import java.io.FileInputStream;
import java.io.FileOutputStream;
import java.io.ObjectInputStream;
import java.io.ObjectOutputStream;
import java.io.Serializable;
import java.util.BitSet;
import java.util.concurrent.atomic.AtomicInteger;
 
public class BloomFileter implements Serializable {
 private static final long serialVersionUID = -5221305273707291280L;
 private final int[] seeds;
 private final int size;
 private final BitSet notebook;
 private final MisjudgmentRate rate;
 private final AtomicInteger useCount = new AtomicInteger(0);
 private final Double autoClearRate;
 
 /**
 * 默認(rèn)中等程序的誤判率：MisjudgmentRate.MIDDLE 以及不自動(dòng)清空數(shù)據(jù)（性能會(huì)有少許提升）
 * 
 * @param dataCount
 *      預(yù)期處理的數(shù)據(jù)規(guī)模，如預(yù)期用于處理1百萬數(shù)據(jù)的查重，這里則填寫1000000
 */
 public BloomFileter(int dataCount) {
 this(MisjudgmentRate.MIDDLE, dataCount, null);
 }
 
 /**
 * 
 * @param rate
 *      一個(gè)枚舉類型的誤判率
 * @param dataCount
 *      預(yù)期處理的數(shù)據(jù)規(guī)模，如預(yù)期用于處理1百萬數(shù)據(jù)的查重，這里則填寫1000000
 * @param autoClearRate
 *      自動(dòng)清空過濾器內(nèi)部信息的使用比率，傳null則表示不會(huì)自動(dòng)清理，
 *      當(dāng)過濾器使用率達(dá)到100%時(shí)，則無論傳入什么數(shù)據(jù)，都會(huì)認(rèn)為在數(shù)據(jù)已經(jīng)存在了
 *      當(dāng)希望過濾器使用率達(dá)到80%時(shí)自動(dòng)清空重新使用，則傳入0.8
 */
 public BloomFileter(MisjudgmentRate rate, int dataCount, Double autoClearRate) {
 long bitSize = rate.seeds.length * dataCount;
 if (bitSize < 0 || bitSize > Integer.MAX_VALUE) {
  throw new RuntimeException("位數(shù)太大溢出了，請降低誤判率或者降低數(shù)據(jù)大小");
 }
 this.rate = rate;
 seeds = rate.seeds;
 size = (int) bitSize;
 notebook = new BitSet(size);
 this.autoClearRate = autoClearRate;
 }
 
 public void add(String data) {
 checkNeedClear();
 
 for (int i = 0; i < seeds.length; i++) {
  int index = hash(data, seeds[i]);
  setTrue(index);
 }
 }
 
 public boolean check(String data) {
 for (int i = 0; i < seeds.length; i++) {
  int index = hash(data, seeds[i]);
  if (!notebook.get(index)) {
  return false;
  }
 }
 return true;
 }
 
 /**
 * 如果不存在就進(jìn)行記錄并返回false，如果存在了就返回true
 * 
 * @param data
 * @return
 */
 public boolean addIfNotExist(String data) {
 checkNeedClear();
 
 int[] indexs = new int[seeds.length];
 // 先假定存在
 boolean exist = true;
 int index;
 
 for (int i = 0; i < seeds.length; i++) {
  indexs[i] = index = hash(data, seeds[i]);
 
  if (exist) {
  if (!notebook.get(index)) {
   // 只要有一個(gè)不存在，就可以認(rèn)為整個(gè)字符串都是第一次出現(xiàn)的
   exist = false;
   // 補(bǔ)充之前的信息
   for (int j = 0; j <= i; j++) {
   setTrue(indexs[j]);
   }
  }
  } else {
  setTrue(index);
  }
 }
 
 return exist;
 
