Re: select .... where 1 = 2

aaron@silo-alcazar.com (Aaron Pelton) writes: > I have an ace report which is used to print an invoice based on 2 > different scenarios. In one case, I want to print all unprinted > invoices. In another, I want to print a specific invoice. At the > moment I'm trying to be clever and use the same report for both > instances. So, what I ended up with is (grossly oversimplified): > select field1, field2 from a,b > where f3 = f4 and printed_date is null and "A" = $pass_param dummy_inv > union > select field-from-as-above > where f3 = f4 and invoice num = $inv and "S" = $pass_param 312 > My thought was: with the the report called as "sacego blah A #", I'd > get the first select to work and "sacego blah S #" the second select > would work, the joins creating a single set without temp tables. Perhaps I'm missing something, but couldn't you do something silly like what's below? I agree though with the person who suggested splitting it into two separate reports. The code below should work, but it's highly inefficient for what (I think) you're trying to accomplish. define param[1] blah char(1) variable is_group integer end select 1 which_param, field1, field2 from a, b where printed_date is null... union select 2 which_param, field1, field2 from a, b ... where invoice_num=$inv end format first page header let is_group=0 before group of which_param if blah = "A" then let is_group=1 else if blah = "B" then let is_group=2 on every row if which_param=is_group then print field1, field2 ... end -- -------------------------------------------------------------------------------- Daniel James - Database Administrator djames@statsci.com StatSci, a div of MathSoft, Inc. (206) 283-8802 x284