= dm-rails This gem provides the railtie that allows {datamapper}[http://github.com/datamapper/dm-core] to hook into {rails3}[http://github.com/rails/rails] and thus behave like a rails framework component. Just like activerecord does in rails, {dm-rails}[http://github.com/datamapper/dm-rails] uses the railtie API to hook into rails. The two are actually hooked into rails almost identically. Creating new datamapper apps on rails3 from scratch is actually really easy. The following will guide you through the process. == Generating a new application from scratch To go from zero gems to a working rails3 app using datamapper, all you need are the latest {rubygems}[http://docs.rubygems.org/read/chapter/3] release, {bundler}[http://gembundler.com] and {rails}[http://github.com/rails/rails]. After possibly updating your rubygems, issue the following: gem install rails Once you have {rails}[http://github.com/rails/rails] and thus {bundler}[http://github.com/carlhuda/bundler] installed, you can bootstrap a rails application with a single command. rails new project_name -m http://datamapper.org/templates/rails.rb When run, the command will print out some options on how to proceed with your newly generated application. Be aware that this command executes code loaded from the internet! It is currently the simplest way to bootstrap a new application, and a commonly used way to reuse rails application templates. Have a look at {rails templates}[http://github.com/datamapper/datamapper.github.com/tree/master/templates/] to know exactly what we'll do to your system, but be aware that a man-in-the-middle attack could alter that. == rspec support I haven't yet tested rspec support extensively, but the basics are working after a few modifications to the spec/spec_helper.rb file that the necessary rails generate rspec:install gives us. First you need to uncomment/remove a seemingly invalid rspec configuration option that gets set by default, and you also should turn off transactional_fixtures at least for now. It's pretty common to run `DataMapper.auto_migrate!` before all specs too. My spec_helper.rb file currently looks like the following, and model specs run fine. # This file is copied to ~/spec when you run 'ruby script/generate rspec' # from the project root directory. ENV["RAILS_ENV"] ||= 'test' require File.expand_path("../../config/environment", __FILE__) require 'rspec/rails' # Requires supporting files with custom matchers and macros, etc, # in ./support/ and its subdirectories. Dir["#{File.dirname(__FILE__)}/support/**/*.rb"].each {|f| require f} RSpec.configure do |config| # == Mock Framework # # If you prefer to use mocha, flexmock or RR, uncomment the appropriate line: # # config.mock_with :mocha # config.mock_with :flexmock # config.mock_with :rr config.mock_with :rspec config.before(:suite) { DataMapper.auto_migrate! } end With this in place, running ./bin/rake spec ./bin/rake spec:models ... should work fine. In order to have ./bin/rake available, the bundle needs to be installed with bundle install --binstubs If you haven't done so for whatever reason, use bundle exec rake spec bundle exec rake spec:models instead. You may have noticed that rspec is included into both the `:development` and `:test` groups in the Gemfile the application template generated. I couldn't find another way to make it so that the above mentioned way of running the specs work. If rspec is not included into the `:development` group, rails must be informed explicitly about the environment it should run in, by prefixing the above rake commands with `RAILS_ENV=test`. == Developing dm-rails Issue the following commands to get a fully functional development environment including datamapper and rails up and running within a minute. gem install bundler # if this is new for you bundle install Whenever you want to update your dependencies because you want to make sure you develop against master branches, just issue bundle install again and {bundler}[http://github.com/carlhuda/bundler] will go ahead and fetch the latest commits from the gems you depend on. == Sample Gemfile Using {bundler}[http://github.com/carlhuda/bundler] it's really easy to get an app going with datamapper and rails3. Just use a Gemfile like this, and bundler will pull in everything needed to run your app. Note that you also must add any additional datamapper plugin or any other gem that you'd like to use to the Gemfile. This makes sure that bundler is able to provide a complete environment containing all required dependencies for your app. source 'http://rubygems.org' RAILS_VERSION = '~> 3.0.4' DM_VERSION = '~> 1.1.0' RSPEC_VERSION = '~> 2.5.0' gem 'activesupport', RAILS_VERSION, :require => 'active_support' gem 'actionpack', RAILS_VERSION, :require => 'action_pack' gem 'actionmailer', RAILS_VERSION, :require => 'action_mailer' gem 'railties', RAILS_VERSION, :require => 