*!
Normalizes the text \a $text so that it is easily parsable
\param $isMetaData If \c true then it expects the text to be meta data from objects,
if not it is the search text and needs special handling.
| public normalizeText ( $text, $isMetaData = false ) |
/**
* Normalizes the text so that it is easily parsable
*
* Used only by legacy search engine.
*
* @param string $text
*
* @return string
*/
public function normalizeText($text)
{
if ($this->searchEngine == 'legacy') {
$searchEngine = new eZSearchEngine();
return $searchEngine->normalizeText($text);
}
return $text;
}
/**
* Adds object $contentObject to the search database.
*
* @param eZContentObject $contentObject Object to add to search engine
* @param bool $commit Whether to commit after adding the object
* @return bool True if the operation succeed.
*/
public function addObject($contentObject, $commit = true)
{
$contentObjectID = $contentObject->attribute('id');
$currentVersion = $contentObject->currentVersion();
if (!$currentVersion) {
$errCurrentVersion = $contentObject->attribute('current_version');
eZDebug::writeError("Failed to fetch \"current version\" ({$errCurrentVersion})" . " of content object (ID: {$contentObjectID})", 'eZSearchEngine');
return false;
}
$indexArray = array();
$indexArrayOnlyWords = array();
$wordCount = 0;
$placement = 0;
$previousWord = '';
eZContentObject::recursionProtectionStart();
foreach ($currentVersion->contentObjectAttributes() as $attribute) {
$metaData = array();
$classAttribute = $attribute->contentClassAttribute();
if ($classAttribute->attribute("is_searchable") == 1) {
// Fetch attribute translations
$attributeTranslations = $attribute->fetchAttributeTranslations();
foreach ($attributeTranslations as $translation) {
$tmpMetaData = $translation->metaData();
if (!is_array($tmpMetaData)) {
$tmpMetaData = array(array('id' => $attribute->attribute('contentclass_attribute_identifier'), 'text' => $tmpMetaData));
}
$metaData = array_merge($metaData, $tmpMetaData);
}
foreach ($metaData as $metaDataPart) {
$text = eZSearchEngine::normalizeText(htmlspecialchars($metaDataPart['text'], ENT_NOQUOTES, 'UTF-8'), true);
// Split text on whitespace
if (is_numeric(trim($text))) {
$integerValue = (int) $text;
} else {
$integerValue = 0;
}
$wordArray = explode(' ', $text);
foreach ($wordArray as $word) {
if (trim($word) != "") {
// words stored in search index are limited to 150 characters
if (strlen($word) > 150) {
$word = substr($word, 0, 150);
}
$indexArray[] = array('Word' => $word, 'ContentClassAttributeID' => $attribute->attribute('contentclassattribute_id'), 'identifier' => $metaDataPart['id'], 'integer_value' => $integerValue);
$indexArrayOnlyWords[$word] = 1;
$wordCount++;
//if we have "www." before word than
//treat it as url and add additional entry to the index
if (substr(strtolower($word), 0, 4) == 'www.') {
$additionalUrlWord = substr($word, 4);
$indexArray[] = array('Word' => $additionalUrlWord, 'ContentClassAttributeID' => $attribute->attribute('contentclassattribute_id'), 'identifier' => $metaDataPart['id'], 'integer_value' => $integerValue);
$indexArrayOnlyWords[$additionalUrlWord] = 1;
$wordCount++;
}
}
}
}
}
}
eZContentObject::recursionProtectionEnd();
$wordIDArray = $this->buildWordIDArray(array_keys($indexArrayOnlyWords));
$db = eZDB::instance();
$db->begin();
for ($arrayCount = 0; $arrayCount < $wordCount; $arrayCount += 1000) {
$placement = $this->indexWords($contentObject, array_slice($indexArray, $arrayCount, 1000), $wordIDArray, $placement);
}
$db->commit();
return true;
}
/**
* Normalizes the text so that it is easily parsable
*
* Used only by legacy search engine (eZSearchEngine class).
*
* @param string $text
*
* @return string
*/
public function normalizeText($text)
{
if ($this->searchHandler instanceof LegacyHandler) {
$searchEngine = new eZSearchEngine();
return $searchEngine->normalizeText($text);
}
return $text;
}
