In itself it is best to keep code clean all the time. The best way to do this is to adopt the following scheme:
- Prototype: create a draft
- Test: test the functionality
- Refactor: clean up the code
- Ship: deliver the code to the production environment
Unfortunately, the third point is often overlooked, especially when the decision is made by a manager with no coding experience. The reasoning is: the code works, so why invest any more effort in it? Over time, the code begins to « stink »; technical debts accumulate.
If an existing codebase contains code smells, it is not possible to start from scratch. The code must then be refactored. It is best totake an incremental approach and delineate areas within the codebase whoare relatively easy to clean. You separate operational or upgradeable components from “dark corners”, which are best left alone.
Separating less affected areas of code from areas most affected by code smells reduces complexity. That makes it easier to test code and perform debugging steps. Automated code review tools that track code smells and offer suggestions and/or cleanup aids can help with this process.
The refactoring approaches are effective in eliminating code smells. It’s about encapsulating or breaking down functionality into functions. Removing unused code sections and improving naming clears up the codebase. We can rely here on the basic principles « Don’t Repeat Yourself » (DRY) or « ne pas se repeat » and « You ain’t gonna need it » (YAGNI) or « Tu n’en auras pas nes « .
