Java使用DFA算法实现过滤多家公司自定义敏感字功能详解

本文实例讲述了java使用dfa算法实现过滤多家公司自定义敏感字功能。分享给大家供大家参考，具体如下：

背景

因为最近有通讯有个需求，说需要让多家客户公司可以自定义敏感词过滤掉他们自定义的规则，选择了dfa算法来做，不过和以前传统了dfa写法不太一样了

模式图

直接上代码

public class keywordfilter {
//  private static reentrantreadwritelock lock = new reentrantreadwritelock();
  public static map<string, hashmap> currentmap = new concurrenthashmap<string, hashmap>();
  public static map nowhash = null;
  public static object wordmap;// map子节点
  // 不建立对象
  private keywordfilter() {
  }
  private static string getkey(int companyid) {
    return "companyid" + companyid;
  }
  /*
   * <p>说明:清扫内容</p>
   *
   * @author:姚旭民
   *
   * @data:2017-8-22 上午10:13:11
   */
  public static void clear() {
    try {
      currentmap.clear();
    } catch (exception e) {
      e.printstacktrace();
    } finally {
    }
  }
  /*
   * <p>说明:各个渠道的过滤字符</p>
   *
   * @author:姚旭民
   *
   * @data:2017-8-20 下午2:55:06
   */
  public static void savekeywords(int companyid, list<string> keywords) {
    try {
      map tempallmap = currentmap;
      string key = getkey(companyid);
      int l = keywords.size();
      int il;
      map tempmap;
      for (int i = 0; i < l; i++) {
        string key2 = keywords.get(i).trim();// 去掉空白
        nowhash = currentmap;
        il = key2.length();
        for (int j = 0; j < il; j++) {
          char word = key2.charat(j);
          tempmap = (map) nowhash.get(word);
          wordmap = nowhash.get(word);
          if (wordmap != null) {// 检查数据
            if (!tempmap.containskey(key)) {
              nowhash.put(key, 0);
            }
            nowhash = (hashmap) wordmap;
          } else {
            hashmap<string, string> newwordhash = new hashmap<string, string>();
            newwordhash.put(key, "0");
            nowhash.put(word, newwordhash);
            nowhash = newwordhash;
          }
          if (j == il - 1) {
            nowhash.put(key, "1");
          }
        }
      }
    } catch (exception e) {
      e.printstacktrace();
    } finally {
      nowhash = null;
      wordmap = null;
    }
  }
  /*
   * <p>说明:替换掉对应的渠道规定掉敏感字</p>
   *
   * @author:姚旭民
   *
   * @data:2017-8-20 上午11:41:47
   */
  public static list<string> repword(int companyid, string txt) {
    map tempmap = currentmap;
    list<string> result = new arraylist<string>();
    string key = getkey(companyid);
    nowhash = currentmap;
    int l = txt.length();
    char word;
    string keywordstr = "";
    string keystatu;
    stringbuilder keyword = new stringbuilder();// 敏感字
    for (int i = 0; i < l; i++) {
      word = txt.charat(i);
      wordmap = nowhash.get(word);
      if (wordmap != null) {// 找到类似敏感字的字体，开始查询
        keyword.append(word);
        object te = nowhash = (hashmap) wordmap;
        // 遍历到这一步，就符合完整的关键字模板
        if (nowhash.get(key) != null
            && nowhash.get(key).tostring().equals("1")) {// 确定是敏感字，开始替换
          if (i < l - 1 && nowhash.get(txt.charat(i + 1)) != null) {// 优先过滤长敏感词，去掉就槟城了优先过滤段敏感词
            continue;
          }
          txt = txt.replaceall(keyword.tostring(), "*");
          nowhash = currentmap;
          keywordstr += keyword.tostring() + ",";
          i = i - keyword.length() + 1;
          l = txt.length();// 重新获取字符长度
          keyword.delete(0, keyword.length());// 清空数据
        }
      } else {// 这个字不是敏感字，直接排除
        nowhash = currentmap;
        keyword.delete(0, keyword.length());// 清空数据
        continue;
      }
    }
    // 清除内存指向
    nowhash = null;
    wordmap = null;
    result.add(txt);
    result.add(keywordstr.length() - 1 > 0 ? keywordstr.substring(0,
        keywordstr.length() - 1) : keywordstr);
    return result;
  }
  /*
   * <p>说明:检查是否存在敏感字</p>
   *
   * @author:姚旭民
   *
   * @data:2017-8-20 下午3:00:06 专门设计成私有的，如果没有理由，别改动他
   */
  private static int checkkeywords(string txt, int companyid, int begin) {
    int result = 0;
    string key = getkey(companyid);
    try {
      nowhash = currentmap;
      int l = txt.length();
      char word = 0;
      for (int i = begin; i < l; i++) {
        word = txt.charat(i);
        wordmap = nowhash.get(word);
        if (wordmap != null) {
          result++;
          nowhash = (hashmap) wordmap;
          if (((string) nowhash.get(key)).equals("1")) {
            nowhash = null;
            wordmap = null;
            return result;
          }
        } else {
          result = 0;
          break;
        }
      }
    } catch (exception e) {
      e.printstacktrace();
    } finally {
      nowhash = null;
      wordmap = null;
      return result;
    }
  }
  /*
   * <p>说明:返回检查的文本中包含的敏感字</p>
   *
   * @author:姚旭民
   *
   * @data:2017-8-20 下午3:32:53
   */
  public static string gettxtkeywords(string txt, int companyid) {
    string result = null;
    stringbuilder temp = new stringbuilder();
    string key;
    int l = txt.length();
    for (int i = 0; i < l;) {
      int len = checkkeywords(txt, companyid, i);
      if (len > 0) {
        key = (txt.substring(i, i + len));// 挑选出来的关键字
        temp.append(key + ",");
        txt = txt.replaceall(key, "");// 挑选出来的关键字替换成空白，加快挑选速度
        l = txt.length();
      } else {
        i++;
      }
    }
    if (temp.length() > 0) {
      result = temp.substring(0, temp.length() - 1);
    }
    return result;
  }
  /*
   * <p>说明:判断文中是否包含渠道规定的敏感字</p>
   *
   * @author:姚旭民
   *
   * @data:2017-8-20 下午3:33:19
   */
  public boolean iskeywords(string txt, int companyid) {
    for (int i = 0; i < txt.length(); i++) {
      int len = checkkeywords(txt, companyid, i);
      if (len > 0) {
        return true;
      }
    }
    return false;
  }
  public static void main(string[] arg) {
    list<string> keywords = new arraylist<string>();
    keywords.add("傻×");
    keywords.add("汉奸");
    keywords.add("草");
    keywords.add("*");
    keywordfilter.savekeywords(1, keywords);
    string txt = "是傻×汉奸傻a傻b傻c傻d汉奸傻×*";
    list<string> list = repword(1, txt);
    system.out.println("文中包含的敏感字为:" + list.get(1));
    system.out.println("原文:" + txt);
    system.out.println("敏感字过滤后:" + list.get(0));
  }
}

更多关于java算法相关内容感兴趣的读者可查看本站专题：《java数据结构与算法教程》、《java字符与字符串操作技巧总结》、《java操作dom节点技巧总结》、《java文件与目录操作技巧汇总》和《java缓存操作技巧汇总》

希望本文所述对大家java程序设计有所帮助。