From: Peter Dimov (pdimov_at_[hidden])
Date: 2002-08-14 14:07:36
From: "David Bergman" <davidb_at_[hidden]>
> Peter,
>
> One "subtle" note, though. You are aware that what you threw in 'f' was
> actually the original object?
Yes, that was the point I was trying to make; I said in the original post
that
> > >Well, I am not a compiler writer, but it seems to me that to
> implement
> > >"throw;" and "catch", the compiler already needs a way to copy the
> > >exception, complete with its original type. :-)
The exception is sliced by the catch() clause, but the original is retained
and rethrown.
> The copy-constructor of X *does* indeed
> slice (they always do, them copy-constructors ;-). If you want to see
> the effect of this X object being passed up, and caught as an Y, change
> the code to:
This calls abort() as expected. :-) You can't catch X as Y.
Either way, it seems clear to me that a compiler can transport an exception
in the way we need (since it can generate clone() and throw_this()
functions - or an equivalent - for the correct type at the throw point, even
if the actual exception handling implementation doesn't already require
them.)