'rails' gem 'dm-rails', '~> 1.1.0' gem 'dm-sqlite-adapter', DM_VERSION # You can use any of the other available database adapters. # This is only a small excerpt of the list of all available adapters # Have a look at # # http://wiki.github.com/datamapper/dm-core/adapters # http://wiki.github.com/datamapper/dm-core/community-plugins # # for a rather complete list of available datamapper adapters and plugins # gem 'dm-sqlite-adapter', DM_VERSION # gem 'dm-mysql-adapter', DM_VERSION # gem 'dm-postgres-adapter', DM_VERSION # gem 'dm-oracle-adapter', DM_VERSION # gem 'dm-sqlserver-adapter', DM_VERSION gem 'dm-migrations', DM_VERSION gem 'dm-types', DM_VERSION gem 'dm-validations', DM_VERSION gem 'dm-constraints', DM_VERSION gem 'dm-transactions', DM_VERSION gem 'dm-aggregates', DM_VERSION gem 'dm-timestamps', DM_VERSION gem 'dm-observer', DM_VERSION group(:test) do gem 'rspec-rails', RSPEC_VERSION # To get a detailed overview about what queries get issued and how long they take # have a look at rails_metrics. Once you bundled it, you can run # # rails g rails_metrics Metric # rake db:automigrate # # to generate a model that stores the metrics. You can access them by visiting # # /rails_metrics # # in your rails application. # gem 'rails_metrics', '~> 0.1', :git => 'git://github.com/engineyard/rails_metrics' end == Sample database.yml files DataMapper supports connecting to and working with multiple repositories easily. In order to be able to take full advantage of that feature in rails, you can configure as many repositories for your different environments as you wish. All you need to do is follow some simple naming conventions and you're good to go. Have a look at the #{adapter}_defaults declarations in the sample files below. That's the only convention you need to follow. Your default declarations should always end with "defaults". This is necessary for dm-rails to not confuse these with any of your environment declarations. In fact, dm-rails looks at the content of your database.yml and rejects every key that matches /defaults/. The remaining entries represent the repository configurations for the available environments. An example for setting up a single repository for every environment. This will setup an implicit 'default' repository. defaults: &defaults adapter: mysql username: root password: host: localhost development: database: rails3_app_development <<: *defaults test: database: rails3_app_test <<: *defaults production: database: rails3_app_production <<: *defaults An example of setting up a single repository with a URI, which overrides all other configuration for that environment: development: uri: mysql://localhost/rails3_app_development?user=mysql An example for setting up multiple repositories for every environment. mysql_defaults: &mysql_defaults adapter: mysql username: mysql_user password: mysql_sekret host: localhost postgres_defaults: &postgres_defaults adapter: postgres username: postgres_user password: postgres_sekret host: postgres_host oracle_defaults: &oracle_defaults adapter: oracle username: oracle_user password: oracle_sekret host: oracle_host development: database: rails3_mysql_development <<: *mysql_defaults repositories: foo: database: rails3_postgres_development <<: *postgres_defaults bar: database: rails3_oracle_development <<: *oracle_defaults test: database: rails3_mysql_test <<: *mysql_defaults repositories: foo: database: rails3_postgres_test <<: *postgres_defaults bar: database: rails3_oracle_test <<: *oracle_defaults production: database: rails3_mysql_production <<: *mysql_defaults repositories: foo: database: rails3_postgres_production <<: *postgres_defaults bar: database: rails3_oracle_production <<: *oracle_defaults Once you have defined your database.yml file, dm-rails's rake tasks will be able to create, drop, auto_migrate! and auto_upgrade! all your defined repositories. == Available generators Since the new generators provide hooks that allow the dm-rails gem to provide generators that hook into parts like model, scaffolds or test generation, everything just works like it does with active_record. The following generators are available to help you get started with the typical components of any rails application. ... rails generate controller rails generate generator rails generate helper rails generate integration_test rails generate migration rails generate model rails generate observer rails generate performance_test rails generate plugin rails generate resource rails generate scaffold rails generate scaffold_controller rails generate session_migration rails generate stylesheets ... For a complete list run the following in your project's directory rails generate == Available datamapper specific rake tasks To get a list of all available rake tasks in your rails3 app, issue the usual rake -T Once you do that, you will see the following rake tasks among others. These are the