[ITK] New Bruker module not exposed correctly

[Matt McCormick]

Hi Toby,

Thanks again for the contribution.

Yes, a small follow-up patch in UseITK.cmake will do the trick. For a
large addition like this, some follow-up effort is expected.

In my opinion, "2dseq" is obscure enough that the addition of "Bruker"
helps identify the class.

Thanks,
Matt

On Wed, Sep 20, 2017 at 8:43 AM, Wood, Tobias <tobias.wood at kcl.ac.uk> wrote:
> Hello,
>
> I updated my local copy of ITK to master to test my Bruker patch that was merged a couple of weeks ago. Unfortunately my patch did not change the UseITK.cmake to follow the new name of the module (it was suggested to me that the module name be shortened in case of anymore Bruker formats), so the Bruker reader is not correctly registered.
>
> I think I have worked out the required changes to UseITK.cmake, so can submit a patch. However, this raises the question of whether the naming scheme could be rationalized further. Before, both module and class were called “Bruker2DSEQ”. The module is now “Bruker”, and the class is “Bruker2dseq”, which is the correct capitalisation for the format. Should the class/format be shortened to just “2dseq”?
>
> I’m sorry not to catch this before my original patch was merged. Changing the name was one of the last things I did, and because the tests explicitly register the ImageIOFactory, they would not pick this up.
>
> Best wishes,
> Toby
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/mailman/listinfo/community