I.e. something equivalent to
extension Publisher {
func send(_ subject: some Subject<Output, Failure>) -> AnyCancellable {
sink(
receiveCompletion: subject.send,
receiveValue: subject.send
)
}
}
There is a subscribe(_:) method which does that.
Here is an example:
private let originalSubject = PassthroughSubject<Void, Never>()
private let myRepublishingSubject = PassthroughSubject<Void, Never>()
private var cancellables: Set<AnyCancellable> = []
originalSubject
.subscribe(myRepublishingSubject)
.store(in: &cancellables)
myRepublishingSubject
.sink {
print("Republished")
}
.store(in: &cancellables)
originalSubject.send()
Replace originalSubject with your publisher.
However, this is probably not the best way to use Combine in your code, because Subjects are mostly used as a glue between imperative code and Combine code. Why can't you use the original Publisher directly?
Related
firebase method is working on worker thread automatically. but I have used coroutine and callbackflow to implement firebase listener code synchronously or get return from the listener.
below is my code that I explained
coroutine await with firebase for one shot
override suspend fun checkNickName(nickName: String): Results<Int> {
lateinit var result : Results<Int>
fireStore.collection("database")
.document("user")
.get()
.addOnCompleteListener { document ->
if (document.isSuccessful) {
val list = document.result.data?.get("nickNameList") as List<String>
if (list.contains(nickName))
result = Results.Exist(1)
else
result = Results.No(0)
//document.getResult().get("nickNameList")
}
else {
}
}.await()
return result
}
callbackflow with firebase listener
override fun getOwnUser(): Flow<UserEntity> = callbackFlow{
val document = fireStore.collection("database/user/userList/")
.document("test!!!!!")
val subscription = document.addSnapshotListener { snapshot,_ ->
if (snapshot!!.exists()) {
val ownUser = snapshot.toObject<UserEntity>()
if (ownUser != null) {
trySend(ownUser)
}
}
}
awaitClose { subscription.remove() }
}
so I really wonder these way is good or bad practice and its reason
Do not combine addOnCompleteListener with coroutines await(). There is no guarantee that the listener gets called before or after await(), so it is possible the code in the listener won't be called until after the whole suspend function returns. Also, one of the major reasons to use coroutines in the first place is to avoid using callbacks. So your first function should look like:
override suspend fun checkNickName(nickName: String): Results<Int> {
try {
val userList = fireStore.collection("database")
.document("user")
.get()
.await()
.get("nickNameList") as List<String>
return if (userList.contains(nickName)) Results.Exist(1) else Results.No(0)
} catch (e: Exception) {
// return a failure result here
}
}
Your use of callbackFlow looks fine, except you should add a buffer() call to the flow you're returning so you can specify how to handle backpressure. However, it's possible you will want to handle that downstream instead.
override fun getOwnUser(): Flow<UserEntity> = callbackFlow {
//...
}.buffer(/* Customize backpressure behavior here */)
I'm trying to run a Firebase Transaction under a suspended function in Kotlin and i see no documentation about it.
I'm using
implementation 'org.jetbrains.kotlinx:kotlinx-coroutines-play-services:1.5.2'
for coroutines with firebase (eg: setValue(*).await() ) but there seems to be no await function for runTransaction(*)
override suspend fun modifyProductStock(
product: ProductModel,
valueToModify: Long,
replace: Boolean
) {
CoroutineScope(Dispatchers.Main).launch {
val restaurantId = authRepository.restaurantId.value ?: throw Exception("No restaurant!")
val productId = product.id ?: throw Exception("No Product ID!")
val reference = FirebaseDatabase.getInstance().getReference("database/$restaurantId").child("products")
if (replace) {
reference.child(productId).child("stock").setValue(valueToModify).await()
} else {
reference.child(productId).child("stock")
.runTransaction(object : Transaction.Handler {
override fun doTransaction(p0: MutableData): Transaction.Result {
//any operation
return Transaction.success(p0)
}
override fun onComplete(p0: DatabaseError?, p1: Boolean, p2: DataSnapshot?) {
}
})
}
}
}
You could wrap it in suspendCoroutine:
val result: DataSnapshot? = suspendCoroutine { c ->
reference.child(productId).child("stock")
.runTransaction(object : Transaction.Handler {
override fun doTransaction(p0: MutableData): Transaction.Result {
//any operation
return Transaction.success(p0)
}
override fun onComplete(error: DatabaseError?, p1: Boolean, snapshot: DataSnapshot?) {
c.resume(snapshot)
}
})
}
suspendCoroutine
Obtains the current continuation instance inside suspend functions and suspends the currently running coroutine.
