BrowserUk has asked for the wisdom of the Perl Monks concerning the following question:

Is this another reduce bug or am I missing something? 

#! perl -slw
use strict;
use List::Util qw[ reduce ];

my $s = 'this is a test';

reduce{
    print "$a:$b";
    $b;
} split ' ', $s;
print "\n---\n";

my @p;
reduce{
    print "$a:$b";
    push @p, [ 1, 2 ];
    $b;
} split ' ', $s;
print "\n---\n";

my @q;
reduce{
    print "$a:$b";
    push @q, [ $a, 2 ];
    $b;
} split ' ', $s;
print "\n---\n";

my @r;
reduce{
    print "$a:$b";
    push @r, [ $a, $b ];
    $b;
} split ' ', $s;

__END__

c:\test>junk3
this:is
is:a
a:test

---

this:is
is:a
a:test

---

this:is
is:a
a:test

---

this:is
Use of uninitialized value in concatenation (.) or string at c:\test\j
+unk3.pl line 31.
:a
Use of uninitialized value in concatenation (.) or string at c:\test\j
+unk3.pl line 31.
:test

[download]
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.
"Too many [] have been sedated by an oppressive environment of political correctness and risk aversion."

Replies are listed 'Best First'.
Re: Another reduce bug?
by wind (Priest) on Aug 16, 2007 at 02:31 UTC

    You're not missing anything. I observe the same bug. The following two variations make the problem disappear.

    Using pure perl version of List::Util 

    #! perl -slw
use strict;

BEGIN {$List::Util::TESTING_PERL_ONLY = 1};

use List::Util qw[ reduce ];

my $s = 'this is a test';

my @r;
reduce {
    print "$a:$b";
    push @r, [ $a, $b ];
    $b;
} split ' ', $s;

    [download]

    Or passing an array instead of the split: 

    #! perl -slw
use strict;

use List::Util qw[ reduce ];

my $s = 'this is a test';
my @s = split ' ', $s;

my @r;
reduce {
    print "$a:$b";
    push @r, [ $a, $b ];
    $b;
} @s;

    [download]
    - Miller
[reply]
[d/l]
[select]

      Ha! Thanks for the confirmation. It is a strange one. Following on from your use an array solution, I tried the ol' put it in a list context trick--which is probably a erroneous name, but it serves-- and voila. That also fixes it. Maybe an interation between two different optimisations? Time for a perlbug I think. 

      my @r;
reduce{
    print "$a:$b";
    push @r, [ $a, $b ];
    $b;
} () = split ' ', $s;

      [download]
      Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
      "Science is about questioning the status quo. Questioning authority".
      In the absence of evidence, opinion is indistinguishable from prejudice.
[reply]
[d/l]

        I expect that to make reduce() loop over an empty list and, for me, that is what it does. So, at least in my environment, I don't see how that qualifies as "fixing".

        - tye        

[reply]
Re: Another reduce bug? (noinc)
by tye (Sage) on Aug 16, 2007 at 02:32 UTC

    I'd guess a ref-counting bug, yes (which includes bugs in use of mortalization / FREETMPS).

    - tye        

[reply]

Back to Seekers of Perl Wisdom