java 使用ConcurrentHashMap和计数器实现锁
在某些场景下,我们想让线程根据某些业务数据进行排队,简单代码如下：
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.atomic.AtomicInteger;
public class TestServiceImpl {
private static ConcurrentHashMap<Long, LockObj> lockMap = new ConcurrentHashMap<Long, LockObj>(40);
public void test(Long userId){
LockObj lock = tryLock(userId);
synchronized (lock) {
try{
//处理业务
}
finally{
unLock(lock);
}
}
}
private LockObj tryLock(Long key) {
LockObj curVal = new LockObj(key);
LockObj preVal = lockMap.putIfAbsent(key, curVal);
if (null == preVal) {
curVal.inc();
return curVal;
}
else{
preVal.inc();
}
return preVal;
}
private void unLock(LockObj lock){
if (lock.dec() <= 0){
lockMap.remove(lock.getKey());
}
}
public class LockObj {
private long key = 0;
private AtomicInteger count = new AtomicInteger(0);
public LockObj(long key){
this.key = key;
}
public int inc(){
return count.incrementAndGet();
}
public int dec(){
return count.decrementAndGet();
}
public long getKey(){
return key;
}
@Override
public String toString() {
return "LockObj [key=" + key + ", count=" + count + "]";
}
}
}
按照userId来排队,如果每个线程处理数据后不释放锁的话，那么可以不利用计数器。但是加了释放锁的操作，则必须加上计算器。因为当线程把锁释放掉后，还没来得及退出synchronized 代码块时，另外一个线程调用了tryLock方法，那该线程将拿到另外一个对象的锁，导致利用synchronized 关键字进行userId排队失败。
也可以利用guava的API来实现。
import com.google.common.collect.Interner;
import com.google.common.collect.Interners;
public class TestServiceImpl {
Interner<String> pool = Interners.newWeakInterner();
public void test(Long userId) throws OspException {
synchronized ( pool.intern(String.valueOf(userId))){
//处理业务操作
}
}
}
感谢阅读，希望能帮助到大家，谢谢大家对本站的支持！