In this function both Continuation.resume and Continuation.resumeWithException can be used either synchronously in the same stack-frame where the suspension function is run or asynchronously later in the same thread or from a different thread of execution.
Given that the Kotlin example in the Firebase documentation on transactions uses the same callback style that you have, it seems indeed that there is no specific support for co-routines there.
It might be worth posting an issue on the Android SDK repo to get it added, or hear why it wasn't added.
I am new with Axon and maybe I missed something, but need help to understand.
I have a simple food cart aggregate.
Here is example:
#Aggregate
class FoodCard {
#AggregateIdentifier
private lateinit var foodCardId: UUID
private lateinit var selectedProduct: MutableMap<UUID, Int>
constructor()
#CommandHandler
constructor(command: CreateFoodCartCommand) {
AggregateLifecycle.apply(FoodCartCreateEvent(
UUID.randomUUID()
))
}
#CommandHandler
fun handle(command: SelectProductCommand) {
AggregateLifecycle
.apply(ProductSelectedEvent(foodCardId, command.productId, command.quantity))
}
#CommandHandler
fun handle(command: DeleteFoodCartCommand) {
AggregateLifecycle
.apply(FoodCartDeleteEvent(foodCardId))
}
#CommandHandler
fun handle(command: DeselectProductCommand) {
val productId = command.productId
if (!selectedProduct.containsKey(productId)) {
throw ProductDeselectionException("ProductDeselectionException")
}
AggregateLifecycle
.apply(ProductDeselectEvent(foodCardId, productId, command.quantity))
}
#EventSourcingHandler
fun on(event: FoodCartCreateEvent) {
foodCardId = event.foodCardId
selectedProduct = mutableMapOf()
}
#EventSourcingHandler
fun on(event: ProductSelectedEvent) {
selectedProduct.merge(
event.productId,
event.quantity
) {a, b -> a + b}
}
}
As ES I am using Axon Server.
For FoodCard projector I am using JPA repository that connects to DB.
I want to get all foodcards that contain special product (concrete UUID) and change quantity to -1 for all of them.
I understood there are two types of actions -> read and write
So the question how to correctly implement this flow with Axon?
Thanks
from your explanation and code I feel that you will probably need to complete your implementation of DeselectProductCommand introducing an EventSourcingHandler for ProductDeselectEvent. If I understood correctly your "quantity" information is stored into the selectProduct Map. In this case, based on your code, I see that the information of the quantity that should be subtracted to your product is in the command.
You will also need a Query, such as FindAllFoodCardByProductId, that will retrieve the foodCardId aggregate identifier that contains a certain productId: this operation will be performed on your Projection through the jpa repository.
As a reference you can have a look at the ref guide here https://docs.axoniq.io/reference-guide/implementing-domain-logic/query-handling on how to use QueryGateway into your controller and implement a QueryHandler into your Projection.
Corrado.
I'm dealing with some asynchronous functions and trying to update views. In short I have function 1 with asynchronous function that will return a string to be passed to function 2. I am updating views in both functions, on main thread. It all works but I need to understand if this is correct way.
class A {
var varA = ""
var varB = ""
func f1 (_ completion: #escaping (String) -> void ){
some asynchronous call ... { in
...
DispatchQueue.main.async {
self.varA = "something"
sef.labelA.text = self.varA
completion(self.varA)
}
}
}
func f2 (_ string: String){
another asynchronous call ... { in
...
DispatchQueue.main.async {
self.varB = string
sef.labelB.text = self.varB
}
}
}
// funcation call
f1(completion: f2)
}
Three questions, 1) What is the right way to run a dependent function where there is wait for an asynchronous callback?
2) Is DispatchQueue.main.async needed to update views?
3) Is it ok to call async func in another async callback? Isn't there chance self may be nil in some cases if you are updating views in some escaping function?
I'm going to try helping you according to your questions:
Question 1) There are many right ways and each developer can have its own logic, but in this case, what I personally would probably do is something like this:
class A {
func f1 (_ completion: #escaping (String) -> void ){
some asynchronous call ... { in
...
DispatchQueue.main.async { [weak self] in // 1
guard let strongSelf = self else { return } // 2
let varA = "something" // 3
strongSelf.label.text = varA
completion(varA) // 4
}
}
}
func f2 (_ string: String){
another asynchronous call ... { in
...
