How should you address long functions and excessive parameter lists (code smells)?

• A. Use global variables
• B. Increase function length
• C. Break into smaller functions, extract classes, and group parameters with objects or builders to reduce coupling
• D. Use magic numbers

Answer: (C)

Long functions and long parameter lists make code hard to read, test, and modify. The best fix is to reduce complexity by breaking the work into smaller pieces and tightening the way data is passed between them.

Break into smaller functions by identifying distinct responsibilities inside the big function and pulling each one into its own, focused function. This makes each piece easier to understand, reuse, and verify with tests. It also helps you localize changes when requirements shift.

Extracting classes helps organize related data and behavior, turning a jumble of values into cohesive objects. When you group related parameters into a single object, callers pass fewer items and the callee operates on a well-defined interface. This reduces coupling because the function no longer depends on a long list of individual inputs.

Grouping parameters with objects or builders is especially helpful for optional or configuration-like data. A builder lets you assemble the needed settings in steps and then pass a single, clearly named object to the function. This clarifies intent, makes it easier to extend with new options, and keeps the function signature stable.

In contrast, using global variables introduces hidden dependencies and side effects, increasing coupling and making reasoning about code harder. Increasing function length keeps the problem worse, spreading logic across a longer, more complex block. Magic numbers hide intent and require readers to guess what a value means. These options undermine readability and maintainability.

So the recommended approach—break tasks into smaller units, encapsulate related data, and pass well-formed objects (possibly built with a builder)—addresses the smells directly and leads to cleaner, more maintainable code.