dataset 1.3.1

data/CHANGELOG

@@ -0,0 +1,59 @@
+*1.3.0 [Rails 2.3.2] (??)
+
+* Supporting Rails 2.3.2 [mhawkins, underlog]
+* Added tasks to load path so folks can require 'dataset.rake' [aiwilliams]
+* Updated gemspec to include dependencies [aiwilliams]
+
+*1.2.0 [Cucumber] (April 8, 2009)
+
+* Support for cucumber [jgarber, seancribbs]
+
+*1.1.0 [STI, belongs_to] (February 14, 2009)
+
+* STI is better supported for inserting, naming and finding records [aiwilliams]
+
+  class Place < ActiveRecord::Base; end
+  class State < Place; end
+  class NorthCarolina < State; end
+
+  create_record(NorthCarolina, :state)                        # no need to define the 'type' column value
+  states(:state) == places(:state) == north_carolinas(:state) # read with the class names pluralized
+
+* Moved to jeweler for much cleaner, github embracing gem building [aiwilliams]
+* Support for simple belongs to associations [aiwilliams]
+
+  class Person < ActiveRecord::Base; end
+  class Note < ActiveRecord::Base
+    belongs_to :person
+  end
+  
+  person_id = create_record Person, :myguy
+  create_record Note, :person => :myguy
+  Note.last.person_id == person_id
+
+* Models inside modules are supported a little more conveniently [aiwilliams]
+
+  module MList
+    class Message < ActiveRecord::Base
+    end
+  end
+
+  # We'll get rid of the underscore in 'm_list_messages'
+  create_record MList::Message, :test
+  mlist_messages(:test)
+
+* Helper method for converting strings to useful symbols for finding records [siannopollo]
+
+  This is useful if you write creator methods of your own.
+
+  def create_person(attributes)
+    create_record Person, name_to_sym(attributes[:name]), attributes
+  end
+
+  create_person(:name => 'Little John')
+  people(:little_john)
+
+
+*1.0.0 [Scenarios Replacement] (December 15, 2008)
+
+* Drop-in replacement for Scenarios plugin of old [aiwilliams]

data/LICENSE

@@ -0,0 +1,19 @@
+Copyright (c) 2008-2009, Adam Williams
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

data/README

@@ -0,0 +1,111 @@
+= Dataset
+
+Dataset provides a simple API for creating and finding sets of data in your database. Check out Dataset::RecordMethods and Dataset::ModelFinders.
+
+Dataset loads data intelligently if you use 'nested contexts' in your tests (RSpec, anything that uses Test::Unit::TestCase subclassing for creating nested contexts):
+
+    describe Something do
+      dataset :a              => Dataset :a is loaded (at the right time)
+                              
+      it 'should whatever'    
+      end                     
+                              
+      describe More do        
+        dataset :b            => Dataset :b is loaded. :a data is still there
+                              
+        it 'should'           
+        end                   
+      end                     
+                              
+      describe Another do     => Database is restored to :a, without re-running :a logic
+        it 'should'
+        end
+      end
+    end
+  
+The goal is to see a marked improvement in overall test run speed, basing this on the assumption that it is faster to have the OS copy a file or mySQL dump and load. Of course, we may find this to be a false assumption, but there were plenty of bugs in the former 'Scenarios' - addressing that afforded the opportunity to test the assumption.
+
+
+Dataset does not prevent you from using other libraries like Machinist or factory_girl. If you were to used either of those, you could have a dataset like this:
+  
+    require 'faker'
+    
+    class OrganizationsDataset < Dataset::Base
+      Sham.name  { Faker::Name.name }
+      
+      Organization.blueprint do
+        name { Sham.name }
+      end
+      
+      def load
+        name_model Organization.make, :org_one
+      end
+    end
+  
+The benefit is that you can reuse interesting sets of data, without sacrificing the utility of those other libraries.
+
+    describe Organization, 'stuff' do
+      dataset :organizations
+    end
+    
+    describe Organization, 'other stuff' do
+      dataset :organizations
+    end
+  
+
+Get things installed, then read more in the Dataset documentation at http://aiwilliams.github.com/dataset
+
+
+== Installation
+
+Install the plugin:
+
+    ./script/plugin install git://github.com/aiwilliams/dataset.git
+
+In your test_helper.rb/spec_helper.rb:
+
+    require 'dataset'
+    class Test::Unit::TestCase
+      include Dataset
+      datasets_directory "#{RAILS_ROOT}/spec/datasets"
+    end
+
+If you don't use rspec_on_rails, or you have specs that aren't of the RailsExampleGroup type, you should do this in spec_helper.rb:
+
+    require 'dataset'
+    class Spec::Example::ExampleGroup
+      include Dataset
+      datasets_directory "#{RAILS_ROOT}/spec/datasets"
+    end
+
+If you were a user of the Scenarios plugin, and want to do as little as possible to get going (assumes you are using rspec_on_rails):
+
+    require 'dataset'
+    Scenario = Scenarios = Dataset
+    class Test::Unit::TestCase
+      include Dataset
+      class << self
+        alias_method :scenario, :dataset
+      end
+    end
+    class ScenariosResolver < Dataset::DirectoryResolver
+      def suffix
+        @suffix ||= 'Scenario'
+      end
+    end
+    Dataset::Resolver.default = ScenariosResolver.new("#{RAILS_ROOT}/spec/scenarios")
+
+
+== Credits
+
+Written by [Adam Williams](http://github.com/aiwilliams).
+    
+Contributors:
+
+- [Saturn Flyer](http://www.saturnflyer.com) [github](http://github.com/saturnflyer)
+- [Steve Iannopollo](http://github.com/siannopollo)
+- [John Long](http://github.com/jlong)
+
+---
+
+Dataset is released under the MIT-License and is Copyright (c)2008 Adam Williams.

