/**
* Detect is this page are frontpage?
*
* @return boolean Is frontpage?
*/
public static function isHome()
{
$langPath = null;
$tag = null;
$lang = \JFactory::getLanguage();
// For multi language
if (\JPluginHelper::isEnabled('system', 'languagefilter')) {
$tag = $lang->getTag();
$langCodes = \JLanguageHelper::getLanguages('lang_code');
$langPath = $langCodes[$tag]->sef;
}
$uri = \JUri::getInstance();
$root = $uri::root(true);
// Get site route
$route = StringHelper::substr($uri->getPath(), StringHelper::strlen($root));
// Remove index.php
$route = str_replace('index.php', '', $route);
// If Multiple language enabled, we check first part of URI is language code or not.
if ($langPath) {
$params = ExtensionHelper::getParams('plg_system_languagefilter');
if ($tag == $lang->getDefault() && $params->get('remove_default_prefix', 0)) {
$langPath = '';
}
// If route equals lang path, means it is home route.
if (trim($route, '/') == $langPath && !$uri->getVar('option')) {
return true;
}
} else {
if (!trim($route, '/') && !$uri->getVar('option')) {
return true;
}
}
return false;
}
/**
* Returns an array of options
*
* @param string $query SQL with 'ordering' AS value and 'name field' AS text
* @param integer $chop The length of the truncated headline
*
* @return array An array of objects formatted for JHtml list processing
*
* @since 1.5
*/
public static function genericordering($query, $chop = 30)
{
$db = JFactory::getDbo();
$options = array();
$db->setQuery($query);
$items = $db->loadObjectList();
if (empty($items)) {
$options[] = JHtml::_('select.option', 1, JText::_('JOPTION_ORDER_FIRST'));
return $options;
}
$options[] = JHtml::_('select.option', 0, '0 ' . JText::_('JOPTION_ORDER_FIRST'));
for ($i = 0, $n = count($items); $i < $n; $i++) {
$items[$i]->text = JText::_($items[$i]->text);
if (StringHelper::strlen($items[$i]->text) > $chop) {
$text = StringHelper::substr($items[$i]->text, 0, $chop) . "...";
} else {
$text = $items[$i]->text;
}
$options[] = JHtml::_('select.option', $items[$i]->value, $items[$i]->value . '. ' . $text);
}
$options[] = JHtml::_('select.option', $items[$i - 1]->value + 1, $items[$i - 1]->value + 1 . ' ' . JText::_('JOPTION_ORDER_LAST'));
return $options;
}
/**
* Method to tokenize a text string.
*
* @param string $input The input to tokenize.
* @param string $lang The language of the input.
* @param boolean $phrase Flag to indicate whether input could be a phrase. [optional]
*
* @return array An array of FinderIndexerToken objects.
*
* @since 2.5
*/
public static function tokenize($input, $lang, $phrase = false)
{
static $cache;
$store = StringHelper::strlen($input) < 128 ? md5($input . '::' . $lang . '::' . $phrase) : null;
// Check if the string has been tokenized already.
if ($store && isset($cache[$store])) {
return $cache[$store];
}
$tokens = array();
$quotes = html_entity_decode('‘’'', ENT_QUOTES, 'UTF-8');
// Get the simple language key.
$lang = static::getPrimaryLanguage($lang);
/*
* Parsing the string input into terms is a multi-step process.
*
* Regexes:
* 1. Remove everything except letters, numbers, quotes, apostrophe, plus, dash, period, and comma.
* 2. Remove plus, dash, period, and comma characters located before letter characters.
* 3. Remove plus, dash, period, and comma characters located after other characters.
* 4. Remove plus, period, and comma characters enclosed in alphabetical characters. Ungreedy.
* 5. Remove orphaned apostrophe, plus, dash, period, and comma characters.
* 6. Remove orphaned quote characters.
* 7. Replace the assorted single quotation marks with the ASCII standard single quotation.
* 8. Remove multiple space characters and replaces with a single space.
*/
$input = StringHelper::strtolower($input);
$input = preg_replace('#[^\\pL\\pM\\pN\\p{Pi}\\p{Pf}\'+-.,]+#mui', ' ', $input);
$input = preg_replace('#(^|\\s)[+-.,]+([\\pL\\pM]+)#mui', ' $1', $input);
$input = preg_replace('#([\\pL\\pM\\pN]+)[+-.,]+(\\s|$)#mui', '$1 ', $input);
$input = preg_replace('#([\\pL\\pM]+)[+.,]+([\\pL\\pM]+)#muiU', '$1 $2', $input);
$input = preg_replace('#(^|\\s)[\'+-.,]+(\\s|$)#mui', ' ', $input);
$input = preg_replace('#(^|\\s)[\\p{Pi}\\p{Pf}]+(\\s|$)#mui', ' ', $input);
$input = preg_replace('#[' . $quotes . ']+#mui', '\'', $input);
$input = preg_replace('#\\s+#mui', ' ', $input);
$input = StringHelper::trim($input);
// Explode the normalized string to get the terms.
$terms = explode(' ', $input);
/*
* If we have Unicode support and are dealing with Chinese text, Chinese
* has to be handled specially because there are not necessarily any spaces
* between the "words". So, we have to test if the words belong to the Chinese
* character set and if so, explode them into single glyphs or "words".
*/
if ($lang === 'zh') {
// Iterate through the terms and test if they contain Chinese.
for ($i = 0, $n = count($terms); $i < $n; $i++) {
$charMatches = array();
$charCount = preg_match_all('#[\\p{Han}]#mui', $terms[$i], $charMatches);
// Split apart any groups of Chinese characters.
for ($j = 0; $j < $charCount; $j++) {
$tSplit = StringHelper::str_ireplace($charMatches[0][$j], '', $terms[$i], false);
if (!empty($tSplit)) {
$terms[$i] = $tSplit;
} else {
unset($terms[$i]);
}
$terms[] = $charMatches[0][$j];
}
}
// Reset array keys.
$terms = array_values($terms);
}
/*
* If we have to handle the input as a phrase, that means we don't
* tokenize the individual terms and we do not create the two and three
* term combinations. The phrase must contain more than one word!
*/
if ($phrase === true && count($terms) > 1) {
// Create tokens from the phrase.
$tokens[] = new FinderIndexerToken($terms, $lang);
} else {
// Create tokens from the terms.
for ($i = 0, $n = count($terms); $i < $n; $i++) {
$tokens[] = new FinderIndexerToken($terms[$i], $lang);
}
// Create two and three word phrase tokens from the individual words.
for ($i = 0, $n = count($tokens); $i < $n; $i++) {
// Setup the phrase positions.
$i2 = $i + 1;
$i3 = $i + 2;
// Create the two word phrase.
if ($i2 < $n && isset($tokens[$i2])) {
// Tokenize the two word phrase.
$token = new FinderIndexerToken(array($tokens[$i]->term, $tokens[$i2]->term), $lang, $lang === 'zh' ? '' : ' ');
$token->derived = true;
// Add the token to the stack.
$tokens[] = $token;
}
// Create the three word phrase.
if ($i3 < $n && isset($tokens[$i3])) {
// Tokenize the three word phrase.
$token = new FinderIndexerToken(array($tokens[$i]->term, $tokens[$i2]->term, $tokens[$i3]->term), $lang, $lang === 'zh' ? '' : ' ');
$token->derived = true;
// Add the token to the stack.
$tokens[] = $token;
}
}
}
if ($store) {
$cache[$store] = count($tokens) > 1 ? $tokens : array_shift($tokens);
return $cache[$store];
} else {
return count($tokens) > 1 ? $tokens : array_shift($tokens);
}
}
/**
* Method to construct the token object.
*
* @param mixed $term The term as a string for words or an array for phrases.
* @param string $lang The simple language identifier.
* @param string $spacer The space separator for phrases. [optional]
*
* @since 2.5
*/
public function __construct($term, $lang, $spacer = ' ')
{
$this->language = $lang;
// Tokens can be a single word or an array of words representing a phrase.
if (is_array($term)) {
// Populate the token instance.
$this->term = implode($spacer, $term);
$this->stem = implode($spacer, array_map(array('FinderIndexerHelper', 'stem'), $term, array($lang)));
$this->numeric = false;
$this->common = false;
$this->phrase = true;
$this->length = StringHelper::strlen($this->term);
/*
* Calculate the weight of the token.
*
* 1. Length of the token up to 30 and divide by 30, add 1.
* 2. Round weight to 4 decimal points.
*/
$this->weight = ($this->length >= 30 ? 30 : $this->length) / 30 + 1;
$this->weight = round($this->weight, 4);
} else {
// Populate the token instance.
$this->term = $term;
$this->stem = FinderIndexerHelper::stem($this->term, $lang);
$this->numeric = is_numeric($this->term) || (bool) preg_match('#^[0-9,.\\-\\+]+$#', $this->term);
$this->common = $this->numeric ? false : FinderIndexerHelper::isCommon($this->term, $lang);
$this->phrase = false;
$this->length = StringHelper::strlen($this->term);
/*
* Calculate the weight of the token.
*
* 1. Length of the token up to 15 and divide by 15.
* 2. If common term, divide weight by 8.
* 3. If numeric, multiply weight by 1.5.
* 4. Round weight to 4 decimal points.
*/
$this->weight = ($this->length >= 15 ? 15 : $this->length) / 15;
$this->weight = $this->common == true ? $this->weight / 8 : $this->weight;
$this->weight = $this->numeric == true ? $this->weight * 1.5 : $this->weight;
$this->weight = round($this->weight, 4);
}
}
/**
* Execute and display a template script.
*
* @param string $tpl The name of the template file to parse; automatically searches through the template paths.
*
* @return mixed A string if successful, otherwise an Error object.
