UNIX for Dummies Questions & Answers

No, the variables PCF and SCF in the shell and pcf and scf in awk are the names of temp files used by the script to store sorted lists of products and sorted lists of skews for a selected product, respectively, while producing output in the END actions. The input data comes from a single file named file as marked in red in this excerpt from about eight lines from the end of the shell script:

Code:

        close(pcf)
        exit(ec + 0)
}' file
if [ $? -eq 0 ]
then    # awk completed successfully...

Obviously, you can change file to any other filename you want to use. Or you could change it to "$1" and pass the name of the file you want to process as the only argument to the shell script.

This script doesn't need to sort the entire input file, but with millions of input lines these temporary sort result files could still be large. If they are too big to save in the directory where you run this script, you could add an option to the shell script to specify a different directory for these temp files.

Let me know if you're still confused.

---------- Post updated at 19:08 ---------- Previous update was at 18:45 ----------

The code you have above will create the .Product_Counts file used by the script (before sorting it in reverse order by the number of hits for the product and increasing order of product name), but the script produces a .Skew_Counts file for each product entry in the top 3 list. The script never produces the entire list of skews. (You did say that some skews could appear with more than one product.) The lists of skews produced by my script, only show the skew counts for each of the displayed products; skipping all occurrences of the skew for other products.

Note that I developed and tested this on my MacBook Pro. You should have the KornShell available as /bin/ksh on any recent version of OS X.

No kidding, I wasn't aware of that ksh was on mac, never really looked though either.

1 question though for understanding. Would you mind explaining this piece, I know it's calling the products.txt file, just trying to understand it a bit better.

Code:

{       # Process input data...
        # Increment # of times we have seen this product.
        p[$1]++
        if(!(($1, $2) in s))
                # Add a new skew for this product.
                pl[$1, ++plc[$1]] = $2
        # Increment # of times we have seen this skew with this product.
        s[$1, $2]++
}

Thanks again!

I added magenta line numbers to the code above to make it easier to refer to lines for this discussion.

Thank you again!