README

MathExec is a php library for parsing and evaluating math expressions like this:

echo Evaluator::mathExec("5 ^ 2 + 36 / 2 - 2 * 0.5");

Or you could prepare an expression template and then execute it, with specified precision, when needed, by providing needed variables

$evaluator = Evaluator::mathPrepare("a + 5 / 3 * b - c");

foreach([[1,2,3,4],[4,5,6,4],[7,8,9,0]] as $values) {
    $evaluator->exec(['a' => $values[0], 'b' => $values[1], 'c' => $values[2]], $values[3]);
}

or like this:

$a = 1.5E+2;
$b = 18;

return Evaluator::mathExec("$a * 4 + 30 * 2 + $b / 3");

and this also is true:

var_dump(Evaluator::mathExec("floor((0.7+0.1)*10)") == 8);
// boolean true

var_dump([floor((0.7+0.1)*10) == 8]);
// boolean false

It utilizes a Pratt parser, and uses ext-bcmath extension to evaluate operations. For example this expression

$a = 4;
$b = 4; 
$c = 4;

Evaluator::mathExec("$a + $b / $c", 40);

internally turns into this

bcadd($a, bcdiv($b, $c, 40), 40);

it is very nice and convenient to write expressions naturally, without using bcadd, bcdiv, etc.

Installation

Use the package manager composer to install it

composer install ibelousov/math-exec

Usage

First of all import class Evaluator:

use \Ibelousov\MathExec\Evaluator\Evaluator;

Supported things

// Root  
Evaluator::mathExec("\\2"); 
// "1.1892071150027210667174999705604759152929"

(you could set inner precision, like this:

Evaluator::mathExec("\\2", 1000); 

// "1.1892071150027210667174999705604759152929720924638174130190022
// 2471946666822691715987078134453813767371603739477476921318606372
// 6361789847756785360862538017775070151511403557092273162342868889
// 9241754460719087105038499725591050098371044920154845735674580904
// 8399409309000349779590803848965884300504119871700937907982098462
// 5235373981281740818113780828552014842210060958932412445931035057
// 5191963029413832634742802798244080228008217292720586153666393704
// 0023820730854565306744771485988873345762718678381165470458727612
// 7111269988678434930175861424970170054131455143891998743766762178
// 5161783177987307048236318734734842180537156986842636482761056228
// 4779958628963329392816878747586560347379199645940075615444371574
// 1890303986971294306248625351734129153597531121544674615908647760
// 6517445957055930979119465756398917686972170262497475333629918606
// 5311570834936807698049481706074376847467855865282550141846497924
// 8909951563378299859508764353239662147789654791045418693466186139
// 614521856391702634160435422985610854932687"

by default it is 40 signs after point )

// Multiplication

Evaluator::mathExec("2 * 2", 0); 
//  "4"

// Division

Evaluator::mathExec("2 / 2", 0); 
// "1"

// Power
Evaluator::mathExec("2 ^ 3", 0); 
// "8" (left and right should be whole numbers)

// Modul
Evaluator::mathExec("7 % 2", 0); 
// "1"

// Whole division
Evaluator::mathExec("3.1415 // 2", 0); 
// "1"

// Associativity
Evaluator::mathExec("2 + 2 * 2", 0);  
// "6"

// Parenthesis
Evaluator::mathExec("(2 + 2) * 2", 0); 
// "8"

// Float to string convertion number formats
$a = 0.1415E-10;
$b = 0.1415E-10;
Evaluator::mathExec("$a * $b"); 
// "0.0000000000000000000002002225000000000000"

// comparison
Evaluator::mathExec("4 ^ 128 > 4 ^ 64"); 
// "1"

Evaluator::mathExec("4 ^ 128 < 4 ^ 64"); 
// "0"

Evaluator::mathExec("4 ^ 64 + 1 >= 4 ^ 64"); 
// "1"

Evaluator::mathExec("4 ^ 64 <= 4 ^ 64"); 
// "1"

Evaluator::mathExec("4 == 4"); 
// "1"

Evaluator::mathExec("4 != 4"); 
// "0"

Functions

Evaluator::mathExec("floor(3.1415)"); 
// "3"

Evaluator::mathExec("ceil(3.1415)");
// "4"

Evaluator::mathExec("format(ceil(3.1415) + floor(3.1415), 2)");
// "7.00"

Also you can add your own functions like this:

\Ibelousov\MathExec\Evaluator\BuiltinCollection::addBuiltin('inc', function($args) {
    $number_parts = explode('.', $args[0]->value);
    $precision = isset($number_parts[1]) ? strlen($number_parts[1]) : 0;

    return new \Ibelousov\MathExec\Evaluator\NumberObj(bcadd($args[0]->value, '1', $precision));
});

echo Evaluator::mathExec('inc(inc(2))',40);

Use with cautiousness

Inner representation of numbers

For example if you call this

Evaluator::mathExec("4/4 == 6/5", 0);

it evaluates to 1, because inner representation in this case is 0, and when you divide 6/5 you get 1 and not 1.2

For accurate numbers comparison you should set precision properly. In case shown above, you should do this, to properly compare numbers

Evaluator::mathExec("4/4 == 6/5", 1);

Converting to native PHP number formats

(int)Evaluator::mathExec((string)PHP_INT_MAX); 
// Evaluates to PHP_INT_MAX number

(int)Evaluator::mathExec((string)PHP_INT_MIN);
// Evaluates to PHP_INT_MIN number

(float)Evaluator::mathExec('\\2 + \\2', 30);
// cuts result 2.8284271247461900976033774484193961571392 to 2.8284271247462

(float)Evaluator::mathExec((string)PHP_FLOAT_MIN);
// Evaluates to PHP_FLOAT_MIN number

(float)Evaluator::mathExec('1.7976931348623157E+308');
// Evaluates to PHP_FLOAT_MAX. Today i have no time to figure out, why is
// (float)\Ibelousov\MathExec\Evaluator\Evaluator::mathExec((string)PHP_FLOAT_MAX)
// doesnt return correct value, but it is a fact

Other convertions, like this

(int)Evaluator::mathExec((string)PHP_INT_MIN . ' - 1');

are unpredictable

To evaluate and then compare numbers with high precision and/or big numbers, you should prefer storing values in strings instead of converting them into float or int.

REPL

You can use REPL to test expression evaluation:

$ ./vendor/ibelousov/math-exec/src/REPL.php
>> 4+4*4
20
Executed in 0.0025420188903809s.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

License

MIT

ibelousov / math-exec

Maintainers

Details