regex question

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

while going through input file with each line as below format with below script

--file--
server1 M.22 s.23 2 AB 192.16.1.1 0 42 BBS 50
server3 M.22 s.33 2 AB 192.16.1.2 0 42 BBS 50 MK
server3 M.49 s.49 3 BB 10.1.1.1 0 42 BS 50
and more............
Program runs fine but I am only trying to replace 10th field's 50 with something else and since data is random, sometimes field 3 has 50 or 150 on it and messing up the program.
Can someone please give me a pointer for me to write this better ?
I am guessing

 if ($cou) {
               $orig =~ s/\b$itemY\b/$default/;
               print "$orig\n";
[download]

is just looking at whichever 50 that occurs first and doing the substituion.
Can someone please give me a pointer on this


while (<>) {
   chomp;
   my $orig = $_;

   if ($orig =~ /^#/) {
        print "$_\n";
        next;
   }
   if ($orig =~ /^$/) {
        print "$_\n";
        next;
   }
   if (/(\w+)\s+M\.\d+\s+(s\.\d+)\s+(\S+\s+){6}(\d+)(.+)?$/) {
   #server1  yahoo.22  s.50   2   ABCCD     192.168.1.1    0   42  BBS
+  50
       my $cou = 1;
       my $item = join('@', $2, $1);
       my $itemY = $4;
       for my $i_carriers (keys %s_carriers) {
           if ("$item" eq "$i_carriers") {
               my $itemX = $s_carriers{$item};
               $orig =~ s/\b$itemY\b/$itemX/;
               print "$orig\n";
               $cou = 0;
           }
           last if ! $cou;
       }
       if ($cou) {
               $orig =~ s/\b$itemY\b/$default/;
               print "$orig\n";
       }
   }
}
[download]

Comment on regex question Select or Download Code

Replies are listed 'Best First'.
Re: regex question by moritz (Cardinal) on Nov 12, 2007 at 08:50 UTC
My suggestion is to split at the whitespace, and take or modify those parts you need. Afterwards you can join them again. For example: `my @fields = split m/ /, $line; $fields[9] = 50; print join ' ', @fields;` [download]	[reply] [d/l]
Re: regex question by narainhere (Monk) on Nov 12, 2007 at 10:34 UTC
If you want to be precise on which field you are replacing, Anchoring(i.e Look-ahead and look-behind) would help. The following snippet illustrates how to replace the nth field. This would replace the 2nd field `use strict; use warnings; my $str="1 2 3 4 5 6 7 8 9 10"; print "string is $str\n"; $str=~s/(?<=(\d\s){1})\d/I am replaced/; print "string is $str";` [download] Whereas if you want to replace 1st field use `$str=~s/(?<=(\d\s){0})\d/I am replaced/;` [download] Guesses for 10 th field ?? <Reveal this spoiler or all in this thread> More info on look-ahead and look-behind here. The world is so big for any individual to conquer	[reply] [d/l] [select]
Re^2: regex question by johngg (Canon) on Nov 12, 2007 at 11:25 UTC
The problem with the approach you advocate here is that look-behind assertions are fixed width so you can't do something like `(?<=\d+\s)`. If you declare the string like this `my $str="1 2 3 4 5 6 7 8 9 10 11 12 13 14 15";` [download] how do you replace the 12th field? I'm all for look-around assertions but I don't think they are suited to this particular problem. Cheers, JohnGG	[reply] [d/l] [select]
Re^3: regex question by convenientstore (Pilgrim) on Nov 12, 2007 at 15:55 UTC
I thought of look behind but decide against it since field 9 can and will have dynamic variables.. still working on this.	[reply]
Re^4: regex question by convenientstore (Pilgrim) on Nov 12, 2007 at 17:06 UTC
Re^5: regex question by johngg (Canon) on Nov 12, 2007 at 19:21 UTC
Re: regex question by siva kumar (Pilgrim) on Nov 12, 2007 at 09:08 UTC
Split and join should be the good solution.	[reply]

The world is so big for any individual to conquer