modify slice in array and write back to file 1

[LinguaFranca]

Hope my question isn't too stupid but I just cannot get it working.

I have a database output like this:

# ---------------------------------------------
# 
# Exported Dictionaries: 
# 
#    C:\TRANSIT.WIN\DB\CODA\INTELL
# 
# Date of Export: 10.01.2005
# 
# Format of Export: (ansi)
# 
#    Concept Number;[];Hdr->Datasource e-Procurement/e-Billing/e-Invoice Matching;Hdr->Datasource CODA-Intelligence;Hdr->OLE;Hdr->Entered By;Hdr->Date of Entry;Hdr->Last Update By;Hdr->Date of Last Update;Hdr->Datasource Workflow;Hdr->Subject Area;Hdr->Datasource Financials;Hdr->Datasource e-Finance;Hdr->Datasource e-Assets;
#    Concept Number;[Language];Term;Definition;Index;Abbr./Acronym;Alternatives;Synonyms;Introduced as of version;Remarks;Hotkey;Irregular Plural;Cross-Reference;Part of Speech;Gender;Context;Source;Status;Entered By;Date of Entry;Last Update By;Date of Last Update;User 1;Valid upto/incl.;Validated by;Additional Status;Former Term;
# 
# ---------------------------------------------

1; []; ; companyparameters.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
1; [ENG]; 'From' cannot be empty.; ; ; ; ; ; v10.1; ; ; ; ; ; ; ; ; confirmed; ; ; CAL; 11.11.2004; ; ; ; ; ; 

2; []; ; cpkpi.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
2; [ENG]; 'From' date or period is after 'To' date or period.; ; ; ; ; ; v10.1; ; ; ; ; ; ; ; ; confirmed; ; ; CAL; 11.11.2004; ; ; ; ; ; 

3; []; ; companyparameters.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
3; [ENG]; 'Group' cannot be empty.; ; ; ; ; ; v10.1; ; ; ; ; ; ; ; ; confirmed; ; ; CAL; 11.11.2004; ; ; ; ; ;

and I need to modify a slice of data within these lines:

1; []; ; companyparameters.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
2; []; ; cpkpi.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
3; []; ; companyparameters.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
5; []; ; kpiwizard.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
6; []; ; genericlists.res, genericlists.res; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ; 
7; []; ; dimensioncostcentre.res, dimensioncustomer.res, dimensionemployee.res, dimensiongeneral.res, dimensionnominal.res, dimensionproduct.res, dimensionproject.res, dimensionsubaccount.res, dimensionsupplier.res, eventviewer.res, dimensioncostcentre.res, dimensioncustomer.res, dimensionemployee.res, dimensiongeneral.res, dimensionnominal.res, dimensionproduct.res, dimensionproject.res, dimensionsubaccount.res, dimensionsupplier.res, eventviewer.res, dimensioncostcentre.res, dimensioncustomer.res, dimensionemployee.res, dimensiongeneral.res, dimensionnominal.res, dimensionproduct.res, dimensionproject.res, dimensionsubaccount.res, dimensionsupplier.res, eventviewer.res, dimensioncostcentre.res, dimensioncustomer.res, dimensionemployee.res, dimensiongeneral.res, dimensionnominal.res, dimensionproduct.res, dimensionproject.res, dimensionsubaccount.res, dimensionsupplier.res, eventviewer.res, dimensioncostcentre.res, dimensioncustomer.res, dimensionemployee.res, dimensiongeneral.res, dimensionnominal.res, dimensionprod; ; ; ; CAL; 11.11.2004; ; CODA-Intelligence; ; ; ;

I'll have to uniquely sort all filenames in $files[3] of the lines pointed out above. When done I need the original output but then with the revised sorted data in it.

How do I define this? Up to now it only outputs the slice with the manipulated data. And my $_ only contains the first line of my input data file.

Here's what I have so far:

#!usr/bin/perl

use warnings;
use locale;
use strict;

open TERMSTAR, "INTELL.ANS" or die;
open OUTPUT, ">intell.output.ans" or die;

while (<TERMSTAR>) {
	chomp;
	
	#grab all header lines from export file
	my @headers=grep /^[0-9]+; \[\]/, <TERMSTAR>; 
	
	#grab 4th slice of a header line
	foreach my $fields (@headers){
          my @files = split(/\; /, $fields);
					my @filelist = split (/, /, $files[3]);
										
					#uniq sort on filenames of slice 4
						my %seen=();
						foreach (@filelist) {
						$seen{$_}=1;
						}
						my @uniqfilenames = sort keys %seen;
						print OUTPUT join (", ", @uniqfilenames), "\n";
						
	}	
}							
close OUTPUT;
close TERMSTAR;
exit;

Not sure whether I am clear enough. Hope one of you guys can help.

 

[rharsh]

This should give you a direction to go in:

while (<DATA>) {
    if (/\d+;\s*\[];/) {
        my @record = split(/;/, $_);
        my %files;
        map {$files{$_}++} split(',', $record[3]);
        $record[3] = join(',', sort keys %files);
        print join(';', @record);
    }
}

 

[LinguaFranca]

Thanks for your quick response. I am sorry for not having answered earlier but I had a very urgent project to set up.
I have a follow up question and will ask you when I have time again to work on my script. Thanks again.

 

[LinguaFranca]

Ha, you put me on the right track. It worked. Thanks you very much for helping me!