package org.andy.sensitivewdfilter; 
 
import java.io.BufferedReader; 
import java.io.IOException; 
import java.io.InputStreamReader; 
import java.util.ArrayList; 
import java.util.HashMap; 
import java.util.HashSet; 
import java.util.List; 
import java.util.Map; 
import java.util.Set; 
 
import org.andy.sensitivewdfilter.util.BCConvert; 
 
/** 
 * 創(chuàng)建時間：2016年8月30日 下午3:01:12 
 * 
 * 思路： 創(chuàng)建一個FilterSet，枚舉了0~65535的所有char是否是某個敏感詞開頭的狀態(tài) 
 * 
 * 判斷是否是 敏感詞開頭 | | 是 不是 獲取頭節(jié)點(diǎn) OK--下一個字 然后逐級遍歷，DFA算法 
 * 
 * @author andy 
 * @version 2.2 
 */ 
public class WordFilter { 
 
  private static final FilterSet set = new FilterSet(); // 存儲首字 
  private static final Map<Integer, WordNode> nodes = new HashMap<Integer, WordNode>(1024, 1); // 存儲節(jié)點(diǎn) 
  private static final Set<Integer> stopwdSet = new HashSet<>(); // 停頓詞 
  private static final char SIGN = '*'; // 敏感詞過濾替換 
 
  static { 
    try { 
      long a = System.nanoTime(); 
      init(); 
      a = System.nanoTime() - a; 
      System.out.println("加載時間 : " + a + "ns"); 
      System.out.println("加載時間 : " + a / 1000000 + "ms"); 
    } catch (Exception e) { 
      throw new RuntimeException("初始化過濾器失敗"); 
    } 
  } 
 
  private static void init() { 
    // 獲取敏感詞 
    addSensitiveWord(readWordFromFile("wd.txt")); 
    addStopWord(readWordFromFile("stopwd.txt")); 
  } 
 
  /** 
   * 增加敏感詞 
   * @param path 
   * @return 
   */ 
  private static List<String> readWordFromFile(String path) { 
    List<String> words; 
    BufferedReader br = null; 
    try { 
      br = new BufferedReader(new InputStreamReader(WordFilter.class.getClassLoader().getResourceAsStream(path))); 
      words = new ArrayList<String>(1200); 
      for (String buf = ""; (buf = br.readLine()) != null;) { 
        if (buf == null || buf.trim().equals("")) 
          continue; 
        words.add(buf); 
      } 
    } catch (Exception e) { 
      throw new RuntimeException(e); 
    } finally { 
      try { 
        if (br != null) 
          br.close(); 
      } catch (IOException e) { 
      } 
    } 
    return words; 
  } 
 
  /** 
   * 增加停頓詞 
   * 
   * @param words 
   */ 
  private static void addStopWord(final List<String> words) { 
    if (words != null && words.size() > 0) { 
      char[] chs; 
      for (String curr : words) { 
        chs = curr.toCharArray(); 
        for (char c : chs) { 
          stopwdSet.add(charConvert(c)); 
        } 
      } 
    } 
  } 
 
  /** 
   * 添加DFA節(jié)點(diǎn) 
   * @param words 
   */ 
  private static void addSensitiveWord(final List<String> words) { 
    if (words != null && words.size() > 0) { 
      char[] chs; 
      int fchar; 
      int lastIndex; 
      WordNode fnode; // 首字母節(jié)點(diǎn) 
      for (String curr : words) { 
        chs = curr.toCharArray(); 
        fchar = charConvert(chs[0]); 
        if (!set.contains(fchar)) {// 沒有首字定義 
          set.add(fchar);// 首字標(biāo)志位 可重復(fù)add,反正判斷了，不重復(fù)了 
          fnode = new WordNode(fchar, chs.length == 1); 
          nodes.put(fchar, fnode); 
        } else { 
          fnode = nodes.get(fchar); 
          if (!fnode.isLast() && chs.length == 1) 
            fnode.setLast(true); 
        } 
        lastIndex = chs.length - 1; 
        for (int i = 1; i < chs.length; i++) { 
          fnode = fnode.addIfNoExist(charConvert(chs[i]), i == lastIndex); 
        } 
      } 
    } 
  } 
 
