Fixed bug #14279: Some non-XHTML tags in Backend (thanks to Steffen Gebert)
[Packages/TYPO3.CMS.git] / typo3 / sysext / indexed_search / class.indexer.php
1 <?php
2 /***************************************************************
3 * Copyright notice
4 *
5 * (c) 2001-2010 Kasper Skaarhoj (kasperYYYY@typo3.com)
6 * All rights reserved
7 *
8 * This script is part of the TYPO3 project. The TYPO3 project is
9 * free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License as published by
11 * the Free Software Foundation; either version 2 of the License, or
12 * (at your option) any later version.
13 *
14 * The GNU General Public License can be found at
15 * http://www.gnu.org/copyleft/gpl.html.
16 * A copy is found in the textfile GPL.txt and important notices to the license
17 * from the author is found in LICENSE.txt distributed with these scripts.
18 *
19 *
20 * This script is distributed in the hope that it will be useful,
21 * but WITHOUT ANY WARRANTY; without even the implied warranty of
22 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
23 * GNU General Public License for more details.
24 *
25 * This copyright notice MUST APPEAR in all copies of the script!
26 ***************************************************************/
27 /**
28 * This class is a search indexer for TYPO3
29 *
30 * @author Kasper Skaarhoj <kasperYYYY@typo3.com>
31 * Originally Christian Jul Jensen <christian@jul.net> helped as well.
32 */
33 /**
34 * [CLASS/FUNCTION INDEX of SCRIPT]
35 *
36 *
37 *
38 * 141: class tx_indexedsearch_indexer
39 * 207: function hook_indexContent(&$pObj)
40 *
41 * SECTION: Backend API
42 * 308: function backend_initIndexer($id, $type, $sys_language_uid, $MP, $uidRL, $cHash_array=array(), $createCHash=FALSE)
43 * 347: function backend_setFreeIndexUid($freeIndexUid, $freeIndexSetId=0)
44 * 365: function backend_indexAsTYPO3Page($title, $keywords, $description, $content, $charset, $mtime, $crdate=0, $recordUid=0)
45 *
46 * SECTION: Initialization
47 * 416: function init()
48 * 468: function initializeExternalParsers()
49 *
50 * SECTION: Indexing; TYPO3 pages (HTML content)
51 * 509: function indexTypo3PageContent()
52 * 596: function splitHTMLContent($content)
53 * 642: function getHTMLcharset($content)
54 * 657: function convertHTMLToUtf8($content,$charset='')
55 * 685: function embracingTags($string,$tagName,&$tagContent,&$stringAfter,&$paramList)
56 * 712: function typoSearchTags(&$body)
57 * 741: function extractLinks($content)
58 * 812: function extractHyperLinks($string)
59 *
60 * SECTION: Indexing; external URL
61 * 871: function indexExternalUrl($externalUrl)
62 * 902: function getUrlHeaders($url)
63 *
64 * SECTION: Indexing; external files (PDF, DOC, etc)
65 * 948: function indexRegularDocument($file, $force=FALSE, $contentTmpFile='', $altExtension='')
66 * 1054: function readFileContent($ext,$absFile,$cPKey)
67 * 1071: function fileContentParts($ext,$absFile)
68 * 1089: function splitRegularContent($content)
69 *
70 * SECTION: Analysing content, Extracting words
71 * 1122: function charsetEntity2utf8(&$contentArr, $charset)
72 * 1145: function processWordsInArrays($contentArr)
73 * 1170: function procesWordsInArrays($contentArr)
74 * 1180: function bodyDescription($contentArr)
75 * 1202: function indexAnalyze($content)
76 * 1223: function analyzeHeaderinfo(&$retArr,$content,$key,$offset)
77 * 1242: function analyzeBody(&$retArr,$content)
78 * 1262: function metaphone($word,$retRaw=FALSE)
79 *
80 * SECTION: SQL; TYPO3 Pages
81 * 1304: function submitPage()
82 * 1378: function submit_grlist($hash,$phash_x)
83 * 1398: function submit_section($hash,$hash_t3)
84 * 1416: function removeOldIndexedPages($phash)
85 *
86 * SECTION: SQL; External media
87 * 1459: function submitFilePage($hash,$file,$subinfo,$ext,$mtime,$ctime,$size,$content_md5h,$contentParts)
88 * 1525: function submitFile_grlist($hash)
89 * 1539: function submitFile_section($hash)
90 * 1553: function removeOldIndexedFiles($phash)
91 *
92 * SECTION: SQL Helper functions
93 * 1589: function checkMtimeTstamp($mtime,$phash)
94 * 1625: function checkContentHash()
95 * 1642: function checkExternalDocContentHash($hashGr,$content_md5h)
96 * 1656: function is_grlist_set($phash_x)
97 * 1669: function update_grlist($phash,$phash_x)
98 * 1684: function updateTstamp($phash,$mtime=0)
99 * 1699: function updateSetId($phash)
100 * 1714: function updateParsetime($phash,$parsetime)
101 * 1727: function updateRootline()
102 * 1742: function getRootLineFields(&$fieldArr)
103 * 1761: function removeLoginpagesWithContentHash()
104 * 1778: function includeCrawlerClass()
105 *
106 * SECTION: SQL; Submitting words
107 * 1805: function checkWordList($wl)
108 * 1842: function submitWords($wl,$phash)
109 * 1866: function freqMap($freq)
110 *
111 * SECTION: Hashing
112 * 1899: function setT3Hashes()
113 * 1925: function setExtHashes($file,$subinfo=array())
114 * 1949: function md5inthash($str)
115 * 1959: function makeCHash($paramArray)
116 *
117 * SECTION: Internal logging functions
118 * 1991: function log_push($msg,$key)
119 * 2000: function log_pull()
120 * 2011: function log_setTSlogMessage($msg, $errorNum=0)
121 *
122 * SECTION: tslib_fe hooks:
123 * 2036: function fe_headerNoCache(&$params, $ref)
124 *
125 * TOTAL FUNCTIONS: 59
126 * (This index is automatically created/updated by the extension "extdeveval")
127 *
128 */
129 /**
130 * Indexing class for TYPO3 frontend
131 *
132 * @author Kasper Skaarhoj <kasperYYYY@typo3.com>
133 * @package TYPO3
134 * @subpackage tx_indexedsearch
135 */
136 class tx_indexedsearch_indexer {
137
138 // Messages:
139 var $reasons = array(
140 -1 => 'mtime matched the document, so no changes detected and no content updated',
141 -2 => 'The minimum age was not exceeded',
142 1 => "The configured max-age was exceeded for the document and thus it's indexed.",
143 2 => 'The minimum age was exceed and mtime was set and the mtime was different, so the page was indexed.',
144 3 => 'The minimum age was exceed, but mtime was not set, so the page was indexed.',
145 4 => 'Page has never been indexed (is not represented in the index_phash table).'
146 );
147
148 // HTML code blocks to exclude from indexing:
149 var $excludeSections = 'script,style';
150
151 // Supported Extensions for external files:
152 var $external_parsers = array(); // External parser objects, keys are file extension names. Values are objects with certain methods.
153
154 // Fe-group list (pages might be indexed separately for each usergroup combination to support search in access limited pages!)
155 var $defaultGrList = '0,-1';
156
157 // Min/Max times:
158 var $tstamp_maxAge = 0; // If set, this tells a number of seconds that is the maximum age of an indexed document. Regardless of mtime the document will be re-indexed if this limit is exceeded.
159 var $tstamp_minAge = 0; // If set, this tells a minimum limit before a document can be indexed again. This is regardless of mtime.
160 var $maxExternalFiles = 0; // Max number of external files to index.
161
162 var $forceIndexing = FALSE; // If true, indexing is forced despite of hashes etc.
163 var $crawlerActive = FALSE; // Set when crawler is detected (internal)
164
165 // INTERNALS:
166 var $defaultContentArray=array(
167 'title' => '',
168 'description' => '',
169 'keywords' => '',
170 'body' => '',
171 );
172 var $wordcount = 0;
173 var $externalFileCounter = 0;
174
175 var $conf = array(); // Configuration set internally (see init functions for required keys and their meaning)
176 var $indexerConfig = array(); // Indexer configuration, coming from $GLOBALS['TYPO3_CONF_VARS']['EXT']['extConf']['indexed_search']
177 var $hash = array(); // Hash array, contains phash and phash_grouping
178 var $file_phash_arr = array(); // Hash array for files
179 var $contentParts = array(); // Content of TYPO3 page
180 var $content_md5h = '';
181 var $internal_log = array(); // Internal log
182 var $indexExternalUrl_content = '';
183
184 var $cHashParams = array(); // cHashparams array
185
186 var $freqRange = 32000;
187 var $freqMax = 0.1;
188
189 // Objects:
190 /**
191 * Charset class object
192 *
193 * @var t3lib_cs
194 */
195 var $csObj;
196
197 /**
198 * Metaphone object, if any
199 *
200 * @var user_DoubleMetaPhone
201 */
202 var $metaphoneObj;
203
204 /**
205 * Lexer object for word splitting
206 *
207 * @var tx_indexedsearch_lexer
208 */
209 var $lexerObj;
210
211
212
213 /**
214 * Parent Object (TSFE) Initialization
215 *
216 * @param object Parent Object (frontend TSFE object), passed by reference
217 * @return void
218 */
219 function hook_indexContent(&$pObj) {
220
221 // Indexer configuration from Extension Manager interface:
222 $indexerConfig = unserialize($GLOBALS['TYPO3_CONF_VARS']['EXT']['extConf']['indexed_search']);
223
224 // Crawler activation:
225 // Requirements are that the crawler is loaded, a crawler session is running and re-indexing requested as processing instruction:
226 if (t3lib_extMgm::isLoaded('crawler')
227 && $pObj->applicationData['tx_crawler']['running']
228 && in_array('tx_indexedsearch_reindex', $pObj->applicationData['tx_crawler']['parameters']['procInstructions'])) {
229
230 // Setting simple log message:
231 $pObj->applicationData['tx_crawler']['log'][] = 'Forced Re-indexing enabled';
232
233 // Setting variables:
234 $this->crawlerActive = TRUE; // Crawler active flag
235 $this->forceIndexing = TRUE; // Force indexing despite timestamps etc.
