Re: Virtual generated columns

On Thu, Aug 29, 2024 at 9:35 PM jian he <jian(dot)universality(at)gmail(dot)com> wrote:
>
> On Thu, Aug 29, 2024 at 8:15 PM Peter Eisentraut <peter(at)eisentraut(dot)org> wrote:
> >

> >
> > The new patch does some rebasing and contains various fixes to the
> > issues you presented. As I mentioned, I'll look into improving the
> > rewriting.
>
>
> based on your latest patch (v4-0001-Virtual-generated-columns.patch),
> I did some minor cosmetic code change
> and tried to address get_attgenerated overhead.
>
> basically in expand_generated_columns_in_query
> and expand_generated_columns_in_expr preliminary collect (reloid,attnum)
> that have generated_virtual flag into expand_generated_context.
> later in expand_generated_columns_mutator use the collected information.
>
> deal with wholerow within the expand_generated_columns_mutator seems
> tricky, will try later.

please just ignore v4-0001-Virtual-generated-columns_minorchange.no-cfbot,
which I made some mistakes, but the tests still passed.

please checking this mail attached
v5-0001-Virtual-generated-wholerow-var-and-virtual-che.no-cfbot

It solves:
1. minor cosmetic changes.
2. virtual generated column wholerow var reference, tests added.
3. optimize get_attgenerated overhead, instead of for each var call
get_attgenerated.
walk through the query tree, collect the virtual column's relation
oid, and the virtual generated column's attnum
and use this information later.

I will check the view insert case later.