On Fri, 23 May 2025 15:53:22 GMT, Johannes Döbler <d...@openjdk.org> wrote:
>> While it could be argued that unbounded Spliterators should not report SIZED
>> / SUBSIZED, GatherSink should report an unknown emission size, so switching
>> to downstream.begin(-1) rather than downstream.begin(size).
>>
>> Includes a regression test which yields an OOME if this change is omitted.
>
> src/java.base/share/classes/java/util/stream/GathererOp.java line 153:
>
>> 151: if (initializer != Gatherer.defaultInitializer()) //
>> Optimization
>> 152: state = initializer.get();
>> 153: sink.begin(-1); // GathererOp does not know the size of the
>> output
>
> The Integrator of the SO question which motivated this PR _does_ know the
> size of the output. In general would it be worth to enhance Integrator with
> something like `default long estimateSize(long streamEstimate) { return -1;
> }` to allow Integrator implementations to weigh in?
@jdlib I guess that would depend on the definition of *value* and *cost* in
this context.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/25418#discussion_r2104911643
