Project Reactor expand method

One of my colleagues at work recently introduced me to the expand operator of the Project Reactor types and in this post I want to cover a few ways in which I have used it.

Unrolling a Paginated Result

Consider a Spring Data based repository on a model called City:

import org.springframework.data.jpa.repository.JpaRepository;
import samples.geo.domain.City;

public interface CityRepo extends JpaRepository<City, Long> {
}

This repository provides a way to retrieve the paginated result, along the following lines:

cityRepo.findAll(PageRequest.of(0, 5))

Now, if I were to unroll multiple pages into a result, the way to do it would be the following kind of a loop:

var pageable: Pageable = PageRequest.of(0, 5)
do {
    var page: Page<City> = cityRepo.findAll(pageable)
    page.content.forEach { city -> LOGGER.info("City $city") }
    pageable = page.nextPageable()
} while (page.hasNext())

An equivalent unroll of a paginated result can be done using the Reactor expand operator the following way:

val result: Flux<City> =
    Mono
        .fromSupplier { cityRepo.findAll(PageRequest.of(0, 5)) }
        .expand { page ->
            if (page.hasNext())
                Mono.fromSupplier { cityRepo.findAll(page.nextPageable()) }
            else
                Mono.empty()
        }
        .flatMap { page -> Flux.fromIterable(page.content) }

result.subscribe(
    { page -> LOGGER.info("City ${page}") },
    { t -> t.printStackTrace() }
)

Here the first page of results expands to the second page, the second page to the third page and so on until there are no pages to retrieve.

Traversing a Tree

Consider a node in a tree structure represented by the following model:

data class Node(
    val id: String,
    val nodeRefs: List<String>,
)

A sample data which looks like this:

can be traversed using a call which looks like this:

val rootMono: Mono<Node> = nodeService.getNode("1")
val expanded: Flux<Node> = rootMono.expand { node ->
    Flux.fromIterable(node.childRefs)
        .flatMap { nodeRef -> nodeService.getNode(nodeRef) }
}
expanded.subscribe { node -> println(node) }

This is a breadth-first expansion, the output looks like this:

Node-1
Node-1-1
Node-1-2
Node-1-1-1
Node-1-1-2
Node-1-2-1
Node-1-2-2

An expandDeep variation would traverse it depth-first