Spliterator: 병렬 처리를 위한 분할

Spliterator란 무엇입니까

Spliterator의 핵심 메서드

Spliterator<Integer> sp = List.of(1,2,3,4,5,6).spliterator();
Spliterator<Integer> half = sp.trySplit(); // splits off first ~half
half.forEachRemaining(System.out::println); // 1 2 3
sp.forEachRemaining(System.out::println);   // 4 5 6

Spliterator의 특성

Spliterator<String> sp = List.of("a","b","c").spliterator();
System.out.println(Integer.toBinaryString(sp.characteristics()));
// Characteristics include ORDERED, SIZED, SUBSIZED

trySplit의 작동 방식

// ArrayList Spliterator splits efficiently at midpoint:
// [0,1,2,3,4,5] -> [0,1,2] (new) + [3,4,5] (original)

사용자 지정 Spliterator 만들기

public class RangeSpliterator implements Spliterator<Integer> {
    private int start, end;
    public RangeSpliterator(int start, int end) { this.start=start; this.end=end; }
    public boolean tryAdvance(Consumer<? super Integer> action) {
        if (start >= end) return false;
        action.accept(start++); return true;
    }
    public Spliterator<Integer> trySplit() {
        int mid = (start + end) / 2;
        if (mid <= start) return null;
        RangeSpliterator prefix = new RangeSpliterator(start, mid);
        this.start = mid; return prefix;
    }
    public long estimateSize() { return end - start; }
    public int characteristics() { return ORDERED | SIZED | SUBSIZED | IMMUTABLE; }
}

Spliterator에서 스트림 만들기

Spliterator<Integer> sp = new RangeSpliterator(0, 1_000_000);
Stream<Integer> stream = StreamSupport.stream(sp, true); // true = parallel
long count = stream.filter(n -> n % 2 == 0).count();
System.out.println(count); // 500000

이진 트리를 위한 Spliterator

병렬 처리 효율과 분할 품질

대량 처리를 위한 forEachRemaining

@Override
public void forEachRemaining(Consumer<? super Integer> action) {
    for (int i = start; i < end; i++) action.accept(i);
    start = end; // mark as exhausted
}

Spliterator와 Iterator 비교

확인 문제

복습