BUG: Fix superclass name in itkTypeMacro#3062
BUG: Fix superclass name in itkTypeMacro#3062dzenanz merged 2 commits intoInsightSoftwareConsortium:masterfrom
itkTypeMacro#3062Conversation
e2b0ea3 to
57686ff
Compare
Fix superclass name in `itkTypeMacro`.
57686ff to
6ea0640
Compare
Use the superclass name instead of the `Superclass` alias in `itkTypeMacro`.
6ea0640 to
0ee4882
Compare
| using OctreeType = itk::Octree<unsigned int, 16384, IdentityMap<unsigned int, 16384>>; | ||
| auto octree = OctreeType::New(); | ||
|
|
||
| // ITK_EXERCISE_BASIC_OBJECT_METHODS(octree, Octree, OctreeBase); |
There was a problem hiding this comment.
I have commented this. It seems to upset Windows because the test is telling that the superclass name is Object:
https://dev.azure.com/itkrobotwindow/ITK.Windows/_build/results?buildId=7506&view=logs&j=2d2b3007-3c5c-5840-9bb0-2b1ea49925f3&t=77aad734-2057-5694-9ae2-ee1f5f26eae8&l=55016
which should not to be the case; Object is the superclass of OctreeBase.
Both Linux and macOS seemed to be happy for the very same commit 9ed699b:
https://dev.azure.com/itkrobotlinux/ITK.Linux/_build/results?buildId=7524&view=results
https://dev.azure.com/itkrobotmacos/ITK.macOS/_build/results?buildId=7455&view=results
I do not see anything wrong, and it reminds me of other past cases that are yet to be solved (they were commented as they could not be solved after some investigation), e.g. #2286 (comment)
So I preferred to comment the statement here as well, as it seems to deserve a separate investigation.
Might be loosely linked to #3008, although the latter was mainly opened due to the FFT backend name of class and GCC issues.
There was a problem hiding this comment.
Another one I found yesterday is itk::StdStreamLogOutput, whose parent class is itk::LogOutput; when trying to exercise basic object methods ITK_EXERCISE_BASIC_OBJECT_METHODS, it expects itk::Object as the superclass, which is itk::LogOutput's parent. Not sure if all these are somehow related to symbol exportation macros, i.e. ITKCommon_EXPORT. Just guessing.
|
For the records, after testing, the changes in 0ee4882 did not belong to the |
itkTypeMacroitkTypeMacroPR Checklist