/sudoku

A Sudoku generator and solver for Java and Android.

Primary LanguageJavaMIT LicenseMIT

#Sudoku

A Sudoku library for Java and Android.

It features a Generator to generate random Sudoku Grids of various complexity as well as a Solver to solve any provided Grid using backtracking.

##What is Sudoku?

Sudoku is a logic-based, combinatorial number-placement puzzle. The objective is to fill a 9×9 grid with digits so that each column, each row, and each of the nine 3×3 sub-grids that compose the grid (also called "boxes") contains all of the digits from 1 to 9. (Source Wikipedia)

###Example

╔═══╤═══╤═══╦═══╤═══╤═══╦═══╤═══╤═══╗
║   │   │   ║ 6 │   │ 2 ║   │   │   ║
╟───┼───┼───╫───┼───┼───╫───┼───┼───╢
║   │ 9 │ 2 ║   │   │   ║ 3 │ 1 │   ║
╟───┼───┼───╫───┼───┼───╫───┼───┼───╢
║ 8 │   │ 5 ║   │ 1 │ 4 ║ 6 │ 9 │   ║
╠═══╪═══╪═══╬═══╪═══╪═══╬═══╪═══╪═══╣
║ 1 │ 3 │   ║ 4 │   │   ║ 7 │   │   ║
╟───┼───┼───╫───┼───┼───╫───┼───┼───╢
║ 4 │   │ 9 ║   │   │ 8 ║ 1 │   │   ║
╟───┼───┼───╫───┼───┼───╫───┼───┼───╢
║   │ 5 │   ║ 1 │ 3 │ 6 ║ 4 │   │ 9 ║
╠═══╪═══╪═══╬═══╪═══╪═══╬═══╪═══╪═══╣
║   │   │   ║   │   │ 1 ║ 5 │ 8 │ 7 ║
╟───┼───┼───╫───┼───┼───╫───┼───┼───╢
║ 2 │   │ 1 ║ 5 │   │   ║   │   │   ║
╟───┼───┼───╫───┼───┼───╫───┼───┼───╢
║   │ 8 │ 7 ║ 9 │ 4 │ 3 ║ 2 │   │   ║
╚═══╧═══╧═══╩═══╧═══╧═══╩═══╧═══╧═══╝

##API This library basically features three classes

  • Generator: Generates random Sudoku Grids of various complexity.
  • Grid: Represents a Sudoku Grid and features a variety of convenient methods.
  • Solver: Solves any provided Grid using backtracking.

For detailed information check the javadoc.

###How to use Generator? In order to generate a new, random Sudoku Grid the Generator class features a straightforward generate() method. This method takes the numberOfEmptyCells as parameter. This parameter controls the complexity of the resulting Grid for human players. The complexity increases with a higher amount of empty cells. The example Grid shown above contains 42 empty cells.

Generator generator = new Generator();
Grid grid = generator.generate(42);

###How to use Grid? In order to display the Grid in your application you have to iterate over the Grid utilizing its getSize() and getCell() functions. For console-based application you may also use its convenient toString() method which produces a textual representation as shown in the example grid above.

int size = grid.getSize();
for (int row = 0; row < size; row++) {
  for (int column = 0; column < size; column++) {
    Cell cell = grid.getCell(row, column);
    //do something with cell
  }
}

In order to create a Grid using standard Java data types you can use a two-dimensional array of integers and Grid's static factory method of.

//9x9
int[][] rawGrid = new int[][]{
  {0, 0, 0, 6, 0, 2, 0, 0, 0},
  ...
  {0, 8, 7, 9, 4, 3, 2, 0, 0}
};

Grid grid = Grid.of(rawGrid);

###How to use Solver? The solver solves any valid Grid using backtracking.

Solver solver = new Solver();
Grid grid = ...

solver.solve(grid);

####Backtracking

#####Motivation The reason I created this library was that I wanted to understand how to solve a Sudoku puzzle programmatically.

I was afraid that a sort of "brute force" technique will take too long. Thus, my first solution followed the concept of "candidate search". This lead to many lines of code and only worked for simple - medium level Sudokus. At a certain point I needed to make assumptions anyway, so I started to use backtracking. Very quickly I've thrown away my solution and started to use backtracking completely, which simplified the algorithm's code a lot.

#####Algorithm Backtracking is brute force. The algorithm takes the first valid value and goes on with this approach until the puzzle is solved or there is no more valid value left. If there is no more valid value left (but the puzzle is not solved yet) it tracks back to the last known valid value and tries another value.

That way the algorithm will always find a valid solution unless the initially provided puzzle is invalid! It will even fill an empty puzzle (with no pre-filled cells at all) which makes it also ideal to be used to generate new puzzles.

However, backtracking is a perfect example for recursion. Below is all the code it takes so solve any Sudoku puzzle.

private final int[] values = { 1, 2, 3, 4, 5, 6, 7, 8, 9 };

private boolean solve(Grid grid, Optional<Grid.Cell> cell) {
    if (!cell.isPresent()) {
      return true;
    }

    for (int value : values) {
      if (grid.isValidValueForCell(cell.get(), value)) {
        cell.get().setValue(value);
        if (solve(grid, grid.getNextEmptyCellOf(cell.get()))) return true;
        cell.get().setValue(0);
      }
    }

    return false;
  }