Mercurial > defr > drupal > core

comparison modules/search/search.module @ 1:c1f4ac30525a 6.0

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Drupal 6.0
author Franck Deroche <webmaster@defr.org>
date Tue, 23 Dec 2008 14:28:28 +0100
parents
children 2427550111ae
comparison
equal deleted inserted replaced
0:5a113a1c4740 1:c1f4ac30525a
1 <?php
2 // $Id: search.module,v 1.250.2.1 2008/02/07 16:42:03 goba Exp $
3
4 /**
5 * @file
6 * Enables site-wide keyword searching.
7 */
8
9 /**
10 * Matches Unicode character classes to exclude from the search index.
11 *
12 * See: http://www.unicode.org/Public/UNIDATA/UCD.html#General_Category_Values
13 *
14 * The index only contains the following character classes:
15 * Lu Letter, Uppercase
16 * Ll Letter, Lowercase
17 * Lt Letter, Titlecase
18 * Lo Letter, Other
19 * Nd Number, Decimal Digit
20 * No Number, Other
21 */
22 define('PREG_CLASS_SEARCH_EXCLUDE',
23 '\x{0}-\x{2f}\x{3a}-\x{40}\x{5b}-\x{60}\x{7b}-\x{bf}\x{d7}\x{f7}\x{2b0}-'.
24 '\x{385}\x{387}\x{3f6}\x{482}-\x{489}\x{559}-\x{55f}\x{589}-\x{5c7}\x{5f3}-'.
25 '\x{61f}\x{640}\x{64b}-\x{65e}\x{66a}-\x{66d}\x{670}\x{6d4}\x{6d6}-\x{6ed}'.
26 '\x{6fd}\x{6fe}\x{700}-\x{70f}\x{711}\x{730}-\x{74a}\x{7a6}-\x{7b0}\x{901}-'.
27 '\x{903}\x{93c}\x{93e}-\x{94d}\x{951}-\x{954}\x{962}-\x{965}\x{970}\x{981}-'.
28 '\x{983}\x{9bc}\x{9be}-\x{9cd}\x{9d7}\x{9e2}\x{9e3}\x{9f2}-\x{a03}\x{a3c}-'.
29 '\x{a4d}\x{a70}\x{a71}\x{a81}-\x{a83}\x{abc}\x{abe}-\x{acd}\x{ae2}\x{ae3}'.
30 '\x{af1}-\x{b03}\x{b3c}\x{b3e}-\x{b57}\x{b70}\x{b82}\x{bbe}-\x{bd7}\x{bf0}-'.
31 '\x{c03}\x{c3e}-\x{c56}\x{c82}\x{c83}\x{cbc}\x{cbe}-\x{cd6}\x{d02}\x{d03}'.
32 '\x{d3e}-\x{d57}\x{d82}\x{d83}\x{dca}-\x{df4}\x{e31}\x{e34}-\x{e3f}\x{e46}-'.
33 '\x{e4f}\x{e5a}\x{e5b}\x{eb1}\x{eb4}-\x{ebc}\x{ec6}-\x{ecd}\x{f01}-\x{f1f}'.
34 '\x{f2a}-\x{f3f}\x{f71}-\x{f87}\x{f90}-\x{fd1}\x{102c}-\x{1039}\x{104a}-'.
35 '\x{104f}\x{1056}-\x{1059}\x{10fb}\x{10fc}\x{135f}-\x{137c}\x{1390}-\x{1399}'.
36 '\x{166d}\x{166e}\x{1680}\x{169b}\x{169c}\x{16eb}-\x{16f0}\x{1712}-\x{1714}'.
37 '\x{1732}-\x{1736}\x{1752}\x{1753}\x{1772}\x{1773}\x{17b4}-\x{17db}\x{17dd}'.
38 '\x{17f0}-\x{180e}\x{1843}\x{18a9}\x{1920}-\x{1945}\x{19b0}-\x{19c0}\x{19c8}'.
39 '\x{19c9}\x{19de}-\x{19ff}\x{1a17}-\x{1a1f}\x{1d2c}-\x{1d61}\x{1d78}\x{1d9b}-'.
40 '\x{1dc3}\x{1fbd}\x{1fbf}-\x{1fc1}\x{1fcd}-\x{1fcf}\x{1fdd}-\x{1fdf}\x{1fed}-'.
41 '\x{1fef}\x{1ffd}-\x{2070}\x{2074}-\x{207e}\x{2080}-\x{2101}\x{2103}-\x{2106}'.
42 '\x{2108}\x{2109}\x{2114}\x{2116}-\x{2118}\x{211e}-\x{2123}\x{2125}\x{2127}'.
43 '\x{2129}\x{212e}\x{2132}\x{213a}\x{213b}\x{2140}-\x{2144}\x{214a}-\x{2b13}'.
44 '\x{2ce5}-\x{2cff}\x{2d6f}\x{2e00}-\x{3005}\x{3007}-\x{303b}\x{303d}-\x{303f}'.
45 '\x{3099}-\x{309e}\x{30a0}\x{30fb}-\x{30fe}\x{3190}-\x{319f}\x{31c0}-\x{31cf}'.
46 '\x{3200}-\x{33ff}\x{4dc0}-\x{4dff}\x{a015}\x{a490}-\x{a716}\x{a802}\x{a806}'.
47 '\x{a80b}\x{a823}-\x{a82b}\x{d800}-\x{f8ff}\x{fb1e}\x{fb29}\x{fd3e}\x{fd3f}'.
48 '\x{fdfc}-\x{fe6b}\x{feff}-\x{ff0f}\x{ff1a}-\x{ff20}\x{ff3b}-\x{ff40}\x{ff5b}-'.
49 '\x{ff65}\x{ff70}\x{ff9e}\x{ff9f}\x{ffe0}-\x{fffd}');
50
51 /**
52 * Matches all 'N' Unicode character classes (numbers)
53 */
54 define('PREG_CLASS_NUMBERS',
55 '\x{30}-\x{39}\x{b2}\x{b3}\x{b9}\x{bc}-\x{be}\x{660}-\x{669}\x{6f0}-\x{6f9}'.
56 '\x{966}-\x{96f}\x{9e6}-\x{9ef}\x{9f4}-\x{9f9}\x{a66}-\x{a6f}\x{ae6}-\x{aef}'.
57 '\x{b66}-\x{b6f}\x{be7}-\x{bf2}\x{c66}-\x{c6f}\x{ce6}-\x{cef}\x{d66}-\x{d6f}'.
58 '\x{e50}-\x{e59}\x{ed0}-\x{ed9}\x{f20}-\x{f33}\x{1040}-\x{1049}\x{1369}-'.
59 '\x{137c}\x{16ee}-\x{16f0}\x{17e0}-\x{17e9}\x{17f0}-\x{17f9}\x{1810}-\x{1819}'.
60 '\x{1946}-\x{194f}\x{2070}\x{2074}-\x{2079}\x{2080}-\x{2089}\x{2153}-\x{2183}'.
61 '\x{2460}-\x{249b}\x{24ea}-\x{24ff}\x{2776}-\x{2793}\x{3007}\x{3021}-\x{3029}'.
62 '\x{3038}-\x{303a}\x{3192}-\x{3195}\x{3220}-\x{3229}\x{3251}-\x{325f}\x{3280}-'.
63 '\x{3289}\x{32b1}-\x{32bf}\x{ff10}-\x{ff19}');
64
65 /**
66 * Matches all 'P' Unicode character classes (punctuation)
67 */
68 define('PREG_CLASS_PUNCTUATION',
69 '\x{21}-\x{23}\x{25}-\x{2a}\x{2c}-\x{2f}\x{3a}\x{3b}\x{3f}\x{40}\x{5b}-\x{5d}'.
70 '\x{5f}\x{7b}\x{7d}\x{a1}\x{ab}\x{b7}\x{bb}\x{bf}\x{37e}\x{387}\x{55a}-\x{55f}'.
71 '\x{589}\x{58a}\x{5be}\x{5c0}\x{5c3}\x{5f3}\x{5f4}\x{60c}\x{60d}\x{61b}\x{61f}'.
72 '\x{66a}-\x{66d}\x{6d4}\x{700}-\x{70d}\x{964}\x{965}\x{970}\x{df4}\x{e4f}'.
73 '\x{e5a}\x{e5b}\x{f04}-\x{f12}\x{f3a}-\x{f3d}\x{f85}\x{104a}-\x{104f}\x{10fb}'.
74 '\x{1361}-\x{1368}\x{166d}\x{166e}\x{169b}\x{169c}\x{16eb}-\x{16ed}\x{1735}'.
75 '\x{1736}\x{17d4}-\x{17d6}\x{17d8}-\x{17da}\x{1800}-\x{180a}\x{1944}\x{1945}'.
76 '\x{2010}-\x{2027}\x{2030}-\x{2043}\x{2045}-\x{2051}\x{2053}\x{2054}\x{2057}'.
77 '\x{207d}\x{207e}\x{208d}\x{208e}\x{2329}\x{232a}\x{23b4}-\x{23b6}\x{2768}-'.
78 '\x{2775}\x{27e6}-\x{27eb}\x{2983}-\x{2998}\x{29d8}-\x{29db}\x{29fc}\x{29fd}'.
79 '\x{3001}-\x{3003}\x{3008}-\x{3011}\x{3014}-\x{301f}\x{3030}\x{303d}\x{30a0}'.
80 '\x{30fb}\x{fd3e}\x{fd3f}\x{fe30}-\x{fe52}\x{fe54}-\x{fe61}\x{fe63}\x{fe68}'.
81 '\x{fe6a}\x{fe6b}\x{ff01}-\x{ff03}\x{ff05}-\x{ff0a}\x{ff0c}-\x{ff0f}\x{ff1a}'.
82 '\x{ff1b}\x{ff1f}\x{ff20}\x{ff3b}-\x{ff3d}\x{ff3f}\x{ff5b}\x{ff5d}\x{ff5f}-'.
83 '\x{ff65}');
84
85 /**
86 * Matches all CJK characters that are candidates for auto-splitting
87 * (Chinese, Japanese, Korean).
88 * Contains kana and BMP ideographs.
89 */
90 define('PREG_CLASS_CJK', '\x{3041}-\x{30ff}\x{31f0}-\x{31ff}\x{3400}-\x{4db5}'.
91 '\x{4e00}-\x{9fbb}\x{f900}-\x{fad9}');
92
93 /**
94 * Implementation of hook_help().
95 */
96 function search_help($path, $arg) {
97 switch ($path) {
98 case 'admin/help#search':
99 $output = '<p>'. t('The search module adds the ability to search for content by keywords. Search is often the only practical way to find content on a large site, and is useful for finding both users and posts.') .'</p>';
100 $output .= '<p>'. t('To provide keyword searching, the search engine maintains an index of words found in your site\'s content. To build and maintain this index, a correctly configured <a href="@cron">cron maintenance task</a> is required. Indexing behavior can be adjusted using the <a href="@searchsettings">search settings page</a>; for example, the <em>Number of items to index per cron run</em> sets the maximum number of items indexed in each pass of a <a href="@cron">cron maintenance task</a>. If necessary, reduce this number to prevent timeouts and memory errors when indexing.', array('@cron' => url('admin/reports/status'), '@searchsettings' => url('admin/settings/search'))) .'</p>';
101 $output .= '<p>'. t('For more information, see the online handbook entry for <a href="@search">Search module</a>.', array('@search' => 'http://drupal.org/handbook/modules/search/')) .'</p>';
102 return $output;
103 case 'admin/settings/search':
104 return '<p>'. t('The search engine maintains an index of words found in your site\'s content. To build and maintain this index, a correctly configured <a href="@cron">cron maintenance task</a> is required. Indexing behavior can be adjusted using the settings below.', array('@cron' => url('admin/reports/status'))) .'</p>';
105 case 'search#noresults':
106 return t('<ul>
107 <li>Check if your spelling is correct.</li>
108 <li>Remove quotes around phrases to match each word individually: <em>"blue smurf"</em> will match less than <em>blue smurf</em>.</li>
109 <li>Consider loosening your query with <em>OR</em>: <em>blue smurf</em> will match less than <em>blue OR smurf</em>.</li>
110 </ul>');
111 }
112 }
113
114 /**
115 * Implementation of hook_theme()
116 */
117 function search_theme() {
118 return array(
119 'search_theme_form' => array(
120 'arguments' => array('form' => NULL),
121 'template' => 'search-theme-form',
122 ),
123 'search_block_form' => array(
124 'arguments' => array('form' => NULL),
125 'template' => 'search-block-form',
126 ),
127 'search_result' => array(
128 'arguments' => array('result' => NULL, 'type' => NULL),
129 'file' => 'search.pages.inc',
130 'template' => 'search-result',
131 ),
132 'search_results' => array(
133 'arguments' => array('results' => NULL, 'type' => NULL),
134 'file' => 'search.pages.inc',
135 'template' => 'search-results',
136 ),
137 );
138 }
139
140 /**
141 * Implementation of hook_perm().
142 */
143 function search_perm() {
144 return array('search content', 'use advanced search', 'administer search');
145 }
146
147 /**
148 * Implementation of hook_block().
149 */
150 function search_block($op = 'list', $delta = 0) {
151 if ($op == 'list') {
152 $blocks[0]['info'] = t('Search form');
153 // Not worth caching.
154 $blocks[0]['cache'] = BLOCK_NO_CACHE;
155 return $blocks;
156 }
157 else if ($op == 'view' && user_access('search content')) {
158 $block['content'] = drupal_get_form('search_block_form');
159 $block['subject'] = t('Search');
160 return $block;
161 }
162 }
163
164 /**
165 * Implementation of hook_menu().
166 */
167 function search_menu() {
168 $items['search'] = array(
169 'title' => 'Search',
170 'page callback' => 'search_view',
171 'access arguments' => array('search content'),
172 'type' => MENU_SUGGESTED_ITEM,
173 'file' => 'search.pages.inc',
174 );
175 $items['admin/settings/search'] = array(
176 'title' => 'Search settings',
177 'description' => 'Configure relevance settings for search and other indexing options',
178 'page callback' => 'drupal_get_form',
179 'page arguments' => array('search_admin_settings'),
180 'access arguments' => array('administer search'),
181 'type' => MENU_NORMAL_ITEM,
182 'file' => 'search.admin.inc',
183 );
184 $items['admin/settings/search/wipe'] = array(
185 'title' => 'Clear index',
186 'page callback' => 'drupal_get_form',
187 'page arguments' => array('search_wipe_confirm'),
188 'access arguments' => array('administer search'),
189 'type' => MENU_CALLBACK,
190 'file' => 'search.admin.inc',
191 );
192 $items['admin/reports/search'] = array(
193 'title' => 'Top search phrases',
194 'description' => 'View most popular search phrases.',
195 'page callback' => 'dblog_top',
196 'page arguments' => array('search'),
197 'file' => 'dblog.admin.inc',
198 'file path' => drupal_get_path('module', 'dblog'),
199 );
200
201 foreach (module_implements('search') as $name) {
202 $items['search/'. $name .'/%menu_tail'] = array(
203 'title callback' => 'module_invoke',
204 'title arguments' => array($name, 'search', 'name', TRUE),
205 'page callback' => 'search_view',
206 'page arguments' => array($name),
207 'access callback' => '_search_menu',
208 'access arguments' => array($name),
209 'type' => MENU_LOCAL_TASK,
210 'parent' => 'search',
211 'file' => 'search.pages.inc',
212 );
213 }
214 return $items;
215 }
216
217 function _search_menu($name) {
218 return user_access('search content') && module_invoke($name, 'search', 'name');
219 }
220
221 /**
222 * Wipes a part of or the entire search index.
223 *
224 * @param $sid
225 * (optional) The SID of the item to wipe. If specified, $type must be passed
226 * too.
227 * @param $type
228 * (optional) The type of item to wipe.
229 */
230 function search_wipe($sid = NULL, $type = NULL, $reindex = FALSE) {
231 if ($type == NULL && $sid == NULL) {
232 module_invoke_all('search', 'reset');
233 }
234 else {
235 db_query("DELETE FROM {search_dataset} WHERE sid = %d AND type = '%s'", $sid, $type);
236 db_query("DELETE FROM {search_index} WHERE sid = %d AND type = '%s'", $sid, $type);
237 // Don't remove links if re-indexing.
238 if (!$reindex) {
239 db_query("DELETE FROM {search_node_links} WHERE sid = %d AND type = '%s'", $sid, $type);
240 }
241 }
242 }
243
244 /**
245 * Marks a word as dirty (or retrieves the list of dirty words). This is used
246 * during indexing (cron). Words which are dirty have outdated total counts in
247 * the search_total table, and need to be recounted.
248 */
249 function search_dirty($word = NULL) {
250 static $dirty = array();
251 if ($word !== NULL) {
252 $dirty[$word] = TRUE;
253 }
254 else {
255 return $dirty;
256 }
257 }
258
259 /**
260 * Implementation of hook_cron().
261 *
262 * Fires hook_update_index() in all modules and cleans up dirty words (see
263 * search_dirty).
264 */
265 function search_cron() {
266 // We register a shutdown function to ensure that search_total is always up
267 // to date.
268 register_shutdown_function('search_update_totals');
269
270 // Update word index
271 foreach (module_list() as $module) {
272 module_invoke($module, 'update_index');
273 }
274 }
275
276 /**
277 * This function is called on shutdown to ensure that search_total is always
278 * up to date (even if cron times out or otherwise fails).
279 */
280 function search_update_totals() {
281 // Update word IDF (Inverse Document Frequency) counts for new/changed words
282 foreach (search_dirty() as $word => $dummy) {
283 // Get total count
284 $total = db_result(db_query("SELECT SUM(score) FROM {search_index} WHERE word = '%s'", $word));
285 // Apply Zipf's law to equalize the probability distribution
286 $total = log10(1 + 1/(max(1, $total)));
287 db_query("UPDATE {search_total} SET count = %f WHERE word = '%s'", $total, $word);
288 if (!db_affected_rows()) {
289 db_query("INSERT INTO {search_total} (word, count) VALUES ('%s', %f)", $word, $total);
290 }
291 }
292 // Find words that were deleted from search_index, but are still in
293 // search_total. We use a LEFT JOIN between the two tables and keep only the
294 // rows which fail to join.
295 $result = db_query("SELECT t.word AS realword, i.word FROM {search_total} t LEFT JOIN {search_index} i ON t.word = i.word WHERE i.word IS NULL");
296 while ($word = db_fetch_object($result)) {
297 db_query("DELETE FROM {search_total} WHERE word = '%s'", $word->realword);
298 }
299 }
300
301 /**
302 * Simplifies a string according to indexing rules.
303 */
304 function search_simplify($text) {
305 // Decode entities to UTF-8
306 $text = decode_entities($text);
307
308 // Lowercase
309 $text = drupal_strtolower($text);
310
311 // Call an external processor for word handling.
312 search_invoke_preprocess($text);
313
314 // Simple CJK handling
315 if (variable_get('overlap_cjk', TRUE)) {
316 $text = preg_replace_callback('/['. PREG_CLASS_CJK .']+/u', 'search_expand_cjk', $text);
317 }
318
319 // To improve searching for numerical data such as dates, IP addresses
320 // or version numbers, we consider a group of numerical characters
321 // separated only by punctuation characters to be one piece.
322 // This also means that searching for e.g. '20/03/1984' also returns
323 // results with '20-03-1984' in them.
324 // Readable regexp: ([number]+)[punctuation]+(?=[number])
325 $text = preg_replace('/(['. PREG_CLASS_NUMBERS .']+)['. PREG_CLASS_PUNCTUATION .']+(?=['. PREG_CLASS_NUMBERS .'])/u', '\1', $text);
326
327 // The dot, underscore and dash are simply removed. This allows meaningful
328 // search behavior with acronyms and URLs.
329 $text = preg_replace('/[._-]+/', '', $text);
330
331 // With the exception of the rules above, we consider all punctuation,
332 // marks, spacers, etc, to be a word boundary.
333 $text = preg_replace('/['. PREG_CLASS_SEARCH_EXCLUDE .']+/u', ' ', $text);
334
335 return $text;
336 }
337
338 /**
339 * Basic CJK tokenizer. Simply splits a string into consecutive, overlapping
340 * sequences of characters ('minimum_word_size' long).
