#!/usr/bin./perl

use strict;
use warnings;

my @in=("TFS100","TFS 100","TFS-CR100",
        "TFS-CR 100","TFS_100","TFS ID 100",
        "TFS CR 100","TFS ID100","TFS-ID 100",
        "TFS ID:100","TFS-ID: 100",
        "- TFS CR634: STRESS: H 17326,21600,",
        "CR0080588","TFS0080588"
       );
for (@in){
    print "$1\n" if /TFS[-_\s]?(?:CR|ID)?\s?([1-9]\d+)/;
}
[download]

Your criteria are a little under specified...

Amen to that, brother!

In most cases it works fine...

Further to choroba's post: MARVion: Your regex doesn't really work at all because almost everything in it is optional (i.e., has the  ? quantifier) except the  (\d+) bit at the end, so it is happy to match with and capture the first group of decimal digits in anything. What are you really after here?

>perl -wMstrict -le
"my @in = (
   'TFS100',
   'TFS 100',
   'TFS-CR100',
   'TFS-CR 100',
   'TFS_100',
   'TFS ID 100',
   'TFS CR 100',
   'TFS ID100',
   'TFS-ID 100',
   'TFS ID:100',
   'TFS-ID: 100',
   '- TFS CR634: STRESS: H 17326,21600,',
   'CR0080588',
   'TFS0080588',
   qw(1234  xxx1234  123xxx  xxx123xxx  123xxx456  xxx123xxx456),
   );
 ;;
 for my $s (@in) {
   $s =~ m{
     [Tt]? [Ff]? [Ss]? [-_\s]? [Cc]? [Rr]? [Ii]? [Dd]? :? \s? (\d+)
     }xms;
   print qq{'$1' <- '$s'};
   }
"
'100' <- 'TFS100'
'100' <- 'TFS 100'
'100' <- 'TFS-CR100'
'100' <- 'TFS-CR 100'
'100' <- 'TFS_100'
'100' <- 'TFS ID 100'
'100' <- 'TFS CR 100'
'100' <- 'TFS ID100'
'100' <- 'TFS-ID 100'
'100' <- 'TFS ID:100'
'100' <- 'TFS-ID: 100'
'634' <- '- TFS CR634: STRESS: H 17326,21600,'
'0080588' <- 'CR0080588'
'0080588' <- 'TFS0080588'
'1234' <- '1234'
'1234' <- 'xxx1234'
'123' <- '123xxx'
'123' <- 'xxx123xxx'
'123' <- '123xxx456'
'123' <- 'xxx123xxx456'
[download]

Hi again,

well, with a litte change Utilitarian's Code works perfectly for me.

TFS[-_\s]?(?:CR|ID)?:?\s?([1-9]\d+)
[download]

oops, I missed the one with a colon :(

(?: # begin non capturing group
CR|ID # alternation accepts either pattern group 'CR' or 'ID'
)? # group is optional
[download]

print "Good ",qw(night morning afternoon evening)[(localtime)[2]/6]," fellow monks."

I would apply one or more regular expressions to identify whether-or-not this a string that you are interested in.   Then, write another regex to match only what you need from each.   Use a subroutine or subroutines; “regex golf” only goes so far.   Treat the two problems separately, and break each one down into as many cases as you need.

Obviously, entry #11 is the odd-man-out from all the rest.   You need to build a suite of examples of everything that the string could possibly be, and determine exactly what should be used to extract the data-of-interest from each one.   Then, I strongly recommend putting the whole thing to the Test::Most.   Write a test suite that proves the correct operation of the program.   Also, write the program itself so that it is suspicious of all its inputs and will, say, die() if it encounters anything that it cannot handle.   (Otherwise, you have no practical way to realize that a problem exists, either in the data or in the program or both.   The fact that the program ran cleanly should be a strong indicator that both the data and the software are correct.)

Hi,

thanks for all your help! I'm sorry that I'm not precise enough in describing my problem!
Well, I will try your suggestions and solutions and if I can fix my problem I will tell you so that this thread has a solution to it's problem!
Thank you for all your replies!

Greetz

MARV