How do I use grep in a script

[Flintlock]

I am trying to extract a specific portion of a line found in a file to create a file name. I have this working for command line

grep Acct: Facs_Data.txt |cut -d":" -f2 |cut -d" " -f1

it returns the right half of the search i.e. Acct:1234 returns 1234. and I would then use this 1234 to create a file named 1234_2017.txt

So my question is: How can I achieve the same thing in Perl from opening a file that has more than one entry of Acct: ?

[Corion]

Converting the steps from a shell pipeline to Perl is fairly easy. The steps are:

1. grep Acct: Facs_Data.txt - Read a file line by line and select the lines matching Acct:
2. cut -d":" -f2 - Take the side to the right of : of the line
3. cut -d" " -f1 - take the side to the left of the blank of the line

If we convert each step to Perl, we get the following parts:

1. my $filename = 'Facs_Data.txt';
   open my $fh, '<', $filename
       or die "Couldn't open '$filename': $!";
   
   my @lines = grep { /Acct:/ } <$fh>; # Read a file line by line and sel
   +ect the lines matching Acct:
   

2. @lines = map { [ split $_, /:/ ]->[1] } @lines; # Take the side to the
   + right of : of the line
   

3. @lines = map { [ split $_, /:/ ]->[0] } @lines; # take the side to the
   + left of the blank of the line
   

But a more perlish approach would be to do all that in one go:

my $filename = 'Facs_Data.txt';
open my $fh, '<', $filename
    or die "Couldn't open '$filename': $!";

my @lines = map  { /:([^\s]+)/ ? $1 : () } # take the stuff between th
+e : and the first blank
            grep { /Acct:/ } <$fh>; # Read a file line by line and sel
+ect the lines matching Acct:

# do whatever with the values in @lines
print "$_\n" for @lines;

Update:: Fixed missing = in step three, spotted by AnomalousMonk, thanks.

[ikegami]

map  { /:([^\s]+)/ ? $1 : () } grep { /Acct:/ }

is a long way to write

map  { /Acct:([^\s]+)/ }

[Flintlock]

thank you - using your example, I have the following

#! / usr/bin/perl -w

use strict;
use warnings;

use POSIX 'strftime';

my $nm = strftime('%Y', localtime).".txt";

my $filename = 'Facs_Data.txt';
open my $fh, '<', $filename
    or die "Couldn't open '$filename': $!";

my @lines = map  { /:([^\s]+)/ ? $1 : () } # take the stuff between th
+e : and the first blank
            grep { /Acct:/ } <$fh>; # Read a file line by line and sel
+ect the lines matching Acct:
        my $outfile = $_."_"."$nm" for @lines ;
        print $outfile;
        #open (OUTFILE, ">$outfile");
# do whatever with the values in @lines
#print "$_\n" for @lines;
close($filename);

root@localhost:~/GoldenLFiles# perl facs.pl Use of uninitialized value $outfile in print at facs.pl line 17, <$fh> line 168801.

If I comment the print $outfile to see the output I get the error above. What am I missing here? Thanks in advance

[ForgotPasswordAgain]

I think you're looking for backticks (`...` or qx): Quote Like Operators

[Laurent_R]

The OP wants to do it in Perl. I'm afraid it is somewhat counterproductive to suggest shelling out of Perl with backticks or qx//.

[ForgotPasswordAgain]

Though there's value in learning how to do it in Perl, I respectfully disagree about the "counterproductive" part. The one-liner being done, just wrap backticks around it, use a split /\n/, and that's that (unless there are other constraints). I wouldn't be surprised if it were also faster, though that depends and might not matter.