Chess Engine

A UCI-compliant chess engine written in C using bitboards for efficient move generation and evaluation.

Features

• UCI Protocol: Full UCI (Universal Chess Interface) compliance
• Bitboard Representation: Efficient 64-bit board representation
• Move Generation: Legal move generation for all piece types
• Advanced Search: Alpha-beta search with quiescence, null move pruning, futility pruning, and razor pruning
• Evaluation: Material and positional evaluation with piece-square tables
• Move Ordering: MVV-LVA move ordering with history heuristic
• Transposition Table: 1M entry hash table for position caching
• Opening Book: Human-editable opening book support
• Time Management: Basic time control and move time management
• Iterative Deepening: Progressive depth search with time constraints

Building

Prerequisites

• GCC compiler (version 4.9 or later)
• Make (for Unix/Linux/macOS)
• Windows Command Prompt or PowerShell (for Windows)
• Git (installation)

Build Instructions

Windows

git clone https://github.com/reposketched/ryska.git # or download the code directly
# Using the provided batch file
./build.bat

# Or manually with GCC
gcc -Wall -Wextra -O2 -std=c99 -c src/*.c -o obj/*.o
gcc obj/*.o -o chess_engine.exe

Unix/Linux/macOS

# Clone the repository
git clone https://github.com/reposketched/ryska.git
cd chess-engine

# Build using Make
make

# Or manually with GCC
mkdir -p obj
gcc -Wall -Wextra -O2 -std=c99 -c src/*.c -o obj/*.o
gcc obj/*.o -o chess_engine

Build Output

The build process creates:

• chess_engine (Unix/Linux/macOS) or chess_engine.exe (Windows)
• obj/ directory with compiled object files
• book.txt opening book file (if missing)

Usage

Running the Engine

# Start the engine
./chess_engine          # Unix/Linux/macOS
chess_engine.exe        # Windows

The engine will start and wait for UCI commands on stdin.

Testing with a Chess GUI

You can use the engine with any UCI-compatible chess GUI:

• Arena - Free chess GUI
• Cutechess - Cross-platform chess GUI
• Fritz - Commercial chess software
• Chessbase - Professional chess database
• Lichess - Online chess platform (via UCI engine)

Manual Testing

# Test basic UCI commands
echo "uci" | ./chess_engine
echo "isready" | ./chess_engine
echo "position startpos" | ./chess_engine
echo "go depth 4" | ./chess_engine

UCI Commands Supported

• uci - Engine identification
• isready - Engine ready status
• ucinewgame - Start new game
• position [fen <fenstring> | startpos] moves <move1> ... <movei> - Set position
• go [depth <x>] [movetime <x>] [wtime <x>] [btime <x>] [winc <x>] [binc <x>] - Start search
• stop - Stop search
• quit - Quit engine

Architecture

Core Components

1. Bitboards (bitboard.h/c): 64-bit board representation and operations
2. Board (board.h/c): Board state management and FEN parsing
3. Move Generation (movegen.h/c): Legal move generation for all pieces
4. Evaluation (evaluation.h/c): Position evaluation with piece-square tables and mobility
5. Search (search.h/c): Advanced search with multiple pruning techniques
6. Transposition Table (transposition.h/c): Position caching for improved performance
7. Opening Book (book.h/c): Human-editable opening move database
8. UCI Interface (uci.h/c): UCI protocol implementation

Advanced Features

• Null Move Pruning: Reduces search depth for null moves
• Futility Pruning: Prunes moves unlikely to improve alpha
• Razor Pruning: Aggressive pruning for deep searches
• History Heuristic: Tracks move success for better ordering
• Iterative Deepening: Progressive depth search with time management
• Piece-Square Tables: Positional bonuses for piece placement
• Mobility Evaluation: Considers piece mobility in evaluation

Search Parameters

• Maximum Depth: 64 plies
• Transposition Table: 1M entries
• Null Move Reduction: 2 plies
• Futility Margin: 150 centipawns
• Razor Margin: 400 centipawns

Performance

The engine is optimized for:

• Speed: Efficient bitboard operations and move generation
• Memory: Compact data structures and efficient caching
• Accuracy: Legal move generation and proper evaluation

Limitations

Current limitations include:

• No parallel search (single-threaded)
• Basic time management
• No endgame tablebases
• Simplified evaluation compared to commercial engines

Future Improvements

Planned enhancements:

• Multi-threading support
• Advanced time management
• Endgame tablebase integration
• Neural network evaluation
• Improved opening book format
• Better move ordering heuristics

Development

Project Structure

chess-engine/
├── src/           # Source code
├── obj/           # Compiled objects
├── build.bat      # Windows build script
├── Makefile       # Unix build configuration
├── book.txt       # Opening book
└── README.md      # This file

Compilation Flags

• -Wall -Wextra: Enable all warnings
• -O2: Optimize for speed
• -std=c99: Use C99 standard

License

This project is open source and available under the MIT License.