d3lph1/boollet

Boolean algebra toolkit for PHP

0.1.0 2022-03-12 12:17 UTC

This package is auto-updated.

Last update: 2024-06-12 17:20:36 UTC


README

logo.png

Boollet

Boollet is a boolean algebra toolkit for PHP.

PHP Version Require License

Features

Requirements

  • PHP >= 8.1

Installation

composer require d3lph1/boollet

Usage

Expression object API

You can create either UnaryExpression or BinaryExpression with one or two operands respectively:

use D3lph1\Boollet\Structure\Expression\{Variable, UnaryExpression, BinaryExpression};
use D3lph1\Boollet\Structure\Operator\{UnaryOperators, BinaryOperators};

$expr = new BinaryExpression(
    new UnaryExpression(UnaryOperators::NOT, new Variable(false)),
    BinaryOperators::AND,
    new BinaryExpression(new Variable(true), BinaryOperators::OR, new Variable(false, label: 'Z'))
);

echo $expr; // (!A ⋀ (B ⋁ Z))

If there is no label for variable specified, it will be assigned with sequentially autogenerated symbols.

Evaluate the expression with initial variable values:

$val = $expr->evaluate(); // true

Evaluate the expression with overwritten variable values (It can be partially overwritten):

$val = $expr->evaluate(['A' => true, 'B' => true, 'Z' => true]) // false

Value binding

In the example above there are only static variable values which could not be changed dynamically without expression reconstructing.

To change value of variable at runtime you should use Variable::set() method. For convenient batch value setup there is ValueBinder class:

use D3lph1\Boollet\ValueBinder;

$a = new Variable(false);
$b = new Variable(true);
$z = new Variable(false, label: 'Z');

$expr = new BinaryExpression(
    new UnaryExpression(UnaryOperators::NOT, $a),
    BinaryOperators::AND,
    new BinaryExpression($b, BinaryOperators::OR, $z)
);

$binder = new ValueBinder();
$binder->bind($a);
$binder->bindAll([$b, $z]);

$binder->set([
    'A' => true,
    'B' => true,
    'Z' => true
])

$expr->evaluate(); // true

Expression parser

For parsing stringed expressions uses ShuntingYardParser parser implementation. Under the hood it uses Dijkstra's algorithm of the same name.

use D3lph1\Boollet\Parser\{Lexer, Reader\StringInputReader, ShuntingYardParser};

$lexer = Lexer::default();
$input = new StringInputReader('X ⊕ Y → (X ⋀ Z)');
$parser = new ShuntingYardParser($lexer);

$expr = $parser->parse($input);

echo $expr; // ((X ⊕ Y) → (X ⋀ Z))

Building truth table

use D3lph1\Boollet\TruthTable;

$table = TruthTable::tabulate($expr);
$table->setLabel('f(X ⊕ Y → (X ⋀ Z))');

echo $table;
+---+---+---+--------------------+
| X | Y | Z | f(X ⊕ Y → (X ⋀ Z)) |
+---+---+---+--------------------+
| 0 | 0 | 0 |                  1 |
| 0 | 0 | 0 |                  1 |
| 0 | 1 | 0 |                  0 |
| 0 | 1 | 0 |                  0 |
| 1 | 0 | 0 |                  0 |
| 1 | 0 | 0 |                  1 |
| 1 | 1 | 0 |                  1 |
| 1 | 1 | 0 |                  1 |
+---+---+---+--------------------+

Complete conjunctive/disjunctive normal form calculation

Class NormalForms provides utility methods to find complete conjunctive (or disjunctive ) normal form representations.

use D3lph1\Boollet\NormalForm\NormalForms;

// $expr ~ ((X ⊕ Y) → (X ⋀ Z))

$ccnf = NormalForms::calculateCompleteConjunctive($expr); // ((X ⋁ (!Y ⋁ Z)) ⋀ ((X ⋁ (!Y ⋁ !Z)) ⋀ (!X ⋁ (Y ⋁ Z))))
$cdnf = NormalForms::calculateCompleteDisjunctive($expr); // ((!X ⋀ (!Y ⋀ !Z)) ⋁ ((!X ⋀ (!Y ⋀ Z)) ⋁ ((X ⋀ (!Y ⋀ Z)) ⋁ ((X ⋀ (Y ⋀ !Z)) ⋁ (X ⋀ (Y ⋀ Z))))))

Zhegalkin Polynomial calculation

For such needs you can use ZhegalkinPolynomial utility class:

use \D3lph1\Boollet\ZhegalkinPolynomial;

// $expr ~ (!X → ((!Y ⊕ X) ⋀ !Z))

$polynomial = ZhegalkinPolynomial::calculate($expr);

echo $polynomial; // ((Z ⋀ (Y ⋀ X)) ⊕ ((Y ⋀ X) ⊕ ((Z ⋀ X) ⊕ ((Z ⋀ Y) ⊕ (Y ⊕ (Z ⊕ 1))))))

SAT and UNSAT solvers

Boollet provides naive algorithm implementations to solve boolean (un)satisfiability problem.

SAT is the problem of determining if there exists an interpretation that satisfies a given boolean formula (formula becomes true).

UNSAT is the problem of determining if there exists an interpretation that not satisfies a given boolean formula (formula becomes false).

CompleteDisjunctiveNormalFormSATSolver works only with expressions in complete disjunctive normal form. Whereas CompleteConjunctiveNormalFormUNSATSolver uses only expressions in complete conjunctive normal form.

The second argument of the method findAllPossibleSolutions() takes an array of variables with respect to which it is required to solve the problem. Other variables whose labels are not passed to this argument must have values (in the example below y is such variable).

use \D3lph1\Boollet\SAT\CompleteDisjunctiveNormalFormSATSolver;
// $expr ~ X ⋁ (Y ⋀ Z)

$y->set(false);

$cdnf = NormalForms::calculateCompleteDisjunctive($expr);

$sat = new CompleteDisjunctiveNormalFormSATSolver();
$solutions = $sat->findAllPossibleSolutions($cdnf, ['X', 'Z']);
$solutions = $sat->findAllPossibleSolutions($cdnf, ['X', 'Z']);

$solutions will look like this:

^ array:2 [▼
  0 => array:2 [▼
    "X" => true
    "Z" => false
  ]
  1 => array:2 [▼
    "X" => true
    "Z" => true
  ]
]

To conveniently define results constraints, you can use findAllPossibleSolutionsWithConstraints():

use D3lph1\Boollet\Constraints\Constraints;

$solutions = $sat->findAllPossibleSolutionsWithConstraints($cdnf, ['X', 'Z'], new class() implements Constraints {
    public function isSatisfy(array $values): bool
    {
        return $values['X'];
    }
});

$solutions will look like this:

^ array:1 [▼
  0 => array:2 [▼
    "X" => true
    "Z" => false
  ]
]

License

This code is published under the MIT license. This means you can do almost anything with it, as long as the copyright notice and the accompanying license file is left intact.