From 546d3a9994d9f365f75e55f2f22601dc4941f5d5 Mon Sep 17 00:00:00 2001
From: Andreas Gohr <andi@splitbrain.org>
Date: Mon, 10 Mar 2008 22:49:39 +0100
Subject: another change in highlight handling

Now higlighting phrases are passed as an array which then is quoted correctly
when used in a regexp. This should make phrase highlighting work completely
correct.

Please everyone test it.

darcs-hash:20080310214939-7ad00-1abefb02dde40edeead50b4fa5c866c46b95ca3a.gz
---
 inc/fulltext.php | 26 ++++++++++++--------------
 1 file changed, 12 insertions(+), 14 deletions(-)

(limited to 'inc/fulltext.php')

diff --git a/inc/fulltext.php b/inc/fulltext.php
index a0be280bf..59053cb59 100644
--- a/inc/fulltext.php
+++ b/inc/fulltext.php
@@ -10,24 +10,21 @@
   require_once(DOKU_INC.'inc/indexer.php');
 
 
-/**
- * Wrapper around preg_quote adding the default delimiter
- */
-function ft_preg_quote_cb($string){
-    return preg_quote($string,'/');
-}
-
 /**
  * The fulltext search
  *
  * Returns a list of matching documents for the given query
  *
  */
-function ft_pageSearch($query,&$regex){
+function ft_pageSearch($query,&$highlight){
     $q = ft_queryParser($query);
 
+    $highlight = array();
+
     // remember for hilighting later
-    $regex = str_replace('*','',join('|',$q['words']));
+    foreach($q['words'] as $wrd){
+        $highlight[] =  str_replace('*','',$wrd);
+    }
 
     // lookup all words found in the query
     $words  = array_merge($q['and'],$q['not']);
@@ -76,12 +73,10 @@ function ft_pageSearch($query,&$regex){
     if(!count($docs)) return array();
     // handle phrases
     if(count($q['phrases'])){
-        //build a regexp
         $q['phrases'] = array_map('utf8_strtolower',$q['phrases']);
-        $q['phrases'] = array_map('ft_preg_quote_cb',$q['phrases']);
         // use this for higlighting later:
-        if($regex !== '') $regex .= '|';
-        $regex .= join('|',$q['phrases']);
+        $highlight = array_merge($highlight,$q['phrases']);
+        $q['phrases'] = array_map('preg_quote_cb',$q['phrases']);
         // check the source of all documents for the exact phrases
         foreach(array_keys($docs) as $id){
             $text  = utf8_strtolower(rawWiki($id));
@@ -250,13 +245,16 @@ function ft_pageLookup($id,$pageonly=true){
  *
  * @author Andreas Gohr <andi@splitbrain.org>
  */
-function ft_snippet($id,$re){
+function ft_snippet($id,$highlight){
     $text     = rawWiki($id);
     $match = array();
     $snippets = array();
     $utf8_offset = $offset = $end = 0;
     $len = utf8_strlen($text);
 
+    // build a regexp from the phrases to highlight
+    $re = join('|',array_map('preg_quote_cb',array_filter((array) $highlight)));
+
     for ($cnt=3; $cnt--;) {
       if (!preg_match('#('.$re.')#iu',$text,$match,PREG_OFFSET_CAPTURE,$offset)) break;
 
-- 
cgit v1.2.3