blob: d21518abf6f14d575756157840fd830ebd38a59a [file] [log] [blame]
/*
* Copyright (c) 2012, 2013, Oracle and/or its affiliates. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
* This code is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License version 2 only, as
* published by the Free Software Foundation. Oracle designates this
* particular file as subject to the "Classpath" exception as provided
* by Oracle in the LICENSE file that accompanied this code.
*
* This code is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
* version 2 for more details (a copy is included in the LICENSE file that
* accompanied this code).
*
* You should have received a copy of the GNU General Public License version
* 2 along with this work; if not, write to the Free Software Foundation,
* Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
*
* Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
* or visit www.oracle.com if you need additional information or have any
* questions.
*/
package java.util.stream;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Iterator;
import java.util.List;
import java.util.Objects;
import java.util.PrimitiveIterator;
import java.util.Spliterator;
import java.util.Spliterators;
import java.util.function.Consumer;
import java.util.function.DoubleConsumer;
import java.util.function.IntConsumer;
import java.util.function.IntFunction;
import java.util.function.LongConsumer;
/**
* An ordered collection of elements. Elements can be added, but not removed.
* Goes through a building phase, during which elements can be added, and a
* traversal phase, during which elements can be traversed in order but no
* further modifications are possible.
*
* <p> One or more arrays are used to store elements. The use of a multiple
* arrays has better performance characteristics than a single array used by
* {@link ArrayList}, as when the capacity of the list needs to be increased
* no copying of elements is required. This is usually beneficial in the case
* where the results will be traversed a small number of times.
*
* @param <E> the type of elements in this list
* @since 1.8
* @hide Visible for CTS testing only (OpenJDK8 tests).
*/
// Android-changed: Made public for CTS tests only.
public class SpinedBuffer<E>
extends AbstractSpinedBuffer
implements Consumer<E>, Iterable<E> {
/*
* We optimistically hope that all the data will fit into the first chunk,
* so we try to avoid inflating the spine[] and priorElementCount[] arrays
* prematurely. So methods must be prepared to deal with these arrays being
* null. If spine is non-null, then spineIndex points to the current chunk
* within the spine, otherwise it is zero. The spine and priorElementCount
* arrays are always the same size, and for any i <= spineIndex,
* priorElementCount[i] is the sum of the sizes of all the prior chunks.
*
* The curChunk pointer is always valid. The elementIndex is the index of
* the next element to be written in curChunk; this may be past the end of
* curChunk so we have to check before writing. When we inflate the spine
* array, curChunk becomes the first element in it. When we clear the
* buffer, we discard all chunks except the first one, which we clear,
* restoring it to the initial single-chunk state.
*/
/**
* Chunk that we're currently writing into; may or may not be aliased with
* the first element of the spine.
*/
protected E[] curChunk;
/**
* All chunks, or null if there is only one chunk.
*/
protected E[][] spine;
/**
* Constructs an empty list with the specified initial capacity.
*
* @param initialCapacity the initial capacity of the list
* @throws IllegalArgumentException if the specified initial capacity
* is negative
*/
@SuppressWarnings("unchecked")
// Android-changed: Made public for CTS tests only.
public SpinedBuffer(int initialCapacity) {
super(initialCapacity);
curChunk = (E[]) new Object[1 << initialChunkPower];
}
/**
* Constructs an empty list with an initial capacity of sixteen.
*/
@SuppressWarnings("unchecked")
// Android-changed: Made public for CTS tests only.
public SpinedBuffer() {
super();
curChunk = (E[]) new Object[1 << initialChunkPower];
}
/**
* Returns the current capacity of the buffer
*/
protected long capacity() {
return (spineIndex == 0)
? curChunk.length
: priorElementCount[spineIndex] + spine[spineIndex].length;
}
@SuppressWarnings("unchecked")
private void inflateSpine() {
if (spine == null) {
spine = (E[][]) new Object[MIN_SPINE_SIZE][];
priorElementCount = new long[MIN_SPINE_SIZE];
spine[0] = curChunk;
}
}
/**
* Ensure that the buffer has at least capacity to hold the target size
*/
@SuppressWarnings("unchecked")
protected final void ensureCapacity(long targetSize) {
long capacity = capacity();
if (targetSize > capacity) {
inflateSpine();
for (int i=spineIndex+1; targetSize > capacity; i++) {
if (i >= spine.length) {
int newSpineSize = spine.length * 2;
spine = Arrays.copyOf(spine, newSpineSize);
priorElementCount = Arrays.copyOf(priorElementCount, newSpineSize);
}
int nextChunkSize = chunkSize(i);
spine[i] = (E[]) new Object[nextChunkSize];
priorElementCount[i] = priorElementCount[i-1] + spine[i-1].length;
capacity += nextChunkSize;
}
}
}
/**
* Force the buffer to increase its capacity.
*/
protected void increaseCapacity() {
ensureCapacity(capacity() + 1);
}
/**
* Retrieve the element at the specified index.
*/
public E get(long index) {
// @@@ can further optimize by caching last seen spineIndex,
// which is going to be right most of the time
// Casts to int are safe since the spine array index is the index minus
// the prior element count from the current spine
if (spineIndex == 0) {
if (index < elementIndex)
return curChunk[((int) index)];
else
throw new IndexOutOfBoundsException(Long.toString(index));
}
if (index >= count())
throw new IndexOutOfBoundsException(Long.toString(index));
for (int j=0; j <= spineIndex; j++)
if (index < priorElementCount[j] + spine[j].length)
return spine[j][((int) (index - priorElementCount[j]))];
throw new IndexOutOfBoundsException(Long.toString(index));
}
/**
* Copy the elements, starting at the specified offset, into the specified
* array.
*/
public void copyInto(E[] array, int offset) {
long finalOffset = offset + count();
if (finalOffset > array.length || finalOffset < offset) {
throw new IndexOutOfBoundsException("does not fit");
}
if (spineIndex == 0)
System.arraycopy(curChunk, 0, array, offset, elementIndex);
else {
// full chunks
for (int i=0; i < spineIndex; i++) {
System.arraycopy(spine[i], 0, array, offset, spine[i].length);
offset += spine[i].length;
}
if (elementIndex > 0)
System.arraycopy(curChunk, 0, array, offset, elementIndex);
}
}
/**
* Create a new array using the specified array factory, and copy the
* elements into it.
*/
public E[] asArray(IntFunction<E[]> arrayFactory) {
long size = count();
if (size >= Nodes.MAX_ARRAY_SIZE)
throw new IllegalArgumentException(Nodes.BAD_SIZE);
E[] result = arrayFactory.apply((int) size);
copyInto(result, 0);
return result;
}
@Override
public void clear() {
if (spine != null) {
curChunk = spine[0];
for (int i=0; i<curChunk.length; i++)
curChunk[i] = null;
spine = null;
priorElementCount = null;
}
else {
for (int i=0; i<elementIndex; i++)
curChunk[i] = null;
}
elementIndex = 0;
spineIndex = 0;
}
@Override
public Iterator<E> iterator() {
return Spliterators.iterator(spliterator());
}
@Override
public void forEach(Consumer<? super E> consumer) {
// completed chunks, if any
for (int j = 0; j < spineIndex; j++)
for (E t : spine[j])
consumer.accept(t);
// current chunk
for (int i=0; i<elementIndex; i++)
consumer.accept(curChunk[i]);
}
@Override
public void accept(E e) {
if (elementIndex == curChunk.length) {
inflateSpine();
if (spineIndex+1 >= spine.length || spine[spineIndex+1] == null)
increaseCapacity();
elementIndex = 0;
++spineIndex;
curChunk = spine[spineIndex];
}
curChunk[elementIndex++] = e;
}
@Override
public String toString() {
List<E> list = new ArrayList<>();
forEach(list::add);
return "SpinedBuffer:" + list.toString();
}
private static final int SPLITERATOR_CHARACTERISTICS
= Spliterator.SIZED | Spliterator.ORDERED | Spliterator.SUBSIZED;
/**
* Return a {@link Spliterator} describing the contents of the buffer.