*/
public function display($tpl = null)
{
JLoader::register('SearchHelper', JPATH_COMPONENT_ADMINISTRATOR . '/helpers/search.php');
$app = JFactory::getApplication();
$uri = JUri::getInstance();
$error = null;
$rows = null;
$results = null;
$total = 0;
// Get some data from the model
$areas = $this->get('areas');
$state = $this->get('state');
$searchword = $state->get('keyword');
$params = $app->getParams();
$menus = $app->getMenu();
$menu = $menus->getActive();
// Because the application sets a default page title, we need to get it right from the menu item itself
if (is_object($menu)) {
$menu_params = new Registry($menu->params);
if (!$menu_params->get('page_title')) {
$params->set('page_title', JText::_('COM_SEARCH_SEARCH'));
}
} else {
$params->set('page_title', JText::_('COM_SEARCH_SEARCH'));
}
$title = $params->get('page_title');
if ($app->get('sitename_pagetitles', 0) == 1) {
$title = JText::sprintf('JPAGETITLE', $app->get('sitename'), $title);
} elseif ($app->get('sitename_pagetitles', 0) == 2) {
$title = JText::sprintf('JPAGETITLE', $title, $app->get('sitename'));
}
$this->document->setTitle($title);
if ($params->get('menu-meta_description')) {
$this->document->setDescription($params->get('menu-meta_description'));
}
if ($params->get('menu-meta_keywords')) {
$this->document->setMetadata('keywords', $params->get('menu-meta_keywords'));
}
if ($params->get('robots')) {
$this->document->setMetadata('robots', $params->get('robots'));
}
// Built select lists
$orders = array();
$orders[] = JHtml::_('select.option', 'newest', JText::_('COM_SEARCH_NEWEST_FIRST'));
$orders[] = JHtml::_('select.option', 'oldest', JText::_('COM_SEARCH_OLDEST_FIRST'));
$orders[] = JHtml::_('select.option', 'popular', JText::_('COM_SEARCH_MOST_POPULAR'));
$orders[] = JHtml::_('select.option', 'alpha', JText::_('COM_SEARCH_ALPHABETICAL'));
$orders[] = JHtml::_('select.option', 'category', JText::_('JCATEGORY'));
$lists = array();
$lists['ordering'] = JHtml::_('select.genericlist', $orders, 'ordering', 'class="inputbox"', 'value', 'text', $state->get('ordering'));
$searchphrases = array();
$searchphrases[] = JHtml::_('select.option', 'all', JText::_('COM_SEARCH_ALL_WORDS'));
$searchphrases[] = JHtml::_('select.option', 'any', JText::_('COM_SEARCH_ANY_WORDS'));
$searchphrases[] = JHtml::_('select.option', 'exact', JText::_('COM_SEARCH_EXACT_PHRASE'));
$lists['searchphrase'] = JHtml::_('select.radiolist', $searchphrases, 'searchphrase', '', 'value', 'text', $state->get('match'));
// Log the search
JSearchHelper::logSearch($searchword, 'com_search');
// Limit searchword
$lang = JFactory::getLanguage();
$upper_limit = $lang->getUpperLimitSearchWord();
$lower_limit = $lang->getLowerLimitSearchWord();
if (SearchHelper::limitSearchWord($searchword)) {
$error = JText::sprintf('COM_SEARCH_ERROR_SEARCH_MESSAGE', $lower_limit, $upper_limit);
}
// Sanitise searchword
if (SearchHelper::santiseSearchWord($searchword, $state->get('match'))) {
$error = JText::_('COM_SEARCH_ERROR_IGNOREKEYWORD');
}
if (!$searchword && !empty($this->input) && count($this->input->post)) {
// $error = JText::_('COM_SEARCH_ERROR_ENTERKEYWORD');
}
// Put the filtered results back into the model
// for next release, the checks should be done in the model perhaps...
$state->set('keyword', $searchword);
if ($error == null) {
$results = $this->get('data');
$total = $this->get('total');
$pagination = $this->get('pagination');
JLoader::register('ContentHelperRoute', JPATH_SITE . '/components/com_content/helpers/route.php');
for ($i = 0, $count = count($results); $i < $count; $i++) {
$row =& $results[$i]->text;
if ($state->get('match') == 'exact') {
$searchwords = array($searchword);
$needle = $searchword;
} else {
$searchworda = preg_replace('#\\xE3\\x80\\x80#s', ' ', $searchword);
$searchwords = preg_split("/\\s+/u", $searchworda);
$needle = $searchwords[0];
}
$row = SearchHelper::prepareSearchContent($row, $needle);
$searchwords = array_values(array_unique($searchwords));
$srow = strtolower(SearchHelper::remove_accents($row));
$hl1 = '<span class="highlight">';
$hl2 = '</span>';
$posCollector = array();
$mbString = extension_loaded('mbstring');
if ($mbString) {
// E.g. german umlauts like ä are converted to ae and so
// $pos calculated with $srow doesn't match for $row
$correctPos = mb_strlen($srow) > mb_strlen($row);
$highlighterLen = mb_strlen($hl1 . $hl2);
} else {
// E.g. german umlauts like ä are converted to ae and so
// $pos calculated with $srow desn't match for $row
$correctPos = StringHelper::strlen($srow) > StringHelper::strlen($row);
$highlighterLen = StringHelper::strlen($hl1 . $hl2);
}
foreach ($searchwords as $hlword) {
if ($mbString) {
if (($pos = mb_strpos($srow, strtolower(SearchHelper::remove_accents($hlword)))) !== false) {
// Iconv transliterates '€' to 'EUR'
// TODO: add other expanding translations?
$eur_compensation = $pos > 0 ? substr_count($row, "€", 0, $pos) * 2 : 0;
$pos -= $eur_compensation;
if ($correctPos) {
// Calculate necessary corrections from 0 to current $pos
$ChkRow = mb_substr($row, 0, $pos);
$sChkRowLen = mb_strlen(strtolower(SearchHelper::remove_accents($ChkRow)));
$ChkRowLen = mb_strlen($ChkRow);
// Correct $pos
$pos -= $sChkRowLen - $ChkRowLen;
}
// Collect pos and searchword
$posCollector[$pos] = $hlword;
}
} else {
if (($pos = StringHelper::strpos($srow, strtolower(SearchHelper::remove_accents($hlword)))) !== false) {
// Iconv transliterates '€' to 'EUR'
// TODO: add other expanding translations?
$eur_compensation = $pos > 0 ? substr_count($row, "€", 0, $pos) * 2 : 0;
$pos -= $eur_compensation;
if ($correctPos) {
// Calculate necessary corrections from 0 to current $pos
$ChkRow = StringHelper::substr($row, 0, $pos);
$sChkRowLen = StringHelper::strlen(strtolower(SearchHelper::remove_accents($ChkRow)));
$ChkRowLen = StringHelper::strlen($ChkRow);
// Correct $pos
$pos -= $sChkRowLen - $ChkRowLen;
}
// Collect pos and searchword
$posCollector[$pos] = $hlword;
}
}
}
if (count($posCollector)) {
// Sort by pos. Easier to handle overlapping highlighter-spans
ksort($posCollector);
$cnt = 0;
$lastHighlighterEnd = -1;
foreach ($posCollector as $pos => $hlword) {
$pos += $cnt * $highlighterLen;
/* Avoid overlapping/corrupted highlighter-spans
* TODO $chkOverlap could be used to highlight remaining part
* of searchword outside last highlighter-span.
* At the moment no additional highlighter is set.*/
$chkOverlap = $pos - $lastHighlighterEnd;
if ($chkOverlap >= 0) {
// Set highlighter around searchword
if ($mbString) {
$hlwordLen = mb_strlen($hlword);
$row = mb_substr($row, 0, $pos) . $hl1 . mb_substr($row, $pos, $hlwordLen) . $hl2 . mb_substr($row, $pos + $hlwordLen);
} else {
$hlwordLen = StringHelper::strlen($hlword);
$row = StringHelper::substr($row, 0, $pos) . $hl1 . StringHelper::substr($row, $pos, StringHelper::strlen($hlword)) . $hl2 . StringHelper::substr($row, $pos + StringHelper::strlen($hlword));
}
$cnt++;
$lastHighlighterEnd = $pos + $hlwordLen + $highlighterLen;
}
}
}
$result =& $results[$i];
if ($result->created) {
$created = JHtml::_('date', $result->created, JText::_('DATE_FORMAT_LC3'));
} else {
$created = '';
}
$result->text = JHtml::_('content.prepare', $result->text, '', 'com_search.search');
$result->created = $created;
$result->count = $i + 1;
}
}
// Check for layout override
$active = JFactory::getApplication()->getMenu()->getActive();
if (isset($active->query['layout'])) {
$this->setLayout($active->query['layout']);
}
// Escape strings for HTML output
$this->pageclass_sfx = htmlspecialchars($params->get('pageclass_sfx'));
$this->pagination =& $pagination;
$this->results =& $results;
$this->lists =& $lists;
$this->params =& $params;
$this->ordering = $state->get('ordering');
$this->searchword = $searchword;
$this->origkeyword = $state->get('origkeyword');
$this->searchphrase = $state->get('match');
$this->searchareas = $areas;
$this->total = $total;
$this->error = $error;
$this->action = $uri;
parent::display($tpl);
}
/**
* Returns substring of characters around a searchword.
*
* @param string $text The source string.
* @param integer $searchword Number of chars to return.
*
* @return string
*
* @since 1.5
*/
public static function _smartSubstr($text, $searchword)
{
$lang = JFactory::getLanguage();
$length = $lang->getSearchDisplayedCharactersNumber();
$ltext = self::remove_accents($text);
$textlen = StringHelper::strlen($ltext);
$lsearchword = StringHelper::strtolower(self::remove_accents($searchword));
$wordfound = false;
$pos = 0;
while ($wordfound === false && $pos < $textlen) {
if (($wordpos = @StringHelper::strpos($ltext, ' ', $pos + $length)) !== false) {
$chunk_size = $wordpos - $pos;
} else {
$chunk_size = $length;
}
$chunk = StringHelper::substr($ltext, $pos, $chunk_size);
$wordfound = StringHelper::strpos(StringHelper::strtolower($chunk), $lsearchword);
if ($wordfound === false) {
$pos += $chunk_size + 1;
}
}
if ($wordfound !== false) {
return ($pos > 0 ? '... ' : '') . StringHelper::substr($text, $pos, $chunk_size) . ' ...';
} else {
if (($wordpos = @StringHelper::strpos($text, ' ', $length)) !== false) {
return StringHelper::substr($text, 0, $wordpos) . ' ...';
} else {
return StringHelper::substr($text, 0, $length);
}
}
}
/**
* Method to process the query input string and extract required, optional,
* and excluded tokens; taxonomy filters; and date filters.
