Regular expressions

A regular expression is a pattern for searching text. It checks whether a string looks like a phone number or an email, pulls numbers out of text, replaces fragments by a rule. In PHP patterns are handled by functions whose names start with preg_.

Matching a pattern

preg_match() returns 1 if the pattern is found and 0 if not. The pattern is written between slashes:

<?php
 
echo preg_match('/cat/', 'My cat sleeps'), "\n"; // 1
echo preg_match('/dog/', 'My cat sleeps');       // 0

The letter u after the pattern enables Unicode support, i makes the search case-insensitive: /cat/ui.

Notation

Checking a format

The ^ and $ signs require the whole string to match the pattern, not just a part of it:

<?php
 
$pattern = '/^\d{4}-\d{2}-\d{2}$/';
 
echo preg_match($pattern, '2025-01-30'), "\n"; // 1
echo preg_match($pattern, '30.01.2025');       // 0

Groups

A part of the pattern in parentheses is a group. The matched groups go into the array passed as the third argument:

<?php
 
preg_match('/(\d+) USD/', 'Price: 250 USD.', $matches);
 
echo $matches[0], "\n"; // 250 USD
echo $matches[1];       // 250

The element at index 0 is the whole match, followed by the groups in order.

Every match

preg_match_all() finds all the matches rather than the first one:

<?php
 
preg_match_all('/\d+/', 'Order 15, items 3, total 4200', $matches);
 
echo implode(', ', $matches[0]); // 15, 3, 4200

Replacing

preg_replace() replaces everything that matched the pattern:

<?php
 
echo preg_replace('/\s+/', ' ', 'too    many   spaces'), "\n";
echo preg_replace('/\D/', '', '+1 (555) 123-45-67');

The first line turns runs of spaces into a single space, the second leaves only the digits of the phone number: 15551234567.

Splitting

preg_split() splits a string by a pattern — for when there is more than one separator:

<?php
 
$parts = preg_split('/[,;]\s*/', 'tea, coffee; juice');
 
echo count($parts), ' ', $parts[2]; // 3 juice

Common mistake

Do not write complex patterns where a ready-made check exists. An email address or a URL is more reliably checked with filter_var():

<?php
 
var_dump(filter_var('anna@example.com', FILTER_VALIDATE_EMAIL) !== false); // bool(true)
var_dump(filter_var('not an email', FILTER_VALIDATE_EMAIL) !== false);     // bool(false)

And for a plain substring search str_contains() is enough — it is faster and clearer.