*/
public Spliterator<E> spliterator() {
class Splitr implements Spliterator<E> {
// The current spine index
int splSpineIndex;
// Last spine index
final int lastSpineIndex;
// The current element index into the current spine
int splElementIndex;
// Last spine's last element index + 1
final int lastSpineElementFence;
// When splSpineIndex >= lastSpineIndex and
// splElementIndex >= lastSpineElementFence then
// this spliterator is fully traversed
// tryAdvance can set splSpineIndex > spineIndex if the last spine is full
// The current spine array
E[] splChunk;
Splitr(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
this.splSpineIndex = firstSpineIndex;
this.lastSpineIndex = lastSpineIndex;
this.splElementIndex = firstSpineElementIndex;
this.lastSpineElementFence = lastSpineElementFence;
assert spine != null || firstSpineIndex == 0 && lastSpineIndex == 0;
splChunk = (spine == null) ? curChunk : spine[firstSpineIndex];
}
@Override
public long estimateSize() {
return (splSpineIndex == lastSpineIndex)
? (long) lastSpineElementFence - splElementIndex
: // # of elements prior to end -
priorElementCount[lastSpineIndex] + lastSpineElementFence -
// # of elements prior to current
priorElementCount[splSpineIndex] - splElementIndex;
}
@Override
public int characteristics() {
return SPLITERATOR_CHARACTERISTICS;
}
@Override
public boolean tryAdvance(Consumer<? super E> consumer) {
Objects.requireNonNull(consumer);
if (splSpineIndex < lastSpineIndex
|| (splSpineIndex == lastSpineIndex && splElementIndex < lastSpineElementFence)) {
consumer.accept(splChunk[splElementIndex++]);
if (splElementIndex == splChunk.length) {
splElementIndex = 0;
++splSpineIndex;
if (spine != null && splSpineIndex <= lastSpineIndex)
splChunk = spine[splSpineIndex];
}
return true;
}
return false;
}
@Override
public void forEachRemaining(Consumer<? super E> consumer) {
Objects.requireNonNull(consumer);
if (splSpineIndex < lastSpineIndex
|| (splSpineIndex == lastSpineIndex && splElementIndex < lastSpineElementFence)) {
int i = splElementIndex;
// completed chunks, if any
for (int sp = splSpineIndex; sp < lastSpineIndex; sp++) {
E[] chunk = spine[sp];
for (; i < chunk.length; i++) {
consumer.accept(chunk[i]);
}
i = 0;
}
// last (or current uncompleted) chunk
E[] chunk = (splSpineIndex == lastSpineIndex) ? splChunk : spine[lastSpineIndex];
int hElementIndex = lastSpineElementFence;
for (; i < hElementIndex; i++) {
consumer.accept(chunk[i]);
}
// mark consumed
splSpineIndex = lastSpineIndex;
splElementIndex = lastSpineElementFence;
}
}
@Override
public Spliterator<E> trySplit() {
if (splSpineIndex < lastSpineIndex) {
// split just before last chunk (if it is full this means 50:50 split)
Spliterator<E> ret = new Splitr(splSpineIndex, lastSpineIndex - 1,
splElementIndex, spine[lastSpineIndex-1].length);
// position to start of last chunk
splSpineIndex = lastSpineIndex;
splElementIndex = 0;
splChunk = spine[splSpineIndex];
return ret;
}
else if (splSpineIndex == lastSpineIndex) {
int t = (lastSpineElementFence - splElementIndex) / 2;
if (t == 0)
return null;
else {
Spliterator<E> ret = Arrays.spliterator(splChunk, splElementIndex, splElementIndex + t);
splElementIndex += t;
return ret;
}
}
else {
return null;
}
}
}
return new Splitr(0, spineIndex, 0, elementIndex);
}
/**
* An ordered collection of primitive values. Elements can be added, but
* not removed. Goes through a building phase, during which elements can be
* added, and a traversal phase, during which elements can be traversed in
* order but no further modifications are possible.
