I think you are misinterpreting the results of your test. (No, I was misunderstanding what point was being made. ;)

Scalar::Util::isdual returns a true value if the variable being tested is a dualvar. Your test indicates that isdual is returning 1 when tested against $!, which is a true value, indicating that $! is a dualvar.

Update: A list of "dualvar" special variables, as I find them.

• $( The documentation eludes to this; it may contain a space-delimited list (a string), but you can only assign a number to it.
• $) The documentation is similar on this one, and it too is a dualvar.
• $^E The POD says "Caveats mentioned in the description of $! generally apply to $^E , also."

Also, since $_ implicitly topicalizes a foreach loop, and since it's an alias to an iteration's topic, it can be made to appear to be a dualvar:

for ( $! ) {
  print Scalar::Util::isdual($_), "\n";
}
[download]

...outputs '1', because in this case $_ is an alias to a dualvar. So we can add to the list:

• Any variable that aliases a dualvar becomes a dualvar while the alias is in effect (consider foreach ( $! ) { # $_ is a dualvar here })

That's my point. dualvar indicates that $! is indeed a dualvar but the documentation of Scalar::Util say $! is not a dualvar. Or did I misunderstand what the documentation claims?

Oh, I see what you're saying. It's probably a bug in the documentation for isdual, though relying on how a feature is implemented internally always carries risks.




Dave

Not a variable, but a value: the value for false, which you can generate with !1.

It return '' in string context and 0 in numeric context, and doesn't warn in numeric context like the empty string normally does:

$ perl -wE 'say 0+""'
Argument "" isn't numeric in addition (+) at -e line 1.
0
$ perl -wE 'say 0+!1'
0
[download]

My version of Scalar::Util doesn't export an is_dual function, so I can't check. It helps to type that without the _ :-)

Update: I should have mentioned it earlier: even though the thing is called a "dualvar", it's not actually variables that are tested; values are being tested. It makes no difference if those values are return values, stored in array, or are stored in a scalar variable.

A value for "true" is also dualvar:

perl -MScalar::Util -E 'say Scalar::Util::isdual(!0)'

...outputs '1', indicating the test was true.




Dave

wow !1. I have not heard of that earlier. greping the pods of 5.18.2 - it is not mentioned there either. But it works as you showed.

Are there other dualvars in perl core?

$ perl -MScalar::Util=isdual -E'
   $x=1; say isdual($x)?1:0;
   "$x"; say isdual($x)?1:0;
'
0
1
[download]

The stringification is cached.

IMHO, that's a rather stupid interpretation of "dualvar". If the string value equals the stringification of the other value (IV or NV), then it shouldn't be considered a "dualvar". But then, that's what one would expect from a lazy XS function. A Perl version would just do 0+$x eq $x instead.

!1 doesn't need to be implemented as a dualvar. $x=''; {no warnings; 0+$x;} is also enough to get you what is needed for !1 (a string value of '' and a numeric value of 0 w/o a warning). The canonical value for !1 that Perl uses happens to be a real dualvar, though. But "both an empty string and zero" isn't enough to imply "dualvar".

Win32::TieRegistry can use dualvars.

Update: To be clearer, a Perl version of isdual() would be:

sub isdual {
    my( $v ) = @_;
    no warnings;
    return 0
        if  int $v ne $v^'0'    # No IV nor NV in value
        ||  $v eq 0+$v          # Cached string just same as IV/NV
        ||  $v == "$v"          # Cached number just same as 0+PV
    ;
    return 1;
}
[download]

- tye        

The stringification is cached.

This is a good point. Is this good rule of thumb (bound to be imperfect but close enough)? Anytime an SV with an IV or NV creates a PV representation, that will remain in effect (making it a dualvar) until the container receives a new assignment.




Dave

say "!",$x,"!"; doesn't, but "everything else" does.

It is possible (in XS) to put anything in the specific slots:

$ perl -MData::Peek -wC3 \
       -E'my $tv = Data::Peek::triplevar ("\N{GREEK SMALL LETTER PI}", 3, 3.1415);' \
       -e'say for map { $_ // "<undef>" } DDual ($tv)'
π
3
3.1415
<undef>
0

Data::Peek's DDual will return all of those: my ($pv, $iv, $nv, $rv, $hm) = DDual ($var [, $getmagic])


Enjoy, Have FUN! H.Merijn

This post is wrong.

Whether the docs are correct or not depends on the version of Perl.

perl -MDevel::Peek -e'$!=1; print("$!\n"); Dump($!);'
[download]

5.8.9:

...
  FLAGS = (GMG,SMG,pNOK,pPOK)
...
[download]

No public values. Not a dualvar.

5.18.1:

...
  FLAGS = (GMG,SMG,NOK,POK,pNOK,pPOK)
...
[download]

Two public values (NOK = has NV = has floating point number, POK = has PV = has string). Dualvar.

The latest Scalar::Util::isdual considers FLAGS = (GMG,SMG,pNOK,pPOK) to be a dualvar. (No idea about earlier versions, but I suspect it's the same.) The docs are wrong. The following returns 1 for both 5.8.9 and 5.18.1, while the docs say it returns false:

perl -MScalar::Util=isdual -MDevel::Peek \
   -e'$!=1; print("$!\n"); print(isdual($!)?1:0,"\n"); Dump($!);'
[download]

Nit: $! is magical (like tied variables), but it's not a tied variable like the docs say.