/**
* Transforms a generic Query object into an Elastic Search query DSL
*
* @param Query $query
*/
public function mapQuery(Query $query)
{
$arguments = array();
$map = new Boolean();
if ($query->hasQueryString()) {
Lucene::setDefaultSearchField($query->getQueryString()->getDefaultField());
QueryParser::setDefaultOperator($query->getQueryString()->getDefaultOperator() == Query::OPERATOR_AND ? QueryParser::B_AND : QueryParser::B_OR);
$keyword = $query->getQueryString()->getQuery();
if ("*" === $keyword) {
$subQuery = new Wildcard(new Term($keyword));
$subQuery->setMinPrefixLength(0);
} else {
$subQuery = QueryParser::parse($keyword);
}
$map->addSubquery($subQuery, true);
}
$arguments[] = $map;
foreach ($query->getSort() as $sort) {
$arguments[] = key($sort);
$arguments[] = SORT_REGULAR;
$arguments[] = current($sort) == 'asc' ? SORT_ASC : SORT_DESC;
}
return $arguments;
}
/**
* @returns \Zend\Search\Lucene\SearchIndexInterface
*/
public function get()
{
if (!$this->checkEsists()) {
return Lucene::create($this->path);
} else {
return Lucene::open($this->path);
}
}
/**
* @covers Zend\Search\Lucene\MultiSearcher::find
* @covers Zend\Search\Lucene\Search\QueryHit::getDocument
*/
public function testFind()
{
$index = new Lucene\MultiSearcher(array(Lucene\Lucene::open(__DIR__ . '/_indexSample/_files'), Lucene\Lucene::open(__DIR__ . '/_indexSample/_files')));
$hits = $index->find('submitting');
$this->assertEquals(count($hits), 2 * 3);
foreach ($hits as $hit) {
$document = $hit->getDocument();
$this->assertTrue($document instanceof Lucene\Document);
}
}
/**
* Transforms given objects into a bulk add operation directive
*
* @param ClassMetadata $metadata
* @param array $objects
* @param array bulk commands
*/
public function run(ClassMetadata $metadata, array $objects)
{
$index = $metadata->getIndex()->getName();
$index = Lucene::open("/tmp/index_$index");
foreach ($objects as $object) {
$document = $this->exportObject($metadata, $object);
$index->addDocument($document);
}
}
/**
* Re-write query into primitive queries in the context of specified index
*
* @param \Zend\Search\Lucene\SearchIndexInterface $index
* @throws \Zend\Search\Lucene\Exception\RuntimeException
* @throws \Zend\Search\Lucene\Exception\OutOfBoundsException
* @return \Zend\Search\Lucene\Search\Query\AbstractQuery
*/
public function rewrite(Lucene\SearchIndexInterface $index)
{
$this->_matches = array();
if ($this->_pattern->field === null) {
// Search through all fields
$fields = $index->getFieldNames(true);
} else {
$fields = array($this->_pattern->field);
}
$prefix = self::_getPrefix($this->_pattern->text);
$prefixLength = strlen($prefix);
$matchExpression = '/^' . str_replace(array('\\?', '\\*'), array('.', '.*'), preg_quote($this->_pattern->text, '/')) . '$/';
if ($prefixLength < self::$_minPrefixLength) {
throw new RuntimeException('At least ' . self::$_minPrefixLength . ' non-wildcard characters are required at the beginning of pattern.');
}
/** @todo check for PCRE unicode support may be performed through Zend_Environment in some future */
if (@preg_match('/\\pL/u', 'a') == 1) {
// PCRE unicode support is turned on
// add Unicode modifier to the match expression
$matchExpression .= 'u';
}
$maxTerms = Lucene\Lucene::getTermsPerQueryLimit();
foreach ($fields as $field) {
$index->resetTermsStream();
if ($prefix != '') {
$index->skipTo(new Index\Term($prefix, $field));
