in reply to How to fetch the value of unix command result

I am trying to fetch the number 3 from 3 CHECK and 1 from 1 KCEHC,i want store the numbers in a variable separately

Do you mean you want to keep track of how many occurrences of 'CHECK' there are, versus how many of 'KCEHC' there are? A hash would be particularly well suited to this. I will give an example of that in a second, but I'd also like to recommend the following pure-Perl way to accomplish the same thing in your case, as grabbing the same data out of your unix command pipeline output would actually be just as difficult as just generating it in Perl, except it would be longer, less portable, and (probably) less efficient:

use warnings;
use strict;
use IO::Zlib;
 
my $fh = IO::Zlib->new('test.txt.gz', 'rb') or die "Zlib failed: $!";
 
my %count;
$count{$_}++ for map { (split /\|/)[2] } <$fh>;

printf "%-10s %4d\n", $_, $count{$_} for sort keys %count;

[download]

Output:

CHECK         3
KCEHC         1

[download]

The following line is the one that answers your question most directly:

    $count{$_}++ for map { (split /\|/)[2] } <$fh>;

Basically, this can be translated in English as, read a line from $fh, split it on | characters, grab the 3rd field from that (0-based numbering) and then increment (++) the corresponding hash element in %count. ($_ is an automatic variable used in many Perl loop constructs).

Replies are listed 'Best First'.
Re^2: How to fetch the value of unix command result
by rajsai28 (Novice) on Jun 30, 2013 at 11:46 UTC
    Hi Experts,

    I got the below output when i run the unix command on zip file, this count of CHECK and RECHECK

     2
   4 CHECK
   7 RECHECK

    [download]

    I want to store the value 4 and 6 in a scalar variable as follows

    my $c = 4;
my $c1 =6;

    [download]

    This is my code shown below

    my $r = `gunzip -c test.txt.gz|cut -f3 -d'|'|sort|uniq -c`;

    [download]

    I want to store the values 4 and 6 separately in two separate variables

[reply]
[d/l]
[select]

      I want to store the value 4 and 6 in a scalar variable as follows

      my $c = 4;
my $c1 =6;

      [download]

      Why would you want to go to all the trouble to declare cryptic, arbitrary variables like $c and $c1 when you've already been shown how to dynamically create hash entries that directly map to the keywords in your input file? More to the point, what part of my suggestion was either unacceptable or unclear?

      This is my code shown below

      my $r = `gunzip -c test.txt.gz|cut -f3 -d'|'|sort|uniq -c`;

      This is also the same code you posted in your original node, which makes me wonder how much of an effort you made to either work with my suggestion or continue on your own.

[reply]
[d/l]
[select]
      Hi Experts,

      Below is my code, unix command executes on zip file and displays the count of field 3, based on the value in field 3, i have 10 records in zip file, the field 3 contains word CHECK and KCEHC, the command which is used counts the number of CHECK and KCEHC

      my $result1 = `gunzip -c $file|cut -f3 -d'|'|sort|uniq -c`;

      [download]

      The input zip file contains, the 11 records

      ABC|123|CHECK|1|
DEF|456|CHECK|1|
GHI|789|CHECK|1|
JKL|101|CHECK|1|
ABC|123|CHECK|1|
DEF|456|CHECK|1|
GHI|789|CHECK|1|
JKL|101|KCEHC|1|
ABC|123|KCEHC|1|
DEF|456|KCEHC|1|
GHI|789|KCEHC|1|

      [download]

      gives me this error message / faulty output:

      2
   4 CHECK
   6 KCEHC

      [download]

      I need to store the values 4 in $c and 6 in $c2, could you please suggest how can we do that

      I also tired the below one....

      my $r = `gunzip -c $file|cut -f3 -d'|'|sort|uniq -c`;
my $c1  = substr( $r, 16,-2);
my $c2  = substr( $, 9,-9);

      [download]

      This gives me the output as...

      <code> $c1 = 4 $c1 = 6 <code>

      I have a question to experts, was this correct method

[reply]
[d/l]
[select]

        I also tired the below one....

        my $r = `gunzip -c $file|cut -f3 -d'|'|sort|uniq -c`; 
my $c1 = substr( $r, 16,-2); 
my $c2 = substr( $, 9,-9);

        [download]

        This gives me the output as...
        <code> $c1 = 4 $c1 = 6 <code>

        I have a question to experts, was this correct method

        In a word, no. It might (and I emphasize, might) work for your current input. What happens, though, when:

        • You need to add, remove, or change one of the keywords you are interested in?
        • Someone gives you a data file containing an entry like this: GHI|789|DHECK|1|
        • Pretty much anything changes with the format of the output from that pipeline?
        • Someone asks you, "hey, why doesn't this work on Windows?"
        • You forget whether $c1 has the value for CHECK...or was it $c2?
        • ...

        Maybe you'll be ridiculously lucky and not ever encounter any of the above scenarios. But I guarantee, if you continue writing code that takes inherently variable width text input and parses it with fixed offsets, it will blow up horribly at some point.

[reply]
[d/l]
[select]

In Section Seekers of Perl Wisdom