morcadiss has asked for the wisdom of the Perl Monks concerning the following question:

I have a short perl script which I'd like to open another file for reading in data, then create a new file for each unique identifier (caseid). My script currently does this adequately. However, I can't get the $counter variable to reset with each new file.

I suspect there might be a better way to do this, perhaps a hash or arrays, or a hash of hashes. I'm still trying to figure out the best method, but will settle for something that works.

Sample input file (from test.txt)

CASE ID,ITEM,ENTITY,TYPE,ORDER,TEXT

10001,comments@txt,2,os,1,Monday

10001,comments@txt,7,os,1,Tuesday

10001,comments@txt,8,os,1,Wednesday

10001,comments@txt,10,os,1,

Code is as follows 
#!/usr/bin/perl

use strict;
use warnings;
use Text::CSV;
use Text::Wrap;

$Text::Wrap::columns=100;

my $file = "test.txt";
my @line_list; #list of csid on each line, contains duplicates
my @csid_list; #list of unique csids
#my $oldcsid = 0;

my $csv = Text::CSV->new or  die "Cannot use CSV: ".Text::CSV->error_d
+iag ();
 
open (my $fh, '<', "$file") or  die "Cannot open $file: $!";
 
 $csv->getline ($fh); #skip header row
 
 while ( my $row = $csv->getline( $fh ) ) {
  push @line_list, $row->[0];
 }

 @csid_list = sort {$a <=> $b} uniq(@line_list);

 # Adds title to each file
 foreach my $csid (@csid_list) {
  open (my $csid_fh, '>>', "$csid.txt");
  print $csid_fh "####################\n\nCalYouth Case ID: $csid\n\n#
+###################";
  close $csid_fh;
 }
close $fh;


my $csv2 = Text::CSV->new or die "Cannot use CSV: ".Text::CSV->error_d
+iag ();
 
open (my $fh2, '<', "$file") or 
 die "Cannot open $file: $!";
 
$csv2->getline ($fh2);

# Prints notes to correct csid file 
my $counter = 1; #counts iterations through loop
while ( my $row = $csv2->getline($fh2)) {
 
 foreach my $csid (@csid_list) {
   if ($row->[0] == $csid) {
    if ($row->[5]) {
     my $temp_string = wrap("\t","\t","$row->[5]"); #wrap() is from Te
+xt::Wrap
     open (my $csid_fh, '>>', "$csid.txt");
     
     print $csid_fh "\n\n\t------------------------------\n\nNote Entr
+y $counter:\n\n$temp_string\n\n"; #$counter should count each note in
+ a file
     ++$counter; #increments $counter
     close $csid_fh;
     
    } else {
    next;
    }
    
   }
   
 }
}


# Sorts out duplicates from an array
sub uniq {
 my %seen;
 grep !$seen{$_}++, @_;
}





#More experimental code. Sigh.
 #Reset the counter for new files
#if (-e "$csid.txt"){
# $counter++; 
# print "$csid.txt exists $counter \n";                               
+                                                            
#}else{
# $counter=0;
#}



#Experimental code to handle counter resets. Isn't working.
#if ($oldcsid == 0){
# $oldcsid = $csid;
# $counter++;
#}
#elsif ($oldcsid == $csid) {
# $counter++;
#}
#else{
#$counter = 1;              Endif for ($row->[0] == $csid). I think th
+is will never be false, except on a blank line, which an output file 
+probably won't have, unless data export >1000 chars. However, removin
+g this check broke stuff.
#$oldcsid = $csid;
#}

[download]

Replies are listed 'Best First'.
Re: Unable to get counter to reset
by roboticus (Chancellor) on Jan 13, 2015 at 17:11 UTC

    morcadiss:

    As you guessed, using a hash or array can help out. Here's how I'd do it with a hash. The hash key will be the filename I want to write to:

    my %Files;

sub open_file {
    my $file_name = shift;
    die "$file_name is already open" if exists $Files{$file_name};
    open my $FH, '>', $file_name or die "$file_name open error: $!";
    $Files{$file_name} = { COUNT=>0, FH=>$FH };
}

sub file_handle {
    my $file_name = shift;
    die "$file_name: Hasn't been opened yet" if ! exists $Files{$file_
+name};
    return $Files{$file_name}{FH};
}

sub file_counter {
    my $file_name = shift;
    die "$file_name: Hasn't been opened yet" if ! exists $Files{$file_
+name};
    return $Files{$file_name}{COUNT};
}
    
sub write_to_file {
    my $file_name = shift;
    my @stuff_to_print = @_;
    die "$file_name: Hasn't been opened yet" if ! exists $Files{$file_
+name};
    my $file = $Files{$file_name};
    my $FH = $file->{FH};
    print $FH @stuff_to_print;
    ++$file->{COUNT};
}

for my $file ('foo', 'bar', 'baz') {
    open_file($file);
}

for my $file ('foo', 'foo', 'baz', 'foo', 'bar') {
    my $count = file_counter($file);
    write_to_file($file, "Count is $count\n");
}

    [download]

    Note: untested, yadda yadda...

    ...roboticus

    When your only tool is a hammer, all problems look like your thumb.

[reply]
[d/l]
      Thank you!
[reply]
Re: Unable to get counter to reset
by nlwhittle (Beadle) on Jan 13, 2015 at 17:05 UTC

    There are a few changes I'd like to suggest (such as breaking this code up into functions), but answering your immediate problem, I would create a hash with each key named with a csid. Then every time you process a line, you can increment the hash entry that matches the csid. You'll end up with the hash having a unique count for each file. To create the hash, you can do this:

    my %line_count;
for my $csid (@csid_list) {
    $line_count{$csid} = 0;
}

    [download]

    Then, just before you use the counter variable in your print statement, you can do this:

    ++$line_count{$csid};

    [download]

    This will allow you to keep count on each file, even if the csids in the source file are not in sequence.

    --Nick
[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom