Reactive Programming with Reactor 3


Open Source Your Knowledge, Become a Contributor

Technology knowledge has to be shared and made accessible for free. Join the movement.

Create Content



Reactor ships with several operators that can be used to transform data.


In the first place, we will capitalize a String. Since this is a simple 1-1 transformation with no expected latency, we can use the map operator with a lambda transforming a T into a U.

Capitalize data on Mono

We can use exactly the same code on a Flux, applying the mapping to each element as it becomes available.

Capitalize data on Flux

Now imagine that we have to call a webservice to capitalize our String. This new call can have latency so we cannot use the synchronous map anymore. Instead, we want to represent the asynchronous call as a Flux or Mono, and use a different operator: flatMap.

flatMap takes a transformation Function that returns a Publisher<U> instead of a U. This publisher represents the asynchronous transformation to apply to each element. If we were using it with map, we'd obtain a stream of Flux<Publisher<U>>. Not very useful.

But flatMap on the other hand knows how to deal with these inner publishers: it will subscribe to them then merge all of them into a single global output, a much more useful Flux<U>. Note that if values from inner publishers arrive at different times, they can interleave in the resulting Flux.

Async transformation
Open Source Your Knowledge: become a Contributor and help others learn. Create New Content
package io.pivotal.literx;