*
* @param string $input The query input string.
* @param string $lang The query input language.
* @param string $mode The query matching mode.
*
* @return boolean True on success.
*
* @since 2.5
* @throws Exception on database error.
*/
protected function processString($input, $lang, $mode)
{
// Clean up the input string.
$input = html_entity_decode($input, ENT_QUOTES, 'UTF-8');
$input = StringHelper::strtolower($input);
$input = preg_replace('#\\s+#mi', ' ', $input);
$input = StringHelper::trim($input);
$debug = JFactory::getConfig()->get('debug_lang');
/*
* First, we need to handle string based modifiers. String based
* modifiers could potentially include things like "category:blah" or
* "before:2009-10-21" or "type:article", etc.
*/
$patterns = array('before' => JText::_('COM_FINDER_FILTER_WHEN_BEFORE'), 'after' => JText::_('COM_FINDER_FILTER_WHEN_AFTER'));
// Add the taxonomy branch titles to the possible patterns.
foreach (FinderIndexerTaxonomy::getBranchTitles() as $branch) {
// Add the pattern.
$patterns[$branch] = StringHelper::strtolower(JText::_(FinderHelperLanguage::branchSingular($branch)));
}
// Container for search terms and phrases.
$terms = array();
$phrases = array();
// Cleared filter branches.
$cleared = array();
/*
* Compile the suffix pattern. This is used to match the values of the
* filter input string. Single words can be input directly, multi-word
* values have to be wrapped in double quotes.
*/
$quotes = html_entity_decode('‘’'', ENT_QUOTES, 'UTF-8');
$suffix = '(([\\w\\d' . $quotes . '-]+)|\\"([\\w\\d\\s' . $quotes . '-]+)\\")';
/*
* Iterate through the possible filter patterns and search for matches.
* We need to match the key, colon, and a value pattern for the match
* to be valid.
*/
foreach ($patterns as $modifier => $pattern) {
$matches = array();
if ($debug) {
$pattern = substr($pattern, 2, -2);
}
// Check if the filter pattern is in the input string.
if (preg_match('#' . $pattern . '\\s*:\\s*' . $suffix . '#mi', $input, $matches)) {
// Get the value given to the modifier.
$value = isset($matches[3]) ? $matches[3] : $matches[1];
// Now we have to handle the filter string.
switch ($modifier) {
// Handle a before and after date filters.
case 'before':
case 'after':
// Get the time offset.
$offset = JFactory::getApplication()->get('offset');
// Array of allowed when values.
$whens = array('before', 'after', 'exact');
// The value of 'today' is a special case that we need to handle.
if ($value === StringHelper::strtolower(JText::_('COM_FINDER_QUERY_FILTER_TODAY'))) {
$value = JFactory::getDate('now', $offset)->format('%Y-%m-%d');
}
// Try to parse the date string.
$date = JFactory::getDate($value, $offset);
// Check if the date was parsed successfully.
if ($date->toUnix() !== null) {
// Set the date filter.
$this->date1 = $date->toSql();
$this->when1 = in_array($modifier, $whens) ? $modifier : 'before';
}
break;
// Handle a taxonomy branch filter.
// Handle a taxonomy branch filter.
default:
// Try to find the node id.
$return = FinderIndexerTaxonomy::getNodeByTitle($modifier, $value);
// Check if the node id was found.
if ($return) {
// Check if the branch has been cleared.
if (!in_array($modifier, $cleared)) {
// Clear the branch.
$this->filters[$modifier] = array();
// Add the branch to the cleared list.
$cleared[] = $modifier;
}
// Add the filter to the list.
$this->filters[$modifier][$return->title] = (int) $return->id;
}
break;
}
// Clean up the input string again.
$input = str_replace($matches[0], '', $input);
$input = preg_replace('#\\s+#mi', ' ', $input);
$input = StringHelper::trim($input);
}
}
/*
* Extract the tokens enclosed in double quotes so that we can handle
* them as phrases.
*/
if (StringHelper::strpos($input, '"') !== false) {
$matches = array();
// Extract the tokens enclosed in double quotes.
if (preg_match_all('#\\"([^"]+)\\"#mi', $input, $matches)) {
/*
* One or more phrases were found so we need to iterate through
* them, tokenize them as phrases, and remove them from the raw
* input string before we move on to the next processing step.
*/
foreach ($matches[1] as $key => $match) {
// Find the complete phrase in the input string.
$pos = StringHelper::strpos($input, $matches[0][$key]);
$len = StringHelper::strlen($matches[0][$key]);
// Add any terms that are before this phrase to the stack.
if (StringHelper::trim(StringHelper::substr($input, 0, $pos))) {
$terms = array_merge($terms, explode(' ', StringHelper::trim(StringHelper::substr($input, 0, $pos))));
}
// Strip out everything up to and including the phrase.
$input = StringHelper::substr($input, $pos + $len);
// Clean up the input string again.
