Seedbank
Seedbank allows you to structure your apps seed data instead of having it all dumped into one large file. I find my seed data tended to fall into two categories:
- Stuff that the entire application requires.
- Stuff to populate my development and staging environments.
Seedbank assumes that your common seed data is kept under db/seeds and any directories under db/seeds/
are specific to an environment, so db/seeds/development
contains all your development-only seed data.
The reason behind Seedbank is laziness. When I checkout or re-visit a project I don't want to mess around getting my environment setup I just want the code and a database loaded with data in a known state. Since the Rails core team were good enough to give us rake db:setup it would be rude not to use it.
rake db:setup # Create the database, load the schema, and initialize with the seed data (use db:reset to also drop the db first)
To achieve this slothful aim, Seedbank renames the original db:seed rake task to db:seed:original, makes it a dependency for all the Seedbank seeds and adds a new db:seed task that loads all the common seeds in db/seeds plus all the seeds for the current Rails environment.
Although originally built for Rails, Seedbank can work stand alone thanks to Aleksey Ivanov.
Example
Seedbank seeds follow this structure;
db/seeds/
.seeds.rb
development/
users.seeds.rb
foo.seeds.rb
This would generate the following Rake tasks
rake db:seed # Load the seed data from db/seeds.rb, db/seeds/*.seeds.rb and db/seeds/ENVIRONMENT/*.seeds.rb. ENVIRONMENT is the current environment in Rails.env.
rake db:seed:bar # Load the seed data from db/seeds/bar.seeds.rb
rake db:seed:common # Load the seed data from db/seeds.rb and db/seeds/*.seeds.rb.
rake db:seed:development # Load the seed data from db/seeds.rb, db/seeds/*.seeds.rb and db/seeds/development/*.seeds.rb.
rake db:seed:development:users # Load the seed data from db/seeds/development/users.seeds.rb
rake db:seed:original # Load the seed data from db/seeds.rb
Therefore, assuming RAILS_ENV
is not set or it is "development":
$ rake db:seed
will load the seeds in db/seeds.rb
, db/seeds/bar.seeds.rb
, db/seeds/foo.seeds.rb
and db/seeds/development/users.seeds.rb
. Whereas, setting the RAILS_ENV
variable, like so:
$ RAILS_ENV=production rake db:seed
will load the seeds in db/seeds.rb
, db/seeds/bar.seeds.rb
and db/seeds/foo.seeds.rb
.
Installation
Seedbank > 0.5.0 uses refinements and no longer supports rubies below 2.x. If you are using an older Ruby you'll have to stick with 0.4.0 and below.
I have also dropped support for Rubinius and JRuby. I'm happy to accept pull requests for them, but don't have the time to hack together the test environment. If you want to contribute, please ensure that he travis.yml is in line as it's the only way I will test these two environments.
Rails 5.x
Seedbank has not been updated to work with Rails 5. I've used it with 5.x apps and am working on a new version specifically for 5.x. Other people are also reporting using it with no problems.
Rails 4.x and above
Add the seedbank gem to your Gemfile. In Gemfile:
gem "seedbank"
That's it!
Non Rails apps
Although originally built for Rails, Seedbank should work fine in other environments such as Padrino, Grape or the new new hotness. please let us know how you get on.
Rails 3.x
Seedbank 0.5.0 onwards is no longer tested against Rails 3.x, that isn't to say it will not work. I will not fix issues against Rails 3.x, but will accept tested pull requests.
Rails 2.x
Seedbank hasn't supported Rails 2.x for some time. You'll need to use the 0.2.1 version. In your Gemfile:
gem "seedbank", '~> 0.2.1'
Then in the bottom of your application's Rakefile:
require 'seedbank'
Seedbank.load_tasks if defined?(Seedbank)
If you vendor the gem you'll need to change the require to the specific path.
Usage
Seeds files are just plain old Ruby executed in your application environment so anything you could type into the console will work in your seeds. Seeds files have to be named with the '.seeds.rb' extension.
db/seeds/companies.seeds.rb
Company.find_or_create_by_name('Hatch', :url => 'http://thisishatch.co.uk' )
The seed files under db/seeds are run first in alphanumeric order followed by the ones in the db/seeds/RAILS_ENV. You can add dependencies to your seed files to enforce the run order. for example;
db/seeds/users.seeds.rb
after :companies do
company = Company.find_by_name('Hatch')
company.users.create(:first_name => 'James', :last_name => 'McCarthy')
end
db/seeds/projects.seeds.rb
after :companies do
company = Company.find_by_name('Hatch')
company.projects.create(:title => 'Seedbank')
end
db/seeds/tasks.seeds.rb
after :projects, :users do
project = Project.find_by_name('Seedbank')
user = User.find_by_first_name_and_last_name('James', 'McCarthy')
project.tasks.create(:owner => user, :title => 'Document seed dependencies in the README.md')
end
If the dependencies are in one of the environment folders, you need to namespace the parent task:
db/seeds/development/users.seeds.rb
after "development:companies" do
company = Company.find_by_name('Hatch')
company.users.create(:first_name => 'James', :last_name => 'McCarthy')
end
Note - If you experience any errors like Don't know how to build task 'db:seed:users'
. Ensure you are specifying after 'development:companies'
like the above example. This is the usual culprit (YMMV).
Defining and using methods
As seed files are evaluated within a single runner in dependency order, any methods defined earlier in the run will be available across dependent tasks. I recommend keeping method definitions in the seed file that uses them. Alternatively if you have many common methods, put them into a module and extend the runner with the module.
db/seeds/support.rb
module Support
def notify(filename)
puts "Seeding: #{filename}"
end
end
db/seeds/common.seeds.rb
require_relative 'support'
extend Support
notify(__FILE__)
To keep this dry you could make the seeds dependent on a support seed that extends the runner.
db/seeds/users.seeds.rb
after :common do
notify(__FILE__)
end
Contributors
git log | grep Author | sort | uniq
- Ahmad Sherif
- Andy Triggs
- Corey Purcell
- James McCarthy
- Joost Baaij
- Justin Smestad
- Peter Suschlik
- Philip Arndt
- Tim Galeckas
- lulalala
- pivotal-cloudplanner
- vkill
- Aleksey Ivanov
Note on Patches/Pull Request
- 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 it when I pull)
- Send me a pull request. Bonus points for topic branches.
Copyright
Copyright (c) 2011-2017 James McCarthy, released under the MIT license