/**
* Return the document object for this hit
*
* @return Zend_Search_Lucene_Document
*/
public function getDocument()
{
if (!$this->_document instanceof Zend_Search_Lucene_Document) {
$this->_document = $this->_index->getDocument($this->id);
}
return $this->_document;
}
function search($pQuery)
{
$this->mResults = array();
if (!empty($pQuery) && $this->verifySearchIndex()) {
parent::search($pQuery);
require_zend_file('Search/Lucene.php');
$index = new ZSearch($this->getField('index_path'));
$fields = explode(',', $this->getField('index_fields'));
$query = '';
$lowQuery = strtolower($pQuery);
foreach ($fields as $f) {
$query .= "{$f}:{$lowQuery} OR ";
}
$query = preg_replace('/ OR $/', '', $query);
$this->mResults = $index->find($query);
}
return count($this->mResults);
}
/**
* The sum of squared weights of contained query clauses.
*
* @return float
*/
public function sumOfSquaredWeights()
{
// compute idf
$this->_idf = $this->_reader->getSimilarity()->idf($this->_term, $this->_reader);
// compute query weight
$this->_queryWeight = $this->_idf * $this->_query->getBoost();
// square it
return $this->_queryWeight * $this->_queryWeight;
}
/**
* Score calculator for sloppy phrase queries (terms sequence is fixed)
*
* @param integer $docId
* @param ZSearch $reader
* @return float
*/
public function _sloppyPhraseFreq($docId, ZSearch $reader)
{
$freq = 0;
$phraseQueue = array();
$phraseQueue[0] = array();
// empty phrase
$lastTerm = null;
// Walk through the terms to create phrases.
foreach ($this->_terms as $termId => $term) {
$queueSize = count($phraseQueue);
$firstPass = true;
// Walk through the term positions.
// Each term position produces a set of phrases.
foreach ($this->_termsPositions[$termId][$docId] as $termPosition) {
if ($firstPass) {
for ($count = 0; $count < $queueSize; $count++) {
$phraseQueue[$count][$termId] = $termPosition;
}
} else {
for ($count = 0; $count < $queueSize; $count++) {
if ($lastTerm !== null && abs($termPosition - $phraseQueue[$count][$lastTerm] - ($this->_offsets[$termId] - $this->_offsets[$lastTerm])) > $this->_slop) {
continue;
}
$newPhraseId = count($phraseQueue);
$phraseQueue[$newPhraseId] = $phraseQueue[$count];
$phraseQueue[$newPhraseId][$termId] = $termPosition;
}
}
$firstPass = false;
}
$lastTerm = $termId;
}
foreach ($phraseQueue as $phrasePos) {
$minDistance = null;
for ($shift = -$this->_slop; $shift <= $this->_slop; $shift++) {
$distance = 0;
$start = reset($phrasePos) - reset($this->_offsets) + $shift;
foreach ($this->_terms as $termId => $term) {
$distance += abs($phrasePos[$termId] - $this->_offsets[$termId] - $start);
if ($distance > $this->_slop) {
break;
}
}
if ($minDistance === null || $distance < $minDistance) {
$minDistance = $distance;
}
}
if ($minDistance <= $this->_slop) {
$freq += $reader->getSimilarity()->sloppyFreq($minDistance);
}
}
return $freq;
}
/**
* Score specified document
*
* @param integer $docId
* @param ZSearch $reader
* @return float
*/
public function score($docId, $reader)
{
if ($this->_docVector === null) {
if (extension_loaded('bitset')) {
$this->_docVector = bitset_from_array($reader->termDocs($this->_term));
} else {
$this->_docVector = array_flip($reader->termDocs($this->_term));
}
$this->_termPositions = $reader->termPositions($this->_term);
$this->_initWeight($reader);
}
$match = extension_loaded('bitset') ? bitset_in($this->_docVector, $docId) : isset($this->_docVector[$docId]);
if ($this->_sign && $match) {
return $reader->getSimilarity()->tf(count($this->_termPositions[$docId])) * $this->_weight->getValue() * $reader->norm($docId, $this->_term->field);
} else {
return 0;
}
}
/**
* Score calculator for non conjunction queries (not all terms are required)
*
* @param integer $docId
* @param ZSearch $reader
* @return float
*/
public function _nonConjunctionScore($docId, $reader)
{
if ($this->_coord === null) {
$this->_coord = array();
$maxCoord = 0;
foreach ($this->_signs as $sign) {
if ($sign !== false) {
$maxCoord++;
}
}
for ($count = 0; $count <= $maxCoord; $count++) {
$this->_coord[$count] = $reader->getSimilarity()->coord($count, $maxCoord);
}
}
$score = 0.0;
$matchedTerms = 0;
foreach ($this->_terms as $termId => $term) {
// Check if term is
if ($this->_signs[$termId] !== false && isset($this->_termsPositions[$termId][$docId])) {
$matchedTerms++;
$score += $reader->getSimilarity()->tf(count($this->_termsPositions[$termId][$docId])) * $this->_weights[$termId]->getValue() * $reader->norm($docId, $term->field);
}
}
return $score * $this->_coord[$matchedTerms];
}
$byteCount = filesize($fileName);
$data = '';
while ($byteCount > 0 && ($nextBlock = fread($f, $byteCount)) != false) {
$data .= $nextBlock;
$byteCount -= strlen($nextBlock);
}
fclose($f);
if ($storeContent) {
$this->addField(Zend_Search_Lucene_Field::Text('contents', $data));
} else {
$this->addField(Zend_Search_Lucene_Field::UnStored('contents', $data));
}
}
}
// Create index
$index = new ZSearch('index', true);
// Uncomment next line if you want to have case sensitive index
// ZSearchAnalyzer::setDefault(new ZSearchTextAnalyzer());
$indexSourceDir = 'IndexSource';
$dir = opendir($indexSourceDir);
while ($file = readdir($dir)) {
if (is_dir($file)) {
continue;
}
if (strcasecmp(substr($file, strlen($file) - 5), '.html') != 0) {
continue;
}
// Create new Document from a file
$doc = new FileDocument($indexSourceDir . '/' . $file, true);
// Add document to the index
$index->addDocument($doc);
<?php
/**
* ZSearch Indexing Example - Create a new index and add a
* document to search. [Under Construction]
*
* ZSearch is an implementation of the Java Lucene engine
* written entirely in PHP 5. It is fully compatible with the
* binary format of the Lucene index files.
*
* Base installation of ZSearch requires no database server or
* even installing a PHP extension, although added performance and
* features are available if you can use one or both.
*/
/**
* The second argument tells ZSearch to create a new index
* instead of opening an existing one.
*/
$index = new ZSearch('/tmp/my_index', true);
/**
* Create a new document with two fields: title and body. Each
* field will be independently searchable.
*/
$myDoc = new ZSearchDocument();
$myDoc->setField('title', 'ZSearch Example Title');
$myDoc->setField('body', 'The body of your document goes here.');
/**
* Add the document to the index. It is now searchable.
*/
$index->addDocument($myDoc);
/**
* Computes a score factor for a simple term or a phrase.
*
* The default implementation is:
* return idfFreq(searcher.docFreq(term), searcher.maxDoc());
*
* input - the term in question or array of terms
* reader - reader the document collection being searched
* Returns a score factor for the term
*
* @param mixed $input
* @param ZSearch $reader
* @return a score factor for the term
*/
public function idf($input, $reader)
{
if (!is_array($input)) {
return $this->idfFreq($reader->docFreq($input), $reader->count());
} else {
$idf = 0.0;
foreach ($input as $term) {
$idf += $this->idfFreq($reader->docFreq($term), $reader->count());
}
return $idf;
}
}
<?php
require_once 'ZSearch/ZSearch.php';
/**
* ZSearch Simple Search Example
*
* ZSearch allows you to search all fields of all documents
*/
/**
* Open the index for searching.
*/
$index = new ZSearch('/tmp/index');
/**
* Query the index for documents that contain the term "zend studio"
* but do not contain the term "install".
*/
$hits = $index->find('nntp');
/**
* List titles of each matching document.
*
* All three methods shown below are equivalent, you only need to
* get the objects if you really need them.
*/
foreach ($hits as $hit) {
// get the value of the document's "title" field directly from the hit (shortcut)
// this should be the most common usage
echo $hit->title;
// get the document object and the "title" field's value from it (shortcut)
echo $hit->getDocument()->title . "\n";
// get the document object and then the "title" object and its value.
echo $hit->getDocument()->getField('title')->getFieldValue() . "\n";
