Re^2: sorting a directories content

print join "\n", sort {-M $a <=> -M $b} @DirList;

This will result in a flood of stat calls. This would be more efficient:

print join "\n", map $_->[0], sort {$a->[1] <=> $b->[1]} map [$_, -M], @DirList;

Plus, one might want to do something if -M returns undef (due to permission problems or broken symlinks).

Update: corrected my silly typing errors, thanks to GrandFather

Comment on Re^2: sorting a directories content Download Code

Replies are listed 'Best First'.
Re^3: sorting a directories content by GrandFather (Saint) on Apr 13, 2007 at 21:48 UTC
The Schwartzian Transform is indeed most appropriate in this case. However your sort code is a little buggy. OP may have better luck with: `print join "\n", map {"$_->[0]"} sort {$a->[1] <=> $b->[1]} map {[$_, +-M]} @DirList;` [download] Note that the replies to What is "Schwarzian Transform" (aka Schwartzian) contain many useful comments and references pertaining to the Schwartzian Transform too. DWIM is Perl's answer to Gödel	[reply] [d/l]
Re^4: sorting a directories content by ikegami (Patriarch) on Apr 13, 2007 at 22:53 UTC
An even faster solution that sorts files with identical timestamps by name: `# Assumes 32-bit timestamps. print map { substr($_, 4) . "\n" } sort map { pack('Na*', (stat($_))[9], $_) } @DirList;` [download]	[reply] [d/l]