while ($index->currentTerm() !== null && $index->currentTerm()->field == $field && substr($index->currentTerm()->text, 0, $prefixLength) == $prefix) {
if (preg_match($matchExpression, $index->currentTerm()->text) === 1) {
$this->_matches[] = $index->currentTerm();
if ($maxTerms != 0 && count($this->_matches) > $maxTerms) {
throw new OutOfBoundsException('Terms per query limit is reached.');
}
}
$index->nextTerm();
}
} else {
$index->skipTo(new Index\Term('', $field));
while ($index->currentTerm() !== null && $index->currentTerm()->field == $field) {
if (preg_match($matchExpression, $index->currentTerm()->text) === 1) {
$this->_matches[] = $index->currentTerm();
if ($maxTerms != 0 && count($this->_matches) > $maxTerms) {
throw new OutOfBoundsException('Terms per query limit is reached.');
}
}
$index->nextTerm();
}
}
$index->closeTermsStream();
}
if (count($this->_matches) == 0) {
return new EmptyResult();
} elseif (count($this->_matches) == 1) {
return new Term(reset($this->_matches));
} else {
$rewrittenQuery = new MultiTerm();
foreach ($this->_matches as $matchedTerm) {
$rewrittenQuery->addTerm($matchedTerm);
}
return $rewrittenQuery;
}
}
/**
* Re-write query into primitive queries in the context of specified index
*
* @param \Zend\Search\Lucene\SearchIndex $index
* @throws \Zend\Search\Lucene\Exception\OutOfBoundsException
* @return \Zend\Search\Lucene\Search\Query\AbstractQuery
*/
public function rewrite(Lucene\SearchIndex $index)
{
$this->_matches = array();
if ($this->_field === null) {
// Search through all fields
$fields = $index->getFieldNames(true);
} else {
$fields = array($this->_field);
}
$maxTerms = Lucene\Lucene::getTermsPerQueryLimit();
foreach ($fields as $field) {
$index->resetTermsStream();
if ($this->_lowerTerm !== null) {
$lowerTerm = new Index\Term($this->_lowerTerm->text, $field);
$index->skipTo($lowerTerm);
if (!$this->_inclusive && $index->currentTerm() == $lowerTerm) {
// Skip lower term
$index->nextTerm();
}
} else {
$index->skipTo(new Index\Term('', $field));
}
if ($this->_upperTerm !== null) {
// Walk up to the upper term
$upperTerm = new Index\Term($this->_upperTerm->text, $field);
while ($index->currentTerm() !== null && $index->currentTerm()->field == $field && $index->currentTerm()->text < $upperTerm->text) {
$this->_matches[] = $index->currentTerm();
if ($maxTerms != 0 && count($this->_matches) > $maxTerms) {
throw new OutOfBoundsException('Terms per query limit is reached.');
}
$index->nextTerm();
}
if ($this->_inclusive && $index->currentTerm() == $upperTerm) {
// Include upper term into result
$this->_matches[] = $upperTerm;
}
} else {
// Walk up to the end of field data
while ($index->currentTerm() !== null && $index->currentTerm()->field == $field) {
$this->_matches[] = $index->currentTerm();
if ($maxTerms != 0 && count($this->_matches) > $maxTerms) {
throw new OutOfBoundsException('Terms per query limit is reached.');
}
$index->nextTerm();
}
}
$index->closeTermsStream();
}
if (count($this->_matches) == 0) {
return new EmptyResult();
} else {
if (count($this->_matches) == 1) {
return new Term(reset($this->_matches));
} else {
$rewrittenQuery = new MultiTerm();
foreach ($this->_matches as $matchedTerm) {
$rewrittenQuery->addTerm($matchedTerm);
}
return $rewrittenQuery;
}
}
}
/**
* Re-write query into primitive queries in the context of specified index
*
* @param \Zend\Search\Lucene\SearchIndexInterface $index
* @return \Zend\Search\Lucene\Search\Query\AbstractQuery
*/
public function rewrite(Lucene\SearchIndexInterface $index)
