C++ Essentials

C++ and object oriented programing (OOP)

Note: this post is a work in progress

Template

Templates are the foundation of generic programming, which involves writing code in a way that is independent of any particular type.

Templates allow us to define a blueprint (or formula) for creating a generic class or function. Many STL containers and algorithms are implemented using templates, which allows a great flexibility in the types of supported elements.

For example, in the STL, there is a single definition of the vector container, but we can define many different kinds of vectors: vector<int>, vector<float>, vector<string>, and vector<T> where T is any type.

Generic programming depends on the ability for classes, methods, and functions to accept generic parameters. In C++ this is achieved via template parameters. A template parameter is a special kind of parameter that can be used to pass a type as an argument. Just like regular function parameters allow us to pass values to a function, template parameters allow us to pass types to a function. Templates can use these parameters as if they where any regular type.

Below is a function template that returns the greater of object a and b:

1
2
3
4
  template <class myType>
  myType GetMax (myType a, myType b) {
    return (a>b?a:b);
  }

Here I created a template with myType as the template parameter. The template parameter represents a type that has not been specified yet, but can be used in the tamplate function as though it were a regular type.

Container

A container is a holder object that stores a collection of other objects. Containers are implemented as class templates, which allows for greater flexibility in the types of supported elements.

Lambdas

Lambdas exist in lots of other programming languages that are coming from the functional programming family (e.g., Scheme). They also exist in C based languages like Java, C++, Objective C. We will start by looking at a very primitive form of lambda because in reality even the C language had the embryo of an idea of what lambdas could provide to programmers however it is called something different. So in C we refer to these as pointers to functions.

Object Oriented Programmin (OOP)

The four pillars of OOP:

  1. Polymorphism
  2. Abstraction
  3. Encapsulation
  4. Inheritance

Encapsulation: the act of hiding unnecessary details from the user.

  • As long as the usage contract remains the same, implementation details can change without affecting all users
  • Wrap data and methods into a unit, data is not accessible to the outside world
  • Insulating data from direct access is known as data hiding
  • Example: ssize_t send(int sockfd, const void *buf, size_t len, int flags);. User need not know entire network stack and how send() transmits data reliably from one end host to another.

Inheritance: process by which objects of one class aquire properties of another class.

Polymorphism: when you can treat an object as a generic version of something, but when you access it, the code determines which exact type it is and calls the associated code. Two types:

  1. Static polymorphism: determined at compile time (e.g., method overloading)
  2. Dynamic polymorphism: determined at run time (e.g, method overriding)

Class: a blueprint of data and methods that work with that data

Object: a specific instance of that class

Instance variable(s): belong to a specific instance of a class

Class variable(s): also known as static member variables and there is only one copy of these variables that is shared with all instances of that class

Misc

A formal parameter is a variable which is part of the method signature. An argument is the variable being passed in to the function or method.

Written on January 30, 2019