Unit testing code that uses Task.Factory.StartNew().ContinueWith() - c#

so I have some code
Task.Factory.StartNew(() => this.listener.Start()).ContinueWith(
(task) =>
{
if (task.IsCompleted)
{
this.status = WorkerStatus.Started;
this.RaiseStatusChanged();
this.LogInformationMessage("Worker Started.");
}
});
When I am testing I am mocking all the dependant objects (namley this.listener.Start()). the problem is that the test finishes executing before ContinueWith can be called. When I debug it gets called fine due to the extra delay of me stepping through code.
so how can I - from the test code in a different assembly - ensure that the code is run before my test hits its asserts?
I could just use Thread.Sleep ... but this seems like a really hacky way of doing it.
I guess I am looking for the Task version of Thread.Join.

Consider the following:
public class SomeClass
{
public void Foo()
{
var a = new Random().Next();
}
}
public class MyUnitTest
{
public void MyTestMethod()
{
var target = new SomeClass();
target.Foo(); // What to assert, what is the result?..
}
}
What is the value assigned to a? You cannot tell, unless the result is returned outside the method Foo() (as the return value, a public property, an event, etc.).
The process of "coordinating the actions of threads for a predictable outcome" is called Synchronization.
One of the easiest solutions in your case might be to return the instance of Task class and the use its Wait() method:
var task = Task.Factory.StartNew(() => Method1())
.ContinueWith(() => Method2());
No need to wait for the first task, because ContinueWith() creates a continuation that executes asynchronously when the target Task completes (MSDN):
task.Wait();

I don't think there is an easy-yet-practical way of doing this. Ran into the same problem myself just now and Thread.Sleep(X) is by far the simplest (if not elegant) way of getting around the problem.
The only other solution that I considered is hiding the Task.Factory.StartNew() call behind an interface that you can mock from your test thus removing the actual execution of the task entirely in the test scenario (but still have an expectation that the interface method will be called. For example:
public interface ITaskWrapper
{
void TaskMethod();
}
And your concrete implementation:
public class MyTask : ITaskWrapper
{
public void TaskMethod()
{
Task.Factory.StartNew(() => DoSomeWork());
}
}
Then just mock ITaskWrapper in your test method and set an expectation on TaskMethod being called.

If there's any way for you to be notified of when the processing has ended (can you add a handler for that StatusChanged event?), use a ManualResetEvent and wait on it with a reasonable timeout. If the timeout expired fail the test, otherwise go on and perform your assertions.
E.g.
var waitHandle = new ManualResetEvent(false);
sut.StatusChanged += (s, e) => waitHandle.Set();
sut.DoStuff();
Assert.IsTrue(waitHandle.WaitOne(someTimeout), "timeout expired");
// do asserts here

The continuation task will still run regardless of whether the initial task completed before the ContinueWith() call or not. I double checked this with the following:
// Task immediately exits
var task = Task.Factory.StartNew(() => { });
Thread.Sleep(100);
// Continuation on already-completed task
task.ContinueWith(t => { MessageBox.Show("!"); });
Debug further. Maybe your task is failing.

When dealing with asynchronous processes during code under test that use Reactive Extensions, one approach is to use a TestScheduler. The TestScheduler can be moved forward in time, drained of all shceduled tasks, etc. So your code under test can take an IScheduler, which you provide a TestScheduler instance for. Then your test can manipulate time without needing to actually sleep, wait or synchronize. An improvement on this approach is Lee Campbell's ISchedulerProvider approach.
If you use Observable.Start instead of Task.Factory.StartNew in your code, you can then use your TestScheduler in the unit test to push through all the scheduled tasks.
For example, your code under test could look something like this:
//Task.Factory.StartNew(() => DoSomething())
// .ContinueWith(t => DoSomethingElse())
Observable.Start(() => DoSomething(), schedulerProvider.ThreadPool)
.ToTask()
.ContinueWith(t => DoSomethingElse())
and in your unit test:
// ... test code to execute the code under test
// run the tasks on the ThreadPool scheduler
testSchedulers.ThreadPool.Start();
// assertion code can now run

Related

Should I be using await inside my Task.Run()?

