[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


JD Smith wrote:
>... Unfortunately, this implements, albeit in a
> more manageable and less ugly way, pretty much what I'd already
> accomplished with macros.  That is, each and every of the functions
> BYTE_foo, INT_foo, ULONG_foo, etc., get compiled and included in the
> executeable separately, and you essentially choose among them with the
> run-time type information.  This works, but leads to extreme code bloat
> if you're replicating a large function 10's of times.

   Unless RSI has a Wonderland solution, they probably do the same.
Just a "reverse" idea: Looking at one of your recent examples, it
seemed as if all the code but the definition of the out array was type
independent. So, why not put that code in a macro? Something like

if(maxQ) {switch( type ) { 
 case IDL_TYP_BYTE: 
     UCHAR *tin,*tout,tmp;
     tout=( UCHAR *)out;
     tin=( UCHAR *)arg[0]->value.arr->data;
#include max_fun.h
 case IDL_TYP_INT: 
     short *tin,*tout,tmp;
     tout=( short *)out;
     tin =( short *)arg[0]->value.arr->data ;
#include max_fun.h

Doesn't solve the problem but may make the code easier to read and
maintain (just imagine you want to fix a bug in 10 copies of almost
the same code)



[[ Dr. Martin Schultz   Max-Planck-Institut fuer Meteorologie    [[
[[                      Bundesstr. 55, 20146 Hamburg             [[
[[                      phone: +49 40 41173-308                  [[
[[                      fax:   +49 40 41173-298                  [[
[[ martin.schultz@dkrz.de                                        [[