//MERGE branche release
This commit is contained in:
+39
-13
@@ -93,8 +93,12 @@ define('PREG_CLASS_CJK', '\x{3041}-\x{30ff}\x{31f0}-\x{31ff}\x{3400}-\x{4db5}\x{
|
||||
|
||||
class SearchCore
|
||||
{
|
||||
public static function sanitize($string, $id_lang, $indexation = false)
|
||||
public static function sanitize($string, $id_lang, $indexation = false, $iso_code = false)
|
||||
{
|
||||
$string = trim($string);
|
||||
if (empty($string))
|
||||
return '';
|
||||
|
||||
$string = Tools::strtolower(strip_tags($string));
|
||||
$string = html_entity_decode($string, ENT_NOQUOTES, 'utf-8');
|
||||
|
||||
@@ -138,14 +142,34 @@ class SearchCore
|
||||
|
||||
if ($indexation)
|
||||
{
|
||||
$minWordLen = (int)Configuration::get('PS_SEARCH_MINWORDLEN');
|
||||
if ($minWordLen > 1)
|
||||
// If the language is constituted with symbol and there is no "words", then split every chars
|
||||
if (in_array($iso_code, array('zh', 'tw', 'ja')) && function_exists('mb_strlen'))
|
||||
{
|
||||
$minWordLen -= 1;
|
||||
$string = preg_replace('/(?<=\s)[^\s]{1,'.$minWordLen.'}(?=\s)/Su', ' ', $string);
|
||||
$string = preg_replace('/^[^\s]{1,'.$minWordLen.'}(?=\s)/Su', '', $string);
|
||||
$string = preg_replace('/(?<=\s)[^\s]{1,'.$minWordLen.'}$/Su', '', $string);
|
||||
$string = preg_replace('/^[^\s]{1,'.$minWordLen.'}$/Su', '', $string);
|
||||
// Cut symbols from letters
|
||||
$symbols = '';
|
||||
$letters = '';
|
||||
foreach (explode(' ', $string) as $mb_word)
|
||||
if (strlen(Tools::replaceAccentedChars($mb_word)) == mb_strlen(Tools::replaceAccentedChars($mb_word)))
|
||||
$letters .= $mb_word.' ';
|
||||
else
|
||||
$symbols .= $mb_word.' ';
|
||||
|
||||
if (preg_match_all('/./u', $symbols, $matches))
|
||||
$symbols = implode(' ', $matches[0]);
|
||||
|
||||
$string = $letters.$symbols;
|
||||
}
|
||||
else
|
||||
{
|
||||
$minWordLen = (int)Configuration::get('PS_SEARCH_MINWORDLEN');
|
||||
if ($minWordLen > 1)
|
||||
{
|
||||
$minWordLen -= 1;
|
||||
$string = preg_replace('/(?<=\s)[^\s]{1,'.$minWordLen.'}(?=\s)/Su', ' ', $string);
|
||||
$string = preg_replace('/^[^\s]{1,'.$minWordLen.'}(?=\s)/Su', '', $string);
|
||||
$string = preg_replace('/(?<=\s)[^\s]{1,'.$minWordLen.'}$/Su', '', $string);
|
||||
$string = preg_replace('/^[^\s]{1,'.$minWordLen.'}$/Su', '', $string);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
@@ -175,7 +199,7 @@ class SearchCore
|
||||
|
||||
$intersect_array = array();
|
||||
$score_array = array();
|
||||
$words = explode(' ', Search::sanitize($expr, $id_lang));
|
||||
$words = explode(' ', Search::sanitize($expr, $id_lang, false, $context->language->iso_code));
|
||||
|
||||
foreach ($words as $key => $word)
|
||||
if (!empty($word) && strlen($word) >= (int)Configuration::get('PS_SEARCH_MINWORDLEN'))
|
||||
@@ -383,7 +407,7 @@ class SearchCore
|
||||
|
||||
return Db::getInstance()->executeS('
|
||||
SELECT p.id_product, pl.id_lang, pl.id_shop, pl.name pname, p.reference, p.ean13, p.upc,
|
||||
pl.description_short, pl.description, cl.name cname, m.name mname
|
||||
pl.description_short, pl.description, cl.name cname, m.name mname, l.iso_code
|
||||
FROM '._DB_PREFIX_.'product p
|
||||
LEFT JOIN '._DB_PREFIX_.'product_lang pl
|
||||
ON p.id_product = pl.id_product
|
||||
@@ -392,6 +416,8 @@ class SearchCore
|
||||
ON (cl.id_category = product_shop.id_category_default AND pl.id_lang = cl.id_lang AND cl.id_shop = product_shop.id_shop)
|
||||
LEFT JOIN '._DB_PREFIX_.'manufacturer m
|
||||
ON m.id_manufacturer = p.id_manufacturer
|
||||
LEFT JOIN '._DB_PREFIX_.'lang l
|
||||
ON l.id_lang = pl.id_lang
|
||||
WHERE product_shop.indexed = 0
|
||||
AND product_shop.visibility IN ("both", "search")
|
||||
'.($id_product ? 'AND p.id_product = '.(int)$id_product : '').'
|
||||
@@ -452,7 +478,6 @@ class SearchCore
|
||||
// Those are kind of global variables required to save the processed data in the database every X occurrences, in order to avoid overloading MySQL
|
||||
$count_words = 0;
|
||||
$query_array3 = array();
|
||||
$products_array = array();
|
||||
|
||||
// Every indexed words are cached into a PHP array
|
||||
$word_ids = $db->executeS('
|
||||
@@ -472,6 +497,7 @@ class SearchCore
|
||||
// Products are processed 50 by 50 in order to avoid overloading MySQL
|
||||
while (($products = Search::getProductsToIndex($total_languages, $id_product, 50)) && (count($products) > 0))
|
||||
{
|
||||
$products_array = array();
|
||||
// Now each non-indexed product is processed one by one, langage by langage
|
||||
foreach ($products as $product)
|
||||
{
|
||||
@@ -482,9 +508,9 @@ class SearchCore
|
||||
// Data must be cleaned of html, bad characters, spaces and anything, then if the resulting words are long enough, they're added to the array
|
||||
$product_array = array();
|
||||
foreach ($product as $key => $value)
|
||||
if (strncmp($key, 'id_', 3))
|
||||
if (strncmp($key, 'id_', 3) && isset($weight_array[$key]))
|
||||
{
|
||||
$words = explode(' ', Search::sanitize($value, (int)$product['id_lang'], true));
|
||||
$words = explode(' ', Search::sanitize($value, (int)$product['id_lang'], true, $product['iso_code']));
|
||||
foreach ($words as $word)
|
||||
if (!empty($word))
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user