Quicken didn’t want to do what I needed, which was to total all my expenses by place of origin for the year to date. Quicken only sorts on category, but I long since gave up keeping Quicken up to date since I don’t want to message around with Web access to fetch my expenses from multiple places.

Enter the shell.

First I want all my comma separated value files in a single file.

for e in $(ls *.csv) ; do cat $e >> all.csv ; done

Next, I want to process them, sorting them uniquely by point of sale location.

cat all.csv | grep '\-[0-9]' | sort -T , -k 3 | awk -F , '{ print $3,",",$5 }' > my.out 

What does that process?

03/27/2004,2354,"DOMINO'S PIZZA #05 ","GAINESVILLE   FL ",   -9.99
03/27/2004,7041,"PERKINS REST &08","GAINESVILLE   FL ",  -18.00
03/29/2004,8148,"KASH N KARRY #6","GAINESVILLE   FL ",  -16.23

Finally, some very sloppy Perl code will clean up my mess. (Thanks to WP, replace any [ you see with an actual greater than sign.)

while(<>) {
 $_ =~ m/"([^,]*)"&\s+,\s+-(\d+&\.\;d+)/;
 $a{$1}+=$2;
 print "$1 $2 $a{$1}&\n";
}
while(my($a,$b)=each %a) {
        format STDOUT =
         @[[[[[[[[[[[[[[[[[@#####.##
         $a, $b
.
        write;
}

To those who’s nasty looking Perl I comment on all the time… Don’t even start with me… My sloppy code owns you. (And apparently WP likes to eat front slashes and other characters, so I had to hunt down the appropriate HTML escape sequence.)

The output is rather nice, though. (Amounts changed to protect my favourite merchants’ feelings.)

KASH N KARRY #1786    14.71
CABLE & CONNECTOR     60.00
ITALIAN GATOR PIZZ    40.20
OFFICE DEPOT #2145    57.94
NEWEGG COMPUTERS    2031.47