PHP Cookbook

Chapter 2. Numbers

Introduction

In everyday life, numbers are easy to identify. They’re 3:00 P.M., as in the current time, or $1.29, as in the cost of a pint of milk. Maybe they’re like π, the ratio of the circumference to the diameter of a circle. They can be pretty large, like Avogadro’s number, which is about 6 x 10²³. In PHP, numbers can be all these things.

However, PHP doesn’t treat all these numbers as “numbers.” Instead, it breaks them down into two groups: integers and floating-point numbers. Integers are whole numbers, such as -4, 0, 5, and 1,975. Floating-point numbers are decimal numbers, such as -1.23, 0.0, 3.14159, and 9.9999999999.

Conveniently, most of the time PHP doesn’t make you worry about the differences between the two because it automatically converts integers to floating-point numbers and floating-point numbers to integers. This conveniently allows you to ignore the underlying details. It also means 3/2 is 1.5, not 1, as it would be in some programming languages. PHP also automatically converts from strings to numbers and back. For instance, 1+"1" is 2.

However, sometimes this blissful ignorance can cause trouble. First, numbers can’t be infinitely large or small; there’s a minimum size of 2.2e-308 and a maximum size of about 1.8e308.^[1] If you need larger (or smaller) numbers, you must use the BCMath or GMP libraries, which are discussed in Recipe 2.14.

Next, floating-point numbers aren’t guaranteed to be exactly correct but only correct plus or a minus ...