Re: Regular Expression

You can do this a variety of ways, but here are three:

$string =~ m/abc.*xyz|xyz.*abc/i
$string =~ m/(?=.*abc)(?=.*xyz)/i;
$string =~ /abc/i and $string =~ /xyz/i
[download]

The ampersands inside the regular expression will be matched as ampersands and not used as logical operators.

-enlil

Comment on Re: Regular Expression Download Code

Replies are listed 'Best First'.
Re^2: Regular Expression by nobull (Friar) on Feb 09, 2005 at 20:14 UTC
`$string =~ m/(?=.abc)(?=.xyz)/i;` [download] That unanchored pattern is pathologically inficient. If the string does not match it will try and fail at every character position. Even with a relatively short strings this has a considerable effect. `use Benchmark; my $string = 'x' x 100; timethese 1_000_000 => { unanchored => sub { $string =~ /(?=.abc)(?=.xyz)/i }, anchored => sub { $string =~ /^(?=.abc)(?=.xyz)/i }, };` [download] Benchmark: timing 1000000 iterations of anchored, unanchored... anchored: 1 wallclock secs ( 1.30 usr + 0.00 sys = 1.30 CPU) @ 768049.16/s (n=1000000) unanchored: 70 wallclock secs (70.12 usr + 0.00 sys = 70.12 CPU) @ 14261.06/s (n=1000000)	[reply] [d/l] [select]

Replies are listed 'Best First'.

Re^2: Regular Expression
by nobull (Friar) on Feb 09, 2005 at 20:14 UTC

$string =~ m/(?=.*abc)(?=.*xyz)/i;
[download]

That unanchored pattern is pathologically inficient. If the string does not match it will try and fail at every character position. Even with a relatively short strings this has a considerable effect.

use Benchmark;
my $string = 'x' x 100;
timethese 1_000_000 => {
    unanchored => sub { $string =~ /(?=.*abc)(?=.*xyz)/i },
    anchored => sub { $string =~ /^(?=.*abc)(?=.*xyz)/i },
};
[download]

Benchmark: timing 1000000 iterations of anchored, unanchored...
anchored: 1 wallclock secs ( 1.30 usr + 0.00 sys = 1.30 CPU) @ 768049.16/s (n=1000000)
unanchored: 70 wallclock secs (70.12 usr + 0.00 sys = 70.12 CPU) @ 14261.06/s (n=1000000)

[reply]
[d/l]
[select]