While following red/green/refactor in application code is great, developers often forget tests are code too and need to be refactored as such. Page Objects provide a layer of abstraction throughout tests which simplify interaction through well-named methods and good object-orientation; this pattern allows developers to manage dependencies, selectors, and leverage RSpec's matcher syntax to write clean, focused tests which encapsulate DOM interaction and page concepts concisely. This talk will cover the progression of existing code bases to use Page Objects, Page Objects' benefits and practical applications, and why Page Objects will improve your acceptance tests.
I wrote a brief introduction on Page Objects within the context of a Rails application recently on the thoughtbot blog.
In this talk, I will briefly discuss how Cucumber and Gherkin were used "back in the day" to write acceptance tests and how we have been migrating towards more abstract tests with RSpec and Capybara. With this foundation laid, I will introduce Page Objects and their benefits, how I discovered them through a progression of refactoring my tests, how to refactor existing tests using Page Objects, and patterns that emerged after migrating test suites to use this pattern. Finally, I will cover caveats, tricks to using Page Objects effectively, and when to avoid using Page Objects.