Generate a PHP compatible regular expression to match emoji from the most recent unicode data.

build-emoji-regexp.php

<?php

/**
 * Uses the data from unicode.org's emoji-data.txt to build a PHP compatible Regular Expression
 * along the lines of:
 * (?:\p{Emoji_Modifier_Base}\p{Emoji_Modifier}?|\p{Emoji_Presentation}|\p{Emoji}\x{FE0F}?)
 *
 * To use: php build-hashtag-regexp.php <emoji-data.txt>
 * Output will be the generated regular expression.
 *
 * Get a current copy of emoji-data.txt from http://www.unicode.org/Public/emoji/latest/emoji-data.txt
  */

if (!isset($argv[1]) || !file_exists($argv[1])) {
	print "Usage: build-hashtag-regexp.php <emoji-data.txt>\n";
	print "\n";
	print "Prints the generated PHP compatible Regular Expression to STDOUT.\n";
	print "\n";
	print "Grab emoji-data.txt from http://www.unicode.org/Public/emoji/latest/emoji-data.txt\n";
	die(1);
}

$emojiFilename = $argv[1];
$emojiFile = file($emojiFilename);
$emojiData = array();

$emojiClasses = array();
foreach ($emojiFile as $line) {
	$pos = strpos($line,'#');
	if ($pos!==false) {
		$line = substr($line,0,$pos);
	}
	$line = trim($line);
	if (!$line) {
		continue;
	}

	$line = explode(';',$line);
	if (count($line)!=2) {
		continue;
	}

	$range = strtoupper(trim($line[0]));
	$class = trim($line[1]);
	if (!isset($emojiClasses[$class])) {
		$emojiClasses[$class] = array();
	}

	$range = explode('..',$range);
	if (count($range)==1) {
		$emojiClasses[$class][] = '\\x{'.$range[0].'}';
	}
	else {
		$emojiClasses[$class][] = '[\\x{'.$range[0].'}-\\x{'.$range[1].'}]';
	}

}

$emojiRegexp = '(?:\\p{Emoji_Modifier_Base}\\p{Emoji_Modifier}?|\\p{Emoji_Presentation}|\\p{Emoji}\\x{FE0F}?)';

foreach ($emojiClasses as $class=>$components) {
	$emojiRegexp = str_replace('\\p{'.$class.'}','(?:'.implode('|',$components).')',$emojiRegexp);
}

print $emojiRegexp;