[bitc-dev] Bug in DEFREPR specification (and implementation?)

[Jonathan S. Shapiro]

I think you mean all *scalar* literals.

It definitely needs to be extended to boolean. I can make a case for
char, but I've never seen that actually used in the wild.

shap
On Fri, 2007-05-04 at 11:53 -0400, Swaroop Sridhar wrote:
> Jonathan S. Shapiro wrote:
> > The list of constraints in section 3.6.3 is not quite right.
> > 
> > Two fields in two legs having the same name must not only have the same
> > bit position. They must have the same TYPE!
> > 
> > I suspect this error is present in the implementation as well (?)
> 
> True. I have fixed this problem in the specification and in the 
> implementation.
> 
> Currently, the values in the discriminating values (for fields within 
> the where clause) must be integer literals. Consequently, the 
> discriminating fields must be of integer/bitfield type.  I have noted 
> this restriction also in the spec. It is conceivable that this rule will 
> be relaxed include to all literals.
> 
> Thanks,
> Swaroop.
> 
> _______________________________________________
> bitc-dev mailing list
> bitc-dev at coyotos.org
> http://www.coyotos.org/mailman/listinfo/bitc-dev