*
* <p> One or more arrays are used to store elements. The use of a multiple
* arrays has better performance characteristics than a single array used by
* {@link ArrayList}, as when the capacity of the list needs to be increased
* no copying of elements is required. This is usually beneficial in the case
* where the results will be traversed a small number of times.
*
* @param <E> the wrapper type for this primitive type
* @param <T_ARR> the array type for this primitive type
* @param <T_CONS> the Consumer type for this primitive type
*/
abstract static class OfPrimitive<E, T_ARR, T_CONS>
extends AbstractSpinedBuffer implements Iterable<E> {
/*
* We optimistically hope that all the data will fit into the first chunk,
* so we try to avoid inflating the spine[] and priorElementCount[] arrays
* prematurely. So methods must be prepared to deal with these arrays being
* null. If spine is non-null, then spineIndex points to the current chunk
* within the spine, otherwise it is zero. The spine and priorElementCount
* arrays are always the same size, and for any i <= spineIndex,
* priorElementCount[i] is the sum of the sizes of all the prior chunks.
*
* The curChunk pointer is always valid. The elementIndex is the index of
* the next element to be written in curChunk; this may be past the end of
* curChunk so we have to check before writing. When we inflate the spine
* array, curChunk becomes the first element in it. When we clear the
* buffer, we discard all chunks except the first one, which we clear,
* restoring it to the initial single-chunk state.
*/
// The chunk we're currently writing into
T_ARR curChunk;
// All chunks, or null if there is only one chunk
T_ARR[] spine;
/**
* Constructs an empty list with the specified initial capacity.
*
* @param initialCapacity the initial capacity of the list
* @throws IllegalArgumentException if the specified initial capacity
* is negative
*/
OfPrimitive(int initialCapacity) {
super(initialCapacity);
curChunk = newArray(1 << initialChunkPower);
}
/**
* Constructs an empty list with an initial capacity of sixteen.
*/
OfPrimitive() {
super();
curChunk = newArray(1 << initialChunkPower);
}
@Override
public abstract Iterator<E> iterator();
@Override
public abstract void forEach(Consumer<? super E> consumer);
/** Create a new array-of-array of the proper type and size */
protected abstract T_ARR[] newArrayArray(int size);
/** Create a new array of the proper type and size */
public abstract T_ARR newArray(int size);
/** Get the length of an array */
protected abstract int arrayLength(T_ARR array);
/** Iterate an array with the provided consumer */
protected abstract void arrayForEach(T_ARR array, int from, int to,
T_CONS consumer);
protected long capacity() {
return (spineIndex == 0)
? arrayLength(curChunk)
: priorElementCount[spineIndex] + arrayLength(spine[spineIndex]);
}
private void inflateSpine() {
if (spine == null) {
spine = newArrayArray(MIN_SPINE_SIZE);
priorElementCount = new long[MIN_SPINE_SIZE];
spine[0] = curChunk;
}
}
protected final void ensureCapacity(long targetSize) {
long capacity = capacity();
