A strange CGAL error

Hello everyone!I try to use CGAL to deal with OFF model and then I find this strange error:

WARNING: Geometry is not solid, will make it fat. ->
CGAL error: assertion violation!
Expression : !GM0.empty() && !GM1.empty()
File : C:\dev\CGAL-4.13\include\CGAL/Minkowski_sum_3/bipartite_nary_union_sorted_combined.h
Line : 211
Explanation: one operand must be full-dimensional
Refer to the bug-reporting instructions at https://www.cgal.org/bug_report.html

I tried 10 different models, and 6 of them gave the same error message. So I want to consult experienced developers how to solve this problem.

Tagged:

Comments

  • @shanmama if this is specific to IfcOpenShell's CGAL branch, you might get better results creating a bugreport or pinging @aothms . I'm not sure who here are experienced in CGAL.

    shanmama
  • @Moult said:
    @shanmama if this is specific to IfcOpenShell's CGAL branch, you might get better results creating a bugreport or pinging @aothms . I'm not sure who here are experienced in CGAL.

    Thank you very much for your help, and congratulations again on your achievement!

    Moult
  • This appears to be related to https://github.com/tudelft3d/ifc2citygml/blob/master/off2citygml/FileIO.cpp#L114

    I can't provide support for this code, but I'm working on a replacement that will be open source soon.

    Maybe check the return value of sstream2nef() here https://github.com/tudelft3d/ifc2citygml/blob/master/off2citygml/FileIO.cpp#L100 and if false goto next()?

    Also, but not related to this. In that code base every invocation of regularization() needs to be replaced with extract_regularization() as the first is a non-void const function. So it doesn't make sense to call it without capturing the return value.

  • Or, maybe in addition to that. Add a call to is_empty() and also go to next() if true

  • @aothms said:
    Or, maybe in addition to that. Add a call to is_empty() and also go to next() if true

    Fine. It is a "disrepaired" project, and many resources cannot be found. I am trying to rewrite this geometric algorithm.

  • I am trying to rewrite this geometric algorithm.

    Like I said, I'm working on a newer version of the code directly on IFC. I can probably add you to the repo given that it'll be open source soon. Send me a message with your Github username

  • @aothms said:

    I am trying to rewrite this geometric algorithm.

    Like I said, I'm working on a newer version of the code directly on IFC. I can probably add you to the repo given that it'll be open source soon. Send me a message with your Github username

    OK, my username is "keleisi2001"

Sign In or Register to comment.