c# “finally” block that only runs on exceptions

Question

Edit: I have looked at the answers code: NONE of them do what I want (I've checked). It would seem that there is no way to do what I want in native c#. I guess that's not a disaster just a shame given that .NET does support it (see accepted answer).

Thanks all.

---

I have c# code (part of a test framework that will never be run except under a debugger) like this who's point it to avoid actually catching the exception as that makes debugging the code in the unwound part of the stack a royal pain.

Bool bad = true;
try
{
   MightThrow();
   bad = false;
}
finally
{
   if(bad) DoSomeLoggingOnFailure();

   //// Does not catch!!!! 
   //// exception continues to unwind stack.

   //// Note that re throwing the exception is NOT
   //// the same as not catching it in the first place
}

is their a better way to do this?

A solution would have to behave exactly like that under the debugger with regards to un-caught exceptions. It would have to result in the only one first chance exception and the debugger breaking at the point that the exception was originally thrown, not in a catch block.

Specifically I need the debugger on un-caught exceptions to stop a in side MightThrow.

The following doesn't work because it fails to have the debugger break in the correct place

try { ... } catch { throw; }

And this doesn't work because it loses stack info (and also breaks in the wrong place).

try { ... } catch(Excption e) { throw e; }

I known that in D I could use a scope(failure) block

Answer 1

So, in .NET what you're asking for is theoretically possible, but it's not going to be easy.

CIL actually defines five types of exception handling block! The try, catch and finally ones you're used to in C#, and two others:

• filter - similar to a catch block but can run arbitrary code to determine whether it wants to handle the error, rather than just matching on type. This block has access to the exception object, and has the same effect on the exception stack trace as a catch block.

• fault - similar to a finally block, however it is only run when an exception occurs. This block does not have access to the exception object, and has no effect on the exception stack trace (just like a finally block).

filter is available in some .NET languages (e.g. VB.NET, C++/CLI) but is not available in C#, unfortunately. However I don't know of any language, other than CIL, that allows the fault block to be expressed.

Because it can be done in IL means not all is lost, though. In theory you could use Reflection.Emit to dynamically emit a function that has a fault block and then pass the code you want to run in as lambda expressions (i.e. one for the try part, one for the fault part, and so on), however (a) this isn't easy, and (b) I'm unconvinced that this will actually give you a more useful stack trace than you're currently getting.

Sorry the answer isn't a "here's how to do it" type thing, but at least now you know! What you're doing now is probably the best approach IMHO.

---

Note to those saying that the approach used in the question is 'bad practice', it really isn't. When you implement a catch block you're saying "I need to do something with the exception object when an exception occurs" and when you implement a finally you're saying "I don't need the exception object, but I need to do something before the end of the function".

If what you're actually trying to say is "I don't need the exception object, but I need to do something when an exception occurs" then you're half way between the two, i.e. you want a fault block. As this isn't available in C#, you don't have an ideal option, so you may as well choose the option that is less likely to cause bugs by forgetting to re-throw, and which doesn't corrupt the stack trace.

Answer 2

How about this:

try
{
  MightThrow();
}
catch
{
  DoSomethingOnFailure();
  throw; // added based on new information in the original question
}

Really, that's all you did. Finally is for things that must run regardless of whether an exception occurred.

[Edit: Clarification]

Based on the comments you've been mentioning, you want the exception to continue being thrown without modifying its original stack trace. In that case, you want to use the unadorned throw that I've added. This will allow the exception to continue up the stack and still allow you to handle part of the exception. Typical cases might be to close network connections or files.

[Second edit: Regarding your clarification]

Specifically I need the debugger on un-caught exceptions to stop at the original point of the throw (in MightThrow) not in the catch block.

I would argue against ever breaking a best-practice (and yes, this is a best-practice for partially handling exceptions) to add some minor value to your debugging. You can easily inspect the exception to determine the location of the exception throw.

[Final edit: You have your answer]

kronoz has thoughtfully provided you with the answer you sought. Don't break best practices -- use Visual Studio properly! You can set Visual Studio to break exactly when an exception is thrown. Here's official info on the subject.

I was actually unaware of the feature, so go give him the accepted answer. But please, don't go trying to handle exceptions in some funky way just to give yourself a hand debugging. All you do is open yourself up to more bugs.

