aiwilliams-dataset 1.2.0

data/CHANGELOG

@@ -0,0 +1,54 @@
+*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 +1,28 @@
+$:.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.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 @@
+--- 
+:minor: 2
+:patch: 0
+:major: 1

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

data/lib/dataset/collection.rb

@@ -0,0 +1,19 @@
+require 'set'
+
+module Dataset
+  class Collection < Array # :nodoc:
+    def initialize(parent)
+      concat parent
+    end
+    
+    def <<(dataset)
+      super
+      uniq!
+      self
+    end
+    
+    def subset?(other)
+      Set.new(self).subset?(Set.new(other))
+    end
+  end
+end

data/lib/dataset/database/base.rb

@@ -0,0 +1,30 @@
+require 'fileutils'
+
+module Dataset
+  module Database # :nodoc:
+    
+    # Provides Dataset a way to clear, dump and load databases.
+    class Base
+      include FileUtils
+      
+      def clear
+        connection = ActiveRecord::Base.connection
+        ActiveRecord::Base.silence do
+          connection.tables.each do |table_name|
+            connection.delete "DELETE FROM #{connection.quote_table_name(table_name)}",
+              "Dataset::Database#clear" unless table_name == ActiveRecord::Migrator.schema_migrations_table_name
+          end
+        end
+      end
+      
+      def record_meta(record_class)
+        record_metas[record_class] ||= Dataset::Record::Meta.new(record_class)
+      end
+      
+      protected
+        def record_metas
+          @record_metas ||= Hash.new
+        end
+    end
+  end
+end

data/lib/dataset/database/mysql.rb

@@ -0,0 +1,34 @@
+module Dataset
+  module Database # :nodoc:
+    
+    # The interface to a mySQL database, this will capture by creating a dump
+    # file and restore by loading one of the same.
+    #
+    class Mysql < Base
+      def initialize(database_spec, storage_path)
+        @database = database_spec[:database]
+        @username = database_spec[:username]
+        @password = database_spec[:password]
+        @storage_path = storage_path
+        FileUtils.mkdir_p(@storage_path)
+      end
+      
+      def capture(datasets)
+        return if datasets.nil? || datasets.empty?
+        `mysqldump -u #{@username} --password=#{@password} --compact --extended-insert --no-create-db --add-drop-table --quick --quote-names #{@database} > #{storage_path(datasets)}`
+      end
+      
+      def restore(datasets)
+        store = storage_path(datasets)
+        if File.file?(store)
+          `mysql -u #{@username} --password=#{@password} --database=#{@database} < #{store}`
+          true
+        end
+      end
+      
+      def storage_path(datasets)
+        "#{@storage_path}/#{datasets.collect {|c| c.__id__}.join('_')}.sql"
+      end
+    end
+  end
+end

data/lib/dataset/database/postgresql.rb

@@ -0,0 +1,34 @@
+module Dataset
+  module Database # :nodoc:
+    
+    # The interface to a PostgreSQL database, this will capture by creating a dump
+    # file and restore by loading one of the same.
+    #
+    class Postgresql < Base
+      def initialize(database_spec, storage_path)
+        @database = database_spec[:database]
+        @username = database_spec[:username]
+        @password = database_spec[:password]
+        @storage_path = storage_path
+        FileUtils.mkdir_p(@storage_path)
+      end
+      
+      def capture(datasets)
+        return if datasets.nil? || datasets.empty?
+        `pg_dump -c #{@database} > #{storage_path(datasets)}`
+      end
+      
+      def restore(datasets)
+        store = storage_path(datasets)
+        if File.file?(store)
+          `psql -U #{@username} -p #{@password} -e #{@database} < #{store}`
+          true
+        end
+      end
+      
+      def storage_path(datasets)
+        "#{@storage_path}/#{datasets.collect {|c| c.__id__}.join('_')}.sql"
+      end
+    end
+  end
+end