341 */
342 function search_expand_cjk($matches) {
343 $min = variable_get('minimum_word_size', 3);
344 $str = $matches[0];
345 $l = drupal_strlen($str);
346 // Passthrough short words
347 if ($l <= $min) {
348 return ' '. $str .' ';
349 }
350 $tokens = ' ';
351 // FIFO queue of characters
352 $chars = array();
353 // Begin loop
354 for ($i = 0; $i < $l; ++$i) {
355 // Grab next character
356 $current = drupal_substr($str, 0, 1);
357 $str = substr($str, strlen($current));
358 $chars[] = $current;
359 if ($i >= $min - 1) {
360 $tokens .= implode('', $chars) .' ';
361 array_shift($chars);
362 }
363 }
364 return $tokens;
365 }
366
367 /**
368 * Splits a string into tokens for indexing.
369 */
370 function search_index_split($text) {
371 static $last = NULL;
372 static $lastsplit = NULL;
373
374 if ($last == $text) {
375 return $lastsplit;
376 }
377 // Process words
378 $text = search_simplify($text);
379 $words = explode(' ', $text);
380 array_walk($words, '_search_index_truncate');
381
382 // Save last keyword result
383 $last = $text;
384 $lastsplit = $words;
385
386 return $words;
387 }
388
389 /**
390 * Helper function for array_walk in search_index_split.
391 */
392 function _search_index_truncate(&$text) {
393 $text = truncate_utf8($text, 50);
394 }
395
396 /**
397 * Invokes hook_search_preprocess() in modules.
398 */
399 function search_invoke_preprocess(&$text) {
400 foreach (module_implements('search_preprocess') as $module) {
401 $text = module_invoke($module, 'search_preprocess', $text);
402 }
403 }
404
405 /**
406 * Update the full-text search index for a particular item.
407 *
408 * @param $sid
409 * A number identifying this particular item (e.g. node id).
410 *
411 * @param $type
412 * A string defining this type of item (e.g. 'node')
413 *
414 * @param $text
415 * The content of this item. Must be a piece of HTML text.
416 *
417 * @ingroup search
418 */
419 function search_index($sid, $type, $text) {
420 $minimum_word_size = variable_get('minimum_word_size', 3);
421
422 // Link matching
423 global $base_url;
424 $node_regexp = '@href=[\'"]?(?:'. preg_quote($base_url, '@') .'/|'. preg_quote(base_path(), '@') .')(?:\?q=)?/?((?![a-z]+:)[^\'">]+)[\'">]@i';
425
426 // Multipliers for scores of words inside certain HTML tags.
427 // Note: 'a' must be included for link ranking to work.
428 $tags = array('h1' => 25,
429 'h2' => 18,
430 'h3' => 15,
431 'h4' => 12,
432 'h5' => 9,
433 'h6' => 6,
434 'u' => 3,
435 'b' => 3,
436 'i' => 3,
437 'strong' => 3,
438 'em' => 3,
439 'a' => 10);
440
441 // Strip off all ignored tags to speed up processing, but insert space before/after
442 // them to keep word boundaries.
443 $text = str_replace(array('<', '>'), array(' <', '> '), $text);
444 $text = strip_tags($text, '<'. implode('><', array_keys($tags)) .'>');
445
446 // Split HTML tags from plain text.
447 $split = preg_split('/\s*<([^>]+?)>\s*/', $text, -1, PREG_SPLIT_DELIM_CAPTURE);
448 // Note: PHP ensures the array consists of alternating delimiters and literals
449 // and begins and ends with a literal (inserting $null as required).
450
451 $tag = FALSE; // Odd/even counter. Tag or no tag.
452 $link = FALSE; // State variable for link analyser
453 $score = 1; // Starting score per word
454 $accum = ' '; // Accumulator for cleaned up data
455 $tagstack = array(); // Stack with open tags
456 $tagwords = 0; // Counter for consecutive words
457 $focus = 1; // Focus state
458
459 $results = array(0 => array()); // Accumulator for words for index
460
461 foreach ($split as $value) {
462 if ($tag) {
463 // Increase or decrease score per word based on tag
464 list($tagname) = explode(' ', $value, 2);
465 $tagname = drupal_strtolower($tagname);
466 // Closing or opening tag?
467 if ($tagname[0] == '/') {
468 $tagname = substr($tagname, 1);
469 // If we encounter unexpected tags, reset score to avoid incorrect boosting.
470 if (!count($tagstack) || $tagstack[0] != $tagname) {
471 $tagstack = array();
472 $score = 1;
473 }
474 else {
475 // Remove from tag stack and decrement score
476 $score = max(1, $score - $tags[array_shift($tagstack)]);
477 }
478 if ($tagname == 'a') {
479 $link = FALSE;
480 }
481 }
482 else {
483 if (isset($tagstack[0]) && $tagstack[0] == $tagname) {
484 // None of the tags we look for make sense when nested identically.
485 // If they are, it's probably broken HTML.
486 $tagstack = array();
487 $score = 1;
488 }
489 else {
490 // Add to open tag stack and increment score
491 array_unshift($tagstack, $tagname);
492 $score += $tags[$tagname];
493 }
494 if ($tagname == 'a') {
495 // Check if link points to a node on this site
496 if (preg_match($node_regexp, $value, $match)) {
497 $path = drupal_get_normal_path($match[1]);
498 if (preg_match('!(?:node|book)/(?:view/)?([0-9]+)!i', $path, $match)) {
499 $linknid = $match[1];
500 if ($linknid > 0) {
501 // Note: ignore links to uncachable nodes to avoid redirect bugs.
502 $node = db_fetch_object(db_query('SELECT n.title, n.nid, n.vid, r.format FROM {node} n INNER JOIN {node_revisions} r ON n.vid = r.vid WHERE n.nid = %d', $linknid));
503 if (filter_format_allowcache($node->format)) {
504 $link = TRUE;
505 $linktitle = $node->title;
506 }
507 }
508 }
509 }
510 }
511 }
512 // A tag change occurred, reset counter.
513 $tagwords = 0;
514 }
515 else {
516 // Note: use of PREG_SPLIT_DELIM_CAPTURE above will introduce empty values
517 if ($value != '') {
518 if ($link) {
519 // Check to see if the node link text is its URL. If so, we use the target node title instead.
520 if (preg_match('!^https?://!i', $value)) {
521 $value = $linktitle;
522 }
523 }
524 $words = search_index_split($value);
525 foreach ($words as $word) {
526 // Add word to accumulator
527 $accum .= $word .' ';
528 $num = is_numeric($word);
529 // Check wordlength
530 if ($num || drupal_strlen($word) >= $minimum_word_size) {
531 // Normalize numbers
532 if ($num) {
533 $word = (int)ltrim($word, '-0');
534 }
535
536 // Links score mainly for the target.
537 if ($link) {
538 if (!isset($results[$linknid])) {
539 $results[$linknid] = array();
540 }
541 $results[$linknid][] = $word;
542 // Reduce score of the link caption in the source.
543 $focus *= 0.2;
544 }
545 // Fall-through
546 if (!isset($results[0][$word])) {
547 $results[0][$word] = 0;
548 }
549 $results[0][$word] += $score * $focus;
550
551 // Focus is a decaying value in terms of the amount of unique words up to this point.
