Edgestitch allows engines to define partial structure-self.sql files to be stitched into a single structure.sql file by the umbrella application. This allows big teams to develop large Cobra Applications without much conflict happening in a single structure.sql file. Instead, each team will more likely update one component or two structure-self.sql files while other teams concerned with other areas of the application will be changing different files.
Add this line to your application's Gemfile:
# Gemfile
gem "edgestitch", require: falseAnd then execute:
$ bundle install
Then require the railtie in the umbrella's config/application.rb:
# config/application.rb
require "edgestitch/railtie"If your umbrella app also has migrations or even models, you'll have to also install the engine task to your Rakefile:
# Rakefile
Edgestitch.define_engine(::My::Application)Each internal engine will also have a development dependency on edgestitch, and can install it the same way:
# crazy.gemspec
spec.add_development_dependency "edgestitch"And then execute:
$ bundle install
And install the helper tasks:
Edgestitch.define_engine(::Crazy::Engine)You'll also have to add the railtie to the dummy app (as they're dummy umbrella apps), in case the engine has external database dependencies.
# spec/dummy/config/application.rb
require "edgestitch/railtie"Edgestitch will enhance the default rails tasks, so nothing special has to be done in order to use it. Once edgestitch is correctly installed things should Just Work™️ as explained in the rails manual, but it will be generating structure-self.sql along with structure.sql files. Important: It's recommended that structure.sql files are added to .gitignore.
Edgestitch works based on table and migrations ownership. Based on these ownerships, Edgestitch can export a structure-self.sql file defining the DDL owned by a specific engine.
The stitching process then takes all loaded engines (assuming the dependency between engines is defined correctly) and assembles a structure.sql file.
A model is owned by an engine when it inherits from the Engine's ApplicationModel, and thus the table is owned by that engine. A migration is owned by the engine if it is defined within its db/migrate directory.
When an external dependency brings in extra tables (i.e. acts_as_taggable) that are not defined in any structure-self.sql. To be part of the ecosystem, the new tables should be owned by the engine adding them. That can be done by adding these tables to <engine>/db/extra_tables. It's a simple text file with a list of table names that do not inherit from <Engine>::AplicationModel, but are owned by that engine and should be included in its structure-self.sql.
An external gem can also define a structure-self.sql file to be adapted in this ecosystem. That can be done using the same approach specified in Installation / Engines.
After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.
To install this gem onto your local machine, run bundle exec rake install.
Bug reports and pull requests are welcome on GitHub at https://github.com/powerhome/power-tools.
The gem is available as open source under the terms of the MIT License.