It’s often helpful to see how our UI’s scale with lots of data in them but seeding data manually in very slow. So using factory_bot_rails and Faker I’ll show you how to make a whole lot of realistic data quickly.
First step install the gems. Add these gems to your Gemfile:
Next we’ll need to setup our factories. A factory is an easily called snippet that creates a model for you. In general they are found in the test/
spec/
and factories/
directories. I generally put mine in spec/factories/model_name.rb
. So for an an example model of a user, we might provide a firstname, lastname, Department (a belongs to relationship), and username.
But this isn’t great as it will create the same user every time. What if we want each user to be unique?
Enter Faker. Faker allows you to generate either deterministic (keeps your test cases consistent) to random-ish data of many many many types. Faker can generate names, addresses, phone numbers, countries, almost whatever you need. Here’s how you call it.
And to see a full list of Faker’s see this list.
Now combining the two ideas we can do something like this:
And with a little bit of ruby-foo outside of Factory Bot or Faker you can randomize the relationship by selecting a random # between 0 and the number of departments - 1.
Okay now that we have factories all set up lets call them from our seeds. We can either build or create a factory. Building a factory creates the model and fills the attributes, creating it does the same but it also saves. Here’s what your seeds.rb could look like: