Can anybody explain flatMapMerge and flatMapConcat...

# getting-started

[sarmad sohaib]

Can anybody explain flatMapMerge and flatMapConcat with some real world example? I have read the docs and seen videos but they are confusing.

Thanks a lot.

[Casey Brooks]

The biggest difference is that flatMapConcat collects the entirety of each Flow returned by the transform lambda, before starting to collect the next flow. It processes each flow in series, and thus one flow that emits slowly or runs infiniately will prevent the later ones from running.

flow<Flow<String>> {
    emit(flowOf("one", "two", "three"))
    emit(flow {
        // this is an infinite flow
        var i = 0
        while (true) {
            i++
            emit(i.toString())
        }
    })
    emit(flowOf("ONE", "TWO", "THREE")) // will never start collecting this flow
}.flatMapConcat { it }
// result: ["one", "two", "three", "1", "2", "3", "4"...]

flatMapMerge collects each flow in parallel. So slow/infinite emitters are not a problem

flow<Flow<String>> {
    emit(flowOf("one", "two", "three"))
    emit(flow {
        var i = 0
        while (true) {
            i++
            emit(i.toString())
        }
    })
    emit(flowOf("five", "six", "seven"))
}.flatMapMerge { it }
// result: ["one", "1", "ONE", "two", "2", "TWO", "three", "3", "THREE", "4", "5"...]

[sarmad sohaib]

Thanks. Can you provide an easy to read and understand example?

[Casey Brooks]

What’s confusing about the examples posted above? I’m not sure how to get any simpler than that.

You have a Flow that emits more Flows, then uses the flatMapConcat or flatMapMerge operators on the result. For the first snippet, using flatMapConcat, since the second Flow is infinite, it will never be able to start processing the third. But using flatMapMerge instead will run them all 3 Flows in parallel, merging each value into the resulting list.