summaryrefslogtreecommitdiff
path: root/program/lib/Roundcube
diff options
context:
space:
mode:
authorThomas Bruederli <bruederli@kolabsys.com>2014-12-28 16:22:08 +0100
committerThomas Bruederli <bruederli@kolabsys.com>2014-12-28 16:22:08 +0100
commite8b82c2e7b0ae2e5d45ecb600813b8990568feb9 (patch)
tree9522b21d1161fdb9ae37e1a8572b3d3670836b96 /program/lib/Roundcube
parent09c58d1adde92a60a3e7cd67f4e66c8b1a56be6a (diff)
Fix rcube_utils::normalize_string() to support unicode characters + add argument for minimum token length
Diffstat (limited to 'program/lib/Roundcube')
-rw-r--r--program/lib/Roundcube/rcube_utils.php22
1 files changed, 15 insertions, 7 deletions
diff --git a/program/lib/Roundcube/rcube_utils.php b/program/lib/Roundcube/rcube_utils.php
index 2e4aa323e..f4c0e90ca 100644
--- a/program/lib/Roundcube/rcube_utils.php
+++ b/program/lib/Roundcube/rcube_utils.php
@@ -912,14 +912,21 @@ class rcube_utils
* Split the given string into word tokens
*
* @param string Input to tokenize
+ * @param integer Minimum length of a single token
* @return array List of tokens
*/
- public static function tokenize_string($str)
+ public static function tokenize_string($str, $minlen = 2)
{
- return explode(" ", preg_replace(
- array('/[\s;\/+-]+/i', '/(\d)[-.\s]+(\d)/', '/\s\w{1,3}\s/u'),
- array(' ', '\\1\\2', ' '),
- $str));
+ $expr = array('/[\s;\/+-]+/ui', '/(\d)[-.\s]+(\d)/u');
+ $repl = array(' ', '\\1\\2');
+
+ if ($minlen > 1) {
+ $minlen--;
+ $expr[] = "/(^|\s+)\w{1,$minlen}(\s+|$)/u";
+ $repl[] = ' ';
+ }
+
+ return array_filter(explode(" ", preg_replace($expr, $repl, $str)));
}
/**
@@ -928,10 +935,11 @@ class rcube_utils
*
* @param string Input string (UTF-8)
* @param boolean True to return list of words as array
+ * @param integer Minimum length of tokens
*
* @return mixed Normalized string or a list of normalized tokens
*/
- public static function normalize_string($str, $as_array = false)
+ public static function normalize_string($str, $as_array = false, $minlen = 2)
{
// replace 4-byte unicode characters with '?' character,
// these are not supported in default utf-8 charset on mysql,
@@ -943,7 +951,7 @@ class rcube_utils
. ')/', '?', $str);
// split by words
- $arr = self::tokenize_string($str);
+ $arr = self::tokenize_string($str, $minlen);
// detect character set
if (utf8_encode(utf8_decode($str)) == $str) {