Re: Expected Unhandled Exceptions

This is a knotty problem and goes a bit beyond what most would call unit
testing.
 
If I understand rightly, you want to test that
1) An exception is thrown,
2) That it is caught by the unhandled exception handler.
 
Since the test is passing, I assume that you are creating a thread - or your
SUT
is creating one. I say that because it's impoissilbe to have an unhandled
exception
in the test without creating a thread, since NUnit handles all exceptions on
the
test thread.
 
I was surprised to hear that the exception was being handled correctly by
your handler,
since I was under the impression that an unhandled exception handler had to
be 
in the primary appdomain to work. It turns out - as you probably know - that
MS
changed that beginning with 2.0.
 
Per the documentation for AppDomain.UnhandledExceptionEvent, your unhandled
exception should be reported to a handler in the domain where the thread was
created and /also/ to the handler in the primary AppDomain, which is created
by NUnit. 
 
That's exactly what you seem to be seeing. Your handler takes care of
things,
so the test passes, but NUnit's handler also gets the exception. Since NUnit
has no way to know that you already took care of the exception, it reports
it.
 
Unfortunately, I don't see any way around this, but I'm open to suggestions.
 
Charlie
 
 


  _____  

From: [email protected]
[mailto:[email protected]] On Behalf Of Wade Barnes
Sent: Friday, July 11, 2008 9:20 AM
To: [email protected]
Subject: [Nunit-users] Expected Unhandled Exceptions



Hi, 

Is there any way to tell a particular test to expect an unhandled exception;
one that would normally be indicated through the
AppDomain.UnhandledException event?



I am testing code that uses AppDomains to isolate code.  One of the tests
purposely causes an unhandled exception in one of these AppDomains to ensure
it is properly reported and acted upon by the code managing the AppDomains;
therefore I am expecting an unhandled exception to occur. 

The issue:

The test itself passes. 

However,

When using the GUI runner the test bar goes Red and the unhandled exception
is reported in the output list.

When using the Console runner this same unhandled exception causes the
runner to return a code of -100; this ends up failing our nightly build.



What I would like to do:

I would like to be able to indicate, for a particular test, that an
unhandled exception is expected.  I would then expect the test to behave
similarly to any other test expecting an exception; if received it happy
goes about the remained of its testing business, otherwise it fails because
it did not receive the expected exception.  Additionally when such a test is
encountered I would expect the test bar in the GUI not to turn Red and the
unhandled exception not to be reported in the output list; similarly I would
expect the Console runner not to return -100.



Wade

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08

_______________________________________________
Nunit-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/nunit-users