[cctbxbb] sginfo and sgtbx

[Pascal]

On 15/10/16 19:31, Luc Bourhis wrote:
>
>> On 15 Oct 2016, at 15:05, Pascal <pascal22p at parois.net
>> <mailto:pascal22p at parois.net>> wrote:
>>
>> Crystals is lacking the capability to get the Hall symbol and space
>> group from symmetry operators. sginfo works fine but no longer
>> maintained and I am wondering if a better solution exists.
>>
>> Another concern when looking at cctbx is it is written in C++. It  is
>> really a hassle as I would need to write C wrapper that I could access
>> in Fortran.
>
> Writing a C function taking mere strings and returning a mere strings,
> and whose body creates cctbx C++ objects and calls their methods, that
> is not that difficult to write, is it? Or am I missing something?

No, that should be it.
More complicated objects could be flatten as a C struct and then pass to 
fortran.

>
> What I would try to do would be to compile the cctbx with its native
> build system. Then compile your C function, something like
> g++ -I path/to/cctbx_project ...
> on Linux so that it finds the sgtbx headers, where path/to/cctbx_project
> is the path to the root of the cctbx source code. Then copy the cctbx
> shared library to your build directory, you will find it at
> cctbx_build/lib/libcctbx.so, and then link your object against
> libcctbx.so. You would then ship libcctbx.so within Crystals. If you
> prefer, you can configure the cctbx build system to produce static
> libraries if you don’t want to ship shared libraries with Crystals.
> Mmhmmm… I have the feeling I am forgetting another library to also copy
> from cctbx_build/lib…

Ok, if it just libcctbx and maybe an other one it's not too bad.

Thanks,
Pascal

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5985 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://phenix-online.org/pipermail/cctbxbb/attachments/20161021/d71e700f/attachment.p7s>