2 rcs_id('$Id: FuzzyPages.php,v 1.11 2004-02-17 12:11:36 rurban Exp $');
4 Copyright 1999, 2000, 2001, 2002 $ThePhpWikiProgrammingTeam
6 This file is part of PhpWiki.
8 PhpWiki is free software; you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation; either version 2 of the License, or
11 (at your option) any later version.
13 PhpWiki is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with PhpWiki; if not, write to the Free Software
20 Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23 //require_once('lib/PageList.php');
26 * FuzzyPages is plugin which searches for similar page titles.
28 * Pages are considered similar by averaging the similarity scores of
29 * the spelling comparison and the metaphone comparison for each page
30 * title in the database (php's metaphone() is an improved soundex
33 * http://www.php.net/manual/en/function.similar-text.php
34 * http://www.php.net/manual/en/function.metaphone.php
36 class WikiPlugin_FuzzyPages
40 return _("FuzzyPages");
43 function getDescription() {
44 return sprintf(_("Search for page titles similar to %s."),
48 function getVersion() {
49 return preg_replace("/[Revision: $]/", '',
50 "\$Revision: 1.11 $");
53 function getDefaultArguments() {
54 return array('s' => false,
58 function spelling_similarity($subject) {
59 $spelling_similarity_score = 0;
60 similar_text($subject, $this->_searchterm,
61 $spelling_similarity_score);
62 return $spelling_similarity_score;
65 function sound_similarity($subject) {
66 $sound_similarity_score = 0;
67 similar_text(metaphone($subject), $this->_searchterm_metaphone,
68 $sound_similarity_score);
69 return $sound_similarity_score;
72 function averageSimilarities($subject) {
73 return ($this->spelling_similarity($subject)
74 + $this->sound_similarity($subject)) / 2;
77 function collectSimilarPages(&$list, &$dbi) {
78 if (! defined('MIN_SCORE_CUTOFF'))
79 define('MIN_SCORE_CUTOFF', 33);
81 $this->_searchterm_metaphone = metaphone($this->_searchterm);
83 $allPages = $dbi->getAllPages();
85 while ($pagehandle = $allPages->next()) {
86 $pagename = $pagehandle->getName();
87 $similarity_score = $this->averageSimilarities($pagename);
88 if ($similarity_score > MIN_SCORE_CUTOFF)
89 $list[$pagename] = $similarity_score;
93 function sortCollectedPages(&$list) {
94 arsort($list, SORT_NUMERIC);
97 function addTableCaption(&$table, &$dbi) {
98 if ($dbi->isWikiPage($this->_searchterm))
99 $link = WikiLink($this->_searchterm, 'auto');
101 $link = $this->_searchterm;
102 $caption = fmt("These page titles match fuzzy with '%s'", $link);
103 $table->pushContent(HTML::caption(array('align'=>'top'), $caption));
106 function addTableHead(&$table) {
107 $row = HTML::tr(HTML::th(_("Name")),
108 HTML::th(array('align' => 'right'), _("Score")));
110 $this->_pushDebugHeadingTDinto($row);
112 $table->pushContent(HTML::thead($row));
115 function addTableBody(&$list, &$table) {
116 if (! defined('HIGHLIGHT_ROWS_CUTOFF_SCORE'))
117 define('HIGHLIGHT_ROWS_CUTOFF_SCORE', 60);
119 $tbody = HTML::tbody();
120 foreach ($list as $found_pagename => $score) {
121 $row = HTML::tr(array('class' =>
122 $score > HIGHLIGHT_ROWS_CUTOFF_SCORE
123 ? 'evenrow' : 'oddrow'),
124 HTML::td(WikiLink($found_pagename)),
125 HTML::td(array('align' => 'right'),
129 $this->_pushDebugTDinto($row, $found_pagename);
131 $tbody->pushContent($row);
133 $table->pushContent($tbody);
136 function formatTable(&$list, &$dbi) {
138 $table = HTML::table(array('cellpadding' => 2,
141 'class' => 'pagelist'));
142 $this->addTableCaption($table, $dbi);
143 $this->addTableHead($table);
144 $this->addTableBody($list, $table);
149 function run($dbi, $argstr, &$request, $basepage) {
150 $args = $this->getArgs($argstr, $request);
154 $this->debug = $debug;
156 $this->_searchterm = $s;
157 $this->_list = array();
159 $this->collectSimilarPages($this->_list, $dbi);
161 $this->sortCollectedPages($this->_list);
163 return $this->formatTable($this->_list, $dbi);
168 function _pushDebugHeadingTDinto(&$row) {
169 $row->pushContent(HTML::td(_("Spelling Score")),
170 HTML::td(_("Sound Score")),
171 HTML::td('Metaphones'));
174 function _pushDebugTDinto(&$row, $pagename) {
175 // This actually calculates everything a second time for each pagename
176 // so the individual scores can be displayed separately for debugging.
177 $debug_spelling = round($this->spelling_similarity($pagename), 1);
178 $debug_sound = round($this->sound_similarity($pagename), 1);
179 $debug_metaphone = sprintf("(%s, %s)", metaphone($pagename),
180 $this->_searchterm_metaphone);
182 $row->pushcontent(HTML::td(array('align' => 'center'), $debug_spelling),
183 HTML::td(array('align' => 'center'), $debug_sound),
184 HTML::td($debug_metaphone));
188 // $Log: not supported by cvs2svn $
189 // Revision 1.10 2003/02/22 20:49:55 dairiki
190 // Fixes for "Call-time pass by reference has been deprecated" errors.
192 // Revision 1.9 2003/01/18 21:41:02 carstenklapp
194 // Reformatting & tabs to spaces;
195 // Added copyleft, getVersion, getDescription, rcs_id.
202 // c-hanging-comment-ender-p: nil
203 // indent-tabs-mode: nil