Create ActiveRecord objects without step definitions

cucumber_factory allows you to create ActiveRecord objects directly from your Cucumber features. No step definitions required.

Basic usage

To create a new record with default attributes, begin any step with Given there is:

Given there is a movie

To create the record, cucumber_factory will call FactoryBot.create(:movie), FactoryGirl.create(:movie), Movie.make, Movie.create! or Movie.new, depending on what's available.

Quoted strings and numbers denote attribute values:

Given there is a movie with the title "Sunshine" and the year 2007

To update an existing record, specify the record and the changes:

Given the movie above has the title "Sunset" and the year 2008
Given the movie "Sunrise" has the year 2009

A record can be specified by the above keyword, which uses the last created record of this class, or by any string that was used during its creation.

Setting boolean attributes

Boolean attributes can be set by appending which, that or who at the end:

Given there is a movie which is awesome
And there is a movie with the name "Sunshine" that is not a comedy
And there is a director who is popular

Instead of and you can also use but and commas to join sentences:

Given there is a movie which is awesome, popular and successful but not science fiction
And there is a director with the income "500000" but with the account balance "-30000"

To update boolean attributes use the keyword is:

Given the movie above is awesome but not popular
Given the movie above has the year 1979 but is not science fiction

Setting many attributes with a table

If you have many attribute assignments you can use doc string or data table:

Given there is a movie with these attributes:
  """
  name: Sunshine
  comedy: false
  """

Given there is a movie with these attributes:
  | name   | Sunshine |
  | comedy | false    |

Given the movie above has these attributes:
  """
  name: Sunshine
  comedy: false
  """

Setting associations

You can set belongs_to and transient associations by referring to the last created record of as above:

Given there is a movie with the title "Before Sunrise"
And there is a movie with the prequel above

The example above also shows how to set has_many associations - you simply set the belongs_to association on the other side.

You can also refer to a previously created record using any string attribute used in its creation:

Given there is a movie with the title "Before Sunrise"
And there is a movie with the title "Limitless"
And there is a movie with the prequel "Before Sunrise"

You can also explicitly give a record a name and use it to set a belongs_to association below:

Given "Before Sunrise" is a movie
And there is a movie with the title "Limitless"
And there is a movie with the prequel "Before Sunrise"

Note that in the example above, "Before Sunrise" is only a name you can use to refer to the record. The name is not actually used for the movie title, or any other attribute value.

It is not possible to define associations in doc string or data table, but you can combine them in one step:

Given there is a movie with the prequel above and these attributes:
  """
  name: Sunshine
  comedy: false
  """

Given there is a movie with the prequel above and these attributes:
  | name   | Sunshine |
  | comedy | false    |

Setting array attributes or has_many associations

You can set has_many associations by referring to multiple named records in square brackets:

Given there is a movie with the title "Sunshine"
And there is a movie with the title "Limitless"
And there is a movie with the title "Salt"
And there is a user with the favorite movies ["Sunshine", "Limitless" and "Salt"]

When using PostgreSQL array columns, you can set an array attribute to a value with square brackets:

Given there is a movie with the tags ["comedy", "drama" and "action"]

Setting file attributes

You can set an attribute to a file object with the following syntax:

Given there is a movie with the image file:'path/to/image.jpg'

All paths are relative to the project root, absolute paths are not supported. Please note that file attributes must follow the syntax file:"PATH", both single and double quotes are allowed.

Using named factories and traits

You can use a FactoryBot child factory or Machinist named blueprint by putting the variant name in parentheses:

Given there is a movie (comedy) with the title "Groundhog Day"

You can use FactoryBot traits by putting the traits in parentheses, as a comma-separated list:

Given there is a movie (moody, dark) with the title "Interstellar"

Overriding factory steps

If you want to override a factory step with your own version, just do so:

Given /^there is a movie with good actors$/ do
  movie = Movie.make
  movie.actors << Actor.make(:name => 'Clive Owen')
  movie.actors << Actor.make(:name => 'Denzel Washington')
end

Custom steps will always be preferred over factory steps. Also Cucumber will not raise a warning about ambiguous steps if the only other matching step is a factory step. Thanks, cucumber_priority!

Supported Cucumber versions

cucumber_factory is tested against Cucumber 1.3, 2.4, 3.0, 3.1, 4.1, 5.3, 9.2 and 10.1.

Installation

In your Gemfile say:

gem 'cucumber_factory'

Now create a file features/step_definitions/factory_steps.rb, which just says

require 'cucumber_factory/add_steps'

Now run bundle install and restart your server.

Development

There are tests in spec. We only accept PRs with tests. To run tests:

• Install the Ruby version stated in .ruby-version
• Create a local PostgreSQL database:

$ sudo -u postgres psql -c 'create database cucumber_factory_test;'

• Copy spec/support/database.sample.yml to spec/support/database.yml and enter your local credentials for the test databases
• Install development dependencies using bundle install
• Run tests with the default symlinked Gemfile using bundle exec rspec or explicit with BUNDLE_GEMFILE=Gemfile.cucumber-x.x bundle exec rspec spec

We recommend to test large changes against multiple versions of Ruby and multiple dependency sets. Supported combinations are configured in .github/workflows/test.yml. We provide some rake tasks to help with this:

For each ruby version do (you need to change it manually):

• Install development dependencies using rake matrix:install
• Run tests using rake matrix:spec
• Shorthand:

  for version in "2.5.3" "2.6.6" "2.7.2" "3.2.0" "3.4.1"; do rbenv shell $version && bundle install && bundle exec rake matrix:install && bundle exec rake matrix:spec; done

Note that we have configured GitHub Actions to automatically run tests in all supported Ruby versions and dependency sets after each push. We will only merge pull requests after a green workflow build.

If you would like to contribute:

• Fork the repository.
• Push your changes with passing specs.
• Send us a pull request.

I'm very eager to keep this gem leightweight and on topic. If you're unsure whether a change would make it into the gem talk to me beforehand.

Credits

Henning Koch from makandra