Class CompactHashMap<K,V>
- All Implemented Interfaces:
Serializable,Map<K,V>
- Direct Known Subclasses:
CompactLinkedHashMap
containsKey(k), put(k, v) and 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 java.util.HashMap, iteration is only proportional to the actual size(),
which is optimal, and not the size of the internal hashtable, which could be much larger
than 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 entrySet(), keySet(), and
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 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.
-
Nested Class Summary
Nested ClassesModifier and TypeClassDescription(package private) classprivate class(package private) class(package private) final class(package private) classNested classes/interfaces inherited from class java.util.AbstractMap
AbstractMap.SimpleEntry<K extends Object,V extends Object>, AbstractMap.SimpleImmutableEntry<K extends Object, V extends Object> -
Field Summary
FieldsModifier and TypeFieldDescription(package private) int[]Contains the logical entries, in the range of [0, size()).(package private) static final doubleMaximum allowed false positive probability of detecting a hash flooding attack given random input.(package private) Object[]The keys of the entries in the map, in the range of [0, size()).private static final intMaximum allowed length of a hash table bucket before falling back to a j.u.LinkedHashMap-based implementation.private intKeeps 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).private static final Objectprivate intThe number of elements contained in the set.private ObjectThe hashtable object.(package private) Object[]The values of the entries in the map, in the range of [0, size()).private Collection<V> -
Constructor Summary
ConstructorsConstructorDescriptionConstructs a new empty instance ofCompactHashMap.CompactHashMap(int expectedSize) Constructs a new instance ofCompactHashMapwith the specified capacity. -
Method Summary
Modifier and TypeMethodDescription(package private) voidaccessEntry(int index) Mark an access of the specified entry.(package private) intadjustAfterRemove(int indexBeforeRemove, int indexRemoved) 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.(package private) intHandle lazy allocation of arrays.voidclear()booleancontainsKey(Object key) booleancontainsValue(Object value) static <K,V> CompactHashMap<K, V> create()Creates an emptyCompactHashMapinstance.createHashFloodingResistantDelegate(int tableSize) (package private) Collection<V>static <K,V> CompactHashMap<K, V> createWithExpectedSize(int expectedSize) Creates aCompactHashMapinstance, with a high enough "initial capacity" that it should holdexpectedSizeelements without growth.private intentry(int i) entrySet()(package private) intvoidforEach(BiConsumer<? super K, ? super V> action) (package private) intgetSuccessor(int entryIndex) private intGets the hash table mask using the stored number of hash table bits.(package private) voidprivate int(package private) voidinit(int expectedSize) Pseudoconstructor for serialization support.(package private) voidinsertEntry(int entryIndex, K key, V value, int hash, int mask) Creates a fresh entry with the specified object at the specified position in the entry arrays.booleanisEmpty()private Kkey(int i) keySet()(package private) voidmoveLastEntry(int dstIndex, int mask) Moves the last entry in the entry array intodstIndex, and nulls out its old position.(package private) booleanReturns whether arrays need to be allocated.private voidreadObject(ObjectInputStream stream) private ObjectremoveHelper(Object key) voidreplaceAll(BiFunction<? super K, ? super V, ? extends V> function) private int[]private Object[]private Objectprivate Object[](package private) voidresizeEntries(int newCapacity) Resizes the internal entries array to the specified capacity, which may be greater or less than the current capacity.private voidresizeMeMaybe(int newSize) Resizes the entries storage if necessary.private intresizeTable(int oldMask, int newCapacity, int targetHash, int targetEntryIndex) private voidsetEntry(int i, int value) private voidsetHashTableMask(int mask) Stores the hash table mask as the number of bits needed to represent an index.private voidprivate voidintsize()voidEnsures that thisCompactHashMaphas the smallest representation in memory, given its current size.private Vvalue(int i) values()private voidwriteObject(ObjectOutputStream stream) Methods inherited from class java.lang.Object
finalize, getClass, notify, notifyAll, wait, wait, waitMethods inherited from interface java.util.Map
compute, computeIfAbsent, computeIfPresent, getOrDefault, merge, putIfAbsent, remove, replace, replace
-
Field Details
-
NOT_FOUND
-
HASH_FLOODING_FPP
static final double HASH_FLOODING_FPPMaximum allowed false positive probability of detecting a hash flooding attack given random input.- See Also:
-
MAX_HASH_BUCKET_LENGTH
private static final int MAX_HASH_BUCKET_LENGTHMaximum allowed length of a hash table bucket before falling back to a j.u.LinkedHashMap-based implementation. Experimentally determined.- See Also:
-
table
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
- a byte[], short[], or int[], with size a power of two, created by
CompactHashing.createTable, whose values are either
-
entries
@CheckForNull transient int[] entriesContains 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 = 00000fff next = 00000bbb entry = aaaaabbb
The pointers in [size(), entries.length) are all "null" (UNSET).
