Java TTL Cache

http://www.1point3acres.com/bbs/thread-116171-1-1.html
一直在和他讨论如何实现。。。题目是实现一个hashtable能够对每个entry设置ttl(time to live)，ttl一过就删除。一开始感觉有点像LRUCache，就往linkedlist上想了，但是纠结于什么时候check是否要删除entry，后来又想每个entry插入的时候开一个thread，然后ttl后这个thread把对应entry删除，问了有什么优缺点，race condition，如果系统thread没有准时删除怎么办？反正各种跪。

X.
https://gist.github.com/pcan/16faf4e59942678377e0

    private final DelayQueue<ExpiringKey> delayQueue = new DelayQueue<ExpiringKey>();

    public SelfExpiringHashMap(final long defaultMaxLifeTimeMillis) {

        internalMap = new ConcurrentHashMap<K, V>();

        expiringKeys = new WeakHashMap<K, ExpiringKey<K>>();

        this.maxLifeTimeMillis = defaultMaxLifeTimeMillis;

    }

    public V put(final K key, final V value, final long lifeTimeMillis) {

        cleanup();

        final ExpiringKey delayedKey = new ExpiringKey(key, lifeTimeMillis);

        expireKey(expiringKeys.put(key, delayedKey));

        delayQueue.offer(delayedKey);

        return internalMap.put(key, value);

    }

X.  Use ConcurrentHashMap and a thread
http://www.java2s.com/Code/Java/Collections-Data-Structure/ExpiringMap.htm

    private final ConcurrentHashMap<K, ExpiringObject> delegate;

    private final CopyOnWriteArrayList<ExpirationListener<V>> expirationListeners;

    private final Expirer expirer;

    private ExpiringMap(final ConcurrentHashMap<K, ExpiringObject> delegate, final CopyOnWriteArrayList<ExpirationListener<V>> expirationListeners, final int timeToLive, final int expirationInterval) {

        this.delegate = delegate;

        this.expirationListeners = expirationListeners;

        this.expirer = new Expirer();

        expirer.setTimeToLive(timeToLive);

        expirer.setExpirationInterval(expirationInterval);

    }

    public V put(final K key, final V value) {

        final ExpiringObject answer = delegate.put(key, new ExpiringObject(key, value, System.currentTimeMillis()));

        if (answer == null) {

            return null;

        }

        return answer.getValue();

    }

    public V get(final Object key) {

        final ExpiringObject object = delegate.get(key);

        if (object != null) {

            object.setLastAccessTime(System.currentTimeMillis());

            return object.getValue();

        }

        return null;

    }

        private void processExpires() {

            final long timeNow = System.currentTimeMillis();

            for (final ExpiringObject o : delegate.values()) {

                if (timeToLiveMillis <= 0) {

                    continue;

                }

                final long timeIdle = timeNow - o.getLastAccessTime();

                if (timeIdle >= timeToLiveMillis) {

                    delegate.remove(o.getKey());

                    for (final ExpirationListener<V> listener : expirationListeners) {

                        listener.expired(o.getValue());

                    }

                }

            }

        }

X. Improve outdated element when get or iterate element.
==>  we can also remove (one or all) outdated element when size > threshold.
https://git.eclipse.org/r/#/c/41522/1/org.eclipse.scout.commons/src/org/eclipse/scout/commons/TTLCache.java
http://nirlevy.blogspot.com/2007/11/java-ttl-cache-quick-and-dirty-way.html
===> not good implementation.
http://www.javaworld.com/article/2071825/build-ci-sdlc/the-timestamp-based-caching-framework--current-data-with-peak-performance.html

Cache<Key, Graph> graphs = CacheBuilder.newBuilder()
    .concurrencyLevel(4)
    .weakKeys()
    .maximumSize(10000)
    .expireAfterWrite(10, TimeUnit.MINUTES)
    .build(
        new CacheLoader<Key, Graph>() {
          public Graph load(Key key) throws AnyException {
            return createExpensiveGraph(key);
          }
        });