漏桶算法

实现思路：

1. 定义一个数据结构实现漏桶功能，当有外部流量涌入时把流量放入桶中
2. 实现一个请求处理器，以恒定的速率获取桶中流量进行处理

public class LeakyBucketRateLimit implements RateLimit,Runnable{

    //出口限制qps
    private Integer limitSecond;

    //漏桶队列
    private BlockingDeque<Thread> leakyBucket;
    private ScheduledExecutorService scheduledExecutorService;

    public LeakyBucketRateLimit(Integer bucketSize, Integer limitSecond) {
        this.limitSecond = limitSecond;
        this.leakyBucket = new LinkedBlockingDeque<>(bucketSize);
        scheduledExecutorService = Executors.newSingleThreadScheduledExecutor();
        //周期秒数，TimeUnit.NANOSECONDS(毫秒数，1秒=1000*1000*1000)毫微妙
        long interval = (1000 * 1000 * 1000) / limitSecond;
        scheduledExecutorService.scheduleAtFixedRate(this, 0, interval, TimeUnit.NANOSECONDS);
    }

    @Override
    public boolean canPass() throws RuntimeException {
        if(leakyBucket.remainingCapacity() == 0){
            throw new RuntimeException();
        }
        leakyBucket.offer(Thread.currentThread());
        LockSupport.park();
        return true;
    }

    @Override
    public void run() {
        Thread thread = leakyBucket.poll();
        if(Objects.nonNull(thread)){
            LockSupport.unpark(thread);
        }
    }

    public static void main(String[] args) throws RuntimeException {
        LeakyBucketRateLimit leakyBucketRateLimit = new LeakyBucketRateLimit(100, 1);
        for (int i = 0; i < 10000; i++){
            if(leakyBucketRateLimit.canPass()){
                System.out.println("通过第【" + i + "】个请求");
            }
        }
    }
}