We should implement a Triangulation::from_facets function, which computes all subentites on it's own and stores them.
This would then replace the need for computing the subentity relation (see #2), as it is automatically given by the creation of the subentities.
This would then give us a complete simplicial complex, in contrast to what gmsh might give us.
Gmsh only includes the simplicies that are directly or indirectly mentioned in the .geo file. ALl others are not listed as entities in the .msh file.
It's therefore (at this point) pretty pointless to get any subentities from gmsh. We should only get the facets from there and all then we rely on Triangulation::from_facets.
We might need to revert back to using the gmsh subentities at some point, because the indexing order or something else matters decided by gmsh matters to us.
We should implement a
Triangulation::from_facets
function, which computes all subentites on it's own and stores them. This would then replace the need for computing the subentity relation (see #2), as it is automatically given by the creation of the subentities.This would then give us a complete simplicial complex, in contrast to what gmsh might give us. Gmsh only includes the simplicies that are directly or indirectly mentioned in the
.geo
file. ALl others are not listed as entities in the.msh
file. It's therefore (at this point) pretty pointless to get any subentities from gmsh. We should only get the facets from there and all then we rely onTriangulation::from_facets
. We might need to revert back to using the gmsh subentities at some point, because the indexing order or something else matters decided by gmsh matters to us.