$input = preg_replace('#\\s+#mi', ' ', $input);
$input = StringHelper::trim($input);
// Get the number of words in the phrase.
$parts = explode(' ', $match);
// Check if the phrase is longer than three words.
if (count($parts) > 3) {
/*
* If the phrase is longer than three words, we need to
* break it down into smaller chunks of phrases that
* are less than or equal to three words. We overlap
* the chunks so that we can ensure that a match is
* found for the complete phrase and not just portions
* of it.
*/
for ($i = 0, $c = count($parts); $i < $c; $i += 2) {
// Set up the chunk.
$chunk = array();
// The chunk has to be assembled based on how many
// pieces are available to use.
switch ($c - $i) {
/*
* If only one word is left, we can break from
* the switch and loop because the last word
* was already used at the end of the last
* chunk.
*/
case 1:
break 2;
// If there words are left, we use them both as
// the last chunk of the phrase and we're done.
// If there words are left, we use them both as
// the last chunk of the phrase and we're done.
case 2:
$chunk[] = $parts[$i];
$chunk[] = $parts[$i + 1];
break;
// If there are three or more words left, we
// build a three word chunk and continue on.
// If there are three or more words left, we
// build a three word chunk and continue on.
default:
$chunk[] = $parts[$i];
$chunk[] = $parts[$i + 1];
$chunk[] = $parts[$i + 2];
break;
}
// If the chunk is not empty, add it as a phrase.
if (count($chunk)) {
$phrases[] = implode(' ', $chunk);
$terms[] = implode(' ', $chunk);
}
}
} else {
// The phrase is <= 3 words so we can use it as is.
$phrases[] = $match;
$terms[] = $match;
}
}
}
}
// Add the remaining terms if present.
if (!empty($input)) {
$terms = array_merge($terms, explode(' ', $input));
}
// An array of our boolean operators. $operator => $translation
$operators = array('AND' => StringHelper::strtolower(JText::_('COM_FINDER_QUERY_OPERATOR_AND')), 'OR' => StringHelper::strtolower(JText::_('COM_FINDER_QUERY_OPERATOR_OR')), 'NOT' => StringHelper::strtolower(JText::_('COM_FINDER_QUERY_OPERATOR_NOT')));
// If language debugging is enabled you need to ignore the debug strings in matching.
if (JDEBUG) {
$debugStrings = array('**', '??');
$operators = str_replace($debugStrings, '', $operators);
}
/*
* Iterate through the terms and perform any sorting that needs to be
* done based on boolean search operators. Terms that are before an
* and/or/not modifier have to be handled in relation to their operator.
*/
for ($i = 0, $c = count($terms); $i < $c; $i++) {
// Check if the term is followed by an operator that we understand.
if (isset($terms[$i + 1]) && in_array($terms[$i + 1], $operators)) {
// Get the operator mode.
$op = array_search($terms[$i + 1], $operators);
// Handle the AND operator.
if ($op === 'AND' && isset($terms[$i + 2])) {
// Tokenize the current term.
$token = FinderIndexerHelper::tokenize($terms[$i], $lang, true);
$token = $this->getTokenData($token);
// Set the required flag.
$token->required = true;
// Add the current token to the stack.
$this->included[] = $token;
$this->highlight = array_merge($this->highlight, array_keys($token->matches));
// Skip the next token (the mode operator).
$this->operators[] = $terms[$i + 1];
// Tokenize the term after the next term (current plus two).
$other = FinderIndexerHelper::tokenize($terms[$i + 2], $lang, true);
$other = $this->getTokenData($other);
// Set the required flag.
$other->required = true;
// Add the token after the next token to the stack.
$this->included[] = $other;
$this->highlight = array_merge($this->highlight, array_keys($other->matches));
// Remove the processed phrases if possible.
if (($pk = array_search($terms[$i], $phrases)) !== false) {
unset($phrases[$pk]);
}
if (($pk = array_search($terms[$i + 2], $phrases)) !== false) {
unset($phrases[$pk]);
}
// Remove the processed terms.
unset($terms[$i]);
unset($terms[$i + 1]);
unset($terms[$i + 2]);
// Adjust the loop.
$i += 2;
continue;
} elseif ($op === 'OR' && isset($terms[$i + 2])) {
// Tokenize the current term.
$token = FinderIndexerHelper::tokenize($terms[$i], $lang, true);
$token = $this->getTokenData($token);
// Set the required flag.
$token->required = false;
// Add the current token to the stack.
if (count($token->matches)) {
$this->included[] = $token;
$this->highlight = array_merge($this->highlight, array_keys($token->matches));
} else {
$this->ignored[] = $token;
}
// Skip the next token (the mode operator).
$this->operators[] = $terms[$i + 1];
// Tokenize the term after the next term (current plus two).
$other = FinderIndexerHelper::tokenize($terms[$i + 2], $lang, true);
$other = $this->getTokenData($other);
// Set the required flag.
