Re: Virtual generated columns

Peter Eisentraut <peter@eisentraut.org>

From: Peter Eisentraut <peter@eisentraut.org>
To: jian he <jian.universality@gmail.com>
Cc: pgsql-hackers <pgsql-hackers@postgresql.org>, Dean Rasheed <dean.a.rasheed@gmail.com>
Date: 2024-11-29T09:50:19Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →
  1. Expand virtual generated columns for ALTER COLUMN TYPE

  2. Eliminate code duplication in replace_rte_variables callbacks

  3. Expand virtual generated columns in the planner

  4. Virtual generated columns

  5. Additional tests for stored generated columns

  6. Improve generated_stored test

  7. Fix handling of CREATE DOMAIN with GENERATED constraint syntax

  8. Add pg_constraint rows for not-null constraints

  9. Put generated_stored test objects in a schema

  10. Rename regress test generated to generated_stored

  11. Small code simplification

  12. Remove useless code

  13. Remove useless initializations

  14. doc: Clarify that pg_attrdef also stores generation expressions

  15. Clean out column-level pg_init_privs entries when dropping tables.

  16. Re-implement the ereport() macro using __VA_ARGS__.

On 13.11.24 04:30, jian he wrote:
> in
> transformColumnDefinition
> we can add parser_errposition for the error report.
>          if (column->is_not_null && column->generated ==
> ATTRIBUTE_GENERATED_VIRTUAL)
>              ereport(ERROR,
>                      (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
>                       errmsg("not-null constraints are not supported on
> virtual generated columns"),
>                       parser_errposition(cxt->pstate,
> constraint->location)));
> sometimes, it points to the word "generated", sometimes "not". I guess
> this should be fine.
> example:
> create table t13 (a int, b bool generated always as ((true )) VIRTUAL not null);
> create table t13 (a int, b bool not null generated always as ((true )) VIRTUAL);

Ok, done in v10.

> These 3 functions will call StoreRelNotNull to store the not-null constraint.
> StoreConstraints
> AddRelationNotNullConstraints
> AddRelationNewConstraints

> related tests:
> create table t12(b int, a int generated always as (11) virtual,
> constraint nn not null a);
> create table t12(b int, constraint nn not null a, a int generated
> always as (11) virtual);
> 
> drop table if exists t14;
> create table t14(b int, a int generated always as (11) virtual);
> alter table t14 add constraint nn not null a;
> alter table t14 add constraint nn not null a no inherit;

Ok, I have added the missing checks and added these test cases to v10.

I didn't put the checks in StoreRelNotNull(), I think that is too late 
in the process.  (It's already trying to store it.  The checking should 
come earlier.)  I put the checks into AddRelationNewConstraints() and 
AddRelationNotNullConstraints(), which already have similar checks.