236 }
237
238 // Determine if page should be indexed, and if so, configure and initialize indexer
239 if ($pObj->config['config']['index_enable']) {
240 $this->log_push('Index page','');
241
242 if (!$indexerConfig['disableFrontendIndexing'] || $this->crawlerActive) {
243 if (!$pObj->page['no_search']) {
244 if (!$pObj->no_cache) {
245 if (!strcmp($pObj->sys_language_uid,$pObj->sys_language_content)) {
246
247 // Setting up internal configuration from config array:
248 $this->conf = array();
249
250 // Information about page for which the indexing takes place
251 $this->conf['id'] = $pObj->id; // Page id
252 $this->conf['type'] = $pObj->type; // Page type
253 $this->conf['sys_language_uid'] = $pObj->sys_language_uid; // sys_language UID of the language of the indexing.
254 $this->conf['MP'] = $pObj->MP; // MP variable, if any (Mount Points)
255 $this->conf['gr_list'] = $pObj->gr_list; // Group list
256
257 $this->conf['cHash'] = $pObj->cHash; // cHash string for additional parameters
258 $this->conf['cHash_array'] = $pObj->cHash_array; // Array of the additional parameters
259
260 $this->conf['crdate'] = $pObj->page['crdate']; // The creation date of the TYPO3 page
261 $this->conf['page_cache_reg1'] = $pObj->page_cache_reg1; // reg1 of the caching table. Not known what practical use this has.
262
263 // Root line uids
264 $this->conf['rootline_uids'] = array();
265 foreach($pObj->config['rootLine'] as $rlkey => $rldat) {
266 $this->conf['rootline_uids'][$rlkey] = $rldat['uid'];
267 }
268
269 // Content of page:
270 $this->conf['content'] = $pObj->content; // Content string (HTML of TYPO3 page)
271 $this->conf['indexedDocTitle'] = $pObj->convOutputCharset($pObj->indexedDocTitle); // Alternative title for indexing
272 $this->conf['metaCharset'] = $pObj->metaCharset; // Character set of content (will be converted to utf-8 during indexing)
273 $this->conf['mtime'] = $pObj->register['SYS_LASTCHANGED']; // Most recent modification time (seconds) of the content on the page. Used to evaluate whether it should be re-indexed.
274
275 // Configuration of behavior:
276 $this->conf['index_externals'] = $pObj->config['config']['index_externals']; // Whether to index external documents like PDF, DOC etc. (if possible)
277 $this->conf['index_descrLgd'] = $pObj->config['config']['index_descrLgd']; // Length of description text (max 250, default 200)
278 $this->conf['index_metatags'] = isset($pObj->config['config']['index_metatags']) ? $pObj->config['config']['index_metatags'] : true;
279
280 // Set to zero:
281 $this->conf['recordUid'] = 0;
282 $this->conf['freeIndexUid'] = 0;
283 $this->conf['freeIndexSetId'] = 0;
284
285 // Init and start indexing:
286 $this->init();
287 $this->indexTypo3PageContent();
288 } else $this->log_setTSlogMessage('Index page? No, ->sys_language_uid was different from sys_language_content which indicates that the page contains fall-back content and that would be falsely indexed as localized content.');
289 } else $this->log_setTSlogMessage('Index page? No, page was set to "no_cache" and so cannot be indexed.');
290 } else $this->log_setTSlogMessage('Index page? No, The "No Search" flag has been set in the page properties!');
291 } else $this->log_setTSlogMessage('Index page? No, Ordinary Frontend indexing during rendering is disabled.');
292 $this->log_pull();
293 }
294 }
295
296
297
298
299
300
301
302
303 /****************************
304 *
305 * Backend API
306 *
307 ****************************/
308
309 /**
310 * Initializing the "combined ID" of the page (phash) being indexed (or for which external media is attached)
311 *
312 * @param integer The page uid, &id=
313 * @param integer The page type, &type=
314 * @param integer sys_language uid, typically &L=
315 * @param string The MP variable (Mount Points), &MP=
316 * @param array Rootline array of only UIDs.
317 * @param array Array of GET variables to register with this indexing
318 * @param boolean If set, calculates a cHash value from the $cHash_array. Probably you will not do that since such cases are indexed through the frontend and the idea of this interface is to index non-cachable pages from the backend!
319 * @return void
320 */
321 function backend_initIndexer($id, $type, $sys_language_uid, $MP, $uidRL, $cHash_array=array(), $createCHash=FALSE) {
322
323 // Setting up internal configuration from config array:
324 $this->conf = array();
325
326 // Information about page for which the indexing takes place
327 $this->conf['id'] = $id; // Page id (integer)
328 $this->conf['type'] = $type; // Page type (integer)
329 $this->conf['sys_language_uid'] = $sys_language_uid; // sys_language UID of the language of the indexing (integer)
330 $this->conf['MP'] = $MP; // MP variable, if any (Mount Points) (string)
331 $this->conf['gr_list'] = '0,-1'; // Group list (hardcoded for now...)
332
333 // cHash values:
334 $this->conf['cHash'] = $createCHash ? t3lib_div::generateCHash(t3lib_div::implodeArrayForUrl('', $cHash_array)) : ''; // cHash string for additional parameters
335 $this->conf['cHash_array'] = $cHash_array; // Array of the additional parameters
336
337 // Set to defaults
338 $this->conf['freeIndexUid'] = 0;
339 $this->conf['freeIndexSetId'] = 0;
340 $this->conf['page_cache_reg1'] = '';
341
342 // Root line uids
343 $this->conf['rootline_uids'] = $uidRL;
344
345 // Configuration of behavior:
346 $this->conf['index_externals'] = 1; // Whether to index external documents like PDF, DOC etc. (if possible)
347 $this->conf['index_descrLgd'] = 200; // Length of description text (max 250, default 200)
348 $this->conf['index_metatags'] = true; // Whether to index document keywords and description (if present)
349
350 // Init and start indexing:
351 $this->init();
352 }
353
354 /**
355 * Sets the free-index uid. Can be called right after backend_initIndexer()
356 *
357 * @param integer Free index UID
358 * @param integer Set id - an integer identifying the "set" of indexing operations.
359 * @return void
360 */
361 function backend_setFreeIndexUid($freeIndexUid, $freeIndexSetId=0) {
362 $this->conf['freeIndexUid'] = $freeIndexUid;
363 $this->conf['freeIndexSetId'] = $freeIndexSetId;
364 }
365
366 /**
367 * Indexing records as the content of a TYPO3 page.
368 *
369 * @param string Title equivalent
370 * @param string Keywords equivalent
371 * @param string Description equivalent
372 * @param string The main content to index
373 * @param string The charset of the title, keyword, description and body-content. MUST BE VALID, otherwise nothing is indexed!
374 * @param integer Last modification time, in seconds
375 * @param integer The creation date of the content, in seconds
376 * @param integer The record UID that the content comes from (for registration with the indexed rows)
377 * @return void
378 */
379 function backend_indexAsTYPO3Page($title, $keywords, $description, $content, $charset, $mtime, $crdate=0, $recordUid=0) {
380
381 // Content of page:
382 $this->conf['mtime'] = $mtime; // Most recent modification time (seconds) of the content
383 $this->conf['crdate'] = $crdate; // The creation date of the TYPO3 content
384 $this->conf['recordUid'] = $recordUid; // UID of the record, if applicable
385
386 // Construct fake HTML for parsing:
387 $this->conf['content'] = '
388 <html>
389 <head>
390 <title>'.htmlspecialchars($title).'</title>
391 <meta name="keywords" content="'.htmlspecialchars($keywords).'" />
392 <meta name="description" content="'.htmlspecialchars($description).'" />
393 </head>
394 <body>
395 '.htmlspecialchars($content).'
396 </body>
397 </html>'; // Content string (HTML of TYPO3 page)
398
399 // Initializing charset:
400 $this->conf['metaCharset'] = $charset; // Character set of content (will be converted to utf-8 during indexing)
401 $this->conf['indexedDocTitle'] = ''; // Alternative title for indexing
402
403 // Index content as if it was a TYPO3 page:
404 $this->indexTypo3PageContent();
405 }
406
407
408
409
410
411
412
413
414
415
416
417
418
419 /********************************
420 *
421 * Initialization
422 *
423 *******************************/
424
425 /**
426 * Initializes the object. $this->conf MUST be set with proper values prior to this call!!!
427 *
428 * @return void
429 */
430 function init() {
431 global $TYPO3_CONF_VARS;
432
433 // Initializing:
434 $this->cHashParams = $this->conf['cHash_array'];
435 if (is_array($this->cHashParams) && count($this->cHashParams)) {
436 if ($this->conf['cHash']) $this->cHashParams['cHash'] = $this->conf['cHash']; // Add this so that URL's come out right...
437 unset($this->cHashParams['encryptionKey']); // encryptionKey is added inside TSFE in order to calculate the cHash value and it should NOT be a part of this array!!! If it is it will be exposed in links!!!
438 }
439
440 // Setting phash / phash_grouping which identifies the indexed page based on some of these variables:
441 $this->setT3Hashes();
442
443 // Indexer configuration from Extension Manager interface:
444 $this->indexerConfig = unserialize($GLOBALS['TYPO3_CONF_VARS']['EXT']['extConf']['indexed_search']);
445 $this->tstamp_minAge = t3lib_div::intInRange($this->indexerConfig['minAge']*3600,0);
446 $this->tstamp_maxAge = t3lib_div::intInRange($this->indexerConfig['maxAge']*3600,0);
447 $this->maxExternalFiles = t3lib_div::intInRange($this->indexerConfig['maxExternalFiles'],0,1000,5);
448 $this->flagBitMask = t3lib_div::intInRange($this->indexerConfig['flagBitMask'],0,255);
449
450 // Initialize external document parsers:
451 // Example configuration, see ext_localconf.php of this file!
452 if ($this->conf['index_externals']) {
453 $this->initializeExternalParsers();
454 }
455
456 // Initialize lexer (class that deconstructs the text into words):
457 // Example configuration (localconf.php) for this hook: $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['lexer'] = 'EXT:indexed_search/class.lexer.php:&tx_indexedsearch_lexer';
458 $lexerObjRef = $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['lexer'] ?