$other->required = false;
// Add the token after the next token to the stack.
if (count($other->matches)) {
$this->included[] = $other;
$this->highlight = array_merge($this->highlight, array_keys($other->matches));
} else {
$this->ignored[] = $other;
}
// Remove the processed phrases if possible.
if (($pk = array_search($terms[$i], $phrases)) !== false) {
unset($phrases[$pk]);
}
if (($pk = array_search($terms[$i + 2], $phrases)) !== false) {
unset($phrases[$pk]);
}
// Remove the processed terms.
unset($terms[$i]);
unset($terms[$i + 1]);
unset($terms[$i + 2]);
// Adjust the loop.
$i += 2;
continue;
}
} elseif (isset($terms[$i + 1]) && array_search($terms[$i], $operators) === 'OR') {
// Skip the next token (the mode operator).
$this->operators[] = $terms[$i];
// Tokenize the next term (current plus one).
$other = FinderIndexerHelper::tokenize($terms[$i + 1], $lang, true);
$other = $this->getTokenData($other);
// Set the required flag.
$other->required = false;
// Add the token after the next token to the stack.
if (count($other->matches)) {
$this->included[] = $other;
$this->highlight = array_merge($this->highlight, array_keys($other->matches));
} else {
$this->ignored[] = $other;
}
// Remove the processed phrase if possible.
if (($pk = array_search($terms[$i + 1], $phrases)) !== false) {
unset($phrases[$pk]);
}
// Remove the processed terms.
unset($terms[$i]);
unset($terms[$i + 1]);
// Adjust the loop.
$i++;
continue;
} elseif (isset($terms[$i + 1]) && array_search($terms[$i], $operators) === 'NOT') {
// Skip the next token (the mode operator).
$this->operators[] = $terms[$i];
// Tokenize the next term (current plus one).
$other = FinderIndexerHelper::tokenize($terms[$i + 1], $lang, true);
$other = $this->getTokenData($other);
// Set the required flag.
$other->required = false;
// Add the next token to the stack.
if (count($other->matches)) {
$this->excluded[] = $other;
} else {
$this->ignored[] = $other;
}
// Remove the processed phrase if possible.
if (($pk = array_search($terms[$i + 1], $phrases)) !== false) {
unset($phrases[$pk]);
}
// Remove the processed terms.
unset($terms[$i]);
unset($terms[$i + 1]);
// Adjust the loop.
$i++;
continue;
}
}
/*
* Iterate through any search phrases and tokenize them. We handle
* phrases as autonomous units and do not break them down into two and
* three word combinations.
*/
for ($i = 0, $c = count($phrases); $i < $c; $i++) {
// Tokenize the phrase.
$token = FinderIndexerHelper::tokenize($phrases[$i], $lang, true);
$token = $this->getTokenData($token);
// Set the required flag.
$token->required = true;
// Add the current token to the stack.
$this->included[] = $token;
$this->highlight = array_merge($this->highlight, array_keys($token->matches));
// Remove the processed term if possible.
if (($pk = array_search($phrases[$i], $terms)) !== false) {
unset($terms[$pk]);
}
// Remove the processed phrase.
unset($phrases[$i]);
}
/*
* Handle any remaining tokens using the standard processing mechanism.
*/
if (!empty($terms)) {
// Tokenize the terms.
$terms = implode(' ', $terms);
$tokens = FinderIndexerHelper::tokenize($terms, $lang, false);
// Make sure we are working with an array.
$tokens = is_array($tokens) ? $tokens : array($tokens);
// Get the token data and required state for all the tokens.
foreach ($tokens as $token) {
// Get the token data.
$token = $this->getTokenData($token);
// Set the required flag for the token.
$token->required = $mode === 'AND' ? $token->phrase ? false : true : false;
// Add the token to the appropriate stack.
if (count($token->matches) || $token->required) {
$this->included[] = $token;
$this->highlight = array_merge($this->highlight, array_keys($token->matches));
} else {
$this->ignored[] = $token;
}
}
}
return true;
}
/**
* Abridges text strings over the specified character limit. The
* behavior will insert an ellipsis into the text replacing a section
* of variable size to ensure the string does not exceed the defined
* maximum length. This method is UTF-8 safe.
*
* For example, it transforms "Really long title" to "Really...title".
*
* Note that this method does not scan for HTML tags so will potentially break them.
*
* @param string $text The text to abridge.
* @param integer $length The maximum length of the text (default is 50).
* @param integer $intro The maximum length of the intro text (default is 30).
*
* @return string The abridged text.
*
* @since 1.6
*/
public static function abridge($text, $length = 50, $intro = 30)
{
// Abridge the item text if it is too long.
if (StringHelper::strlen($text) > $length) {
// Determine the remaining text length.
$remainder = $length - ($intro + 3);
// Extract the beginning and ending text sections.
$beg = StringHelper::substr($text, 0, $intro);
$end = StringHelper::substr($text, StringHelper::strlen($text) - $remainder);
// Build the resulting string.
