| static public regexp_split ( $pattern, $subject, $limit ) : array | ||
| $pattern | string Regular expression | |
| $subject | string String to apply regular expression to | |
| $limit | int Number of times to match; -1 for unlimited | |
| return | array | Resulting string segments |
/**
* Split a string into a clean array of keywords
* @param $text string
* @param $allowWildcards boolean
* @return array of keywords
*/
static function filterKeywords($text, $allowWildcards = false)
{
$minLength = Config::getVar('search', 'min_word_length');
$stopwords = self::_loadStopwords();
// Join multiple lines into a single string
if (is_array($text)) {
$text = join("\n", $text);
}
$cleanText = Core::cleanVar($text);
// Remove punctuation
$cleanText = PKPString::regexp_replace('/[!"\\#\\$%\'\\(\\)\\.\\?@\\[\\]\\^`\\{\\}~]/', '', $cleanText);
$cleanText = PKPString::regexp_replace('/[\\+,:;&\\/<=>\\|\\\\]/', ' ', $cleanText);
$cleanText = PKPString::regexp_replace('/[\\*]/', $allowWildcards ? '%' : ' ', $cleanText);
$cleanText = PKPString::strtolower($cleanText);
// Split into words
$words = PKPString::regexp_split('/\\s+/', $cleanText);
// FIXME Do not perform further filtering for some fields, e.g., author names?
// Remove stopwords
$keywords = array();
foreach ($words as $k) {
if (!isset($stopwords[$k]) && PKPString::strlen($k) >= $minLength && !is_numeric($k)) {
$keywords[] = PKPString::substr($k, 0, SEARCH_KEYWORD_MAX_LENGTH);
}
}
return $keywords;
}
