>>584は試してみてからそういってるのか？
手動フォーマットよりsprintf()の方が圧倒的に高速だが…俺のやり方がまずいんだろうか。
それともオーバーヘッドの意味が俺の考えている小物と違うのかな。
WinME perl5.6.1 ActivePerl 631
メモリ320MB CPU Duron 800MHｚ
#!perl -w
use Benchmark;
my @tm = localtime time;
sub timef1{
($sec, $min, $hour, $day, $mon, $year) = @tm;
$year += 1900;
$mon++;

$_ = sprintf '%d/%02d/%02d %02d:%02d:%02d', $year, $mon, $day, $hour, $min, $sec;
}
sub timef2{
($sec, $min, $hour, $day, $mon, $year) = @tm;
$year += 1900;
$mon++;

($mon < 10) and ($mon = "0$mon");
($day < 10) and ($day = "0$day");
($hour < 10) and ($hour = "0$hour");
($min < 10) and ($min = "0$min");
($sec < 10) and ($sec = "0$sec");
$_ = "$year/$mon/$day $hour:$min:$sec";
}
timethese 100000 => {
sprintf => \&timef1,
handformat => \&timef2,
};
__END__
結果
Benchmark: timing 100000 iterations of handformat, sprintf...
handformat: 6 wallclock secs ( 5.38 usr + 0.00 sys = 5.38 CPU) @ 18587.36/s (n=100000)
sprintf: 1 wallclock secs ( 0.82 usr + 0.00 sys = 0.82 CPU) @ 121951.22/s(n=100000)