All Downloads are FREE. Search and download functionalities are using the official Maven repository.

com.google.common.collect.CompactHashMap Maven / Gradle / Ivy

Go to download

Guava is a suite of core and expanded libraries that include utility classes, google's collections, io classes, and much much more.

There is a newer version: 62
Show newest version
/*
 * Copyright (C) 2012 The Guava Authors
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 * http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.google.common.collect;

import static com.google.common.base.Preconditions.checkNotNull;
import static com.google.common.collect.CollectPreconditions.checkRemove;
import static com.google.common.collect.CompactHashing.UNSET;
import static com.google.common.collect.Hashing.smearedHash;

import com.google.common.annotations.GwtIncompatible;
import com.google.common.annotations.VisibleForTesting;
import com.google.common.base.Objects;
import com.google.common.base.Preconditions;
import com.google.common.primitives.Ints;
import com.google.errorprone.annotations.CanIgnoreReturnValue;
import com.google.j2objc.annotations.WeakOuter;
import java.io.IOException;
import java.io.InvalidObjectException;
import java.io.ObjectInputStream;
import java.io.ObjectOutputStream;
import java.io.Serializable;
import java.util.AbstractMap;
import java.util.Arrays;
import java.util.Collection;
import java.util.ConcurrentModificationException;
import java.util.Iterator;
import java.util.LinkedHashMap;
import java.util.Map;
import java.util.NoSuchElementException;
import java.util.Set;
import java.util.Spliterator;
import java.util.Spliterators;
import java.util.function.BiConsumer;
import java.util.function.BiFunction;
import java.util.function.Consumer;
import org.checkerframework.checker.nullness.qual.Nullable;

/**
 * CompactHashMap is an implementation of a Map. All optional operations (put and remove) are
 * supported. Null keys and values are supported.
 *
 * 

{@code containsKey(k)}, {@code put(k, v)} and {@code remove(k)} are all (expected and * amortized) constant time operations. Expected in the hashtable sense (depends on the hash * function doing a good job of distributing the elements to the buckets to a distribution not far * from uniform), and amortized since some operations can trigger a hash table resize. * *

Unlike {@code java.util.HashMap}, iteration is only proportional to the actual {@code size()}, * which is optimal, and not the size of the internal hashtable, which could be much larger * than {@code size()}. Furthermore, this structure places significantly reduced load on the garbage * collector by only using a constant number of internal objects. * *

If there are no removals, then iteration order for the {@link #entrySet}, {@link #keySet}, and * {@link #values} views is the same as insertion order. Any removal invalidates any ordering * guarantees. * *

This class should not be assumed to be universally superior to {@code java.util.HashMap}. * Generally speaking, this class reduces object allocation and memory consumption at the price of * moderately increased constant factors of CPU. Only use this class when there is a specific reason * to prioritize memory over CPU. * * @author Louis Wasserman * @author Jon Noack */ @GwtIncompatible // not worth using in GWT for now class CompactHashMap extends AbstractMap implements Serializable { /* * TODO: Make this a drop-in replacement for j.u. versions, actually drop them in, and test the * world. Figure out what sort of space-time tradeoff we're actually going to get here with the * *Map variants. This class is particularly hard to benchmark, because the benefit is not only in * less allocation, but also having the GC do less work to scan the heap because of fewer * references, which is particularly hard to quantify. */ /** Creates an empty {@code CompactHashMap} instance. */ public static CompactHashMap create() { return new CompactHashMap<>(); } /** * Creates a {@code CompactHashMap} instance, with a high enough "initial capacity" that it * should hold {@code expectedSize} elements without growth. * * @param expectedSize the number of elements you expect to add to the returned set * @return a new, empty {@code CompactHashMap} with enough capacity to hold {@code expectedSize} * elements without resizing * @throws IllegalArgumentException if {@code expectedSize} is negative */ public static CompactHashMap createWithExpectedSize(int expectedSize) { return new CompactHashMap<>(expectedSize); } private static final Object NOT_FOUND = new Object(); /** * Maximum allowed false positive probability of detecting a hash flooding attack given random * input. */ @VisibleForTesting( ) static final double HASH_FLOODING_FPP = 0.001; /** * Maximum allowed length of a hash table bucket before falling back to a j.u.LinkedHashMap-based * implementation. Experimentally determined. */ private static final int MAX_HASH_BUCKET_LENGTH = 9; /** * The hashtable object. This can be either: * *