  /** 
   * 過濾判斷 將敏感詞轉(zhuǎn)化為成屏蔽詞 
   * @param src 
   * @return 
   */ 
  public static final String doFilter(final String src) { 
    char[] chs = src.toCharArray(); 
    int length = chs.length; 
    int currc; 
    int k; 
    WordNode node; 
    for (int i = 0; i < length; i++) { 
      currc = charConvert(chs[i]); 
      if (!set.contains(currc)) { 
        continue; 
      } 
      node = nodes.get(currc);// 日 2 
      if (node == null)// 其實(shí)不會發(fā)生，習(xí)慣性寫上了 
        continue; 
      boolean couldMark = false; 
      int markNum = -1; 
      if (node.isLast()) {// 單字匹配（日） 
        couldMark = true; 
        markNum = 0; 
      } 
      // 繼續(xù)匹配（日你/日你妹），以長的優(yōu)先 
      // 你-3 妹-4 夫-5 
      k = i; 
      for (; ++k < length;) { 
        int temp = charConvert(chs[k]); 
        if (stopwdSet.contains(temp)) 
          continue; 
        node = node.querySub(temp); 
        if (node == null)// 沒有了 
          break; 
        if (node.isLast()) { 
          couldMark = true; 
          markNum = k - i;// 3-2 
        } 
      } 
      if (couldMark) { 
        for (k = 0; k <= markNum; k++) { 
          chs[k + i] = SIGN; 
        } 
        i = i + markNum; 
      } 
    } 
 
    return new String(chs); 
  } 
   
  /** 
   * 是否包含敏感詞 
   * @param src 
   * @return 
   */ 
  public static final boolean isContains(final String src) { 
    char[] chs = src.toCharArray(); 
    int length = chs.length; 
    int currc; 
    int k; 
    WordNode node; 
    for (int i = 0; i < length; i++) { 
      currc = charConvert(chs[i]); 
      if (!set.contains(currc)) { 
        continue; 
      } 
      node = nodes.get(currc);// 日 2 
      if (node == null)// 其實(shí)不會發(fā)生，習(xí)慣性寫上了 
        continue; 
      boolean couldMark = false; 
      if (node.isLast()) {// 單字匹配（日） 
        couldMark = true; 
      } 
      // 繼續(xù)匹配（日你/日你妹），以長的優(yōu)先 
      // 你-3 妹-4 夫-5 
      k = i; 
      for (; ++k < length;) { 
        int temp = charConvert(chs[k]); 
        if (stopwdSet.contains(temp)) 
          continue; 
        node = node.querySub(temp); 
        if (node == null)// 沒有了 
          break; 
        if (node.isLast()) { 
          couldMark = true; 
        } 
      } 
      if (couldMark) { 
        return true; 
      } 
    } 
 
    return false; 
  } 
 
  /** 
   * 大寫轉(zhuǎn)化為小寫 全角轉(zhuǎn)化為半角 
   * 
   * @param src 
   * @return 
   */ 
  private static int charConvert(char src) { 
    int r = BCConvert.qj2bj(src); 
    return (r >= 'A' && r <= 'Z') ? r + 32 : r; 
  } 
 
} 

package org.andy.sensitivewdfilter; 
 
import java.util.LinkedList; 
import java.util.List; 
 
/** 
 * 創(chuàng)建時間：2016年8月30日 下午3:07:45 
 * 
 * @author andy 
 * @version 2.2 
 */ 
public class WordNode { 
 
  private int value; // 節(jié)點(diǎn)名稱 
 
  private List<WordNode> subNodes; // 子節(jié)點(diǎn) 
 
  private boolean isLast;// 默認(rèn)false 
 
  public WordNode(int value) { 
    this.value = value; 
  } 
 
  public WordNode(int value, boolean isLast) { 
    this.value = value; 
    this.isLast = isLast; 
  } 
 
  /** 
   * 
   * @param subNode 
   * @return 就是傳入的subNode 
   */ 
  private WordNode addSubNode(final WordNode subNode) { 
    if (subNodes == null) 
      subNodes = new LinkedList<WordNode>(); 
    subNodes.add(subNode); 
    return subNode; 
  } 
 
  /** 
   * 有就直接返回該子節(jié)點(diǎn)， 沒有就創(chuàng)建添加并返回該子節(jié)點(diǎn) 
   * 
   * @param value 
   * @return 
   */ 
  public WordNode addIfNoExist(final int value, final boolean isLast) { 
    if (subNodes == null) { 
      return addSubNode(new WordNode(value, isLast)); 
    } 
    for (WordNode subNode : subNodes) { 
      if (subNode.value == value) { 
        if (!subNode.isLast && isLast) 
          subNode.isLast = true; 
        return subNode; 
      } 
    } 
    return addSubNode(new WordNode(value, isLast)); 
  } 
 
  public WordNode querySub(final int value) { 
    if (subNodes == null) { 
      return null; 
    } 
    for (WordNode subNode : subNodes) { 
      if (subNode.value == value) 
        return subNode; 
    } 
    return null; 
  } 
 
  public boolean isLast() { 
    return isLast; 
  } 
 
  public void setLast(boolean isLast) { 
    this.isLast = isLast; 
  } 
 
  @Override 
  public int hashCode() { 
    return value; 
  } 
 
}