/**
* Convert character set and HTML entities in the value of input content array keys
*
* @param array Standard content array
* @param string Charset of the input content (converted to utf-8)
* @return void
*/
public function charsetEntity2utf8(&$contentArr, $charset)
{
// Convert charset if necessary
foreach ($contentArr as $key => $value) {
if ((string) $contentArr[$key] !== '') {
if ($charset !== 'utf-8') {
$contentArr[$key] = $this->csObj->utf8_encode($contentArr[$key], $charset);
}
// decode all numeric / html-entities in the string to real characters:
$contentArr[$key] = $this->csObj->entities_to_utf8($contentArr[$key], TRUE);
}
}
}
/**
* Splits the search word input into an array where each word is represented by an array with key "sword" holding the search word and key "oper" holding the SQL operator (eg. AND, OR)
*
* Only words with 2 or more characters are accepted
* Max 200 chars total
* Space is used to split words, "" can be used search for a whole string
* AND, OR and NOT are prefix words, overruling the default operator
* +/|/- equals AND, OR and NOT as operators.
* All search words are converted to lowercase.
*
* $defOp is the default operator. 1=OR, 0=AND
*
* @param bool $defOp If TRUE, the default operator will be OR, not AND
* @return array Returns array with search words if any found
*/
public function getSearchWords($defOp)
{
// Shorten search-word string to max 200 bytes (does NOT take multibyte charsets into account - but never mind, shortening the string here is only a run-away feature!)
$inSW = substr($this->piVars['sword'], 0, 200);
// Convert to UTF-8 + conv. entities (was also converted during indexing!)
$inSW = $this->charsetConverter->conv($inSW, $this->frontendController->metaCharset, 'utf-8');
$inSW = $this->charsetConverter->entities_to_utf8($inSW);
$sWordArray = false;
if ($hookObj = $this->hookRequest('getSearchWords')) {
$sWordArray = $hookObj->getSearchWords_splitSWords($inSW, $defOp);
} else {
if ($this->piVars['type'] == 20) {
// type = Sentence
$sWordArray = array(array('sword' => trim($inSW), 'oper' => 'AND'));
} else {
$searchWords = \TYPO3\CMS\IndexedSearch\Utility\IndexedSearchUtility::getExplodedSearchString($inSW, $defOp == 1 ? 'OR' : 'AND', $this->operator_translate_table);
if (is_array($searchWords)) {
$sWordArray = $this->procSearchWordsByLexer($searchWords);
}
}
}
return $sWordArray;
}
/**
* Splits the search word input into an array where each word is represented by an array with key "sword"
* holding the search word and key "oper" holding the SQL operator (eg. AND, OR)
*
* Only words with 2 or more characters are accepted
* Max 200 chars total
* Space is used to split words, "" can be used search for a whole string
* AND, OR and NOT are prefix words, overruling the default operator
* +/|/- equals AND, OR and NOT as operators.
* All search words are converted to lowercase.
*
* $defOp is the default operator. 1=OR, 0=AND
*
* @param bool $defaultOperator If TRUE, the default operator will be OR, not AND
* @return array Search words if any found
*/
protected function getSearchWords($defaultOperator)
{
// Shorten search-word string to max 200 bytes (does NOT take multibyte charsets into account - but never mind,
// shortening the string here is only a run-away feature!)
$searchWords = substr($this->sword, 0, 200);
// Convert to UTF-8 + conv. entities (was also converted during indexing!)
$searchWords = $this->charsetConverter->conv($searchWords, $GLOBALS['TSFE']->metaCharset, 'utf-8');
$searchWords = $this->charsetConverter->entities_to_utf8($searchWords);
$sWordArray = false;
if ($hookObj = $this->hookRequest('getSearchWords')) {
$sWordArray = $hookObj->getSearchWords_splitSWords($searchWords, $defaultOperator);
} else {
// sentence
if ($this->searchData['searchType'] == 20) {
$sWordArray = array(array('sword' => trim($searchWords), 'oper' => 'AND'));
} else {
// case-sensitive. Defines the words, which will be
// operators between words
$operatorTranslateTable = array(array('+', 'AND'), array('|', 'OR'), array('-', 'AND NOT'), array($this->charsetConverter->conv_case('utf-8', LocalizationUtility::translate('localizedOperandAnd', 'IndexedSearch'), 'toLower'), 'AND'), array($this->charsetConverter->conv_case('utf-8', LocalizationUtility::translate('localizedOperandOr', 'IndexedSearch'), 'toLower'), 'OR'), array($this->charsetConverter->conv_case('utf-8', LocalizationUtility::translate('localizedOperandNot', 'IndexedSearch'), 'toLower'), 'AND NOT'));
$swordArray = \TYPO3\CMS\IndexedSearch\Utility\IndexedSearchUtility::getExplodedSearchString($searchWords, $defaultOperator == 1 ? 'OR' : 'AND', $operatorTranslateTable);
if (is_array($swordArray)) {
$sWordArray = $this->procSearchWordsByLexer($swordArray);
}
}
}
return $sWordArray;
}
