it::convertregex() sets /u if default_charset is 'utf-8' (can be overridden)

1 files changed, 8 insertions, 19 deletions

diff --git a/it.class b/it.class
index 818c9e0..292c154 100644
--- a/it.class
+++ b/it.class
@@ -310,35 +310,24 @@ static function toascii($text)
  * @param $p['casesensitive'] Regex is case sensitive (omit modifier i)
  * @param $p['multiline'] add modifier m: ^ and $ match \n
  * @param $p['singleline'] add modifier s: . matches \n
- * @param $p['utf8'] add modifier u
+ * @param $p['utf8'] add modifier u. This is the default if default_charset is utf-8, override with $p['utf8'] = false
  * @param $p['extended'] add modifier x (non signifcant whitespace)
  * @return converted regex to use with preg
  */
 static function convertregex($pattern, $p = array())
 {
-	$pattern = preg_replace('|/|', '\/', $pattern); 
-	$modifiers = '';
-
-	if (!$p['casesensitive'])
-		$modifiers .= 'i';
-
 	if ($p['exec'])
 		it::fatal("Option exec to it::replace has been removed for security reasons");
 
-	foreach (array(
-			'multiline'  => 'm',
-			'singleline' => 's',
-			'utf8'       => 'u',
-			'extended'   => 'x',
-		) as $key => $mod)
-	{
-		if ($p[$key])
-			$modifiers .= $mod;
-	}
-
-	return  "/$pattern/$modifiers";
+	return '/' . strtr($pattern, array('/' => '\/')) . '/' .
+		(!$p['casesensitive'] ? 'i' : '') .
+		($p['multiline'] ? 'm' : '') .
+		($p['singleline'] ? 's' : '') .
+		($p['extended'] ? 'x' : '') .
+		((!isset($p['utf8']) && ini_get('default_charset') == 'utf-8' || $p['utf8']) ? 'u' : '');
 }
 
+
 /**
  * Try to match string against regex. Case insensitive by default.
  * @param $pattern Regex to match against