Seed Do

This project is a fork of seed-fu.

Seed Do is an attempt to once and for all solve the problem of inserting and maintaining seed data in a database. It uses a variety of techniques gathered from various places around the web and combines them to create what is hopefully the most robust seed data system around.

Just add gem 'seed-do' to your Gemfile

For detailed API documentation, please visit: https://rubydoc.info/github/willnet/seed-do

If you are migrating from seed-fu to seed-do, you'll need to make the following changes:

1. Rename all occurrences of SeedFu to SeedDo in your code
2. Update rake task references from rake db:seed_fu to rake db:seed_do

For example:

# Before
SeedFu.seed
rake db:seed_fu

# After
SeedDo.seed
rake db:seed_do

Capistrano integration also needs to be updated:

# Before
require 'seed-fu/capistrano'
# or for Capistrano3
require 'seed-fu/capistrano3'

# After
require 'seed-do/capistrano'
# or for Capistrano3
require 'seed-do/capistrano3'

User.seed do |s|
  s.id    = 1
  s.login = "jon"
  s.email = "jon@example.com"
  s.name  = "Jon"
end

User.seed do |s|
  s.id    = 2
  s.login = "emily"
  s.email = "emily@example.com"
  s.name  = "Emily"
end

$ rake db:seed_do
== Seed from /path/to/app/db/fixtures/users.rb
 - User {:id=>1, :login=>"jon", :email=>"jon@example.com", :name=>"Jon"}
 - User {:id=>2, :login=>"emily", :email=>"emily@example.com", :name=>"Emily"}

Constraints are used to identify seeds, so that they can be updated if necessary. For example:

Point.seed(:x, :y) do |s|
  s.x = 4
  s.y = 7
  s.name = "Home"
end

The first time this seed is loaded, a Point record will be created. Now suppose the name is changed:

Point.seed(:x, :y) do |s|
  s.x = 4
  s.y = 7
  s.name = "Work"
end

When this is run, Seed Do will look for a Point based on the :x and :y constraints provided. It will see that a matching Point already exists and so update its attributes rather than create a new record.

If you do not want seeds to be updated after they have been created, use seed_once:

Point.seed_once(:x, :y) do |s|
  s.x = 4
  s.y = 7
  s.name = "Home"
end

The default constraint just checks the id of the record.

By default, seed files are looked for in the following locations:

• #{Rails.root}/db/fixtures and #{Rails.root}/db/fixtures/#{Rails.env} in a Rails app
• ./db/fixtures when loaded without Rails

You can change these defaults by modifying the SeedDo.fixture_paths array.

Seed files can be named whatever you like, and are loaded in alphabetical order.

When loading lots of records, the above block-based syntax can be quite verbose. You can use the following instead:

User.seed(:id,
  { :id => 1, :login => "jon",   :email => "jon@example.com",   :name => "Jon"   },
  { :id => 2, :login => "emily", :email => "emily@example.com", :name => "Emily" }
)

Seed files can be run automatically using rake db:seed_do. There are two options which you can pass:

• rake db:seed_do FIXTURE_PATH=path/to/fixtures -- Where to find the fixtures
• rake db:seed_do FILTER=users,articles -- Only run seed files with a filename matching the FILTER

You can also do a similar thing in your code by calling SeedDo.seed(fixture_paths, filter).

If you load a large amount of seed data, you can enable bulk upsert mode:

SeedDo.seed(bulk: true)

In bulk mode, SeedDo buffers seed data for each file and writes it with upsert_all, which can significantly reduce the number of queries.

You can also control the batch size per upsert_all call. The default is 1000.

SeedDo.seed(bulk: { batch_size: 100 })

• Bulk mode relies on Rails upsert_all
• The seed constraints are passed to unique_by, so the related columns must have a unique index
• Bulk mode is supported only on databases that support conflict targets, such as PostgreSQL and SQLite
• Bulk mode is not available on MySQL
• SeedDo.seed and SeedDo.seed(bulk: true) may not always produce the same record order, so verify the behavior carefully before enabling it in an existing production environment

To disable output from Seed Do, set SeedDo.quiet = true.

Seed files can be huge. To handle large files (over a million rows), try these tricks:

• Gzip your fixtures. Seed Do will read .rb.gz files happily. gzip -9 gives the best compression, and with Seed Do's repetitive syntax, a 160M file can shrink to 16M.
• Add lines reading # BREAK EVAL in your big fixtures, and Seed Do will avoid loading the whole file into memory. If you use SeedDo::Writer, these breaks are built into your generated fixtures.
• Load a single fixture at a time with the FILTER environment variable
• If you don't need Seed Do's ability to update seed with new data, then you may find that activerecord-import is faster

If you need to programmatically generate seed files, for example to convert a CSV file into a seed file, then you can use SeedDo::Writer.

SeedDo has included Capistrano deploy script, you just need require that in config/deploy.rb:

require 'seed-do/capistrano'

# Trigger the task after update_code
after 'deploy:update_code', 'db:seed_do'

If you use Capistrano3, you should require another file.

require 'seed-do/capistrano3'

# Trigger the task before publishing
before 'deploy:publishing', 'db:seed_do'

bundle install      # Install the dependencies
bundle exec rspec   # Run the specs

By default an sqlite3 database is used.

DB=mysql2 bundle exec rspec
DB=postgresql bundle exec rspec

The connection paramaters for each of these are specified in spec/connections/, which you can edit if necessary (for example to change the username/password).

This repository includes a Dev Container setup for running the test suite locally with Ruby, SQLite, MySQL, and PostgreSQL ready to use.

1. Open the repository in a Dev Container.
2. Wait for bundle install to finish.
3. Run the specs you want:

bundle exec rspec
DB=mysql2 bundle exec rspec
DB=postgresql bundle exec rspec

The Dev Container config also sets MYSQL_HOST=mysql and POSTGRES_HOST=postgres automatically, so the database-backed specs work without editing local machine settings.

Michael Bleigh is the original author

Copyright © 2008-2010 Michael Bleigh, released under the MIT license