On Fri, Jul 31, 2009 at 8:58 PM, Konstantinos Poulios<[email protected]> wrote: > Hi all! > > As an attachment I sent a patch adding a new option for the BDF file format. > Up to now the property id of elements exported in the BDF format corresponds > to id of the elementary entity to which the element belongs. In order to add > some flexibility to the translation of the gmsh native entities organization > to the bdf format, I added the option Mesh.BdfElementPID. I consider the > adoption of such an option very important, since it will help essentially > the definition of element groups in the exported bdf-file. For the option > values I followed the convention by the Mesh.ColorCarousel option: > > [...]
Hey, this is great. I've been meaning to do the same thing for the MEDIT .mesh export; mimicking this will make that much easier for me. Thank you. I agree that the current output methods for MEDIT and BDF (and probably some other formats?) don't seem to be respecting the principle set forth in `6.2 Elementary vs. physical entities' <http://geuz.org/gmsh/doc/texinfo/gmsh_7.html#SEC42>. I suppose this is because those formats, unlike Gmsh's .msh, only allow one physical entity per element? Anyway, I think that by far the most common use case is one physical entity per element; at least with your modification, the user has the option of getting the first physical entity rather than having them all ignored and discarded as happens at present. By the way, I note that there are a few white-space-only changes in your patch. Thanks again, and I vote for the inclusion of this functionality. As I say I will be mimicking this in my personal Gmsh installation for MEDIT .mesh output, and will try and submit a similar patch for that format if yours is accepted. Best wishes, Geordie McBain _______________________________________________ gmsh mailing list [email protected] http://www.geuz.org/mailman/listinfo/gmsh
