Inoxis Interpreter

Inoxis is a memory-safe, statically-typed programming language inspired by Rust's ownership model, but with C++-like syntax. This repository contains a full interpreter for the language, built using C/C++ and ANTLR, and running on a custom virtual machine.

---

Features

• Rust-style memory safety (ownership, borrowing, mutability)
• C++-like syntax
• Stack- and heap-allocated arrays
• Custom instruction set for a stack-based virtual machine
• Manual lexer, ANTLR-generated parser
• Three-pass AST walker:
  • Symbol resolution
  • Memory safety checking
  • Intermediate representation (IR) generation
• Detailed error reporting with line numbers and messages

---

Example

int main()
{
    int mut *x = new int;
    *x = 10;
    int &y = & mut x;
    cout << "*y = " << *y << endl;
    return 0;
}

Project Structure

• main.cpp – Entry point: loads source and starts interpretation

• Interpreter.* – Core logic for parsing and execution flow

• myLexer.* – Custom lexer implementation

• symbolTable., MemSafetyPass., VMInputPass.* – Three AST passes

• VMInput., VirtualMachine., VMInstruction.* – Virtual machine backend

• .antlr/ – ANTLR-generated parser files

• Inoxis.g4, lexRules.g4 – Grammar definition for parser and lexer

Installation

Clone the repository:

git clone https://github.com/mbclause/Inoxis-Interpreter

Install GLib using vcpkg:

./vcpkg install glib:x64-windows-static

Open the solution in Visual Studio 2022 and build it.

Note: The ANTLR runtime and generated parser files are included in the repo, but GLib must be installed separately.

Usage

From the directory containing the executable:

Inoxis-Interpreter.exe <input-file.txt>

Ensure iconv-2.dll and your input file are in the same directory as the executable.

Dependencies

• Visual Studio 2022

• ANTLR4 (requires Java 8+, JDK 23, and Python 3)

• GLib (via vcpkg)

Future Work

Planned Improvements

• Add data types: float, double, bool, char, string, etc.

• Support void functions, and more flexible parameter lists

• Add operators: multiplication, division, modulus, and expression parentheses

• Implement optimization passes (peephole, loop)

• Improve scoping using parse tree metadata instead of string search

• Improve interpretation speed with better scope-resolution algorithms

• Fix evaluation order issues in arithmetic expressions

License

MIT License