data/lib/dataset/database/sqlite3.rb

@@ -0,0 +1,32 @@
+module Dataset
+  module Database # :nodoc:
+    
+    # The interface to a sqlite3 database, this will capture by copying the db
+    # file and restore by replacing and reconnecting to one of the same.
+    #
+    class Sqlite3 < Base
+      def initialize(database_spec, storage_path)
+        @database_path, @storage_path = database_spec[:database], storage_path
+        FileUtils.mkdir_p(@storage_path)
+      end
+      
+      def capture(datasets)
+        return if datasets.nil? || datasets.empty?
+        cp @database_path, storage_path(datasets)
+      end
+      
+      def restore(datasets)
+        store = storage_path(datasets)
+        if File.file?(store)
+          mv store, @database_path
+          ActiveRecord::Base.establish_connection 'test'
+          true
+        end
+      end
+      
+      def storage_path(datasets)
+        "#{@storage_path}/#{datasets.collect {|c| c.__id__}.join('_')}.sqlite3.db"
+      end
+    end
+  end
+end

data/lib/dataset/extensions/cucumber.rb

@@ -0,0 +1,20 @@
+module Dataset
+  module Extensions # :nodoc:
+
+    module CucumberWorld # :nodoc:
+      def dataset(*datasets, &block)
+        add_dataset(*datasets, &block)
+
+        load = nil
+        $__cucumber_toplevel.Before do
+          load = dataset_session.load_datasets_for(self.class)
+          extend_from_dataset_load(load)
+        end
+        # Makes sure the datasets are reloaded after each scenario
+        Cucumber::Rails.use_transactional_fixtures
+      end
+    end
+
+  end
+end
+Cucumber::Rails::World.extend Dataset::Extensions::CucumberWorld

data/lib/dataset/extensions/rspec.rb

@@ -0,0 +1,21 @@
+module Dataset
+  module Extensions # :nodoc:
+    
+    module RSpecExampleGroup # :nodoc:
+      def dataset(*datasets, &block)
+        add_dataset(*datasets, &block)
+        
+        load = nil
+        before(:all) do
+          load = dataset_session.load_datasets_for(self.class)
+          extend_from_dataset_load(load)
+        end
+        before(:each) do
+          extend_from_dataset_load(load)
+        end
+      end
+    end
+    
+  end
+end
+Spec::Example::ExampleGroup.extend Dataset::Extensions::RSpecExampleGroup

data/lib/dataset/extensions/test_unit.rb

@@ -0,0 +1,60 @@
+module Dataset
+  class TestSuite # :nodoc:
+    def initialize(suite, test_class)
+      @suite = suite
+      @test_class = test_class
+    end
+    
+    def dataset_session
+      @test_class.dataset_session
+    end
+    
+    def run(result, &progress_block)
+      if dataset_session
+        load = dataset_session.load_datasets_for(@test_class)
+        @suite.tests.each { |e| e.extend_from_dataset_load(load) }
+      end
+      @suite.run(result, &progress_block)
+    end
+    
+    def method_missing(method_symbol, *args)
+      @suite.send(method_symbol, *args)
+    end
+  end
+  
+  module Extensions # :nodoc:
+    
+    module TestUnitTestCase # :nodoc:
+      def self.extended(test_case)
+        class << test_case
+          alias_method_chain :suite, :dataset
+        end
+      end
+      
+      def suite_with_dataset
+        Dataset::TestSuite.new(suite_without_dataset, self)
+      end
+      
+      def dataset(*datasets, &block)
+        add_dataset(*datasets, &block)
+        
+        # Unfortunately, if we have rspec loaded, TestCase has it's suite method
+        # modified for the test/unit runners, but uses a different mechanism to
+        # collect tests if the rspec runners are used.
+        if included_modules.find {|m| m.name =~ /ExampleMethods\Z/}
+          load = nil
+          before(:all) do
+            load = dataset_session.load_datasets_for(self.class)
+            extend_from_dataset_load(load)
+          end
+          before(:each) do
+            extend_from_dataset_load(load)
+          end
+        end
+      end
+    end
+    
+  end
+end
+
+Test::Unit::TestCase.extend Dataset::Extensions::TestUnitTestCase

data/lib/dataset/instance_methods.rb

@@ -0,0 +1,10 @@
+module Dataset
+  module InstanceMethods # :nodoc:
+    def extend_from_dataset_load(load)
+      load.dataset_binding.install_block_variables(self)
+      self.extend load.dataset_binding.record_methods
+      self.extend load.dataset_binding.model_finders
+      self.extend load.helper_methods
+    end
+  end
+end