end
-  end
-end

data/readme2.md

@@ -1,78 +0,0 @@
-Spec Discovery
---------------
-
-ChemistryKit is built on top of RSpec. All specs are in the _beaker_ directory and end in _beaker.rb. Rather than being discovered via class or file name as some systems they are by identified by tag. 
-
-```ruby
-it 'with invalid credentials', :depth => 'shallow' do
-
-end
-
-it 'with invalid credentials', :depth => 'deep' do
-
-end
-````
-All specs should have at least a :depth tag. The depth should either be 'shallow' or 'deep'. Shallow specs are the ones that are the absolute-must-pass ones. And there will only be a few of them typically. Deep ones are everything else.
-
-You can add multiple tags as well.
-
-```ruby
-it 'with invalid credentials', :depth => 'shallow', :authentication => true do
-
-end
-````
-
-By default ChemistryKit will discover and run the _:depth => 'shallow'_ scripts. To run different ones you use the --tag option.
-
-    ckit brew --tag authentication
-
-    ckit brew --tag depth:shallow --tag authentication
-
-To exlude a tag, put a ~ in front of it.
-
-    ckit brew --tag depth:shallow --tag ~authentication
-
-A useful trick when developing a script is to use a custom tag.
-
-```ruby
-it 'with invalid credentials', :depth => 'shallow', :flyingmonkeybutt => true do
-
-end
-````
-
-Execution Order
----------------
-
-Chemistry Kit executes specs in a random order. This is intentional. Knowing the order a spec will be executed in allows for dependencies between them to creep in. Sometimes unintentionally. By having them go in a random order parallelization becomes a much easier.
-
-Facade all the Things!
-----------------------
-
-Chemistry Kit injects itself between you and WebDriver and various other future components. You should also inject something between your project and Chemistry Kit. Chemistry Kit has started this for you in the following ways:
-
-- _config/requires.rb: @driver inside your scripts comes from the first line in this file
-
-Configuration
--------------
-
-Configuration should not be in your script. Nor should it be commit to version control. 
-
-Before and After
-----------------
-
-Chemistry Kit uses the 4-phase model for scripts with a chunk of code that gets run before and after each method. By default, it does nothing more than launch a browser instance that your configuration says you want. If you want to do something more than that, just add it to your spec.
-
-```ruby
-before(:each) do
-    # something here
-end
-```
-
-You can even nest them inside different describe/context blocks and they will get executed from the outside-in.
-
-Logs and CI Integration
------------------------
-
-Each run of Chemistry Kit creates a timestamped directory inside the _evidence_ directory. And in there will be the full set of JUnit Ant XML files. You don't point your CI server at this timestamped directory. Instead you want to point at the _latest_ directory which is a symlink to the latest timestamp directory.
-
-