552 // From 100 words and more, it decays, to e.g. 0.5 at 500 words and 0.3 at 1000 words.
553 $focus = min(1, .01 + 3.5 / (2 + count($results[0]) * .015));
554 }
555 $tagwords++;
556 // Too many words inside a single tag probably mean a tag was accidentally left open.
557 if (count($tagstack) && $tagwords >= 15) {
558 $tagstack = array();
559 $score = 1;
560 }
561 }
562 }
563 }
564 $tag = !$tag;
565 }
566
567 search_wipe($sid, $type, TRUE);
568
569 // Insert cleaned up data into dataset
570 db_query("INSERT INTO {search_dataset} (sid, type, data, reindex) VALUES (%d, '%s', '%s', %d)", $sid, $type, $accum, 0);
571
572 // Insert results into search index
573 foreach ($results[0] as $word => $score) {
574 // The database will collate similar words (accented and non-accented forms, etc.),
575 // and the score is additive, so first add and then insert.
576 db_query("UPDATE {search_index} SET score = score + %d WHERE word = '%s' AND sid = '%d' AND type = '%s'", $score, $word, $sid, $type);
577 if (!db_affected_rows()) {
578 db_query("INSERT INTO {search_index} (word, sid, type, score) VALUES ('%s', %d, '%s', %f)", $word, $sid, $type, $score);
579 }
580 search_dirty($word);
581 }
582 unset($results[0]);
583
584 // Get all previous links from this item.
585 $result = db_query("SELECT nid, caption FROM {search_node_links} WHERE sid = %d AND type = '%s'", $sid, $type);
586 $links = array();
587 while ($link = db_fetch_object($result)) {
588 $links[$link->nid] = $link->caption;
589 }
590
591 // Now store links to nodes.
592 foreach ($results as $nid => $words) {
593 $caption = implode(' ', $words);
594 if (isset($links[$nid])) {
595 if ($links[$nid] != $caption) {
596 // Update the existing link and mark the node for reindexing.
597 db_query("UPDATE {search_node_links} SET caption = '%s' WHERE sid = %d AND type = '%s' AND nid = %d", $caption, $sid, $type, $nid);
598 search_touch_node($nid);
599 }
600 // Unset the link to mark it as processed.
601 unset($links[$nid]);
602 }
603 else {
604 // Insert the existing link and mark the node for reindexing.
605 db_query("INSERT INTO {search_node_links} (caption, sid, type, nid) VALUES ('%s', %d, '%s', %d)", $caption, $sid, $type, $nid);
606 search_touch_node($nid);
607 }
608 }
609 // Any left-over links in $links no longer exist. Delete them and mark the nodes for reindexing.
610 foreach ($links as $nid => $caption) {
611 db_query("DELETE FROM {search_node_links} WHERE sid = %d AND type = '%s' AND nid = %d", $sid, $type, $nid);
612 search_touch_node($nid);
613 }
614 }
615
616 /**
617 * Change a node's changed timestamp to 'now' to force reindexing.
618 *
619 * @param $nid
620 * The nid of the node that needs reindexing.
621 */
622 function search_touch_node($nid) {
623 db_query("UPDATE {search_dataset} SET reindex = %d WHERE sid = %d AND type = 'node'", time(), $nid);
624 }
625
626 /**
627 * Implementation of hook_nodeapi().
628 */
629 function search_nodeapi(&$node, $op, $teaser = NULL, $page = NULL) {
630 switch ($op) {
631 // Transplant links to a node into the target node.
632 case 'update index':
633 $result = db_query("SELECT caption FROM {search_node_links} WHERE nid = %d", $node->nid);
634 $output = array();
635 while ($link = db_fetch_object($result)) {
636 $output[] = $link->caption;
637 }
638 return '<a>('. implode(', ', $output) .')</a>';
639 // Reindex the node when it is updated. The node is automatically indexed
640 // when it is added, simply by being added to the node table.
641 case 'update':
642 search_touch_node($node->nid);
643 break;
644 }
645 }
646
647 /**
648 * Implementation of hook_comment().
649 */
650 function search_comment($a1, $op) {
651 switch ($op) {
652 // Reindex the node when comments are added or changed
653 case 'insert':
654 case 'update':
655 case 'delete':
656 case 'publish':
657 case 'unpublish':
658 search_touch_node(is_array($a1) ? $a1['nid'] : $a1->nid);
659 break;
660 }
661 }
662
663 /**
664 * Extract a module-specific search option from a search query. e.g. 'type:book'
665 */
666 function search_query_extract($keys, $option) {
667 if (preg_match('/(^| )'. $option .':([^ ]*)( |$)/i', $keys, $matches)) {
668 return $matches[2];
669 }
670 }
671
672 /**
673 * Return a query with the given module-specific search option inserted in.
674 * e.g. 'type:book'.
675 */
676 function search_query_insert($keys, $option, $value = '') {
677 if (search_query_extract($keys, $option)) {
678 $keys = trim(preg_replace('/(^| )'. $option .':[^ ]*/i', '', $keys));
679 }
680 if ($value != '') {
681 $keys .= ' '. $option .':'. $value;
682 }
683 return $keys;
684 }
685
686 /**
687 * Parse a search query into SQL conditions.
688 *
689 * We build two queries that matches the dataset bodies. @See do_search for
690 * more about these.
691 *
692 * @param $text
693 * The search keys.
694 * @return
695 * A list of six elements.
696 * * A series of statements AND'd together which will be used to provide all
697 * possible matches.
698 * * Arguments for this query part.
699 * * A series of exact word matches OR'd together.
700 * * Arguments for this query part.
701 * * A bool indicating whether this is a simple query or not. Negative
702 * terms, presence of both AND / OR make this FALSE.
703 * * A bool indicating the presence of a lowercase or. Maybe the user
704 * wanted to use OR.
705 */
706 function search_parse_query($text) {
707 $keys = array('positive' => array(), 'negative' => array());
708
709 // Tokenize query string
710 preg_match_all('/ (-?)("[^"]+"|[^" ]+)/i', ' '. $text, $matches, PREG_SET_ORDER);
711
712 if (count($matches) < 1) {
713 return NULL;
714 }
715
716 // Classify tokens
717 $or = FALSE;
718 $warning = '';
719 $simple = TRUE;
720 foreach ($matches as $match) {
721 $phrase = FALSE;
722 // Strip off phrase quotes
723 if ($match[2]{0} == '"') {
724 $match[2] = substr($match[2], 1, -1);
725 $phrase = TRUE;
726 $simple = FALSE;
727 }
728 // Simplify keyword according to indexing rules and external preprocessors
729 $words = search_simplify($match[2]);
730 // Re-explode in case simplification added more words, except when matching a phrase
731 $words = $phrase ? array($words) : preg_split('/ /', $words, -1, PREG_SPLIT_NO_EMPTY);
732 // Negative matches
733 if ($match[1] == '-') {
734 $keys['negative'] = array_merge($keys['negative'], $words);
735 }
736 // OR operator: instead of a single keyword, we store an array of all
737 // OR'd keywords.
