99405a6dc867762b95badc7cf7ffcf306f881576
2 /***************************************************************
5 * (c) 2001-2005 Kasper Skaarhoj (kasperYYYY@typo3.com)
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.
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.
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.
25 * This copyright notice MUST APPEAR in all copies of the script!
26 ***************************************************************/
28 * This class is a search indexer for TYPO3
30 * @author Kasper Skårhøj <kasperYYYY@typo3.com>
31 * Originally Christian Jul Jensen <christian@jul.net> helped as well.
34 * [CLASS/FUNCTION INDEX of SCRIPT]
38 * 141: class tx_indexedsearch_indexer
39 * 207: function hook_indexContent(&$pObj)
41 * SECTION: Backend API
42 * 308: function backend_initIndexer($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, $recordUid=0)
46 * SECTION: Initialization
47 * 416: function init()
48 * 468: function initializeExternalParsers()
50 * SECTION: Indexing; TYPO3 pages (HTML content)
51 * 509: function indexTypo3PageContent()
52 * 596: function splitHTMLContent($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($string)
60 * SECTION: Indexing; external URL
61 * 871: function indexExternalUrl($externalUrl)
62 * 902: function getUrlHeaders($url)
64 * SECTION: Indexing; external files (PDF, DOC, etc)
65 * 948: function indexRegularDocument($file, $force=FALSE, $contentTmpFile='', $altExtension='')
66 * 1054: function readFileContent($ext,$absFile,$cPKey)
67 * 1071: function fileContentParts($ext,$absFile)
68 * 1089: function splitRegularContent($content)
70 * SECTION: Analysing content, Extracting words
71 * 1122: function charsetEntity2utf8(&$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,$retRaw=FALSE)
80 * SECTION: SQL; TYPO3 Pages
81 * 1304: function submitPage()
82 * 1378: function submit_grlist($hash,$phash_x)
83 * 1398: function submit_section($hash,$hash_t3)
84 * 1416: function removeOldIndexedPages($phash)
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)
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(&$fieldArr)
103 * 1761: function removeLoginpagesWithContentHash()
104 * 1778: function includeCrawlerClass()
106 * SECTION: SQL; Submitting words
107 * 1805: function checkWordList($wl)
108 * 1842: function submitWords($wl,$phash)
109 * 1866: function freqMap($freq)
112 * 1899: function setT3Hashes()
113 * 1925: function setExtHashes($file,$subinfo=array())
114 * 1949: function md5inthash($str)
115 * 1959: function makeCHash($paramArray)
117 * SECTION: Internal logging functions
118 * 1991: function log_push($msg,$key)
119 * 2000: function log_pull()
120 * 2011: function log_setTSlogMessage($msg, $errorNum=0)
122 * SECTION: tslib_fe hooks:
123 * 2036: function fe_headerNoCache(&$params, $ref)
125 * TOTAL FUNCTIONS: 59
126 * (This index is automatically created/updated by the extension "extdeveval")
131 require_once(PATH_t3lib
.'class.t3lib_parsehtml.php');
135 * Indexing class for TYPO3 frontend
137 * @author Kasper Skaarhoj <kasperYYYY@typo3.com>
139 * @subpackage tx_indexedsearch
141 class tx_indexedsearch_indexer
{
144 var $reasons = array(
145 -1 => 'mtime matched the document, so no changes detected and no content updated',
146 -2 => 'The minimum age was not exceeded',
147 1 => "The configured max-age was exceeded for the document and thus it's indexed.",
148 2 => 'The minimum age was exceed and mtime was set and the mtime was different, so the page was indexed.',
149 3 => 'The minimum age was exceed, but mtime was not set, so the page was indexed.',
150 4 => 'Page has never been indexed (is not represented in the index_phash table).'
153 // HTML code blocks to exclude from indexing:
154 var $excludeSections = 'script,style';
156 // Supported Extensions for external files:
157 var $external_parsers = array(); // External parser objects, keys are file extension names. Values are objects with certain methods.
159 // Fe-group list (pages might be indexed separately for each usergroup combination to support search in access limited pages!)
160 var $defaultGrList = '0,-1';
163 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.
164 var $tstamp_minAge = 0; // If set, this tells a minimum limit before a document can be indexed again. This is regardless of mtime.
165 var $maxExternalFiles = 0; // Max number of external files to index.
167 var $forceIndexing = FALSE; // If true, indexing is forced despite of hashes etc.
168 var $crawlerActive = FALSE; // Set when crawler is detected (internal)
171 var $defaultContentArray=array(
178 var $externalFileCounter = 0;
180 var $conf = array(); // Configuration set internally (see init functions for required keys and their meaning)
181 var $indexerConfig = array(); // Indexer configuration, coming from $GLOBALS['TYPO3_CONF_VARS']['EXT']['extConf']['indexed_search']
182 var $hash = array(); // Hash array, contains phash and phash_grouping
183 var $file_phash_arr = array(); // Hash array for files
184 var $contentParts = array(); // Content of TYPO3 page
185 var $content_md5h = '';
186 var $internal_log = array(); // Internal log
187 var $indexExternalUrl_content = '';
189 var $cHashParams = array(); // cHashparams array
191 var $freqRange = 32000;
196 * Charset class object
203 * Metaphone object, if any
205 * @var user_DoubleMetaPhone
210 * Lexer object for word splitting
212 * @var tx_indexedsearch_lexer
219 * Parent Object (TSFE) Initialization
221 * @param object Parent Object (frontend TSFE object), passed by reference
224 function hook_indexContent(&$pObj) {
226 // Indexer configuration from Extension Manager interface:
227 $indexerConfig = unserialize($GLOBALS['TYPO3_CONF_VARS']['EXT']['extConf']['indexed_search']);
229 // Crawler activation:
230 // Requirements are that the crawler is loaded, a crawler session is running and re-indexing requested as processing instruction:
231 if (t3lib_extMgm
::isLoaded('crawler')
232 && $pObj->applicationData
['tx_crawler']['running']
233 && in_array('tx_indexedsearch_reindex', $pObj->applicationData
['tx_crawler']['parameters']['procInstructions'])) {
235 // Setting simple log message:
236 $pObj->applicationData
['tx_crawler']['log'][] = 'Forced Re-indexing enabled';
238 // Setting variables:
239 $this->crawlerActive
= TRUE; // Crawler active flag
240 $this->forceIndexing
= TRUE; // Force indexing despite timestamps etc.
243 // Determine if page should be indexed, and if so, configure and initialize indexer
244 if ($pObj->config
['config']['index_enable']) {
245 $this->log_push('Index page','');
247 if (!$indexerConfig['disableFrontendIndexing'] ||
$this->crawlerActive
) {
248 if (!$pObj->page
['no_search']) {
249 if (!$pObj->no_cache
) {
250 if (!strcmp($pObj->sys_language_uid
,$pObj->sys_language_content
)) {
252 // Setting up internal configuration from config array:
253 $this->conf
= array();
255 // Information about page for which the indexing takes place
256 $this->conf
['id'] = $pObj->id
; // Page id
257 $this->conf
['type'] = $pObj->type
; // Page type
258 $this->conf
['sys_language_uid'] = $pObj->sys_language_uid
; // sys_language UID of the language of the indexing.
259 $this->conf
['MP'] = $pObj->MP
; // MP variable, if any (Mount Points)
260 $this->conf
['gr_list'] = $pObj->gr_list
; // Group list
262 $this->conf
['cHash'] = $pObj->cHash
; // cHash string for additional parameters
263 $this->conf
['cHash_array'] = $pObj->cHash_array
; // Array of the additional parameters
265 $this->conf
['crdate'] = $pObj->page
['crdate']; // The creation date of the TYPO3 page
266 $this->conf
['page_cache_reg1'] = $pObj->page_cache_reg1
; // reg1 of the caching table. Not known what practical use this has.
269 $this->conf
['rootline_uids'] = array();
270 foreach($pObj->config
['rootLine'] as $rlkey => $rldat) {
271 $this->conf
['rootline_uids'][$rlkey] = $rldat['uid'];
275 $this->conf
['content'] = $pObj->content
; // Content string (HTML of TYPO3 page)
276 $this->conf
['indexedDocTitle'] = $pObj->convOutputCharset($pObj->indexedDocTitle
); // Alternative title for indexing
277 $this->conf
['metaCharset'] = $pObj->metaCharset
; // Character set of content (will be converted to utf-8 during indexing)
278 $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.
280 // Configuration of behavior:
281 $this->conf
['index_externals'] = $pObj->config
['config']['index_externals']; // Whether to index external documents like PDF, DOC etc. (if possible)
282 $this->conf
['index_descrLgd'] = $pObj->config
['config']['index_descrLgd']; // Length of description text (max 250, default 200)
285 $this->conf
['recordUid'] = 0;
286 $this->conf
['freeIndexUid'] = 0;
287 $this->conf
['freeIndexSetId'] = 0;
289 // Init and start indexing:
291 $this->indexTypo3PageContent();
292 } 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.');
293 } else $this->log_setTSlogMessage('Index page? No, page was set to "no_cache" and so cannot be indexed.');
294 } else $this->log_setTSlogMessage('Index page? No, The "No Search" flag has been set in the page properties!');
295 } else $this->log_setTSlogMessage('Index page? No, Ordinary Frontend indexing during rendering is disabled.');
307 /****************************
311 ****************************/
314 * Initializing the "combined ID" of the page (phash) being indexed (or for which external media is attached)
316 * @param integer The page uid, &id=
317 * @param integer The page type, &type=
318 * @param integer sys_language uid, typically &L=
319 * @param string The MP variable (Mount Points), &MP=
320 * @param array Rootline array of only UIDs.
321 * @param array Array of GET variables to register with this indexing
322 * @param boolean If set, calculates a cHash value from the $cHash_array. Probably you will not do that since such cases are indexed through the frontend and the idea of this interface is to index non-cachable pages from the backend!
325 function backend_initIndexer($id, $type, $sys_language_uid, $MP, $uidRL, $cHash_array=array(), $createCHash=FALSE) {
327 // Setting up internal configuration from config array:
328 $this->conf
= array();
330 // Information about page for which the indexing takes place
331 $this->conf
['id'] = $id; // Page id (integer)
332 $this->conf
['type'] = $type; // Page type (integer)
333 $this->conf
['sys_language_uid'] = $sys_language_uid; // sys_language UID of the language of the indexing (integer)
334 $this->conf
['MP'] = $MP; // MP variable, if any (Mount Points) (string)
335 $this->conf
['gr_list'] = '0,-1'; // Group list (hardcoded for now...)
338 $this->conf
['cHash'] = $createCHash ?
$this->makeCHash($cHash_array) : ''; // cHash string for additional parameters
339 $this->conf
['cHash_array'] = $cHash_array; // Array of the additional parameters
342 $this->conf
['freeIndexUid'] = 0;
343 $this->conf
['freeIndexSetId'] = 0;
344 $this->conf
['page_cache_reg1'] = '';
347 $this->conf
['rootline_uids'] = $uidRL;
349 // Configuration of behavior:
350 $this->conf
['index_externals'] = 1; // Whether to index external documents like PDF, DOC etc. (if possible)
351 $this->conf
['index_descrLgd'] = 200; // Length of description text (max 250, default 200)
353 // Init and start indexing:
358 * Sets the free-index uid. Can be called right after backend_initIndexer()
360 * @param integer Free index UID
361 * @param integer Set id - an integer identifying the "set" of indexing operations.
364 function backend_setFreeIndexUid($freeIndexUid, $freeIndexSetId=0) {
365 $this->conf
['freeIndexUid'] = $freeIndexUid;
366 $this->conf
['freeIndexSetId'] = $freeIndexSetId;
370 * Indexing records as the content of a TYPO3 page.
372 * @param string Title equivalent
373 * @param string Keywords equivalent
374 * @param string Description equivalent
375 * @param string The main content to index
376 * @param string The charset of the title, keyword, description and body-content. MUST BE VALID, otherwise nothing is indexed!
377 * @param integer Last modification time, in seconds
378 * @param integer The creation date of the content, in seconds
379 * @param integer The record UID that the content comes from (for registration with the indexed rows)
382 function backend_indexAsTYPO3Page($title, $keywords, $description, $content, $charset, $mtime, $crdate=0, $recordUid=0) {
385 $this->conf
['mtime'] = $mtime; // Most recent modification time (seconds) of the content
386 $this->conf
['crdate'] = $crdate; // The creation date of the TYPO3 content
387 $this->conf
['recordUid'] = $recordUid; // UID of the record, if applicable
389 // Construct fake HTML for parsing:
390 $this->conf
['content'] = '
393 <title>'.htmlspecialchars($title).'</title>
394 <meta name="keywords" content="'.htmlspecialchars($keywords).'" />
395 <meta name="description" content="'.htmlspecialchars($description).'" />
398 '.htmlspecialchars($content).'
400 </html>'; // Content string (HTML of TYPO3 page)
402 // Initializing charset:
403 $this->conf
['metaCharset'] = $charset; // Character set of content (will be converted to utf-8 during indexing)
404 $this->conf
['indexedDocTitle'] = ''; // Alternative title for indexing
406 // Index content as if it was a TYPO3 page:
407 $this->indexTypo3PageContent();
422 /********************************
426 *******************************/
429 * Initializes the object. $this->conf MUST be set with proper values prior to this call!!!
434 global $TYPO3_CONF_VARS;
437 $this->cHashParams
= $this->conf
['cHash_array'];
438 if (is_array($this->cHashParams
) && count($this->cHashParams
)) {
439 if ($this->conf
['cHash']) $this->cHashParams
['cHash'] = $this->conf
['cHash']; // Add this so that URL's come out right...
440 unset($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!!!
443 // Setting phash / phash_grouping which identifies the indexed page based on some of these variables:
444 $this->setT3Hashes();
446 // Indexer configuration from Extension Manager interface:
447 $this->indexerConfig
= unserialize($GLOBALS['TYPO3_CONF_VARS']['EXT']['extConf']['indexed_search']);
448 $this->tstamp_minAge
= t3lib_div
::intInRange($this->indexerConfig
['minAge']*3600,0);
449 $this->tstamp_maxAge
= t3lib_div
::intInRange($this->indexerConfig
['maxAge']*3600,0);
450 $this->maxExternalFiles
= t3lib_div
::intInRange($this->indexerConfig
['maxExternalFiles'],0,1000,5);
451 $this->flagBitMask
= t3lib_div
::intInRange($this->indexerConfig
['flagBitMask'],0,255);
453 // Initialize external document parsers:
454 // Example configuration, see ext_localconf.php of this file!
455 if ($this->conf
['index_externals']) {
456 $this->initializeExternalParsers();
459 // Initialize lexer (class that deconstructs the text into words):
460 // Example configuration (localconf.php) for this hook: $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['lexer'] = 'EXT:indexed_search/class.lexer.php:&tx_indexedsearch_lexer';
461 $lexerObjRef = $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['lexer'] ?
462 $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['lexer'] :
463 'EXT:indexed_search/class.lexer.php:&tx_indexedsearch_lexer';
464 $this->lexerObj
= &t3lib_div
::getUserObj($lexerObjRef);
465 $this->lexerObj
->debug
= $this->indexerConfig
['debugMode'];
467 // Initialize metaphone hook:
468 // Example configuration (localconf.php) for this hook: $TYPO3_CONF_VARS['EXTCONF']['indexed_search']['metaphone'] = 'EXT:indexed_search/class.doublemetaphone.php:&user_DoubleMetaPhone';
469 if ($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['metaphone']) {
470 $this->metaphoneObj
= &t3lib_div
::getUserObj($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['metaphone']);
471 $this->metaphoneObj
->pObj
= &$this;
474 // Init charset class:
475 $this->csObj
= &t3lib_div
::makeInstance('t3lib_cs');
479 * Initialize external parsers
485 function initializeExternalParsers() {
486 global $TYPO3_CONF_VARS;
488 if (is_array($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['external_parsers'])) {
489 foreach($TYPO3_CONF_VARS['EXTCONF']['indexed_search']['external_parsers'] as $extension => $_objRef) {
490 $this->external_parsers
[$extension] = &t3lib_div
::getUserObj($_objRef);
491 $this->external_parsers
[$extension]->pObj
= &$this;
493 // Init parser and if it returns false, unset its entry again:
494 if (!$this->external_parsers
[$extension]->initParser($extension)) {
495 unset($this->external_parsers
[$extension]);
515 /********************************
517 * Indexing; TYPO3 pages (HTML content)
519 *******************************/
522 * Start indexing of the TYPO3 page
526 function indexTypo3PageContent() {
528 $check = $this->checkMtimeTstamp($this->conf
['mtime'], $this->hash
['phash']);
529 $is_grlist = $this->is_grlist_set($this->hash
['phash']);
531 if ($check > 0 ||
!$is_grlist ||
$this->forceIndexing
) {
534 if ($this->forceIndexing
) {
535 $this->log_setTSlogMessage('Indexing needed, reason: Forced',1);
536 } elseif ($check > 0) {
537 $this->log_setTSlogMessage('Indexing needed, reason: '.$this->reasons
[$check],1);
539 $this->log_setTSlogMessage('Indexing needed, reason: Updates gr_list!',1);
542 // Divide into title,keywords,description and body:
543 $this->log_push('Split content','');
544 $this->contentParts
= $this->splitHTMLContent($this->conf
['content']);
545 if ($this->conf
['indexedDocTitle']) {
546 $this->contentParts
['title'] = $this->conf
['indexedDocTitle'];
550 // 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!)
551 $this->content_md5h
= $this->md5inthash(implode($this->contentParts
,''));
553 // This function checks if there is already a page (with gr_list = 0,-1) indexed and if that page has the very same contentHash.
554 // 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.
555 // 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.
556 $checkCHash = $this->checkContentHash();
557 if (!is_array($checkCHash) ||
$check===1) {
558 $Pstart=t3lib_div
::milliseconds();
560 $this->log_push('Converting charset of content ('.$this->conf
['metaCharset'].') to utf-8','');
561 $this->charsetEntity2utf8($this->contentParts
,$this->conf
['metaCharset']);
565 $this->log_push('Extract words from content','');
566 $splitInWords = $this->processWordsInArrays($this->contentParts
);
569 // Analyse the indexed words.
570 $this->log_push('Analyse the extracted words','');
571 $indexArr = $this->indexAnalyze($splitInWords);
574 // Submitting page (phash) record
575 $this->log_push('Submitting page','');
579 // Check words and submit to word list if not there
580 $this->log_push('Check word list and submit words','');
581 $this->checkWordList($indexArr);
582 $this->submitWords($indexArr,$this->hash
['phash']);
586 $this->updateParsetime($this->hash
['phash'],t3lib_div
::milliseconds()-$Pstart);
588 // Checking external files if configured for.
589 $this->log_push('Checking external files','');
590 if ($this->conf
['index_externals']) {
591 $this->extractLinks($this->conf
['content']);
595 $this->updateTstamp($this->hash
['phash'],$this->conf
['mtime']); // Update the timestatmp
596 $this->updateSetId($this->hash
['phash']);
597 $this->update_grlist($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.
598 $this->updateRootline();
599 $this->log_setTSlogMessage('Indexing not needed, the contentHash, '.$this->content_md5h
.', has not changed. Timestamp, grlist and rootline updated if necessary.');
602 $this->log_setTSlogMessage('Indexing not needed, reason: '.$this->reasons
[$check]);
607 * Splits HTML content and returns an associative array, with title, a list of metatags, and a list of words in the body.
609 * @param string HTML content to index. To some degree expected to be made by TYPO3 (ei. splitting the header by ":")
610 * @return array Array of content, having keys "title", "body", "keywords" and "description" set.
611 * @see splitRegularContent()
613 function splitHTMLContent($content) {
615 // divide head from body ( u-ouh :) )
616 $contentArr = $this->defaultContentArray
;
617 $contentArr['body'] = stristr($content,'<body');
618 $headPart = substr($content,0,-strlen($contentArr['body']));
621 $this->embracingTags($headPart,'TITLE',$contentArr['title'],$dummy2,$dummy);
622 $titleParts = explode(':',$contentArr['title'],2);
623 $contentArr['title'] = trim(isset($titleParts[1]) ?
$titleParts[1] : $titleParts[0]);
625 // get keywords and description metatags
626 for($i=0;$this->embracingTags($headPart,'meta',$dummy,$headPart,$meta[$i]);$i++
) { /*nothing*/ }
627 for($i=0;isset($meta[$i]);$i++
) {
628 $meta[$i] = t3lib_div
::get_tag_attributes($meta[$i]);
629 if(stristr($meta[$i]['name'],'keywords')) $contentArr['keywords'].=','.$meta[$i]['content'];
630 if(stristr($meta[$i]['name'],'description')) $contentArr['description'].=','.$meta[$i]['content'];
633 // Process <!--TYPO3SEARCH_begin--> or <!--TYPO3SEARCH_end--> tags:
634 $this->typoSearchTags($contentArr['body']);
636 // Get rid of unwanted sections (ie. scripting and style stuff) in body
637 $tagList = explode(',',$this->excludeSections
);
638 foreach($tagList as $tag) {
639 while($this->embracingTags($contentArr['body'],$tag,$dummy,$contentArr['body'],$dummy2));
642 // remove tags, but first make sure we don't concatenate words by doing it
643 $contentArr['body'] = str_replace('<',' <',$contentArr['body']);
644 $contentArr['body'] = trim(strip_tags($contentArr['body']));
646 $contentArr['keywords'] = trim($contentArr['keywords']);
647 $contentArr['description'] = trim($contentArr['description']);
654 * Extract the charset value from HTML meta tag.
656 * @param string HTML content
657 * @return string The charset value if found.
659 function getHTMLcharset($content) {
660 if (eregi('<meta[[:space:]]+[^>]*http-equiv[[:space:]]*=[[:space:]]*["\']CONTENT-TYPE["\'][^>]*>',$content,$reg)) {
661 if (eregi('charset[[:space:]]*=[[:space:]]*([[:alnum:]-]+)',$reg[0],$reg2)) {
668 * Converts a HTML document to utf-8
670 * @param string HTML content, any charset
671 * @param string Optional charset (otherwise extracted from HTML)
672 * @return string Converted HTML
674 function convertHTMLToUtf8($content,$charset='') {
677 $charset = $charset ?
$charset : $this->getHTMLcharset($content);
678 $charset = $this->csObj
->parse_charset($charset);
681 if ($charset && $charset!=='utf-8') {
682 $content = $this->csObj
->utf8_encode($content, $charset);
684 // Convert entities, assuming document is now UTF-8:
685 $content = $this->csObj
->entities_to_utf8($content, TRUE);
691 * Finds first occurence of embracing tags and returns the embraced content and the original string with
692 * the tag removed in the two passed variables. Returns false if no match found. ie. useful for finding
693 * <title> of document or removing <script>-sections
695 * @param string String to search in
696 * @param string Tag name, eg. "script"
697 * @param string Passed by reference: Content inside found tag
698 * @param string Passed by reference: Content after found tag
699 * @param string Passed by reference: Attributes of the found tag.
700 * @return boolean Returns false if tag was not found, otherwise true.
702 function embracingTags($string,$tagName,&$tagContent,&$stringAfter,&$paramList) {
703 $endTag = '</'.$tagName.'>';
704 $startTag = '<'.$tagName;
706 $isTagInText = stristr($string,$startTag); // stristr used because we want a case-insensitive search for the tag.
707 if(!$isTagInText) return false; // if the tag was not found, return false
709 list($paramList,$isTagInText) = explode('>',substr($isTagInText,strlen($startTag)),2);
710 $afterTagInText = stristr($isTagInText,$endTag);
711 if ($afterTagInText) {
712 $stringBefore = substr($string, 0, strpos(strtolower($string), strtolower($startTag)));
713 $tagContent = substr($isTagInText,0,strlen($isTagInText)-strlen($afterTagInText));
714 $stringAfter = $stringBefore.substr($afterTagInText,strlen($endTag));
715 } else { // If there was no ending tag, the tagContent is blank and anything after the tag it self is returned.
717 $stringAfter = $isTagInText;
724 * Removes content that shouldn't be indexed according to TYPO3SEARCH-tags.
726 * @param string HTML Content, passed by reference
727 * @return boolean Returns true if a TYPOSEARCH_ tag was found, otherwise false.
729 function typoSearchTags(&$body) {
730 $expBody = preg_split('/\<\!\-\-[\s]?TYPO3SEARCH_/',$body);
732 if(count($expBody)>1) {
735 foreach($expBody as $val) {
736 $part = explode('-->',$val,2);
737 if(trim($part[0])=='begin') {
740 } elseif(trim($part[0])=='end') {
753 * Extract links (hrefs) from HTML content and if indexable media is found, it is indexed.
755 * @param string HTML content
758 function extractLinks($content) {
761 $list = $this->extractHyperLinks($content);
763 if ($this->indexerConfig
['useCrawlerForExternalFiles'] && t3lib_extMgm
::isLoaded('crawler')) {
764 $this->includeCrawlerClass();
765 $crawler = t3lib_div
::makeInstance('tx_crawler_lib');
769 foreach($list as $linkInfo) {
772 if ($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!
773 $linkSource = t3lib_div
::htmlspecialchars_decode($linkInfo['localPath']);
775 $linkSource = t3lib_div
::htmlspecialchars_decode($linkInfo['href']);
779 $qParts = parse_url($linkSource);
781 // Check for jumpurl (TYPO3 specific thing...)
782 if ($qParts['query'] && strstr($qParts['query'],'jumpurl=')) {
783 parse_str($qParts['query'],$getP);
784 $linkSource = $getP['jumpurl'];
785 $qParts = parse_url($linkSource); // parse again due to new linkSource!
788 if ($qParts['scheme']) {
789 if ($this->indexerConfig
['indexExternalURLs']) {
790 // Index external URL (http or otherwise)
791 $this->indexExternalUrl($linkSource);
793 } elseif (!$qParts['query']) {
794 if (t3lib_div
::isAllowedAbsPath($linkSource)) {
795 $localFile = $linkSource;
797 $localFile = t3lib_div
::getFileAbsFileName(PATH_site
.$linkSource);
799 if ($localFile && @is_file
($localFile)) {
802 if ($linkInfo['localPath']) {
804 $fI = pathinfo($linkSource);
805 $ext = strtolower($fI['extension']);
806 if (is_object($crawler)) {
808 'document' => $linkSource,
809 'alturl' => $linkInfo['href'],
810 'conf' => $this->conf
812 unset($params['conf']['content']);
814 $crawler->addQueueEntry_callBack(0,$params,'EXT:indexed_search/class.crawler.php:&tx_indexedsearch_files',$this->conf
['id']);
815 $this->log_setTSlogMessage('media "'.$params['document'].'" added to "crawler" queue.',1);
817 $this->indexRegularDocument($linkInfo['href'], false, $linkSource, $ext);
820 if (is_object($crawler)) {
822 'document' => $linkSource,
823 'conf' => $this->conf
825 unset($params['conf']['content']);
826 $crawler->addQueueEntry_callBack(0,$params,'EXT:indexed_search/class.crawler.php:&tx_indexedsearch_files',$this->conf
['id']);
827 $this->log_setTSlogMessage('media "'.$params['document'].'" added to "crawler" queue.',1);
829 $this->indexRegularDocument($linkSource);
838 * Extracts all links to external documents from content string.
840 * @param string Content to analyse
841 * @return array Array of hyperlinks
842 * @see extractLinks()
844 function extractHyperLinks($string) {
845 if (!is_object($this->htmlParser
)) {
846 $this->htmlParser
= t3lib_div
::makeInstance('t3lib_parseHtml');
849 $parts = $this->htmlParser
->splitTags('a',$string);
851 foreach ($parts as $k => $v) {
853 $params = $this->htmlParser
->get_tag_attributes($v,1);
854 $firstTagName = $this->htmlParser
->getFirstTagName($v); // The 'name' of the first tag
856 switch (strtolower($firstTagName)) {
858 $src = $params[0]['href'];
860 // Check if a local path to that file has been set - useful if you are using a download script.
861 $md5 = t3lib_div
::shortMD5($src);
862 if (is_array($indexLocalFiles=$GLOBALS['T3_VAR']['ext']['indexed_search']['indexLocalFiles'])) {
863 $localPath = isset($indexLocalFiles[$md5]) ?
$indexLocalFiles[$md5] : '';
864 } else $localPath=false;
868 'href' => $params[0]['href'],
869 'localPath' => $localPath
890 /******************************************
892 * Indexing; external URL
894 ******************************************/
897 * Index External URLs HTML content
899 * @param string URL, eg. "http://typo3.org/"
901 * @see indexRegularDocument()
903 function indexExternalUrl($externalUrl) {
905 // Parse External URL:
906 $qParts = parse_url($externalUrl);
907 $fI = pathinfo($qParts['path']);
908 $ext = strtolower($fI['extension']);
911 $urlHeaders = $this->getUrlHeaders($externalUrl);
912 if (stristr($urlHeaders['Content-Type'],'text/html')) {
913 $content = $this->indexExternalUrl_content
= t3lib_div
::getUrl($externalUrl);
914 if (strlen($content)) {
916 // Create temporary file:
917 $tmpFile = t3lib_div
::tempnam('EXTERNAL_URL').'.html';
918 t3lib_div
::writeFile($tmpFile, $content);
921 $this->indexRegularDocument($externalUrl, TRUE, $tmpFile, 'html'); // Using "TRUE" for second parameter to force indexing of external URLs (mtime doesn't make sense, does it?)
928 * Getting HTTP request headers of URL
930 * @param string The URL
931 * @param integer Timeout (seconds?)
932 * @return mixed If no answer, returns false. Otherwise an array where HTTP headers are keys
934 function getUrlHeaders($url) {
935 $content = t3lib_div
::getURL($url,2); // Try to get the headers only
937 if (strlen($content)) {
939 $headers = t3lib_div
::trimExplode(chr(10),$content,1);
941 foreach($headers as $line) {
942 if (!strlen(trim($line))) {
943 break; // Stop at the first empty line (= end of header)
946 list($headKey, $headValue) = explode(':', $line, 2);
947 $retVal[$headKey] = $headValue;
965 /******************************************
967 * Indexing; external files (PDF, DOC, etc)
969 ******************************************/
972 * Indexing a regular document given as $file (relative to PATH_site, local file)
974 * @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
975 * @param boolean If set, indexing is forced (despite content hashes, mtime etc).
976 * @param string Temporary file with the content to read it from (instead of $file). Used when the $file is a URL.
977 * @param string File extension for temporary file.
980 function indexRegularDocument($file, $force=FALSE, $contentTmpFile='', $altExtension='') {
983 $fI = pathinfo($file);
984 $ext = $altExtension ?
$altExtension : strtolower($fI['extension']);
987 if (!$contentTmpFile) {
988 if (!t3lib_div
::isAbsPath($file)) { // Relative, prepend PATH_site:
989 $absFile = t3lib_div
::getFileAbsFileName(PATH_site
.$file);
990 } else { // Absolute, pass-through:
993 $absFile = t3lib_div
::isAllowedAbsPath($absFile) ?
$absFile : '';
995 $absFile = $contentTmpFile;
998 // Indexing the document:
999 if ($absFile && @is_file
($absFile)) {
1000 if ($this->external_parsers
[$ext]) {
1001 $mtime = filemtime($absFile);
1002 $cParts = $this->fileContentParts($ext,$absFile);
1004 foreach($cParts as $cPKey) {
1005 $this->internal_log
= array();
1006 $this->log_push('Index: '.str_replace('.','_',basename($file)).($cPKey?
'#'.$cPKey:''),'');
1007 $Pstart = t3lib_div
::milliseconds();
1008 $subinfo = array('key' => $cPKey); // Setting page range. This is "0" (zero) when no division is made, otherwise a range like "1-3"
1009 $phash_arr = $this->file_phash_arr
= $this->setExtHashes($file,$subinfo);
1010 $check = $this->checkMtimeTstamp($mtime, $phash_arr['phash']);
1011 if ($check > 0 ||
$force) {
1013 $this->log_setTSlogMessage('Indexing needed, reason: '.$this->reasons
[$check],1);
1015 $this->log_setTSlogMessage('Indexing forced by flag',1);
1018 // Check external file counter:
1019 if ($this->externalFileCounter
< $this->maxExternalFiles ||
$force) {
1021 // Divide into title,keywords,description and body:
1022 $this->log_push('Split content','');
1023 $contentParts = $this->readFileContent($ext,$absFile,$cPKey);
1026 if (is_array($contentParts)) {
1027 // Calculating a hash over what is to be the actual content. (see indexTypo3PageContent())
1028 $content_md5h = $this->md5inthash(implode($contentParts,''));
1030 if ($this->checkExternalDocContentHash($phash_arr['phash_grouping'], $content_md5h) ||
$force) {
1032 // Increment counter:
1033 $this->externalFileCounter++
;
1036 $this->log_push('Extract words from content','');
1037 $splitInWords = $this->processWordsInArrays($contentParts);
1040 // Analyse the indexed words.
1041 $this->log_push('Analyse the extracted words','');
1042 $indexArr = $this->indexAnalyze($splitInWords);
1045 // Submitting page (phash) record
1046 $this->log_push('Submitting page','');
1047 $size = filesize($absFile);
1048 $ctime = filemtime($absFile); // Unfortunately I cannot determine WHEN a file is originally made - so I must return the modification time...
1049 $this->submitFilePage($phash_arr,$file,$subinfo,$ext,$mtime,$ctime,$size,$content_md5h,$contentParts);
1052 // Check words and submit to word list if not there
1053 $this->log_push('Check word list and submit words','');
1054 $this->checkWordList($indexArr);
1055 $this->submitWords($indexArr,$phash_arr['phash']);
1059 $this->updateParsetime($phash_arr['phash'],t3lib_div
::milliseconds()-$Pstart);
1061 $this->updateTstamp($phash_arr['phash'],$mtime); // Update the timestamp
1062 $this->log_setTSlogMessage('Indexing not needed, the contentHash, '.$content_md5h.', has not changed. Timestamp updated.');
1064 } else $this->log_setTSlogMessage('Could not index file! Unsupported extension.');
1065 } else $this->log_setTSlogMessage('The limit of '.$this->maxExternalFiles
.' has already been exceeded, so no indexing will take place this time.');
1066 } else $this->log_setTSlogMessage('Indexing not needed, reason: '.$this->reasons
[$check]);
1068 // Checking and setting sections:
1069 # $this->submitFile_grlist($phash_arr['phash']); // Setting a gr_list record if there is none already (set for default fe_group)
1070 $this->submitFile_section($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.
1073 } else $this->log_setTSlogMessage('Indexing not possible; The extension "'.$ext.'" was not supported.');
1074 } else $this->log_setTSlogMessage('Indexing not possible; File "'.$absFile.'" not found or valid.');
1078 * Reads the content of an external file being indexed.
1079 * The content from the external parser MUST be returned in utf-8!
1081 * @param string File extension, eg. "pdf", "doc" etc.
1082 * @param string Absolute filename of file (must exist and be validated OK before calling function)
1083 * @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.)
1084 * @return array Standard content array (title, description, keywords, body keys)
1086 function readFileContent($ext,$absFile,$cPKey) {
1088 // Consult relevant external document parser:
1089 if (is_object($this->external_parsers
[$ext])) {
1090 $contentArr = $this->external_parsers
[$ext]->readFileContent($ext,$absFile,$cPKey);
1097 * Creates an array with pointers to divisions of document.
1099 * @param string File extension
1100 * @param string Absolute filename (must exist and be validated OK before calling function)
1101 * @return array Array of pointers to sections that the document should be divided into
1103 function fileContentParts($ext,$absFile) {
1106 // Consult relevant external document parser:
1107 if (is_object($this->external_parsers
[$ext])) {
1108 $cParts = $this->external_parsers
[$ext]->fileContentParts($ext,$absFile);
1115 * Splits non-HTML content (from external files for instance)
1117 * @param string Input content (non-HTML) to index.
1118 * @return array Array of content, having the key "body" set (plus "title", "description" and "keywords", but empty)
1119 * @see splitHTMLContent()
1121 function splitRegularContent($content) {
1122 $contentArr = $this->defaultContentArray
;
1123 $contentArr['body'] = $content;
1141 /**********************************
1143 * Analysing content, Extracting words
1145 **********************************/
1148 * Convert character set and HTML entities in the value of input content array keys
1150 * @param array Standard content array
1151 * @param string Charset of the input content (converted to utf-8)
1154 function charsetEntity2utf8(&$contentArr, $charset) {
1156 // Convert charset if necessary
1158 while(list($key,)=each($contentArr)) {
1159 if (strlen($contentArr[$key])) {
1161 if ($charset!=='utf-8') {
1162 $contentArr[$key] = $this->csObj
->utf8_encode($contentArr[$key], $charset);
1165 // decode all numeric / html-entities in the string to real characters:
1166 $contentArr[$key] = $this->csObj
->entities_to_utf8($contentArr[$key],TRUE);
1172 * Processing words in the array from split*Content -functions
1174 * @param array Array of content to index, see splitHTMLContent() and splitRegularContent()
1175 * @return array Content input array modified so each key is not a unique array of words
1177 function processWordsInArrays($contentArr) {
1179 // split all parts to words
1181 while(list($key,)=each($contentArr)) {
1182 $contentArr[$key] = $this->lexerObj
->split2Words($contentArr[$key]);
1185 // For title, keywords, and description we don't want duplicates:
1186 $contentArr['title'] = array_unique($contentArr['title']);
1187 $contentArr['keywords'] = array_unique($contentArr['keywords']);
1188 $contentArr['description'] = array_unique($contentArr['description']);
1190 // Return modified array:
1195 * Processing words in the array from split*Content -functions
1196 * This function is only a wrapper because the function has been removed (see above).
1198 * @param array Array of content to index, see splitHTMLContent() and splitRegularContent()
1199 * @return array Content input array modified so each key is not a unique array of words
1202 function procesWordsInArrays($contentArr) {
1203 return $this->processWordsInArrays($contentArr);
1207 * Extracts the sample description text from the content array.
1209 * @param array Content array
1210 * @return string Description string
1212 function bodyDescription($contentArr) {
1214 // Setting description
1215 $maxL = t3lib_div
::intInRange($this->conf
['index_descrLgd'],0,255,200);
1217 // Takes the quadruple lenght first, because whitespace and entities may be removed and thus shorten the string more yet.
1218 # $bodyDescription = implode(' ',split('[[:space:],]+',substr(trim($contentArr['body']),0,$maxL*4)));
1219 $bodyDescription = str_replace(array(' ',"\t","\r","\n"),' ',$contentArr['body']);
1221 // Shorten the string:
1222 $bodyDescription = $this->csObj
->strtrunc('utf-8', $bodyDescription, $maxL);
1225 return $bodyDescription;
1229 * Analyzes content to use for indexing,
1231 * @param array Standard content array: an array with the keys title,keywords,description and body, which all contain an array of words.
1232 * @return array Index Array (whatever that is...)
1234 function indexAnalyze($content) {
1235 $indexArr = Array();
1238 $this->analyzeHeaderinfo($indexArr,$content,'title',7);
1239 $this->analyzeHeaderinfo($indexArr,$content,'keywords',6);
1240 $this->analyzeHeaderinfo($indexArr,$content,'description',5);
1241 $this->analyzeBody($indexArr,$content);
1247 * Calculates relevant information for headercontent
1249 * @param array Index array, passed by reference
1250 * @param array Standard content array
1251 * @param string Key from standard content array
1252 * @param integer Bit-wise priority to type
1255 function analyzeHeaderinfo(&$retArr,$content,$key,$offset) {
1256 reset($content[$key]);
1257 while(list(,$val)=each($content[$key])) {
1258 $val = substr($val,0,60); // Max 60 - because the baseword varchar IS 60. This MUST be the same.
1259 $retArr[$val]['cmp'] = $retArr[$val]['cmp']|
pow(2,$offset);
1260 $retArr[$val]['count'] = $retArr[$val]['count']+
1;
1261 $retArr[$val]['hash'] = hexdec(substr(md5($val),0,7));
1262 $retArr[$val]['metaphone'] = $this->metaphone($val);
1268 * Calculates relevant information for bodycontent
1270 * @param array Index array, passed by reference
1271 * @param array Standard content array
1274 function analyzeBody(&$retArr,$content) {
1275 foreach($content['body'] as $key => $val) {
1276 $val = substr($val,0,60); // Max 60 - because the baseword varchar IS 60. This MUST be the same.
1277 if(!isset($retArr[$val])) {
1278 $retArr[$val]['first'] = $key;
1279 $retArr[$val]['hash'] = hexdec(substr(md5($val),0,7));
1280 $retArr[$val]['metaphone'] = $this->metaphone($val);
1282 $retArr[$val]['count'] = $retArr[$val]['count']+
1;
1288 * Creating metaphone based hash from input word
1290 * @param string Word to convert
1291 * @param boolean If set, returns the raw metaphone value (not hashed)
1292 * @return mixed Metaphone hash integer (or raw value, string)
1294 function metaphone($word,$retRaw=FALSE) {
1296 if (is_object($this->metaphoneObj
)) {
1297 $tmp = $this->metaphoneObj
->metaphone($word, $this->conf
['sys_language_uid']);
1299 $tmp = metaphone($word);
1302 // Return raw value?
1303 if ($retRaw) return $tmp;
1305 // Otherwise create hash and return integer
1306 if($tmp=='') $ret=0; else $ret=hexdec(substr(md5($tmp),0,7));
1325 /********************************
1329 *******************************/
1332 * Updates db with information about the page (TYPO3 page, not external media)
1336 function submitPage() {
1338 // Remove any current data for this phash:
1339 $this->removeOldIndexedPages($this->hash
['phash']);
1341 // setting new phash_row
1343 'phash' => $this->hash
['phash'],
1344 'phash_grouping' => $this->hash
['phash_grouping'],
1345 'cHashParams' => serialize($this->cHashParams
),
1346 'contentHash' => $this->content_md5h
,
1347 'data_page_id' => $this->conf
['id'],
1348 'data_page_reg1' => $this->conf
['page_cache_reg1'],
1349 'data_page_type' => $this->conf
['type'],
1350 'data_page_mp' => $this->conf
['MP'],
1351 'gr_list' => $this->conf
['gr_list'],
1352 'item_type' => 0, // TYPO3 page
1353 'item_title' => $this->contentParts
['title'],
1354 'item_description' => $this->bodyDescription($this->contentParts
),
1355 'item_mtime' => $this->conf
['mtime'],
1356 'item_size' => strlen($this->conf
['content']),
1359 'item_crdate' => $this->conf
['crdate'], // Creation date of page
1360 'sys_language_uid' => $this->conf
['sys_language_uid'], // Sys language uid of the page. Should reflect which language it DOES actually display!
1362 'recordUid' => intval($this->conf
['recordUid']),
1363 'freeIndexUid' => intval($this->conf
['freeIndexUid']),
1364 'freeIndexSetId' => intval($this->conf
['freeIndexSetId']),
1367 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_phash', $fields);
1369 // PROCESSING index_section
1370 $this->submit_section($this->hash
['phash'],$this->hash
['phash']);
1372 // PROCESSING index_grlist
1373 $this->submit_grlist($this->hash
['phash'],$this->hash
['phash']);
1375 // PROCESSING index_fulltext
1377 'phash' => $this->hash
['phash'],
1378 'fulltextdata' => implode(' ', $this->contentParts
)
1380 if ($this->indexerConfig
['fullTextDataLength']>0) {
1381 $fields['fulltextdata'] = substr($fields['fulltextdata'],0,$this->indexerConfig
['fullTextDataLength']);
1383 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_fulltext', $fields);
1385 // PROCESSING index_debug
1386 if ($this->indexerConfig
['debugMode']) {
1388 'phash' => $this->hash
['phash'],
1389 'debuginfo' => serialize(array(
1390 'cHashParams' => $this->cHashParams
,
1391 'external_parsers initialized' => array_keys($this->external_parsers
),
1392 'conf' => array_merge($this->conf
,array('content'=>substr($this->conf
['content'],0,1000))),
1393 'contentParts' => array_merge($this->contentParts
,array('body' => substr($this->contentParts
['body'],0,1000))),
1394 'logs' => $this->internal_log
,
1395 'lexer' => $this->lexerObj
->debugString
,
1398 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_debug', $fields);
1403 * Stores gr_list in the database.
1405 * @param integer Search result record phash
1406 * @param integer Actual phash of current content
1408 * @see update_grlist()
1410 function submit_grlist($hash,$phash_x) {
1412 // Setting the gr_list record
1415 'phash_x' => $phash_x,
1416 'hash_gr_list' => $this->md5inthash($this->conf
['gr_list']),
1417 'gr_list' => $this->conf
['gr_list']
1419 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_grlist', $fields);
1424 * $hash and $hash_t3 are the same for TYPO3 pages, but different when it is external files.
1426 * @param integer phash of TYPO3 parent search result record
1427 * @param integer phash of the file indexation search record
1430 function submit_section($hash,$hash_t3) {
1433 'phash_t3' => $hash_t3,
1434 'page_id' => intval($this->conf
['id'])
1437 $this->getRootLineFields($fields);
1439 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_section', $fields);
1443 * Removes records for the indexed page, $phash
1445 * @param integer phash value to flush
1448 function removeOldIndexedPages($phash) {
1449 // Removing old registrations for all tables. Because the pages are TYPO3 pages there can be nothing else than 1-1 relations here.
1450 $tableArr = explode(',','index_phash,index_section,index_grlist,index_fulltext,index_debug');
1451 foreach($tableArr as $table) {
1452 $GLOBALS['TYPO3_DB']->exec_DELETEquery($table, 'phash='.intval($phash));
1454 // Removing all index_section records with hash_t3 set to this hash (this includes such records set for external media on the page as well!). The re-insert of these records are done in indexRegularDocument($file).
1455 $GLOBALS['TYPO3_DB']->exec_DELETEquery('index_section', 'phash_t3='.intval($phash));
1470 /********************************
1472 * SQL; External media
1474 *******************************/
1478 * Updates db with information about the file
1480 * @param array Array with phash and phash_grouping keys for file
1481 * @param string File name
1482 * @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)
1483 * @param string File extension determining the type of media.
1484 * @param integer Modification time of file.
1485 * @param integer Creation time of file.
1486 * @param integer Size of file in bytes
1487 * @param integer Content HASH value.
1488 * @param array Standard content array (using only title and body for a file)
1491 function submitFilePage($hash,$file,$subinfo,$ext,$mtime,$ctime,$size,$content_md5h,$contentParts) {
1494 $storeItemType = $this->external_parsers
[$ext]->ext2itemtype_map
[$ext];
1495 $storeItemType = $storeItemType ?
$storeItemType : $ext;
1497 // Remove any current data for this phash:
1498 $this->removeOldIndexedFiles($hash['phash']);
1501 $fileParts = parse_url($file);
1505 'phash' => $hash['phash'],
1506 'phash_grouping' => $hash['phash_grouping'],
1507 'cHashParams' => serialize($subinfo),
1508 'contentHash' => $content_md5h,
1509 'data_filename' => $file,
1510 'item_type' => $storeItemType,
1511 'item_title' => trim($contentParts['title']) ?
$contentParts['title'] : basename($file),
1512 'item_description' => $this->bodyDescription($contentParts),
1513 'item_mtime' => $mtime,
1514 'item_size' => $size,
1515 'item_crdate' => $ctime,
1518 'gr_list' => $this->conf
['gr_list'],
1519 'externalUrl' => $fileParts['scheme'] ?
1 : 0,
1520 'recordUid' => intval($this->conf
['recordUid']),
1521 'freeIndexUid' => intval($this->conf
['freeIndexUid']),
1522 'freeIndexSetId' => intval($this->conf
['freeIndexSetId']),
1524 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_phash', $fields);
1526 // PROCESSING index_fulltext
1528 'phash' => $hash['phash'],
1529 'fulltextdata' => implode(' ', $contentParts)
1531 if ($this->indexerConfig
['fullTextDataLength']>0) {
1532 $fields['fulltextdata'] = substr($fields['fulltextdata'],0,$this->indexerConfig
['fullTextDataLength']);
1534 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_fulltext', $fields);
1536 // PROCESSING index_debug
1537 if ($this->indexerConfig
['debugMode']) {
1539 'phash' => $hash['phash'],
1540 'debuginfo' => serialize(array(
1541 'cHashParams' => $subinfo,
1542 'contentParts' => array_merge($contentParts,array('body' => substr($contentParts['body'],0,1000))),
1543 'logs' => $this->internal_log
,
1544 'lexer' => $this->lexerObj
->debugString
,
1547 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_debug', $fields);
1552 * Stores file gr_list for a file IF it does not exist already
1554 * @param integer phash value of file
1557 function submitFile_grlist($hash) {
1558 // 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.
1559 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_grlist', 'phash='.intval($hash).' AND (hash_gr_list='.$this->md5inthash($this->defaultGrList
).' OR hash_gr_list='.$this->md5inthash($this->conf
['gr_list']).')');
1560 if (!$GLOBALS['TYPO3_DB']->sql_num_rows($res)) {
1561 $this->submit_grlist($hash,$hash);
1566 * Stores file section for a file IF it does not exist
1568 * @param integer phash value of file
1571 function submitFile_section($hash) {
1572 // Testing if there is a section
1573 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_section', 'phash='.intval($hash).' AND page_id='.intval($this->conf
['id']));
1574 if (!$GLOBALS['TYPO3_DB']->sql_num_rows($res)) {
1575 $this->submit_section($hash,$this->hash
['phash']);
1580 * Removes records for the indexed page, $phash
1582 * @param integer phash value to flush
1585 function removeOldIndexedFiles($phash) {
1587 // Removing old registrations for tables.
1588 $tableArr = explode(',','index_phash,index_grlist,index_fulltext,index_debug');
1589 foreach($tableArr as $table) {
1590 $GLOBALS['TYPO3_DB']->exec_DELETEquery($table, 'phash='.intval($phash));
1607 /********************************
1609 * SQL Helper functions
1611 *******************************/
1614 * Check the mtime / tstamp of the currently indexed page/file (based on phash)
1615 * Return positive integer if the page needs to be indexed
1617 * @param integer mtime value to test against limits and indexed page (usually this is the mtime of the cached document)
1618 * @param integer "phash" used to select any already indexed page to see what its mtime is.
1619 * @return integer Result integer: Generally: <0 = No indexing, >0 = Do indexing (see $this->reasons): -2) Min age was NOT exceeded and so indexing cannot occur. -1) mtime matched so no need to reindex page. 0) N/A 1) Max age exceeded, page must be indexed again. 2) mtime of indexed page doesn't match mtime given for current content and we must index page. 3) No mtime was set, so we will index... 4) No indexed page found, so of course we will index.
1621 function checkMtimeTstamp($mtime,$phash) {
1623 // Select indexed page:
1624 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('item_mtime,tstamp', 'index_phash', 'phash='.intval($phash));
1627 // If there was an indexing of the page...:
1628 if ($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1629 if ($this->tstamp_maxAge
&& ($row['tstamp']+
$this->tstamp_maxAge
) < time()) { // If max age is exceeded, index the page
1630 $out = 1; // The configured max-age was exceeded for the document and thus it's indexed.
1632 if (!$this->tstamp_minAge ||
($row['tstamp']+
$this->tstamp_minAge
)<time()) { // if minAge is not set or if minAge is exceeded, consider at mtime
1633 if ($mtime) { // It mtime is set, then it's tested. If not, the page must clearly be indexed.
1634 if ($row['item_mtime'] != $mtime) { // And if mtime is different from the index_phash mtime, it's about time to re-index.
1635 $out = 2; // The minimum age was exceed and mtime was set and the mtime was different, so the page was indexed.
1637 $out = -1; // mtime matched the document, so no changes detected and no content updated
1638 if ($this->tstamp_maxAge
) {
1639 $this->log_setTSlogMessage('mtime matched, timestamp NOT updated because a maxAge is set ('.($row['tstamp'] +
$this->tstamp_maxAge
- time()).' seconds to expire time).',1);
1641 $this->updateTstamp($phash); // Update the timestatmp
1642 $this->log_setTSlogMessage('mtime matched, timestamp updated.',1);
1645 } else {$out = 3; } // The minimum age was exceed, but mtime was not set, so the page was indexed.
1646 } else {$out = -2;} // The minimum age was not exceeded
1648 } else {$out = 4;} // Page has never been indexed (is not represented in the index_phash table).
1653 * Check content hash in phash table
1655 * @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!
1657 function checkContentHash() {
1658 // With this query the page will only be indexed if it's content is different from the same "phash_grouping" -page.
1659 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_phash A', 'A.phash_grouping='.intval($this->hash
['phash_grouping']).' AND A.contentHash='.intval($this->content_md5h
));
1660 if ($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1667 * Check content hash for external documents
1668 * Returns true if the document needs to be indexed (that is, there was no result)
1670 * @param integer phash value to check (phash_grouping)
1671 * @param integer Content hash to check
1672 * @return boolean Returns true if the document needs to be indexed (that is, there was no result)
1674 function checkExternalDocContentHash($hashGr,$content_md5h) {
1675 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('*', 'index_phash A', 'A.phash_grouping='.intval($hashGr).' AND A.contentHash='.intval($content_md5h));
1676 if ($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1683 * 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)
1685 * @param integer Phash integer to test.
1688 function is_grlist_set($phash_x) {
1689 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash_x', 'index_grlist', 'phash_x='.intval($phash_x));
1690 return $GLOBALS['TYPO3_DB']->sql_num_rows($res);
1694 * Check if an grlist-entry for this hash exists and if not so, write one.
1696 * @param integer phash of the search result that should be found
1697 * @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...
1699 * @see submit_grlist()
1701 function update_grlist($phash,$phash_x) {
1702 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('phash', 'index_grlist', 'phash='.intval($phash).' AND hash_gr_list='.$this->md5inthash($this->conf
['gr_list']));
1703 if (!$GLOBALS['TYPO3_DB']->sql_num_rows($res)) {
1704 $this->submit_grlist($phash,$phash_x);
1705 $this->log_setTSlogMessage("Inserted gr_list '".$this->conf
['gr_list']."' for phash '".$phash."'",1);
1710 * Update tstamp for a phash row.
1712 * @param integer phash value
1713 * @param integer If set, update the mtime field to this value.
1716 function updateTstamp($phash,$mtime=0) {
1717 $updateFields = array(
1720 if ($mtime) { $updateFields['item_mtime'] = intval($mtime); }
1722 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_phash', 'phash='.intval($phash), $updateFields);
1726 * Update SetID of the index_phash record.
1728 * @param integer phash value
1731 function updateSetId($phash) {
1732 $updateFields = array(
1733 'freeIndexSetId' => intval($this->conf
['freeIndexSetId'])
1736 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_phash', 'phash='.intval($phash), $updateFields);
1740 * Update parsetime for phash row.
1742 * @param integer phash value.
1743 * @param integer Parsetime value to set.
1746 function updateParsetime($phash,$parsetime) {
1747 $updateFields = array(
1748 'parsetime' => intval($parsetime)
1751 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_phash', 'phash='.intval($phash), $updateFields);
1755 * Update section rootline for the page
1759 function updateRootline() {
1761 $updateFields = array();
1762 $this->getRootLineFields($updateFields);
1764 $GLOBALS['TYPO3_DB']->exec_UPDATEquery('index_section', 'page_id='.intval($this->conf
['id']), $updateFields);
1768 * Adding values for root-line fields.
1769 * rl0, rl1 and rl2 are standard. A hook might add more.
1771 * @param array Field array, passed by reference
1774 function getRootLineFields(&$fieldArr) {
1776 $fieldArr['rl0'] = intval($this->conf
['rootline_uids'][0]);
1777 $fieldArr['rl1'] = intval($this->conf
['rootline_uids'][1]);
1778 $fieldArr['rl2'] = intval($this->conf
['rootline_uids'][2]);
1780 if (is_array($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['indexed_search']['addRootLineFields'])) {
1781 foreach($GLOBALS['TYPO3_CONF_VARS']['EXTCONF']['indexed_search']['addRootLineFields'] as $fieldName => $rootLineLevel) {
1782 $fieldArr[$fieldName] = intval($this->conf
['rootline_uids'][$rootLineLevel]);
1788 * Removes any indexed pages with userlogins which has the same contentHash
1789 * NOT USED anywhere inside this class!
1793 function removeLoginpagesWithContentHash() {
1794 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('*', 'index_phash A,index_grlist B', '
1796 AND A.phash_grouping='.intval($this->hash
['phash_grouping']).'
1797 AND B.hash_gr_list!='.$this->md5inthash($this->defaultGrList
).'
1798 AND A.contentHash='.intval($this->content_md5h
));
1799 while($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1800 $this->log_setTSlogMessage("The currently indexed page was indexed under no user-login and apparently this page has been indexed under login conditions earlier, but with the SAME content. Therefore the old similar page with phash='".$row['phash']."' are now removed.",1);
1801 $this->removeOldIndexedPages($row['phash']);
1806 * Includes the crawler class
1810 function includeCrawlerClass() {
1811 global $TYPO3_CONF_VARS;
1813 require_once(t3lib_extMgm
::extPath('crawler').'class.tx_crawler_lib.php');
1825 /********************************
1827 * SQL; Submitting words
1829 *******************************/
1832 * Adds new words to db
1834 * @param array Word List array (where each word has information about position etc).
1837 function checkWordList($wl) {
1839 $phashArr = array();
1840 while(list($key,) = each($wl)) {
1841 $phashArr[] = $wl[$key]['hash'];
1843 if (count($phashArr)) {
1844 $cwl = implode(',',$phashArr);
1845 $res = $GLOBALS['TYPO3_DB']->exec_SELECTquery('baseword', 'index_words', 'wid IN ('.$cwl.')');
1847 if($GLOBALS['TYPO3_DB']->sql_num_rows($res)!=count($wl)) {
1848 $this->log_setTSlogMessage('Inserting words: '.(count($wl)-$GLOBALS['TYPO3_DB']->sql_num_rows($res)),1);
1849 while($row = $GLOBALS['TYPO3_DB']->sql_fetch_assoc($res)) {
1850 unset($wl[$row['baseword']]);
1854 while(list($key,$val)=each($wl)) {
1855 $insertFields = array(
1856 'wid' => $val['hash'],
1858 'metaphone' => $val['metaphone']
1860 // 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.
1861 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_words', $insertFields);
1868 * Submits RELATIONS between words and phash
1870 * @param array Word list array
1871 * @param integer phash value
1874 function submitWords($wl,$phash) {
1875 $GLOBALS['TYPO3_DB']->exec_DELETEquery('index_rel', 'phash='.intval($phash));
1877 foreach($wl as $val) {
1878 $insertFields = array(
1880 'wid' => $val['hash'],
1881 'count' => $val['count'],
1882 'first' => $val['first'],
1883 'freq' => $this->freqMap(($val['count']/$this->wordcount
)),
1884 'flags' => ($val['cmp'] & $this->flagBitMask
)
1887 $GLOBALS['TYPO3_DB']->exec_INSERTquery('index_rel', $insertFields);
1892 * maps frequency from a real number in [0;1] to an integer in [0;$this->freqRange] with anything above $this->freqMax as 1
1895 * @param double Frequency
1896 * @return integer Frequency in range.
1898 function freqMap($freq) {
1899 $mapFactor = $this->freqMax
*100*$this->freqRange
;
1901 $newFreq = $freq*$mapFactor;
1902 $newFreq = $newFreq>$this->freqRange?
$this->freqRange
:$newFreq;
1904 $newFreq = $freq/$mapFactor;
1920 /********************************
1924 *******************************/
1927 * Get search hash, T3 pages
1931 function setT3Hashes() {
1935 'id' => (integer)$this->conf
['id'],
1936 'type' => (integer)$this->conf
['type'],
1937 'sys_lang' => (integer)$this->conf
['sys_language_uid'],
1938 'MP' => (string)$this->conf
['MP'],
1939 'cHash' => $this->cHashParams
1942 // Set grouping hash (Identifies a "page" combined of id, type, language, mountpoint and cHash parameters):
1943 $this->hash
['phash_grouping'] = $this->md5inthash(serialize($hArray));
1945 // 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.)
1946 $hArray['gr_list'] = (string)$this->conf
['gr_list'];
1947 $this->hash
['phash'] = $this->md5inthash(serialize($hArray));
1951 * Get search hash, external files
1953 * @param string File name / path which identifies it on the server
1954 * @param array Additional content identifying the (subpart of) content. For instance; PDF files are divided into groups of pages for indexing.
1955 * @return array Array with "phash_grouping" and "phash" inside.
1957 function setExtHashes($file,$subinfo=array()) {
1964 // Set grouping hash:
1965 $hash['phash_grouping'] = $this->md5inthash(serialize($hArray));
1968 $hArray['subinfo'] = $subinfo;
1969 $hash['phash'] = $this->md5inthash(serialize($hArray));
1976 * 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.
1978 * @param string String to hash
1979 * @return integer Integer intepretation of the md5 hash of input string.
1981 function md5inthash($str) {
1982 return hexdec(substr(md5($str),0,7));
1986 * Calculates the cHash value of input GET array (for constructing cHash values if needed)
1988 * @param array Array of GET parameters to encode
1991 function makeCHash($paramArray) {
1992 $addQueryParams = t3lib_div
::implodeArrayForUrl('', $paramArray);
1994 $pA = t3lib_div
::cHashParams($addQueryParams);
1996 return t3lib_div
::shortMD5(serialize($pA));
2010 /*********************************
2012 * Internal logging functions
2014 *********************************/
2017 * Push function wrapper for TT logging
2019 * @param string Title to set
2020 * @param string Key (?)
2023 function log_push($msg,$key) {
2024 if (is_object($GLOBALS['TT'])) $GLOBALS['TT']->push($msg,$key);
2028 * Pull function wrapper for TT logging
2032 function log_pull() {
2033 if (is_object($GLOBALS['TT'])) $GLOBALS['TT']->pull();
2037 * Set log message function wrapper for TT logging
2039 * @param string Message to set
2040 * @param integer Error number
2043 function log_setTSlogMessage($msg, $errorNum=0) {
2044 if (is_object($GLOBALS['TT'])) $GLOBALS['TT']->setTSlogMessage($msg,$errorNum);
2045 $this->internal_log
[] = $msg;
2055 /**************************
2059 **************************/
2062 * Frontend hook: If the page is not being re-generated this is our chance to force it to be (because re-generation of the page is required in order to have the indexer called!)
2064 * @param array Parameters from frontend
2065 * @param object TSFE object (reference under PHP5)
2068 function fe_headerNoCache(&$params, $ref) {
2070 // Requirements are that the crawler is loaded, a crawler session is running and re-indexing requested as processing instruction:
2071 if (t3lib_extMgm
::isLoaded('crawler')
2072 && $params['pObj']->applicationData
['tx_crawler']['running']
2073 && in_array('tx_indexedsearch_reindex', $params['pObj']->applicationData
['tx_crawler']['parameters']['procInstructions'])) {
2075 // Setting simple log entry:
2076 $params['pObj']->applicationData
['tx_crawler']['log'][] = 'RE_CACHE (indexed), old status: '.$params['disableAcquireCacheData'];
2078 // Disables a look-up for cached page data - thus resulting in re-generation of the page even if cached.
2079 $params['disableAcquireCacheData'] = TRUE;
2085 if (defined('TYPO3_MODE') && $TYPO3_CONF_VARS[TYPO3_MODE
]['XCLASS']['ext/indexed_search/class.indexer.php']) {
2086 include_once($TYPO3_CONF_VARS[TYPO3_MODE
]['XCLASS']['ext/indexed_search/class.indexer.php']);