Here’s a technique I sometimes use when I’m doing test-driven development (which is my preferred coding technique). Let’s say I’ve got a bunch of test cases that are all basically the same, but they have slightly different inputs and slightly different outputs. A typical test case might look like this: