美文网首页Java
Spliterator并行遍历迭代器(JDK8)

Spliterator并行遍历迭代器(JDK8)

作者: maxzhao_ | 来源:发表于2018-12-11 10:45 被阅读26次

    文章前半部分转自:https://blog.csdn.net/lh513828570/article/details/56673804

    Spliterator是什么?

    public interface Spliterator<T> 
    

    Spliterator是一个可分割迭代器(splitable iterator),可以和iterator顺序遍历迭代器一起看。jdk1.8发布后,对于并行处理的能力大大增强,Spliterator就是为了并行遍历元素而设计的一个迭代器,jdk1.8中的集合框架中的数据结构都默认实现了spliterator,后面我们也会结合ArrayList中的spliterator()一起解析。

     //单个对元素执行给定的动作,如果有剩下元素未处理返回true,否则返回false
     boolean tryAdvance(Consumer<? super T> action);
    
     //对每个剩余元素执行给定的动作,依次处理,直到所有元素已被处理或被异常终止。默认方法调用tryAdvance方法
     default void forEachRemaining(Consumer<? super T> action) {
        do { } while (tryAdvance(action));
     }
    
     //对任务分割,返回一个新的Spliterator迭代器
     Spliterator<T> trySplit();
    
     //用于估算还剩下多少个元素需要遍历
     long estimateSize();
    
     //当迭代器拥有SIZED特征时,返回剩余元素个数;否则返回-1
     default long getExactSizeIfKnown() {
        return (characteristics() & SIZED) == 0 ? -1L : estimateSize();
     }
    
      //返回当前对象有哪些特征值
     int characteristics();
    
     //是否具有当前特征值
     default boolean hasCharacteristics(int characteristics) {
        return (characteristics() & characteristics) == characteristics;
     }
     //如果Spliterator的list是通过Comparator排序的,则返回Comparator
     //如果Spliterator的list是自然排序的 ,则返回null
     //其他情况下抛错
     default Comparator<? super T> getComparator() {
         throw new IllegalStateException();
     }
    
    

    特征值其实就是为表示该Spliterator有哪些特性,用于可以更好控制和优化Spliterator的使用。关于获取比较器getComparator这一个方法,目前我还没看到具体使用的地方,所以可能理解有些误差。(源玛里有这里就不展示了)

    ArrayList的例子

    ArrayListSpliterator在ArrayList的源码里

    static final class ArrayListSpliterator<E> implements Spliterator<E> {
        //用于存放ArrayList对象
       private final ArrayList<E> list;
       //起始位置(包含),advance/split操作时会修改
       private int index; 
       //结束位置(不包含),-1 表示到最后一个元素
       private int fence; 
       //用于存放list的modCount
       private int expectedModCount; 
    
       ArrayListSpliterator(ArrayList<E> list, int origin, int fence,
                                 int expectedModCount) {
                this.list = list; 
                this.index = origin;
                this.fence = fence;
                this.expectedModCount = expectedModCount;
            }
        //获取结束位置(存在意义:首次初始化石需对fence和expectedModCount进行赋值)
       private int getFence() { 
            int hi; 
            ArrayList<E> lst;
            //fence<0时(第一次初始化时,fence才会小于0):
            if ((hi = fence) < 0) {
                //list 为 null时,fence=0
                if ((lst = list) == null)
                    hi = fence = 0;
                else {
                //否则,fence = list的长度。
                    expectedModCount = lst.modCount;
                    hi = fence = lst.size;
                }
            }
            return hi;
        }
        //分割list,返回一个新分割出的spliterator实例
        public ArrayListSpliterator<E> trySplit() {
            //hi为当前的结束位置
            //lo 为起始位置
            //计算中间的位置
            int hi = getFence(), lo = index, mid = (lo + hi) >>> 1;
            //当lo>=mid,表示不能在分割,返回null
            //当lo<mid时,可分割,切割(lo,mid)出去,同时更新index=mid
            return (lo >= mid) ? null : 
                new ArrayListSpliterator<E>(list, lo, index = mid,                                         expectedModCount);
        }
        //返回true 时,只表示可能还有元素未处理
        //返回false 时,没有剩余元素处理了。。。
        public boolean tryAdvance(Consumer<? super E> action) {
             if (action == null)
                 throw new NullPointerException();
             //hi为当前的结束位置
             //i 为起始位置
             int hi = getFence(), i = index;
             //还有剩余元素未处理时
             if (i < hi) {
                 //处理i位置,index+1
                 index = i + 1;
                 @SuppressWarnings("unchecked") E e = (E)list.elementData[i];
                 action.accept(e);
                 //遍历时,结构发生变更,抛错
                 if (list.modCount != expectedModCount)
                     throw new ConcurrentModificationException();
                 return true;
             }
             return false;
         }
        //顺序遍历处理所有剩下的元素
       public void forEachRemaining(Consumer<? super E> action) {
           int i, hi, mc; // hoist accesses and checks from loop
           ArrayList<E> lst; Object[] a;
           if (action == null)
               throw new NullPointerException();
           if ((lst = list) != null && (a = lst.elementData) != null) {
               //当fence<0时,表示fence和expectedModCount未初始化,可以思考一下这里能否直接调用getFence(),嘿嘿?
               if ((hi = fence) < 0) {
                   mc = lst.modCount;
                   hi = lst.size;
               }
               else
                   mc = expectedModCount;
               if ((i = index) >= 0 && (index = hi) <= a.length) {
                   for (; i < hi; ++i) {
                       @SuppressWarnings("unchecked") E e = (E) a[i];
                       //调用action.accept处理元素
                       action.accept(e);
                   }
                   //遍历时发生结构变更时抛出异常
                   if (lst.modCount == mc)
                       return;
               }
           }
           throw new ConcurrentModificationException();
       }
    
       public long estimateSize() {
            return (long) (getFence() - index);
        }
    
        public int characteristics() {
            //打上特征值:、可以返回size
            return Spliterator.ORDERED | Spliterator.SIZED | Spliterator.SUBSIZED;
        }
    }
    
    测试代码如下:
     List<String> arrs = new ArrayList<>();
            arrs.add("a");
            arrs.add("b");
            arrs.add("c");
            arrs.add("d");
            arrs.add("e");
            arrs.add("f");
            arrs.add("h");
            arrs.add("i");
            arrs.add("j");
            Spliterator<String> a = arrs.spliterator();
            System.out.println(a);
            //此时结果:a:0-9(index-fence)
    
            Spliterator<String> b = a.trySplit();
            System.out.println(b.toString());
            //此时结果:b:4-9,a:0-4
    
            Spliterator<String> c = a.trySplit();
            //此时结果:c:4-6,b:4-9,a:6-9
            System.out.println(c.toString());
    
            Spliterator<String> d = a.trySplit();
            System.out.println(d.toString());
            //此时结果:d:6-7,c:4-6,b:4-9,a:7-9
    

    可以看到每次分割,都会分割剩余的前一半,fence之不变,index后移。同时也发现:
    1.ArrayListSpliterator本质上还是对原list进行操作,只是通过index和fence来控制每次处理范围
    2.也可以得出,ArrayListSpliterator在遍历元素时,不能对list进行结构变更操作,否则抛错。

    衍生接口OfPrimitive

    源码:

    public interface OfPrimitive<T, T_CONS, T_SPLITR extends Spliterator.OfPrimitive<T, T_CONS, T_SPLITR>>
                extends Spliterator<T> {
            @Override
            T_SPLITR trySplit();
            @SuppressWarnings("overloads")
            boolean tryAdvance(T_CONS action);
            @SuppressWarnings("overloads")
            default void forEachRemaining(T_CONS action) {
                do { } while (tryAdvance(action));
            }
        }
    
     public interface OfPrimitive<T, T_CONS, T_SPLITR extends Spliterator.OfPrimitive<T, T_CONS, T_SPLITR>> extends Spliterator<T> {
            T_SPLITR trySplit();
            boolean tryAdvance(T_CONS var1);
            default void forEachRemaining(T_CONS var1) {
                while(this.tryAdvance(var1)) {
                    ;
                }
    
            }
        }
    

    可以看到,这个接口基本没有变动,这是多增加两个泛型声明而已,本质上和Spliterator没有太大的区别,只不过,它限制tryAdvance的参数action类型T_CONS和trySplit的返回参数T_SPLITR必须在实现接口时先声明类型。
    基于OfPrimitive接口,又衍生出了OfInt、OfLong、OfDouble等专门用来处理int、Long、double等分割迭代器接口(在Spliterators有具体的实现)。
    LinkedHashSet中的Spliterator方法的实现,内部使用到了Sqlieterator接口的常量值!

    简单的并发测试(jdk1.8+)

        private AtomicInteger count = new AtomicInteger(0);
        private List<String> strList = createList();
        private Spliterator spliterator = strList.spliterator();
    
        /**
         * 多线程计算list中数值的和
         * 测试spliterator遍历
         */
        @Test
        public void mytest() {
            for (int i = 0; i < 4; i++) {
                new Thread(() -> {
                    String threadName = Thread.currentThread().getName();
                    System.out.println("   " + threadName + " start ");
                    spliterator.trySplit().forEachRemaining((o) -> {
                        if (isInteger((String) o)) {
                            int num = Integer.parseInt(o + "");
                            count.addAndGet(num);
                            System.out.println("数值:" + num + "     " + threadName);
                            try {
                                Thread.sleep(2000);
                            } catch (InterruptedException e) {
                                e.printStackTrace();
                            }
                        }
                    });
                    System.out.println("     " + threadName + " end");
                }).start();
            }
            try {
                Thread.sleep(15000);
            } catch (InterruptedException e) {
                e.printStackTrace();
            }
            System.out.println("结果为:" + count);
        }
    
        private List<String> createList() {
            List<String> result = new ArrayList<>();
            for (int i = 0; i < 100; i++) {
                if (i % 10 == 0) {
                    result.add(i + "");
                } else {
                    result.add("=");
                }
            }
            return result;
        }
    
        public static boolean isInteger(String str) {
            Pattern pattern = Pattern.compile("^[-\\+]?[\\d]*$");
            return pattern.matcher(str).matches();
        }
    

    相关文章

      网友评论

        本文标题:Spliterator并行遍历迭代器(JDK8)

        本文链接:https://www.haomeiwen.com/subject/hmfshqtx.html