Clean the output of the vtkMaterialInterface filter
requested to merge 66-crash-de-themys-lors-de-l-application-d-un-filter-de-reflection-apres-application-du-filtre-d into main
Application of a Reflect
filter (or a PlotOverLine
filter) after the vtkMaterialInterface
may lead to:
- wrong result (with the https://gitlab.com/themys/readers/-/blob/469c8403939def6608bc6c0d7c409677954bbdc4/Plugin/Testing/Data/Interfaces/euler_1proc/HDep-n=Temps_u=s.p-0001 database)
- crash (with the https://gitlab.com/themys/readers/-/blob/469c8403939def6608bc6c0d7c409677954bbdc4/Plugin/Testing/Data/Interfaces/ale_8procs/HDep-n=Temps_u=s+n=NumSDom_g=0000x0007.p-0001 database)
This MR fixes this by applying a vtkCleanPolyData
filter at the end of the vtkMaterialInterface
algorithm.
Two things are to be noticed:
- the case where the output of the
vtkMaterialInterface
is avtkUnstructuredGrid
is not handled yet because we don't know if it is problematic too - no test can be added right here to check this fix because the problems described above do not occur with the vtm/vtu databases of this repo. They only appear when manipulating original Hercule databases. That's why another MR (globalintegrationtests!19 (merged)) is planned in globalintegrationtest to add a test that check this fix while working on a Hercule database.
Closes #66 (closed) Closes #64 (closed)
Edited by guillaume Peillex