4 Copyright 1999,2000,2001,2002,2004,2005 $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/TextSearchQuery.php');
24 require_once("lib/PageList.php");
27 * Case insensitive fulltext search
28 * Options: case_exact, regex, hilight
31 * TODO: Hooks to search in external documents: ExternalTextSearch
32 * Only uploaded: textfiles, PDF, HTML, DOC, XLS, ... or
33 * External apps: xapian-omages seems to be the best, over lucene.net,
34 * swish, nakamazu, ...
36 * See http://sf.net/tracker/index.php?aid=927395&group_id=6121&atid=106121
37 * Wordaround to let the dead locks occur somewhat later:
38 * increased the memory limit of PHP4 from 8 MB to 32 MB
39 * php.ini: memory_limit = 32 MB
41 class WikiPlugin_FullTextSearch
45 return _("FullTextSearch");
48 function getDescription() {
49 return _("Search the content of all pages in this wiki.");
52 function getVersion() {
53 return preg_replace("/[Revision: $]/", '',
57 function getDefaultArguments() {
60 PageList::supportedArgs(), // paging and more.
63 'case_exact' => false,
65 'sortby' => '-hi_content',
67 'exclude' => false, // comma-seperated list of glob
68 'quiet' => true)); // be less verbose
71 function run($dbi, $argstr, &$request, $basepage) {
73 $args = $this->getArgs($argstr, $request);
74 if (empty($args['s'])) {
75 return HTML::div(array('class' => "error"), "Please provide 's' argument to the plugin.");
79 $query = new TextSearchQuery($s, $case_exact, $regex);
80 $pages = $dbi->fullSearch($query, $sortby, $limit, $exclude);
82 $hilight_re = $hilight ? $query->getHighlightRegexp() : false;
85 if ($quiet) { // see how easy it is with PageList...
87 $args['listtype'] = 'dl';
88 $args['types'] = array(new _PageList_Column_content
89 ('rev:hi_content', _("Content"), "left", $s));
90 $list = new PageList(false, $exclude, $args);
91 $list->setCaption(fmt("Full text search results for '%s'", $s));
92 while ($page = $pages->next()) {
93 $list->addPage( $page );
98 // Todo: we should better define a new PageListDL class for dl/dt/dd lists
99 // But the new column types must have a callback then. (showhits)
100 // See e.g. WikiAdminSearchReplace for custom pagelist columns
102 if (!$limit or !is_int($limit))
104 // expand all page wildcards to a list of pages which should be ignored
105 if ($exclude) $exclude = explodePageList($exclude);
106 while ($page = $pages->next() and (!$limit or ($count < $limit))) {
107 $name = $page->getName();
108 if ($exclude and in_array($name,$exclude)) continue;
110 $list->pushContent(HTML::dt(WikiLink($page)));
112 $list->pushContent($this->showhits($page, $hilight_re));
115 if ($limit and $count >= $limit) //todo: pager link to list of next matches
116 $list->pushContent(HTML::dd(fmt("only %d pages displayed",$limit)));
117 if (!$list->getContent())
118 $list->pushContent(HTML::dd(_("<no matches>")));
120 if (!empty($pages->stoplisted))
121 $list = HTML(HTML::p(fmt(_("Ignored stoplist words '%s'"),
122 join(', ', $pages->stoplisted))),
126 return HTML(HTML::p(fmt("Full text search results for '%s'", $s)),
130 function showhits($page, $hilight_re) {
131 $current = $page->getCurrentRevision();
132 $matches = preg_grep("/$hilight_re/i", $current->getContent());
134 foreach ($matches as $line) {
135 $line = $this->highlight_line($line, $hilight_re);
136 $html[] = HTML::dd(HTML::small(array('class' => 'search-context'),
142 function highlight_line ($line, $hilight_re) {
143 while (preg_match("/^(.*?)($hilight_re)/i", $line, $m)) {
144 $line = substr($line, strlen($m[0]));
145 $html[] = $m[1]; // prematch
146 $html[] = HTML::strong(array('class' => 'search-term'), $m[2]); // match
148 $html[] = $line; // postmatch
154 * List of Links and link to ListLinks
156 class _PageList_Column_hilight extends _PageList_Column {
157 function _PageList_Column_WantedPages_links (&$params) {
158 $this->parentobj =& $params[3];
159 $this->_PageList_Column($params[0],$params[1],$params[2]);
161 function _getValue(&$page, $revision_handle) {
163 $pagename = $page->getName();
164 $count = count($this->parentobj->_wpagelist[$pagename]);
165 return LinkURL(WikiURL($page, array('action' => 'BackLinks'), false),
166 fmt("(%d Links)", $count));
174 // c-hanging-comment-ender-p: nil
175 // indent-tabs-mode: nil