Kudos to the voters... awarding Abigail (with the slowest regex) the most number of votes. (at least as I write this) There is a difference between wanting to do it with a regex and wanting to do it as slowly as possible. I took the above regex's from grinder, roger, posix_guy, and Abigail and ran them through a benchmark doing 10,000 iterations each.
First, Roger: might want to add a $ on the end there because yours matches 19534 and 20010 etc... and posix_guy: yours matched 19534 for some reason. Anyway, I took grinder's data set and put it in an array (as with the multiple subroutines all accessing __DATA__ it wasn't properly displaying things). Then I seperated each regex into a different subroutine with the name of the author. I've attached the code if anyone wants to run it themselves. All results are from my Powerbook G4 667Mhz.
Benchmark: timing 10000 iterations of abigail, grinder, posixguy, roger...
grinder: 1 wallclock secs ( 1.06 usr + 0.03 sys = 1.09 CPU) @ 9174.31/s (n=10000)
roger: 2 wallclock secs ( 1.02 usr + 0.07 sys = 1.09 CPU) @ 9174.31/s (n=10000)
posixguy: 1 wallclock secs ( 1.37 usr + 0.03 sys = 1.40 CPU) @ 7142.86/s (n=10000)
abigail: 57 wallclock secs (51.45 usr + 0.38 sys = 51.83 CPU) @ 192.94/s (n=10000)
#!/usr/bin/perl -w
use strict;
use Benchmark;
my @data = qw{1949 1950 1951 1999 2000 2001 2010 2049 2050 2051 2102 2
+2102 19534 19080 20010};
sub grinder() {
foreach( @data ) {
chomp;
print "$_ ", /^(19[5-9]\d|20([0-4]\d|50))$/ ? "ok\n" : "n
+ok\n";
}
}
sub roger() {
foreach (@data) {
chomp;
print "$_ ", /^(?:19|20)(?:(?:(?<=19)[5-9]|(?<=20)[0-4
+])[0-9]|50)/x ? "Ok\n" : "not ok\n";
}
}
sub posixguy() {
foreach (@data) {
chomp;
print "$_ ", /^(19[5-9]\d)|(20([0-4]\d)|50)$/ ? "ok\n"
+ : "nok\n";
}
}
sub abigail() {
foreach (@data) {
chomp;
local $" = "|";
print "$_ ", /^(?:@{[1950 .. 2050]})$/ ? "ok\n" : "nok
+\n";
}
}
timethese(10000, {grinder=>'grinder()', roger=>'roger()', posixguy=>'p
+osixguy()', abigail=>'abigail()'});