README

php实现基于确定有穷自动机算法的铭感词过滤

安装扩展

composer require phpch/dfa-sensitive-word

如果你需要手动引入

require './vendor/autoload.php';

use Phpch\SensitiveWord\SensitiveWordFilter;

构建敏感词库树

场景一: 可以拿到不同（用户）词库数组

 $wordData = array(
        '察象蚂',
        '拆迁灭',
        '车牌隐',
        '成人电',
        '成人卡通',
        ......
    );
    
    $handle = SensitiveWordFilter::init()->setArray($wordData)->buildTree();

场景二: 全站使用一套敏感词库

    // 获取感词库文件路径
    $wordFilePath = 'tests/words/key2.txt';
    
    $handle = SensitiveWordFilter::init()->setFilePath($wordFilePath)->buildTree();
    
     // 设置缓存后可以避免重复构建铭感词树
      $handle = SensitiveWordFilter::init()->setFilePath($wordFilePath)->buildTree(true);
      
      // 删除缓存
    $handle->clearCache();

设置干扰因子

$wordFilePath = 'tests/words/key2.txt';
$handle = SensitiveWordFilter::init()->setFilePath($wordFilePath)->setDisturbList(['-'])->buildTree();
$filterContent = $handle->replace('我是中--国人'); // 我是*****

设置匹配规则

默认是最大匹配

// 最小匹配规则
$wordData = ["中国","中国人"];
$handle = SensitiveWordFilter::init()->setArray($wordData)->setMatchTypeMin()->buildTree();
$filterContent = $handle->replace('我是中国人'); // 我是**人

// 最大匹配规则
$wordData = ["中国","中国人"];
$handle = SensitiveWordFilter::init()->setArray($wordData)->setMatchTypeMax()->buildTree();
$filterContent = $handle->replace('我是中国人'); // 我是***

检测是否含有敏感词

$islegal = $handle->contains($content);

敏感词过滤

// 敏感词替换为*为例（会替换为相同字符长度的*）
$filterContent = $handle->replace($content, '*');

 // 或敏感词替换为***为例
 $filterContent = $handle->replace($content, '***', false);

标记敏感词

 $markedContent =  $handle->mark($content, '<mark>', '</mark>');

获取文字中的敏感词

// 获取内容中所有的敏感词
$sensitiveWordGroup = $handle->getBadWord($content);
// 仅且获取一个敏感词
$sensitiveWordGroup =  $handle->getBadWord($content, 1);