Issue 535:  WWW Form output (issues)
Source:  (, )
Nature:  Uncategorized
Severity: 
Summary: Summary: Issue regarding implementation of the Query IDL specifications on a Java ORB. Issue involves implementing following idl definition from the CosQueryCollection module

Resolution: 
Revised Text: 
Actions taken:
March 6, 1997: received issue
Discussion: 

End of Annotations:=====
Return-Path: <soley>
Date: Thu, 6 Mar 1997 08:21:23 -0500
From: Richard Mark Soley <soley@emerald.omg.org>
To: issues
cc: msecrist@fc.hp.com
Subject: [www: WWW Form output]
Reply-to: soley@omg.org
Errors-to: soley_errors@omg.org

------- Start of forwarded message -------
Return-Path: <www>
Date: Wed, 5 Mar 1997 13:08:39 -0500
From: www
To: experts@omg.org, web-incoming@amethyst.omg.org
Subject: WWW Form output
Content-Length: 1456

Name:: Mark Secrist
Email: msecrist@fc.hp.com
Questions: Hi,
  I ran into an issue regarding implementation
of the Query idl specification on a Java ORB.
The issue involves implementing the following
idl definition from the CosQueryCollection module:

typedef boolean Null;
union FieldValue switch(Null) {
case false : Value v;
};

Because the discriminant of the above union is defined
as a boolean, the Java idl compiler will actually
use the Java boolean primitive type.  The problem is
that switch only allows an int, therefore, the Java
compiler aborts with an error.
In my estimation, the jidl compiler is doing the
right thing by attempting to use the boolean primative,
but the use of a boolean as the discrimenant of a
union is incorrect.  This only works for C++ 
(for now).
My short term solution was to re-define the above idl
as follows:

enum FVType { TypeFalse,TypeTrue };
union FieldValue switch(FVType) {
        case TypeFalse : Value v;
        case TypeTrue : octet nothing;
};

This seems to work for both C++ and Java.  I don't
know the proper mechanism for feeding this back
other than sending this to the Experts.  If this
is the incorrect forum for this, could you tell
me what the correct one is?  Also, if I have missed
something or come to an incorrect conclusion, I
would appreciate feedback there as well.

Thank you,

Mark Secrist
msecrist@fc.hp.com

nextpage: http://www.omg.org/library/exthnks.htm
emailto: experts@omg.org

------- End of forwarded message -------