** I've summarised this question at the bottom with an edit **
This has been asked before but I think my circumstances are different.
I am processing multiple requests simultaneously.
This is my code to do that, it runs in a loop. I've removed a bit of code that handles the taskAllocated variable for brevity.
while (!taskAllocated)
{
lock (_lock)
{
// Find an empty slot in the task queue to insert this task
for (i = 0; i < MaxNumTasks; i++)
{
if (_taskQueue[i] == null)
{
_taskQueue[i] = Task.Run(() => Process());
_taskQueue[i].ContinueWith(ProcessCompleted);
break;
}
}
}
}
Process is a typical async Task Process() { CpuIntensiveStuff(); } method.
I've been running the above code, and it has been working fine. It multithreads nicely. Whenever an item comes in, it will find an empty slot in the task queue, and kick it off. When the task completes, the ProcessCompleted method runs, and frees up the slot.
But then I thought, shouldn't I be using await inside my Task.Run? Something like:
_taskQueue[i] = Task.Run(async () => await Process());
After thinking about it, I'm not sure. ContinueWith triggers correctly, when the task has completed, so perhaps it's not necessary.
I ask because I wanted to monitor and log how long each task takes to complete.
So Instead of Process(), I would make another method like:
async Task DoProcess()
{
var sw = Stopwatch.StartNew();
Process();
sw.Stop();
Log(sw.ElapsedMilliseconds);
}
And it occurred to me that if I did that, I wasn't sure if I'd need to await Process(); or not, in addition to not knowing if I should await inside the Task.Run()
I'm in a bit of a tizz about this. Can anyone offer guidance?
Edit:
To summarise:
If Somemethod is:
void SomeMethod() { }
Then
Task.Run(() => SomeMethod()); is great, calls SomeMethod on a new 'thread' (not technically, but you know what I mean).
However, my SomeMethod is actually:
async Task SomeMethod() { }
Do you need to do anything special with Task.Run()?
My code, I am not, I am just straight up ignoring that it's an async Task, and that seems to work:
Task.Run(() => SomeMethod()); // SomeMethod is async Task but I am ignoring that
But I'm not convinced that it a) should work or b) is a good idea. The alternative could be to do:
Task.Run(async() => await SomeMethod());
But is there any point? And this is compounded by the fact I want to really do:
Task.Run(() =>
{
someCode();
var x = startTimer();
SomeMethod();
var y = stopTimer();
someMoreCode()
});
but without await I'm not sure it will wait for somemethod to finish and the timer will be wrong.
Things become more clear if you do not use anonymous methods. For example,
Task.Run(() => Process())
is equivalent to this:
Task.Run(DoSomething);
Task DoSomething() {
return Process();
}
Whereas
Task.Run(async () => await Process())
is equivalent to this:
Task.Run(DoSomething);
async Task DoSomething() {
await Process();
}
In most cases, there is no functional difference between return SomethingThatReturnsATask() and return await SomethingThatReturnsATask(), and you usually want to return the Task directly and not use await (for reasons described here). When used inside Task.Run, things could easily go bad if the .NET team didn't have your back.
It is important to note that asynchronous methods start running on the same thread just like any other method. The magic happens at the first await that acts on an incomplete Task. At that point, await returns its own incomplete Task. That's important - it returns, with a promise to do the rest later.
This could have meant that the Task returned from Task.Run would complete whenever Process() returns a Task. And since Process() returns a Task at the first await, that would happen when it has not yet totally completed.
The .NET team has your back
That is not the case however, because Task.Run has a specific overload for when you give it a method returning a Task. And if you look at the code, it returns a Task *that is tied to the Task you return.
That means that the Task returned from Task.Run(() => Process()) will not complete until the Task returned from Process() has completed.
So your code is fine the way it is.

What's the difference between Foo().Result and Task.Run(() => Foo()).Result in C#?

In C# what is the difference between these two statements? If I use the first one in my constructor in my test classes I get a deadlock, or something similar, and the tests never finish. With the second one the code works.
// Deadlock.
var r = MyMethod().Result;
// Works.
var r = Task.Run(() => MyMethod()).Result;
Update: There is a bit more context in this commit: https://github.com/webCRMdotcom/erp-integrations/pull/92/commits/dd8af89899ce1de837ef6e34f0688a685a5cea3b.
The difference is the starting thread context.
Here a simple sample
using System;
using System.Threading.Tasks;
public class Program
{
public static void Main()
{
string r;
OutputThreadInfo("Main");
r = MyMethod().Result;
r = Task.Run( () => MyMethod() ).Result;
}
public static async Task<string> MyMethod()
{
OutputThreadInfo("MyMethod");
await Task.Delay(50);
return "finished";
}
private static void OutputThreadInfo(string context)
{
Console.WriteLine("{0} {1}",context,System.Threading.Thread.CurrentThread.ManagedThreadId);
}
}
.net fiddle
which will output
Main 32
MyMethod 32
MyMethod 63
The first call of MyMethod will start at the same thread as Main and if started from a thread with a synchronization context it will block.
The second call of MyMethod will start from a different thread (worker thread from thread pool) as Main which does not have a synchronization context and therefor will not block.
PS You should keep in mind that Console applications do not have a synchronization context as default but WinForms, WPF, UWP application do have and so will behave somehow different on async/await
Task.Result and Task.Wait block the current thread you should use await for this to work without any problems. (Though they only block if not already completed).
The second line will create a task and will start it's execution on a available thread in the Thread Pool and that's why it doesn't block.
This is because the Task construct when used with async-await will generate a State Machine that keeps track of all the awaits used in the code block and when all finishes then it can return the result. Keep in mind thought that depending on the Synchronization Context you are in, the code after await may run on a different thread then the one the task started.
So what I do when I have to execute synchronous an async method I use a small piece of code like this:
private static readonly TaskFactory _tf = new TaskFactory(
CancellationToken.None, TaskCreationOptions.None,
TaskContinuationOptions.None, TaskScheduler.Default);
public static TResult RunSync<TResult>(Func<Task<TResult>> func)
{
return _tf.StartNew<Task<TResult>>((Func<Task<TResult>>) (() =>
{
return func();
})).Unwrap<TResult>().GetAwaiter().GetResult();
}
Keep in mind that, if needed, you have to use the same CultureInfo inside the RunSync StarNew task factory call so you won't have this kind of problems.

Verify that task is being awaited

I have the following code which i'd like to test:
private Task _keepAliveTask; // get's assigned by object initializer
public async Task EndSession()
{
_cancellationTokenSource.Cancel(); // cancels the _keepAliveTask
await _logOutCommand.LogOutIfPossible();
await _keepAliveTask;
}
It is important that the EndSession Task only ends once the `_keepAliveTask' ended. However, I'm struggling to find a way to test it reliably.
Question: How do i unit test the EndSession method and verify that the Task returned by EndSession awaits the _keepAliveTask.
For demonstration purposes, the unit test could look like that:
public async Task EndSession_MustWaitForKeepAliveTaskToEnd()
{
var keepAliveTask = new Mock<Task>();
// for simplicity sake i slightly differ from the other examples
// by passing the task as method parameter
await EndSession(keepAliveTask);
keepAliveTask.VerifyAwaited(); // this is what i want to achieve
}
Further criterias:
- reliable test (always passes when implementation is correct, always fails when implementation is wrong)
- cannot take longer than a few milliseconds (it's a unit test, after all).
I have already taken several alternatives into considerations which i'm documenting below:
non-async method
If there wouldn't be the call to _logOutCommand.LogOutIfPossible() it would be quite simple: i'd just remove the async and return _keepAliveTask instead of awaiting it:
public Task EndSession()
{
_cancellationTokenSource.Cancel();
return _keepAliveTask;
}
The unit test would look (simplified):
public void EndSession_MustWaitForKeepAliveTaskToEnd()
{
var keepAliveTask = new Mock<Task>();
// for simplicity sake i slightly differ from the other examples
// by passing the task as method parameter
Task returnedTask = EndSession(keepAliveTask);
returnedTask.Should().be(keepAliveTask);
}
However, there's two arguments against this:
i have multiple task which need awaiting (i'm considering Task.WhenAll further down)
doing so only moves the responsibility to await the task to the caller of EndSession. Still will have to test it there.
non-async method, sync over async
Of course, I could do something similar:
public Task EndSession()
{
_cancellationTokenSource.Cancel(); // cancels the _keepAliveTask
_logOutCommand.LogOutIfPossible().Wait();
return _keepAliveTask;
}
But that is a no-go (sync over async). Plus it still has the problems of the previous approach.
non-async method using Task.WhenAll(...)
Is a (valid) performance improvement but introduces more complexity:
- difficult to get right without hiding a second exception (when both fail)
- allows parallel execution
Since performance isn't key here i'd like to avoid the extra complexity. Also, previously mentioned issue that it just moves the (verification) problem to the caller of the EndSession method applies here, too.
observing effects instead of verifying calls
Now of course instead of "unit" testing method calls etc. I could always observe effects. Which is: As long as _keepAliveTask hasn't ended the EndSession Task mustn't end either. But since I can't wait indefinite one has to settle for a timeout. The tests should be fast so a timeout like 5 seconds is a no go. So what I've done is:
[Test]
public void EndSession_MustWaitForKeepAliveTaskToEnd()
{
var keepAlive = new TaskCompletionSource<bool>();
_cancelableLoopingTaskFactory
.Setup(x => x.Start(It.IsAny<ICancelableLoopStep>(), It.IsAny<CancellationToken>()))
.Returns(keepAlive.Task);
_testee.StartSendingKeepAlive();
_testee.EndSession()
.Wait(TimeSpan.FromMilliseconds(20))
.Should().BeFalse();
}
But I really really dislike this approach:
hard to understand
unreliable
or - when it's quite reliable - it takes a long time (which unit tests shouldn't).
If all you want is to verify that EndSession is awaiting _keepAliveTask (and you really have full control over _keepAliveTask) then you can create your own awaitable type instead of Task the signals when it's awaited and check that:
public class MyAwaitable
{
public bool IsAwaited;
public MyAwaiter GetAwaiter()
{
return new MyAwaiter(this);
}
}
public class MyAwaiter
{
private readonly MyAwaitable _awaitable;
public MyAwaiter(MyAwaitable awaitable)
{
_awaitable = awaitable;
}
public bool IsCompleted
{
get { return false; }
}
public void GetResult() {}
public void OnCompleted(Action continuation)
{
_awaitable.IsAwaited = true;
}
}
Since all you need to await something is that has a GetAwaiter method that returns something with IsCompleted, OnCompleted and GetResult you can use the dummy awaitable to make sure _keepAliveTask is being awaited:
_keepAliveTask = new MyAwaitable();
EndSession();
_keepAliveTask.IsAwaited.Should().BeTrue();
If you use some mocking framework you can instead make Task's GetAwaiter return our MyAwaiter.
Use TaskCompletionSource and set its result at a known time.
Verify that before setting the result, the await on EndSession hasn't completed.
Verify that after setting the result, the await on EndSession has completed.
A simplified version could look like the following (using nunit):
[Test]
public async Task VerifyTask()
{
var tcs = new TaskCompletionSource<bool>();
var keepAliveTask = tcs.Task;
// verify pre-condition
Assert.IsFalse(keepAliveTask.IsCompleted);
var waitTask = Task.Run(async () => await keepAliveTask);
tcs.SetResult(true);
await waitTask;
// verify keepAliveTask has finished, and as such has been awaited
Assert.IsTrue(keepAliveTask.IsCompleted);
Assert.IsTrue(waitTask.IsCompleted); // not needed, but to make a point
}
You can also add a short delay at the waitTask to ensure any synchronous execution would be faster, something like:
var waitTask = Task.Run(async () =>
{
await Task.Delay(1);
await keepAliveTask;
});
And if you don't trust your unit test framework to deal correctly with async, you can set a completed flag as part of the waitTask, and check for that in the end. Something like:
bool completed = false;
var waitTask = Task.Run(async () =>
{
await Task.Delay(1);
await keepAliveTask;
completed = true;
});
// { .... }
// at the end of the method
Assert.IsTrue(completed);

Awaiting an empty Task spins forever (await new Task(() => { }))

I'm trying to get my head around this code:
[TestFixture]
public class ExampleTest
{
[Test]
public void Example()
{
AwaitEmptyTask().Wait();
}
public async Task AwaitEmptyTask()
{
await new Task(() => { });
}
}
The method Example never ends and blocks forever. Why??
The fix (from Stubbing Task returning method in async unit test) is to replace await new Task( () => {}) with return Task.FromResult<object>(null); but again, why is this necessary?
I know there are a bunch of questions similar to this one, but none that I've seen seem to explain why this is happening:
Await method that returns Task - spins forever?: I included the await keyword AFAIK correctly
Stubbing Task returning method in async unit test: Doesn't explain why.
Why will an empty .NET Task not complete if started and waited for from a static constructor?: I'm not running in a static context and as far as I can tell the context is consistent
Async methods return null: also gives the solution, but doesn't explain why
async await return Task: discusses Task.FromResult but not why await new Task(() => {}) doesn't work
You're creating a task and never starting it, so it never finishes.
You should be using Task.Run to create a task that you want to start executing immediately, rather than using the Task constructor.
You need to call Start() on your task.
Otherwise, it will never finish.
Even if you changed it to
await Task.Run(() => { });
it may still never complete. When you hit this line, the program creates a continuation task that will continue when the task completes. The fact that the task is empty is irrelevant. At that point, control flow goes back to
AwaitEmptyTask().Wait();
which waits for the task to complete and blocks the current thread.
When the task does complete, it attempts to continue on the current thread. However, it can't, because the thread is blocked by the call to Wait().
Change the awaited line to
await Task.Run(() => { }).ConfigureAwait(false);
This is a common deadlocking problem. Don't mix Wait() and await. See http://msdn.microsoft.com/en-us/magazine/jj991977.aspx

