Dan,

Here's the soggy history of this:

Step 1: define a base class @Before that protocols with derived classes
to read in some beans.

Step 2: Get completely hosed by the engine lifecycle department.

Step 3: Get the idea that all the @Tests in the class could really just
share a set of bean-defined clients and services, they don't need them
to be reconstructed for each @Test. 

Step 4: discover that the object is constructed over and over for each
@Test, so setting up in a constructor won't work.

Step 5: resort to a static boolean.

If we had the engine lifecycle set up so that a Before/After class could
start/stop http services reliably, then I could, as it were, go back to
step 1.

 

> -----Original Message-----
> From: Daniel Kulp [mailto:[EMAIL PROTECTED] 
> Sent: Tuesday, November 20, 2007 4:34 PM
> To: [email protected]
> Cc: Benson Margulies
> Subject: Re: Junit 4 versus us
> 
> 
> Umm...  
> 
> Couldn't it be in a @BeforeClass annotated method?   Why have 
> the static 
> things?
> 
> Dan
> 
> 
> On Monday 19 November 2007, Benson Margulies wrote:
> > And if one wants a protocol using overriden functions, one 
> has to have 
> > an @Before that uses a static boolean to make it run once.
> >
> > > -----Original Message-----
> > > From: Christopher Moesel [mailto:[EMAIL PROTECTED]
> > > Sent: Monday, November 19, 2007 7:26 AM
> > > To: [email protected]
> > > Subject: RE: Junit 4 versus us
> > >
> > > Yes, I've found that any code I want only to run once (per test 
> > > class) needs to go in a static block outside the 
> constructor.  That 
> > > way it only gets processed the first time the JVM loads the class.
> > >
> > > -Chris
> > >
> > > -----Original Message-----
> > > From: Benson Margulies [mailto:[EMAIL PROTECTED]
> > > Sent: Saturday, November 17, 2007 1:02 PM
> > > To: [email protected]
> > > Subject: Junit 4 versus us
> > >
> > >
> > > We have a fair number of junit tests that have code in their 
> > > constructor that should only be run once.
> > >
> > > Imagine my surprise to discover that this is really not a 
> good idea, 
> > > since JUnit reconstructs for each test function.
> 
> 
> 
> --
> J. Daniel Kulp
> Principal Engineer
> IONA
> P: 781-902-8727    C: 508-380-7194
> [EMAIL PROTECTED]
> http://www.dankulp.com/blog
> 

Reply via email to