459 $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['lexer'] :
460 'EXT:indexed_search/class.lexer.php:&tx_indexedsearch_lexer';
461 $this->lexerObj = t3lib_div::getUserObj($lexerObjRef);
462 $this->lexerObj->debug = $this->indexerConfig['debugMode'];
463
464 // Initialize metaphone hook:
465 // Example configuration (localconf.php) for this hook: $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['metaphone'] = 'EXT:indexed_search/class.doublemetaphone.php:&user_DoubleMetaPhone';
466 if ($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['metaphone']) {
467 $this->metaphoneObj = t3lib_div::getUserObj($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['metaphone']);
468 $this->metaphoneObj->pObj = $this;
469 }
470
471 // Init charset class:
472 $this->csObj = t3lib_div::makeInstance('t3lib_cs');
473 }
474
475 /**
476 * Initialize external parsers
477 *
478 * @return void
479 * @access private
480 * @see init()
481 */
482 function initializeExternalParsers() {
483 global $TYPO3_CONF_VARS;
484
485 if (is_array($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['external_parsers'])) {
486 foreach($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['external_parsers'] as $extension => $_objRef) {
487 $this->external_parsers[$extension] = t3lib_div::getUserObj($_objRef);
488 $this->external_parsers[$extension]->pObj = $this;
489
490 // Init parser and if it returns false, unset its entry again:
491 if (!$this->external_parsers[$extension]->initParser($extension)) {
492 unset($this->external_parsers[$extension]);
493 }
494 }
495 }
496 }
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512 /********************************
513 *
514 * Indexing; TYPO3 pages (HTML content)
515 *
516 *******************************/
517
518 /**
519 * Start indexing of the TYPO3 page
520 *
521 * @return void
522 */
523 function indexTypo3PageContent() {
524
525 $check = $this->checkMtimeTstamp($this->conf['mtime'], $this->hash['phash']);
526 $is_grlist = $this->is_grlist_set($this->hash['phash']);
527
528 if ($check > 0 || !$is_grlist || $this->forceIndexing) {
529
530 // Setting message:
531 if ($this->forceIndexing) {
532 $this->log_setTSlogMessage('Indexing needed, reason: Forced',1);
533 } elseif ($check > 0) {
534 $this->log_setTSlogMessage('Indexing needed, reason: '.$this->reasons[$check],1);
535 } else {
536 $this->log_setTSlogMessage('Indexing needed, reason: Updates gr_list!',1);
537 }
538
539 // Divide into title,keywords,description and body:
540 $this->log_push('Split content','');
541 $this->contentParts = $this->splitHTMLContent($this->conf['content']);
542 if ($this->conf['indexedDocTitle']) {
543 $this->contentParts['title'] = $this->conf['indexedDocTitle'];
544 }
545 $this->log_pull();
546
547 // Calculating a hash over what is to be the actual page content. Maybe this hash should not include title,description and keywords? The bodytext is the primary concern. (on the other hand a changed page-title would make no difference then, so dont!)
548 $this->content_md5h = $this->md5inthash(implode($this->contentParts,''));
549
550 // This function checks if there is already a page (with gr_list = 0,-1) indexed and if that page has the very same contentHash.
551 // If the contentHash is the same, then we can rest assured that this page is already indexed and regardless of mtime and origContent we don't need to do anything more.
552 // This will also prevent pages from being indexed if a fe_users has logged in and it turns out that the page content is not changed anyway. fe_users logged in should always search with hash_gr_list = "0,-1" OR "[their_group_list]". This situation will be prevented only if the page has been indexed with no user login on before hand. Else the page will be indexed by users until that event. However that does not present a serious problem.
553 $checkCHash = $this->checkContentHash();
554 if (!is_array($checkCHash) || $check===1) {
555 $Pstart=t3lib_div::milliseconds();
556
557 $this->log_push('Converting charset of content ('.$this->conf['metaCharset'].') to utf-8','');
558 $this->charsetEntity2utf8($this->contentParts,$this->conf['metaCharset']);
559 $this->log_pull();
560
561 // Splitting words
562 $this->log_push('Extract words from content','');
563 $splitInWords = $this->processWordsInArrays($this->contentParts);
564 $this->log_pull();
565
566 // Analyse the indexed words.
567 $this->log_push('Analyse the extracted words','');
568 $indexArr = $this->indexAnalyze($splitInWords);
569 $this->log_pull();
570
571 // Submitting page (phash) record
572 $this->log_push('Submitting page','');
573 $this->submitPage();
574 $this->log_pull();
575
576 // Check words and submit to word list if not there
577 $this->log_push('Check word list and submit words','');
578 $this->checkWordList($indexArr);
579 $this->submitWords($indexArr,$this->hash['phash']);
580 $this->log_pull();
581
582 // Set parsetime
583 $this->updateParsetime($this->hash['phash'],t3lib_div::milliseconds()-$Pstart);
584
585 // Checking external files if configured for.
586 $this->log_push('Checking external files','');
587 if ($this->conf['index_externals']) {
588 $this->extractLinks($this->conf['content']);
589 }
590 $this->log_pull();
591 } else {
592 $this->updateTstamp($this->hash['phash'],$this->conf['mtime']); // Update the timestatmp
593 $this->updateSetId($this->hash['phash']);
594 $this->update_grlist($checkCHash['phash'],$this->hash['phash']); // $checkCHash['phash'] is the phash of the result row that is similar to the current phash regarding the content hash.
595 $this->updateRootline();
596 $this->log_setTSlogMessage('Indexing not needed, the contentHash, '.$this->content_md5h.', has not changed. Timestamp, grlist and rootline updated if necessary.');
597 }
598 } else {
599 $this->log_setTSlogMessage('Indexing not needed, reason: '.$this->reasons[$check]);
600 }
601 }
602
603 /**
604 * Splits HTML content and returns an associative array, with title, a list of metatags, and a list of words in the body.
605 *
606 * @param string HTML content to index. To some degree expected to be made by TYPO3 (ei. splitting the header by ":")
607 * @return array Array of content, having keys "title", "body", "keywords" and "description" set.
608 * @see splitRegularContent()
609 */
610 function splitHTMLContent($content) {
611
612 // divide head from body ( u-ouh :) )
613 $contentArr = $this->defaultContentArray;
614 $contentArr['body'] = stristr($content,'<body');
615 $headPart = substr($content,0,-strlen($contentArr['body']));
616
617 // get title
618 $this->embracingTags($headPart,'TITLE',$contentArr['title'],$dummy2,$dummy);
619 $titleParts = explode(':',$contentArr['title'],2);
620 $contentArr['title'] = trim(isset($titleParts[1]) ? $titleParts[1] : $titleParts[0]);
621
622 // get keywords and description metatags
623 if($this->conf['index_metatags']) {
624 for($i=0;$this->embracingTags($headPart,'meta',$dummy,$headPart,$meta[$i]);$i++) { /*nothing*/ }
625 for($i=0;isset($meta[$i]);$i++) {
626 $meta[$i] = t3lib_div::get_tag_attributes($meta[$i]);
627 if (stristr($meta[$i]['name'], 'keywords')) {
628 $contentArr['keywords'] .= ',' . $this->addSpacesToKeywordList($meta[$i]['content']);
629 }
630 if (stristr($meta[$i]['name'], 'description')) {
631 $contentArr['description'] .= ',' . $meta[$i]['content'];
632 }
633 }
634 }
635
636 // Process <!--TYPO3SEARCH_begin--> or <!--TYPO3SEARCH_end--> tags:
637 $this->typoSearchTags($contentArr['body']);
638
639 // Get rid of unwanted sections (ie. scripting and style stuff) in body
640 $tagList = explode(',',$this->excludeSections);
641 foreach($tagList as $tag) {
642 while($this->embracingTags($contentArr['body'],$tag,$dummy,$contentArr['body'],$dummy2));
643 }
644
645 // remove tags, but first make sure we don't concatenate words by doing it
646 $contentArr['body'] = str_replace('<',' <',$contentArr['body']);
647 $contentArr['body'] = trim(strip_tags($contentArr['body']));
648
649 $contentArr['keywords'] = trim($contentArr['keywords']);
650 $contentArr['description'] = trim($contentArr['description']);
651
652 // Return array
653 return $contentArr;
654 }
655
656 /**
657 * Extract the charset value from HTML meta tag.
658 *
659 * @param string HTML content
660 * @return string The charset value if found.
661 */
662 function getHTMLcharset($content) {
663 if (preg_match('/<meta[[:space:]]+[^>]*http-equiv[[:space:]]*=[[:space:]]*["\']CONTENT-TYPE["\'][^>]*>/i',$content,$reg)) {
664 if (preg_match('/charset[[:space:]]*=[[:space:]]*([[:alnum:]-]+)/i',$reg[0],$reg2)) {
665 return $reg2[1];
666 }
667 }
668 }
669
670 /**
671 * Converts a HTML document to utf-8
672 *
673 * @param string HTML content, any charset
674 * @param string Optional charset (otherwise extracted from HTML)
675 * @return string Converted HTML
676 */
677 function convertHTMLToUtf8($content,$charset='') {
678
679 // Find charset:
680 $charset = $charset ? $charset : $this->getHTMLcharset($content);
681 $charset = $this->csObj->parse_charset($charset);
682
683 // Convert charset:
684 if ($charset && $charset!=='utf-8') {
685 $content = $this->csObj->utf8_encode($content, $charset);
686 }
687 // Convert entities, assuming document is now UTF-8:
688 $content = $this->csObj->entities_to_utf8($content, TRUE);
689
690 return $content;
691 }
692
693 /**
694 * Finds first occurence of embracing tags and returns the embraced content and the original string with
695 * the tag removed in the two passed variables. Returns false if no match found. ie. useful for finding
696 * <title> of document or removing <script>-sections
697 *
698 * @param string String to search in
699 * @param string Tag name, eg. "script"
700 * @param string Passed by reference: Content inside found tag
701 * @param string Passed by reference: Content after found tag
702 * @param string Passed by reference: Attributes of the found tag.
703 * @return boolean Returns false if tag was not found, otherwise true.
704 */
705 function embracingTags($string,$tagName,&$tagContent,&$stringAfter,&$paramList) {
706 $endTag = '</'.$tagName.'>';
707 $startTag = '<'.$tagName;
708
709 $isTagInText = stristr($string,$startTag); // stristr used because we want a case-insensitive search for the tag.
710 if(!$isTagInText) return false; // if the tag was not found, return false
711
712 list($paramList,$isTagInText) = explode('>',substr($isTagInText,strlen($startTag)),2);
713 $afterTagInText = stristr($isTagInText,$endTag);
714 if ($afterTagInText) {
715 $stringBefore = substr($string, 0, strpos(strtolower($string), strtolower($startTag)));
716 $tagContent = substr($isTagInText,0,strlen($isTagInText)-strlen($afterTagInText));
717 $stringAfter = $stringBefore.substr($afterTagInText,strlen($endTag));
718 } else { // If there was no ending tag, the tagContent is blank and anything after the tag it self is returned.
