Book Image

Metaprogramming in C#

By : Einar Ingebrigtsen
Book Image

Metaprogramming in C#

By: Einar Ingebrigtsen

Overview of this book

Metaprogramming is an advanced technique that helps developers to automate repetitive tasks, generate scalable code, and enhance productivity in software development. Metaprogramming in C# is a comprehensive guide that will help you reap the full potential of metaprogramming in .NET runtime. You’ll start by learning about the .NET runtime environment and how you can use it to become a more productive developer. You'll learn how to infer types using reflection, use attributes, and create dynamic proxies. You’ll also explore the use of expressions to create and execute code and how to take advantage of Dynamic Language Runtime. But that's not all! You’ll also learn to go beyond inheritance and use method signature conventions to create easily maintainable code. Finally, you’ll dive into the world of compiler magic with Roslyn, where you'll discover how to use Roslyn to generate code, perform static code analysis, and write your own compiler extensions. By the end of this book, you’ll have a deep understanding of metaprogramming concepts and how to apply them to your C# code. You’ll be able to think about types, use attributes and expressions to generate code, and apply crosscutting concerns to improve code quality.
Table of Contents (25 chapters)
1
Part 1:Why Metaprogramming?
5
Part 2:Leveraging the Runtime
12
Part 3:Increasing Productivity, Consistency, and Quality
18
Part 4:Compiler Magic Using Roslyn

Summary

Recipes are pieces of code that clearly specify, in a linear manner, what they do, are a great tool for new developers. The developer can really see what’s going on, reason about the code, and find errors. As the developer, team, and project mature, the recipes start to feel like unnecessary chores or at least become very repetitive. Not only is this something that potentially affects productivity, but these kinds of repetitive tasks are easy to get wrong. Getting it wrong could pose multiple risks to the system:

  • Security risks
  • The risk of persisting invalid data
  • The risk of allowing operations that are not allowed
  • The risk of losing operational insight due to a lack of logging

The tradeoff between imperative procedural code that a developer can quickly reason about and the need for a consistent system is one you should consider. In smaller projects, it might not be worth the cognitive load of being “different” in applying things...