flow sharedFlow in kotlin complete tutorial

1. What is sharedFlow

SharedFlowIt's in the Kotlin coroutineFlowA kind ofHot Flow, used to be between multiple subscribersShared events or data streams. It's suitable for handlingOne-time event(such as navigation, pop-up windows, Toast, refresh notifications, etc.), rather than persistent state.

✅ What is SharedFlow?

SharedFlowyesFlowAn extension of the company has the following characteristics:

It's essentially aEvent broadcaster”。

✅ Common usage scenarios

📍 1. One-time UI event

• Toast pop-up window
• SnackBar Tips
• Navigation Jump
• Close the page
• Dialog display/cancel

These events are "one-time" and do not need to be saved and should not be triggered repeatedly, so they are suitable forSharedFlow。

sealed class UiEvent {
    data class ShowToast(val message: String) : UiEvent()
    object NavigateToHome : UiEvent()
}

// ViewModelprivate val _uiEvent = MutableSharedFlow<UiEvent>()
val uiEvent = _uiEvent.asSharedFlow()
fun loginSuccess() {
     {
        _uiEvent.emit()
    }
}

// Fragment {
     { event ->
        when (event) {
            is  -> showToast()
            is  -> navigate()
        }
    }
}

📍 2. Streaming notification

• Notify other modules to refresh data
• Updated page for data pull completion notification
• ViewModel sends a signal to the UI

📍 3. Replace LiveData<Event> to solve the problem of repeated consumption

Traditional useLiveData<Event<T>>orSingleLiveEventHandle one-time events, the code is complex and inelegant, andSharedFlowIt is an officially recommended alternative.

✅ Typical usage of SharedFlow in company projects

Example: Jump + pop up after login

// ViewModel
val eventFlow = MutableSharedFlow<UiEvent>()
suspend fun login(username: String, pwd: String) {
    if (doLogin(username, pwd)) {
        (("Login successfully"))
        ()
    }
}

✅ Comparison with other Flow types

✅ In a summary sentence:

SharedFlow= KotlinHandle one-time eventsRecommended tools, suitable forViewModel → UI layer passes toast, navigation, pop-ups and other short-term behaviors,CompareLiveDataMore modern and controllable.

2. How to deal with backpressure in sharedFlow?

✅ 1. How does SharedFlow handle backpressure?

SharedFlowyesHot stream, meaning that the data will be sent immediately, not likeFlowWait for subscribers to collect that way. This means:

• If the launch is too fast (such as multiple consecutive emits)
• And subscribers haven't had time to collect
• Data may be discarded or cached to wait

This requires aCache Policy"To decide how to process these "unable time to process" data - This is what SharedFlowbufferandOverflow policy (onBufferOverflow)。

✅ 2. The meaning of replay = 2, extraBufferCapacity = 5

val sharedFlow = MutableSharedFlow<Int>(
    replay = 2,
    extraBufferCapacity = 5
)

These two control two cache areas:

💡 Total buffer size =replay + extraBufferCapacity

That is, the above configuration can be buffered in totalUp to 7 data。

This means that without collect, up to 7 pieces of data can be emitted without failure or loss.

✅ 3. onBufferOverflow = DROP_OLDEST / DROP_LATEST / SUSPEND What is?

This is controlThe buffer is fullWhen continuing the strategy of emit, what will be done with the strategy.

Supported policies:

Example description:

val flow = MutableSharedFlow<Int>(
    replay = 1,
    extraBufferCapacity = 2,
    onBufferOverflow = BufferOverflow.DROP_OLDEST
)

At this time, the total buffers are 3:

• If you continuously emit the 1, 2, 3rd item → you can enter the buffer.
• If no one is collect → buffer is full when emit Article 4.
• DROP_OLDESTStrategy: WillArticle 1 Value Removal, retain 2, 3, 4.

🔄 emit() and tryEmit():

• emit()yesSuspend function, maysuspend(If the buffer is full and the policy isSUSPEND）。
• tryEmit()yesNon-hanging,returntrue/falseIndicates whether the launch is successful.

✅ In a summary sentence:

This is the end of this article about the complete tutorial of flow sharedFlow in kotlin. For more related content of kotlin sharedFlow, please search for my previous articles or continue browsing the related articles below. I hope everyone will support me in the future!