-
keys
The keys of the entries in the map, in the range of [0, size()). The keys in [size(), keys.length) are allnull. -
values
The values of the entries in the map, in the range of [0, size()). The values in [size(), values.length) are allnull. -
metadata
private transient int metadataKeeps 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.For a new instance, where the arrays above have not yet been allocated, the value of
metadatais the size that the arrays should be allocated with. Once the arrays have been allocated, the value ofmetadatacombines the number of bits in the "short hash", in its bottom 5 bits, with a modification count in the remaining bits that is used to detect concurrent modification during iteration. -
size
private transient int sizeThe number of elements contained in the set. -
keySetView
-
entrySetView
-
valuesView
-
-
Constructor Details
-
CompactHashMap
CompactHashMap()Constructs a new empty instance ofCompactHashMap. -
CompactHashMap
CompactHashMap(int expectedSize) Constructs a new instance ofCompactHashMapwith the specified capacity.- Parameters:
expectedSize- the initial capacity of thisCompactHashMap.
-
-
Method Details
-
create
Creates an emptyCompactHashMapinstance. -
createWithExpectedSize
Creates aCompactHashMapinstance, with a high enough "initial capacity" that it should holdexpectedSizeelements without growth.- Parameters:
expectedSize- the number of elements you expect to add to the returned set- Returns:
- a new, empty
CompactHashMapwith enough capacity to holdexpectedSizeelements without resizing - Throws:
IllegalArgumentException- ifexpectedSizeis negative
-
init
void init(int expectedSize) Pseudoconstructor for serialization support. -
needsAllocArrays
boolean needsAllocArrays()Returns whether arrays need to be allocated. -
allocArrays
int allocArrays()Handle lazy allocation of arrays. -
delegateOrNull
-
createHashFloodingResistantDelegate
-
convertToHashFloodingResistantImplementation
-
setHashTableMask
private void setHashTableMask(int mask) Stores the hash table mask as the number of bits needed to represent an index. -
hashTableMask
private int hashTableMask()Gets the hash table mask using the stored number of hash table bits. -
incrementModCount
void incrementModCount() -
accessEntry
void accessEntry(int index) Mark an access of the specified entry. Used only inCompactLinkedHashMapfor LRU ordering. -
put
-
insertEntry
Creates a fresh entry with the specified object at the specified position in the entry arrays. -
resizeMeMaybe
private void resizeMeMaybe(int newSize) Resizes the entries storage if necessary. -
resizeEntries
void resizeEntries(int newCapacity) Resizes the internal entries array to the specified capacity, which may be greater or less than the current capacity. -
resizeTable
private int resizeTable(int oldMask, int newCapacity, int targetHash, int targetEntryIndex) -
indexOf
-
containsKey
- Specified by:
containsKeyin interfaceMap<K,V> - Overrides:
containsKeyin classAbstractMap<K,V>
-
get
-
remove
-
removeHelper
-
moveLastEntry
void moveLastEntry(int dstIndex, int mask) Moves the last entry in the entry array intodstIndex, and nulls out its old position. -
firstEntryIndex
int firstEntryIndex() -
getSuccessor
int getSuccessor(int entryIndex) -
adjustAfterRemove
int adjustAfterRemove(int indexBeforeRemove, int indexRemoved) 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. -
replaceAll
- Specified by:
replaceAllin interfaceMap<K,V>
-
keySet
-
createKeySet
-
keySetIterator
-
forEach
-
entrySet
-
createEntrySet
-
entrySetIterator
-
size
public int size() -
isEmpty
public boolean isEmpty() -
containsValue
- Specified by:
containsValuein interfaceMap<K,V> - Overrides:
containsValuein classAbstractMap<K,V>
-
values
-
createValues
Collection<V> createValues() -
valuesIterator
-
trimToSize
public void trimToSize()Ensures that thisCompactHashMaphas the smallest representation in memory, given its current size. -
clear
public void clear() -
writeObject
- Throws:
IOException
-
readObject
- Throws:
IOExceptionClassNotFoundException
-
requireTable
-
requireEntries
private int[] requireEntries() -
requireKeys
-
requireValues
-
key
-
value
-
entry
private int entry(int i) -
setKey
-
setValue
-
setEntry
private void setEntry(int i, int value)
-