719 $tagContent='';
720 $stringAfter = $isTagInText;
721 }
722
723 return true;
724 }
725
726 /**
727 * Removes content that shouldn't be indexed according to TYPO3SEARCH-tags.
728 *
729 * @param string HTML Content, passed by reference
730 * @return boolean Returns true if a TYPOSEARCH_ tag was found, otherwise false.
731 */
732 function typoSearchTags(&$body) {
733 $expBody = preg_split('/\<\!\-\-[\s]?TYPO3SEARCH_/',$body);
734
735 if(count($expBody)>1) {
736 $body = '';
737
738 foreach($expBody as $val) {
739 $part = explode('-->',$val,2);
740 if(trim($part[0])=='begin') {
741 $body.= $part[1];
742 $prev = '';
743 } elseif(trim($part[0])=='end') {
744 $body.= $prev;
745 } else {
746 $prev = $val;
747 }
748 }
749 return true;
750 } else {
751 return false;
752 }
753 }
754
755 /**
756 * Extract links (hrefs) from HTML content and if indexable media is found, it is indexed.
757 *
758 * @param string HTML content
759 * @return void
760 */
761 function extractLinks($content) {
762
763 // Get links:
764 $list = $this->extractHyperLinks($content);
765
766 if ($this->indexerConfig['useCrawlerForExternalFiles'] && t3lib_extMgm::isLoaded('crawler')) {
767 $this->includeCrawlerClass();
768 $crawler = t3lib_div::makeInstance('tx_crawler_lib');
769 }
770
771 // Traverse links:
772 foreach($list as $linkInfo) {
773
774 // Decode entities:
775 if ($linkInfo['localPath']) { // localPath means: This file is sent by a download script. While the indexed URL has to point to $linkInfo['href'], the absolute path to the file is specified here!
776 $linkSource = t3lib_div::htmlspecialchars_decode($linkInfo['localPath']);
777 } else {
778 $linkSource = t3lib_div::htmlspecialchars_decode($linkInfo['href']);
779 }
780
781 // Parse URL:
782 $qParts = parse_url($linkSource);
783
784 // Check for jumpurl (TYPO3 specific thing...)
785 if ($qParts['query'] && strstr($qParts['query'],'jumpurl=')) {
786 parse_str($qParts['query'],$getP);
787 $linkSource = $getP['jumpurl'];
788 $qParts = parse_url($linkSource); // parse again due to new linkSource!
789 }
790
791 if ($qParts['scheme']) {
792 if ($this->indexerConfig['indexExternalURLs']) {
793 // Index external URL (http or otherwise)
794 $this->indexExternalUrl($linkSource);
795 }
796 } elseif (!$qParts['query']) {
797 $linkSource = urldecode($linkSource);
798 if (t3lib_div::isAllowedAbsPath($linkSource)) {
799 $localFile = $linkSource;
800 } else {
801 $localFile = t3lib_div::getFileAbsFileName(PATH_site.$linkSource);
802 }
803 if ($localFile && @is_file($localFile)) {
804
805 // Index local file:
806 if ($linkInfo['localPath']) {
807
808 $fI = pathinfo($linkSource);
809 $ext = strtolower($fI['extension']);
810 if (is_object($crawler)) {
811 $params = array(
812 'document' => $linkSource,
813 'alturl' => $linkInfo['href'],
814 'conf' => $this->conf
815 );
816 unset($params['conf']['content']);
817
818 $crawler->addQueueEntry_callBack(0,$params,'EXT:indexed_search/class.crawler.php:&tx_indexedsearch_files',$this->conf['id']);
819 $this->log_setTSlogMessage('media "'.$params['document'].'" added to "crawler" queue.',1);
820 } else {
821 $this->indexRegularDocument($linkInfo['href'], false, $linkSource, $ext);
822 }
823 } else {
824 if (is_object($crawler)) {
825 $params = array(
826 'document' => $linkSource,
827 'conf' => $this->conf
828 );
829 unset($params['conf']['content']);
830 $crawler->addQueueEntry_callBack(0,$params,'EXT:indexed_search/class.crawler.php:&tx_indexedsearch_files',$this->conf['id']);
831 $this->log_setTSlogMessage('media "'.$params['document'].'" added to "crawler" queue.',1);
832 } else {
833 $this->indexRegularDocument($linkSource);
834 }
835 }
836 }
837 }
838 }
839 }
840
841 /**
842 * Extracts all links to external documents from content string.
843 *
844 * @param string Content to analyse
845 * @return array Array of hyperlinks
846 * @see extractLinks()
847 */
848 function extractHyperLinks($string) {
849 if (!is_object($this->htmlParser)) {
850 $this->htmlParser = t3lib_div::makeInstance('t3lib_parseHtml');
851 }
852
853 $parts = $this->htmlParser->splitTags('a',$string);
854 $list = array();
855 foreach ($parts as $k => $v) {
856 if ($k%2) {
857 $params = $this->htmlParser->get_tag_attributes($v,1);
858 $firstTagName = $this->htmlParser->getFirstTagName($v); // The 'name' of the first tag
859
860 switch (strtolower($firstTagName)) {
861 case 'a':
862 $src = $params[0]['href'];
863 if ($src) {
864 // Check if a local path to that file has been set - useful if you are using a download script.
865 $md5 = t3lib_div::shortMD5($src);
866 if (is_array($indexLocalFiles=$GLOBALS['T3_VAR']['ext']['indexed_search']['indexLocalFiles'])) {
867 $localPath = isset($indexLocalFiles[$md5]) ? $indexLocalFiles[$md5] : '';
868 } else $localPath=false;
869
870 $list[] = array(
871 'tag' => $v,
872 'href' => $params[0]['href'],
873 'localPath' => $localPath
874 );
875 }
876 break;
877 }
878 }
879 }
880
881 return $list;
882 }
883
884 /**
885 * Extracts the "base href" from content string.
886 *
887 * @param string Content to analyze
888 * @return string The base href or an empty string if not found
889 */
890 public function extractBaseHref($string) {
891 if (!is_object($this->htmlParser)) {
892 $this->htmlParser = t3lib_div::makeInstance('t3lib_parseHtml');
893 }
894
895 $parts = $this->htmlParser->splitTags('base', $string);
896 foreach ($parts as $key => $value) {
897 if ($key % 2) {
898 $params = $this->htmlParser->get_tag_attributes($value, 1);
899 $firstTagName = $this->htmlParser->getFirstTagName($value); // The 'name' of the first tag
900
901 switch (strtolower($firstTagName)) {
902 case 'base':
903 $href = $params[0]['href'];
904 if ($href) {
905 // Return the first "base href" found (a single one should be present anyway)
906 return $href;
907 }
908 }
909 }
910 }
911
912 return '';
913 }
914
915
916
917
918
919
920
921 /******************************************
922 *
923 * Indexing; external URL
924 *
925 ******************************************/
926
927 /**
928 * Index External URLs HTML content
929 *
930 * @param string URL, eg. "http://typo3.org/"
931 * @return void
932 * @see indexRegularDocument()
933 */
934 function indexExternalUrl($externalUrl) {
935
936 // Parse External URL:
937 $qParts = parse_url($externalUrl);
938 $fI = pathinfo($qParts['path']);
939 $ext = strtolower($fI['extension']);
940
941 // Get headers:
942 $urlHeaders = $this->getUrlHeaders($externalUrl);
943 if (stristr($urlHeaders['Content-Type'],'text/html')) {
944 $content = $this->indexExternalUrl_content = t3lib_div::getUrl($externalUrl);
945 if (strlen($content)) {
946
947 // Create temporary file:
948 $tmpFile = t3lib_div::tempnam('EXTERNAL_URL');
949 if ($tmpFile) {
950 t3lib_div::writeFile($tmpFile, $content);
951
952 // Index that file:
953 $this->indexRegularDocument($externalUrl, TRUE, $tmpFile, 'html'); // Using "TRUE" for second parameter to force indexing of external URLs (mtime doesn't make sense, does it?)
954 unlink($tmpFile);
955 }
956 }
957 }
958 }
959
960 /**
961 * Getting HTTP request headers of URL
962 *
963 * @param string The URL
964 * @param integer Timeout (seconds?)
965 * @return mixed If no answer, returns false. Otherwise an array where HTTP headers are keys
966 */
967 function getUrlHeaders($url) {
968 $content = t3lib_div::getURL($url,2); // Try to get the headers only
969
970 if (strlen($content)) {
971 // Compile headers:
972 $headers = t3lib_div::trimExplode(LF,$content,1);
973 $retVal = array();
974 foreach($headers as $line) {
975 if (!strlen(trim($line))) {
976 break; // Stop at the first empty line (= end of header)
977 }
978
979 list($headKey, $headValue) = explode(':', $line, 2);
980 $retVal[$headKey] = $headValue;
981 }
982 return $retVal;
983 }
984 }
985
986
987
988
989
990
991
992
993
994
995
996
997
998 /******************************************
999 *
1000 * Indexing; external files (PDF, DOC, etc)
1001 *
1002 ******************************************/
1003
1004 /**
1005 * Indexing a regular document given as $file (relative to PATH_site, local file)
1006 *
1007 * @param string Relative Filename, relative to PATH_site. It can also be an absolute path as long as it is inside the lockRootPath (validated with t3lib_div::isAbsPath()). Finally, if $contentTmpFile is set, this value can be anything, most likely a URL
1008 * @param boolean If set, indexing is forced (despite content hashes, mtime etc).
1009 * @param string Temporary file with the content to read it from (instead of $file). Used when the $file is a URL.
1010 * @param string File extension for temporary file.
