README

This library is a small wrapper around PHP's BCMath extension. BCMath implements arbitrary precision mathematics, it's usage is not very comfortable.

This library aims to implement an easy to use interface for the most commonly needed mathematical operations using BCMath internally.

Installation

Install the package using composer:

composer require mehr-it/php-decimals

Introduction

Mathematical operations using PHP's built-in number types (float and integer) suffer from two drawbacks:

• limited range
• floating point inaccuracy

Whenever dealing with numbers out of the build-in number types' range or whenever floating point inaccuracy can become a problem (eg. currency calculations) an alternative solution is required.

In the same way as the BCMath extension, this library uses strings to represent numbers. Unlike PHP's float to string conversion (which uses a locale depended decimal separator), the decimal separator is always ".".

One thing you have to worry about when dealing with arbitrary precision mathematics is the precision you need. You have to tell how much precision you need for your results. Whenever possible, this library will do this for you. But for some operations you might have to manually set the required precision when you need extraordinary accuracy.

Usage

Any operations can be accessed using the static methods of the Decimals class:

Decimals::add('5.78', '7.1');

Basic mathematical operations

// add: 5.78 + 7.1
Decimals::add('5.78', '7.1');

// substract: 5.32 - 1.08
Decimals::sub('5.32', '1.08');

// multiply: 1.1 * 2.87
Decimals::mul('1.1', '2.87');

// divide: 5 / 2.5
Decimals::div('5', '2.5');

// modulus: 5 % 2
Decimals::mod('5', '2');

// pow: 5 ** 2
Decimals::pow('5', '2');

For add() and sub() the result precision is automatically set to the largest number of decimals of either operands. This is sufficient for any operation.

For mul(), div() and mod() the result precision is automatically set to the double of the largest number of decimals of either operands but at least 16 decimals. If you need higher precision, you may specify it as third parameter. The following example outputs 32 decimals:

Decimals::div('1', '3', 32);

The sum() function sums up all given operands:

Decimals::sum('1', '2.1', '-0.3'); 

Converting and parsing

Sometimes you need to convert a floating number to a string. As mentioned above, casting a floating number to a string uses the locale dependent decimal separator (e.g. "." for "en" and "," for "de"). This can cause a lot of problems when you have no control which locale your PHP interpreter uses (especially when writing libraries).

Convert native type to number string

Luckily this library ships with a function, converting these local strings to the BCMath compatible number strings (with "." as decimal separator). You may simply pass your number to the parse() function:

$float = 1.5;

$sNumber = Decimals::parse($float);

Always use the parse() function when converting native number types to string to write platform independent code!

Parse string with any decimal separator

The parse() function is also useful when you have numbers from external sources with (maybe different) decimal separator than the locale. You can safely convert these strings by passing the decimal separator as second argument:

$sNumber = Decimals::parse('1,67', ',');

// $sNumber: '1.67'

The parse function also validates the input and throws an InvalidArguemntException if no a valid number is given. (Whitespaces around a valid number string are gracefully ignored and number is parsed without exception)

Convert number string to float or int

In cases when you need to convert the value back to a native data type you can simply cast the string to the type or use the toNative() function which will return the correct type (float or int) automatically:

$float = Decimals::toNative('1.56'); 

$int = Decimals::toNative('167'); 

Normalize numbers

Often, a number can be represented in multiple forms:

• 0 can be expressed as "0.0", "0", "0." ".0", "-0", ...
• 5.8 can be expressed as "005.8", "5.8000", "+5.8" ...

The norm() function strips of any unneeded zeros before and after the number, remove any "+" signs and represent 0 always as "0":

Decimals::norm('+045.89');

Rounding and truncating numbers

These functions are pretty self explaining. round() rounds the number to the given number of decimals, while truncate() simply strips of decimals:

Decimals::round('5.4591', 0); // = '5'
Decimals::round('5.4591', 1); // = '5.5'
Decimals::round('5.4591', 2); // = '5.46'

Decimals::truncate('5.4591', 0); // = '5'
Decimals::truncate('5.4591', 1); // = '5.4'
Decimals::truncate('5.4591', 2); // = '5.45'

Comparison

To compare numbers, the comp() function can be used. It acts like the <=> operator, returning -1 if the left operand is lower, 0 if both operands are equal and 1 if the right operand is greater:

Decimals::comp('1.5', '6');   // = -1
Decimals::comp('1.5', '1.5'); // = 0
Decimals::comp('6', '1.5');   // = 1

To make code more readable, the following functions handle specific comparison cases:

Decimals::isEqual('1.5', '6');               // = false
Decimals::isNotEqual('1.5', '6');            // = true
Decimals::isGreaterThan('1.5', '6');         // = false
Decimals::isGreaterThanOrEqual('1.5', '6');  // = false
Decimals::isLessThan('1.5', '6');            // = true
Decimals::isLessThanOrEqual('1.5', '6');     // = true

To find the maximum or minimum of a given value set, the max() and min() functions exist:

Decimals::max('1', '2', '3');   // '3'
Decimals::min('1', '2', '3');   // '1'

Mathematical functions

// returns the absolute value
Decimals::abs('-1.5'); // = '1.5'
Decimals::abs('1.5');  // = '1.5'

Other functions

// returns the number of decimals of the given number
Decimals::decimals('78.8');   // = 1
Decimals::decimals('78.889'); // = 3
Decimals::decimals('78.800'); // = 3

Using expressions

Applying multiple operations can lead to unreadable code:

$result = Decimals::add(Decimals::mul($a, $b), $c);

The Decimals::expr() method (or the helper expr()) can help here. The same operations as above can be written as follows:

$result = Decimals::expr($a, '*', $b, '+', $c);

// or using helper
$result = expr($a, '*', $b, '+', $c);

This comes with the drawback of little performance overhead, but is very easy to read.

Expressions are always evaluated from left to right. If an expression is given which would break mathematical or logical operator precedence, an exception is thrown.

TODO

• add wrapper for bcpowmod() and bcsqrt()