coroutines | Exploring C20 Coroutines | Android library

MainCoroutineRule and runBlocking are seemingly similar but there are clear differences.

Defenition of MainCoroutineRule:

MainCoroutineRule installs a TestCoroutineDispatcher for Disptachers.Main. Since it extends TestCoroutineScope, you can directly launch coroutines on the MainCoroutineRule as a [CoroutineScope]...

Defenition of runBlocking:

Runs a new coroutine and blocks the current thread interruptibly until its completion. This function should not be used from a coroutine. It is designed to bridge regular blocking code to libraries that are written in suspending style, to be used in main functions and in tests. ...

In terms of definition, runBlocking is used for the sole purpose of synchronous execution in coroutines. It is not only used in tests but also used in UI management, Data Management, and several other areas in Android Development.

Whereas runBlocking is used more generally, MainCoroutineRule is used only in tests and it has the same synchronous execution behavior found in runBlocking. However, MainCoroutineRule has more specialized testing features not found in runBlocking such as control flow management. In addition, using MainCoroutineRule will significantly make your testing code cleaner.

In a much more detailed manner, MainCoroutineRule is directly related to JUnit Rule. If you have used JUnit before, you might remember filling out @Before @After to provide a testing environment for the test cases. If you have multiple test cases, you would eventually have to write multiple @Before @After to provide multiple testing environments, which could possibly lead to redundant boilerplate code. Now, this is where MainCoroutineRule shines! With MainCoroutineRule you can declare a testing environment once and reuse them for multiple test cases.

Compare two sample codes below:

Test case without using MainCoroutineRule

Your code works, the issue why you are not able to create res successfully is because the code does not raise just the normal Exception class. Since the task fails it ends up calling asyncio.exceptions.CancelledError which if we take a look in the documentation inherits from BaseException not Exception. This change is new as of Python 3.8 and since you are using Python 3.9 that change is live. Changing your code slightly to the following yields:

One idea could be to return the Job in method1 like the following:

I don't think MutableState is designed to be used in the ViewModel layer, since it's an observable integrated with the compose runtime. You could create a MutableStateFlow instead and use collectAsState() from the view layer.

In your case the issue is probably, because of the state is captured in a coroutine invoked outside composition.

the question is, what is its type?

That question is settled in the same way as any variable declaration with initialization: by the type of what's on the right-hand side of the equal sign.

For example, if you have a method that returns a String (let's name it getMyString), then when you say

lua_yield is a C function, and C does not have a mechanism to magically jump back into a function that has halted. lua_yield uses the C standard library longjmp function to arbitrarily jump out of the function that called it. But it can't come back.

So what happens is that your C function yields, exiting the function and returning control to the Lua code that called coroutine.resume. The resume was successful, so true is printed. You then resume the coroutine again, which begins execution at the site in Lua code that called the C function that yielded. That code then exits the coroutine normally. Since the resume was also successful, true is printed again.

But the coroutine is now exhausted and therefore cannot be resumed.

C functions don't have a clean way to be "resumed". Indeed, the Lua 5.1 documentation explicitly states:

This function should only be called as the return expression of a C function, as follows:

Keeping the whole discussion aside of LiveData vs SharedFlow or StateFlow. Coming onto ViewModels as you asked. If we are to go by documentation

The ViewModel class is designed to store and manage UI-related data in a lifecycle conscious way. The ViewModel class allows data to survive configuration changes such as screen rotations.

UI controllers such as activities and fragments are primarily intended to display UI data, react to user actions, or handle operating system communication, such as permission requests. Requiring UI controllers to also be responsible for loading data from a database or network adds bloat to the class. Assigning excessive responsibility to UI controllers can result in a single class that tries to handle all of an app's work by itself, instead of delegating work to other classes. Assigning excessive responsibility to the UI controllers in this way also makes testing a lot harder.

It's easier and more efficient to separate out view data ownership from UI controller logic.

I guess this sums it up quite well. It is true that lifeCycleScope can eliminate the need of ViewModel in a way, but ViewModel does more than just being a holder for LiveData.

Even if you want to use SharedFlow or StateFlow over LiveData I would suggest you still make use of ViewModel and inside it use a viewModelScope instead to still perform the usual and required separation of concerns between UI and data.

it looks like you are using wrong dependency. change to below line :

Leaving my understanding here:

Question 1: LaunchedEffect should be used when you want that some action must be taken when your composable is first launched. For example, when you want to request some data from your ViewModel or run some sort of animation...
rememberCoroutineScope on the other hand, is specific to store the Coroutine scope allowing the code to launch some suspend function... imho, the only relation between them is that you can also use a LaunchedEffect to launch a coroutine...

Question 2: As you can see in the docs, rememberCoroutineScope will keep the reference of the coroutine's scope in a specific point of the composition. Therefore, if a given composable is removed from the recomposition, that coroutine will be cancelled automatically. For instance, you have the following composable calls A -> B -> C. If you remember the coroutine scope in C and it is removed from the composition, the coroutine is automatically cancelled. But if you remember from A, pass the scope through B and C, use this scope in C, and then C is removed, the coroutine will continue running (because it was remembered in A)...