1011 * @return void
1012 */
1013 function indexRegularDocument($file, $force=FALSE, $contentTmpFile='', $altExtension='') {
1014
1015 // Init
1016 $fI = pathinfo($file);
1017 $ext = $altExtension ? $altExtension : strtolower($fI['extension']);
1018
1019 // Create abs-path:
1020 if (!$contentTmpFile) {
1021 if (!t3lib_div::isAbsPath($file)) { // Relative, prepend PATH_site:
1022 $absFile = t3lib_div::getFileAbsFileName(PATH_site.$file);
1023 } else { // Absolute, pass-through:
1024 $absFile = $file;
1025 }
1026 $absFile = t3lib_div::isAllowedAbsPath($absFile) ? $absFile : '';
1027 } else {
1028 $absFile = $contentTmpFile;
1029 }
1030
1031 // Indexing the document:
1032 if ($absFile && @is_file($absFile)) {
1033 if ($this->external_parsers[$ext]) {
1034 $mtime = filemtime($absFile);
1035 $cParts = $this->fileContentParts($ext,$absFile);
1036
1037 foreach($cParts as $cPKey) {
1038 $this->internal_log = array();
1039 $this->log_push('Index: '.str_replace('.','_',basename($file)).($cPKey?'#'.$cPKey:''),'');
1040 $Pstart = t3lib_div::milliseconds();
1041 $subinfo = array('key' => $cPKey); // Setting page range. This is "0" (zero) when no division is made, otherwise a range like "1-3"
1042 $phash_arr = $this->file_phash_arr = $this->setExtHashes($file,$subinfo);
1043 $check = $this->checkMtimeTstamp($mtime, $phash_arr['phash']);
1044 if ($check > 0 || $force) {
1045 if ($check > 0) {
1046 $this->log_setTSlogMessage('Indexing needed, reason: '.$this->reasons[$check],1);
1047 } else {
1048 $this->log_setTSlogMessage('Indexing forced by flag',1);
1049 }
1050
1051 // Check external file counter:
1052 if ($this->externalFileCounter < $this->maxExternalFiles || $force) {
1053
1054 // Divide into title,keywords,description and body:
1055 $this->log_push('Split content','');
1056 $contentParts = $this->readFileContent($ext,$absFile,$cPKey);
1057 $this->log_pull();
1058
1059 if (is_array($contentParts)) {
1060 // Calculating a hash over what is to be the actual content. (see indexTypo3PageContent())
1061 $content_md5h = $this->md5inthash(implode($contentParts,''));
1062
1063 if ($this->checkExternalDocContentHash($phash_arr['phash_grouping'], $content_md5h) || $force) {
1064
1065 // Increment counter:
1066 $this->externalFileCounter++;
1067
1068 // Splitting words
1069 $this->log_push('Extract words from content','');
1070 $splitInWords = $this->processWordsInArrays($contentParts);
1071 $this->log_pull();
1072
1073 // Analyse the indexed words.
1074 $this->log_push('Analyse the extracted words','');
1075 $indexArr = $this->indexAnalyze($splitInWords);
1076 $this->log_pull();
1077
1078 // Submitting page (phash) record
1079 $this->log_push('Submitting page','');
1080 $size = filesize($absFile);
1081 $ctime = filemtime($absFile); // Unfortunately I cannot determine WHEN a file is originally made - so I must return the modification time...
1082 $this->submitFilePage($phash_arr,$file,$subinfo,$ext,$mtime,$ctime,$size,$content_md5h,$contentParts);
1083 $this->log_pull();
1084
1085 // Check words and submit to word list if not there
1086 $this->log_push('Check word list and submit words','');
1087 $this->checkWordList($indexArr);
1088 $this->submitWords($indexArr,$phash_arr['phash']);
1089 $this->log_pull();
1090
1091 // Set parsetime
1092 $this->updateParsetime($phash_arr['phash'],t3lib_div::milliseconds()-$Pstart);
1093 } else {
1094 $this->updateTstamp($phash_arr['phash'],$mtime); // Update the timestamp
1095 $this->log_setTSlogMessage('Indexing not needed, the contentHash, '.$content_md5h.', has not changed. Timestamp updated.');
1096 }
1097 } else $this->log_setTSlogMessage('Could not index file! Unsupported extension.');
1098 } else $this->log_setTSlogMessage('The limit of '.$this->maxExternalFiles.' has already been exceeded, so no indexing will take place this time.');
1099 } else $this->log_setTSlogMessage('Indexing not needed, reason: '.$this->reasons[$check]);
1100
1101 // Checking and setting sections:
1102 # $this->submitFile_grlist($phash_arr['phash']); // Setting a gr_list record if there is none already (set for default fe_group)
1103 $this->submitFile_section($phash_arr['phash']); // Setting a section-record for the file. This is done also if the file is not indexed. Notice that section records are deleted when the page is indexed.
1104 $this->log_pull();
1105 }
1106 } else $this->log_setTSlogMessage('Indexing not possible; The extension "'.$ext.'" was not supported.');
1107 } else $this->log_setTSlogMessage('Indexing not possible; File "'.$absFile.'" not found or valid.');
1108 }
1109
1110 /**
1111 * Reads the content of an external file being indexed.
1112 * The content from the external parser MUST be returned in utf-8!
1113 *
1114 * @param string File extension, eg. "pdf", "doc" etc.
1115 * @param string Absolute filename of file (must exist and be validated OK before calling function)
1116 * @param string Pointer to section (zero for all other than PDF which will have an indication of pages into which the document should be splitted.)
1117 * @return array Standard content array (title, description, keywords, body keys)
1118 */
1119 function readFileContent($ext,$absFile,$cPKey) {
1120
1121 // Consult relevant external document parser:
1122 if (is_object($this->external_parsers[$ext])) {
1123 $contentArr = $this->external_parsers[$ext]->readFileContent($ext,$absFile,$cPKey);
1124 }
1125
1126 return $contentArr;
1127 }
1128
1129 /**
1130 * Creates an array with pointers to divisions of document.
1131 *
1132 * @param string File extension
1133 * @param string Absolute filename (must exist and be validated OK before calling function)
1134 * @return array Array of pointers to sections that the document should be divided into
1135 */
1136 function fileContentParts($ext,$absFile) {
1137 $cParts = array(0);
1138
1139 // Consult relevant external document parser:
1140 if (is_object($this->external_parsers[$ext])) {
1141 $cParts = $this->external_parsers[$ext]->fileContentParts($ext,$absFile);
1142 }
1143
1144 return $cParts;
1145 }
1146
1147 /**
1148 * Splits non-HTML content (from external files for instance)
1149 *
1150 * @param string Input content (non-HTML) to index.
1151 * @return array Array of content, having the key "body" set (plus "title", "description" and "keywords", but empty)
1152 * @see splitHTMLContent()
1153 */
1154 function splitRegularContent($content) {
1155 $contentArr = $this->defaultContentArray;
1156 $contentArr['body'] = $content;
1157
1158 return $contentArr;
1159 }
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174 /**********************************
1175 *
1176 * Analysing content, Extracting words
1177 *
1178 **********************************/
1179
1180 /**
1181 * Convert character set and HTML entities in the value of input content array keys
1182 *
1183 * @param array Standard content array
1184 * @param string Charset of the input content (converted to utf-8)
1185 * @return void
1186 */
1187 function charsetEntity2utf8(&$contentArr, $charset) {
1188
1189 // Convert charset if necessary
1190 reset($contentArr);
1191 while(list($key,)=each($contentArr)) {
1192 if (strlen($contentArr[$key])) {
1193
1194 if ($charset!=='utf-8') {
1195 $contentArr[$key] = $this->csObj->utf8_encode($contentArr[$key], $charset);
1196 }
1197
1198 // decode all numeric / html-entities in the string to real characters:
1199 $contentArr[$key] = $this->csObj->entities_to_utf8($contentArr[$key],TRUE);
1200 }
1201 }
1202 }
1203
1204 /**
1205 * Processing words in the array from split*Content -functions
1206 *
1207 * @param array Array of content to index, see splitHTMLContent() and splitRegularContent()
1208 * @return array Content input array modified so each key is not a unique array of words
1209 */
1210 function processWordsInArrays($contentArr) {
1211
1212 // split all parts to words
1213 reset($contentArr);
1214 while(list($key,)=each($contentArr)) {
1215 $contentArr[$key] = $this->lexerObj->split2Words($contentArr[$key]);
1216 }
1217
1218 // For title, keywords, and description we don't want duplicates:
1219 $contentArr['title'] = array_unique($contentArr['title']);
1220 $contentArr['keywords'] = array_unique($contentArr['keywords']);
1221 $contentArr['description'] = array_unique($contentArr['description']);
1222
1223 // Return modified array:
1224 return $contentArr;
1225 }
1226
1227 /**
1228 * Processing words in the array from split*Content -functions
1229 * This function is only a wrapper because the function has been removed (see above).
1230 *
1231 * @param array Array of content to index, see splitHTMLContent() and splitRegularContent()
1232 * @return array Content input array modified so each key is not a unique array of words
1233 * @deprecated since TYPO3 4.0, this function will be removed in TYPO3 4.5.
1234 */
1235 function procesWordsInArrays($contentArr) {
1236 t3lib_div::logDeprecatedFunction();
1237
1238 return $this->processWordsInArrays($contentArr);
1239 }
1240
1241 /**
1242 * Extracts the sample description text from the content array.
1243 *
1244 * @param array Content array
1245 * @return string Description string
1246 */
1247 function bodyDescription($contentArr) {
1248
1249 // Setting description
1250 $maxL = t3lib_div::intInRange($this->conf['index_descrLgd'],0,255,200);
1251 if ($maxL) {
1252 // Takes the quadruple lenght first, because whitespace and entities may be removed and thus shorten the string more yet.
1253 # $bodyDescription = implode(' ',split('[[:space:],]+',substr(trim($contentArr['body']),0,$maxL*4)));
1254 $bodyDescription = str_replace(array(' ',TAB,CR,LF),' ',$contentArr['body']);
1255
1256 // Shorten the string:
1257 $bodyDescription = $this->csObj->strtrunc('utf-8', $bodyDescription, $maxL);
1258 }
1259
1260 return $bodyDescription;
1261 }
1262
1263 /**
1264 * Analyzes content to use for indexing,
1265 *
1266 * @param array Standard content array: an array with the keys title,keywords,description and body, which all contain an array of words.
1267 * @return array Index Array (whatever that is...)