data/Rakefile

@@ -0,0 +1,31 @@
+$:.unshift(File.join(File.dirname(__FILE__), 'lib'))
+
+require File.join(File.dirname(__FILE__), 'plugit/descriptor')
+require 'rubygems'
+require 'spec/rake/spectask'
+
+task :default => :spec
+
+desc "Run all specs"
+Spec::Rake::SpecTask.new do |t|
+  t.spec_files = FileList['spec/**/*_spec.rb']
+  t.spec_opts = ['--options', 'spec/spec.opts']
+end
+
+begin
+  require 'jeweler'
+  Jeweler::Tasks.new do |s|
+    s.name = 'dataset'
+    s.summary = 'A simple API for creating and finding sets of data in your database, built on ActiveRecord.'
+    s.email = 'adam@thewilliams.ws'
+    s.files = FileList["[A-Z]*", "{lib,tasks}/**/*", "plugit/descriptor.rb"].exclude("tmp")
+    s.require_paths = ["lib", "tasks"]
+    s.add_dependency('activesupport', '>= 2.3.0')
+    s.add_dependency('activerecord', '>= 2.3.0')
+    s.homepage = "http://github.com/aiwilliams/dataset"
+    s.description = s.summary
+    s.authors = ['Adam Williams']
+  end
+rescue LoadError
+  puts "Jeweler not available. Install it with: sudo gem install technicalpickles-jeweler -s http://gems.github.com"
+end

data/TODO

@@ -0,0 +1,15 @@
+take any instance variables already in context and make them available to dataset blocks - this is for nested describes
+  I'm not sure about this one. It can be very frustrating to lose context of when the state of an iv is modified.
+  
+add ability to clear the database (some tests wanted to guarantee a clear db)
+  This is acheived with "dataset {}"
+
+clear database completely at beginning of session, only tables where data was created within a session??
+
+clear all dumps on new run of tests
+be sure we are capturing a dataset if it has already be captured before during a run
+describe what happens when someone has a fixtures file - they get loaded after our datasets, thereby causing all the data in the table of the fixture file (like things.yml) to be deleted - the fixtures are then loaded
+look into truncating database instead individual table deletes
+allow configuration of dataset
+  * permatable / global scope
+re-evaluation location of some tests that depend on TestCase in non-test/unit tests

