[Gmsh] New Mesh.BdfElementPID Option

[Geordie McBain]

On Fri, Jul 31, 2009 at 8:58 PM, Konstantinos
Poulios<logari81 at googlemail.com> 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