java8新特性-引用流-reduce

reduce操作用于对数据进行聚合，比如求和等。

一、reduce(BinaryOperator accumulator)

例子：

List<User> users = new ArrayList<>();
users.add(new User("张三",30));
users.add(new User("李四",39));
users.add(new User("王五",20));

Optional<Integer> reduce = users.stream().map(User::getAge).reduce((a, b) -> a + b);
System.out.println(reduce.get());

输出：

源码分析：
ReferencePipeline#reduce(BinaryOperator<P_OUT> accumulator)

 public final Optional<P_OUT> reduce(BinaryOperator<P_OUT> accumulator) {
    return evaluate(ReduceOps.makeRef(accumulator));
}

进入ReduceOps.makeRef(BinaryOperator operator)方法。

ReduceOps#makeRef(BinaryOperator operator)

public static <T> TerminalOp<T, Optional<T>>
makeRef(BinaryOperator<T> operator) {
    Objects.requireNonNull(operator);
    class ReducingSink
            implements AccumulatingSink<T, Optional<T>, ReducingSink> {
        private boolean empty;
        private T state;

        public void begin(long size) {
            empty = true;
            state = null;
        }

        @Override
        public void accept(T t) {
            if (empty) {
                empty = false;
                state = t;
            } else {
                state = operator.apply(state, t);
            }
        }

        @Override
        public Optional<T> get() {
            return empty ? Optional.empty() : Optional.of(state);
        }

        @Override
        public void combine(ReducingSink other) {
            if (!other.empty)
                accept(other.state);
        }
    }
    return new ReduceOp<T, Optional<T>, ReducingSink>(StreamShape.REFERENCE) {
        @Override
        public ReducingSink makeSink() {
            return new ReducingSink();
        }
    };
}

ReduceOps会调用makeSink获取Sink。begin方法设置empty和state的初始值。看看ReducingSink里的accept方法。第一次进入，empty为true，将state设置t，以后将t和state运行BinaryOperator聚合数据且将结果赋给state。如果Stream中没有元素则empty为true，state为null。

二、 reduce(T identity, BinaryOperator accumulator)

identity是初始值，功能和reduce(BinaryOperator accumulator )类似。

例子：

   List<User> users = new ArrayList<>();
   users.add(new User("张三",30));
   users.add(new User("李四",39));
   users.add(new User("王五",20));

   Integer reduce = users.stream().map(User::getAge).reduce(0,(a, b) -> a + b);
   System.out.println(reduce);

输出：

源码分析：
ReferencePipeline#reduce(final P_OUT identity, final BinaryOperator<P_OUT> accumulator)

 public final P_OUT reduce(final P_OUT identity, final BinaryOperator<P_OUT> accumulator) {
    return evaluate(ReduceOps.makeRef(identity, accumulator, accumulator));
}

进入ReduceOps.makeRef(U seed, BiFunction<U, ? super T, U> reducer, BinaryOperator<U> combiner )方法

  makeRef(U seed, BiFunction<U, ? super T, U> reducer, BinaryOperator<U> combiner) {
      Objects.requireNonNull(reducer);
      Objects.requireNonNull(combiner);
      class ReducingSink extends Box<U> implements AccumulatingSink<T, U, ReducingSink> {
          @Override
          public void begin(long size) {
              state = seed;
          }

          @Override
          public void accept(T t) {
              state = reducer.apply(state, t);
          }

          @Override
          public void combine(ReducingSink other) {
              state = combiner.apply(state, other.state);
          }
      }
      return new ReduceOp<T, U, ReducingSink>(StreamShape.REFERENCE) {
          @Override
          public ReducingSink makeSink() {
              return new ReducingSink();
          }
      };
  }

ReduceOps会调用makeSink获取Sink。ReducingSink的begin将state设置为seed，seed就是传进去的初始值。每次调用accept都会运行BiFunction将state 和 t执行聚合。

三、reduce(U identity, BiFunction<U, ? super T, U> accumulator, BinaryOperator<U> combiner)
第一个参数是返回值的初始值，accumulator是对Stream中每个元素执行的聚合操作，combiner是在并行流中对每个线程的执行结果进行合并。

例子

 List<User> users = new ArrayList<>();
 users.add(new User("张三",30));
 users.add(new User("李四",34));
 users.add(new User("王五",20));

ArrayList<User> reduce1 = users.stream().reduce(new ArrayList<User>(), (list, user) -> {
        if (user.getAge() >= 30){
            list.add(user);
        }
        System.out.println("执行accumulator");
        return list;
    }, ((users1, users2) -> {
        System.out.println("执行combiner");
        users1.addAll(users2);
        return users1;
    }));
    System.out.println(reduce1);

输出：

源码分析：
ReferencePipeline#reduce(R identity, BiFunction<R, ? super P_OUT, R> accumulator, BinaryOperator combiner)

public final <R> R reduce(R identity, BiFunction<R, ? super P_OUT, R> accumulator, BinaryOperator<R> combiner) {
    return evaluate(ReduceOps.makeRef(identity, accumulator, combiner));
}

进入ReduceOps.makeRef。

ReduceOps#makeRef

    public static <T, U> TerminalOp<T, U>
makeRef(U seed, BiFunction<U, ? super T, U> reducer, BinaryOperator<U> combiner) {
    Objects.requireNonNull(reducer);
    Objects.requireNonNull(combiner);
    class ReducingSink extends Box<U> implements AccumulatingSink<T, U, ReducingSink> {
        @Override
        public void begin(long size) {
            state = seed;
        }

        @Override
        public void accept(T t) {
            state = reducer.apply(state, t);
        }

        @Override
        public void combine(ReducingSink other) {
            state = combiner.apply(state, other.state);
        }
    }
    return new ReduceOp<T, U, ReducingSink>(StreamShape.REFERENCE) {
        @Override
        public ReducingSink makeSink() {
            return new ReducingSink();
        }
    };
}