[omniORB] IDL Case

[Richard Hardgrave]

> From owner-omniorb-list@uk.research.att.com Tue Jun 19 12:27 CDT 2001
> 
> On Tue, Jun 19, 2001 at 11:58:48AM -0500, Stephen P. White wrote:
> > I have old IDL code which was written with Fnorb.   We would like to
> > convert to omniORBpy but have a case issue.  Under Fnorb the IDL shown
> > below was legal.
> > 
> > module m {
> >  interface M {
> >  };
> >  };
> > 
> > Now under 2.3 spec this will not be allowed - and omniidl correctly
> > catches it.   Does omniidl have any switch which will allow it to
> > accept  the above idl?
> > 
> > I suspect we will have to bite the bullet and convert all the IDL, but
> > ya know ya gotta ask.  :-)
> > 
> > 
> 
> To get a list of all the command-line flags to omniidl, type:
>  omniidl -u

_Now_, ya' tell me.  Just curious - does this part of the specification
exist because there are backend compilers that are case insensitive?
I had to filter our Orbix IDL that I am converting to omniORB because
its creator liked doing things such as:

typedef string NameString;
NameString namestring;	// without the -nc flag, omniidl will flag as a conflict

I don't really subscribe to writing IDLs or any C++ header file this
way, but I know the SUNWspro compiler doesn't mind.

> 
> You will see a flag:
>   -nc             Do not treat identifiers differing only in case as an error
> 
> -- 
> Craig Rodrigues        
> http://www.gis.net/~craigr    
> rodrigc@mediaone.net          
>