From 4f0030dd466f56b3dc0c864656fb1bf0e76d2932 Mon Sep 17 00:00:00 2001
From: Andreas Gohr <andi@splitbrain.org>
Date: Sun, 6 Feb 2011 19:07:31 +0100
Subject: ignore soft-hyphens for search FS#2049

This makes it possible to find words that include soft-hyphens. However,
search higlighting will not work and I have no idea how to make it work.
---
 doku.php         | 1 +
 inc/fulltext.php | 1 +
 inc/indexer.php  | 9 ++++++++-
 3 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/doku.php b/doku.php
index 6cd0c0e0c..dc5e0ec66 100644
--- a/doku.php
+++ b/doku.php
@@ -27,6 +27,7 @@ if (isset($_SERVER['HTTP_X_DOKUWIKI_DO'])){
 require_once(DOKU_INC.'inc/init.php');
 
 //import variables
+$_REQUEST['id'] = str_replace("\xC2\xAD",'',$_REQUEST['id']); //soft-hyphen
 $QUERY = trim($_REQUEST['id']);
 $ID    = getID();
 
diff --git a/inc/fulltext.php b/inc/fulltext.php
index be3938cac..0f2414213 100644
--- a/inc/fulltext.php
+++ b/inc/fulltext.php
@@ -304,6 +304,7 @@ function ft_pagesorter($a, $b){
  */
 function ft_snippet($id,$highlight){
     $text = rawWiki($id);
+    $text = str_replace("\xC2\xAD",'',$text); // remove soft-hyphens
     $evdata = array(
             'id'        => $id,
             'text'      => &$text,
diff --git a/inc/indexer.php b/inc/indexer.php
index 9cf079261..526c8db05 100644
--- a/inc/indexer.php
+++ b/inc/indexer.php
@@ -221,7 +221,14 @@ function idx_getPageWords($page){
 
     list($page,$body) = $data;
 
-    $body   = strtr($body, "\r\n\t", '   ');
+    $body   = strtr($body,
+                    array(
+                        "\r" => ' ',
+                        "\n" => ' ',
+                        "\t" => ' ',
+                        "\xC2\xAD" => '', //soft-hyphen
+                    )
+                   );
     $tokens = explode(' ', $body);
     $tokens = array_count_values($tokens);   // count the frequency of each token
 
-- 
cgit v1.2.3