RE: [tcltk-perl] return of 1=>1.0 transformation error

[Jeff H. <je...@Ac...>]

This is the exact code being run:

my $row = 0;
foreach my $leaf('Tk.xbm','Xcamel.gif')
 {
  my $file = "./t/$leaf"; #Tk->findINC($leaf);
  my $src = $mw->Photo(-file => $file);
  ok(defined($src),1," Cannot load $file");
  my $kind = 'Initial';
  my $col = 0;
  $mw->Label(-text  => 'Initial')
	->grid(-row => $row, -column => $col);
  $mw->Label(-background => 'white',-image => $src)
	->grid(-row => $row+1, -column => $col++);
	...

We see that the problem is at -row 1, so that last line is the
issue.  That this becomes SvNOK w/o being SvIOK in Perl 5.6
seems "just wrong" to me (and obviously to others as 5.8 corrects
the issue).  You guys are right that I should try without the
SvIOK as well, but still, this is a wacky promotion behavior.

Jeff

> > > Tcl error 'bad grid value "1.0": must be a non-negative integer at 
> > > d:/Perl56/site/lib/Tcl.pm line 363. ' while invoking scalar result 
> > > call:
> > >       "grid .lbl02 -row 1 -column 0" at
> > 	...
> > > perl-5.8.2 passes test without this error, 5.6.0 and 5.6.1 fail.
> > 
> > This is specifically supposed to be handled by this code in
> > TclObjFromSv:
> > 
> >     else if (SvNOK(sv)) {
> > 	double dval = SvNV(sv);
> > 	int ival;
> > 	/*
> > 	 * Account for some perl versions aggressive NOK-ness where
> > 	 * an int was all that was intented.
> > 	 */
> > 	if (SvIOK(sv) && ((double)(ival = SvIV(sv)) == dval)) {
> > 	    objPtr = Tcl_NewIntObj(ival);
> > 	} else {
> > 	    objPtr = Tcl_NewDoubleObj(dval);
> > 	}
> >     }
> > 
> > as it was an issue that 5.6 decided to aggressively promote "1" with 
> > SvNOK as well as SvIOK, whereas 5.8 is corrected to not do that.
> 
> That code will not catch the case when the floating point 
> value has already been stringified.
> 
> What is weird thought is that Perl would still stringify an 
> NV of value 1.0 as "1" and not "1.0".  So is it possible that 
> the value already got round-tripped through Tcl?