perl matching script

saltydog4791 · 12-03-2007, 11:52 PM

Hello all,

I am new to all of this so be gentle.

I am working on a script that will match 2 files that will always have the same naming scheme and cat one to the other.

Basically I will have files like

102907-cf.txt
103007-cf.txt
103107-cf.txt
110107-cf.txt

and

TWUS 24K 102907-cf.txt
TWUS 24K 103007-cf.txt
TWUS 24K 103107-cf.txt
TWUS 24K 110107-cf.txt

I need to cat the TWUS files to the matching other file based on the month and day. So TWUS 24K 102907-cf.txt would be cat'd to the end of 102907-cf.txt, TWUS 24K 110107-cf.txt would be cat'd to end of 110107-cf.txt, etc.

After the cat'ing is done I have a piece of perl code that "pretties" the file name into a file like "November 01, 2007". That code is this:

Code:

 
#!/usr/bin/perl 
 
#use strict; 
use warnings; 
 
my %mons = ("01", January, "02", February, "03", March, "04", April, "05", May, "06", June, "07", July, "08", August, "09", September, "10", October, "11", November, "12", December); 
 
 
foreach my $filename (@ARGV) 
{ 
    if ($filename =~ /^(\d\d)(\d\d)(\d\d)\-cf.txt$/) 
    { 
        my $month = $1;  
        my $day = $2;  
        my $year = $3;  
		 
        $year += 2000; 
        my $newFilename = "$mons{$month} $day, $year.txt";  
 
        system "mv $filename '$newFilename'"; 
    } 
}

I would love to some how merge this into one perl script. Right now I have two separate folders for each type of file with some bash script that does the cat with a wildcard that can only work if there is only one file in each directory. This clearly is not the right way to do it. I need a way to process multiple files based on matching and then doing the renaming all in the same perl script.

I know it's possible but I am stumped and would love some assistance from the community. Thank you so much.

ghostdog74 · 12-04-2007, 12:45 AM

not a perl script,

Code:

awk 'BEGIN{
    #FS="-"
    month["01"]="January"
    month["02"]="February"
    month["03"]="March"
    month["04"]="April"
    month["05"]="May"
    month["06"]="June"
    month["07"]="July"
    month["08"]="August"
    month["09"]="September"
    month["10"]="October"
    month["11"]="November"
    month["12"]="December"
}
{

    n=split(FILENAME,f,"-")
    m=split(f[1],name," ")
    filename = name[3]"-cf.txt"
    cmd="test -f " filename
    ret = system(cmd)
    if ( ret == 0 ) {
       print $0 >> filename
    }
    m = substr(name[3],0,2)
    d = substr(name[3],3,2)
    y = substr(name[3],5,2)
    newfile = month[m]" "d","y    
    cmd = "mv \047" FILENAME "\047 \047" newfile"\047"
    print cmd
    #system(cmd) #uncomment to rename file.
    nextfile
}
' TWUS*

chrism01 · 12-04-2007, 04:47 PM

The system cmd should be, if I understand the requirement

system(cat 'TWUS 24K '.$filename >> $newFilename");

Can i just say that's a horrible format for a filename btw. Most Unix tools will have problems with spaces in the filename unless you take special precautions. Not homework by any chance?

Quigi · 12-05-2007, 02:01 PM

Agree with chrism01 about spaces, and the system() call for concatenating (his misses the opening double quote.) If you choose a pretty name like "2007-11-01.txt", alphabetical order will also be chronological. For renaming in Perl, you don't need the clunky "system" hammer. Taking this together, the body of your loop becomes

Code:

    if ($filename =~ /^(\d\d)(\d\d)(\d\d)-cf.txt$/) 
    { 
        my $newFilename = "20$3-$1-$2.txt";  
        rename $filename, $newFilename or die $!;
        system("cat 'TWUS 24K '.$filename >> $newFilename");
    }

Take the parts you like :-) Maybe add error checking to the system() call.

chrism01 · 12-05-2007, 04:52 PM

Y, missing double quote was a typo. That's why I always use -w and use strict...