Testing

Why?

• verify that our code does what we want, in a repeatable, automated way
  • alternative is manually running the code -- works fine for small problems but each time you add a feature, you have one more thing to manually test, and you need to test everything to make sure you didn't break anything that was working before

How?

• in Ruby, we use rspec. (other languages have other testing frameworks)
  • rspec is a ruby gem, so we should specify which version we're targeting in a Gemfile, so people running our code later know which version we were using
• rspec looks like this:

  describe "general thing you're testing" do
    it "does specific thing you care about" do
      expect(some_function_call()).to eq expected_result
    end
  end

Things we talked about

• Gemfiles (documenting which gems you used for future developers using your code)
• bundle install to install everything listed in the Gemfile (need to gem install bundler first if bundle install fails)
• .rspec files (need to be in the directory where you're running rspec, put any flags you always want to use there, namely -fd for the documentation)
• passing blocks and yielding
• custom errors for making your errors more intuitive for users of your code