Re: [Libmesh-users] GmshIO class additional functionality

[Vijay S. M. <vi...@gm...>]

Hmm.. Interesting. My sent mail shows 3 files !

I am sending the files again. This time, I'm zipping them all to 1
file. Let me know if you get them this time.

Note: Rename to .zip before unzipping.

Vijay

On Fri, Apr 11, 2008 at 4:14 PM, John Peterson <jwp...@gm...> wrote:
> Hi Vijay,
>
>  The only one that made it through for me was meshex.C.
>
>  -J
>
>
>
>  On Fri, Apr 11, 2008 at 4:12 PM, Vijay S. Mahadevan <vi...@gm...> wrote:
>  > John,
>  >
>  >  I sent you 3 files:
>  >
>  >  1) The patch file with modifications in GmshIO.C, GmshIO.h, and
>  >  UnstructuredMesh.C.
>  >  2) The test code (meshex.C) for GmshIO read and write
>  >  3) The sample .msh file : 2dmultiregion.msh
>  >
>  >  I hope you have received these files. If not, let me know and I'll
>  >  send them again.
>  >
>  >  After applying the modifications, just compile meshex.C with a libMesh
>  >  makefile and run as
>  >
>  >  ./meshex-opt -d 2 2dmultiregion.msh 2dmultiout.msh
>  >
>  >  The idea is that, all the attributes read from 2dmultiregion.msh is
>  >  stored and written back to 2dmultiout.msh. I have tested this with a
>  >  few .msh files and the code works correctly. Please let me know if
>  >  that helps and if you need another .msh file.
>  >
>  >  Vijay
>  >
>  >
>  >
>  >  On Fri, Apr 11, 2008 at 3:57 PM, John Peterson <jwp...@gm...> wrote:
>  >  > Hi,
>  >  >
>  >  >  Did you attach the library modifications as well, or just the test
>  >  >  code?  Also, I'm not sure about others, but I don't really have a
>  >  >  simple .msh file handy so if you could attach one or email it
>  >  >  separately if it's too large that would help us out as well.
>  >  >
>  >  >  Thanks,
>  >  >  John
>  >  >
>  >  >
>  >  >
>  >  >  On Fri, Apr 11, 2008 at 2:26 PM, Vijay S. Mahadevan <vi...@gm...> wrote:
>  >  >  > Hi,
>  >  >  >
>  >  >  >  I have been using Gmsh to generate my multi-dimensional mesh files and
>  >  >  >  was delighted to find that libMesh supports the format. I eventually
>  >  >  >  realized though that the material numbers were not accounted for or
>  >  >  >  stored anywhere even though specified in the input msh file. And I had
>  >  >  >  to dig into the code and eventually modified the GmshIO class to
>  >  >  >  include these data and store it in a MeshData object based on the same
>  >  >  >  concepts from UNVIO class.
>  >  >  >
>  >  >  >  I have attached the patch and a sample test code along with the mail.
>  >  >  >  I would appreciate it if you can include this code in the current
>  >  >  >  developer version so that I do not have to maintain this separately
>  >  >  >  and worry about losing changes every time I update my codebase. Please
>  >  >  >  do verify that there are no bugs in the modifications also. Also, I
>  >  >  >  hope that other people will find this functionality useful since the
>  >  >  >  .unv file generation from Gmsh has its own issues and this IMO is a
>  >  >  >  little cleaner way of passing physical and elementary data to the end
>  >  >  >  user along with arbitrary number of attributes/element.
>  >  >  >
>  >  >  >  I am also interested in knowing how others have been passing the
>  >  >  >  material data from the mesh to the code for multi-region problems. Do
>  >  >  >  let me know if you have any comments.
>  >  >  >
>  >  >  >  Cheers,
>  >  >  >  Vijay
>  >  >  >
>  >  >  > -------------------------------------------------------------------------
>  >  >  >  This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
>  >  >  >  Don't miss this year's exciting event. There's still time to save $100.
>  >  >  >  Use priority code J8TL2D2.
>  >  >  >  http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
>  >  >  > _______________________________________________
>  >  >  >  Libmesh-users mailing list
>  >  >  >  Lib...@li...
>  >  >  >  https://lists.sourceforge.net/lists/listinfo/libmesh-users
>  >  >  >
>  >  >  >
>  >  >
>  >
>