Package org.apache.uima.internal.util

Class WeakIdentityMap<K,V>

java.lang.Object
org.apache.uima.internal.util.WeakIdentityMap<K,V>
Type Parameters:
K - the key type.
V - the value type.
public final class WeakIdentityMap<K,V> extends Object
Implements a combination of WeakHashMap and IdentityHashMap. Useful for caches that need to key off of a == comparison instead of a .equals.

This class is not a general-purpose Map implementation! It intentionally violates Map's general contract, which mandates the use of the equals method when comparing objects. This class is designed for use only in the rare cases wherein reference-equality semantics are required.

This implementation was derived from Apache Lucene Lucene in turn forked this class from from Apache CXF but modified to not implement the Map interface and without any set views on it, as those are error-prone and inefficient, if not implemented carefully. The map only contains Iterator implementations on the values and not-GCed keys. Lucene's implementation also supports null keys, but those are never weak!

The map supports two modes of operation:

  • reapOnRead = true: This behaves identical to a WeakHashMap where it also cleans up the reference queue on every read operation (get(Object), containsKey(Object), size(), valueIterator()), freeing map entries of already GCed keys.
  • reapOnRead = false: This mode does not call reap() on every read operation. In this case, the reference queue is only cleaned up on write operations (like put(Object, Object)). This is ideal for maps with few entries where the keys are unlikely be garbage collected, but there are lots of get(Object) operations. The code can still call reap() to manually clean up the queue without doing a write operation.

  • Method Details

    • newHashMap

      public static <K, V> WeakIdentityMap<K,V> newHashMap()
      Creates a new WeakIdentityMap based on a non-synchronized HashMap. The map cleans up the reference queue on every read operation.
      Returns:
      the new map.

    • newHashMap

      public static <K, V> WeakIdentityMap<K,V> newHashMap(boolean reapOnRead)
      Creates a new WeakIdentityMap based on a non-synchronized HashMap.
      Parameters:
      reapOnRead - controls if the map cleans up the reference queue on every read operation.
      Returns:
      the new map.

    • newConcurrentHashMap

      public static <K, V> WeakIdentityMap<K,V> newConcurrentHashMap()
      Creates a new WeakIdentityMap based on a ConcurrentHashMap. The map cleans up the reference queue on every read operation.
      Returns:
      the new map.

    • newConcurrentHashMap

      public static <K, V> WeakIdentityMap<K,V> newConcurrentHashMap(boolean reapOnRead)
      Creates a new WeakIdentityMap based on a ConcurrentHashMap.
      Parameters:
      reapOnRead - controls if the map cleans up the reference queue on every read operation.
      Returns:
      the new map.

    • clear

      public void clear()
      Removes all of the mappings from this map.

    • containsKey

      public boolean containsKey(Object key)
      Parameters:
      key - the key to check for a mapping.
      Returns:
      true if this map contains a mapping for the specified key.

    • get

      public V get(Object key)
      Parameters:
      key - the key for which to return the associated value.
      Returns:
      the value to which the specified key is mapped.

    • put

      public V put(K key, V value)
      Associates the specified value with the specified key in this map. If the map previously contained a mapping for this key, the old value is replaced.
      Parameters:
      key - the associate the value with.
      value - the value to associate with the key.
      Returns:
      the previous associated value.

    • isEmpty

      public boolean isEmpty()
      Returns:
      true if this map contains no key-value mappings.

    • remove

      public V remove(Object key)
      Removes the mapping for a key from this weak hash map if it is present. Returns the value to which this map previously associated the key, or null if the map contained no mapping for the key. A return value of null does not necessarily indicate that the map contained.
      Parameters:
      key - the key to remove.
      Returns:
      the previous mapping.

    • size

      public int size()
      Returns:
      the number of key-value mappings in this map. This result is a snapshot, and may not reflect unprocessed entries that will be removed before next attempted access because they are no longer referenced.

    • keyIterator

      public Iterator<K> keyIterator()
      Returns:
      an iterator over all weak keys of this map. Keys already garbage collected will not be returned. This Iterator does not support removals.

    • valueIterator

      public Iterator<V> valueIterator()
      Returns:
      an iterator over all values of this map. This iterator may return values whose key is already garbage collected while iterator is consumed, especially if reapOnRead is false.

    • reap

      public void reap()
      This method manually cleans up the reference queue to remove all garbage collected key/value pairs from the map. Calling this method is not needed if reapOnRead = true. Otherwise it might be a good idea to call this method when there is spare time (e.g. from a background thread).
      See Also: