4 * Copyright 2007 Reini Urban
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 along
19 * with PhpWiki; if not, write to the Free Software Foundation, Inc.,
20 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
23 require_once 'lib/plugin/SemanticSearch.php';
26 * Advanced search for relations/attributes and its values.
27 * Parse the query string, which can contain full mathematical expressions
28 * and various logical and mathematical functions and operators.
29 * Support subqueries (for _pagename ...) and temporary variables
32 * Are multiple variables valid for one page only, or is the result
33 * constructed as list of all matches? We'll stick with one page for now.
34 * This the only way I can see semantic meaning for now.
36 * Simple queries, with no variables, only against the pagename (implicit):
37 * "is_a::city and (population < 1.000.000 or population > 10.000.000)"
38 * "(is_a::city or is_a::country) and population < 10.000.000"
40 * Subqueries, with variables bound to the matching pagename, with (for ...):
41 * "works_at::_organization
42 * and (for _organization located_in::_city
43 * and (for _city population>1000000))"
45 * "works_at::_organization
46 * and (for _organization
48 * and (for _city is_a::City
49 * and population>1000000))
50 * or (located_in::_country
51 * and (for _country is_a::Country and population>5000000)))
53 * Relation links may contain wildcards. For relation and attribute names I'm not sure yet.
55 * @author: Reini Urban
58 class WikiPlugin_SemanticSearchAdvanced
59 extends WikiPlugin_SemanticSearch
63 return _("SemanticSearchAdvanced");
66 function getDescription()
68 return _("Parse and execute a full query expression");
71 function getDefaultArguments()
75 PageList::supportedArgs(), // paging and more.
77 's' => "", // query expression
78 'page' => "*", // which pages (glob allowed), default: all
79 'case_exact' => false,
80 'regex' => 'auto', // hmm
81 'noform' => false, // don't show form with results.
82 'noheader' => false // no caption
86 function showForm(&$dbi, &$request, $args, $allrelations)
89 $action = $request->getPostURL();
90 $hiddenfield = HiddenInputs($request->getArgs(), '',
91 array('action', 'page', 's'));
92 $pagefilter = HTML::input(array('name' => 'page',
93 'value' => $args['page'],
94 'title' => _("Search only in these pages. With autocompletion."),
95 'class' => 'dropdown',
96 'acdropdown' => 'true',
97 'autocomplete_complete' => 'true',
98 'autocomplete_matchsubstring' => 'false',
99 'autocomplete_list' => 'xmlrpc:wiki.titleSearch ^[S] 4'
101 $help = Button('submit:semsearch[help]', "?", false);
102 $svalues = empty($allrelations) ? "" : join("','", $allrelations);
103 $reldef = JavaScript("var semsearch_relations = new Array('" . $svalues . "')");
104 $querybox = HTML::textarea(array('name' => 's',
105 'title' => _("Enter a valid query expression"),
107 'acdropdown' => 'true',
108 'autocomplete_complete' => 'true',
109 'autocomplete_assoc' => 'false',
110 'autocomplete_matchsubstring' => 'true',
111 'autocomplete_list' => 'array:semsearch_relations'
113 $submit = Button('submit:semsearch[relations]', _("Search"), false,
114 array('title' => 'Move to help page. No seperate window'));
115 $instructions = _("Search in all specified pages for the expression.");
116 $form = HTML::form(array('action' => $action,
118 'accept-charset' => $GLOBALS['charset']),
120 $hiddenfield, HiddenInputs(array('attribute' => '')),
121 $instructions, HTML::br(),
122 HTML::table(array('border' => '0', 'width' => '100%'),
123 HTML::tr(HTML::td(_("Pagename(s): "), $pagefilter),
124 HTML::td(array('align' => 'right'),
126 HTML::tr(HTML::td(array('colspan' => 2),
129 HTML::div(array('align' => 'center'), $submit));
133 function run($dbi, $argstr, &$request, $basepage)
137 $this->_supported_operators = array(':=', '<', '<=', '>', '>=', '!=', '==', '=~');
138 $args = $this->getArgs($argstr, $request);
139 $posted = $request->getArg('semsearch');
140 $request->setArg('semsearch', false);
141 if ($request->isPost() and isset($posted['help'])) {
142 $request->redirect(WikiURL(_("Help/SemanticSearchAdvancedPlugin"),
143 array('redirectfrom' => $basepage), true));
145 $allrelations = $dbi->listRelations();
146 $form = $this->showForm($dbi, $request, $args, $allrelations);
147 if (isset($this->_norelations_warning))
148 $form->pushContent(HTML::div(array('class' => 'warning'),
149 _("Warning:") . $this->_norelations_warning));
151 // For convenience, peace and harmony we allow GET requests also.
152 if (!$args['s']) // check for good GET request
153 return $form; // nobody called us, so just display our form
155 // In reality we have to iterate over all found pages.
156 // To makes things shorter extract the next AND required expr and
157 // iterate only over this, then recurse into the next AND expr.
158 // => Split into an AND and OR expression tree.
160 $parsed_relations = $this->detectRelationsAndAttributes($args['s']);
162 if ($parsed_relations)
163 $regex = preg_grep("/[\*\?]/", $parsed_relations);
164 // Check that all those do exist.
166 $this->error("Invalid query: No relations or attributes in the query $s found");
167 $pagelist = new PageList($args['info'], $args['exclude'], $args);
169 $pagelist->setCaption
170 (HTML($noform ? '' : HTML($form, HTML::hr()),
171 fmt("Semantic %s Search Result for \"%s\" in pages \"%s\"",
174 if (!$regex and $missing = array_diff($parsed_relations, $allrelations))
176 $relquery = new TextSearchQuery(join(" ", $parsed_relations));
177 if (!$relquery->match(join(" ", $allrelations)))
179 $pagequery = new TextSearchQuery($page, $args['case_exact'], $args['regex']);
180 // if we have only numeric or text ops we can optimize.
181 //$parsed_attr_ops = $this->detectAttrOps($args['s']);
184 $linkquery = new TextSearchQuery($s, $args['case_exact'], $args['regex']);
185 $links = $dbi->linkSearch($pagequery, $linkquery, 'relation', $relquery);
186 $pagelist->_links = array();
187 while ($link = $links->next()) {
188 $pagelist->addPage($link['pagename']);
189 $pagelist->_links[] = $link;
191 $pagelist->addColumnObject
192 (new _PageList_Column_SemanticSearch_relation('relation', _("Relation"), $pagelist));
193 $pagelist->addColumnObject
194 (new _PageList_Column_SemanticSearch_link('link', _("Link"), $pagelist));
199 // ... (for _variable subquery) ...
200 function bindSubquery($query)
204 // is_a::city* and (population < 1.000.000 or population > 10.000.000)
205 // => is_a population
206 // Do we support wildcards in relation names also? is_*::city
207 function detectRelationsAndAttributes($subquery)
209 $relations = array();
210 // relations are easy
211 //$reltoken = preg_grep("/::/", preg_split("/\s+/", $query));
212 //$relations = array_map(create_function('$a','list($f,$b)=split("::",$a); return $f'),
214 foreach (preg_split("/\s+/", $query) as $whitetok) {
215 if (preg_match("/^([\w\*\?]+)::/", $whitetok))
216 $relations[] = $m[1];
219 // for attributes we might use the tokenizer. All non-numerics excl. units and non-ops
227 // c-hanging-comment-ender-p: nil
228 // indent-tabs-mode: nil