Need a regex..

littlemonk has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: Need a regex.. by AnomalousMonk (Archbishop) on May 21, 2012 at 14:33 UTC
You don't say what should happen with negative numbers, but it's easy enough to exclude them in the same way four-or-more digit patterns are excluded. Update: A link to Regexp::Common might be helpful. >perl -wMstrict -le "use Regexp::Common; ;; for my $t (qw( 12.12 123.123 1 12 123 1. 12. 123. .1 .12 .123 0.1 1.0 0.0 1234.0 1234 0.1234 .1234 -123 -.12 -0.1 )) { my $ok = $t =~ m{ \A (?! .* \d{4}) $RE{num}{real} \z }xms; printf qq{%s '$t' \n}, $ok ? 'accept' : 'REJECT'; } " accept '12.12' accept '123.123' accept '1' accept '12' accept '123' accept '1.' accept '12.' accept '123.' accept '.1' accept '.12' accept '.123' accept '0.1' accept '1.0' accept '0.0' REJECT '1234.0' REJECT '1234' REJECT '0.1234' REJECT '.1234' accept '-123' accept '-.12' accept '-0.1' [download]	[reply] [d/l]
Re: Need a regex.. by Neighbour (Friar) on May 21, 2012 at 13:34 UTC
Could you specify whether only one number is entered per line, or if multiple numbers are entered per line, how are they separated? You can use `{n,m}` after any matching character/group to specify how often it is allowed to match (minimal n, maximal m). See also QUANTIFIERS.	[reply] [d/l]
Re: Need a regex.. by scorpio17 (Canon) on May 21, 2012 at 13:35 UTC
Something like this, maybe? `use strict; while (my $data = <DATA>) { chomp $data; if ($data =~ /^\d?\d?\d\.\d\d?\d?$/) { print "GOOD\n"; } else { print "BAD\n"; } } __DATA__ 12.12 123.123 0.1 1234.0 1234` [download]	[reply] [d/l]
Re^2: Need a regex.. by littlemonk (Sexton) on May 21, 2012 at 13:55 UTC
it should match non-decimal value also ...example 12,123, and it should not match more than 3 digits before and after decimal..i have one regex(/^\- \+{0,1}0-9{0,3}(\.){0,1}0-9{0,3}\%{0,1}$/) but its accepting 1234(more than 3 digits before decimal) ... plz help me	[reply]
Re^3: Need a regex.. by kennethk (Abbot) on May 21, 2012 at 14:33 UTC
Please wrap code it `<code>` tags to keep things legible. As you can see, your character classes got linkified. 1234 is passing because it's being read as (123)(4). You should set the whole part following and including the decimal in a non-capturing group with a conditional: `/^[-+]?[0-9]{0,3}(?:\.[0-9]{0,3})?%?$/` I've also removed a bunch of unnecessary escaping, and changed your `{0,1}` to the shorter `?` synonym. A read through of perlretut may be helpful. #11929 First ask yourself `How would I do this without a computer?' Then have the computer do it the same way.	[reply] [d/l] [select]
Re^3: Need a regex.. by Neighbour (Friar) on May 21, 2012 at 14:40 UTC
Edit: Lost the race to Re: Need a regex.., which is also slightly better :) Try a more grouped approach: `/^([+]\|[-])?[0-9]{0,3}(\.[0-9]{1,3})?$/` First the start of the line: `^` then an optional + or - sign, with the + enclosed in [] to avoid interpretation: `([+]\|-)?` then 0-3 digits: `[0-9]{0,3}` then an optional group of a period followed by 1-3 digits, with the period escaped in another way to avoid interpretation: `(\.[0-9]{1,3})?` and finally the end of the line: `$`	[reply] [d/l] [select]
Re: Need a regex.. by diddy_perl (Novice) on May 21, 2012 at 14:38 UTC
Just slightly modified the previous post: `use strict; while (defined(my $data = <DATA>)) { chomp $data; print "Got: ", $data, "\n"; if ($data =~ /[-+]?\d+\| # optional sign followed by whole number or [-+]?\d{0,3} # optional sign followed by at most 3 digits (\.\d{1, 3})? # followed by decimal point, and finally by at mo +st 3 digits /x) { print "GOOD\n"; } else { print "BAD\n"; } } __DATA__ 12.12 123.123 0.1 1234.0 1234` [download]	[reply] [d/l]