Re: Aggregate ORDER BY patch

>>>>> "Tom" == Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> writes:

>> What about error handling? If the user specifies agg(distinct x)
>> where x is not sortable, do we leave it to the planner to detect
>> that (which means not reporting the error position?)

Tom> Well, at the moment there's only going to be a sort-based
Tom> implementation, so I don't object to throwing an error for that
Tom> as soon as possible. OTOH I wouldn't recommend expending a lot
Tom> of code to do it there. I would hope that most of the parser's
Tom> work for this can be shared with the existing support for
Tom> query-level ORDER BY/DISTINCT.

The code already uses transformSortClause for most of the work, but
reusing the existing code for DISTINCT would have required more
refactoring than I was happy with, because transformDistinct etc.
all have error message text which is specific to SELECT DISTINCT etc.
Let's see how it falls out in the next patch.

--
Andrew.