Re: Virtual generated columns

On Sat, Feb 22, 2025 at 11:12 PM Richard Guo <guofenglinux(at)gmail(dot)com> wrote:
>
> On Sat, Feb 22, 2025 at 11:55 PM Richard Guo <guofenglinux(at)gmail(dot)com> wrote:
> > Attached are the updated patches to fix all the mentioned issues. I
> > plan to push them early next week after staring at the code for a bit
> > longer, barring any objections.
>
> Sign... I neglected to make the change in 0001 that a Var newnode
> compares its varlevelsup with 0 when deciding to wrap it in a
> ReturningExpr. I made this change in 0002 though, so maybe we're good
> here. Still, I'll fix this later.
>
i also noticed this issue...

some minor comments about v7.

* In order to be able to cache the results, we always generate the
* expansion with varlevelsup = 0. The caller is responsible for
* adjusting it if needed.
*
expandRTE(target_rte,
var->varno, 0 /* not varlevelsup */ ,
var->varreturningtype, var->location,
(var->vartype != RECORDOID),
&colnames, &fields);
the above comments should be put on top of ReplaceVarFromTargetList?
so people can easily catch it.
when using ReplaceVarFromTargetList,
they’ll be aware that they might need to call IncrementVarSublevelsUp
in the caller.

src/include/nodes/primnodes.h
* ReturningExpr nodes never appear in a parsed Query --- they are only ever
* inserted by the rewriter.
*/
typedef struct ReturningExpr
this comment needs to change?

on top of src/test/regress/sql/generated_virtual.sql, we have:
-- keep these tests aligned with generated_stored.sql

but gtest32 is only related to virtual generated column.
maybe add a comment saying gtest32 related tests do not
apply to stored generated column.