puzzlegen 0.1.10

Procedural Match-3 Puzzle Generator and Solver

puzzlegen - Procedural Match-3 Puzzle Generator

Project Links:
GitHub Repository | PyPI Project | OSF Project | Preprint

Overview

puzzlegen is a Python package for generating, visualizing, and solving procedural Match-3 puzzles. It allows you to create single puzzles or batches, visualize them, and automatically find solutions using a Breadth-First Search (BFS) solver.

---

Features

• Random puzzle generation with customizable grid size, colors, and block counts
• Automatic solver (BFS) to check puzzle solvability within a given number of moves
• Batch generation: create and filter many puzzles at once
• Visualization: display puzzles and solutions as images or PDFs
• Export: save puzzles and solutions as PDF or CSV

---

Installation

Option 1

pip install puzzlegen

Option 2

Or, if you want to install from the github repository that may contain more up-to-date versions:

pip install git+https://github.com/carodak/puzzlegen.git

Option 3

Or, if you want to clone and install locally:

git clone https://github.com/carodak/puzzlegen.git
cd puzzlegen
pip install .

---

Basic Usage

Example 1: Randomly generate and solve a single puzzle

from puzzlegen.frontend import SinglePuzzle

# 1. Create a single puzzle with defined parameters
puzzle = SinglePuzzle(nb_blocks=10, colors=['red', 'blue', 'gray'], nb_moves=5, grid_size=(12, 12))

# 2. Randomly populate the puzzle board with blocks
puzzle.generate()

# 3. Visualize the puzzle
puzzle.show()

# 4. Try to solve the puzzle (within the allowed number of moves)
solution = puzzle.solve()

# 5. Visualize the solution (if found)
puzzle.show_solution()

Example 2: Generate a batch of solvable puzzles

from puzzlegen.frontend import PuzzleBatch

# 1. Set up a batch generator with parameter ranges
batch = PuzzleBatch(
    blocks_range=(6, 10),
    colors_range=(2, 4),
    colors_blocks=['blue', 'red', 'gray'],
    nb_moves=5,
    grid_size=(12, 6),
    stack_probability=0.75
)

# 2. Attempt to generate solvable puzzles for all combinations within the given ranges
batch.generate()

# 3. Show stats and save the results
batch.show_stats()
batch.save_pdf("batch_puzzles.pdf")
batch.save_csv("batch_puzzles.csv")

Or use without installation

You can also use the package directly in the Jupyter notebook located at examples/basic_usage.ipynb, available online here: https://colab.research.google.com/github/carodak/puzzlegen/blob/main/examples/basic_usage.ipynb — no installation needed.

For additional detailed usage examples, please see:

examples folder

This script demonstrates how to generate, solve, and save puzzles using the puzzlegen package.

---

Puzzle Rules

1. Puzzle Setup: The game board is a grid with colored blocks. Blocks are initially placed in the last row or stacked on top of each other. No more than 2 blocks of the same color are aligned horizontally or vertically. Each color must have at least 3 blocks.
2. GamePlay: The goal is to remove all blocks from the grid in as few moves as possible.
3. Available Moves:
   • Simple Move: Move a block left or right if the target cell is empty.
   • Exchange: Swap a block with its left or right neighbor if the target cell is occupied.
4. Elimination Rule: If 3 or more blocks of the same color are aligned horizontally or vertically, they disappear.
5. Gravity Rule: Blocks fall down if unsupported until they reach another block or the bottom of the grid.
6. Winning Condition: All blocks are removed from the grid within the allowed number of moves.

Executive Function Task Reference: https://osf.io/3pz74/wiki/home/

---

Project Structure

• src/puzzlegen/core/ – Core logic (puzzle generation, solver, batch processing, etc.)
• src/puzzlegen/frontend.py – User-friendly interface for puzzle creation and batch operations
• examples/ – Example notebooks and scripts to help you get started
• tests/ – Unit tests for code reliability
• outputs/ – Folder where generated puzzles and results are saved

---

API Documentation

The API Documentation is available at https://carodak.github.io/puzzlegen/puzzlegen.html or in docs folder.

---

Authors

• Caroline DAKOURE – Development
• Katie LAVIGNE – Code review, feedback, and improvements

---

Contributing

Pull requests and suggestions are welcome!
Please add tests for any new features.

---

Resources

• Solving simplified Candy Crush (Medium)
• EightPuzzle (GitHub)
• BFS explanation (YouTube)
• Visualgo BFS/DFS
• What is BFS? (dev.to)

---

License

GPL-3.0

---

Contact:
[e-mail]