Answer 3

If you're interested in the debugger simply stopping precisely where the exception occurred then have you considered first-chance exceptions?

If you open Tools|Exceptions then tick the Common Language Runtime Exceptions box, the debugger will stop at the point of exception regardless of any try/catch/finally blocks.

Update: You can specify the precise exception you wish to catch by expanding the [+] tree in the Exceptions dialog. Though of course it will fire every time any exception of the specified type[s] occur[s], you can switch it on and off at will even in the middle of a debugging session, so with judicious use of breakpoints you can get it to do your bidding. I used it successfully to get around the 'target of an invocation has thrown an exception' ball ache originating from using reflection to instantiate objects. Very useful tool in such circumstances. Also note the locals and stack trace should be firmly available as far as I recall (just did a quick test and they are available), so no problems there.

Of course if you want to log things then that is outside the scope of an IDE debugger; and in which case first-chance exceptions won't help you!

Give it a go at least; I found them very useful and they might be more appropriate for your issue than you think.

Answer 4

What's wrong with:

try
{
   MightThrow();
}
catch
{
   DoSomthingOnFailure();
   throw;
}

Answer 5

For code that should only run on exceptions, use the catch block:

try
{
   MightThrow();
}
catch (Exception ex)
{
   // this runs only when there was an exception
   DoSomthingOnFailure();
   // pass exception on to caller
   throw; 
}
finally
{
   // this runs everytime
   Cleanup();
}

Answer 6

This is what you want. It will only call this method when an error occurs, and the "throw" statement will re-throw the exception with the callstack intact.

try
{
   MightThrow();
}
catch
{
   DoSomthingOnFailure();
   throw;
}

Answer 7

A "finally" block that runs only on failure is called "catch" (with no parameters). :-)

Now, there is a small caveat. If you want to have a specialised "catch" case for a particular exception type and have a generic "catch" that works for all exceptions, you'll have to do a bit of a custom logic.

Thus, I would do something like:

  try
  {
    MightThrow();
  }
  catch(MyException ex)
  {
    // Runs on MyException
    MySpecificFailureHandler()
    // Since we have handled the exception and can't execute the generic
    // "catch" block below, we need to explicitly run the generic failure handler
    MyGenericFailureHandler()
  }
  catch
  {
    // Runs on any exception hot handled specifically before
    MyGenericFailureHandler()
    // If you want to mimic "finally" behavior and propagate the exception
    // up the call stack
    throw;
  }
  finally
  {
    // Runs on any failure or success
    MyGenericCleanupHandler();
  }

Answer 8

Every example so far is losing the original StackTrace according to my tests. Here's a solution that should work for you.

private static void PreserveStackTrace(Exception exception)
{
  MethodInfo preserveStackTrace = typeof(Exception).GetMethod("InternalPreserveStackTrace",
    BindingFlags.Instance | BindingFlags.NonPublic);
  preserveStackTrace.Invoke(exception, null);
}

try
{
   MightThrow();
}
catch (Exception ex)
{
    DoSomethingOnFailure();
    PreserveStackTrace(ex);
    throw;
}

Answer 9

How about only catching an exception that "MightThrow" does not throw?

Bool bad = true;
try
{
   MightThrow();
   bad = false;
}
catch (SomePrivateMadeUpException foo)
{ 
   //empty
}
finally
{
   if(bad) DoSomeLoggingOnFailure();   
}

Answer 10

Let me recap your requirements the way I understand them:

1. You want some code that is run only when an exception is generated, in order to do some logging.
2. You want to run your test framework under debugger and break at the point at which the exception is thrown.

To meet your first requirement, you should write the code the way everybody suggested - using parameterless catch and throw.

To meet your second requirement while using the parameterless catch, you could configure your debugger to break when an exception is throw, not only when there's an unhandled exception. I suspect you know how to do it, but I'll put it here for answer completeness: in VS you can do that in Debug -> Exception -> Common Language Runtime Exceptions -> check the Thrown checkbox.

If you know that your app throws a lot of handled exceptions, that might not be an option for you. At that point, your only choice left to meet your first requirement is to either write the code to use finally for exception logging purposes or look into the direct IL emitting route as Greg Beech suggests.

