[gmx-developers] Re: C++ Topology Generation

[Roland Schulz]

Hi,

currently all core developers are very busy with the 4.6 release which
will be released this summer. After that it is likely that more
resources are available for the C++ transition. But I don't know when
resources are available for a C++ topology module. Any useful
contribution are always very welcome. If you want to propose a design
for the class(es) you can post it to redmine.gromacs.org. If you want
to contribute code you can do so at gerrit.gromacs.org.

Roland

On Tue, Jul 24, 2012 at 12:13 AM, Andy Somogyi <andy.somogyi at gmail.com> wrote:
> err, I meant gmx_mtop_t, not topo_mol.
>
> At the least, I'd like to create a class that consolidates all the
> gmx_mtop_t read / write / access / modify functions.
>
>
> On Jul 23, 2012, at 9:55 PM, Andy Somogyi wrote:
>
> Hi,
>
> I've looked at the C++ transition for topology generation:
> http://www.gromacs.org/index.php?title=Developer_Zone/C%2b%2b_Transition/Topology_generation_modules
>
> Are there any set plans for c++ topology generation?
>
> Reason I ask is I've been working on a topology editing module that use the
> guts of grompp to fill read a topo_mol struct. The topo_mol is basically the
> core of a python extension, and this can write a modified topology out to
> .top file.
>
> I'm also working on another project (fast multipole) which also needs this
> topology and charge information.
>
> So, basically, what I'm thinking about is basically a set of C++ classes
> that wrap a topo_mol, can populate it via either the args to grompp, or from
> a binary .tpr. Essentially, this same class could do the topology generation
> / editing, and also be used in mdrun for reading the .tpr
>
> The project I'm working on could really use a class like this.
>
> So, any thoughts on the usefulness of such a class(s)?
>
>
>
>
>



-- 
ORNL/UT Center for Molecular Biophysics cmb.ornl.gov
865-241-1537, ORNL PO BOX 2008 MS6309