738 elseif ($match[2] == 'OR' && count($keys['positive'])) {
739 $last = array_pop($keys['positive']);
740 // Starting a new OR?
741 if (!is_array($last)) {
742 $last = array($last);
743 }
744 $keys['positive'][] = $last;
745 $or = TRUE;
746 continue;
747 }
748 // AND operator: implied, so just ignore it
749 elseif ($match[2] == 'AND' || $match[2] == 'and') {
750 $warning = $match[2];
751 continue;
752 }
753
754 // Plain keyword
755 else {
756 if ($match[2] == 'or') {
757 $warning = $match[2];
758 }
759 if ($or) {
760 // Add to last element (which is an array)
761 $keys['positive'][count($keys['positive']) - 1] = array_merge($keys['positive'][count($keys['positive']) - 1], $words);
762 }
763 else {
764 $keys['positive'] = array_merge($keys['positive'], $words);
765 }
766 }
767 $or = FALSE;
768 }
769
770 // Convert keywords into SQL statements.
771 $query = array();
772 $query2 = array();
773 $arguments = array();
774 $arguments2 = array();
775 $matches = 0;
776 $simple_and = FALSE;
777 $simple_or = FALSE;
778 // Positive matches
779 foreach ($keys['positive'] as $key) {
780 // Group of ORed terms
781 if (is_array($key) && count($key)) {
782 $simple_or = TRUE;
783 $queryor = array();
784 $any = FALSE;
785 foreach ($key as $or) {
786 list($q, $num_new_scores) = _search_parse_query($or, $arguments2);
787 $any |= $num_new_scores;
788 if ($q) {
789 $queryor[] = $q;
790 $arguments[] = $or;
791 }
792 }
793 if (count($queryor)) {
794 $query[] = '('. implode(' OR ', $queryor) .')';
795 // A group of OR keywords only needs to match once
796 $matches += ($any > 0);
797 }
798 }
799 // Single ANDed term
800 else {
801 $simple_and = TRUE;
802 list($q, $num_new_scores, $num_valid_words) = _search_parse_query($key, $arguments2);
803 if ($q) {
804 $query[] = $q;
805 $arguments[] = $key;
806 if (!$num_valid_words) {
807 $simple = FALSE;
808 }
809 // Each AND keyword needs to match at least once
810 $matches += $num_new_scores;
811 }
812 }
813 }
814 if ($simple_and && $simple_or) {
815 $simple = FALSE;
816 }
817 // Negative matches
818 foreach ($keys['negative'] as $key) {
819 list($q) = _search_parse_query($key, $arguments2, TRUE);
820 if ($q) {
821 $query[] = $q;
822 $arguments[] = $key;
823 $simple = FALSE;
824 }
825 }
826 $query = implode(' AND ', $query);
827
828 // Build word-index conditions for the first pass
829 $query2 = substr(str_repeat("i.word = '%s' OR ", count($arguments2)), 0, -4);
830
831 return array($query, $arguments, $query2, $arguments2, $matches, $simple, $warning);
832 }
833
834 /**
835 * Helper function for search_parse_query();
836 */
837 function _search_parse_query(&$word, &$scores, $not = FALSE) {
838 $num_new_scores = 0;
839 $num_valid_words = 0;
840 // Determine the scorewords of this word/phrase
841 if (!$not) {
842 $split = explode(' ', $word);
843 foreach ($split as $s) {
844 $num = is_numeric($s);
845 if ($num || drupal_strlen($s) >= variable_get('minimum_word_size', 3)) {
846 $s = $num ? ((int)ltrim($s, '-0')) : $s;
847 if (!isset($scores[$s])) {
848 $scores[$s] = $s;
849 $num_new_scores++;
850 }
851 $num_valid_words++;
852 }
853 }
854 }
855 // Return matching snippet and number of added words
856 return array("d.data ". ($not ? 'NOT ' : '') ."LIKE '%% %s %%'", $num_new_scores, $num_valid_words);
857 }
858
859 /**
860 * Do a query on the full-text search index for a word or words.
861 *
862 * This function is normally only called by each module that support the
863 * indexed search (and thus, implements hook_update_index()).
864 *
865 * Results are retrieved in two logical passes. However, the two passes are
866 * joined together into a single query. And in the case of most simple
867 * queries the second pass is not even used.
868 *
869 * The first pass selects a set of all possible matches, which has the benefit
870 * of also providing the exact result set for simple "AND" or "OR" searches.
871 *
872 * The second portion of the query further refines this set by verifying
873 * advanced text conditions (such negative or phrase matches)
874 *
875 * @param $keywords
876 * A search string as entered by the user.
877 *
878 * @param $type
879 * A string identifying the calling module.
880 *
881 * @param $join1
882 * (optional) Inserted into the JOIN part of the first SQL query.
883 * For example "INNER JOIN {node} n ON n.nid = i.sid".
884 *
885 * @param $where1
886 * (optional) Inserted into the WHERE part of the first SQL query.
887 * For example "(n.status > %d)".
888 *
889 * @param $arguments1
890 * (optional) Extra SQL arguments belonging to the first query.
891 *
892 * @param $columns2
893 * (optional) Inserted into the SELECT pat of the second query. Must contain
894 * a column selected as 'score'.
895 * defaults to 'i.relevance AS score'
896 *
897 * @param $join2
898 * (optional) Inserted into the JOIN par of the second SQL query.
899 * For example "INNER JOIN {node_comment_statistics} n ON n.nid = i.sid"
900 *
901 * @param $arguments2
902 * (optional) Extra SQL arguments belonging to the second query parameter.
903 *
904 * @param $sort_parameters
905 * (optional) SQL arguments for sorting the final results.
906 * Default: 'ORDER BY score DESC'
907 *
908 * @return
909 * An array of SIDs for the search results.
910 *
911 * @ingroup search
912 */
913 function do_search($keywords, $type, $join1 = '', $where1 = '1', $arguments1 = array(), $columns2 = 'i.relevance AS score', $join2 = '', $arguments2 = array(), $sort_parameters = 'ORDER BY score DESC') {
914 $query = search_parse_query($keywords);
915
916 if ($query[2] == '') {
917 form_set_error('keys', t('You must include at least one positive keyword with @count characters or more.', array('@count' => variable_get('minimum_word_size', 3))));
918 }
919 if ($query[6]) {
920 if ($query[6] == 'or') {
921 drupal_set_message(t('Search for either of the two terms with uppercase <strong>OR</strong>. For example, <strong>cats OR dogs</strong>.'));
922 }
923 }
924 if ($query === NULL || $query[0] == '' || $query[2] == '') {
925 return array();
926 }
927
928 // Build query for keyword normalization.