ones that dm-rails added for us. ... rake db:automigrate # Perform destructive automigration of all repositories in the current Rails.env rake db:autoupgrade # Perform non destructive automigration of all repositories in the current Rails.env rake db:create # Create the database(s) defined in config/database.yml for the current Rails.env - also creates the test database(s) if Rails.env.development? rake db:create:all # Create all the local databases defined in config/database.yml rake db:drop # Drops the database(s) for the current Rails.env - also drops the test database(s) if Rails.env.development? rake db:drop:all # Drop all the local databases defined in config/database.yml rake db:migrate # Migrate the database to the latest version rake db:migrate:down[version] # Migrate down using migrations rake db:migrate:up[version] # Migrate up using migrations rake db:seed # Load the seed data from db/seeds.rb rake db:sessions:clear # Clear the sessions table for DataMapperStore rake db:sessions:create # Creates the sessions table for DataMapperStore rake db:setup # Create the database, load the schema, and initialize with the seed data ... == Configuring and introspecting dm-rails Rails3 makes it easy to expose framework component specific configuration to application developers in a uniform and easy to use way. To build on this philosophy, dm-rails exposes its configuration via a single object that is used throughout dm-rails's code to store configuration relevant to datamapper and to rails. You can access it from within your application and of course alter it's settings in your config/application.rb or config/environments files. Here's a quick overview of the API the configuration object exposes. Expect this to grow as we come up with additional useful stuff to configure. Rails::DataMapper.configuration Rails::DataMapper::Configuration.for(database_yml_hash) Rails::DataMapper::Configuration#raw Rails::DataMapper::Configuration#environments Rails::DataMapper::Configuration#repositories As promised, you can easily inspect the configuration from your running rails application. Let's start a rails console and have a look what's configured for our app. ree-1.8.7-2010.01 mungo:alfred snusnu$ rails console Loading development environment (Rails 3.0.0.beta4) ruby-1.8.7-p248 > require 'pp' => ["PP"] ruby-1.8.7-p248 > pp Rails::DataMapper.configuration # {"adapter"=>"mysql", "username"=>"root", "database"=>"alfred_production", "host"=>"localhost", "password"=>nil}, "development"=> {"adapter"=>"mysql", "username"=>"root", "database"=>"alfred_development", "host"=>"localhost", "password"=>nil}, "defaults"=> {"username"=>"root", "adapter"=>"mysql", "host"=>"localhost", "password"=>nil}, "test"=> {"adapter"=>"mysql", "username"=>"root", "database"=>"alfred_test", "host"=>"localhost", "password"=>nil}}, @repositories= {"production"=> {"default"=> {"adapter"=>"mysql", "username"=>"root", "database"=>"alfred_production", "host"=>"localhost", "password"=>nil}}, "development"=> {"default"=> {"adapter"=>"mysql", "username"=>"root", "database"=>"alfred_development", "host"=>"localhost", "password"=>nil}}, "test"=> {"default"=> {"adapter"=>"mysql", "username"=>"root", "database"=>"alfred_test", "host"=>"localhost", "password"=>nil}}}, @root=#> => nil Additionally, you can reach the configuration object via the standard way that rails provides to expose configuration for framework components and plugins. Rails.application.config.data_mapper This will give you the exact same object we inspected in the previous pretty print output from rails console. == Extending dm-rails It's easy to extend or adapt dm-rails to meet your specific needs. Thanks to the railties API it's possible to hook into any part of the initialization process. In order to customize dm-rails, all you need to do is define your own Rails::Railtie class that inherits from Rails::DataMapper::Railtie and require that instead of the standard dm-rails/railtie. During initialization of any rails plugin, the initializers defined by that plugin are run in the order specified. Since dm-rails defines every action that gets called by its initializers as a method on either the Railtie instance or the class, you can just go ahead and overwrite these in your subclass. All those methods get called with the running Rails::Application (always reachable via Rails.application) as single parameter, so you can customize depending on the app's state as much as you wish. Additionally, the initializers are all named, and you can hook your own additional initializers before or after any of the named rails (or dm-rails) initializers. To give you an idea of what you get when inheriting from Rails::DataMapper::Railtie have a look at the list of methods provided by that object. Rails::DataMapper::Railtie.configure_data_mapper(app) Rails::DataMapper::Railtie.setup_i18n_support(app) Rails::DataMapper::Railtie#setup_controller_runtime(app) Rails::DataMapper::Railtie#setup_logger(app) To