testCellGeometry is failing on macOS and linux

[dcoeurjo]

An assert is raised when running the testCellGeometry on macOS (master branch)

 [ERR] Assertion Error - assertion (cIsInside( kp )) failed in typename DGtal::KhalimskySpaceND<dim, TInteger>::Cell     DGtal::KhalimskySpaceND<2>::uCell(const Point &) const [dim = 2, TInteger = int]:     /Users/davidcoeurjolly/Sources/DGtal/src/DGtal/topology/KhalimskySpaceND.ih(725)

It seems to be related to addCellsTouchingPolytope. Can you please have a look, @JacquesOlivierLachaud ?

[dcoeurjo]

The issue is raised from the getIntersectedKPoints method l689 of CellGeometry.ih.

Could you please have a look?

[dcoeurjo]

Also failing on linux

[dcoeurjo]

[ERR] Assertion Error - assertion (cIsInside( kp )) failed in DGtal::KhalimskySpaceND<dim, TInteger>::Cell DGtal::KhalimskySpaceND<dim, TInteger>::uCell(const Point&) const [with unsigned int dim = 2; TInteger = int; DGtal::KhalimskySpaceND<dim, TInteger>::Cell = DGtal::KhalimskyCell<2, int>; DGtal::KhalimskySpaceND<dim, TInteger>::Point = DGtal::PointVector<2, int, std::array<int, 2> >]: /home/davidcoeurjolly/Sources/DGtal/src/DGtal/topology/KhalimskySpaceND.ih(725)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
testCellGeometry is a Catch v2.13.7 host application.
Run with -? for options

-------------------------------------------------------------------------------
Scenario: CellGeometry< Z2 > rational intersections
      Given: A rational simplex P={ Point(0/4,0/4), Point(17/4,8/4), Point(-5/
             4,15/4) }
-------------------------------------------------------------------------------
/home/davidcoeurjolly/Sources/DGtal/tests/geometry/volumes/testCellGeometry.cpp:340
...............................................................................

/home/davidcoeurjolly/Sources/DGtal/tests/geometry/volumes/testCellGeometry.cpp:340: FAILED:
due to a fatal error condition:
  SIGABRT - Abort (abnormal termination) signal

===============================================================================
test cases:  7 |  6 passed | 1 failed
assertions: 50 | 49 passed | 1 failed

[JacquesOlivierLachaud]

I will have a closer look, but I think it is related to BoundedRationalPolytope, with which I am not confident anymore. I wonder if I should not deprecate it, since nobody uses it.

[dcoeurjo]

Yeah that is strange because the test should had passed at some point

[dcoeurjo]

Same kind of issue for testDigitalConvexity