Raven

Raven is a clean, efficient C++ chess engine built from scratch. It uses a bitboard representation for high-performance move generation and board state management, and supports the standard Universal Chess Interface (UCI) protocol.

🚀 Key Features

• Bitboard Engine: High-performance 64-bit representation for all pieces, enabling fast bitwise move generation.
• Full Chess Rules: Supports all legal moves, including:
  • Castling (Kingside and Queenside).
  • En Passant captures.
  • Pawn Promotion (to Queen, Rook, Bishop, or Knight).
• Search & Optimization:
  • Alpha-Beta Pruning: Efficiently narrows the search tree.
  • Iterative Deepening: Ensures a best move is always ready when time runs out.
  • Time Management: Standard UCI time control support (wtime, btime, winc, binc).
• UCI Protocol: Fully compatible with chess GUIs like Cute Chess, Arena, or Stockfish GUI.
• Position Support: Can load games from any state using FEN strings.

📁 Source Overview

• board.h/cpp: Core board representation, bitboard management, and move application logic.
• move_gen.h/cpp: Comprehensive move generator including check detection and legal move validation.
• engine.h/cpp: The search brain, implementing Minimax with Alpha-Beta and Iterative Deepening.
• uci.h/cpp: Communication layer for the UCI protocol.
• main.cpp: Entry point that launches the UCI loop.

🛠️ Getting Started

Prerequisites

• A C++ compiler (GCC or Clang recommended).
• C++17 support.

Compilation

To build Raven, run:

g++ -std=c++17 main.cpp board.cpp move_gen.cpp uci.cpp engine.cpp -o raven

Usage

Once compiled, you can run the engine directly (./raven) and type UCI commands, or plug the executable into your favorite Chess GUI.

Example manual commands:

uci
isready
position startpos moves e2e4
go depth 6

📈 Roadmap (Future Improvements)

• Zobrist Hashing: To support Transposition Tables.
• Transposition Tables: To avoid redundant move calculations.
• Quiescence Search: To improve evaluation stability by searching captures until quiet positions are reached.
• Move Ordering: Heuristics like MVV-LVA or Killer Heuristic to speed up Alpha-Beta pruning.
• Advanced PST: More refined Piece-Square Tables for better positional understanding.