public function testFuzzyQuery()
{
$index = Lucene\Lucene::open(__DIR__ . '/_index23Sample/_files');
$defaultPrefixLength = Query\Fuzzy::getDefaultPrefixLength();
Query\Fuzzy::setDefaultPrefixLength(0);
$hits = $index->find('tesd~0.4');
$this->assertEquals(count($hits), 9);
$expectedResultset = array(array(2, 0.037139, 'IndexSource/contributing.patches.html'), array(0, 0.008735, 'IndexSource/contributing.documentation.html'), array(7, 0.002449, 'IndexSource/contributing.bugs.html'), array(1, 0.000483, 'IndexSource/contributing.wishlist.html'), array(3, 0.000483, 'IndexSource/about-pear.html'), array(9, 0.000483, 'IndexSource/core.html'), array(5, 0.000414, 'IndexSource/authors.html'), array(8, 0.000414, 'IndexSource/contributing.html'), array(4, 0.000345, 'IndexSource/copyright.html'));
foreach ($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]);
}
Query\Fuzzy::setDefaultPrefixLength($defaultPrefixLength);
}
public function tearDown()
{
Query\Wildcard::setMinPrefixLength($this->_wildcardMinPrefix);
Query\Fuzzy::setDefaultPrefixLength($this->_defaultPrefixLength);
}
/**
* Query specific matches highlighting
*
* @param Highlighter $highlighter Highlighter object (also contains doc for highlighting)
*/
protected function _highlightMatches(Highlighter $highlighter)
{
/** Skip fields detection. We don't need it, since we expect all fields presented in the HTML body and don't differentiate them */
/** Skip exact term matching recognition, keyword fields highlighting is not supported */
// -------------------------------------
// Recognize wildcard queries
/**
* @todo check for PCRE unicode support may be performed through Zend_Environment in some future
*/
ErrorHandler::start(E_WARNING);
$result = preg_match('/\\pL/u', 'a');
ErrorHandler::stop();
if ($result == 1) {
$subPatterns = preg_split('/[*?]/u', iconv($this->_encoding, 'UTF-8', $this->_word));
} else {
$subPatterns = preg_split('/[*?]/', $this->_word);
}
if (count($subPatterns) > 1) {
// Do nothing
return;
}
// -------------------------------------
// Recognize one-term multi-term and "insignificant" queries
$tokens = Analyzer\Analyzer::getDefault()->tokenize($this->_word, $this->_encoding);
if (count($tokens) == 0) {
// Do nothing
return;
}
if (count($tokens) == 1) {
$term = new Index\Term($tokens[0]->getTermText(), $this->_field);
$query = new Query\Fuzzy($term, $this->_minimumSimilarity);
$query->_highlightMatches($highlighter);
return;
}
// Word is tokenized into several tokens
// But fuzzy search is supported only for non-multiple word terms
// Do nothing
}
