PHP Regular Expressions Matching Words

PHP Regular Expressions 

Matching Words

Problem

You want to pull out all words from a string.

Solution

The simplest way to do this is to use the PCRE “word character” character type escape sequence, \w:

       $text = "Knock, knock. Who's there? r2d2!";

       $words = preg_match_all('/\w+/', $text, $matches);

       var_dump($matches[0]);

Discussion

The \w escape sequence matches letters, digits, and underscores. It does not include other punctuation. So the output from the preceding code is:

       array(6) {

            [0]=>

            string(5) "Knock"

            [1]=>

            string(5) "knock"

            [2]=>

            string(3) "Who"

            [3]=>

            string(1) "s"

            [4]=>

            string(5) "there"

            [5]=>

            string(4) "r2d2"

       }

This is mostly correct except that Who’s is broken up into Who and s. To extend this pattern to handle English contractions properly, we can match against either a word character or an apostrophe sandwiched by word characters:

       $text = "Knock, knock. Who's there? r2d2!";

       $pattern = "/(?:\w'\w|\w)+/";

       $words = preg_match_all($pattern, $text, $matches);

       var_dump($matches[0]);

(The ?: syntax in this pattern prevents the text that matches the parenthesized subpattern from being “captured.”)

With the addition of the u modifier, a pattern becomes Unicode-aware and will handle words properly in non-ASCII character sets. For example:

       $fr = 'Toc, toc. Qui est là? R2D2!';

       $fr_words = preg_match_all('/\w+/u', $fr, $matches);

       print "The French words are:\n\t";

       print implode(', ', $matches[0]) . "\n";

       $kr = '노크, 노크. 거기 누구입니까? R2D2!';

       $kr_words = preg_match_all('/\w+/u', $kr, $matches);

       print "The Korean words are:\n\t";

       print implode(', ', $matches[0]) . "\n";

This prints:

       The French words are:

                       Toc, toc, Qui, est, là, R2D2

       The Korean words are:

                       노크, 노크, 거기, 누구입니까, R2D2

Without that u at the end of each pattern, the non-ASCII characters would be stripped out of the matches, producing incorrect results.