Re: Ideal BDD using JMock

Which components are causing you trouble?  Most Swing components
interact with their model through a model interface and listener
interfaces.  What are you trying to mock out?

--Nat

On 27/09/2007, Tom Lanning <tom.lanning@...> wrote:
>    I do a lot of Swing-based development and routinely delegate the
> behavior of Swing components using object injection.  Unfortunately,
> many of the Swing components specify (abstract) delegate objects
> instead of interfaces.  Therefore I routinely need to subclass
> unalterable provided (abstract) classes.  I know I can use JMock with
> ClassImposteriser but I often end up mocking most of the base class'
> behavior to verify my over-ridden methods.
>
>    How do "interface-only purists" handle this problem?  (I don't
> mean any offense by the term "interface-only purist".  I am becoming
> one myself; hence the question.)
>
>
> ---------------------------------------------------------------------
> To unsubscribe from this list please visit:
>
>     http://xircles.codehaus.org/manage_email
>
>

---------------------------------------------------------------------
To unsubscribe from this list please visit:

Re: 2.3 release?

I pushed the 2.3.0 release this morning.

--Nat

On 28/09/2007, Erick Dovale <edovale@...> wrote:
> Folks,
> Is there a date for the release of 2.3?? What's holding it?
>
> cheers
>
> Erick.
>

---------------------------------------------------------------------
To unsubscribe from this list please visit:

    http://xircles.codehaus.org/manage_email

RE: 2.3 release?

Hi Nat,

Is there anywhere I can see what is new in this release?

Thanks for your help!
Ash

> -----Original Message-----
> From: Nat Pryce [mailto:nat.pryce@...] 
> Sent: 01 October 2007 08:38
> To: user@...
> Subject: Re: [jmock-user] 2.3 release?
> 
> I pushed the 2.3.0 release this morning.
> 
> --Nat
> 

---------------------------------------------------------------------
To unsubscribe from this list please visit:

    http://xircles.codehaus.org/manage_email

jMock2 and two expectations on same method

Re: jMock2 and two expectations on same method

On 1 Oct 2007, at 08:50, marco.mistroni@... wrote:
> public void retrieveData() {
>
>        List list = businessManager.retrieveData();
>        typeManager.populateListWithTypes(list);
> }
>
> [...]
> i'll have to do same for typeManagerMock.... is this code correct?
>
> context.checking(new Expectations() {{
>     one (businessManagerMock).retrieveData();
>         will(returnValue(myTestList));
>     one(typeManagerMock).populateList(myTestList);
> }});

that should be

   one(typeManagerMock).populateListWithTypes(myTestList)

but yes, if that's the behaviour you want to test.

If you're in Paribas in London, talk to the Sierra team. They should  
have a lot of experience with Mocks.

S.

Steve Freeman
Winner of the Agile Alliance Gordon Pask award 2006

http://www.m3p.co.uk

M3P Limited.
Registered office. 2 Church Street, Burnham, Bucks, SL1 7HZ.
Company registered in England & Wales. Number 03689627

---------------------------------------------------------------------
To unsubscribe from this list please visit:

    http://xircles.codehaus.org/manage_email

Newbie: how do you make a test pass?

One thing keeps puzzling me: when you've specified a test with expectations 
using Mock Objects, then how do you proceed to make the test pass? Following 
the Getting Started with the Publisher and the Subscriber: how do you create a 
Subscriber in the Publisher.publish() method? You don't use jMock stuff here, 
do you? So I'd guess you have to create some implementation of Subscriber 
anyway, and you could use that same implementation in the test case 
alltogether, and what does jMock bring me, then?

Regards, Wouter.

---------------------------------------------------------------------
To unsubscribe from this list please visit:

    http://xircles.codehaus.org/manage_email

Re: Newbie: how do you make a test pass?

I think if you take another look you'll find that the subscriber is  
created outside and then passed to the publisher. We use jMock to  
help us think about the relationships between objects, so what we're  
doing in the example is figuring out what the publisher needs from  
its neighbours.

Of course, sometime later we'd create an implementation of a  
subscriber and test that separately (and together in an integration  
test).

S.

On 1 Oct 2007, at 09:26, Wouter Scherphof wrote:
> One thing keeps puzzling me: when you've specified a test with  
> expectations
> using Mock Objects, then how do you proceed to make the test pass?  
> Following
> the Getting Started with the Publisher and the Subscriber: how do  
> you create a
> Subscriber in the Publisher.publish() method? You don't use jMock  
> stuff here,
> do you? So I'd guess you have to create some implementation of  
> Subscriber
> anyway, and you could use that same implementation in the test case
> alltogether, and what does jMock bring me, then?

Steve Freeman
Winner of the Agile Alliance Gordon Pask award 2006

http://www.m3p.co.uk

M3P Limited.
Registered office. 2 Church Street, Burnham, Bucks, SL1 7HZ.
Company registered in England & Wales. Number 03689627

---------------------------------------------------------------------
To unsubscribe from this list please visit:

    http://xircles.codehaus.org/manage_email

Re: 2.3 release?

Re: 2.3 release?

Re: Ideal BDD using JMock

   Thanks Steve and Nat for your interest.

   There are several variations of this problem, but the an "easy  
one" to describe is the use of TableCellRenderer (http://java.sun.com/ 
javase/6/docs/api/javax/swing/table/TableCellRenderer.html).

   TableCellRenderer describes an interface that returns a properly  
configured Component that a JTable uses to "paint" its cells.   
Although the table cell renderer is specified as an interface, two  
important collaborators, the component and the table, are not.  From  
a JMock point-of-view I want to ensure that appropriate query methods  
are invoked on the JTable and appropriate configuration methods are  
invoked on the Component.  Notice that like all Swing components,  
JTable and the relatively abstract Component, are not defined as  
interfaces.  Since the collaborators are declared as objects I am  
forced to mock objects not interfaces.

   Another example is a group of classes used to implement drag-and- 
drop behavior on JComponents.  The class TransferHandler (http:// 
java.sun.com/javase/6/docs/api/javax/swing/TransferHandler.html)  
collaborates with a variety of objects that are not specified as  
interfaces.  Aside from JComponent, it uses the TransferSupport  
class, a family of classes descended from DropLocation, and the  
DataFlavor class.  I have found it very difficult to test (specify)  
the behavior drag-and-drop classes and I don't think that even what I  
have accomplished is possible without mocking objects.

   Again thanks for taking time to think about these problems.  Any  
insight is welcome.

On Sep 30, 2007, at 11:47 PM, Nat Pryce wrote:

> Which components are causing you trouble?  Most Swing components
> interact with their model through a model interface and listener
> interfaces.  What are you trying to mock out?
>
> --Nat
>
> On 27/09/2007, Tom Lanning <tom.lanning@...> wrote:
>>    I do a lot of Swing-based development and routinely delegate the
>> behavior of Swing components using object injection.  Unfortunately,
>> many of the Swing components specify (abstract) delegate objects
>> instead of interfaces.  Therefore I routinely need to subclass
>> unalterable provided (abstract) classes.  I know I can use JMock with
>> ClassImposteriser but I often end up mocking most of the base class'
>> behavior to verify my over-ridden methods.
>>
>>    How do "interface-only purists" handle this problem?  (I don't
>> mean any offense by the term "interface-only purist".  I am becoming
>> one myself; hence the question.)
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe from this list please visit:
>>
>>     http://xircles.codehaus.org/manage_email
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe from this list please visit:
>
>     http://xircles.codehaus.org/manage_email
>

---------------------------------------------------------------------
To unsubscribe from this list please visit:

    http://xircles.codehaus.org/manage_email