if (targetSize > capacity) {
inflateSpine();
for (int i=spineIndex+1; targetSize > capacity; i++) {
if (i >= spine.length) {
int newSpineSize = spine.length * 2;
spine = Arrays.copyOf(spine, newSpineSize);
priorElementCount = Arrays.copyOf(priorElementCount, newSpineSize);
}
int nextChunkSize = chunkSize(i);
spine[i] = newArray(nextChunkSize);
priorElementCount[i] = priorElementCount[i-1] + arrayLength(spine[i - 1]);
capacity += nextChunkSize;
}
}
}
protected void increaseCapacity() {
ensureCapacity(capacity() + 1);
}
protected int chunkFor(long index) {
if (spineIndex == 0) {
if (index < elementIndex)
return 0;
else
throw new IndexOutOfBoundsException(Long.toString(index));
}
if (index >= count())
throw new IndexOutOfBoundsException(Long.toString(index));
for (int j=0; j <= spineIndex; j++)
if (index < priorElementCount[j] + arrayLength(spine[j]))
return j;
throw new IndexOutOfBoundsException(Long.toString(index));
}
public void copyInto(T_ARR array, int offset) {
long finalOffset = offset + count();
if (finalOffset > arrayLength(array) || finalOffset < offset) {
throw new IndexOutOfBoundsException("does not fit");
}
if (spineIndex == 0)
System.arraycopy(curChunk, 0, array, offset, elementIndex);
else {
// full chunks
for (int i=0; i < spineIndex; i++) {
System.arraycopy(spine[i], 0, array, offset, arrayLength(spine[i]));
offset += arrayLength(spine[i]);
}
if (elementIndex > 0)
System.arraycopy(curChunk, 0, array, offset, elementIndex);
}
}
public T_ARR asPrimitiveArray() {
long size = count();
if (size >= Nodes.MAX_ARRAY_SIZE)
throw new IllegalArgumentException(Nodes.BAD_SIZE);
T_ARR result = newArray((int) size);
copyInto(result, 0);
return result;
}
protected void preAccept() {
if (elementIndex == arrayLength(curChunk)) {
inflateSpine();
if (spineIndex+1 >= spine.length || spine[spineIndex+1] == null)
increaseCapacity();
elementIndex = 0;
++spineIndex;
curChunk = spine[spineIndex];
}
}
public void clear() {
if (spine != null) {
curChunk = spine[0];
spine = null;
priorElementCount = null;
}
elementIndex = 0;
spineIndex = 0;
}
@SuppressWarnings("overloads")
public void forEach(T_CONS consumer) {
// completed chunks, if any
for (int j = 0; j < spineIndex; j++)
arrayForEach(spine[j], 0, arrayLength(spine[j]), consumer);
// current chunk
arrayForEach(curChunk, 0, elementIndex, consumer);
}
abstract class BaseSpliterator<T_SPLITR extends Spliterator.OfPrimitive<E, T_CONS, T_SPLITR>>
implements Spliterator.OfPrimitive<E, T_CONS, T_SPLITR> {
// The current spine index
int splSpineIndex;
// Last spine index
final int lastSpineIndex;
// The current element index into the current spine
int splElementIndex;
// Last spine's last element index + 1
final int lastSpineElementFence;
// When splSpineIndex >= lastSpineIndex and
// splElementIndex >= lastSpineElementFence then
// this spliterator is fully traversed
// tryAdvance can set splSpineIndex > spineIndex if the last spine is full
// The current spine array
T_ARR splChunk;
BaseSpliterator(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
this.splSpineIndex = firstSpineIndex;
