From 12dac4911b91e80cf1d8c85ee8ad1ef191b630cb Mon Sep 17 00:00:00 2001 From: thomascube <thomas@roundcube.net> Date: Tue, 12 Apr 2011 14:01:49 -0400 Subject: [PATCH] Handle unicode strings when normalizing for search (#1487866) --- program/include/rcube_addressbook.php | 24 +++++++++++++++++------- 1 files changed, 17 insertions(+), 7 deletions(-) diff --git a/program/include/rcube_addressbook.php b/program/include/rcube_addressbook.php index 9e8254c..8ec0abb 100644 --- a/program/include/rcube_addressbook.php +++ b/program/include/rcube_addressbook.php @@ -397,14 +397,24 @@ */ protected static function normalize_string($str) { - $norm = strtolower(strtr(utf8_decode($str), - '��������������������������������������������������', - 'ccaaaaeeeeiiiaeooouuuyooaiounnaaaaaeeeiiioooouuuyy')); - - return preg_replace( - array('/[\s;\+\-\/]+/i', '/(\d)\s+(\d)/', '/\s\w{1,3}\s/'), + // split by words + $arr = explode(" ", preg_replace( + array('/[\s;\+\-\/]+/i', '/(\d)[-.\s]+(\d)/', '/\s\w{1,3}\s/'), array(' ', '\\1\\2', ' '), - $norm); + $str)); + + foreach ($arr as $i => $part) { + if (utf8_encode(utf8_decode($part)) == $part) { // is latin-1 ? + $arr[$i] = strtr(strtolower(strtr(utf8_decode($part), + '��������������������������������������������������', + 'ccaaaaeeeeiiiaeooouuuyooaiounnaaaaaeeeiiioooouuuyy')), + array('�' => 'ss', 'ae' => 'a', 'oe' => 'o', 'ue' => 'u')); + } + else + $arr[$i] = strtolower($part); + } + + return join(" ", $arr); } } -- Gitblit v1.9.1