OOP Software Design Principles And Patterns
This guide covers higher-level considerations that every developer should be aware of when coding in C++. They are complementary to the other development guides; if you find a conflict between another guide and this one, that’s a bug–please report it.
The goal of every program is to be useful (solve the right problems), usable (easy to use), and modifiable (easy to maintain). To that end, there are important high level design principles that help developers achieve these goals, discussed below in order. Crucially, all of these principles are valid across different levels of software, and for many different definitions of “module”: function, object, class, architecture, etc.
The principles discussed are all from the perspective of other developers and future you–not the end/customer client!
Generally speaking, Systems should designed such that they can be decomposed into cohesive, loosely coupled modules. A cohesive module has a unified purpose, while loose coupling implies dependencies on other modules are minimized. Taken together, this makes a module easier to reuse, replace, and understand.
Important
Naming is of paramount importance at ALL levels of abstraction and modularity, because how those identifiers hang together (or don’t) at the level of classes, functions, etc., plays a large role in how people understand your code and how easy or difficult it will be for someone to say “Oh I can just add this one other thing even though…”; the more cohesive the code, the greater the cognitive dissonance developers will experience if they try to shoehorn in something unrelated.
Design Principles for Useful Software
TBD.
Design Principles For Usable Software
OOP Design Patterns
Incorporate design patterns into your code explicitly whenever possible. That
is, if you’re going to use the decorator pattern, instead of just having a
member variable and wrapping/extending functionality as needed, inherit from a
decorator<T> class in C++ for example. See here
for a comprehensive list of patterns to be aware of. Not all patterns are
relevant in all cases/languages.
Patterns in this category create objects for you by some mechanism, rather than you having to instantiate objects directly.
Factory
Prototype
Singleton
Patterns in this category concern class and object composition (duh).
Decorator
Composite
PIMPL - This one is dispreferred, because it is easy to abuse as a “cure” for many dependency problems which are better solve by a clean interface + implementation + factory). It is still allowable, and can be useful in some situations with careful design.
Patterns in this category are primarily concerned with communication between objects.
FSM
Visitor
Memento
Strategy
Mediator
Rationale: Improves readability and makes the intent of the code/programmer much clearer, and having reuseable template classes for common design patterns greatly reduces the risk of bugs in your usage of them.
Design Principles For Modifiable Software
SOLID Design Principles
Another overlapping school of thought worth mentioning with respect to design principles discussed above is:
Single responsibility - This mostly maps to The Modularity Principle: Designing Internally Cohesive Classes.
Open/Closed - Software entities (classes, modules, functions, etc.) should be open for extension, but closed for modification. Basically, you should be able to add new features while leaving existing code intact.
Liskov substitution - Objects of a superclass shall be replaceable with objects of its subclasses without breaking the application. That is, subclasses cannot exhibit different behavior at the level of interfaces than their parent classes. Such behaviors might include taking additional parameters in overriden functions, return any data types that parent class methods don’t.
Interface Segregation - Clients should not be forced to rely on interfaces they don’t use. Basically, if you have to stub out a bunch of methods from an interface when inherit from it in a class, your interface is too broad, and needs to be separated.
Dependency Inversion - High-level modules should not depend on low-level modules. Both should depend on abstractions. Abstractions should not depend on details. Details should depend on abstractions. This mostly maps to the Abstraction Principle.