Arion Compiler: Milestone 4 & 5 - Intermediate Code Generator & Interpreter

This project implements a complete custom compiler and virtual machine developed in GNU C/C++ for the Arion programming language. The pipeline includes a lexical analyzer, a Recursive Descent syntax parser, a semantic analyzer, an Intermediate Code (IC) generator, and a custom Stack-based Virtual Machine (Interpreter).

This project is part of the IF2224 - Formal Language and Automata Theory course at STEI ITB.

---

Contributors

NIM	Name	GitHub
13524024	Billie Bhaskara Wibawa	@billie-bytes
13524050	Raysha Erviandika Putra	@Arbane557
13524058	Muhammad Naufal Romi Annafi	@naufalromi
13524084	Dzaki Ahmad Al Hussainy	@HussainDzaki

---

Features

This project contains a full Front-End to Back-End pipeline:

1. Tokenization (Lexer): Converts source text into token streams using a manual DFA-based scanner.
2. Recursive Descent Parsing: Builds Parse Tree structures from Arion grammar rules.
3. Semantic Validation: Performs scope resolution, type compatibility checks, and constructs TAB, ATAB, and BTAB symbol tables.
4. Intermediate Code Generation: Traverses the Decorated AST to synthesize flat, Three-Address Code (TAC) instructions (e.g., LOD, STO, JMP, JPC, OPR).
5. Stack-Based Virtual Machine (Interpreter): A custom execution engine featuring a Program Counter, Stack Pointer, and Base Pointer to manage memory, variable scopes (Static/Dynamic links), and subroutine calls.
6. File-Based Execution: Includes an ICParser capable of loading and executing standalone .ic bytecode files.
7. Runtime Error Handling: Active VM protection against stack overflow, stack underflow, invalid jump targets, and division by zero.
8. Interactive I/O: Full support for readln, read, writeln, and write standard procedures.
9. No External Generators: Implemented entirely from scratch in C++ without tools like Lex or Yacc.

---

Repository Structure

.
├── README.md
├── src
│   ├── codegen
│   │   ├── CodeGenerator.cpp
│   │   └── CodeGenerator.hpp
│   ├── common
│   │   ├── DataType.hpp
│   │   ├── Error.cpp
│   │   ├── Error.hpp
│   │   ├── Instruction.cpp
│   │   ├── Instruction.hpp
│   │   ├── NodeType.hpp
│   │   ├── ObjectType.hpp
│   │   ├── Reader.cpp
│   │   ├── Reader.hpp
│   │   ├── RuntimeValue.cpp
│   │   ├── RuntimeValue.hpp
│   │   ├── Token.cpp
│   │   ├── Token.hpp
│   │   ├── Tree.cpp
│   │   ├── Tree.hpp
│   │   ├── Writer.cpp
│   │   └── Writer.hpp
│   ├── interpreter
│   │   ├── ICParser.cpp
│   │   ├── ICParser.hpp
│   │   ├── Interpreter.cpp
│   │   └── Interpreter.hpp
│   ├── lexer
│   │   ├── Lexer.cpp
│   │   ├── Lexer.hpp
│   │   ├── LexerState.hpp
│   │   ├── SpecificScanners.cpp
│   │   └── SpecificScanners.hpp
│   ├── semantic
│   │   ├── ASTBuilder.cpp
│   │   ├── ASTBuilder.hpp
│   │   ├── ASTNode.hpp
│   │   ├── ASTPrinter.cpp
│   │   ├── ASTPrinter.hpp
│   │   ├── ASTVisitor.hpp
│   │   ├── SemanticAnalyzer.cpp
│   │   ├── SemanticAnalyzer.hpp
│   │   ├── SemanticVisitor.cpp
│   │   ├── SemanticVisitor.hpp
│   │   ├── SymbolTable.cpp
│   │   └── SymbolTable.hpp
│   ├── syntax
│   │   ├── Parser.cpp
│   │   └── Parser.hpp
│   └── main.cpp
└── test
    ├── lexer
    ├── parser
    └── semantic

---

Installation

Requirements

• GNU C/C++ Compiler (GCC/G++)
• Make

---

Installing Dependencies

Windows

1. Download and install MinGW-w64 from an official distributor
2. Ensure gcc, g++, and make (or mingw32-make) are added to your system PATH
3. Verify:

   g++ --version
   make --version

Linux / UNIX

1. Open your terminal
2. Install build essentials:

sudo apt update
sudo apt install build-essential

3. Verify:

g++ --version && make --version

---

How To Run

1. Clone the repository

git clone https://github.com/naufalromi/JEE-Tubes-IF2224-2026.git

2. Navigate to the repository and build

cd JEE-Tubes-IF2224-2026
make clean
make all

3. Run the compiler

make run

4. Choose analysis mode in the interactive menu:

• 1 Lexical Analyzer
• 2 Syntax Analyzer
• 3 Semantic Analyzer
• 4 Intermediate Code Generator
• 5 Interpreter (Run Program)

5. Choose testcase number based on the selected mode folder.

Semantic mode will:

• Build and analyze the decorated AST
• Print semantic errors when present
• Allow semantic display options (Decorated AST, TAB, ATAB, BTAB, or all)

Work Distribution

Name	NIM	Workload	Percentage
Billie Bhaskara Wibawa	13524024	Report and Coding	25%
Raysha Erviandika Putra	13524050	Report and Coding	25%
Muhammad Naufal Romi Annafi	13524058	Report and Coding	25%
Dzaki Ahmad Al Hussainy	13524084	Report and Coding	25%