Chapter 6. Test-driving data access
Information is the currency of democracy.
Thomas Jefferson
Data is everywhere. Think about it. How many systems have you worked with in your professional career where the business logic you implemented did not operate on persistent data? Exceptions do exist, of course, but the majority of systems in today’s enterprises deal with persistent data—and with lots of it. As such, data has a central role in the development of these systems and in the day-to-day lives of software developers. Furthermore, if you’d have to pick one component of an enterprise system as the single most important, it’d have to be the data. Yet software developers have a long tradition of relying on manual, visual inspection for testing their data-access code. Visual inspection is slow and error-prone—not how we as professional craftsmen want to work.
So far, we’ve ignored data as an element in our tests by faking where it’s coming from. In other words, although we have seen how we can implement business logic that needs persistence services using test-driven development and test doubles, we haven’t seen how to implement the persistence layer itself. In this chapter, we’ll pick up the slack and give the data layer our full attention.