i hate to get picky, but your Schwartzian transform doesn't look like merlyn's to me.


perlmonkey's:

my @sorted = 
           map{ $_->[0] }
           sort {$a->[1] cmp $b->[1]}
           map { m/(.*?)\s*(\S+)$/; [$_, "$2 $1" ] }
           @unsorted;
[download]

merlyn's:

my @output =
  map { $_->[1] }
  sort { $a->[0] cmp $b->[1] }
  map { [$_, expensive_func($_)] }
  @input;
[download]

is this because the ST is more general than i thought, or is there a typo in merlyn's ST writeup? i'm not sure i'm comfortable with any sort that compares $a->[0] to $b->[1]... is there ever a good reason to do that?

update: looks like merlyn's changed his to compare index 1 of both arrays. which is almost a shame; i was hoping there was some really cool reason to compare different indices. but anyway, i withdraw the question.