Gregory Stark <stark@enterprisedb.com> writes:
> "Tom Lane" <tgl@sss.pgh.pa.us> writes:
>> Also, the fact that this particular form of the grammar requires
>> reserving the keywords does not prove that there is no way to have the
>> features without that. I'd want to see us try a little harder first.
> At least some of them are unavoidable conflicts:
> select a,b,count(*) from tab GROUP BY cube(a,b)
> select a,b,count(*) from tab GROUP BY rollup(a,b)
In principle this case could be handled by having parse analysis
treat a FuncCall node at the top level of GROUP BY specially when
the function name is CUBE etc.
I'm not saying that might not be uglier than having the grammar
recognize it, just pointing out that there's usually more than
one way to skin a cat.
regards, tom lane