    *
  • a byte[], short[], or int[], with size a power of two, created by * CompactHashing.createTable, whose values are either *
      *
    • UNSET, meaning "null pointer" *
    • one plus an index into the keys, values, and entries arrays *
    *
  • another java.util.Map delegate implementation. In most modern JDKs, normal java.util hash * collections intelligently fall back to a binary search tree if hash table collisions are * detected. Rather than going to all the trouble of reimplementing this ourselves, we * simply switch over to use the JDK implementation wholesale if probable hash flooding is * detected, sacrificing the compactness guarantee in very rare cases in exchange for much * more reliable worst-case behavior. *
  • null, if no entries have yet been added to the map *
*/ @Nullable private transient Object table; /** * Contains the logical entries, in the range of [0, size()). The high bits of each int are the * part of the smeared hash of the key not covered by the hashtable mask, whereas the low bits are * the "next" pointer (pointing to the next entry in the bucket chain), which will always be less * than or equal to the hashtable mask. * *
   * hash  = aaaaaaaa
   * mask  = 0000ffff
   * next  = 0000bbbb
   * entry = aaaabbbb
   * 
* *

The pointers in [size(), entries.length) are all "null" (UNSET). */ @VisibleForTesting transient int @Nullable [] entries; /** * The keys of the entries in the map, in the range of [0, size()). The keys in [size(), * keys.length) are all {@code null}. */ @VisibleForTesting transient Object @Nullable [] keys; /** * The values of the entries in the map, in the range of [0, size()). The values in [size(), * values.length) are all {@code null}. */ @VisibleForTesting transient Object @Nullable [] values; /** * Keeps track of metadata like the number of hash table bits and modifications of this data * structure (to make it possible to throw ConcurrentModificationException in the iterator). Note * that we choose not to make this volatile, so we do less of a "best effort" to track such * errors, for better performance. */ private transient int metadata; /** The number of elements contained in the set. */ private transient int size; /** Constructs a new empty instance of {@code CompactHashMap}. */ CompactHashMap() { init(CompactHashing.DEFAULT_SIZE); } /** * Constructs a new instance of {@code CompactHashMap} with the specified capacity. * * @param expectedSize the initial capacity of this {@code CompactHashMap}. */ CompactHashMap(int expectedSize) { init(expectedSize); } /** Pseudoconstructor for serialization support. */ void init(int expectedSize) { Preconditions.checkArgument(expectedSize >= 0, "Expected size must be >= 0"); // Save expectedSize for use in allocArrays() this.metadata = Ints.constrainToRange(expectedSize, 1, CompactHashing.MAX_SIZE); } /** Returns whether arrays need to be allocated. */ @VisibleForTesting boolean needsAllocArrays() { return table == null; } /** Handle lazy allocation of arrays. */ @CanIgnoreReturnValue int allocArrays() { Preconditions.checkState(needsAllocArrays(), "Arrays already allocated"); int expectedSize = metadata; int buckets = CompactHashing.tableSize(expectedSize); this.table = CompactHashing.createTable(buckets); setHashTableMask(buckets - 1); this.entries = new int[expectedSize]; this.keys = new Object[expectedSize]; this.values = new Object[expectedSize]; return expectedSize; } @SuppressWarnings("unchecked") @VisibleForTesting @Nullable Map delegateOrNull() { if (table instanceof Map) { return (Map) table; } return null; } Map createHashFloodingResistantDelegate(int tableSize) { return new LinkedHashMap<>(tableSize, 1.0f); } @SuppressWarnings("unchecked") @VisibleForTesting @CanIgnoreReturnValue Map convertToHashFloodingResistantImplementation() { Map