Image via Shutterstock

Once in a while I come up with a bit of code I think might be useful to others, and here’s the latest example: a regular expression that parses a string of text looking for Twitter hashtags (preceded with ‘#’) or usernames (preceded with ‘@’), and replaces them with HTML links to those pages at twitter.com. Every bit of text posted at LGF in a comment, page or article gets passed through this code.

It’s written in PHP, and uses preg_replace_callback() to process the matches with an anonymous function.

The regular expression is fairly sophisticated; named capturing groups are used to make it easier to reference the groups (instead of counting up parentheses), and it uses positive lookbehind combined with a lookaround conditional in order to avoid converting patterns that look like hashtags or usernames if they’re already contained within an HTML link.

Don’t panic, I won’t go into any more depth on the regular expression; I’ll leave that as an exercise for those who may be so inclined.

$text = preg_replace_callback(
	'~
		(?<=
			^
			|
			(?<=
				[^a-zA-Z0-9-.&]
			)
		)
		(?(?=
				]*>.+?
			)
			(?:
				]*>.+?
			)
			|
			(?
				[@#]{1}
			)
			(?
				[A-Za-z_]+[A-Za-z0-9_]+
			)
		)
	~xsi',
	function($matches) {
		if ($matches['name']) {
			return (
				'#' 
					: $matches['name'] . '/with_replies">@'
				) . $matches['name'] . ''
			);
		} else {
			return $matches[0];
		}
	},
	$text
);