929 $conditions = "$where1 AND ($query[2]) AND i.type = '%s'";
930 $arguments1 = array_merge($arguments1, $query[3], array($type));
931 $join = "INNER JOIN {search_total} t ON i.word = t.word $join1";
932 if (!$query[5]) {
933 $conditions .= " AND ($query[0])";
934 $arguments1 = array_merge($arguments1, $query[1]);
935 $join .= " INNER JOIN {search_dataset} d ON i.sid = d.sid AND i.type = d.type";
936 }
937
938 // Calculate maximum keyword relevance, to normalize it.
939 $select = "SELECT SUM(i.score * t.count) AS score FROM {search_index} i $join WHERE $conditions GROUP BY i.type, i.sid HAVING COUNT(*) >= %d ORDER BY score DESC";
940 $arguments = array_merge($arguments1, array($query[4]));
941 $normalize = db_result(db_query_range($select, $arguments, 0, 1));
942 if (!$normalize) {
943 return array();
944 }
945 $columns2 = str_replace('i.relevance', '('. (1.0 / $normalize) .' * SUM(i.score * t.count))', $columns2);
946
947 // Build query to retrieve results.
948 $select = "SELECT i.type, i.sid, $columns2 FROM {search_index} i $join $join2 WHERE $conditions GROUP BY i.type, i.sid HAVING COUNT(*) >= %d";
949 $count_select = "SELECT COUNT(*) FROM ($select) n1";
950 $arguments = array_merge($arguments2, $arguments1, array($query[4]));
951
952 // Do actual search query
953 $result = pager_query("$select $sort_parameters", 10, 0, $count_select, $arguments);
954 $results = array();
955 while ($item = db_fetch_object($result)) {
956 $results[] = $item;
957 }
958 return $results;
959 }
960
961 /**
962 * Helper function for grabbing search keys.
963 */
964 function search_get_keys() {
965 static $return;
966 if (!isset($return)) {
967 // Extract keys as remainder of path
968 // Note: support old GET format of searches for existing links.
969 $path = explode('/', $_GET['q'], 3);
970 $keys = empty($_REQUEST['keys']) ? '' : $_REQUEST['keys'];
971 $return = count($path) == 3 ? $path[2] : $keys;
972 }
973 return $return;
974 }
975
976 /**
977 * @defgroup search Search interface
978 * @{
979 * The Drupal search interface manages a global search mechanism.
980 *
981 * Modules may plug into this system to provide searches of different types of
982 * data. Most of the system is handled by search.module, so this must be enabled
983 * for all of the search features to work.
984 *
985 * There are three ways to interact with the search system:
986 * - Specifically for searching nodes, you can implement nodeapi('update index')
987 * and nodeapi('search result'). However, note that the search system already
988 * indexes all visible output of a node, i.e. everything displayed normally
989 * by hook_view() and hook_nodeapi('view'). This is usually sufficient.
990 * You should only use this mechanism if you want additional, non-visible data
991 * to be indexed.
992 * - Implement hook_search(). This will create a search tab for your module on
993 * the /search page with a simple keyword search form. You may optionally
994 * implement hook_search_item() to customize the display of your results.
995 * - Implement hook_update_index(). This allows your module to use Drupal's
996 * HTML indexing mechanism for searching full text efficiently.
997 *
998 * If your module needs to provide a more complicated search form, then you need
999 * to implement it yourself without hook_search(). In that case, you should
1000 * define it as a local task (tab) under the /search page (e.g. /search/mymodule)
1001 * so that users can easily find it.
1002 */
1003
1004 /**
1005 * Render a search form.
1006 *
1007 * @param $action
1008 * Form action. Defaults to "search".
1009 * @param $keys
1010 * The search string entered by the user, containing keywords for the search.
1011 * @param $type
1012 * The type of search to render the node for. Must be the name of module
1013 * which implements hook_search(). Defaults to 'node'.
1014 * @param $prompt
1015 * A piece of text to put before the form (e.g. "Enter your keywords")
1016 * @return
1017 * An HTML string containing the search form.
1018 */
1019 function search_form(&$form_state, $action = '', $keys = '', $type = NULL, $prompt = NULL) {
1020
1021 // Add CSS
1022 drupal_add_css(drupal_get_path('module', 'search') .'/search.css', 'module', 'all', FALSE);
1023
1024 if (!$action) {
1025 $action = url('search/'. $type);
1026 }
1027 if (is_null($prompt)) {
1028 $prompt = t('Enter your keywords');
1029 }
1030
1031 $form = array(
1032 '#action' => $action,
1033 '#attributes' => array('class' => 'search-form'),
1034 );
1035 $form['module'] = array('#type' => 'value', '#value' => $type);
1036 $form['basic'] = array('#type' => 'item', '#title' => $prompt);
1037 $form['basic']['inline'] = array('#prefix' => '<div class="container-inline">', '#suffix' => '</div>');
1038 $form['basic']['inline']['keys'] = array(
1039 '#type' => 'textfield',
1040 '#title' => '',
1041 '#default_value' => $keys,
1042 '#size' => $prompt ? 40 : 20,
1043 '#maxlength' => 255,
1044 );
1045 // processed_keys is used to coordinate keyword passing between other forms
1046 // that hook into the basic search form.
1047 $form['basic']['inline']['processed_keys'] = array('#type' => 'value', '#value' => array());
1048 $form['basic']['inline']['submit'] = array('#type' => 'submit', '#value' => t('Search'));
1049
1050 return $form;
1051 }
1052
1053 /**
1054 * Form builder; Output a search form for the search block and the theme's search box.
1055 *
1056 * @ingroup forms
1057 * @see search_box_form_submit()
1058 * @see theme_search_box_form()
1059 */
1060 function search_box(&$form_state, $form_id) {
1061 $form[$form_id] = array(
1062 '#title' => t('Search this site'),
1063 '#type' => 'textfield',
1064 '#size' => 15,
1065 '#default_value' => '',
1066 '#attributes' => array('title' => t('Enter the terms you wish to search for.')),
1067 );
1068 $form['submit'] = array('#type' => 'submit', '#value' => t('Search'));
1069 $form['#submit'][] = 'search_box_form_submit';
1070 $form['#validate'][] = 'search_box_form_validate';
1071
1072 return $form;
1073 }
1074
1075 /**
1076 * Process a block search form submission.
1077 */
1078 function search_box_form_submit($form, &$form_state) {
1079 $form_id = $form['form_id']['#value'];
1080 $form_state['redirect'] = 'search/node/'. trim($form_state['values'][$form_id]);
1081 }
1082
1083 /**
1084 * Process variables for search-theme-form.tpl.php.
1085 *
1086 * The $variables array contains the following arguments:
1087 * - $form
1088 *
1089 * @see search-theme-form.tpl.php
1090 */
1091 function template_preprocess_search_theme_form(&$variables) {
1092 $variables['search'] = array();
1093 $hidden = array();
1094 // Provide variables named after form keys so themers can print each element independently.
1095 foreach (element_children($variables['form']) as $key) {
1096 $type = $variables['form'][$key]['#type'];
1097 if ($type == 'hidden' || $type == 'token') {
1098 $hidden[] = drupal_render($variables['form'][$key]);
1099 }
1100 else {
1101 $variables['search'][$key] = drupal_render($variables['form'][$key]);
1102 }
1103 }
1104 // Hidden form elements have no value to themers. No need for separation.
