Manipulating comma delimited CSV - newbie help with sed etc

jonnymorris · 09-19-2008, 03:25 AM

Quote:

Originally Posted by jan61

Moin,

I'd assume you have a syntax error at another part of your script. Did you try the script without the END section?

Jan

The script works fine with the original END loop provided by PTrenholme. Tinkering with it, it appears that when I moved the code around I inadvertently removed some brackets, so the culprit was this part of the line:

Code:

(Record[i,"Name"]!="" ? ", " Record[i,"Name"] : "")

Silly me!

Now it runs, but it seems to give me ALL of the fields instead of just the ones I want ($1, $9 split, $10, $11, $12, $13, $14, $15), the Address array element seems to hold the entire rest of the record. Hmm. Not really sure why this is happening. I will continue to tinker.

EDIT: Ah, I think I see why Address contains so much data. The original input file specification has changed, it contains many fields per record but it is still just the fields I mentioned ($1, $9 - 15) that I want, the rest can be ignored. Just noting this as an update to the original problem.

234107 · 09-19-2008, 06:14 AM

Perl is better

Code:

#!/usr/bin/perl

$csv_file=shift;
open(DAT, $csv_file) || die("Could not open $csv_file!");
@raw_data=<DAT>;

foreach $csv_record (@raw_data)
{
 chomp($csv_record);
  ($f_name,$l_name,$age)=split(/,/,$csv_record);
  print "$f_name\t$l_name\t$age\n";
};

close(DAT);

Example comma separated file. If you want to change it... lets say not comma, but semicolon, just replace it in the split function like this

Code:

($f_name,$l_name,$age)=split(/;/,$csv_record);

Remember to use backslash, if you are going to use dot, slash or some other special symbol.

Code:

($f_name,$l_name,$age)=split(/\\/,$csv_record);

One backslash for backslash and one backslash for escape character.

Code:

John,Smith,23
Eva,Eva,18
No,Namer,35

Usage:
$perl_file_name_it_as_you_wish.pl mufile.csv