Global vs. Local Variables Assistance Requested

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

Dear Most Merciful and Benevolent Monks,

I come with utter supplication and freely admit my utter ignorance and sloth. I plead with you to have extreme mercy and patience so as to accept my most basic request.

I have a text file called funtime.dat that is comma separated values. Here is a small snippet:

IM,BEN01,D,0
IM,BEN02,D,0
IM,BEN03, ,0
IM,BEN04, ,0
IM,BEN05, ,0
IM,BEN06,C,0
IM,BEN07, ,0
IM,BEN08, ,0
IM,BEN09, ,0
IM,BEN10,D,0
[download]

My Perl code is thus and works with the following output included below:

#!/usr/bin/perl
open(MYINPUTFILE, "<funtime.dat");
while(<MYINPUTFILE>)
   {
    my($line) = $_;    
    chomp($line);
    @data = split(/,/, $line);
    if ((@data[2] eq "D")||(@data[2] eq "C"))
        { 
            printf ("\n type ");
            print @data[2]." ";
            print @data[1].".txt ";
        }
    else
        {
        print @data[1].".txt ";  
        }
    }
close(MYINPUTFILE);
[download]

My output is

type D BEN01.txt 
type D BEN02.txt BEN03.txt BEN04.txt BEN05.txt
type C BEN06.txt BEN07.txt BEN08.txt BEN09.txt
type D BEN10.txt
[download]

Actually what I want is it to concatenate the files into the first file name:

type D BEN01.txt > BEN01.txt
type D BEN02.txt BEN03.txt BEN04.txt BEN05.txt >BEN02.txt
type C BEN06.txt BEN07.txt BEN08.txt BEN09.txt > BEN06.txt
type D BEN10.txt > BEN10.txt
[download]

Please disregard the C and D as those I use for debugging purposes. Obviously I would then pipe the output of the perl program into a .BAT program which I would execute in MS-DOS.

I believe I should be using some global values instead of local values or something along those lines. In terms of when the value of the line is a "D" or a "C", it gets text file stored as a value until another "D" or "C" value is read and then it prints out a "> textfile.txt" and the makes a newline and prints out "type" and so on.

If this is too difficult how about using the last file before the document has a "D" or "C" to pipe it out to such as this following example:

 
type D BEN01.txt > BEN01.txt
type D BEN02.txt BEN03.txt BEN04.txt BEN05.txt > BEN05.txt
type C BEN06.txt BEN07.txt BEN08.txt BEN09.txt >BEN09.txt 
type D BEN10.txt  > BEN10.txt
[download]

You might ask why don't I just concatenate the files in Perl. The thing is I am getting the .dat file and therefore it is an extra step. These .dat files can actually be 1,000,000+ lines so that is why I am using Perl to automate this process.

Thank you so much for your assistance.

Comment on Global vs. Local Variables Assistance Requested Select or Download Code

Replies are listed 'Best First'.
Re: Global vs. Local Variables Assistance Requested by ikegami (Patriarch) on Dec 04, 2008 at 22:35 UTC
Like this? `#!/usr/bin/perl my $head; open(my $fh, '<', "funtime.dat") or die $!; while(my $line = <$fh>) { chomp($line); my @data = split(/,/, $line); if (($data[2] eq "D")\|\|($data[2] eq "C")) { print(" > $head.txt\n") if defined($head); print("type $data[2]"); $head = $data[1]; } print(" $data[1].txt"); } print(" > $head.txt\n") if defined($head);` [download]	[reply] [d/l]
Re^2: Global vs. Local Variables Assistance Requested by Knoperl (Acolyte) on Dec 04, 2008 at 22:55 UTC
Dear ikegami, That was perfect. Thank you very much! How quick is the wonderful PerlMonks in their righteousness!	[reply]
Re: Global vs. Local Variables Assistance Requested by GrandFather (Saint) on Dec 04, 2008 at 22:53 UTC
Yes, you need a variable that is global to the while loop because you need to preserve some state between iterations of the loop. Consider: use strict; use warnings; my $inFileStr = <<STR; IM,BEN01,D,0 IM,BEN02,D,0 IM,BEN03, ,0 IM,BEN04, ,0 IM,BEN05, ,0 IM,BEN06,C,0 IM,BEN07, ,0 IM,BEN08, ,0 IM,BEN09, ,0 IM,BEN10,D,0 STR my $first; open my $inFile, '<', \$inFileStr or die "Can't open input file: $!\n" +; while (<$inFile>) { chomp; my @data = split /,/; if ($data[2] =~ /^(D\|C)$/) { print " > $first.txt\n" if defined $first; print "type $data[2] "; $first = $data[1]; } print "$data[1].txt "; } close $inFile; print " > $first.txt\n" if defined $first; [download] Prints: `type D BEN01.txt > BEN01.txt type D BEN02.txt BEN03.txt BEN04.txt BEN05.txt > BEN02.txt type C BEN06.txt BEN07.txt BEN08.txt BEN09.txt > BEN06.txt type D BEN10.txt > BEN10.txt` [download] There are a few things that I've tidied up that you should take note of. First off, always use strictures (use strict; use warnings;). Use the three parameter version of open and check the result (using die is pretty standard for that). Use lexical variables (my $inFile) for file handles. printf is not print - do not confuse them or you will likely have an unhappy life. Regular expressions are clearer (when you get used to them) for non-trivial string matching (see perlretut and perlre). Perl's payment curve coincides with its learning curve.	[reply] [d/l] [select]
Re^2: Global vs. Local Variables Assistance Requested by Knoperl (Acolyte) on Dec 04, 2008 at 23:03 UTC
Grandfather, Please accept my sincere gratitude for your very kind advice and assistance. I will be more careful and mindful in the future.	[reply]
Re: Global vs. Local Variables: try using more memory by Narveson (Chaplain) on Dec 05, 2008 at 06:28 UTC
Consider making slightly more use of your program's memory. my %command; my $destination; while(<DATA>) { # assign split results to named scalars my ($ignored, $file, $flag) = split /,/; if ($flag =~ /^[CD]$/) { $destination = $file; $command{$destination} = "type $flag"; } $command{$destination} .= " $file.txt"; } for my $destination (sort keys %command) { # Look, only one print statement! print "$command{$destination} > $destination.txt\n"; } __DATA__ IM,BEN01,D,0 IM,BEN02,D,0 IM,BEN03, ,0 IM,BEN04, ,0 IM,BEN05, ,0 IM,BEN06,C,0 IM,BEN07, ,0 IM,BEN08, ,0 IM,BEN09, ,0 IM,BEN10,D,0 [download]	[reply] [d/l]