Why does the Task.ContinueWith fail to execute in this Unit Test?

I have come across a problem with a unit test that failed because a TPL Task never executed its ContinueWith(x, TaskScheduler.FromCurrentSynchronizationContext()).
The problem turned out to be because a Winforms UI Control was accidentally being created before the Task was started.
Here is an example that reproduces it. You will see that if you run the test as-is, it passes. If you run the test with the Form line uncommented, it fails.
[TestClass]
public class UnitTest1
{
[TestMethod]
public void TestMethod1()
{
// Create new sync context for unit test
SynchronizationContext.SetSynchronizationContext(new SynchronizationContext());
var waitHandle = new ManualResetEvent(false);
var doer = new DoSomethinger();
//Uncommenting this line causes the ContinueWith part of the Task
//below never to execute.
//var f = new Form();
doer.DoSomethingAsync(() => waitHandle.Set());
Assert.IsTrue(waitHandle.WaitOne(10000), "Wait timeout exceeded.");
}
}
public class DoSomethinger
{
public void DoSomethingAsync(Action onCompleted)
{
var task = Task.Factory.StartNew(() => Thread.Sleep(1000));
task.ContinueWith(t =>
{
if (onCompleted != null)
onCompleted();
}, TaskScheduler.FromCurrentSynchronizationContext());
}
}
Can anyone explain why this is the case?
I thought it might have been because the wrong SynchronizationContext is used, but actually, the ContinueWith never executes at all! And besides, in this unit test, whether or not it is the correct SynchronizationContext is irrelevant because as long as the waitHandle.set() is called on any thread, the test should pass.
I overlooked the comments section in your code, Indeed that fails when uncommenting the var f = new Form();
Reason is subtle, Control class will automatically overwrite the synchronization context to WindowsFormsSynchronizationContext if it sees that SynchronizationContext.Current is null or its is of type System.Threading.SynchronizationContext.
As soon as Control class overwrite the SynchronizationContext.Current with WindowsFormsSynchronizationContext, all the calls to Send and Post expects the windows message loop to be running in order to work. That's not going to happen till you created the Handle and you run a message loop.
Relevant part of the problematic code:
internal Control(bool autoInstallSyncContext)
{
...
if (autoInstallSyncContext)
{
//This overwrites your SynchronizationContext
WindowsFormsSynchronizationContext.InstallIfNeeded();
}
}
You can refer the source of WindowsFormsSynchronizationContext.InstallIfNeeded here.
If you want to overwrite the SynchronizationContext, you need your custom implementation of SynchronizationContext to make it work.
Workaround:
internal class MyContext : SynchronizationContext
{
}
[TestMethod]
public void TestMethod1()
{
// Create new sync context for unit test
SynchronizationContext.SetSynchronizationContext(new MyContext());
var waitHandle = new ManualResetEvent(false);
var doer = new DoSomethinger();
var f = new Form();
doer.DoSomethingAsync(() => waitHandle.Set());
Assert.IsTrue(waitHandle.WaitOne(10000), "Wait timeout exceeded.");
}
Above code works as expected :)
Alternatively you could set WindowsFormsSynchronizationContext.AutoInstall to false, that will prevent automatic overwriting of the synchronization context mentioned above.(Thanks for OP #OffHeGoes for mentioning this in comments)
With the line commented out, your SynchronizationContext is the default one you created. This will cause TaskScheduler.FromCurrentSynchrozisationContext() to use the default scheduler, which will run the continuation on the thread pool.
Once you create a Winforms object like your Form, the current SynchronizationContext becomes a WindowsFormsSynchronizationContext, which in turn will return a scheduler that depends on the WinForms message pump to schedule the continuation.
Since there is no WinForms pump in a unit test, the continuation never gets run.

Categories