[Boost-users] MPL preprocessed question

From: Neal Meyer (nmeyer_at_[hidden])
Date: 2008-04-30 00:54:15

Alright I've been through the mailing list a few times, and thanks to
Ovanes for the help on getting some of the mpl stuff working.  Here's
what I've got working so far.

I built a vector100.hpp and placed it in the boost\mpl\vector and
added the following to my header.

#define BOOST_MPL_CFG_NO_PREPROCESSED_HEADERS
#define BOOST_MPL_LIMIT_VECTOR_SIZE 100
#include "boost/mpl/vector.hpp"
#undef BOOST_MPL_CFG_NO_PREPROCESSED_HEADERS
#undef BOOST_MPL_LIMIT_VECTOR_SIZE

Great everything works great, except for the fact that it takes an
extra long time to compile.  So following some threads on the list,
I'd like to pregenerate a header file with the definitions for the
vector that takes 100 elements.

Placed the vector100.cpp in the \libs\mpl\preprocessed\vector and ran
the preproces.py and got the vector100.hpp file generated in
boost\mpl\vector\aux_\preprocessed\plain.

Everthing seems like I'm heading in the right direction to get it
proprocessed.  So know what do I do to get it included in the same way
as the previous code?  I've tried including the new header directly
and playing around with various macros to get it to be included, but
no dice so far.  Is there a way to have the preprocessing scripts
regenerate the vector.hpp in boost\mpl\aux_\preprocessed\plain?  So
that I can use boost::mpl::vector instead of a version with the number
of parameters in it, because that's varies dramatically throughout how
this code is being used and would be a major pain.

-Neal