this.lastSpineIndex = lastSpineIndex;
this.splElementIndex = firstSpineElementIndex;
this.lastSpineElementFence = lastSpineElementFence;
assert spine != null || firstSpineIndex == 0 && lastSpineIndex == 0;
splChunk = (spine == null) ? curChunk : spine[firstSpineIndex];
}
abstract T_SPLITR newSpliterator(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence);
abstract void arrayForOne(T_ARR array, int index, T_CONS consumer);
abstract T_SPLITR arraySpliterator(T_ARR array, int offset, int len);
@Override
public long estimateSize() {
return (splSpineIndex == lastSpineIndex)
? (long) lastSpineElementFence - splElementIndex
: // # of elements prior to end -
priorElementCount[lastSpineIndex] + lastSpineElementFence -
// # of elements prior to current
priorElementCount[splSpineIndex] - splElementIndex;
}
@Override
public int characteristics() {
return SPLITERATOR_CHARACTERISTICS;
}
@Override
public boolean tryAdvance(T_CONS consumer) {
Objects.requireNonNull(consumer);
if (splSpineIndex < lastSpineIndex
|| (splSpineIndex == lastSpineIndex && splElementIndex < lastSpineElementFence)) {
arrayForOne(splChunk, splElementIndex++, consumer);
if (splElementIndex == arrayLength(splChunk)) {
splElementIndex = 0;
++splSpineIndex;
if (spine != null && splSpineIndex <= lastSpineIndex)
splChunk = spine[splSpineIndex];
}
return true;
}
return false;
}
@Override
public void forEachRemaining(T_CONS consumer) {
Objects.requireNonNull(consumer);
if (splSpineIndex < lastSpineIndex
|| (splSpineIndex == lastSpineIndex && splElementIndex < lastSpineElementFence)) {
int i = splElementIndex;
// completed chunks, if any
for (int sp = splSpineIndex; sp < lastSpineIndex; sp++) {
T_ARR chunk = spine[sp];
arrayForEach(chunk, i, arrayLength(chunk), consumer);
i = 0;
}
// last (or current uncompleted) chunk
T_ARR chunk = (splSpineIndex == lastSpineIndex) ? splChunk : spine[lastSpineIndex];
arrayForEach(chunk, i, lastSpineElementFence, consumer);
// mark consumed
splSpineIndex = lastSpineIndex;
splElementIndex = lastSpineElementFence;
}
}
@Override
public T_SPLITR trySplit() {
if (splSpineIndex < lastSpineIndex) {
// split just before last chunk (if it is full this means 50:50 split)
T_SPLITR ret = newSpliterator(splSpineIndex, lastSpineIndex - 1,
splElementIndex, arrayLength(spine[lastSpineIndex - 1]));
// position us to start of last chunk
splSpineIndex = lastSpineIndex;
splElementIndex = 0;
splChunk = spine[splSpineIndex];
return ret;
}
else if (splSpineIndex == lastSpineIndex) {
int t = (lastSpineElementFence - splElementIndex) / 2;
if (t == 0)
return null;
else {
T_SPLITR ret = arraySpliterator(splChunk, splElementIndex, t);
splElementIndex += t;
return ret;
}
}
else {
return null;
}
}
}
}
/**
* An ordered collection of {@code int} values.
* @hide Visible for CTS testing only (OpenJDK8 tests).
*/
// Android-changed: Made public for CTS tests only.
public static class OfInt extends SpinedBuffer.OfPrimitive<Integer, int[], IntConsumer>
implements IntConsumer {
// Android-changed: Made public for CTS tests only.
public OfInt() { }
// Android-changed: Made public for CTS tests only.
