[PolyORB-users] Additional pragmas facilitating migration of legacy Ada to CORBA

[oliver.kellogg at t-online.de]

Hello polyorb-users, 

	In the GNU Ada CORBA Kit
(http://orbitada.cvs.sourceforge.net/viewvc/orbitada/gnack), we
support following pragmas that facilitate the migration of legacy Ada
code to CORBA: 

	#pragma switchname <union_type> <switch_name>
- Instructs idl2ada to use the given name for the union switch instead
of
  "Switch". Example:

   // IDL
   #pragma switchname union_t discriminant
   union union_t switch (boolean) {
      case FALSE:
         char a;
      case TRUE:
         float f;
   };

   -- Ada
   type ind_t is (zero, one, two);
   type union_t
     (discriminant : CORBA.Boolean := FALSE)
    is record
      case discriminant is
         when FALSE =>
            a : CORBA.Char;
         when TRUE =>
            f : CORBA.Float;
      end case;
   end record;

#pragma array_index <array_type> <index_type>
- Instructs idl2ada to use the given type as the index type in the
  declaration of an array mapped from "typedef". Example:

   // IDL
   enum ind_t { zero, one, two };
   #pragma array_index arr_t ind_t
   typedef boolean arr_t[3];

   -- Ada
   type ind_t is (zero, one, two);
   type arr_t is array (ind_t) of CORBA.Boolean; 

	Would these be of interest for PolyORB? 

	If so then I would volunteer to provide their implementation in iac. 

	Regards, 

	Oliver M. Kellogg 

	<mailto:okellogg at users.sourceforge.net>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/polyorb-users/attachments/20110614/0a82bb90/attachment.htm