1268 */
1269 function indexAnalyze($content) {
1270 $indexArr = Array();
1271 $counter = 0;
1272
1273 $this->analyzeHeaderinfo($indexArr,$content,'title',7);
1274 $this->analyzeHeaderinfo($indexArr,$content,'keywords',6);
1275 $this->analyzeHeaderinfo($indexArr,$content,'description',5);
1276 $this->analyzeBody($indexArr,$content);
1277
1278 return ($indexArr);
1279 }
1280
1281 /**
1282 * Calculates relevant information for headercontent
1283 *
1284 * @param array Index array, passed by reference
1285 * @param array Standard content array
1286 * @param string Key from standard content array
1287 * @param integer Bit-wise priority to type
1288 * @return void
1289 */
1290 function analyzeHeaderinfo(&$retArr,$content,$key,$offset) {
1291 reset($content[$key]);
1292 while(list(,$val)=each($content[$key])) {
1293 $val = substr($val,0,60); // Max 60 - because the baseword varchar IS 60. This MUST be the same.
1294 $retArr[$val]['cmp'] = $retArr[$val]['cmp']|pow(2,$offset);
1295 $retArr[$val]['count'] = $retArr[$val]['count']+1;
1296 $retArr[$val]['hash'] = hexdec(substr(md5($val),0,7));
1297 $retArr[$val]['metaphone'] = $this->metaphone($val);
1298 $this->wordcount++;
1299 }
1300 }
1301
1302 /**
1303 * Calculates relevant information for bodycontent
1304 *
1305 * @param array Index array, passed by reference
1306 * @param array Standard content array
1307 * @return void
1308 */
1309 function analyzeBody(&$retArr,$content) {
1310 foreach($content['body'] as $key => $val) {
1311 $val = substr($val,0,60); // Max 60 - because the baseword varchar IS 60. This MUST be the same.
1312 if(!isset($retArr[$val])) {
1313 $retArr[$val]['first'] = $key;
1314 $retArr[$val]['hash'] = hexdec(substr(md5($val),0,7));
1315 $retArr[$val]['metaphone'] = $this->metaphone($val);
1316 }
1317 $retArr[$val]['count'] = $retArr[$val]['count']+1;
1318 $this->wordcount++;
1319 }
1320 }
1321
1322 /**
1323 * Creating metaphone based hash from input word
1324 *
1325 * @param string Word to convert
1326 * @param boolean If set, returns the raw metaphone value (not hashed)
1327 * @return mixed Metaphone hash integer (or raw value, string)
1328 */
1329 function metaphone($word,$retRaw=FALSE) {
1330
1331 if (is_object($this->metaphoneObj)) {
1332 $tmp = $this->metaphoneObj->metaphone($word, $this->conf['sys_language_uid']);
1333 } else {
1334 $tmp = metaphone($word);
1335 }
1336
1337 // Return raw value?
1338 if ($retRaw) return $tmp;
1339
1340 // Otherwise create hash and return integer
1341 if($tmp=='') $ret=0; else $ret=hexdec(substr(md5($tmp),0,7));
1342 return $ret;
1343 }
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360 /********************************
1361 *
1362 * SQL; TYPO3 Pages
1363 *
1364 *******************************/
1365
1366 /**
1367 * Updates db with information about the page (TYPO3 page, not external media)
1368 *
1369 * @return void
1370 */
1371 function submitPage() {
1372
1373 // Remove any current data for this phash:
1374 $this->removeOldIndexedPages($this->hash['phash']);
1375
1376 // setting new phash_row
1377 $fields = array(
1378 'phash' => $this->hash['phash'],
1379 'phash_grouping' => $this->hash['phash_grouping'],
1380 'cHashParams' => serialize($this->cHashParams),
1381 'contentHash' => $this->content_md5h,
1382 'data_page_id' => $this->conf['id'],
1383 'data_page_reg1' => $this->conf['page_cache_reg1'],
1384 'data_page_type' => $this->conf['type'],
1385 'data_page_mp' => $this->conf['MP'],
1386 'gr_list' => $this->conf['gr_list'],
1387 'item_type' => 0, // TYPO3 page
1388 'item_title' => $this->contentParts['title'],
1389 'item_description' => $this->bodyDescription($this->contentParts),
1390 'item_mtime' => $this->conf['mtime'],
1391 'item_size' => strlen($this->conf['content']),
1392 'tstamp' => $GLOBALS['EXEC_TIME'],
1393 'crdate' => $GLOBALS['EXEC_TIME'],
1394 'item_crdate' => $this->conf['crdate'], // Creation date of page
1395 'sys_language_uid' => $this->conf['sys_language_uid'], // Sys language uid of the page. Should reflect which language it DOES actually display!
1396 'externalUrl' => 0,
1397 'recordUid' => intval($this->conf['recordUid']),
1398 'freeIndexUid' => intval($this->conf['freeIndexUid']),
1399 'freeIndexSetId' => intval($this->conf['freeIndexSetId']),
1400 );
1401
1402 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_phash', $fields);
1403
1404 // PROCESSING index_section
1405 $this->submit_section($this->hash['phash'],$this->hash['phash']);
1406
1407 // PROCESSING index_grlist
1408 $this->submit_grlist($this->hash['phash'],$this->hash['phash']);
1409
1410 // PROCESSING index_fulltext
1411 $fields = array(
1412 'phash' => $this->hash['phash'],
1413 'fulltextdata' => implode(' ', $this->contentParts)
1414 );
1415 if ($this->indexerConfig['fullTextDataLength']>0) {
1416 $fields['fulltextdata'] = substr($fields['fulltextdata'],0,$this->indexerConfig['fullTextDataLength']);
1417 }
1418 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_fulltext', $fields);
1419
1420 // PROCESSING index_debug
1421 if ($this->indexerConfig['debugMode']) {
1422 $fields = array(
1423 'phash' => $this->hash['phash'],
1424 'debuginfo' => serialize(array(
1425 'cHashParams' => $this->cHashParams,
1426 'external_parsers initialized' => array_keys($this->external_parsers),
1427 'conf' => array_merge($this->conf,array('content'=>substr($this->conf['content'],0,1000))),
1428 'contentParts' => array_merge($this->contentParts,array('body' => substr($this->contentParts['body'],0,1000))),
1429 'logs' => $this->internal_log,
1430 'lexer' => $this->lexerObj->debugString,
1431 ))
1432 );
1433 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_debug', $fields);
1434 }
1435 }
1436
1437 /**
1438 * Stores gr_list in the database.
1439 *
1440 * @param integer Search result record phash
1441 * @param integer Actual phash of current content
1442 * @return void
1443 * @see update_grlist()
1444 */
1445 function submit_grlist($hash,$phash_x) {
1446
1447 // Setting the gr_list record
1448 $fields = array(
1449 'phash' => $hash,
1450 'phash_x' => $phash_x,
1451 'hash_gr_list' => $this->md5inthash($this->conf['gr_list']),
1452 'gr_list' => $this->conf['gr_list']
1453 );
1454 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_grlist', $fields);
1455 }
1456
1457 /**
1458 * Stores section
1459 * $hash and $hash_t3 are the same for TYPO3 pages, but different when it is external files.
1460 *
1461 * @param integer phash of TYPO3 parent search result record
1462 * @param integer phash of the file indexation search record
1463 * @return void
1464 */
1465 function submit_section($hash,$hash_t3) {
1466 $fields = array(
1467 'phash' => $hash,
1468 'phash_t3' => $hash_t3,
1469 'page_id' => intval($this->conf['id'])
1470 );
1471
1472 $this->getRootLineFields($fields);
1473
1474 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_section', $fields);
1475 }
1476
1477 /**
1478 * Removes records for the indexed page, $phash
1479 *
1480 * @param integer phash value to flush
1481 * @return void
1482 */
1483 function removeOldIndexedPages($phash) {
1484 // Removing old registrations for all tables. Because the pages are TYPO3 pages there can be nothing else than 1-1 relations here.
1485 $tableArr = explode(',','index_phash,index_section,index_grlist,index_fulltext,index_debug');
1486 foreach($tableArr as $table) {
1487 $GLOBALS['TYPO3_DB']->exec_DELETEquery($table, 'phash='.intval($phash));
1488 }
1489 // Removing all index_section records with hash_t3 set to this hash (this includes such records set for external media on the page as well!). The re-insert of these records are done in indexRegularDocument($file).
1490 $GLOBALS['TYPO3_DB']->exec_DELETEquery('index_section', 'phash_t3='.intval($phash));
1491 }
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505 /********************************
1506 *
1507 * SQL; External media
1508 *
1509 *******************************/
1510
1511
1512 /**
1513 * Updates db with information about the file
1514 *
1515 * @param array Array with phash and phash_grouping keys for file
1516 * @param string File name
1517 * @param array Array of "cHashParams" for files: This is for instance the page index for a PDF file (other document types it will be a zero)
1518 * @param string File extension determining the type of media.
1519 * @param integer Modification time of file.
1520 * @param integer Creation time of file.
1521 * @param integer Size of file in bytes
1522 * @param integer Content HASH value.
1523 * @param array Standard content array (using only title and body for a file)
1524 * @return void
1525 */
1526 function submitFilePage($hash,$file,$subinfo,$ext,$mtime,$ctime,$size,$content_md5h,$contentParts) {
1527
1528 // Find item Type:
1529 $storeItemType = $this->external_parsers[$ext]->ext2itemtype_map[$ext];
1530 $storeItemType = $storeItemType ? $storeItemType : $ext;
1531
1532 // Remove any current data for this phash:
1533 $this->removeOldIndexedFiles($hash['phash']);
1534
1535 // Split filename:
1536 $fileParts = parse_url($file);
1537
1538 // Setting new
1539 $fields = array(
1540 'phash' => $hash['phash'],
1541 'phash_grouping' => $hash['phash_grouping'],
1542 'cHashParams' => serialize($subinfo),
1543 'contentHash' => $content_md5h,
1544 'data_filename' => $file,
1545 'item_type' => $storeItemType,
1546 'item_title' => trim($contentParts['title']) ? $contentParts['title'] : basename($file),
1547 'item_description' => $this->bodyDescription($contentParts),
1548 'item_mtime' => $mtime,
1549 'item_size' => $size,
1550 'item_crdate' => $ctime,
1551 'tstamp' => $GLOBALS['EXEC_TIME'],
1552 'crdate' => $GLOBALS['EXEC_TIME'],
1553 'gr_list' => $this->conf['gr_list'],
1554 'externalUrl' => $fileParts['scheme'] ? 1 : 0,
1555 'recordUid' => intval($this->conf['recordUid']),
1556 'freeIndexUid' => intval($this->conf['freeIndexUid']),
1557 'freeIndexSetId' => intval($this->conf['freeIndexSetId']),
1558 );
1559 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_phash', $fields);
1560
1561 // PROCESSING index_fulltext
1562 $fields = array(
1563 'phash' => $hash['phash'],
1564 'fulltextdata' => implode(' ', $contentParts)
1565 );
1566 if ($this->indexerConfig['fullTextDataLength']>0) {
1567 $fields['fulltextdata'] = substr($fields['fulltextdata'],0,$this->indexerConfig['fullTextDataLength']);
1568 }
1569 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_fulltext', $fields);
1570
1571 // PROCESSING index_debug
1572 if ($this->indexerConfig['debugMode']) {
1573 $fields = array(
1574 'phash' => $hash['phash'],
1575 'debuginfo' => serialize(array(
1576 'cHashParams' => $subinfo,
1577 'contentParts' => array_merge($contentParts,array('body' => substr($contentParts['body'],0,1000))),
1578 'logs' => $this->internal_log,
1579 'lexer' => $this->lexerObj->debugString,
1580 ))
1581 );
1582 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_debug', $fields);
1583 }
1584 }
1585
1586 /**
1587 * Stores file gr_list for a file IF it does not exist already
1588 *
1589 * @param integer phash value of file
1590 * @return void
1591 */
1592 function submitFile_grlist($hash) {
1593 // Testing if there is a gr_list record for a non-logged in user and if so, there is no need to place another one.
