TIL, 2017-11-04, Database Cleaner and before all

Database cleaner:
- transaction strategy means the transactions are just rolled back. You can make the tests also read the uncommitted transactions.
- deletion and truncation: DELETE is faster but it doesn’t completely recreate the tables, it just deletes the rows. TRUNCATE is the slowest but it recreates the tables.
- DatabaseCleaner.clean_with(:truncation) to do a truncate delete, then you can put this in an after block.
- DatabaseCleaner.logger = Rails.logger if you want.
RSpec best practices Reference
- before(:all) block is not wrapped in a transaction.
- Only create what you need.
- Mock creating a ton of records.
- Do not over-mock.
Create different test suites to speed up your workflow. That way you can run different test suites.
RSpec order: before(:suite), before(:all), before(:each), after(:each), after(:all), after(:suite).
You can do things like params.fetch(:hello) { 'hello' }
Service object article Reference
- Modules over objects?
- The problem with just a lot of service objects handling business rules: you can end up with a whole basket full of Service Objects, many with implicit data dependencies between them, representing business workflows that have no explicit representation.
- Domain Driven Design:
  - Appropriate domain model to receive the functionality.
  - Infrastructure services should be things like “send an email”. This should be separate from the domain-level services.
  - Domain-level services should be named in terminology that everyone understands.
  - Services should have no persistent state.
  - Usually better to do Service.new(dependencies).call instead of Service.call(dependencies) because it allows breaking down call into multiple private methods that have access to the dependencies without having to pass them around as arguments to all the methods.

class ProcessIpn
  def initialize(*args)
    # Setup
  end

  def process
    # Do stuff...
  end

  def self.with(*args)
    processor = new *args
    processor.process
  end
end

Backing up your own database: create a private repository, git submodule add, then run a script that does the backup thing via the shell.