complete the picture of dm-rails's initialization process, here's an overview of the defined initializers in the order they are called by rails during bootup. Note that every one of these initializers does one single thing; calling one of the methods listed above. This makes it easy to customize each of these steps by overwriting the respective method. initializer 'data_mapper.configuration' do |app| configure_data_mapper(app) end initializer 'data_mapper.logger' do |app| setup_logger(app, Rails.logger) end initializer 'data_mapper.i18n_support' do |app| setup_i18n_support(app) end # Expose database runtime to controller for logging. initializer "data_mapper.log_runtime" do |app| setup_controller_runtime(app) end # Preload all models once in production mode, # and before every request in development mode initializer "datamapper.add_to_prepare" do |app| config.to_prepare { Rails::DataMapper.preload_models(app) } end # Run setup code once in after_initialize to make sure all initializers # are in effect once we setup the connection. Also, this will make sure # that the connection gets set up after all models have been loaded, # because #after_initialize is guaranteed to run after #to_prepare. # Both production and development environment will execute the setup # code only once. config.after_initialize do |app| Rails::DataMapper.setup(Rails.env) end If you want to add additional rake tasks in your extension, you can do so by adding the following to your railtie. rake_tasks do load 'path/to/your/tasks.rake' end == Identity Map support Activating the identity map is achieved by installing a middleware that wraps the whole request inside a block DataMapper.repository { ... } Note that this scopes every call to datamapper to the :default repository specified in your database.yml file. If you need to access a different repository from within your actions, just wrap the calls in another DataMapper.repository block. DataMapper stacks the repositories it uses and the innermost will always win. see also next paragraph. In order to activate the Identity Map in your application, you need to explicitly use the provided middleware (or in fact any other middleware that does the job to your liking) in any of your controllers. For example, if you want to enable the Identity Map for all controllers, you would declare to use the middleware in your ApplicationController. require 'dm-rails/middleware/identity_map' class ApplicationController < ActionController::Base use Rails::DataMapper::Middleware::IdentityMap protect_from_forgery end For a different scope then :default of other repositories replace/add the line with: use Rails::DataMapper::Middleware::IdentityMap, :myscope If you've created your application using the official templates at http://datamapper.org/templates/rails.rb this has already been added for you. If for some reason you don't want to enable the (default) Identity Map globally for all controllers, you can either just use the middleware in a few selected controllers (in case you don't have too many controllers needing it), or you can create a controller class that uses the middleware, and inherit from that controller in cases where you need Identity Map support. # app/controllers/identity_map_controller.rb require 'dm-rails/middleware/identity_map' class IdentityMapController < ApplicationController use Rails::DataMapper::Middleware::IdentityMap end # app/controllers/people_controller.rb class PeopleController < IdentityMapController # ... end == Cucumber testing framework support dm-rails ist working fine with the cucumber testing framework out of the box. In order to have transaction support in cucumber, so that the data generated by the different scenarios doesn't cause problems, create a file "RAILS_APP/features/support/datamapper.rb" with the following contents: Before do repository(:default) do |repository| transaction = DataMapper::Transaction.new(repository) transaction.begin repository.adapter.push_transaction(transaction) end end After do repository(:default).adapter.pop_transaction.rollback end This simply wraps each scenario in a transaction which is rolled back. If your underlying database doesn't support transactions, you can instead use this code to clean/ truncate the whole database after each scenario (not dm-rails specific): require 'database_cleaner' require 'database_cleaner/cucumber' DatabaseCleaner.strategy = :truncation == Using adapter specific Resource naming conventions Say you want to namespace your models and don't want your storage names to reflect your module nesting. DataMapper provides easy ways to use just the naming conventions you like. Basically, all you need to do is tell dm-rails and thus DataMapper that you either want to use a predefined naming convention # in some config/initializers/file.rb