public OfInt(int initialCapacity) {
super(initialCapacity);
}
@Override
public void forEach(Consumer<? super Integer> consumer) {
if (consumer instanceof IntConsumer) {
forEach((IntConsumer) consumer);
}
else {
if (Tripwire.ENABLED)
Tripwire.trip(getClass(), "{0} calling SpinedBuffer.OfInt.forEach(Consumer)");
spliterator().forEachRemaining(consumer);
}
}
@Override
protected int[][] newArrayArray(int size) {
return new int[size][];
}
@Override
public int[] newArray(int size) {
return new int[size];
}
@Override
protected int arrayLength(int[] array) {
return array.length;
}
@Override
protected void arrayForEach(int[] array,
int from, int to,
IntConsumer consumer) {
for (int i = from; i < to; i++)
consumer.accept(array[i]);
}
@Override
public void accept(int i) {
preAccept();
curChunk[elementIndex++] = i;
}
public int get(long index) {
// Casts to int are safe since the spine array index is the index minus
// the prior element count from the current spine
int ch = chunkFor(index);
if (spineIndex == 0 && ch == 0)
return curChunk[(int) index];
else
return spine[ch][(int) (index - priorElementCount[ch])];
}
@Override
public PrimitiveIterator.OfInt iterator() {
return Spliterators.iterator(spliterator());
}
public Spliterator.OfInt spliterator() {
class Splitr extends BaseSpliterator<Spliterator.OfInt>
implements Spliterator.OfInt {
Splitr(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
super(firstSpineIndex, lastSpineIndex,
firstSpineElementIndex, lastSpineElementFence);
}
@Override
Splitr newSpliterator(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
return new Splitr(firstSpineIndex, lastSpineIndex,
firstSpineElementIndex, lastSpineElementFence);
}
@Override
void arrayForOne(int[] array, int index, IntConsumer consumer) {
consumer.accept(array[index]);
}
@Override
Spliterator.OfInt arraySpliterator(int[] array, int offset, int len) {
return Arrays.spliterator(array, offset, offset+len);
}
}
return new Splitr(0, spineIndex, 0, elementIndex);
}
@Override
public String toString() {
int[] array = asPrimitiveArray();
if (array.length < 200) {
return String.format("%s[length=%d, chunks=%d]%s",
getClass().getSimpleName(), array.length,
spineIndex, Arrays.toString(array));
}
else {
int[] array2 = Arrays.copyOf(array, 200);
return String.format("%s[length=%d, chunks=%d]%s...",
getClass().getSimpleName(), array.length,
spineIndex, Arrays.toString(array2));
}
}
}
/**
* An ordered collection of {@code long} values.
* @hide Visible for CTS testing only (OpenJDK8 tests).
*/
// Android-changed: Made public for CTS tests only.
public static class OfLong extends SpinedBuffer.OfPrimitive<Long, long[], LongConsumer>
implements LongConsumer {
// Android-changed: Made public for CTS tests only.
public OfLong() { }
// Android-changed: Made public for CTS tests only.
public OfLong(int initialCapacity) {
super(initialCapacity);
}
@Override
public void forEach(Consumer<? super Long> consumer) {
if (consumer instanceof LongConsumer) {
forEach((LongConsumer) consumer);
}
else {
if (Tripwire.ENABLED)
Tripwire.trip(getClass(), "{0} calling SpinedBuffer.OfLong.forEach(Consumer)");
spliterator().forEachRemaining(consumer);
}
}
@Override
protected long[][] newArrayArray(int size) {
return new long[size][];
}
@Override
public long[] newArray(int size) {
return new long[size];
}
@Override
protected int arrayLength(long[] array) {
return array.length;
}
@Override
protected void arrayForEach(long[] array,
int from, int to,
LongConsumer consumer) {
for (int i = from; i < to; i++)
consumer.accept(array[i]);
}
@Override
public void accept(long i) {
preAccept();
curChunk[elementIndex++] = i;
}
public long get(long index) {
// Casts to int are safe since the spine array index is the index minus
// the prior element count from the current spine
int ch = chunkFor(index);
if (spineIndex == 0 && ch == 0)
return curChunk[(int) index];
else
return spine[ch][(int) (index - priorElementCount[ch])];
}
@Override
public PrimitiveIterator.OfLong iterator() {
return Spliterators.iterator(spliterator());
}
public Spliterator.OfLong spliterator() {
class Splitr extends BaseSpliterator<Spliterator.OfLong>
implements Spliterator.OfLong {
Splitr(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
super(firstSpineIndex, lastSpineIndex,
firstSpineElementIndex, lastSpineElementFence);
}
@Override
Splitr newSpliterator(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
return new Splitr(firstSpineIndex, lastSpineIndex,
firstSpineElementIndex, lastSpineElementFence);
}
@Override
void arrayForOne(long[] array, int index, LongConsumer consumer) {
consumer.accept(array[index]);
}
@Override
Spliterator.OfLong arraySpliterator(long[] array, int offset, int len) {
return Arrays.spliterator(array, offset, offset+len);
}
}
return new Splitr(0, spineIndex, 0, elementIndex);
}
@Override
public String toString() {
long[] array = asPrimitiveArray();
if (array.length < 200) {
return String.format("%s[length=%d, chunks=%d]%s",
getClass().getSimpleName(), array.length,
spineIndex, Arrays.toString(array));
}
else {
long[] array2 = Arrays.copyOf(array, 200);
return String.format("%s[length=%d, chunks=%d]%s...",
getClass().getSimpleName(), array.length,
spineIndex, Arrays.toString(array2));
}
}
}
/**
* An ordered collection of {@code double} values.
