Anonymous one: Variations on this approach are

c:\@Work\Perl\monks\Anonymous Monk\1199083>perl -wMstrict -MData::Dump
+ -le
"my @bin;
 ;;
 while(my $line = <>) {
   push @bin, $line;
   }
 ;;
 dd \@bin;
"  junk_1.txt
["line uno\n", "line two\n", "line 3\n"]

c:\@Work\Perl\monks\Anonymous Monk\1199083>perl -wMstrict -MData::Dump
+ -le
"my @bin = <>;
 ;;
 dd \@bin;
"  junk_1.txt
["line uno\n", "line two\n", "line 3\n"]
[download]

where junk_1.txt contains

line uno
line two
line 3
[download]

Both these variations relieve you of the need to maintain the  $counter counter (assuming it's initialized to 0).