convention = DataMapper::NamingConventions::Resource::UnderscoredAndPluralizedWithoutModule Rails::DataMapper.configuration.resource_naming_convention[:default] = convention or that you want to use your own naming convention, that is implemented in e.g. a lambda # in some config/initializers/file.rb Rails::DataMapper.configuration.resource_naming_convention[:default] = lambda do |value| 'tbl' + value.camelize(true) end Field names can be customized in the same manner using `Rails::DataMapper.configuration.resource_naming_convention`. For more detailed documentation about DataMapper naming conventions and the ones that are available by default, have a look at http://rdoc.info/projects/datamapper/dm-core and search for _NamingConventions_ in the Class List. == Mass assignment protection By default, `dm-rails` doesn't activate any support for mass assignment protection. You can however activate it for your application by including the relevant module either globally into all known models, or on a per model basis (the latter being advised, for reasons explained below). # Global installation (config/application.rb is a good place for adding this) # # Make .attr_protected and .attr_accessible available to all models # NOTE: This won't work if you have code that includes DataMapper::Resource # into any other module. This is done by dm-is-remixable for example, so to # be safe, you should only do that if you really know what you're doing. # Quite some plugins make use of dm-is-remixable, so be sure to check that out # before you go ahead and do the following in your config/application.rb DataMapper::Model.append_inclusions(Rails::DataMapper::MassAssignmentSecurity) # Local installation (recommended) # # Include the mass assignment protection only into models that actually need it # This is the preferred way of doing things, at least for now. You will only # have the functionality available where you actually need it, and you don't run # into problems with third party code including DataMapper::Resource into other # modules. class Person include DataMapper::Resource include DataMapper::MassAssignmentSecurity property :id, Serial property :login, String attr_protected :login end == Using additional datamapper plugins In order to use additional plugins add them to the Gemfile and require them from inside a file in config/initializers. Once you've done that, update your bundle and you should be ready to use the plugin(s) cd /path/to/your/app # edit Gemfile bundle install Have a look at this application's {Gemfile}[http://github.com/snusnu/datamapper_on_rails3/blob/master/Gemfile] for an idea of how to use gems from git repositories. == Rails notification system Currently dm-rails publishes the same benchmarking information like active_record does. This means that you will get output like this in your log files. Completed in 9ms (Views: 7.6ms | Models: 0.6ms) with 200 == TODO (not necessarily in that order) * SPECS !!! * Think about a release strategy supporting both beta releases and master branch * Further README updates * More work on migrations == Credits Big thanks to everyone working on {datamapper}[http://github.com/datamapper/dm-core], {rails}[http://github.com/rails/rails], {bundler}[http://github.com/carlhuda/bundler] and open source in general. This will be (and actually already is) an awesome platform for developing web applications. == Note on Patches/Pull Requests * Fork the project. * Make your feature addition or bug fix. * Add tests for it. This is important so I don't break it in a future version unintentionally. * Commit, do not mess with rakefile, version, or history. (if you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull) * Send me a pull request. Bonus points for topic branches. == The dm-rails team Thx to all contributors, every patch, big or small is very much appreciated! * Martin Gamsjaeger ({snusnu}[http://github.com/snusnu]) * Dan Kubb ({dkubb}[http://github.com/dkubb]) * Alex Coles ({myabc}[http://github.com/myabc]) * Alex Mankuta ({cheba}[http://github.com/cheba]) * Foy Savas ({foysavas}[http://github.com/foysavas]) * Randall Brewer * Josh Huckabee {jhuckabee}[http://github.com/jhuckabee] * Patrik Sundberg ({sundbp}[http://github.com/sundbp]) * Corin Langosch ({gucki}[http://github.com/gucki]) * Jared Morgan ({jm81}[http://github.com/jm81]) * Blake Gentry ({bgentry}[http://github.com/bgentry]) * Kabari Hendrick ({RipTheJacker}[http://github.com/RipTheJacker]) * Nate Mueller ({natemueller}[http://github.com/natemueller]) * Xavier Shay ({xaviershay}[http://github.com/xaviershay]) * Nico Rieck ({gix}[http://github.com/gix]) * Piotr Solnica ({solnic}[http://github.com/solnic]) * Corin Lawson ({maxsum-corin}[http://github.com/maxsum-corin]) == Copyright Copyright (c) 2011-2010 The dm-rails team. See {LICENSE}[http://github.com/datamapper/dm-rails/blob/master/LICENSE] for details.