Re: Slicing & sliding

[matthewsj@deleteme.saic.com (John B. Matthews, M.D.)]

In article <3720EA2F.B359672@averstar.com>, Tucker Taft <stt@averstar.com>
wrote:

> John B. Mattehws, M.D. wrote:
> > 
> > Given the following program
> > 
> > with Text_IO;
> > 
> > procedure Slice is
> > 
> > S : String (1 .. 64) := (others => '.');
> > 
> >     function F return String is
> >     begin
> >         return "result";
> >     end F;
> > 
> > begin
> > 
> >     Text_IO.Put_Line(S);
> >     S(S'Range) := F(S'Range);
> >     Text_IO.Put_Line(S);
> > 
> > end Slice;
> > 
> > The output is
> > ................................................................
> > result
> > 
> > I was suprised to see that the indicated slice of F (F(S'Range)) is allowed
> > to be larger than the result returned by F, despite RM 4.1.2(7). For
> > smaller slices, the remainder of S seems to slide down, and the rest is
> > padded with nulls. Can someone help me understand this? Is this conformant
> > with the RM?
> 
> Sure looks like a compiler bug to me.  Alternatively, you
> are suppressing constraint checks without knowing it.

The latter, of course. Now clear in restrospect:-)

Thanks,

John

> On my favorite compiler, your program prints the line of dots, and
> then dies with an unhandled Constraint_Error on the slice
> assignment.
> 
> > Thanks,
> > 
> > John
> > ----
> > John B. Matthews, M.D.
> > matthewsj@saic.com; jmattthews@nova.wright.edu
> 
> -- 
> -Tucker Taft   stt@averstar.com   http://www.averstar.com/~stt/
> Technical Director, Distributed IT Solutions  (www.averstar.com/tools)
> AverStar (formerly Intermetrics, Inc.)   Burlington, MA  USA