From 4b9792c696658fe0cbedc187198fa463b6ff83fc Mon Sep 17 00:00:00 2001
From: Tom N Harris <tnharris@whoopdedo.org>
Date: Sun, 14 Nov 2010 14:22:08 -0500
Subject: Measure length of multi-character Asian words

---
 inc/indexer.php | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

(limited to 'inc')

diff --git a/inc/indexer.php b/inc/indexer.php
index 7a8bb3ff8..d9eccac76 100644
--- a/inc/indexer.php
+++ b/inc/indexer.php
@@ -52,8 +52,10 @@ function wordlen($w){
     $l = strlen($w);
     // If left alone, all chinese "words" will get put into w3.idx
     // So the "length" of a "word" is faked
-    if(preg_match('/'.IDX_ASIAN2.'/u',$w))
-        $l += ord($w) - 0xE1;  // Lead bytes from 0xE2-0xEF
+    if(preg_match_all('/[\xE2-\xEF]/',$w,$leadbytes)) {
+        foreach($leadbytes[0] as $b)
+            $l += ord($b) - 0xE1;
+    }
     return $l;
 }
 
-- 
cgit v1.2.3