Main Content

Regular Expressions in PHP

Archive - Originally posted on "The Horse's Mouth" - 2008-09-16 07:52:55 - Graham Ellis

Regular Expressions allow you to check if a string of text matches a particular pattern - for example to see if the data that a user has entered into a form looks like a Postcode / zipcode ... and then to extract the vital parts of the string that you're checking into separate variables.

Down the left here, I've written the regular expression to match a British Postcode - this involves describing what a postcode looks like in great detail.

Here are some of the elements:
^ - Starts with
[A-Z] - Capital Letter
[0-9] - Digit
+ - one or more of the previous item
? - previous item is optional
{2} - exactly 2 of the previous item
$ - ends with

The round brackets are known as "capture parentheses" - or in easy English, they're used to indicate which parts of the pattern match against interesting bits of the incoming string that you would like to capture into separate variables when you match.



There are complete books on regular expressions, but this summary to the right shows you the basic types of elements ...

Anchors or "zero width assersions" indicate thinsg like "starts with" and "ends with" so you can match at the beginning, in the middle, at the end of, or a whole string.

Literals match a single character (perhaps modified by a following count) exactly.

Character groups match any one character from a whole group of possibilities (perhaps modified by a following count) exactly.

Counts say how many times the previous element may occur - the default is once.

You can also use (...) to group elements for capture, | which means "or" (a.k.a. alternation) and a few other types.


There are two regular expression engines in PHP - there's the ereg series which are based on POSIX regular expressions, and the preg series which are based on the Perl standard. In easy terms, the preg flavour are shorter to write and more efficient in operation, but harder to understand and maintain.

One of the big questions with Regular Expressions is "are they case sensitive?" By default, yes they are, but you can add an "i" to ignore case - in different places in the syntax depending on if you're using ereg or preg.


If you're matching and capturing, you need to be very careful just which parts of the regular expression engine match which parts of the incoming string, as there can be many alternatives. You should remember that matching always starts as far to the left as it can, and that counts such as "+" always match as MANY characters as possible. If you use counts such as "+?" instead, then you're doing what is called a "sparse" rather than a "greedy" match, and that will match as few characters as possible.

The diagram shows the effect of sparse v greedy on a match to an email address with two @ characters in it. Most of the time, it'll be greedy you want to use ... sparse is useful especially when you're identifying / stripping HTML tags, though.


The boards / notes above come from the PHP Techniques Workshop that I was presenting yesterday ... it's a regular public course and we would welcome YOU on it ;-)