DispatchQueue.main.async {
sef.labelB.text = string // 5
}
}
}
// function call
// 6
f1(completion: { [weak self] text in
guard let strongSelf = self else { return }
strongSelf.f2(text)
})
}
1 - Here I'm using [weak self] to avoid retain cycles.
2 - Just unwrapping the optional self, case it's nil, I'll just return.
3 - In your case, it's not really necessary to have class variables, so I'm just creating local variables inside the block.
4 - Finally, I'm calling the completion with the variable containing the string.
5 - I also don't really need to set a class variable in here, so I'm just updating the label text with the string provided as a paramater.
6 - Then, I just need to call the first function and use the completion block to call the second after the first one completes.
Question 2) Yes, you must call DispatchQueue.main to update the view. This way your making sure that your code will be executed in the main thread that is crucial for things interacting with UI because it allow us to have a sincronization point as you can read in Apple's documentation.
Question 3) Using [weak self] and guard let strongSelf = self else { return }, I'm avoiding retain cycles and the cases where self can be nil.
is it posible to have callable objects on ActionScript? For example:
class Foo extends EventDispatcher
{
Foo() { super(); }
call(world:String):String
{
return "Hello, " + world;
}
}
And later...
var foo:Foo = new Foo();
trace( foo("World!") ); // Will NOT work
Why would you need to do this? (I'm not criticising, just interested!) Functions in AS3 are themselves first-class citizens, and can be passed around as arguments.
e.g.
public function main(foo:Function):void
{
trace(foo("World!")); // Will work, assuming foo = function(str:String):String {...}
}
No, only functions/methods can be called in this way. If the only reason is you want to type fewer characters, then you should shorten the length of the instance names and method names.
One option is to use a closure:
public function Foo():Function {
var bar:String;
return function (world:String):String {
var msg:String;
if (bar) {
msg = bar + ' says "Hello, ' + world + '"';
} else {
msg = "Hello, " + world;
}
bar = world;
return msg;
}
}
...
var foo = Foo();
trace( foo("World!") );
This is a much simplified case of the larger pattern of implementing objects as functions. As such, it's more useful in languages that support FP but not OOP, but does technically give you a callable "object". The syntax may be a little off, but:
public function createFoo(barInit, ...):Function {
var slots = {
greeter: barInit, ...
};
var methods = {
'get': function(name) { return slots[name]; }
'set': function(name, value) { slots[name] = value; }
greet: function(whom) {
var msg = slots.greeter + ' says "Hello, ' + whom + '"'
slots.greeter = whom;
return msg;
},
...
};
return function (method:String):* {
args = Array.splice.call(arguments, 1);
return methods[method].apply(null, args);
}
}
var foo = createFoo('Kermit');
trace(foo('greet', "World"));
trace(foo('greet', "Sailor"));
You probably don't want to do it in AS.
As others had said, you can't have callable objects. However, if for some reason you want to have stateful functions, you can achieve it with help of static class variables and package level functions. For example:
// com/example/foo/Helper.as
package com.example.foo {
public class Helper {
private static var _instance:Foo;
public static var data:String;
public static function get instance():Helper
{
if(!_instance) { _instance = new Helper(); }
return _instance;
}
}
}
// com/example/foo/hello.as
package com.example.foo {
public function hello(world:String):void
{
if(Helper.instance.data)
{
trace("Bye, " + Helper.instance.data);
}
trace("Hello, " + world);
Helper.instance.data = world;
}
}
When used, it will print different things.
hello("World!"); // traces "Hello, World!"
hello("People"); // traces "Bye, World!" and "Hello, People"
note: both the constructor and the method declaration miss the keywords public function to even compile, but I suppose that's not the original code. :)
the answer is: you can't.
my question is: what do you want to accomplish?
Functions are the only callable values. And Functions are primitives in ActionScript, much as ints, or Booleans, so there is no meaningful way to extend them.
If you want it to be an object, do it the Java way, defining an ICallable interface, and actually call a method, or just really use a function. closures provide the most simple and flexible possibility to create stateful functions, if that is what you want.
edit: well, you can do this (as an example):
private var fooInst:Foo = new Foo();
protected var foo:Function = fooInst.call;
and then the following workst as you wish:
<mx:Label text="{foo('Whatever')}"/>
its maybe even a little more flexible, although you lose the benefits of strict typing.
greetz
back2dos