小弟我今天在群里看到有个大牛手写两个一个仿照redis思想的hashMap，自己想着，先记下来，以后慢慢研究，以下是代码。
/**
* 仿redis思想，按照redis dict 实现自己的hashmap
* 去除了原生的Java hashmap一些接口封装，只为实现功能
* 目前实现了自动扩容和渐进式rehash
* 自动收缩待实现
* @author zzc
* @date 2017/11/9
*/
public class MyHashMap<K,V> {
private static int INIT_SIZE = 2;
private static int THREAD_HOLD = 5;
MyHashMap(){
this.entries = new Entry[INIT_SIZE];
this.size = INIT_SIZE;
this.used = 0;
this.mask = size-1;
this.rehashidx = -1;
}
MyHashMap(int size){
this.entries = new Entry[size];
this.size = size;
this.used = 0;
this.mask = size-1;
this.rehashidx = -1;
}
private Entry<K,V>[] entries;
private int size;
private int mask;
private int used;
private int rehashidx;
private MyHashMap<K,V> back;
private int hash(Object key) {
return key.hashCode();
}
private int index(int hash){
return hash&mask;
}
public void put(K k,V v){
//扩展&收缩处理
if((used/size)>=THREAD_HOLD){
//此时需要扩展
//计算新表大小
if (back == null){
int i = 0;
while (used>>(++i)>0){
}
int newSize =1<<i;
back = new MyHashMap<>(newSize);
//标志开始rehash
this.rehashidx = 0;
}
transToNewTable();
putToEntries(back,k,v);
return;
}
putToEntries(this,k,v);
}
private void transToNewTable() {
if (this.rehashidx<this.used){
for (int i = 0 ;i<this.size;i++){
Entry<K, V> tmp = this.entries[i];
if (tmp != null){
this.entries[i] = tmp.next;
int index = back.index(tmp.k.hashCode());
Entry<K, V> btmp = back.entries[index];
tmp.next = btmp;
back.entries[index] = tmp;
back.used++;
this.rehashidx++;
break;
}
}
return;
}
copyEntries();
this.rehashidx = -1;
}
private void copyEntries() {
this.entries = back.entries;
this.size = back.size;
this.used = back.used;
this.mask = back.mask;
back.entries = null;
back = null;
}
private void putToEntries(MyHashMap map, K k, V v) {
//计算索引
int index = map.index(hash(k));
Entry curr = map.entries[index];
if (curr == null){
map.entries[index] = new Entry();
map.entries[index].v=v;
map.entries[index].k = k;
map.used++;
return;
}
if (curr.k.equals(k)){
curr.v = v;
return;
}
while (curr.hasNext()){
curr = curr.next;
if (curr.k.equals(k)) {
curr.v = v;
return;
}
}
Entry<K, V> tmp = new Entry<>();
tmp.k = k;
tmp.v = v;
curr.next = tmp;
map.used++;
}
public V get(K k){
//扩展&收缩处理
if((used/size)>=THREAD_HOLD&&back!=null){
transToNewTable();
V ret = getEntries(this, k);
if (ret!= null){
//rehash移动原来表kv到新表
return ret;
}
return getEntries(back,k);
}
return getEntries(this,k);
}
private V getEntries(MyHashMap map, K k) {
int index = map.index(hash(k));
Entry<K,V> curr = map.entries[index];
if (curr == null){
return null;
}
if (k.equals(curr.k)){
return curr.v;
}
while (curr.hasNext()){
curr = curr.next;
if (curr.k.equals(k)){
return curr.v;
}
}
return null;
}
class Entry<K,V>{
K k;
V v;
Entry<K,V> next;
boolean hasNext(){
return next != null;
}
}
public static void main(String[] args) {
MyHashMap<String, String> myHashMap = new MyHashMap<>(2);
myHashMap.put("name","张三");
myHashMap.put("name1","李四");
myHashMap.put("name2","李四");
myHashMap.put("name3","李四");
myHashMap.put("name4","李四");
myHashMap.put("name5","李四");
myHashMap.put("name6","李四");
myHashMap.put("name7","李四");
myHashMap.put("name8","李四");
myHashMap.put("name9","李四");
myHashMap.put("name10","李四");
myHashMap.put("name11","李四");
myHashMap.put("name12","李四");
myHashMap.put("name13","李四");
myHashMap.put("name14","李四");
myHashMap.put("name15","李四");
myHashMap.put("name16","李四");
myHashMap.put("name17","李四");
myHashMap.put("age","18");
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name1"));
System.out.println(myHashMap.get("name2"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name2"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name"));
System.out.println(myHashMap.get("name1"));
System.out.println(myHashMap.get("name"));
}
}