Laziest Developer

= Usage

== Databases

=== rake db:migrate

Everyone I know forgets to clone the test database when they make changes, so I decided to override rake:db:migrate so that when you run it, it automatically clones the test database, which is really useful if you run tests individually via TextMate or through the command line.

rake db:migrate rake db:migrate:up rake db:migrate:down

are all supported. As of version 1.2.0, this task calls rake db:test:prepare whenever invoked.

=== rake db:migrate:compact Uses schema.rb to construct a brand new clean migration and moves all of your existing migrations out of the way. This might at first seem to violate the spirit of migrations, but honestly sometimes large projects have a LOT of migrations and it can get difficult to manage these files.

=== rake db:remigrate

Sometimes you just need to wipe out your database tables and start over. This task drops your tables and starts over by dropping the tables directly and then re-running your migrations. This provides a great way to test to make sure you haven't broken migrations at some point, which will happen to you at some point.

=== rake db:to_yaml

Dump your database to fixtures. Stores them in Rails.root/production_data. You can then use this to load the data back into another database, even one of a different type. We've used this to move data from SQL Server to MySQL and back again.

Optionally you can specify only the models you want to dump data for by passing a list of model names as a comma delimited list to the MODELS environment variable.

For example, if I wanted to dump data for the User, Role, and RoleUsers models and not the rest of the models in my app, any of the following would work:

rake db:to_yaml MODELS=User,Role,RoleUser #Model names rake db:to_yaml MODELS=user.rb,role.rb,role_user.rb #File names rake db:to_yaml MODELS=user,role,role_user #File names minus their extension rake db:to_yaml MODELS=users,roles,role_users #DB table names

=== rake db:from_yaml Load fixtures from Rails.root/production_data into your database. Loads fixtures dumped by using rake db:export

== Cleaning Up Files

===Nuke The Nuke tasks make cleaning up your project a breeze.

Let's say a previous developer generated some controllers, models, helpers, and views using scaffolding and a healthy mix of broken Test::Unit test cases. You've already generated a nice RESTful controller, correctly pluralized while the original ones are not. When you look in your controllers folder, you see this:

account_controller.rb accounts_controller.rb

If you want to clean this up, you'd have to delete these files like so:

rm -rf app/controllers/account_controller.rb rm -rf app/helpers/account_helper.rb rm -rf app/views/account rm -rf test/functionals/account_controller_test.rb

Well, with the Nuke tasks, you can do this easily:

rake nuke:vc:account

The 'vc' is for view and controller. It'll take out the whole controller and view combination, including the helper too. It'll blow away any specs or tests associated, and it'll detect if you're using svn or git, so instead of deleting the files directly, it will schedule them for deletion via your source control system.

The nuke tasks can be quite granular.

• rake nuke:model:user / rake nuke:m:user
• rake nuke:controller:user / rake nuke:c:user
• rake nuke:views:user / rake nuke:v:user
• rake nuke:helper:user
• rake nuke:vc:user
• rake nuke:mvc:user

The only thing these tasks won't do is remove migrations, and that's cos I think that's dangerous

== Subversion

=== rake svn:root Displays the root of your repository

=== rake svn:tags Displays all the tags. Assumes you use a tags/ folder and a /trunk folder

=== rake svn:tags:last Displays the last tag.

=== rake svn:tag TAG=rel_1-0-0 Creates a new tag from the trunk.

== Rcov We're fans of RCov and we've included some ways to make it easier to use RCov in your projects.

=== Test::Unit

In order to use the Test::Unit coverage tasks, you'll need to install the Rails_rcov plugin

ruby script/plugin install http://svn.codahale.com/rails_rcov/

This is not needed if your project uses Rspec.

==== rake test:models:rcov

Runs coverage on your models

=== rake test:controllers:rcov

Runs coverage on your controllers

==== rake test:rcov:full

Runs coverage on models and controllers

=== RSpec

Rspec already includes the ability to get code coverage, but we made it just as focused and granular as we made the tasks for Test::Unit

==== rake spec:models:rcov

Coverage for models

==== rake spec:controllers:rcov

Coverage for controllers

==== rake spec:views:rcov

Coverage for views

==== rake spec:helpers:rcov

Coverage for helpers

==== rake spec:lib:rcov

Coverage for files in lib/

== Test::Unit

=== Running individual tests

Based on an idea from Geoffrey Grosenbach, you can run all tests in unitsuser_test.rb by doing

rake test:units:user:all

Or run a specific test by specifying all or part of a name. For example, if I wanted to run the "test_create" test case, I would use

rake test:units:user:create

The same rules apply to functional tests

rake test:functionals:users:list

== RSpec

RSpec gets some love here. If you need nice output for your specs, we overrode some of the built-in specs.

=== rake spec:models

Runs all model specs

=== rake spec:controllers

Runs all controller specs

=== Running a specific model or controller spec

rake spec:model:user rake spec:controller:sessions

== Plugins

Create a file in your Home folder called .plugins and you can have all of your favorite plugins easily installed by doing

rake rails:install:plugins

Just put each plugin name or repository on its own line.

Windows users need to set the HOME environment variable.

== Cleaning Up

Run

rake rails:clear

to clean up tmp, logs, and docs in one easy command.

AUTHORS

Brian P. Hogan and Kevin Gisi