[Insight-developers] Compile time errors for umimplemented methods

[Rupert Brooks]

Hi,

Is this related to the conversation about this change?

 http://review.source.kitware.com/2087

If not the two should be aware of each other.

Rupert



--------------------------------------------------------------
Rupert Brooks
rupert.brooks at gmail.com




On Tue, Jul 19, 2011 at 16:09, Bill Lorensen <bill.lorensen at gmail.com> wrote:
> You may be stuck with runtime exceptions. Unless you can rearrange the
> class hierarchy and add another base class.
>
> It's hard to solve with limited info. e-mail is not so good for this.
> I suggest a few of you get in a room and figure it out.
>
>
> On Tue, Jul 19, 2011 at 4:06 PM, Jeffrey Duda <jtduda at seas.upenn.edu> wrote:
>> Right. I now realize that my question was incomplete. What we want is to
>> have these methods be pure virtual in the base class, implement them in the
>> derived classes for which we know how to implement them and have other
>> derived classes, that we are still thinking about how to implement
>> correctly, throw compile-time errors if the method is referenced. For now we
>> could put in run time exceptions for these "in-progress" methods, but we
>> were hoping to avoid that since it seems less informative for the user.
>> -jeff
>>
>> On Tue, Jul 19, 2011 at 3:17 PM, Bill Lorensen <bill.lorensen at gmail.com>
>> wrote:
>>>
>>> You could define those required methods as pure virtual in the base
>>> class. This would require an implementer to think about what to do in
>>> the derived class.
>>>
>>> Bill
>>>
>>> On Tue, Jul 19, 2011 at 2:16 PM, Jeffrey Duda <jtduda at seas.upenn.edu>
>>> wrote:
>>> > Hi,
>>> > I've been working on some of the changes related to Transforms and was
>>> > hoping for some guidance on how to best deal with umimplemented methods
>>> > in
>>> > derived classes. This is best described with an example:
>>> > in the Transform base class we have two methods for transforming a
>>> > vector:
>>> > 1 - TransformVector( vector )
>>> > 2 - TransformVector( vector, point )
>>> > The second is needed for transforms that vary spatially (i.e.
>>> > deformation
>>> > fields). For linear transforms, the point is simply ignored since it is
>>> > unnecessary, however since the point is required for a
>>> > DeformationFieldTransform, the implementation of the first method
>>> > currently
>>> > throws an exception at run time. I was wondering if/how we could have
>>> > this
>>> > result in a compile-time error instead.
>>> > -jeff
>>> > _______________________________________________
>>> > Powered by www.kitware.com
>>> >
>>> > Visit other Kitware open-source projects at
>>> > http://www.kitware.com/opensource/opensource.html
>>> >
>>> > Kitware offers ITK Training Courses, for more information visit:
>>> > http://kitware.com/products/protraining.html
>>> >
>>> > Please keep messages on-topic and check the ITK FAQ at:
>>> > http://www.itk.org/Wiki/ITK_FAQ
>>> >
>>> > Follow this link to subscribe/unsubscribe:
>>> > http://www.itk.org/mailman/listinfo/insight-developers
>>> >
>>> >
>>
>>
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://kitware.com/products/protraining.html
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>