Re: ecpg regression test failures caused by window functions patch

Michael Meskes <meskes@postgresql.org> writes:
> On Sun, Dec 28, 2008 at 03:21:16PM -0500, Tom Lane wrote:
>> It strikes me that it might be a good idea to change cat2_str to not
>> insert a space when obviously not necessary, perhaps along the lines
>> of

> Hey, good idea, will add this too.

I just tried that and it seems you'd need to make another couple of
changes --- the "fetch" test fails for instance, probably because
something somewhere isn't recognizing a fetch command.
        regards, tom lane