Gurusamy Sarathy wrote: > >> I doubt that change#1307 is the problem--we call sv_mortalcopy() in > >> lots of other situations. > > > >Yes, but not where tied magic is involved. > > How do you know? sv_mortalcopy() is called *unconditionally* in lots > of places (when returning values from a subroutine, or in C<($x,@a) = @a>, > for instance). Sorry - what I mean is that the code in pp_aelem specifically checks for magic and then calls sv_mortalcopy only if the sv is magical. If it is not magical the call to sv_mortalcopy doesn't happen. > >Shouldn't tie magic be disabled when retrieving from the array > >'underneath' a tie? Unless this is done, self-ties cannot ever work. > >If self-ties are deemed to be unsupported, then pp_tie should detect and > >attemt to do so and croak. That is certainly the easiest option. I'm > >happy to have a crack at getting them to work correctly, but I'd need > >someone with a sufficiently pointy hat to make some suggestions as to > >how to do it. > > Code always makes people talk. :-) And a little guidance is likely to avoid wasted effort and frustration. At the moment the only patch I'm likely to submit will croak if it detects a self-tie. Is this what is wanted? Alan BurlisonThread Previous | Thread Next