 }
 
 private void checkNeedClear() {
 if (autoClearRate != null) {
  if (getUseRate() >= autoClearRate) {
  synchronized (this) {
   if (getUseRate() >= autoClearRate) {
   notebook.clear();
   useCount.set(0);
   }
  }
  }
 }
 }
 
 public void setTrue(int index) {
 useCount.incrementAndGet();
 notebook.set(index, true);
 }
 
 private int hash(String data, int seeds) {
 char[] value = data.toCharArray();
 int hash = 0;
 if (value.length > 0) {
 
  for (int i = 0; i < value.length; i++) {
  hash = i * hash + value[i];
  }
 }
 
 hash = hash * seeds % size;
 // 防止溢出變成負(fù)數(shù)
 return Math.abs(hash);
 }
 
 public double getUseRate() {
 return (double) useCount.intValue() / (double) size;
 }
 
 public void saveFilterToFile(String path) {
 try (ObjectOutputStream oos = new ObjectOutputStream(new FileOutputStream(path))) {
  oos.writeObject(this);
 } catch (Exception e) {
  throw new RuntimeException(e);
 }
 
 }
 
 public static BloomFileter readFilterFromFile(String path) {
 try (ObjectInputStream ois = new ObjectInputStream(new FileInputStream(path))) {
  return (BloomFileter) ois.readObject();
 } catch (Exception e) {
  throw new RuntimeException(e);
 }
 }
 
 /**
 * 清空過濾器中的記錄信息
 */
 public void clear() {
 useCount.set(0);
 notebook.clear();
 }
 
 public MisjudgmentRate getRate() {
 return rate;
 }
 
 /**
 * 分配的位數(shù)越多，誤判率越低但是越占內(nèi)存
 * 
 * 4個(gè)位誤判率大概是0.14689159766308
 * 
 * 8個(gè)位誤判率大概是0.02157714146322
 * 
 * 16個(gè)位誤判率大概是0.00046557303372
 * 
 * 32個(gè)位誤判率大概是0.00000021167340
 * 
 * @author lianghaohui
 *
 */
 public enum MisjudgmentRate {
 // 這里要選取質(zhì)數(shù)，能很好的降低錯(cuò)誤率
 /**
  * 每個(gè)字符串分配4個(gè)位
  */
 VERY_SMALL(new int[] { 2, 3, 5, 7 }),
 /**
  * 每個(gè)字符串分配8個(gè)位
  */
 SMALL(new int[] { 2, 3, 5, 7, 11, 13, 17, 19 }), //
 /**
  * 每個(gè)字符串分配16個(gè)位
  */
 MIDDLE(new int[] { 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53 }), //
 /**
  * 每個(gè)字符串分配32個(gè)位
  */
 HIGH(new int[] { 2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97,
  101, 103, 107, 109, 113, 127, 131 });
 
 private int[] seeds;
 
 private MisjudgmentRate(int[] seeds) {
  this.seeds = seeds;
 }
 
 public int[] getSeeds() {
  return seeds;
 }
 
 public void setSeeds(int[] seeds) {
  this.seeds = seeds;
 }
 
 }
 
 public static void main(String[] args) {
 BloomFileter fileter = new BloomFileter(7);
 System.out.println(fileter.addIfNotExist("1111111111111"));
 System.out.println(fileter.addIfNotExist("2222222222222222"));
 System.out.println(fileter.addIfNotExist("3333333333333333"));
 System.out.println(fileter.addIfNotExist("444444444444444"));
 System.out.println(fileter.addIfNotExist("5555555555555"));
 System.out.println(fileter.addIfNotExist("6666666666666"));
 System.out.println(fileter.addIfNotExist("1111111111111"));
 fileter.saveFilterToFile("C:\\Users\\john\\Desktop\\1111\\11.obj");
 fileter = readFilterFromFile("C:\\Users\\john\\Desktop\\111\\11.obj");
 System.out.println(fileter.getUseRate());
 System.out.println(fileter.addIfNotExist("1111111111111"));
 }
}