Search and Replace a column of the csv file

[Vasuki Masilamani]

Hi,

I have a csv file which has "1" in the column 5. I need to replace all "1" in column 5 to "1030".

The Pattern which would search for "1" in 5th column is given below.

'^"[^,]*","[^,]*","[^,]*","[^,]*","1"'

Can anyone help me with a Search and Replace step for this in Perl Script?

Any solution will be helpful.

Thanks,
Vasuki.

[Vasuki Masilamani]

Please find the below code.

$f1 = 'C:\Vasuki\Repl acePerl\chm_com mitmt.csv';

open(FILE, "$f1") || die "Could not open file chm_commitmt.cs v \n";

$f2= 'C:\Vasuki\Repl acePerl\chm_com mitmt_new.csv';

my @outLines1;

while($data = <FILE>)

{
if($data =~ '^"[^,]*","[^,]*","[^,]*","[^,]*","1"')

{

$data =~ s/^"[^,]*","[^,]*","[^,]*","[^,]*","1"/"[^,]*","[^,]*","[^,]*","[^,]*","1030"/g;

push(@outLines, $data);

}

}

open(FILE2, ">$f2") || die "Could not open file chm_commitmt_ne w.csv \n";

print (FILE2 @outLines);

close (FILE);

close (FILE2);

undef( @outLines );


This code not properly replacing 1 in 5th column to 1030. Can anyone correct the mistake in this??


Thanks,
Vasuki.

[miller]

For easy parsing of CSV files, use this CPAN module or any of the other ones freely available. This will simplify your code and avoid any needless bugs.

http://search.cpan.org/search?query=Text::CSV

[vmalagi]

Hope this may be help

#!/usr/bin/perl

open(FILE,"test.csv");
open(OUTFILE, "newfile.csv");

while($line=<FILE>){

  chomp($line);
  $line =~ s/(.*)?\,(.*)?\,(.*)?\,(.*)?\,(1)(.*)/$1\,$2\,$3\,$4\,1030$6/g;
  print OUTFILE $line ."\n";
}

close FILE, OUTFILE;

One more way to do this

#!/usr/bin/perl

open(FILE,"test.csv") or die $!;
open(TARGET,"> new.csv") or die $!;

while($line = <FILE>){

  chomp($line);
  next unless($line =~ m/\S+/);

  @array = split(",",$line);

  if($array[4] == 1){
    $array[4] = 1035;
  }

  $line = join(",",@array);

  print TARGET $line ."\n";
}
close FILE;
close TARGET;

[GunnarH]

Hope this may be help

#!/usr/bin/perl

open(FILE,"test.csv");
open(OUTFILE, "newfile.csv");

while($line=<FILE>){

  chomp($line);
  $line =~ s/(.*)?\,(.*)?\,(.*)?\,(.*)?\,(1)(.*)/$1\,$2\,$3\,$4\,1030$6/g;
  print OUTFILE $line ."\n";
}

close FILE, OUTFILE;

I doubt that that buggy code is at all helpful. There is about one error or other kind of obvious shortcoming per line:

• No check of the return value from open

• Opening OUTFILE in read mode, even if it's intended for writing

• Unnecessary escaping of commas in both PATTERN and REPLACEMENT

• Unnecessary /g modifier

• Assuming that each column is delimited with dubbelquotes (which is indicated by Vasuki's pattern), the PATTERN does not match

• Incorrect use of the close() function

This is a better attempt (I hope):

#!/usr/bin/perl
use strict;
use warnings;

my $f1 = 'C:\Vasuki\ReplacePerl\chm_commitmt.csv';
my $f2 = 'C:\Vasuki\ReplacePerl\chm_commitmt_new.csv';

open my $file1, '<', $f1 or die "Couldn't open $f1: $!";
open my $file2, '>', $f2 or die "Couldn't open $f2: $!";

while (<$file1>) {
	print $file2 $_ if s/^((?:"[^,]*",){4})"1"/$1"1030"/;
}