newDelegate = createHashFloodingResistantDelegate(hashTableMask() + 1); for (int i = firstEntryIndex(); i >= 0; i = getSuccessor(i)) { newDelegate.put((K) keys[i], (V) values[i]); } this.table = newDelegate; this.entries = null; this.keys = null; this.values = null; incrementModCount(); return newDelegate; } /** Stores the hash table mask as the number of bits needed to represent an index. */ private void setHashTableMask(int mask) { int hashTableBits = Integer.SIZE - Integer.numberOfLeadingZeros(mask); metadata = CompactHashing.maskCombine(metadata, hashTableBits, CompactHashing.HASH_TABLE_BITS_MASK); } /** Gets the hash table mask using the stored number of hash table bits. */ private int hashTableMask() { return (1 << (metadata & CompactHashing.HASH_TABLE_BITS_MASK)) - 1; } void incrementModCount() { metadata += CompactHashing.MODIFICATION_COUNT_INCREMENT; } /** * Mark an access of the specified entry. Used only in {@code CompactLinkedHashMap} for LRU * ordering. */ void accessEntry(int index) { // no-op by default } @CanIgnoreReturnValue @Override public @Nullable V put(@Nullable K key, @Nullable V value) { if (needsAllocArrays()) { allocArrays(); } @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.put(key, value); } int[] entries = this.entries; Object[] keys = this.keys; Object[] values = this.values; int newEntryIndex = this.size; // current size, and pointer to the entry to be appended int newSize = newEntryIndex + 1; int hash = smearedHash(key); int mask = hashTableMask(); int tableIndex = hash & mask; int next = CompactHashing.tableGet(table, tableIndex); if (next == UNSET) { // uninitialized bucket if (newSize > mask) { // Resize and add new entry mask = resizeTable(mask, CompactHashing.newCapacity(mask), hash, newEntryIndex); } else { CompactHashing.tableSet(table, tableIndex, newEntryIndex + 1); } } else { int entryIndex; int entry; int hashPrefix = CompactHashing.getHashPrefix(hash, mask); int bucketLength = 0; do { entryIndex = next - 1; entry = entries[entryIndex]; if (CompactHashing.getHashPrefix(entry, mask) == hashPrefix && Objects.equal(key, keys[entryIndex])) { @SuppressWarnings("unchecked") // known to be a V @Nullable V oldValue = (V) values[entryIndex]; values[entryIndex] = value; accessEntry(entryIndex); return oldValue; } next = CompactHashing.getNext(entry, mask); bucketLength++; } while (next != UNSET); if (bucketLength >= MAX_HASH_BUCKET_LENGTH) { return convertToHashFloodingResistantImplementation().put(key, value); } if (newSize > mask) { // Resize and add new entry mask = resizeTable(mask, CompactHashing.newCapacity(mask), hash, newEntryIndex); } else { entries[entryIndex] = CompactHashing.maskCombine(entry, newEntryIndex + 1, mask); } } resizeMeMaybe(newSize); insertEntry(newEntryIndex, key, value, hash, mask); this.size = newSize; incrementModCount(); return null; } /** * Creates a fresh entry with the specified object at the specified position in the entry arrays. */ void insertEntry(int entryIndex, @Nullable K key, @Nullable V value, int hash, int mask) { this.entries[entryIndex] = CompactHashing.maskCombine(hash, UNSET, mask); this.keys[entryIndex] = key; this.values[entryIndex] = value; } /** Resizes the entries storage if necessary. */ private void resizeMeMaybe(int newSize) { int entriesSize = entries.length; if (newSize > entriesSize) { // 1.5x but round up to nearest odd (this is optimal for memory consumption on Android) int newCapacity = Math.min(CompactHashing.MAX_SIZE, (entriesSize + Math.max(1, entriesSize >>> 1)) | 1); if (newCapacity != entriesSize) { resizeEntries(newCapacity); } } } /** * Resizes the internal entries array to the specified capacity, which may be greater or less than * the current capacity. */ void resizeEntries(int newCapacity) { this.entries = Arrays.copyOf(entries, newCapacity); this.keys = Arrays.copyOf(keys, newCapacity); this.values = Arrays.copyOf(values, newCapacity); } @CanIgnoreReturnValue private int resizeTable(int mask, int newCapacity, int targetHash, int targetEntryIndex) { Object newTable = CompactHashing.createTable(newCapacity); int newMask = newCapacity - 1; if (targetEntryIndex != UNSET) { // Add target first; it must be last in the chain because its entry hasn't yet been created CompactHashing.tableSet(newTable, targetHash & newMask, targetEntryIndex + 1); } Object table = this.table; int[] entries = this.entries; // Loop over current hashtable for (int tableIndex = 0; tableIndex <= mask; tableIndex++) { int next = CompactHashing.tableGet(table, tableIndex); while (next != UNSET) { int entryIndex = next - 1; int entry = entries[entryIndex]; // Rebuild hash using entry hashPrefix and tableIndex ("hashSuffix") int hash = CompactHashing.getHashPrefix(entry, mask) | tableIndex; int newTableIndex = hash & newMask; int newNext = CompactHashing.tableGet(newTable, newTableIndex); CompactHashing.tableSet(newTable, newTableIndex, next); entries[entryIndex] = CompactHashing.maskCombine(hash, newNext, newMask); next = CompactHashing.getNext(entry, mask); } } this.table = newTable; setHashTableMask(newMask); return newMask; } private int indexOf(@Nullable Object key) { if (needsAllocArrays()) { return -1; } int hash = smearedHash(key); int mask = hashTableMask(); int next = CompactHashing.tableGet(table, hash & mask); if (next == UNSET) { return -1; } int hashPrefix = CompactHashing.getHashPrefix(hash, mask); do { int entryIndex = next - 1; int entry = entries[entryIndex]; if (CompactHashing.getHashPrefix(entry, mask) == hashPrefix && Objects.equal(key, keys[entryIndex])) { return entryIndex; } next = CompactHashing.getNext(entry, mask); } while (next != UNSET); return -1; } @Override public boolean containsKey(@Nullable Object key) { @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.containsKey(key) : indexOf(key) != -1; } @SuppressWarnings("unchecked") // known to be a V @Override public V get(@Nullable Object key) { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.get(key); } int index = indexOf(key); if (index == -1) { return null; } accessEntry(index); return (V) values[index]; } @CanIgnoreReturnValue @SuppressWarnings("unchecked") // known to be a V @Override public @Nullable V remove(@Nullable Object key) { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.remove(key); } Object oldValue = removeHelper(key); return (oldValue == NOT_FOUND) ? null : (V) oldValue; } private @Nullable Object removeHelper(@Nullable Object key) { if (needsAllocArrays()) { return NOT_FOUND; } int mask = hashTableMask(); int index = CompactHashing.remove( key, /* value= */ null, mask, table, entries, keys, /* values= */ null); if (index == -1) { return NOT_FOUND; } @Nullable Object oldValue = values[index]; moveLastEntry(index, mask); size--; incrementModCount(); return oldValue; } /** * Moves the last entry in the entry array into {@code dstIndex}, and nulls out its old position. */ void moveLastEntry(int dstIndex, int mask) { int srcIndex = size() - 1; if (dstIndex < srcIndex) { // move last entry to deleted spot @Nullable Object key = keys[srcIndex]; keys[dstIndex] = key; values[dstIndex] = values[srcIndex]; keys[srcIndex] = null; values[srcIndex] = null; // move the last entry to the removed spot, just like we moved the element entries[dstIndex] = entries[srcIndex]; entries[srcIndex] = 0; // also need to update whoever's "next" pointer was pointing to the last entry place int tableIndex = smearedHash(key) & mask; int next = CompactHashing.tableGet(table, tableIndex); int srcNext = srcIndex + 1; if (next == srcNext) { // we need to update the root pointer