data/VERSION.yml

@@ -0,0 +1,4 @@
+--- 
+:patch: 1
+:major: 1
+:minor: 3

data/lib/dataset.rb

@@ -0,0 +1,128 @@
+require 'activesupport'
+require 'activerecord'
+
+require 'dataset/version'
+require 'dataset/instance_methods'
+require 'dataset/base'
+require 'dataset/database/base'
+require 'dataset/database/mysql'
+require 'dataset/database/sqlite3'
+require 'dataset/database/postgresql'
+require 'dataset/collection'
+require 'dataset/load'
+require 'dataset/resolver'
+require 'dataset/session'
+require 'dataset/session_binding'
+require 'dataset/record/meta'
+require 'dataset/record/fixture'
+require 'dataset/record/model'
+
+# == Quick Start
+# 
+# Write a test. If you want some data in your database, create a dataset.
+# Start simple.
+# 
+#     describe States do
+#       dataset do
+#         [%w(Colorado CO), %w(North\ Carolina NC), %w(South\ Carolina SC)].each do |name,abbrev|
+#           create_record :state, abbrev.downcase, :name => name, :abbrev => abbrev
+#         end
+#       end
+#        
+#       it 'should have an abbreviated name'
+#         states(:nc).abbrev.should be('NC')
+#       end
+#        
+#       it 'should have a name'
+#         states(:nc).name.should be('North Carolin')
+#       end
+#     end
+# 
+# Notice that you won't be using _find_id_ or _find_model_ in your tests. You
+# use methods like _states_ and _state_id_, as in the example above.
+#
+# When you find that you're seeing patterns in the data you are creating, pull it into a class.
+#
+#     spec/datasets/states.rb
+#     class StatesDataset < Dataset::Base
+#       def load
+#         # create useful data
+#       end
+#     end
+#
+#     spec/models/state.rb
+#     describe State do
+#       dataset :states
+#     end
+#
+# == Installation
+#
+# Dataset is installed into your testing environment by requiring the library,
+# then including it into the class that will be the context of your test
+# methods.
+#
+#    require 'dataset'
+#    class Test::Unit::TestCase
+#      include Dataset
+#      datasets_directory "#{RAILS_ROOT}/test/datasets"
+#    end
+#
+# Note that should you desire your Dataset::Base subclasses be
+# auto-discovered, you can set the _datasets_directory_.
+#
+module Dataset
+  def self.included(test_context) # :nodoc:
+    if test_context.name =~ /World\Z/
+      require 'dataset/extensions/cucumber'
+    elsif test_context.name =~ /TestCase\Z/
+      require 'dataset/extensions/test_unit'
+    elsif test_context.name =~ /ExampleGroup\Z/
+      require 'dataset/extensions/rspec'
+    else
+      raise "I don't understand your test framework"
+    end
+    
+    test_context.extend ContextClassMethods
+  end
+  
+  # Methods that are added to the class that Dataset is included in (the test
+  # context class).
+  #
+  module ContextClassMethods
+    def self.extended(context_class) # :nodoc:
+      context_class.module_eval do
+        include InstanceMethods
+        superclass_delegating_accessor :dataset_session
+      end
+    end
+    
+    mattr_accessor :datasets_database_dump_path
+    self.datasets_database_dump_path = File.expand_path(RAILS_ROOT + '/tmp/dataset') if defined?(RAILS_ROOT)
+    
+    # Replaces the default Dataset::Resolver with one that will look for
+    # dataset class definitions in the specified directory. Captures of the
+    # database will be stored in a subdirectory 'tmp' (see
+    # Dataset::Database::Base).
+    def datasets_directory(path)
+      Dataset::Resolver.default = Dataset::DirectoryResolver.new(path)
+      Dataset::ContextClassMethods.datasets_database_dump_path = File.join(path, '/tmp/dataset')
+    end
+    
+    def add_dataset(*datasets, &block) # :nodoc:
+      dataset_session = dataset_session_in_hierarchy
+      datasets.each { |dataset| dataset_session.add_dataset(self, dataset) }
+      dataset_session.add_dataset(self, Class.new(Dataset::Block) {
+        define_method :doload, block
+      }) unless block.nil?
+    end
+    
+    def dataset_session_in_hierarchy # :nodoc:
+      self.dataset_session ||= begin
+        database_spec = ActiveRecord::Base.configurations['test'].with_indifferent_access
+        database_class = Dataset::Database.const_get(database_spec[:adapter].classify)
+        database = database_class.new(database_spec, Dataset::ContextClassMethods.datasets_database_dump_path)
+        Dataset::Session.new(database)
+      end
+    end
+  end
+end

data/lib/dataset/base.rb

@@ -0,0 +1,157 @@
+module Dataset
+  
+  # The superclass of your Dataset classes.
+  #
+  # It is recommended that you create a dataset using the Dataset::Block
+  # method first, then grow into using classes as you recognize patterns in
+  # your test data creation. This will help you to keep simple things simple.
+  #
+  class Base
+    class << self
+      # Allows a subclass to define helper methods that should be made
+      # available to instances of this dataset, to datasets that use this
+      # dataset, and to tests that use this dataset.
+      #
+      # This feature is great for providing any kind of method that would help
+      # test the code around the data your dataset creates. Be careful,
+      # though, to keep from adding business logic to these methods! That
+      # belongs in your production code.
+      #
+      def helpers(&method_definitions)
+        @helper_methods ||= begin
+          mod = Module.new
+          include mod
+          mod
+        end
+        @helper_methods.module_eval &method_definitions
+      end
+      
+      def helper_methods # :nodoc:
+        @helper_methods
+      end
+      
+      # Allows a subsclass to declare which datasets it uses.
+      #
+      # Dataset is designed to promote 'design by composition', rather than
+      # 'design by inheritance'. You should not use class hiearchies to share
+      # data and code in your datasets. Instead, you can write something like
+      # this:
+      #
+      #   class PeopleDataset < Dataset::Base; end
+      #   class DepartmentsDataset < Dataset::Base; end
+      #   class OrganizationsDataset < Dataset::Base
+      #     uses :people, :departments
+      #   end
+      #
+      # When the OrganizationsDataset is loaded, it will have all the data
+      # from the datasets is uses, as well as all of the helper methods
+      # defined by those datasets.
+      #
+      # When a dataset uses other datasets, and those datasets themselves use
+      # datasets, things will be loaded in the order of dependency you would
+      # expect:
+      #
+      #   C uses B
+      #   A uses C
+      #   B, C, A is the load order
+      # 
+      def uses(*datasets)
+        @used_datasets = datasets
+      end
+      
+      def used_datasets # :nodoc:
+        @used_datasets
+      end
+    end
+    
+    # Invoked once before a collection of tests is run. If you use a dataset
+    # in multiple test classes, it will be called once for each of them -
+    # remember that the database will be cleared at the beginning of running a
+    # 'suite' or 'group' of tests, unless you are using nested contexts (as in
+    # nested describe blocks in RSpec).
+    #
+    # Override this method in your subclasses.
+    #
+    def load; end
+  end
+  
+  # The easiest way to create some data before a suite of tests is run is by
+  # using a Dataset::Block. An example works wonders:
+  #
+  #    class PeopleTest < Test::Unit::TestCase
+  #      dataset do
+  #        create_record :person, :billy, :name => 'Billy'
+  #      end
+  #            
+  #      def test_name
+  #        assert_equal 'Billy', people(:billy).name
+  #      end
+  #    end
+  #
+  # The database will be cleared and billy will be inserted once before
+  # running each of the tests within a transaction. All the normal transaction
+  # fixtures stuff will still work.
+  #
+  # One of the great features of Dataset, at least when things get really
+  # interesting in your data needs, is that nested contexts will be additive.
+  # Consider this:
+  #
+  #    describe Something do
+  #      dataset :a              => Dataset :a is loaded (at the right time)
+  #         
+  #      it 'should whatever'
+  #      end
+  #         
+  #      describe More do
+  #        dataset :b            => Dataset :b is loaded. :a data is still there
+  #           
+  #        it 'should'
+  #        end
+  #      end
+  #         
+  #      describe Another do     => Database is restored to :a, without re-running :a logic
+  #        it 'should'
+  #        end
+  #      end
+  #    end
+  #
+  # == Instance Variables
+  #
+  # You may also assign instance variables in a dataset block, and they will
+  # be available to your test methods. You have to be careful with this in a
+  # similar way that you must with an RSpec before :all block. Since the
+  # instance variables are pointing to the same instances accross all tests,
+  # things can get weird if you intend to change their state. It's best use is
+  # for loading objects that you want to read a lot without loading over and
+  # over again for each test.
+  #
+  # == Building on Other Datasets
+  #
+  # You may pass any number of Dataset::Base subclasses - or better yet, their
+  # names - to the dataset method. When you use a block, this adds a lot of
+  # clarity:
+  #
+  #    class PersonTest < Test::Unit::TestCase
+  #      dataset :organization, :people do
+  #        id = create_record :person, :second_admin, :name => 'Admin Three'
+  #        create_record :organization_administratorship, :organization_id => organization_id(:first_bank), :person_id => id
+  #      end
+  #       
+  #      def test_admins
+  #        assert organizations(:first_bank).admins.include?(people(:second_admin))
+  #      end
+  #    end
+  #
+  # == Reusing a Dataset
+  #
+  # When you need to go beyond the block, create a Dataset::Base subclass!
+  class Block < Base
+    include Dataset::InstanceMethods
+    
+    def load # :nodoc:
+      dataset_session_binding.install_block_variables(self)
+      doload
+      dataset_session_binding.copy_block_variables(self)
+    end
+  end
+end