Book Image

Hands-On Functional Programming with C++

By : Alexandru Bolboaca
Book Image

Hands-On Functional Programming with C++

By: Alexandru Bolboaca

Overview of this book

Functional programming enables you to divide your software into smaller, reusable components that are easy to write, debug, and maintain. Combined with the power of C++, you can develop scalable and functional applications for modern software requirements. This book will help you discover the functional features in C++ 17 and C++ 20 to build enterprise-level applications. Starting with the fundamental building blocks of functional programming and how to use them in C++, you’ll explore functions, currying, and lambdas. As you advance, you’ll learn how to improve cohesion and delve into test-driven development, which will enable you in designing better software. In addition to this, the book covers architectural patterns such as event sourcing to help you get to grips with the importance of immutability for data storage. You’ll even understand how to “think in functions” and implement design patterns in a functional way. By the end of this book, you’ll be able to write faster and cleaner production code in C++ with the help of functional programming.
Table of Contents (23 chapters)
Free Chapter
1
Section 1: Functional Building Blocks in C++
7
Section 2: Design with Functions
12
Section 3: Reaping the Benefits of Functional Programming
17
Section 4: The Present and Future of Functional Programming in C++

Optimizing memory usage

The method we've discussed so far, for structuring code in a functional way, involves passing multiple times through a collection that is treated as immutable. As a result, this can lead to copies of the collection. Let's look, for example, at a simple code sample that uses transform to increment all the elements of a vector:

template<typename DestinationType>
auto transformAll = [](const auto source, auto lambda){
DestinationType result;
transform(source.begin(), source.end(), back_inserter(result),
lambda);
return result;
};

TEST_CASE("Memory"){
vector<long long> manyNumbers(size);
fill_n(manyNumbers.begin(), size, 1000L);

auto result = transformAll<vector<long long>>(manyNumbers,
increment);

CHECK_EQ(result[0], 1001);
}

This implementation leads to a lot of memory allocations...