dm-sweatshop is a model factory for DataMapper. It makes it easy & painless to crank out complex pseudo random models – useful for tests and seed data. Production Goals:
-
Easy generation of random models with data that fits the application domain.
-
Simple syntax for declaring and generating model patterns.
-
Add context to model patterns, allowing grouping and
-
Effortlessly generate or fill in associations for creating complex models with few lines of code.
DataMapper Sweatshop is built around idea of storing attribute hashes associated with a particular class. For instance, you can store two attribute hashes named :without_password and :without_email, associated with a Person class. Later, when in the test you need a Person instance without password or email, you use DataMapper Sweatship helper methods to pick an object that has attributes set you need.
So the workflow is usually the following:
-
Figure out what sets of attributes you need for a good test coverage.
-
Name those sets.
-
Store them associated with a particular class.
-
Use them or objects with those attributes in your tests.
But there’s more. Two hard parts of working with Ruby code fixtures are associations and generation of test data. Dummy data like “foo” and “bar” not just very readable and becomes a mess after a while, it’s really annoying to generate a few objects that have, for instance, a title of 20+ characters.
DataMapper Sweatshop to the rescue. It uses RandExp gem to generate you strings from regular expressions. When you need an email that is 60 characters long, you can relax and use something like “#{/w{58}/.gen}@somedomain.info” instead of typing 58 characters long foobar string.
Another nice thing is associations. Say we want to have say 20 tags for a document or 10 orders for account in tests. DataMapper Sweatshop lets us use associations list in attributes hashes described earlier.
Starting off with a simple user model.
class User include DataMapper::Resource property :id, Serial property :username, String property :email, String property :password, String end
A fixture for the user model can be defined using the @fixture@ method.
User.fixture {{ :username => (username = /\w+/.gen), :email => "#{username}@example.com", :password => (password = /\w+/.gen), :password_confirmation => password # The /\w+/.gen notation is part of the randexp gem: # http://github.com/benburkert/randexp/ }}
Notice the double curly brace (@{{@), a quick little way to pass a block that returns a hash to the fixture method. This is important because it ensures the data is random when we generate a new instance of the model, by calling the block every time.
Code snippet above stores a Proc that returns attributes hash in model map for class User. Since you did not explicitly specify fixture name, default name is used (99+% of the cases it is :default).
You can access that attributes hash later in your tests, make objects with those attributes, and use and abuse it any way you want. It’s just a way to memoize attributes set associated with a particular class.
And here’s how you generate said model.
User.generate
That’s it. In fact, it can even be shortened.
User.gen
But what if we want to use some name for that attributes set? Just pass an argument to @fixture@ method like this:
Person.fixture(:valid) {{ :first_name => %w(Michael Adam Guiseppe)[rand(3)], :last_name => %w(Smith Black White)[rand(3)], :email => "#{/\w{10}/.gen}@somedomain.info", :password_salt => (salt = /\w{20}/.gen), :password_hash => Digest::SHA1.hexdigest("#{salt}@--,-`--secret") }}
Now to a model that has given attributes, use
Person.gen(:valid)
@generate@ (or @gen@) method uses @create@ method of DataMapper models. This means that validations are run on the model. There are two other methods you can use to create data - @make@ to build a model that has not been saved, and @generate!@ to force saving of the model even if it is invalid (it uses @create!@ internally).
Person.make(:valid) Person.generate!(:invalid) # You can also use #gen!
The real power of sweatshop is generating working associations.
DataMapper.setup(:default, "sqlite3::memory:") class Tweet include DataMapper::Resource property :id, Serial property :message, String, :length => 140 property :user_id, Integer belongs_to :user has n, :tags, :through => Resource end class Tag include DataMapper::Resource property :id, Serial property :name, String has n, :tweets, :through => Resource end class User include DataMapper::Resource property :id, Serial property :username, String has n, :tweets end DataMapper.auto_migrate! User.fix {{ :username => /\w+/.gen, :tweets => 500.of {Tweet.make} }} Tweet.fix {{ :message => /[:sentence:]/.gen[0..140], :tags => (0..10).of {Tag.make} }} Tag.fix {{ :name => /\w+/.gen }} # now lets generate 100 users, each with 500 tweets. Also, the tweet's have 0 to 10 tags! users = 10.of {User.gen}
That’s going to generate alot of tags, way more than you would see in the production app. Let’s recycle some already generated tags instead.
User.fix {{ :username => /\w+/.gen, :tweets => 500.of {Tweet.make} }} Tweet.fix {{ :message => /[:sentence:]/.gen[0..140], :tags => (0..10).of {Tag.pick} #lets pick, not make this time }} Tag.fix {{ :name => /\w+/.gen }} 50.times {Tag.gen} users = 10.of {User.gen}
You can add multiple fixtures to a mode, dm-sweatshop will randomly pick between the available fixtures when it generates a new model.
Tweet.fix {{ # a @reply for some user :message => /\@#{User.pick.name} [:sentence:]/.gen[0..140], :tags => (0..10).of {Tag.pick} }}
To keep track of all of our new fixtures, we can even give them a context.
Tweet.fix(:at_reply) {{ :message => /\@#{User.pick.name} [:sentence:]/.gen[0..140], :tags => (0..10).of {Tag.pick} }} Tweet.fix(:conversation) {{ :message => /\@#{(tweet = Tweet.pick(:at_reply)).user.name} [:sentence:]/.gen[0..140], :tags => tweet.tags }}
Sometimes you will want to change one of your fixtures a little bit. You can create a new fixture with a whole new context, but this can be overkill. The other option is to specify attributes in the call to generate.
User.gen(:username => 'datamapper') #uses 'datamapper' as the user name instead of the randomly generated word
This works with contexts too.
User.gen(:conversation, :tags => Tag.all) #a very, very broad conversation
Data for fields with a uniqueness constraint (for example, e-mail addresses) can be generated using the @unique@ method. The simplest usage is to guarantee that random data is unique - wrap your generator in a @unique@ block with no parameters, and the block will be repeatedly executed until it generates a unique value (don’t worry, it raises after a few tries).
For repeatable data, provide a block with one parameter. An incrementing value will be passed in on each invocation of that block. You can also name a unique block to override the block’s identity (yeah that sentence is dense, just see the examples).
include DataMapper::Sweatshop::Unique # Use DataMapper::Sweatshop.unique if you don't want to pollute your namespace
User.fix {{
:name => unique { /\w+/.gen }
:email => unique {|x| "person-#{x}@example.com" }
}}
[User.gen.email, User.gen.email]
# => ["[email protected]", "[email protected]"]
names = ['bob', 'tom', 'bob']
Person.fix {{
:name => (name = names.shift)
:email => unique(name) {|x| "#{name}-#{x}@example.com" }
}}
[Person.gen.email, Person.gen.email, Person.gen.email]
# => ["[email protected]", "[email protected]", "[email protected]"]
The suggested way to use dm-sweatshop with test specs is to create a spec/spec_fixtures.rb file, then declare your fixtures in there. Next, @require@ it in your @spec/spec_helper.rb@ file, after your models have loaded.
Merb.start_environment(:testing => true, :adapter => 'runner', :environment => ENV['MERB_ENV'] || 'test') require 'dm-sweatshop' require File.join(File.dirname(__FILE__), 'spec_fixtures')
Add the .generate calls in your before setup. Make sure to clear your tables or auto_migrate your models after each spec!
Enforce validations at generation time, before the call to @new@/@create@.
Add multiple contexts to pick, or an ability to _fall back_ if one context has no generated models.