基于Laravel 敏感词过滤

2019-09-12 必备技能阅读量

敏感词过滤,不能让用户作死带动平台作死.

PhraseFilter

<?php

namespace App\Support;

use App\Concerns\Singleton;

/**
 * 敏感词过滤
 * Class PhraseFilter
 * @package App\Support
 */
class PhraseFilter
{
    use Singleton;

    const MATCH_SIZE = 256;

    public $matched;

    protected $source_txt;

    /**
     * PhraseFilter constructor.
     * @param string|null $source_txt
     */
    public function __construct(string $source_txt = null)
    {
        if (null === $source_txt) {
            $source_txt = database_path() . '/app/filter_word.txt';
        }

        // PCRE的回溯限制
        ini_set('pcre.backtrack_limit', '23001337');
        // PCRE的递归限制
        ini_set('pcre.recursion_limit', '23001337');

        $this->source_txt = $source_txt;
    }

    /**
     * @param string $target
     * @return string
     */
    private function prepare(string $target)
    {
        $target  = strtolower($target);
        $pattern = "/([0-9a-zA-Z\x{4e00}-\x{9fa5}]+)/u";
        preg_match_all($pattern, $target, $matched);
        $result  = implode('', $matched[0]);

        return $result;
    }

    /**
     * 检查是否匹配
     * @param string $target
     * @param callable|null $callback
     * @return bool true:detected, false:not detected.
     */
    public function match(string $target, callable $callback = null)
    {
        static $patterns = null;
        if (null === $patterns) {
            $patterns = $this->getPatterns();
        }
        $this->matched = null;
        $target = $this->prepare($target);

        foreach ($patterns as $pattern) {
            if (preg_match($pattern, $target, $this->matched)) {
                if (strlen($this->matched[0]) > 0) {
                    if (!empty($callback)) {
                        $callback($this->matched);
                    }
                    return true;
                }
            }
        }

        return false;
    }

    /**
     * 敏感词匹配规则
     * @return array
     */
    private function getPatterns()
    {
        $src_txt = file_get_contents($this->source_txt);
        $src_txt = preg_replace('/(\r\n)|(\r\n)|(\r)|(\n)/', '|', $src_txt);

        $word_list  = explode('|', $src_txt);
        $result     = [];
        $word_count = count($word_list);

        for ($i = 0; $i < $word_count; $i += self::MATCH_SIZE) {
            $src = array_splice($word_list, 0, self::MATCH_SIZE);
            $src_count = count($src);
            if (empty($src[$src_count - 1])) {
                $src = array_slice($src, 0, $src_count - 1);
            }
            $result[] = '/' . implode('|', $src) . '/';
        }

        return $result;
    }
}

使用test_filter

/**
 * @param Request $request
 * @param PhraseFilter $filter
 * @return mixed
 * @Get("/test_filter")
 */
public function test_filter(Request $request, PhraseFilter $filter)
{
    $word = $request->get('word');

    if (true == $filter->match($word)) {
        return System::apiResponse(4003, '含有敏感词');
    }

    return System::apiResponse(200, 'Success');
}