One of the ideas that's been talked about in my group is using text
pre-processing to simulate this, e.g., if two queries share the same
select list and group by clauses but differe in having clauses, we
should be able to use comething like the C preprocessor to minimize the
textual repetition at bare minimum.
I've thought about using pipelined functions or returning cursors from
functions, but my gut tells me joining to their output will be slow. I
guess I'll try benchmarking.
Received on Mon Jan 09 2006 - 10:20:27 CST
