对null值友好的Collectors.groupingBy

一个人行走的范围，就是他的世界。——北岛

我们在使用Collectors.groupingBy时会遇到这种情况：

Map<String, List<User>> map = Arrays.asList(new User(), null).stream().collect(Collectors.groupingBy(User::getName));

为了避免这种情况，于是我自己实现了一个：

@SafeVarargs
@SuppressWarnings("unchecked")
public static <T, K, D, A, M extends Map<K, D>> M listGroupBy(List<T> list, Function<T, K> sFunction, Collector<? super T, A, D> downstream, boolean isParallel, Consumer<T>... peeks) {
    boolean hasFinished = downstream.characteristics().contains(Collector.Characteristics.IDENTITY_FINISH);
    return peekStream(list, isParallel, peeks).collect(new Collector<T, HashMap<K, A>, M>() {
        @Override
        public Supplier<HashMap<K, A>> supplier() {
            return HashMap::new;
        }

        @Override
        public BiConsumer<HashMap<K, A>, T> accumulator() {
            return (m, t) -> {
                K key = Optional.ofNullable(t).map(sFunction).orElse(null);
                A container = m.computeIfAbsent(key, k -> downstream.supplier().get());
                downstream.accumulator().accept(container, t);
            };
        }

        @Override
        public BinaryOperator<HashMap<K, A>> combiner() {
            return (m1, m2) -> {
                for (Map.Entry<K, A> e : m2.entrySet()) {
                    m1.merge(e.getKey(), e.getValue(), downstream.combiner());
                }
                return m1;
            };
        }

        @Override
        public Function<HashMap<K, A>, M> finisher() {
            return hasFinished ? i -> (M) i : intermediate -> {
                // a-> a[0]
                intermediate.replaceAll((k, v) -> (A) downstream.finisher().apply(v));
                @SuppressWarnings("unchecked")
                M castResult = (M) intermediate;
                return castResult;
            };
        }

        @Override
        public Set<Characteristics> characteristics() {
            return hasFinished ? Collections.unmodifiableSet(EnumSet.of(Collector.Characteristics.IDENTITY_FINISH)) : Collections.emptySet();
        }
    });
}

使用方式：

Map<String, List<User>> map = listGroupBy(Arrays.asList(new User(), null), User::getName, Collectors.toList(), false)

这样避免了抛出异常，返回了对null值友好的结果(map里包含一个key为null的结果)

我稍作修改放到MP的SimpleQuery和hutool中的CollStreamUtil以及CollectorUtil中去了