Code Smells

Common problems that arise in code

The term “code smell” was introduced in a book by Martin Fowler, in a chapter he co-authored with Kent Beck (the designer of JUnit) in a book called “Refactoring: Improving the Design of Existing Code”.

In Fowler’s book, he introduces twenty-two specific “code smells”, along with specific guidance for fixing the smelly code and replacing it with better code.

Since that time, the phrase “code smell” has taken on a much broader sense: basically anything about your code that someone looking at it doesn’t like.

Ok, maybe that was a bit snarky. The point is that “code smell” originally had a very technical, specific meaning, and it would be nice to to reclaim some of that meaning.

A Refactoring Exercise

You are invited, at various stages in the evolution of your project, to take on a refactoring exercise as a team.

1. Look through your code, and see if you can find the “smelliest” part.

It’s up to you as a team to decide what the criteria are for “smelliness”, but here are some suggestions. For each of these, consider how and whether you can make it specific enough to measure, quantitatively. For example “length of methods” is fairly easy to quantify: you could simply measure the number of lines of code.

For others, it may be more challenging, but see what you can do.

• Length of methods
• Methods that rely on too much global state rather than on parameter values
• Comments instead of self-documenting code (If you replaced the method name like update() with something more description like update_customer_database_from_csv_file, would you be able to eliminate the comment?)
• Classes that are too “tightly coupled” with each other (the Inappropriate Intimacy code smell). (How much intimacy is too much? Can you quantify it in some way?)