{
// Allow to use wildcards within phrases
// They are either removed by text analyzer or used as a part of keyword for keyword fields
//
// if (strpos($this->_phrase, '?') !== false || strpos($this->_phrase, '*') !== false) {
// require_once 'Zend/Search/Lucene/Search/QueryParserException.php';
// throw new Zend_Search_Lucene_Search_QueryParserException('Wildcards are only allowed in a single terms.');
// }
// Split query into subqueries if field name is not specified
if ($this->_field === null) {
$query = new Query\Boolean();
$query->setBoost($this->getBoost());
if (Lucene\Lucene::getDefaultSearchField() === null) {
$searchFields = $index->getFieldNames(true);
} else {
$searchFields = array(Lucene\Lucene::getDefaultSearchField());
}
foreach ($searchFields as $fieldName) {
$subquery = new Phrase($this->_phrase, $this->_phraseEncoding, $fieldName);
$subquery->setSlop($this->getSlop());
$query->addSubquery($subquery->rewrite($index));
}
$this->_matches = $query->getQueryTerms();
return $query;
}
// Recognize exact term matching (it corresponds to Keyword fields stored in the index)
// encoding is not used since we expect binary matching
$term = new Index\Term($this->_phrase, $this->_field);
if ($index->hasTerm($term)) {
$query = new Query\Term($term);
$query->setBoost($this->getBoost());
$this->_matches = $query->getQueryTerms();
return $query;
}
// tokenize phrase using current analyzer and process it as a phrase query
$tokens = Analyzer::getDefault()->tokenize($this->_phrase, $this->_phraseEncoding);
if (count($tokens) == 0) {
$this->_matches = array();
return new Query\Insignificant();
}
if (count($tokens) == 1) {
$term = new Index\Term($tokens[0]->getTermText(), $this->_field);
$query = new Query\Term($term);
$query->setBoost($this->getBoost());
$this->_matches = $query->getQueryTerms();
return $query;
}
//It's non-trivial phrase query
$position = -1;
$query = new Query\Phrase();
foreach ($tokens as $token) {
$position += $token->getPositionIncrement();
$term = new Index\Term($token->getTermText(), $this->_field);
$query->addTerm($term, $position);
$query->setSlop($this->getSlop());
}
$this->_matches = $query->getQueryTerms();
return $query;
}
/**
* @group ZF-4252
*/
public function testHtmlInlineTagsIndexing()
{
$index = Lucene\Lucene::create(__DIR__ . '/_index/_files');
$htmlString = '<html><head><title>Hello World</title></head>' . '<body><b>Zend</b>Framework' . "\n" . ' <div>Foo</div>Bar ' . "\n" . ' <strong>Test</strong></body></html>';
$doc = Document\Html::loadHTML($htmlString);
$index->addDocument($doc);
$hits = $index->find('FooBar');
$this->assertEquals(count($hits), 0);
$hits = $index->find('ZendFramework');
$this->assertEquals(count($hits), 1);
unset($index);
$this->_clearDirectory(__DIR__ . '/_index/_files');
}
/**
* @group ZF-9680
*/
public function testIsDeletedWithoutExplicitCommit()
{
$index = Lucene\Lucene::create(__DIR__ . '/_index/_files');
$document = new Document();
$document->addField(Document\Field::Keyword('_id', 'myId'));
$document->addField(Document\Field::Keyword('bla', 'blubb'));
$index->addDocument($document);
$this->assertFalse($index->isDeleted(0));
}
/**
* Class constructor. Create a new multi-term query object.
*
* if $signs array is omitted then all terms are required
* it differs from addTerm() behavior, but should never be used
*
* @param array $terms Array of \Zend\Search\Lucene\Index\Term objects
* @param array $signs Array of signs. Sign is boolean|null.
