Building a Compiler with C#

Stop treating compilers like a black box. Learn to build your own simple language and virtual machine from scratch using C#.

#1about 4 minutes

Understanding Roslyn as a compiler-as-a-platform

Roslyn is not just a compiler black box but a full platform with APIs that provide language services to IDEs and inspired the Language Server Protocol.

#2about 2 minutes

Visualizing code structure with Roslyn and LINQPad

A practical demonstration in LINQPad shows how Roslyn can expose the intermediate language and abstract syntax tree for any C# code snippet.

#3about 4 minutes

Why build a custom language for game development

The "majors" language was created to meet the performance demands of a game that dynamically generated 3D worlds from user-provided mathematical formulas.

#4about 3 minutes

Defining a new language with a formal grammar

Creating a new language requires a formal specification and grammar that defines its syntax and rules, which serves as the blueprint for the compiler.

#5about 4 minutes

Comparing native and managed compiler architectures

Native compilers produce machine-specific object files, while managed compilers generate platform-independent intermediate language like MSIL or bytecode to be run by a virtual machine.

#6about 5 minutes

The four fundamental stages of compiler parsing

The compiler's front-end processes code through four stages: decoding the byte stream, tokenizing characters into blocks, generating an abstract syntax tree, and performing semantic analysis.

#7about 5 minutes

Representing code structure with an abstract syntax tree

The abstract syntax tree (AST) is a crucial data structure that represents code hierarchically, distinguishing between statements and expressions and enforcing operator precedence.

#8about 3 minutes

Traversing the abstract syntax tree with the visitor pattern

The Visitor design pattern provides a type-safe and extensible method for traversing an abstract syntax tree using a dedicated TreeWalker class for analysis or code generation.

#9about 3 minutes

A live demonstration of a custom language REPL

A live demo showcases the custom "majors" language REPL, including its support for function definitions, currying, and a built-in command for inspecting the AST.

#10about 3 minutes

Implementing operator precedence with Pratt parsing

The Pratt parsing technique simplifies handling operator precedence by using a hierarchy of recursive function calls that naturally builds the correct abstract syntax tree.

#11about 4 minutes

Transforming the AST into a linear intermediate language

The compiler flattens the hierarchical abstract syntax tree into a linear sequence of instructions, known as intermediate language, for efficient execution by a process virtual machine.