Refactoring duplicated code is a trade off: on the one hand you create abstraction and centralizing code logic at the cost of, first, the overhead of learning t

Refactoring duplicated code is a trade off: on the one hand you create abstraction and centralizing code logic at the cost of, first, the overhead of learning that abstraction, and second, by increasing coupling between functions. I've personally found that the coupling is by far the most important factor to consider. If A depends on B, and C is found in both A and B, then you should factor out C. If A and B share C because they are adjacent, without being fundamentally intertwined, then duplicate C, but consider creating a library or module that makes it easy to talk about things that are similar to that duplicated code, C. (I tend to think about modules/libraries as little, independent DSLs)