* @throws \Zend\Search\Lucene\Exception\InvalidArgumentException
*/
public function __construct($terms = null, $signs = null)
{
if (is_array($terms)) {
if (count($terms) > Lucene\Lucene::getTermsPerQueryLimit()) {
throw new InvalidArgumentException('Terms per query limit is reached.');
}
$this->_terms = $terms;
$this->_signs = null;
// Check if all terms are required
if (is_array($signs)) {
foreach ($signs as $sign) {
if ($sign !== true) {
$this->_signs = $signs;
break;
}
}
}
}
}
/**
* Re-write query into primitive queries in the context of specified index
*
* @param \Zend\Search\Lucene\SearchIndex $index
* @throws \Zend\Search\Lucence\Search\Exception\QueryParserException
* @return \Zend\Search\Lucene\Search\Query\AbstractQuery
*/
public function rewrite(Lucene\SearchIndex $index)
{
if ($this->_field === null) {
$query = new Query\MultiTerm();
$query->setBoost($this->getBoost());
$hasInsignificantSubqueries = false;
if (Lucene\Lucene::getDefaultSearchField() === null) {
$searchFields = $index->getFieldNames(true);
} else {
$searchFields = array(Lucene\Lucene::getDefaultSearchField());
}
foreach ($searchFields as $fieldName) {
$subquery = new Term($this->_word, $this->_encoding, $fieldName);
$rewrittenSubquery = $subquery->rewrite($index);
foreach ($rewrittenSubquery->getQueryTerms() as $term) {
$query->addTerm($term);
}
if ($rewrittenSubquery instanceof Query\Insignificant) {
$hasInsignificantSubqueries = true;
}
}
if (count($query->getTerms()) == 0) {
$this->_matches = array();
if ($hasInsignificantSubqueries) {
return new Query\Insignificant();
} else {
return new Query\EmptyResult();
}
}
$this->_matches = $query->getQueryTerms();
return $query;
}
// -------------------------------------
// Recognize exact term matching (it corresponds to Keyword fields stored in the index)
// encoding is not used since we expect binary matching
$term = new Index\Term($this->_word, $this->_field);
if ($index->hasTerm($term)) {
$query = new Query\Term($term);
$query->setBoost($this->getBoost());
$this->_matches = $query->getQueryTerms();
return $query;
}
// -------------------------------------
// Recognize wildcard queries
/** @todo check for PCRE unicode support may be performed through Zend_Environment in some future */
if (@preg_match('/\\pL/u', 'a') == 1) {
$word = iconv($this->_encoding, 'UTF-8', $this->_word);
$wildcardsPattern = '/[*?]/u';
$subPatternsEncoding = 'UTF-8';
} else {
$word = $this->_word;
$wildcardsPattern = '/[*?]/';
$subPatternsEncoding = $this->_encoding;
}
$subPatterns = preg_split($wildcardsPattern, $word, -1, PREG_SPLIT_OFFSET_CAPTURE);
if (count($subPatterns) > 1) {
// Wildcard query is recognized
$pattern = '';
foreach ($subPatterns as $id => $subPattern) {
// Append corresponding wildcard character to the pattern before each sub-pattern (except first)
if ($id != 0) {
$pattern .= $word[$subPattern[1] - 1];
}
// Check if each subputtern is a single word in terms of current analyzer
$tokens = Analyzer\Analyzer::getDefault()->tokenize($subPattern[0], $subPatternsEncoding);
if (count($tokens) > 1) {
throw new QueryParserException('Wildcard search is supported only for non-multiple word terms');
}
foreach ($tokens as $token) {
$pattern .= $token->getTermText();
}
}
$term = new Index\Term($pattern, $this->_field);
$query = new Query\Wildcard($term);
$query->setBoost($this->getBoost());
// Get rewritten query. Important! It also fills terms matching container.
$rewrittenQuery = $query->rewrite($index);
$this->_matches = $query->getQueryTerms();
return $rewrittenQuery;
}
// -------------------------------------
// Recognize one-term multi-term and "insignificant" queries
$tokens = Analyzer\Analyzer::getDefault()->tokenize($this->_word, $this->_encoding);
if (count($tokens) == 0) {
$this->_matches = array();
return new Query\Insignificant();
}
if (count($tokens) == 1) {
$term = new Index\Term($tokens[0]->getTermText(), $this->_field);
$query = new Query\Term($term);
$query->setBoost($this->getBoost());
$this->_matches = $query->getQueryTerms();
return $query;
}
//It's not insignificant or one term query
$query = new Query\MultiTerm();
/**
* @todo Process $token->getPositionIncrement() to support stemming, synonyms and other
* analizer design features
*/
foreach ($tokens as $token) {
$term = new Index\Term($token->getTermText(), $this->_field);
$query->addTerm($term, true);
// all subterms are required
}
$query->setBoost($this->getBoost());
$this->_matches = $query->getQueryTerms();
return $query;
}
/**
* Performs a query against the index and returns an array
* of Zend_Search_Lucene_Search_QueryHit objects.
* Input is a string or Zend_Search_Lucene_Search_Query.
*
* @param \Zend\Search\Lucene\Search\QueryParser|string $query
* @return array \Zend\Search\Lucene\Search\QueryHit
* @throws \Zend\Search\Lucene\Exception\InvalidArgumentException
* @throws \Zend\Search\Lucene\Exception\RuntimeException
*/
public function find($query)
{
if (is_string($query)) {
$query = Search\QueryParser::parse($query);
}
if (!$query instanceof Search\Query\AbstractQuery) {
throw new InvalidArgumentException('Query must be a string or Zend\Search\Lucene\Search\Query object');
}
$this->commit();
$hits = array();
$scores = array();
$ids = array();
$query = $query->rewrite($this)->optimize($this);
$query->execute($this);
$topScore = 0;
$resultSetLimit = Lucene::getResultSetLimit();
foreach ($query->matchedDocs() as $id => $num) {
$docScore = $query->score($id, $this);
if( $docScore != 0 ) {
$hit = new Search\QueryHit($this);
$hit->id = $id;
$hit->score = $docScore;
$hits[] = $hit;
$ids[] = $id;
$scores[] = $docScore;
if ($docScore > $topScore) {
$topScore = $docScore;
}
}
if ($resultSetLimit != 0 && count($hits) >= $resultSetLimit) {
break;
}
}
if (count($hits) == 0) {
// skip sorting, which may cause a error on empty index
return array();
}
if ($topScore > 1) {
foreach ($hits as $hit) {
$hit->score /= $topScore;
}
}
if (func_num_args() == 1) {
// sort by scores
array_multisort($scores, SORT_DESC, SORT_NUMERIC,
$ids, SORT_ASC, SORT_NUMERIC,
$hits);
} else {
// sort by given field names
$argList = func_get_args();
$fieldNames = $this->getFieldNames();
$sortArgs = array();
// PHP 5.3 now expects all arguments to array_multisort be passed by
// reference (if it's invoked through call_user_func_array());
// since constants can't be passed by reference, create some placeholder variables.
$sortReg = SORT_REGULAR;
$sortAsc = SORT_ASC;
$sortNum = SORT_NUMERIC;
$sortFieldValues = array();
for ($count = 1; $count < count($argList); $count++) {
$fieldName = $argList[$count];
if (!is_string($fieldName)) {
throw new RuntimeException('Field name must be a string.');
}
if (strtolower($fieldName) == 'score') {
$sortArgs[] = &$scores;
} else {
if (!in_array($fieldName, $fieldNames)) {
throw new RuntimeException('Wrong field name.');
}
if (!isset($sortFieldValues[$fieldName])) {
$valuesArray = array();
foreach ($hits as $hit) {
try {
$value = $hit->getDocument()->getFieldValue($fieldName);
} catch (\Exception $e) {
if (strpos($e->getMessage(), 'not found') === false) {
throw new RuntimeException($e->getMessage(), $e->getCode(), $e);
} else {
$value = null;
}
}
$valuesArray[] = $value;
}
// Collect loaded values in $sortFieldValues
// Required for PHP 5.3 which translates references into values when source
// variable is destroyed
$sortFieldValues[$fieldName] = $valuesArray;
}
$sortArgs[] = &$sortFieldValues[$fieldName];
}
if ($count + 1 < count($argList) && is_integer($argList[$count+1])) {
$count++;
$sortArgs[] = &$argList[$count];
if ($count + 1 < count($argList) && is_integer($argList[$count+1])) {
$count++;
$sortArgs[] = &$argList[$count];
} else {
if ($argList[$count] == SORT_ASC || $argList[$count] == SORT_DESC) {
$sortArgs[] = &$sortReg;
} else {
$sortArgs[] = &$sortAsc;
}
}
} else {
$sortArgs[] = &$sortAsc;
$sortArgs[] = &$sortReg;
}
}
// Sort by id's if values are equal
$sortArgs[] = &$ids;
$sortArgs[] = &$sortAsc;
$sortArgs[] = &$sortNum;
// Array to be sorted
$sortArgs[] = &$hits;
// Do sort
call_user_func_array('array_multisort', $sortArgs);
}
return $hits;
}
/**
* Re-write query into primitive queries in the context of specified index
*
* @param \Zend\Search\Lucene\IndexInterface $index
* @return \Zend\Search\Lucene\Search\Query\AbstractQuery
* @throws \Zend\Search\Lucene\Exception
*/
public function rewrite(Lucene\IndexInterface $index)
{
$this->_matches = array();
$this->_scores = array();
$this->_termKeys = array();
if ($this->_term->field === null) {
// Search through all fields
$fields = $index->getFieldNames(true);
} else {
$fields = array($this->_term->field);
}
$prefix = Index\Term::getPrefix($this->_term->text, $this->_prefixLength);
$prefixByteLength = strlen($prefix);
$prefixUtf8Length = Index\Term::getLength($prefix);
$termLength = Index\Term::getLength($this->_term->text);
$termRest = substr($this->_term->text, $prefixByteLength);
// we calculate length of the rest in bytes since levenshtein() is not UTF-8 compatible
$termRestLength = strlen($termRest);
$scaleFactor = 1 / (1 - $this->_minimumSimilarity);
$maxTerms = Lucene\Lucene::getTermsPerQueryLimit();
foreach ($fields as $field) {
$index->resetTermsStream();
if ($prefix != '') {
$index->skipTo(new Index\Term($prefix, $field));
while ($index->currentTerm() !== null && $index->currentTerm()->field == $field && substr($index->currentTerm()->text, 0, $prefixByteLength) == $prefix) {
// Calculate similarity
$target = substr($index->currentTerm()->text, $prefixByteLength);
$maxDistance = isset($this->_maxDistances[strlen($target)]) ? $this->_maxDistances[strlen($target)] : $this->_calculateMaxDistance($prefixUtf8Length, $termRestLength, strlen($target));
if ($termRestLength == 0) {
// we don't have anything to compare. That means if we just add
// the letters for current term we get the new word
$similarity = $prefixUtf8Length == 0 ? 0 : 1 - strlen($target) / $prefixUtf8Length;
} else {
if (strlen($target) == 0) {
$similarity = $prefixUtf8Length == 0 ? 0 : 1 - $termRestLength / $prefixUtf8Length;
} else {
if ($maxDistance < abs($termRestLength - strlen($target))) {
//just adding the characters of term to target or vice-versa results in too many edits
//for example "pre" length is 3 and "prefixes" length is 8. We can see that
//given this optimal circumstance, the edit distance cannot be less than 5.
//which is 8-3 or more precisesly abs(3-8).
//if our maximum edit distance is 4, then we can discard this word
//without looking at it.
$similarity = 0;
} else {
$similarity = 1 - levenshtein($termRest, $target) / ($prefixUtf8Length + min($termRestLength, strlen($target)));
}
}
}
if ($similarity > $this->_minimumSimilarity) {
$this->_matches[] = $index->currentTerm();
$this->_termKeys[] = $index->currentTerm()->key();
$this->_scores[] = ($similarity - $this->_minimumSimilarity) * $scaleFactor;
if ($maxTerms != 0 && count($this->_matches) > $maxTerms) {
throw new Lucene\Exception('Terms per query limit is reached.');
}
}
$index->nextTerm();
}
} else {
$index->skipTo(new Index\Term('', $field));
while ($index->currentTerm() !== null && $index->currentTerm()->field == $field) {
// Calculate similarity
$target = $index->currentTerm()->text;
$maxDistance = isset($this->_maxDistances[strlen($target)]) ? $this->_maxDistances[strlen($target)] : $this->_calculateMaxDistance(0, $termRestLength, strlen($target));
if ($maxDistance < abs($termRestLength - strlen($target))) {
//just adding the characters of term to target or vice-versa results in too many edits
//for example "pre" length is 3 and "prefixes" length is 8. We can see that
//given this optimal circumstance, the edit distance cannot be less than 5.
//which is 8-3 or more precisesly abs(3-8).
//if our maximum edit distance is 4, then we can discard this word
//without looking at it.
$similarity = 0;
} else {
$similarity = 1 - levenshtein($termRest, $target) / min($termRestLength, strlen($target));
}
if ($similarity > $this->_minimumSimilarity) {
$this->_matches[] = $index->currentTerm();
$this->_termKeys[] = $index->currentTerm()->key();
$this->_scores[] = ($similarity - $this->_minimumSimilarity) * $scaleFactor;
if ($maxTerms != 0 && count($this->_matches) > $maxTerms) {
throw new Lucene\Exception('Terms per query limit is reached.');
}
}
$index->nextTerm();
}
}
$index->closeTermsStream();
}
if (count($this->_matches) == 0) {
return new EmptyResult();
} else {
if (count($this->_matches) == 1) {
return new Term(reset($this->_matches));
} else {
$rewrittenQuery = new Boolean();
array_multisort($this->_scores, SORT_DESC, SORT_NUMERIC, $this->_termKeys, SORT_ASC, SORT_STRING, $this->_matches);
$termCount = 0;
foreach ($this->_matches as $id => $matchedTerm) {
$subquery = new Term($matchedTerm);
$subquery->setBoost($this->_scores[$id]);
$rewrittenQuery->addSubquery($subquery);
$termCount++;
if ($termCount >= self::MAX_CLAUSE_COUNT) {
break;
}
}
return $rewrittenQuery;
}
}
}
/**
* Re-write query into primitive queries in the context of specified index
*
* @param \Zend\Search\Lucene\SearchIndex $index
* @throws \Zend\Search\Lucence\Search\Exception\QueryParserException
* @return \Zend\Search\Lucene\Search\Query\AbstractQuery
*/
public function rewrite(Lucene\SearchIndex $index)
{
if ($this->_field === null) {
$query = new Search\Query\Boolean();
$hasInsignificantSubqueries = false;
if (Lucene\Lucene::getDefaultSearchField() === null) {
$searchFields = $index->getFieldNames(true);
} else {
$searchFields = array(Lucene\Lucene::getDefaultSearchField());
}
foreach ($searchFields as $fieldName) {
$subquery = new self($this->_word, $this->_encoding, $fieldName, $this->_minimumSimilarity);
$rewrittenSubquery = $subquery->rewrite($index);
if (!($rewrittenSubquery instanceof Query\Insignificant || $rewrittenSubquery instanceof Query\EmptyResult)) {
$query->addSubquery($rewrittenSubquery);
}
if ($rewrittenSubquery instanceof Query\Insignificant) {
$hasInsignificantSubqueries = true;
}
}
$subqueries = $query->getSubqueries();
if (count($subqueries) == 0) {
$this->_matches = array();
if ($hasInsignificantSubqueries) {
return new Query\Insignificant();
} else {
return new Query\EmptyResult();
}
}
if (count($subqueries) == 1) {
$query = reset($subqueries);
}
$query->setBoost($this->getBoost());
$this->_matches = $query->getQueryTerms();
return $query;
}
// -------------------------------------
// Recognize exact term matching (it corresponds to Keyword fields stored in the index)
// encoding is not used since we expect binary matching
$term = new Index\Term($this->_word, $this->_field);
if ($index->hasTerm($term)) {
$query = new Query\Fuzzy($term, $this->_minimumSimilarity);
$query->setBoost($this->getBoost());
// Get rewritten query. Important! It also fills terms matching container.
$rewrittenQuery = $query->rewrite($index);
$this->_matches = $query->getQueryTerms();
return $rewrittenQuery;
}
// -------------------------------------
// Recognize wildcard queries
/** @todo check for PCRE unicode support may be performed through Zend_Environment in some future */
if (@preg_match('/\\pL/u', 'a') == 1) {
$subPatterns = preg_split('/[*?]/u', iconv($this->_encoding, 'UTF-8', $this->_word));
} else {
$subPatterns = preg_split('/[*?]/', $this->_word);
}
if (count($subPatterns) > 1) {
throw new QueryParserException('Fuzzy search doesn\'t support wildcards (except within Keyword fields).');
}
// -------------------------------------
// Recognize one-term multi-term and "insignificant" queries
$tokens = Analyzer\Analyzer::getDefault()->tokenize($this->_word, $this->_encoding);
if (count($tokens) == 0) {
$this->_matches = array();
return new Query\Insignificant();
}
if (count($tokens) == 1) {
$term = new Index\Term($tokens[0]->getTermText(), $this->_field);
$query = new Query\Fuzzy($term, $this->_minimumSimilarity);
$query->setBoost($this->getBoost());
// Get rewritten query. Important! It also fills terms matching container.
$rewrittenQuery = $query->rewrite($index);
$this->_matches = $query->getQueryTerms();
return $rewrittenQuery;
}
// Word is tokenized into several tokens
throw new QueryParserException('Fuzzy search is supported only for non-multiple word terms');
}
/**
* Transforms given objects into a bulk add operation directive
*
* @param ClassMetadata $metadata
* @param array $objects
* @param array bulk commands
*/
public function run(ClassMetadata $metadata)
{
$index = $metadata->getIndex()->getName();
Lucene::create("/tmp/index_$index");
}
public function testLimitingResult()
{
$index = Lucene\Lucene::open(dirname(__FILE__) . '/_index23Sample/_files');
$storedResultSetLimit = Lucene\Lucene::getResultSetLimit();
Lucene\Lucene::setResultSetLimit(3);
$hits = $index->find('"reporting bugs"', 'path');
$this->assertEquals(count($hits), 3);
$expectedResultset = array(array(7, 0.212395, 'IndexSource/contributing.bugs.html'), array(0, 0.247795, 'IndexSource/contributing.documentation.html'), array(2, 0.176996, 'IndexSource/contributing.patches.html'));
foreach ($hits as $resId => $hit) {
$this->assertEquals($hit->id, $expectedResultset[$resId][0]);
$this->assertTrue(abs($hit->score - $expectedResultset[$resId][1]) < 1.0E-6);
$this->assertEquals($hit->path, $expectedResultset[$resId][2]);
}
Lucene\Lucene::setResultSetLimit($storedResultSetLimit);
}
/**
* Object constructor
*/
public function __construct($text, $field = null)
{
$this->field = ($field === null)? Lucene\Lucene::getDefaultSearchField() : $field;
$this->text = $text;
}
public function testTermsStreamInterfaceSkipToTermsRetrievingTwoTermsCase()
{
$index = Lucene\Lucene::create(dirname(__FILE__) . '/_index/_files');
// Zero terms
$doc = new Document\Document();
$doc->addField(Document\Field::Text('contents', 'someterm word'));
$index->addDocument($doc);
unset($index);
$index = Lucene\Lucene::open(dirname(__FILE__) . '/_index/_files');
$index->resetTermsStream();
$index->skipTo(new Index\Term('term', 'contents'));
$this->assertTrue($index->currentTerm() == new Index\Term('word', 'contents'));
$index->closeTermsStream();
$this->_clearDirectory(dirname(__FILE__) . '/_index/_files');
}