However, whether the finally code is being executed depends on the debugger you are using. In particular, VS will break on an unhadled exception before the finally is executed and will not let you continue. Thus, unless you detach from the process at that point, your logging code will never be executed. In other words, the second requirement will interfere with meeting the first requirement.

Answer 11

You could encapsulate your logic in a custom class, something like:

    public  class Executor
{
    private readonly Action mainActionDelegate;
    private readonly Action onFaultDelegate;

    public Executor(Action mainAction, Action onFault)
    {
        mainActionDelegate = mainAction;
        onFaultDelegate = onFault;
    }

    public  void Run()
    {
        bool bad = true;
        try
        {
            mainActionDelegate();
            bad = false;
        }
        finally
        {
            if(bad)
            {
                onFaultDelegate();
            }
        }
    }

}

And use it as:

            new Executor(MightThrow, DoSomeLoggingOnFailure).Run();

Hope this helps.

Answer 12

Isn't this the same as:

try 
{
    MightThrow();
}
catch (Exception e) 
{
    DoSomethingOnFailure();
    throw e;
}

?

Answer 13

You could write, or have someone write for you, a small assembly in VB.net which implements a TryFaultCatchFinally(of T) method that accepts four delegates:

1. TryMethod -- An Action(of T) to perform the "Try" block.
2. FaultMethod -- A Predicate(Of T, Exception) which, if an exception occurs, will be called before any "finally" blocks run; if it returns true the Catch block will run--otherwise it won't.
3. CatchMethod -- An Action(Of T, Exception) to be performed if an exception had occurred and FaultMethod returned true; happens after "finally" blocks run.
4. FinallyMethod -- An Action(OF T, Exception, Boolean) to be performed as a "Finally" block. The passed-in exception will be null if TryMethod ran to completion, or will hold the exception that caused it to exit. The Boolean will be true if the exception was caught, or false otherwise.

Note that when the FaultMethod is executed, one may be able to examine the state of objects that caused the exception, before such state is destroyed by Finally blocks. One must use some care when doing this (any locks that were held when the exception was thrown will continue to be held) but the ability may still sometimes be handy, especially when debugging.

I'd suggest the routine look something like:

    Shared Sub TryFaultCatchFinally(Of T)(ByVal TryProc As Action(Of T), _
                                          ByVal FaultProc As Func(Of T, Exception, Boolean), _
                                          ByVal CatchProc As Action(Of T, Exception), _
                                          ByVal FinallyProc As Action(Of T, Exception, Boolean), _
                                          ByVal Value As T)
        Dim theException As Exception = Nothing
        Dim exceptionCaught As Boolean = False
        Try
            TryProc(Value)
            theException = Nothing
            exceptionCaught = False
        Catch Ex As Exception When CopyExceptionAndReturnFalse(Ex, theException) OrElse FaultProc(Value, Ex)
            exceptionCaught = True
            CatchProc(Value, Ex)
        Finally
            FinallyProc(Value, theException, exceptionCaught)
        End Try
    End Sub

Answer 14

No, I think this is a common idiom the way you have it.

EDIT To be clear, the "catch" then "rethrow" strategies offer the same run-time semantics, however they change the experience when the VS debugger is attached. Tooling and maintenance is important; debugging often requires you to 'catch all first-chance exceptions' and if you end up with lots of 'spurious' first-chance exceptions due to catch-then-rethrow in your code, it really hurts the ability to debug the code. This idiom is about interacting well with the tooling, as well as clearly expressing the intent (you don't want to 'catch', decide can't handle, and rethrow, instead you just want to log that an exception did happen but let it pass on by).

Answer 15

Have you considered using the DebuggerStepThrough attribute? http://msdn.microsoft.com/en-us/library/system.diagnostics.debuggerstepthroughattribute.aspx

[DebuggerStepThrough]
internal void MyHelper(Action someCallback)
{
    try
    {
        someCallback();
    }
    catch(Exception ex)
    {
        // Debugger will not break here
        // because of the DebuggerStepThrough attribute
        DoSomething(ex);
        throw;
    }
}

Answer 16

try
{
   MightThrow();
}
catch
{
   DoSomethingOnFailure();
}