1594 $count = $GLOBALS['TYPO3_DB']->exec_SELECTcountRows(
1595 'phash',
1596 'index_grlist',
1597 'phash=' . intval($hash) .
1598 ' AND (hash_gr_list=' . $this->md5inthash($this->defaultGrList) .
1599 ' OR hash_gr_list=' . $this->md5inthash($this->conf['gr_list']) . ')'
1600 );
1601 if (!$count) {
1602 $this->submit_grlist($hash,$hash);
1603 }
1604 }
1605
1606 /**
1607 * Stores file section for a file IF it does not exist
1608 *
1609 * @param integer phash value of file
1610 * @return void
1611 */
1612 function submitFile_section($hash) {
1613 // Testing if there is a section
1614 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_section', 'phash='.intval($hash).' AND page_id='.intval($this->conf['id']));
1615 if (!$GLOBALS['TYPO3_DB']->sql_num_rows($res)) {
1616 $this->submit_section($hash,$this->hash['phash']);
1617 }
1618 }
1619
1620 /**
1621 * Removes records for the indexed page, $phash
1622 *
1623 * @param integer phash value to flush
1624 * @return void
1625 */
1626 function removeOldIndexedFiles($phash) {
1627
1628 // Removing old registrations for tables.
1629 $tableArr = explode(',','index_phash,index_grlist,index_fulltext,index_debug');
1630 foreach($tableArr as $table) {
1631 $GLOBALS['TYPO3_DB']->exec_DELETEquery($table, 'phash='.intval($phash));
1632 }
1633 }
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648 /********************************
1649 *
1650 * SQL Helper functions
1651 *
1652 *******************************/
1653
1654 /**
1655 * Check the mtime / tstamp of the currently indexed page/file (based on phash)
1656 * Return positive integer if the page needs to be indexed
1657 *
1658 * @param integer mtime value to test against limits and indexed page (usually this is the mtime of the cached document)
1659 * @param integer "phash" used to select any already indexed page to see what its mtime is.
1660 * @return integer Result integer: Generally: <0 = No indexing, >0 = Do indexing (see $this->reasons): -2) Min age was NOT exceeded and so indexing cannot occur. -1) mtime matched so no need to reindex page. 0) N/A 1) Max age exceeded, page must be indexed again. 2) mtime of indexed page doesn't match mtime given for current content and we must index page. 3) No mtime was set, so we will index... 4) No indexed page found, so of course we will index.
1661 */
1662 function checkMtimeTstamp($mtime,$phash) {
1663
1664 // Select indexed page:
1665 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('item_mtime,tstamp', 'index_phash', 'phash='.intval($phash));
1666 $out = 0;
1667
1668 // If there was an indexing of the page...:
1669 if ($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1670 if ($this->tstamp_maxAge && ($row['tstamp'] + $this->tstamp_maxAge) < $GLOBALS['EXEC_TIME']) { // If max age is exceeded, index the page
1671 $out = 1; // The configured max-age was exceeded for the document and thus it's indexed.
1672 } else {
1673 if (!$this->tstamp_minAge || ($row['tstamp'] + $this->tstamp_minAge) < $GLOBALS['EXEC_TIME']) { // if minAge is not set or if minAge is exceeded, consider at mtime
1674 if ($mtime) { // It mtime is set, then it's tested. If not, the page must clearly be indexed.
1675 if ($row['item_mtime'] != $mtime) { // And if mtime is different from the index_phash mtime, it's about time to re-index.
1676 $out = 2; // The minimum age was exceed and mtime was set and the mtime was different, so the page was indexed.
1677 } else {
1678 $out = -1; // mtime matched the document, so no changes detected and no content updated
1679 if ($this->tstamp_maxAge) {
1680 $this->log_setTSlogMessage('mtime matched, timestamp NOT updated because a maxAge is set (' . ($row['tstamp'] + $this->tstamp_maxAge - $GLOBALS['EXEC_TIME']) . ' seconds to expire time).', 1);
1681 } else {
1682 $this->updateTstamp($phash); // Update the timestatmp
1683 $this->log_setTSlogMessage('mtime matched, timestamp updated.',1);
1684 }
1685 }
1686 } else {$out = 3; } // The minimum age was exceed, but mtime was not set, so the page was indexed.
1687 } else {$out = -2;} // The minimum age was not exceeded
1688 }
1689 } else {$out = 4;} // Page has never been indexed (is not represented in the index_phash table).
1690 return $out;
1691 }
1692
1693 /**
1694 * Check content hash in phash table
1695 *
1696 * @return mixed Returns true if the page needs to be indexed (that is, there was no result), otherwise the phash value (in an array) of the phash record to which the grlist_record should be related!
1697 */
1698 function checkContentHash() {
1699 // With this query the page will only be indexed if it's content is different from the same "phash_grouping" -page.
1700 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_phash A', 'A.phash_grouping='.intval($this->hash['phash_grouping']).' AND A.contentHash='.intval($this->content_md5h));
1701 if ($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1702 return $row;
1703 }
1704 return 1;
1705 }
1706
1707 /**
1708 * Check content hash for external documents
1709 * Returns true if the document needs to be indexed (that is, there was no result)
1710 *
1711 * @param integer phash value to check (phash_grouping)
1712 * @param integer Content hash to check
1713 * @return boolean Returns true if the document needs to be indexed (that is, there was no result)
1714 */
1715 function checkExternalDocContentHash($hashGr,$content_md5h) {
1716 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('*', 'index_phash A', 'A.phash_grouping='.intval($hashGr).' AND A.contentHash='.intval($content_md5h));
1717 if ($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1718 return 0;
1719 }
1720 return 1;
1721 }
1722
1723 /**
1724 * Checks if a grlist record has been set for the phash value input (looking at the "real" phash of the current content, not the linked-to phash of the common search result page)
1725 *
1726 * @param integer Phash integer to test.
1727 * @return void
1728 */
1729 function is_grlist_set($phash_x) {
1730 return $GLOBALS['TYPO3_DB']->exec_SELECTcountRows(
1731 'phash_x',
1732 'index_grlist',
1733 'phash_x=' . intval($phash_x)
1734 );
1735 }
1736
1737 /**
1738 * Check if an grlist-entry for this hash exists and if not so, write one.
1739 *
1740 * @param integer phash of the search result that should be found
1741 * @param integer The real phash of the current content. The two values are different when a page with userlogin turns out to contain the exact same content as another already indexed version of the page; This is the whole reason for the grlist table in fact...
1742 * @return void
1743 * @see submit_grlist()
1744 */
1745 function update_grlist($phash,$phash_x) {
1746 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_grlist', 'phash='.intval($phash).' AND hash_gr_list='.$this->md5inthash($this->conf['gr_list']));
1747 if (!$GLOBALS['TYPO3_DB']->sql_num_rows($res)) {
1748 $this->submit_grlist($phash,$phash_x);
1749 $this->log_setTSlogMessage("Inserted gr_list '".$this->conf['gr_list']."' for phash '".$phash."'",1);
1750 }
1751 }
1752
1753 /**
1754 * Update tstamp for a phash row.
1755 *
1756 * @param integer phash value
1757 * @param integer If set, update the mtime field to this value.
1758 * @return void
1759 */
1760 function updateTstamp($phash,$mtime=0) {
1761 $updateFields = array(
1762 'tstamp' => $GLOBALS['EXEC_TIME']
1763 );
1764 if ($mtime) { $updateFields['item_mtime'] = intval($mtime); }
1765
1766 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_phash', 'phash='.intval($phash), $updateFields);
1767 }
1768
1769 /**
1770 * Update SetID of the index_phash record.
1771 *
1772 * @param integer phash value
1773 * @return void
1774 */
1775 function updateSetId($phash) {
1776 $updateFields = array(
1777 'freeIndexSetId' => intval($this->conf['freeIndexSetId'])
1778 );
1779
1780 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_phash', 'phash='.intval($phash), $updateFields);
1781 }
1782
1783 /**
1784 * Update parsetime for phash row.
1785 *
1786 * @param integer phash value.
1787 * @param integer Parsetime value to set.
1788 * @return void
1789 */
1790 function updateParsetime($phash,$parsetime) {
1791 $updateFields = array(
1792 'parsetime' => intval($parsetime)
1793 );
1794
1795 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_phash', 'phash='.intval($phash), $updateFields);
1796 }
1797
1798 /**
1799 * Update section rootline for the page
1800 *
1801 * @return void
1802 */
1803 function updateRootline() {
1804
1805 $updateFields = array();
1806 $this->getRootLineFields($updateFields);
1807
1808 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_section', 'page_id='.intval($this->conf['id']), $updateFields);
1809 }
1810
1811 /**
1812 * Adding values for root-line fields.
1813 * rl0, rl1 and rl2 are standard. A hook might add more.