$text = $beg . '...' . $end;
}
return $text;
}
/**
* Create workarounds for data to be cached
*
* @param string $data Cached data
* @param array $options Array of options
*
* @return string Data to be cached
*
* @since 11.1
*/
public static function setWorkarounds($data, $options = array())
{
$loptions = array('nopathway' => 0, 'nohead' => 0, 'nomodules' => 0, 'modulemode' => 0);
if (isset($options['nopathway'])) {
$loptions['nopathway'] = $options['nopathway'];
}
if (isset($options['nohead'])) {
$loptions['nohead'] = $options['nohead'];
}
if (isset($options['nomodules'])) {
$loptions['nomodules'] = $options['nomodules'];
}
if (isset($options['modulemode'])) {
$loptions['modulemode'] = $options['modulemode'];
}
$app = JFactory::getApplication();
$document = JFactory::getDocument();
if ($loptions['nomodules'] != 1) {
// Get the modules buffer before component execution.
$buffer1 = $document->getBuffer();
if (!is_array($buffer1)) {
$buffer1 = array();
}
// Make sure the module buffer is an array.
if (!isset($buffer1['module']) || !is_array($buffer1['module'])) {
$buffer1['module'] = array();
}
}
// View body data
$cached['body'] = $data;
// Document head data
if ($loptions['nohead'] != 1 && method_exists($document, 'getHeadData')) {
if ($loptions['modulemode'] == 1) {
$headnow = $document->getHeadData();
$unset = array('title', 'description', 'link', 'links', 'metaTags');
foreach ($unset as $un) {
unset($headnow[$un]);
unset($options['headerbefore'][$un]);
}
$cached['head'] = array();
// Only store what this module has added
foreach ($headnow as $now => $value) {
if (isset($options['headerbefore'][$now])) {
// We have to serialize the content of the arrays because the may contain other arrays which is a notice in PHP 5.4 and newer
$nowvalue = array_map('serialize', $headnow[$now]);
$beforevalue = array_map('serialize', $options['headerbefore'][$now]);
$newvalue = array_diff_assoc($nowvalue, $beforevalue);
$newvalue = array_map('unserialize', $newvalue);
// Special treatment for script and style declarations.
if (($now == 'script' || $now == 'style') && is_array($newvalue) && is_array($options['headerbefore'][$now])) {
foreach ($newvalue as $type => $currentScriptStr) {
if (isset($options['headerbefore'][$now][strtolower($type)])) {
$oldScriptStr = $options['headerbefore'][$now][strtolower($type)];
if ($oldScriptStr != $currentScriptStr) {
// Save only the appended declaration.
$newvalue[strtolower($type)] = StringHelper::substr($currentScriptStr, StringHelper::strlen($oldScriptStr));
}
}
}
}
} else {
$newvalue = $headnow[$now];
}
if (!empty($newvalue)) {
$cached['head'][$now] = $newvalue;
}
}
} else {
$cached['head'] = $document->getHeadData();
}
}
// Document MIME encoding
$cached['mime_encoding'] = $document->getMimeEncoding();
// Pathway data
if ($app->isSite() && $loptions['nopathway'] != 1) {
$cached['pathway'] = is_array($data) && isset($data['pathway']) ? $data['pathway'] : $app->getPathway()->getPathway();
}
if ($loptions['nomodules'] != 1) {
// @todo Check if the following is needed, seems like it should be in page cache
// Get the module buffer after component execution.
$buffer2 = $document->getBuffer();
if (!is_array($buffer2)) {
$buffer2 = array();
}
// Make sure the module buffer is an array.
if (!isset($buffer2['module']) || !is_array($buffer2['module'])) {
$buffer2['module'] = array();
}
// Compare the second module buffer against the first buffer.
$cached['module'] = array_diff_assoc($buffer2['module'], $buffer1['module']);
}
// Headers data
if (isset($options['headers']) && $options['headers']) {
$cached['headers'] = $app->getHeaders();
}
return $cached;
}
/**
* @package Joomla.Site
* @subpackage com_finder
*
* @copyright Copyright (C) 2005 - 2016 Open Source Matters, Inc. All rights reserved.
* @license GNU General Public License version 2 or later; see LICENSE.txt
*/
defined('_JEXEC') or die;
use Joomla\String\StringHelper;
// Get the mime type class.
$mime = !empty($this->result->mime) ? 'mime-' . $this->result->mime : null;
$show_description = $this->params->get('show_description', 1);
if ($show_description) {
// Calculate number of characters to display around the result
$term_length = StringHelper::strlen($this->query->input);
$desc_length = $this->params->get('description_length', 255);
$pad_length = $term_length < $desc_length ? (int) floor(($desc_length - $term_length) / 2) : 0;
// Find the position of the search term
$pos = $term_length ? StringHelper::strpos(StringHelper::strtolower($this->result->description), StringHelper::strtolower($this->query->input)) : false;
// Find a potential start point
$start = $pos && $pos > $pad_length ? $pos - $pad_length : 0;
// Find a space between $start and $pos, start right after it.
$space = StringHelper::strpos($this->result->description, ' ', $start > 0 ? $start - 1 : 0);
$start = $space && $space < $pos ? $space + 1 : $start;
$description = JHtml::_('string.truncate', StringHelper::substr($this->result->description, $start), $desc_length, true);
}
$route = $this->result->route;
// Get the route with highlighting information.
if (!empty($this->query->highlight) && empty($this->result->mime) && $this->params->get('highlight_terms', 1) && JPluginHelper::isEnabled('system', 'highlight')) {
$route .= '&highlight=' . base64_encode(json_encode($this->query->highlight));
