The Reporting Problem

[Peter Alcibiades]

We Linux users do not have access to full versions of Quartam, so have to 
contain our disappointment manfully and find another way.

Considering the question of how to get some reports done, the mind  then 
turns inevitably to awk.  Or Perl, but for us antedeluvians, its awk.  
What do you experienced Rev programmers do, because its not possible you 
too turn to awk!?

Here is the problem.  I have a set of reports that come out quite nicely, 
though they take several tedious pages of code.  What they do is put each 
total by revenue/product category/month into a field on a card.  Then we 
print the card, with totals.  So for instance, we have handtools for Jan, 
Feb....etc, and powertools for jan...feb...etc.  Then a subtotal for all 
tools, then a total of all the subs to get total revenue by month and YTD.

This is done from a transaction file of several tens of thousands of 
records, using filter to get the categories and months and subtotal 
elements, then a simple add into a variable.  Brute force, but it works 
reasonably fast on a low spec machine and is accurate and easy to 
understand whats been done.

The problem comes when you want to export it into a word processor or 
spreadsheet for formatting and further analysis.  This is why the 
antiquarian's mind turns nostalgically to awk.  What awk puts out is a 
somewhat formatted text file, which can just be imported into anything you 
like.  What we have generated in Rev is a sort of binary blob!

Now, we could just go through and export each field from the card into a 
text file.  But this is going to make the whole thing even longer and more 
tortuous, and you start to feel, wait a second, if this is the solution to 
the problem why don't I just do it in awk in the first place and get it 
over with?

So what is the approved solution?  Do you indeed go through and export your 
fields?  Or do you generate the report not in fields on a card but as an 
rtf in one text field, and if so how?  Or, he wonders smiling, are there 
actually still some shamefaced users of awk out there, secretly writing 
the stuff in their studies at home where none of their colleagues can see 
what they are up to?  

BEGIN 
{printf("%-15s\t%-40s\t%7s\t%4s\t%5s\n", "Code", "Name" , "Price", "Dept", "Qty")}
{FS="\t"}
{OFS="\t"}
{printf("%-15s\t%-40s\t%7s\t%4s\t%5s\n",$1, $2, $3, $4, $5)}

and so on.  Ah the sweet smell of cordite in the morning!

Peter