1814 *
1815 * @param array Field array, passed by reference
1816 * @return void
1817 */
1818 function getRootLineFields(&$fieldArr) {
1819
1820 $fieldArr['rl0'] = intval($this->conf['rootline_uids'][0]);
1821 $fieldArr['rl1'] = intval($this->conf['rootline_uids'][1]);
1822 $fieldArr['rl2'] = intval($this->conf['rootline_uids'][2]);
1823
1824 if (is_array($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['indexed_search']['addRootLineFields'])) {
1825 foreach($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['indexed_search']['addRootLineFields'] as $fieldName => $rootLineLevel) {
1826 $fieldArr[$fieldName] = intval($this->conf['rootline_uids'][$rootLineLevel]);
1827 }
1828 }
1829 }
1830
1831 /**
1832 * Removes any indexed pages with userlogins which has the same contentHash
1833 * NOT USED anywhere inside this class!
1834 *
1835 * @return void
1836 */
1837 function removeLoginpagesWithContentHash() {
1838 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('*', 'index_phash A,index_grlist B', '
1839 A.phash=B.phash
1840 AND A.phash_grouping='.intval($this->hash['phash_grouping']).'
1841 AND B.hash_gr_list!='.$this->md5inthash($this->defaultGrList).'
1842 AND A.contentHash='.intval($this->content_md5h));
1843 while($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1844 $this->log_setTSlogMessage("The currently indexed page was indexed under no user-login and apparently this page has been indexed under login conditions earlier, but with the SAME content. Therefore the old similar page with phash='".$row['phash']."' are now removed.",1);
1845 $this->removeOldIndexedPages($row['phash']);
1846 }
1847 }
1848
1849 /**
1850 * Includes the crawler class
1851 *
1852 * @return void
1853 */
1854 function includeCrawlerClass() {
1855 global $TYPO3_CONF_VARS;
1856
1857 require_once(t3lib_extMgm::extPath('crawler').'class.tx_crawler_lib.php');
1858 }
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869 /********************************
1870 *
1871 * SQL; Submitting words
1872 *
1873 *******************************/
1874
1875 /**
1876 * Adds new words to db
1877 *
1878 * @param array Word List array (where each word has information about position etc).
1879 * @return void
1880 */
1881 function checkWordList($wl) {
1882 reset($wl);
1883 $phashArr = array();
1884 while(list($key,) = each($wl)) {
1885 $phashArr[] = $wl[$key]['hash'];
1886 }
1887 if (count($phashArr)) {
1888 $cwl = implode(',',$phashArr);
1889 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('baseword', 'index_words', 'wid IN ('.$cwl.')');
1890
1891 if($GLOBALS['TYPO3_DB']->sql_num_rows($res)!=count($wl)) {
1892 $this->log_setTSlogMessage('Inserting words: '.(count($wl)-$GLOBALS['TYPO3_DB']->sql_num_rows($res)),1);
1893 while($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1894 unset($wl[$row['baseword']]);
1895 }
1896
1897 reset($wl);
1898 while(list($key,$val)=each($wl)) {
1899 $insertFields = array(
1900 'wid' => $val['hash'],
1901 'baseword' => $key,
1902 'metaphone' => $val['metaphone']
1903 );
1904 // A duplicate-key error will occur here if a word is NOT unset in the unset() line. However as long as the words in $wl are NOT longer as 60 chars (the baseword varchar is 60 characters...) this is not a problem.
1905 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_words', $insertFields);
1906 }
1907 }
1908 }
1909 }
1910
1911 /**
1912 * Submits RELATIONS between words and phash
1913 *
1914 * @param array Word list array
1915 * @param integer phash value
1916 * @return void
1917 */
1918 function submitWords($wl,$phash) {
1919 $GLOBALS['TYPO3_DB']->exec_DELETEquery('index_rel', 'phash='.intval($phash));
1920
1921 foreach($wl as $val) {
1922 $insertFields = array(
1923 'phash' => $phash,
1924 'wid' => $val['hash'],
1925 'count' => $val['count'],
1926 'first' => $val['first'],
1927 'freq' => $this->freqMap(($val['count']/$this->wordcount)),
1928 'flags' => ($val['cmp'] & $this->flagBitMask)
1929 );
1930
1931 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_rel', $insertFields);
1932 }
1933 }
1934
1935 /**
1936 * maps frequency from a real number in [0;1] to an integer in [0;$this->freqRange] with anything above $this->freqMax as 1
1937 * and back.
1938 *
1939 * @param double Frequency
1940 * @return integer Frequency in range.
1941 */
1942 function freqMap($freq) {
1943 $mapFactor = $this->freqMax*100*$this->freqRange;
1944 if($freq<1) {
1945 $newFreq = $freq*$mapFactor;
1946 $newFreq = $newFreq>$this->freqRange?$this->freqRange:$newFreq;
1947 } else {
1948 $newFreq = $freq/$mapFactor;
1949 }
1950 return $newFreq;
1951
1952 }
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964 /********************************
1965 *
1966 * Hashing
1967 *
1968 *******************************/
1969
1970 /**
1971 * Get search hash, T3 pages
1972 *
1973 * @return void
1974 */
1975 function setT3Hashes() {
1976
1977 // Set main array:
1978 $hArray = array(
1979 'id' => (integer)$this->conf['id'],
1980 'type' => (integer)$this->conf['type'],
1981 'sys_lang' => (integer)$this->conf['sys_language_uid'],
1982 'MP' => (string)$this->conf['MP'],
1983 'cHash' => $this->cHashParams
1984 );
1985
1986 // Set grouping hash (Identifies a "page" combined of id, type, language, mountpoint and cHash parameters):
1987 $this->hash['phash_grouping'] = $this->md5inthash(serialize($hArray));
1988
1989 // Add gr_list and set plain phash (Subdivision where special page composition based on login is taken into account as well. It is expected that such pages are normally similar regardless of the login.)
1990 $hArray['gr_list'] = (string)$this->conf['gr_list'];
1991 $this->hash['phash'] = $this->md5inthash(serialize($hArray));
1992 }
1993
1994 /**
1995 * Get search hash, external files
1996 *
1997 * @param string File name / path which identifies it on the server
1998 * @param array Additional content identifying the (subpart of) content. For instance; PDF files are divided into groups of pages for indexing.
1999 * @return array Array with "phash_grouping" and "phash" inside.
2000 */
2001 function setExtHashes($file,$subinfo=array()) {
2002 // Set main array:
2003 $hash = array();
2004 $hArray = array(
2005 'file' => $file,
2006 );
2007
2008 // Set grouping hash:
2009 $hash['phash_grouping'] = $this->md5inthash(serialize($hArray));
2010
2011 // Add subinfo
2012 $hArray['subinfo'] = $subinfo;
2013 $hash['phash'] = $this->md5inthash(serialize($hArray));
2014
2015 return $hash;
2016 }
2017
2018 /**
2019 * md5 integer hash
2020 * Using 7 instead of 8 just because that makes the integers lower than 32 bit (28 bit) and so they do not interfere with UNSIGNED integers or PHP-versions which has varying output from the hexdec function.
2021 *
2022 * @param string String to hash
2023 * @return integer Integer intepretation of the md5 hash of input string.
2024 */
2025 function md5inthash($str) {
2026 return hexdec(substr(md5($str),0,7));
2027 }
2028
2029 /**
2030 * Calculates the cHash value of input GET array (for constructing cHash values if needed)
2031 *
2032 * @param array Array of GET parameters to encode
2033 * @return void
2034 * @deprecated since TYPO3 4.3, this function will be removed in TYPO3 4.5, use directly t3lib_div::calculateCHash()
2035 */
2036 function makeCHash($paramArray) {
2037 t3lib_div::logDeprecatedFunction();
2038
2039 $addQueryParams = t3lib_div::implodeArrayForUrl('', $paramArray);
2040
2041 $pA = t3lib_div::cHashParams($addQueryParams);
2042
2043 return t3lib_div::shortMD5(serialize($pA));
2044 }
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057 /*********************************
2058 *
2059 * Internal logging functions
2060 *
2061 *********************************/
2062
2063 /**
2064 * Push function wrapper for TT logging
2065 *
2066 * @param string Title to set
2067 * @param string Key (?)
2068 * @return void
2069 */
2070 function log_push($msg,$key) {
2071 if (is_object($GLOBALS['TT'])) $GLOBALS['TT']->push($msg,$key);
2072 }
2073
2074 /**
2075 * Pull function wrapper for TT logging
2076 *
2077 * @return void
2078 */
2079 function log_pull() {
2080 if (is_object($GLOBALS['TT'])) $GLOBALS['TT']->pull();
2081 }
2082
2083 /**
2084 * Set log message function wrapper for TT logging
2085 *
2086 * @param string Message to set
2087 * @param integer Error number
2088 * @return void
2089 */
2090 function log_setTSlogMessage($msg, $errorNum=0) {
2091 if (is_object($GLOBALS['TT'])) $GLOBALS['TT']->setTSlogMessage($msg,$errorNum);
2092 $this->internal_log[] = $msg;
2093 }
2094
2095
2096
2097
2098
2099
2100
2101
2102 /**************************
2103 *
2104 * tslib_fe hooks:
2105 *
2106 **************************/
2107
2108 /**
2109 * Frontend hook: If the page is not being re-generated this is our chance to force it to be (because re-generation of the page is required in order to have the indexer called!)
2110 *
2111 * @param array Parameters from frontend
2112 * @param object TSFE object (reference under PHP5)
2113 * @return void
2114 * @deprecated since TYPO3 4.3, this function will be removed in TYPO3 4.5, the method was extracted to hooks/class.tx_indexedsearch_tslib_fe_hook.php
2115 */
2116 function fe_headerNoCache(&$params, $ref) {
2117 t3lib_div::logDeprecatedFunction();
2118
2119 require_once t3lib_extMgm::extPath('indexed_search') . 'hooks/class.tx_indexedsearch_tslib_fe_hook.php';
2120 t3lib_div::makeInstance('tx_indexedsearch_tslib_fe_hook')->headerNoCache($params, $ref);
2121 }
2122
2123 /**
2124 * Makes sure that keywords are space-separated. This is impotant for their
2125 * proper displaying as a part of fulltext index.
2126 *
2127 * @param string $keywordList
2128 * @return string
2129 * @see http://bugs.typo3.org/view.php?id=1436
2130 */
2131 protected function addSpacesToKeywordList($keywordList) {
2132 $keywords = t3lib_div::trimExplode(',', $keywordList);
2133 return ' ' . implode(', ', $keywords) . ' ';
2134 }
2135 }
2136
2137
2138 if (defined('TYPO3_MODE') && $TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['ext/indexed_search/class.indexer.php']) {
2139 include_once($TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['ext/indexed_search/class.indexer.php']);
2140 }
2141 ?>