Sorry about the delay in my reply. Real life caught up with me for a bit there. I'll try an answer your questions in order:

Spot on, the variable col_list is "22,23,1,2,4" and the split command separates each of the five values into the col array variable, so:

Code:

col[1]="22"
col[2]="23"
col[3]="1"
col[4]="2"
col[5]="4"

The titles array is just for the output headings the 1 split statement could be replaced with:

Code:

titles[1]="sales_date"
titles[2]="cust_id"
titles[3]="UPC"

This stores that first column in the 2nd file into an array named EXC. And only records in the first files where field #7 is not in the EXC array are processed.

Do you really need comma seperator in the 2nd file? if So you might need to do:

Code:

FILENAME != "-" {sub(/,.*$/,"");if(FNR>1)EXC[$0]; next}
        !($7 in EXC) {

Also, I see from your data files that the exclude field appears to be in field 4, not 7 (you might need !($4 in EXC) {)


This strings column values required from the input line together with pipes into val and directly together into key.

So if your col_list was "1,2,3,5,7,6" and the input line was "one,two,three,four,five,six,seven" then key ends up as "fivesevensix" and val ends up with "five|seven|six".

Prints the headings from the titles array for the additional fields you add to the csv (beyond the first 3).

To remove the heading line from the output replace if(NR!=1) { with if(FNR!=1) {