CompactHashing.tableSet(table, tableIndex, dstIndex + 1); } else { // we need to update a pointer in an entry int entryIndex; int entry; do { entryIndex = next - 1; entry = entries[entryIndex]; next = CompactHashing.getNext(entry, mask); } while (next != srcNext); // here, entries[entryIndex] points to the old entry location; update it entries[entryIndex] = CompactHashing.maskCombine(entry, dstIndex + 1, mask); } } else { keys[dstIndex] = null; values[dstIndex] = null; entries[dstIndex] = 0; } } int firstEntryIndex() { return isEmpty() ? -1 : 0; } int getSuccessor(int entryIndex) { return (entryIndex + 1 < size) ? entryIndex + 1 : -1; } /** * Updates the index an iterator is pointing to after a call to remove: returns the index of the * entry that should be looked at after a removal on indexRemoved, with indexBeforeRemove as the * index that *was* the next entry that would be looked at. */ int adjustAfterRemove(int indexBeforeRemove, @SuppressWarnings("unused") int indexRemoved) { return indexBeforeRemove - 1; } private abstract class Itr implements Iterator { int expectedMetadata = metadata; int currentIndex = firstEntryIndex(); int indexToRemove = -1; @Override public boolean hasNext() { return currentIndex >= 0; } abstract T getOutput(int entry); @Override public T next() { checkForConcurrentModification(); if (!hasNext()) { throw new NoSuchElementException(); } indexToRemove = currentIndex; T result = getOutput(currentIndex); currentIndex = getSuccessor(currentIndex); return result; } @Override public void remove() { checkForConcurrentModification(); checkRemove(indexToRemove >= 0); incrementExpectedModCount(); CompactHashMap.this.remove(keys[indexToRemove]); currentIndex = adjustAfterRemove(currentIndex, indexToRemove); indexToRemove = -1; } void incrementExpectedModCount() { expectedMetadata += CompactHashing.MODIFICATION_COUNT_INCREMENT; } private void checkForConcurrentModification() { if (metadata != expectedMetadata) { throw new ConcurrentModificationException(); } } } @SuppressWarnings("unchecked") // known to be Ks and Vs @Override public void replaceAll(BiFunction function) { checkNotNull(function); @Nullable Map delegate = delegateOrNull(); if (delegate != null) { delegate.replaceAll(function); } else { for (int i = 0; i < size; i++) { values[i] = function.apply((K) keys[i], (V) values[i]); } } } private transient @Nullable Set keySetView; @Override public Set keySet() { return (keySetView == null) ? keySetView = createKeySet() : keySetView; } Set createKeySet() { return new KeySetView(); } @WeakOuter class KeySetView extends Maps.KeySet { KeySetView() { super(CompactHashMap.this); } @Override public Object[] toArray() { if (needsAllocArrays()) { return new Object[0]; } @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.keySet().toArray() : ObjectArrays.copyAsObjectArray(keys, 0, size); } @Override public T[] toArray(T[] a) { if (needsAllocArrays()) { if (a.length > 0) { a[0] = null; } return a; } @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.keySet().toArray(a) : ObjectArrays.toArrayImpl(keys, 0, size, a); } @Override public boolean remove(@Nullable Object o) { @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.keySet().remove(o) : CompactHashMap.this.removeHelper(o) != NOT_FOUND; } @Override public Iterator iterator() { return keySetIterator(); } @Override public Spliterator spliterator() { if (needsAllocArrays()) { return Spliterators.spliterator(new Object[0], Spliterator.DISTINCT | Spliterator.ORDERED); } @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.keySet().spliterator() : Spliterators.spliterator(keys, 0, size, Spliterator.DISTINCT | Spliterator.ORDERED); } @SuppressWarnings("unchecked") // known to be Ks @Override public void forEach(Consumer action) { checkNotNull(action); @Nullable