On Feb 12, 2015, at 5:07 PM, Paul Sandoz <paul.san...@oracle.com> wrote:
> On Feb 12, 2015, at 4:51 PM, Peter Levart <peter.lev...@gmail.com> wrote:
>> Hi Paul,
>>
>> Would the following "optimization" make any sense?
>>
>> public static <T, U, A, R>
>> Collector<T, ?, R> flatMapping(Function<? super T, ? extends Stream<?
>> extends U>> mapper,
>> Collector<? super U, A, R> downstream) {
>> BiConsumer<A, ? super U> downstreamAccumulator =
>> downstream.accumulator();
>> return new CollectorImpl<>(downstream.supplier(),
>> (r, t) -> {
>> try (Stream<? extends U> result = mapper.apply(t)) {
>> if (result != null) {
>> ((result.isParallel() &&
>>
>> !downstream.characteristics().contains(Collector.Characteristics.CONCURRENT))
>> ? result.sequential() : result).forEach(u ->
>> downstreamAccumulator.accept(r, u));
>> }
>> }
>> },
>> downstream.combiner(), downstream.finisher(),
>> downstream.characteristics());
>> }
>>
>
> Ah, interesting.
>
> Possibly... i need to think thought the implications on stream and collectors
> (groupingByConcurrent does the right thing for a non-concurrent downstream
> collector and synchronizes on the accumulation).
>
Unfortunately there is not currently enough information propagated down from
the top-level collect to make an informed decision about whether one could go
parallel and collect concurrently within flatMapping.
Paul.
