Reactive Streams is a standard for asynchronous data processing in a streaming fashion with non-blocking backpressure. Starting from Java 9, they have become a part of the JDK in the form of the java.util.concurrent.Flow interfaces. In this live coding session we’re going to go through the basic concepts of reactive stream processing and see how to use the APIs included in JDK 9.

1. OUT-OF-THE-BOX
REACTIVE STREAMS
WITH JAVA 9
JACEK KUNICKI
@rucek
https://github.com/rucek/reactive-streams-java9

2. @rucekOOTB REACTIVE STREAMS WITH JAVA 9
STREAM PROCESSING
PRODUCER CONSUMERPROCESSING
STAGE 1
PROCESSING
STAGE N
DATA
SubscriberPublisher
BACKPRESSURE
request(n)
Processor Processor

3. java.util.concurrent.Flow

4. @rucek
j.u.c.Flow.Publisher<T>
▸ produces items of type T that subscribers are going to consume
▸ multiple subscribers receive items in the same order
▸ subscribers register via subscribe(Subscriber<? super T>)
OOTB REACTIVE STREAMS WITH JAVA 9

5. @rucek
j.u.c.Flow.Subscriber<T>
▸ subscribes to a producer in order to receive:
▸ subscription confirmation via onSubscribe(Subscription)
▸ items via onNext(T)
▸ errors via onError(Throwable)
▸ completion signal via onComplete()
OOTB REACTIVE STREAMS WITH JAVA 9

6. @rucek
j.u.c.Flow.Subscription
▸ connects a single producer to a single subscriber, allows to:
▸ backpressure with request(long)
▸ signal (eventual) termination with cancel()
OOTB REACTIVE STREAMS WITH JAVA 9

7. @rucek
j.u.c.Flow.Processor<T, R>
▸ a combination of a Subscriber<T> and a Publisher<R>
OOTB REACTIVE STREAMS WITH JAVA 9

8. @rucek
j.u.c.SubmissionPublisher<T>
▸ the only concrete implementation available so far
▸ asynchronously issues submitted (non-null) items to current subscribers
▸ can be used as a base for your own components
OOTB REACTIVE STREAMS WITH JAVA 9

9. onSubscribe
onNext*
(onComplete | onError)?
Publisher.subscribe(Subscriber)

10. @rucekOOTB REACTIVE STREAMS WITH JAVA 9

11. @rucek
DEMO 1 - PURE JAVA 9
▸ number publisher based on an IntStream
▸ filtering and mapping processors based on the SubmissionPublisher
▸ a subscriber that prints to System.out
OOTB REACTIVE STREAMS WITH JAVA 9

12. @rucek
DEMO 2 - INTEGRATION
▸ Project Reactor’s Flux as a publisher
▸ Akka Streams Flow as a processor
▸ a pure Java 9 processor
▸ a pure Java 9 subscriber
OOTB REACTIVE STREAMS WITH JAVA 9

13. @rucek
SUMMARY
▸ not a full Reactive Streams implementation
▸ allows for interoperability between other implementations
▸ incubating Reactive Streams support in the new HTTP client
OOTB REACTIVE STREAMS WITH JAVA 9

14. @rucek
THANK YOU!
@rucek
OOTB REACTIVE STREAMS WITH JAVA 9
https://github.com/rucek/reactive-streams-java9