PHP/CSS: Extract all url definitions (e. g. background images) in CSS declarations

gistfile1.aw

/**
 * Extracts all url definitions (e. g. urls to background images) in CSS.
 *
 * This snippet could be usefull within a build process where css files are 
 * distributed in different folder during development and should be merged together.
 * Existing url definitions could be invalid in some cases.
 */

// example css content, could be read from a file
$cssFileContent = <<<CSS
/* background image url without quotes */
div.myBg {
   background-image: url(../images/bg.gif);
   background-color: orange;
}

/* background image url with double quotes */
p.success {
    background: green url("images/bg.gif") fixed no-repeat;
}

/* background image url with double quotes and spaces */
p.success2 {
    background-image: green url( "images/bg.gif" );
}

/* external url to a webfont with single quotes, example from http://www.google.com/webfonts */
@font-face {
    font-family: 'Rammetto One';
    font-style: normal;
    font-weight: 400;
    src: local('Rammetto One Regular'), local('RammettoOne-Regular'),
    url('http://themes.googleusercontent.com/static/fonts/rammettoone/v1/mh0uQ1tV8QgSx9v_KyEYPHhCUOGz7vYGh680lGh-uXM.woff') format('woff');
}
CSS;


// matches all url definitions inside the css file as follows:
// url("*")  URL definitions with double quotes
// url('*')  URL definitions with single quotes
// url(*)    URL definitions without quotes
preg_match_all('/url\(([\s])?([\"|\'])?(.*?)([\"|\'])?([\s])?\)/i', $cssFileContent, $matches, PREG_PATTERN_ORDER);

// in case of found matches, the multi-dimensional array $matches contains following
// important entries:
// $matches[0]  (array)  List containing each string matching the full pattern, e. g. url("images/bg.gif")
// $matches[3]  (array)  List containing matched url definitions, e. g. images/bg.gif
if ($matches) {
   foreach($matches[3] as $match) {
      // do whatever you want with those matches, adapt paths by changing them to absolute or CDN paths
      // - "images/bg.gif" -> "/path_to_css_module/images/bg.gif"
      // - "images/bg.gif" -> "http://cdn.domain.tld/path_to_css_module/images/bg.gif"

      echo $match . "<br/>";
   }
}

but why to match things that are not used? just removed the unnecessary brackets, so it would be $matches[1] not $matches[3] - now it's a waste of memory and waste of computing power