[omniORB] Injecting Enum Name/Values into Enum's Namespace

[Jeff Pitman]

On Wednesday 21 September 2005 21:20, Duncan Grisby wrote:
> > Inline is the only way i can access it in an OO way, it seems. Let
> > me know where I am messed up.
>
> I don't understand what you mean about inline and "an OO way". Can
> you explain?  inline only affects how many stub files you have, not
> the public interface you use to access them.

Hmm. Maybe something is not right with my setup. 

[jeff at kubik foo]$ omniidl -Cstubs -nc -nf -DSUN_ORB=true -bpython 
idl/5.5/*.idl; find stubs | wc -l
109

[jeff at kubik foo]$ omniidl -Cstubs -nc -nf -DSUN_ORB=true -bpython 
-Wbinline idl/5.5/*.idl; find stubs | wc -l
109

Maybe I'm not sending this stuff in right to omniidl.  With fnidl, I 
just had to send in an idl called "All.idl".  The "All.idl" would then 
#include a bunch of other external idls.  However, it appears that the 
omniidl does not recursively include these on parse and does not 
process them without ensuring they're all included on the commandline.

All.idl is like this:

module com {
        module example {
                module interfaces {
			#include "foo.idl"
			#include "bar.idl"
                };
        };
};

> If at all possible, you should avoid using the -nc flag that turns
> off checks about identifiers differing only in case. It permits
> illegal IDL which is not a good idea.

Yeah, sorry. I'm not in control of the IDL.  If I fix it, I'm not sure 
whether I'd damage access to the CORBA calls and whether it diverts 
from their documented "API". 

-- 
-jeff