* @hide Visible for CTS testing only (OpenJDK8 tests).
*/
// Android-changed: Made public for CTS tests only.
public static class OfDouble
extends SpinedBuffer.OfPrimitive<Double, double[], DoubleConsumer>
implements DoubleConsumer {
// Android-changed: Made public for CTS tests only.
public OfDouble() { }
// Android-changed: Made public for CTS tests only.
public OfDouble(int initialCapacity) {
super(initialCapacity);
}
@Override
public void forEach(Consumer<? super Double> consumer) {
if (consumer instanceof DoubleConsumer) {
forEach((DoubleConsumer) consumer);
}
else {
if (Tripwire.ENABLED)
Tripwire.trip(getClass(), "{0} calling SpinedBuffer.OfDouble.forEach(Consumer)");
spliterator().forEachRemaining(consumer);
}
}
@Override
protected double[][] newArrayArray(int size) {
return new double[size][];
}
@Override
public double[] newArray(int size) {
return new double[size];
}
@Override
protected int arrayLength(double[] array) {
return array.length;
}
@Override
protected void arrayForEach(double[] array,
int from, int to,
DoubleConsumer consumer) {
for (int i = from; i < to; i++)
consumer.accept(array[i]);
}
@Override
public void accept(double i) {
preAccept();
curChunk[elementIndex++] = i;
}
public double get(long index) {
// Casts to int are safe since the spine array index is the index minus
// the prior element count from the current spine
int ch = chunkFor(index);
if (spineIndex == 0 && ch == 0)
return curChunk[(int) index];
else
return spine[ch][(int) (index - priorElementCount[ch])];
}
@Override
public PrimitiveIterator.OfDouble iterator() {
return Spliterators.iterator(spliterator());
}
public Spliterator.OfDouble spliterator() {
class Splitr extends BaseSpliterator<Spliterator.OfDouble>
implements Spliterator.OfDouble {
Splitr(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
super(firstSpineIndex, lastSpineIndex,
firstSpineElementIndex, lastSpineElementFence);
}
@Override
Splitr newSpliterator(int firstSpineIndex, int lastSpineIndex,
int firstSpineElementIndex, int lastSpineElementFence) {
return new Splitr(firstSpineIndex, lastSpineIndex,
firstSpineElementIndex, lastSpineElementFence);
}
@Override
void arrayForOne(double[] array, int index, DoubleConsumer consumer) {
consumer.accept(array[index]);
}
@Override
Spliterator.OfDouble arraySpliterator(double[] array, int offset, int len) {
return Arrays.spliterator(array, offset, offset+len);
}
}
return new Splitr(0, spineIndex, 0, elementIndex);
}
@Override
public String toString() {
double[] array = asPrimitiveArray();
if (array.length < 200) {
return String.format("%s[length=%d, chunks=%d]%s",
getClass().getSimpleName(), array.length,
spineIndex, Arrays.toString(array));
}
else {
double[] array2 = Arrays.copyOf(array, 200);
return String.format("%s[length=%d, chunks=%d]%s...",
getClass().getSimpleName(), array.length,
spineIndex, Arrays.toString(array2));
}
}
}
}