PHP Sensitive Word filtering

/** * Forbidden Word Filter * Execution efficiency: 0.05 seconds per piece * @author Liuxu * */ Class Logic_blackword { Const App_forum = 1; Const APP_BLOG = 2; Const APP_VOTE = 3; /** * Filter to get forbidden words * @param unknown $txt * @return ambigous */ Public Function gethitlist ($txt) { $hitList = Array (); Filtering Forbidden words in batches $max = $this->getmax (); if ($max) { $size = 1000; $last = Ceil ($max/$size); for ($page =1; $page <= $last; $page + +) { $result = $this->gethitlistbypage ($txt, $page, $size); if ($result) $hitList = Array_merge ($hitList, $result); } } $hitList 2 = array (); foreach ($hitList as $hit = $type) { $hitList 2[$type] [] = $hit; } return $hitList 2; } Private Function Getmax () { $redis = Rds::factory (); $memKey = ' Blackword_max '; $max = $redis->get ($memKey); if ($max ===false) { $max = 0; $blackWord = new Model_blackword_blackword (); $para [' field '] = "max (ID) as Max"; $result = $blackWord->search ($para); if (Isset ($result [0][' Max ')) $max = $result [0][' Max ']; $redis->setex ($memKey, $max); } return $max; } /** * Batch filter to get banned words * @param unknown $txt * @param number $page * @param number $size * @return Multitype:ambigous */ Private Function Gethitlistbypage ($txt, $page =1, $size =1000) { $hitList = Array (); To get the forbidden word tree in batches $wordTree = $this->getwordtreebypage ($page, $size); $txt = Strip_tags ($txt); $txt = preg_replace ('/[^a-za-z0-9\\x{4e00}-\\x{9fa5}]/iu ', ' ', $txt); $len = Mb_strlen ($txt, ' UTF-8 '); for ($i =0; $i < $len; $i + +) { $char = Mb_substr ($txt, $i, 1, ' UTF-8 '); if (Isset ($wordTree [$char])) { $result = $this->gethitlistbytree (mb_substr ($txt, $i, A, ' UTF-8 '), $wordTree); if ($result) { foreach ($result as $hit = $type) { $hitList [$hit] = $type; } } } } return $hitList; } /** * Whether to ban words * @param str $txt * @param arr $wordTree * @return Multitype:unknown */ Private Function Gethitlistbytree ($txt,& $wordTree) { $len = Mb_strlen ($txt, ' UTF-8 '); $point = & $wordTree; $hit = "; $hitList = Array (); for ($i =0; $i < $len; $i + +) { $char = Mb_substr ($txt, $i, 1, ' UTF-8 '); if (Isset ($point [$char])) { $hit. = $char; $point = & $point [$char]; if (Isset ($point [' type ']))//Match succeeded { $hitList [$hit] = $point [' type ']; } } Else { Break } } return $hitList; } /** * Get banned word trees in batches * @param int $page * @param int $size * @return Arr: */ Private Function Getwordtreebypage ($page =1, $size =1000) { $redis = Rds::factory (); $memKey = ' blackword_tree_ '. $page. ' _ '. $size; $wordTree = $redis->get ($memKey); if ($wordTree ===false) { $wordTree = Array (); $blackWord = new Model_blackword_blackword (); $start = ($page-1) * $size; $end = $start + $size; $para [' where '] = "Status=1 and id>". $start. " and id<= ". $end; $result = $blackWord->search ($para); if ($result) { foreach ($result as $value) { if ($value [' word ']) { $value [' word '] = preg_split ('/(? $point = & $wordTree; foreach ($value [' word '] as $char) { $point = & $point [$char]; } $point [' type '] = $value [' type ']; } } } $redis->setex ($memKey, $wordTree); } return $wordTree; } } Copy Code

Php

