Nilang

A scripting language interpreter designed for Advent of Code 2025!

Overview

Nilang is a dynamically-typed scripting language with a focus on simplicity and expressiveness. It features a custom bytecode compiler with an optimizing IR, register allocation, and a trace and sweep garbage-collected runtime.

Features

• Dynamic typing - Variables can hold any type
• First-class functions - Functions as values, closures, partial functions
• Lists and maps - Built-in collection types
• For/while loops - Standard control flow
• REPL - Interactive development
• And much much more!

Installation

Quick Install (Recommended)

1. First, install Rust and cargo
2. Run the installation script:

curl -sSf https://raw.githubusercontent.com/Nilando/nilang/main/install.sh | bash

This will install nilang via cargo, download the standard library, and place it in ~/.nilang/std/.

Building from Source

git clone https://github.com/Nilando/nilang
cd nilang
cargo build --release
./install.sh

Usage

Run a Nilang program:

nilang program.nl

Start the REPL:

nilang

Pipe input to the interpreter:

echo 'print(1 + 2);' | nilang -s

Quick Start

🏗️ Tutorial document is in progress! 🏗️

Check out the examples/ directory for code samples:

• fibonacci.nl - Fibonacci sequence generator
• fizzbuzz.nl - FizzBuzz implementation
• quicksort.nl - Quicksort algorithm
• game_of_life.nl - Conway's Game of Life
• eratosthenes_sieve.nl - Prime number sieve
• And many more!

Try running an example:

nilang examples/fibonacci.nl

Project Status

Nilang is currently in early development (v0.1.0), and may not be maintained/updated any further. The language is functional but may have rough edges. Future features of interest include FFI through standard C bindings and compilation to WebAssembly, though these may not be implemented in the near future.

Known Limitations

• Error messages need improvement, particularly for parser errors
• Limited standard library
• Performance optimizations ongoing
• No FFI support

Architecture

Nilang uses a multi-stage compilation pipeline:

1. Lexer → Tokens
2. Parser → AST
3. IR Lowering → Three-address code with SSA
4. Optimization → GVN, constant folding, DCE
5. Codegen → Register allocation and bytecode emission
6. VM → Bytecode interpretation with garbage collection

Contributing

Issues and pull requests are welcome! This is a learning project and a work in progress.

License

MIT License - see LICENSE-MIT for details.

Author

Niland Schumacher