1105 $variables['search']['hidden'] = implode($hidden);
1106 // Collect all form elements to make it easier to print the whole form.
1107 $variables['search_form'] = implode($variables['search']);
1108 }
1109
1110 /**
1111 * Process variables for search-block-form.tpl.php.
1112 *
1113 * The $variables array contains the following arguments:
1114 * - $form
1115 *
1116 * @see search-block-form.tpl.php
1117 */
1118 function template_preprocess_search_block_form(&$variables) {
1119 $variables['search'] = array();
1120 $hidden = array();
1121 // Provide variables named after form keys so themers can print each element independently.
1122 foreach (element_children($variables['form']) as $key) {
1123 $type = $variables['form'][$key]['#type'];
1124 if ($type == 'hidden' || $type == 'token') {
1125 $hidden[] = drupal_render($variables['form'][$key]);
1126 }
1127 else {
1128 $variables['search'][$key] = drupal_render($variables['form'][$key]);
1129 }
1130 }
1131 // Hidden form elements have no value to themers. No need for separation.
1132 $variables['search']['hidden'] = implode($hidden);
1133 // Collect all form elements to make it easier to print the whole form.
1134 $variables['search_form'] = implode($variables['search']);
1135 }
1136
1137 /**
1138 * Perform a standard search on the given keys, and return the formatted results.
1139 */
1140 function search_data($keys = NULL, $type = 'node') {
1141
1142 if (isset($keys)) {
1143 if (module_hook($type, 'search')) {
1144 $results = module_invoke($type, 'search', 'search', $keys);
1145 if (isset($results) && is_array($results) && count($results)) {
1146 if (module_hook($type, 'search_page')) {
1147 return module_invoke($type, 'search_page', $results);
1148 }
1149 else {
1150 return theme('search_results', $results, $type);
1151 }
1152 }
1153 }
1154 }
1155 }
1156
1157 /**
1158 * Returns snippets from a piece of text, with certain keywords highlighted.
1159 * Used for formatting search results.
1160 *
1161 * @param $keys
1162 * A string containing a search query.
1163 *
1164 * @param $text
1165 * The text to extract fragments from.
1166 *
1167 * @return
1168 * A string containing HTML for the excerpt.
1169 */
1170 function search_excerpt($keys, $text) {
1171 // We highlight around non-indexable or CJK characters.
1172 $boundary = '(?:(?<=['. PREG_CLASS_SEARCH_EXCLUDE . PREG_CLASS_CJK .'])|(?=['. PREG_CLASS_SEARCH_EXCLUDE . PREG_CLASS_CJK .']))';
1173
1174 // Extract positive keywords and phrases
1175 preg_match_all('/ ("([^"]+)"|(?!OR)([^" ]+))/', ' '. $keys, $matches);
1176 $keys = array_merge($matches[2], $matches[3]);
1177
1178 // Prepare text
1179 $text = ' '. strip_tags(str_replace(array('<', '>'), array(' <', '> '), $text)) .' ';
1180 array_walk($keys, '_search_excerpt_replace');
1181 $workkeys = $keys;
1182
1183 // Extract a fragment per keyword for at most 4 keywords.
1184 // First we collect ranges of text around each keyword, starting/ending
1185 // at spaces.
1186 // If the sum of all fragments is too short, we look for second occurrences.
1187 $ranges = array();
1188 $included = array();
1189 $length = 0;
1190 while ($length < 256 && count($workkeys)) {
1191 foreach ($workkeys as $k => $key) {
1192 if (strlen($key) == 0) {
1193 unset($workkeys[$k]);
1194 unset($keys[$k]);
1195 continue;
1196 }
1197 if ($length >= 256) {
1198 break;
1199 }
1200 // Remember occurrence of key so we can skip over it if more occurrences
1201 // are desired.
1202 if (!isset($included[$key])) {
1203 $included[$key] = 0;
1204 }
1205 // Locate a keyword (position $p), then locate a space in front (position
1206 // $q) and behind it (position $s)
1207 if (preg_match('/'. $boundary . $key . $boundary .'/iu', $text, $match, PREG_OFFSET_CAPTURE, $included[$key])) {
1208 $p = $match[0][1];
1209 if (($q = strpos($text, ' ', max(0, $p - 60))) !== FALSE) {
1210 $end = substr($text, $p, 80);
1211 if (($s = strrpos($end, ' ')) !== FALSE) {
1212 $ranges[$q] = $p + $s;
1213 $length += $p + $s - $q;
1214 $included[$key] = $p + 1;
1215 }
1216 else {
1217 unset($workkeys[$k]);
1218 }
1219 }
1220 else {
1221 unset($workkeys[$k]);
1222 }
1223 }
1224 else {
1225 unset($workkeys[$k]);
1226 }
1227 }
1228 }
1229
1230 // If we didn't find anything, return the beginning.
1231 if (count($ranges) == 0) {
1232 return truncate_utf8($text, 256) .' ...';
1233 }
1234
1235 // Sort the text ranges by starting position.
1236 ksort($ranges);
1237
1238 // Now we collapse overlapping text ranges into one. The sorting makes it O(n).
1239 $newranges = array();
1240 foreach ($ranges as $from2 => $to2) {
1241 if (!isset($from1)) {
1242 $from1 = $from2;
1243 $to1 = $to2;
1244 continue;
1245 }
1246 if ($from2 <= $to1) {
1247 $to1 = max($to1, $to2);
1248 }
1249 else {
1250 $newranges[$from1] = $to1;
1251 $from1 = $from2;
1252 $to1 = $to2;
1253 }
1254 }
1255 $newranges[$from1] = $to1;
1256
1257 // Fetch text
1258 $out = array();
1259 foreach ($newranges as $from => $to) {
1260 $out[] = substr($text, $from, $to - $from);
1261 }
1262 $text = (isset($newranges[0]) ? '' : '... ') . implode(' ... ', $out) .' ...';
1263
1264 // Highlight keywords. Must be done at once to prevent conflicts ('strong' and '<strong>').
1265 $text = preg_replace('/'. $boundary .'('. implode('|', $keys) .')'. $boundary .'/iu', '<strong>\0</strong>', $text);
1266 return $text;
1267 }
1268
1269 /**
1270 * @} End of "defgroup search".
1271 */
1272
1273 /**
1274 * Helper function for array_walk in search_except.
1275 */
1276 function _search_excerpt_replace(&$text) {
1277 $text = preg_quote($text, '/');
1278 }
1279
1280 function search_forms() {
1281 $forms['search_theme_form']= array(
1282 'callback' => 'search_box',
1283 'callback arguments' => array('search_theme_form'),
1284 );
1285 $forms['search_block_form']= array(
1286 'callback' => 'search_box',
1287 'callback arguments' => array('search_block_form'),
1288 );
1289 return $forms;
1290 }