WRT use case: one of the (admittedly few) places I've seen something like this (which is why I mentioned the iterator module from TT) is in templating where you want to "decorate" the first or last item in a list differently than other items. Think maybe pagination where displaying a chunk from the middle and you want those items to have a link to the next or subsequent sets of results (terrible handwavy sample follows, and yes you could do similar stuff with JS or CSS instead).

<table>
[% FOREACH item IN current_chunk %]
<tr>
[% IF loop.first %]<td> &lt;&lt; </td>[% ELSE %]<td>&nbsp;</td>[% END 
+%]
<td>[% item %]</td>
[% IF loop.last%]<td> &gt;&gt; </td>[% ELSE %]<td>&nbsp;</td>[% END %]
</tr>
[% END %]
</table>
[download]

Edit: And I realized I hadn't linked to the TT FOREACH docs which explain the loop magic (which is Template::Iterator under the hood).

Yeah, that's doable. I can also think of doing it without having everything inside the loop ..

#!/usr/bin/perl

use strict;
use warnings;

#  2022-0125: Handle First and List iems differently that everything
#  in between ..

{
    my @keys = qw/q w e r t y u i o p/;

    print "First is $keys[0] ..\n";

    foreach my $k ( @keys[1..(scalar @keys -2)] ) {

      print "In the middle: $k\n";
    }

    print "Finally, we have $keys[-1].\n";
}
[download]

This produces the output

First is q ..
In the middle: w
In the middle: e
In the middle: r
In the middle: t
In the middle: y
In the middle: u
In the middle: i
In the middle: o
Finally, we have p.
[download]

But I like your solution of using an Iterator -- then it has the job of handling the 'first' and 'last' issues.

Alex / talexb / Toronto

Thanks PJ. We owe you so much. Groklaw -- RIP -- 2003 to 2013.

Edge case behaviour might need some changes:

#!/usr/bin/perl

use strict;
use warnings;

#  2022-0125: Handle First and List iems differently that everything
#  in between ..

{
    # my @keys = qw/q w e r t y u i o p/;
    my @keys = qw/q/;

    print "First is $keys[0] ..\n";

    foreach my $k ( @keys[1..(scalar @keys -2)] ) {

      print "In the middle: $k\n";
    }

    print "Finally, we have $keys[-1].\n";
}
[download]

B:\>perl 11140865-single.pl
First is q ..
Finally, we have q.

B:\>
[download]

Alexander

--
Today I will gladly share my knowledge and experience, for there are no sweeter words than "I told you so". ;-)

It's so easy to avoid the limitation. Replace $v eq $values[-1] with \$v == \$values[-1]

> It's so easy to avoid the limitation.

unfortunately no, but good idea nonetheless ... :)

  DB<39> x @input
0  2
1  7
2  11
3  15
  DB<40> use experimental 'refaliasing'; \$input[1] = \$input[-1]

  DB<41> x @input
0  2
1  15
2  11
3  15
  DB<42> for my $v (@input) { say $v if \$v==\$input[-1] }
15
15

  DB<43>
[download]

I'm not even sure that experimental feature is even needed...

IIRC one could somehow use typeglob aliasing on array elements.

Cheers Rolf
_{(addicted to the Perl Programming Language :)
Wikisyntax for the Monastery}

unfortunately no, but good idea nonetheless ... :)

Good catch. I'll add a warning to my answer that first suggested the idea.

I'm not even sure that experimental feature is even needed...

Correct.

my $inputs = sub { \@_ }->( 0, $x, $x );

for my $v (@$inputs) {
   say \$v;
}
[download]

SCALAR(0x559bb5b21960)
SCALAR(0x559bb5b33228)
SCALAR(0x559bb5b33228)
[download]

I imagine that would Data::Alias can also achieve this.