Map delegate = delegateOrNull(); if (delegate != null) { delegate.keySet().forEach(action); } else { for (int i = firstEntryIndex(); i >= 0; i = getSuccessor(i)) { action.accept((K) keys[i]); } } } } Iterator keySetIterator() { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.keySet().iterator(); } return new Itr() { @SuppressWarnings("unchecked") // known to be a K @Override K getOutput(int entry) { return (K) keys[entry]; } }; } @SuppressWarnings("unchecked") // known to be Ks and Vs @Override public void forEach(BiConsumer action) { checkNotNull(action); @Nullable Map delegate = delegateOrNull(); if (delegate != null) { delegate.forEach(action); } else { for (int i = firstEntryIndex(); i >= 0; i = getSuccessor(i)) { action.accept((K) keys[i], (V) values[i]); } } } private transient @Nullable Set> entrySetView; @Override public Set> entrySet() { return (entrySetView == null) ? entrySetView = createEntrySet() : entrySetView; } Set> createEntrySet() { return new EntrySetView(); } @WeakOuter class EntrySetView extends Maps.EntrySet { @Override Map map() { return CompactHashMap.this; } @Override public Iterator> iterator() { return entrySetIterator(); } @Override public Spliterator> spliterator() { @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.entrySet().spliterator() : CollectSpliterators.indexed( size, Spliterator.DISTINCT | Spliterator.ORDERED, MapEntry::new); } @Override public boolean contains(@Nullable Object o) { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.entrySet().contains(o); } else if (o instanceof Entry) { Entry entry = (Entry) o; int index = indexOf(entry.getKey()); return index != -1 && Objects.equal(values[index], entry.getValue()); } return false; } @Override public boolean remove(@Nullable Object o) { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.entrySet().remove(o); } else if (o instanceof Entry) { Entry entry = (Entry) o; if (needsAllocArrays()) { return false; } int mask = hashTableMask(); int index = CompactHashing.remove( entry.getKey(), entry.getValue(), mask, table, entries, keys, values); if (index == -1) { return false; } moveLastEntry(index, mask); size--; incrementModCount(); return true; } return false; } } Iterator> entrySetIterator() { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.entrySet().iterator(); } return new Itr>() { @Override Entry getOutput(int entry) { return new MapEntry(entry); } }; } final class MapEntry extends AbstractMapEntry { private final @Nullable K key; private int lastKnownIndex; @SuppressWarnings("unchecked") // known to be a K MapEntry(int index) { this.key = (K) keys[index]; this.lastKnownIndex = index; } @Nullable @Override public K getKey() { return key; } private void updateLastKnownIndex() { if (lastKnownIndex == -1 || lastKnownIndex >= size() || !Objects.equal(key, keys[lastKnownIndex])) { lastKnownIndex = indexOf(key); } } @SuppressWarnings("unchecked") // known to be a V @Override public @Nullable V getValue() { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.get(key); } updateLastKnownIndex(); return (lastKnownIndex == -1) ? null : (V) values[lastKnownIndex]; } @SuppressWarnings("unchecked") // known to be a V @Override public V setValue(V value) { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.put(key, value); } updateLastKnownIndex(); if (lastKnownIndex == -1) { put(key, value); return null; } else { V old = (V) values[lastKnownIndex]; values[lastKnownIndex] = value; return old; } } } @Override public int size() { @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.size() : size; } @Override public boolean isEmpty() { return size() == 0; } @Override public boolean containsValue(@Nullable Object value) { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.containsValue(value); } for (int i = 0; i < size; i++) { if (Objects.equal(value, values[i])) { return true; } } return false; } private transient @Nullable Collection valuesView; @Override public Collection values() { return (valuesView == null) ? valuesView = createValues() : valuesView; } Collection createValues() { return new ValuesView(); } @WeakOuter class ValuesView extends Maps.Values { ValuesView() { super(CompactHashMap.this); } @Override public Iterator iterator() { return valuesIterator(); } @SuppressWarnings("unchecked") // known to be Vs @Override public void forEach(Consumer action) { checkNotNull(action); @Nullable Map delegate = delegateOrNull(); if (delegate != null) { delegate.values().forEach(action); } else { for (int i = firstEntryIndex(); i >= 0; i = getSuccessor(i)) { action.accept((V) values[i]); } } } @Override public Spliterator spliterator() { if (needsAllocArrays()) { return Spliterators.spliterator(new Object[0], Spliterator.ORDERED); } @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.values().spliterator() : Spliterators.spliterator(values, 0, size, Spliterator.ORDERED); } @Override public Object[] toArray() { if (needsAllocArrays()) { return new Object[0]; } @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.values().toArray() : ObjectArrays.copyAsObjectArray(values, 0, size); } @Override public T[] toArray(T[] a) { if (needsAllocArrays()) { if (a.length > 0) { a[0] = null; } return a; } @Nullable Map delegate = delegateOrNull(); return (delegate != null) ? delegate.values().toArray(a) : ObjectArrays.toArrayImpl(values, 0, size, a); } } Iterator valuesIterator() { @Nullable Map delegate = delegateOrNull(); if (delegate != null) { return delegate.values().iterator(); } return new Itr() { @SuppressWarnings("unchecked") // known to be a V @Override V getOutput(int entry) { return (V) values[entry]; } }; } /** * Ensures that this {@code CompactHashMap} has the smallest representation in memory, given its * current size. */ public void trimToSize() { if (needsAllocArrays()) { return; } @Nullable Map delegate = delegateOrNull(); if (delegate != null) { Map newDelegate = createHashFloodingResistantDelegate(size()); newDelegate.putAll(delegate); this.table = newDelegate; return; } int size = this.size; if (size < entries.length) { resizeEntries(size); } int minimumTableSize = CompactHashing.tableSize(size); int mask = hashTableMask(); if (minimumTableSize < mask) { // smaller table size will always be less than current mask resizeTable(mask, minimumTableSize, UNSET, UNSET); } } @Override public void clear() { if (needsAllocArrays()) { return; } incrementModCount(); @Nullable Map delegate = delegateOrNull(); if (delegate != null) { metadata = Ints.constrainToRange(size(), CompactHashing.DEFAULT_SIZE, CompactHashing.MAX_SIZE); delegate.clear(); // invalidate any iterators left over! table = null; size = 0; } else { Arrays.fill(keys, 0, size, null); Arrays.fill(values, 0, size, null); CompactHashing.tableClear(table); Arrays.fill(entries, 0, size, 0); this.size = 0; } } private void writeObject(ObjectOutputStream stream) throws IOException { stream.defaultWriteObject(); stream.writeInt(size()); Iterator> entryIterator = entrySetIterator(); while (entryIterator.hasNext()) { Entry e = entryIterator.next(); stream.writeObject(e.getKey()); stream.writeObject(e.getValue()); } } @SuppressWarnings("unchecked") private void readObject(ObjectInputStream stream) throws IOException, ClassNotFoundException { stream.defaultReadObject(); int elementCount = stream.readInt(); if (elementCount < 0) { throw new InvalidObjectException("Invalid size: " + elementCount); } init(elementCount); for (int i = 0; i < elementCount; i++) { K key = (K) stream.readObject(); V value = (V) stream.readObject(); put(key, value); } } }





© 2015 - 2024 Weber Informatics LLC | Privacy Policy