"I have seen csv::text cpan , but it does not talk about extraction of any column as individual"

Yes it does! You couldn't have read very far: it's in the first few lines of Text::CSV SYNOPSIS:

    $row->[2] ... # 3rd field ...
[download]

use Data::Table;
my $t = Data::Table::fromCSV("aaa.csv");
my $c = $t->subTable(undef, ['MyColumnHeader']);
my $column_data_ref = $c->data;
[download]

And just to prove it ;o)

use strict;
use warnings;
use utf8;

use Text::CSV;
my $io = \*DATA;
my $csv = Text::CSV->new ({ binary => 1 });
while (my $row = $csv->getline($io)) {
        print $row->[1], "\n";
}

__DATA__
33,44,55,"ABC"
234.33,"ABC ""DEFGHIJ"" KLMNOP","asdfasd asdfasdfasdf "
[download]

44
ABC "DEFGHIJ" KLMNOP
[download]

Hey

what if I define it as any sub routine that extracts a particular column so that i can return the extracted array

Hey

thank you for helping out , I have just one question that here in line

 print $row->[1], "\n";

our 1 is fixed but i want to extract multiple columns , so I wrote a subroutine

sub column_segregation
{
    my $o=$_[0];
    print "$o \n";
my @array_A2=();    
    my $csv = Text::CSV->new ({ binary => 1 });
while (my $row = $csv->getline($io)) {
        push @array_A2, $row->[$o];
}
   return (@array_A2);                                     
}
[download]

but here $o is changing but it constantly returns the same array , which is at location 2. can you please tell me what is wrong in my code

 # get the 4th column
 my @column = map { $_->[3] } @{$csv->getline_all ($fh)};
[download]

hey i did this as per your advice

 my $in_file = 'C:\Documents and Settings\x0199665\My Documents\coll\t
+gb.csv';     #
open my $fh, '<', $in_file or die "could not open $in_file: $!\n"; 
 my @column = map { $_->[4] } @{$csv->getline_all ($fh)};
[download]

but it gives an error "can't call getline on an unidentified value

Please start reading the documentation for Text::CSV_XS or Text::CSV. To me this error indicates that you forgot to initiate the $csv object (or that you are using a very old Text::CSV):

use Text::CSV_XS; # or use Text::CSV;
my $csv = Text::CSV_XS->new ({ binary => 1, auto_diag => 1 });
[download]

update: can I see both the script and the data? Not much has changed for getline_all () in Text::CSV_XS since Dec 2010, when it was introduced


Enjoy, Have FUN! H.Merijn

On one hand, Text::CSV doesn't provide a built-in function to pull out just one specific column from a CSV file. On the other hand, you can use that module to do just that. In fact, I've written code to do something like that for use at work. The basic concepts are to parse the CSV one line at a time within a loop and then pull the desired column to store/print/etc.

I have seen csv::text cpan , but it does not talk about extraction of any column as individual
You mean Text::CSV?
And how are you parsing your csv file?

This is an alternative method to the Text::CSV techniques suggested above.

If you're familiar with databases, you can use DBD::CSV to query the CSV file as a database and select the fourth column of data.

Why pretend? Assuming these two csv lines:

11,22,33,"ABC"
1111,3333,4444,"asdfasd asdfasdfasdf"
[download]

And assuming you have the file_fdw extension installed ( create extension file_fdw; ), you could have Pg query it via:

create server pgcsv foreign data wrapper file_fdw;

create foreign table pgcsv (
   col1 int,
   col2 int,
   col3 int,
   col4 text
) server pgcsv
options ( filename '/tmp/pm-1040691.csv', format 'csv' );

select * from pgcsv; -- or a single column, of course...
[download]

which will output:

CREATE SERVER
CREATE FOREIGN TABLE
 col1 | col2 | col3 |         col4
------+------+------+----------------------
   11 |   22 |   33 | ABC
 1111 | 3333 | 4444 | asdfasd asdfasdfasdf
[download]

Obviously not the easiest way to go about it (and on top of that, postgres is rather picky about the text's precise format). I'm just showing this because you brought up "database" ;)

Well thats one idea. I presume though you'd need the psql client installed on your PC. Your answer is ingenious but not Perl :-)

If you spot any bugs in my solutions, it's because I've deliberately left them in as an exercise for the reader! :-)