MyCCompiler

Overview

MyCCompiler is a C compiler implementation based on the book "Writing a C Compiler" by Nora Sandler. This project is developed in C++ and utilizes LLVM libraries for various compiler infrastructure components, including abstract syntax tree (AST) manipulation, source location tracking, and utility data structures.

The compiler follows a traditional multi-phase compilation approach, transforming C source code through several intermediate representations before generating x86-64 assembly code.

Project Structure

The project is organized into several key components, each responsible for a specific phase of the compilation process:

Core Components

• include/mycc/ - Header files containing interface declarations
• lib/ - Implementation files for all compiler components
• main.cpp - Entry point and command-line interface

Compilation Phases

Lexical Analysis

• Location: include/mycc/Lexer/ and lib/Lexer/
• Files: Lexer.hpp, Token.hpp, Lexer.cpp
• Purpose: Tokenizes input C source code into a stream of tokens

Parsing

• Location: include/mycc/Parser/ and lib/Parser/
• Files: Parser.hpp, Parser.cpp
• Purpose: Constructs an Abstract Syntax Tree (AST) from the token stream

Semantic Analysis

• Location: include/mycc/Sema/ and lib/Sema/
• Files: Sema.hpp, Scope.hpp, Sema.cpp, Scope.cpp
• Purpose: Performs semantic validation and scope resolution

Abstract Syntax Tree

• Location: include/mycc/AST/ and lib/AST/
• Files: AST.hpp, ASTContext.hpp, ASTPrinter.hpp, ASTPrinter.cpp
• Purpose: Defines AST node structures and provides tree manipulation utilities

Intermediate Representation

• Location: include/mycc/IR/ and lib/IR/
• Files: SimpleIR.hpp
• Purpose: Provides a low-level intermediate representation for optimization and code generation

Code Generation

• Location: include/mycc/CodeGen/ and lib/CodeGen/
• Files: IRGen.hpp, IRGen.cpp
• Subdirectory: x64/ containing X64CodeGen.hpp, x64AST.hpp, X64CodeGen.cpp
• Purpose: Transforms IR to target-specific assembly code (x86-64)

Support Components

Basic Utilities

• Location: include/mycc/Basic/ and lib/Basic/
• Files: LLVM.hpp, TokenKinds.def, TokenKinds.hpp, Diagnostic.def, Diagnostic.hpp
• Purpose: Provides fundamental data types, diagnostics, and LLVM integration

Testing

• test/ - Test cases for various compiler components
• writing-a-c-compiler-tests/ - Official test suite from the book

Build System

The project uses CMake as its build system with modular library organization:

• Main executable: mycc
• Component libraries: libLexer.a, libParser.a, libAST.a, libCodeGen.a, libSema.a, libBasic.a

Usage

# Build the project
mkdir build && cd build
cmake ..
make

# Compile a C source file (generates input.s assembly file)
./mycc input.c

# Run specific compilation phases
./mycc --lex input.c          # Lexer only
./mycc --parse input.c        # Lexer and parser
./mycc --validate input.c     # Semantic analysis
./mycc --tacky input.c        # Generate IR
./mycc --codegen input.c      # Full compilation without executable

# Print output to console
./mycc --print input.c        # Show assembly output
./mycc --tacky --print input.c # Show IR output

Dependencies

• LLVM: Provides core data structures (APSInt, StringRef, StringMap) and utility classes
• CMake: Build system
• C++17 or later: Required for modern C++ features used throughout the codebase

Current Implementation Status

The compiler currently supports:

• Basic C syntax parsing
• Variable declarations and assignments with scoped name resolution
• Arithmetic and logical expressions with proper operator precedence
• Unary operators (negation, bitwise complement, logical not)
• Binary operators (arithmetic, bitwise, comparison, logical)
• Ternary conditional operator
• Assignment and compound assignment operators
• Prefix and postfix increment/decrement operators
• Control flow statements:
  • Conditional statements (if/else)
  • Loop statements (while, do-while, for)
  • Jump statements (break, continue, goto)
  • Switch statements with case and default labels
• Labeled statements and goto
• Function definitions with block scope
• Short-circuit evaluation for logical operators
• Comprehensive semantic validation:
  • Variable scope and lifetime management
  • Duplicate declaration detection
  • Break/continue validation in loop contexts
  • Switch statement validation (constant case values, duplicate cases, multiple defaults)
  • Label uniqueness and goto target validation
  • Declaration placement validation (labels must precede statements, not declarations)
• x86-64 assembly generation with Intel syntax

For detailed technical information about the compiler's architecture and implementation, see TECHNICAL.md.