Iterator and Spliterator
2019-02-02 本文已影响0人
gattonero
Iterator
Java中有三种Iterator,Enumeration,Iterator,ListIterator
- Enumeration ,一种用于遍历旧式数据结构的Iterator,比如Vector和Hashtable(一种同步哈希表),也可以用于SequenceInputStream(用于合并输入流,顺序读取)
//add 10 elements for test
Vector v = new Vector();
for (int i = 0; i < 10; i++) v.addElement(i);
Enumeration e = v.elements();
// Checking the next element availability
while (e.hasMoreElements())
{
// moving cursor to next element
int i = (Integer)e.nextElement();
System.out.print(i + " ");
}
- Iterator 通用的Iterator,可用于任何Collection对象,相比于Enumeration,在遍历时可进行读取和删除操作,下面列出了Iterator的基本方法
Iterator iterator = collection.iterator();
iterator.hasNext();//boolean
iterator.next();//Object
iterator.remove()//void
缺陷:不能新增或者替换元素,而且只能单向遍历
- ListIterator
只适用于List集合,比如ArrayList,LinkedList,不过该接口提供双向遍历:
listIterator.hasPrevious();//boolean
listIterator.previous();//Object
listIterator.previousIndex();//int
listIterator.set();//void replace the last element find by next() or previous()
listIterator.add()//void add element before the element which will be returned by next()
ListIterator没有当前元素,当前指针的位置始终在next()和previous()两个元素之间
Spliterators
Spliterators 和其他的 Iterators一样,都是遍历数据源的手段,包括Collections,IO Channel,Generator function
它除了支持顺序遍历之外,添加了Java 8对于并行的支持
每一个Spliterator的特性都不同,可能的值有
ORDERED(0x00000010), DISTINCT(0x00000001), SORTED(0x00000004), SIZED(0x00000040), NONNULL(0x00000100), IMMUTABLE(0x00000400), CONCURRENT(0x00001000), SUBSIZED(0x00004000).
针对不同的特性,characteristics()返回位编码下的特性int值,hasCharacteristics检测是否拥有该特性
//simple bitwise operation
default boolean hasCharacteristics(int characteristics) {
return (characteristics() & characteristics) == characteristics;
}
其中tryAdvance() 方法将next()和hasNext()结合起来,并且可以通过forEachRemaining(Consumer),简单的对Iterator的剩余元素进行遍历
除此之外,还可以使用estimateSize和getExactSizeIfKnown来估计遍历对象的大小
// Create an array list for doubles.
ArrayList<Integer> al = new ArrayList<>();
// Add values to the array list.
al.add(1);
al.add(2);
al.add(-3);
al.add(-3);
al.add(-4);
al.add(5);
Stream<Integer> str1 = al.stream();
Spliterator<Integer> splitr1 = str1.spliterator();
System.out.println(splitr1.characteristics());
//结合hasNext和next方法
splitr1.tryAdvance((a) -> System.out.println(a+"!"));
// trySplit() 将剩余元素分成两部分,取出第一部分
Spliterator<Integer> splitr2 = splitr1.trySplit();
splitr2.forEachRemaining((a) -> System.out.println(a));
splitr1.forEachRemaining((a) -> System.out.println(a));
结果
16464
1!
2
-3
-3
-4
5
Tips
- 16464 = 0x4050 = 0x4000+0x0010+0x0040=SUBSIZED+SIZED+ORDERED
- Even an inexact estimate is often useful and inexpensive to compute. 估计遍历器大小的操作并不昂贵
- The characteristics of a given spliterator before splitting may differ from the characteristics after splitting,For specific examples see the characteristic values SIZED, SUBSIZED and CONCURRENT. spliterator的characteristics在split前后可能不同,详见上述几个特性
- 任何Iterator的引用在初始化时,下标都会指向首元素之前
- 上述的三个Iterator只是不同的接口,需要在具体的类中,通过诸如elements(), iterator(), listIterator()等方法实现,一般是通过匿名内部类来实现的
System.out.println(new java.util.Vector().elements().getClass().getName());//java.util.Vector$1
Ref
https://www.geeksforgeeks.org/java-util-interface-spliterator-java8/
https://blog.csdn.net/java4found/article/details/8656594
https://www.geeksforgeeks.org/iterators-in-java/