Horton

Re: Schema variables - new implementation for Postgres 15

Dmitry Dolgov <9erthalion6@gmail.com>

View thread
From: Dmitry Dolgov <9erthalion6@gmail.com>
To: Pavel Stehule <pavel.stehule@gmail.com>
Cc: Peter Eisentraut <peter.eisentraut@enterprisedb.com>, Sergey Shinderuk <s.shinderuk@postgrespro.ru>, Tomas Vondra <tomas.vondra@enterprisedb.com>, Julien Rouhaud <rjuju123@gmail.com>, dean.a.rasheed@gmail.com, er@xs4all.nl, joel@compiler.org, pgsql-hackers@lists.postgresql.org
Date: 2023-11-17T19:17:32Z
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. Allow underscores in integer and numeric constants.

  2. Remove special outfuncs/readfuncs handling of RangeVar.catalogname.

  3. Remove extra space from dumped ALTER DEFAULT PRIVILEGES.

  4. Create FKs properly when attaching table as partition

  5. psql: improve tab-complete's handling of variant SQL names.

> On Wed, Aug 23, 2023 at 04:02:44PM +0200, Pavel Stehule wrote:
> NameListToString is already buildin function. Do you think NamesFromList?
>
> This is my oversight - there is just `+extern List *NamesFromList(List
> *names); ` line, but sure - it should be in 0002 patch
>
> fixed now

Right, thanks for fixing.

I think there is a wrinkle with pg_session_variables function. It
returns nothing if sessionvars hash table is empty, which has two
consequences:

* One might get confused about whether a variable is created,
  based on the information from the function. An expected behaviour, but
  could be considered a bad UX.

    =# CREATE VARIABLE var1 AS varchar;

    -- empty, is expected
    =# SELECT name, typname, can_select, can_update FROM pg_session_variables();
     name | typname | can_select | can_update
     ------+---------+------------+------------
     (0 rows)

    -- but one can't create a variable
    =# CREATE VARIABLE var1 AS varchar;
    ERROR:  42710: session variable "var1" already exists
    LOCATION:  create_variable, pg_variable.c:102

    -- yet, suddenly after a select...
    =# SELECT var2;
     var2
     ------
      NULL
      (1 row)

    -- ... it's not empty
    =# SELECT name, typname, can_select, can_update FROM pg_sessio
    n_variables();
     name |      typname      | can_select | can_update
     ------+-------------------+------------+------------
      var2 | character varying | t          | t
      (1 row)

* Running a parallel query will end up returning an empty result even
  after accessing the variable.

    -- debug_parallel_query = 1 all the time
    =# CREATE VARIABLE var2 AS varchar;

    -- empty, is expected
    =# SELECT name, typname, can_select, can_update FROM pg_session_variables();
     name | typname | can_select | can_update
     ------+---------+------------+------------
     (0 rows)

    -- but this time an access...
    SELECT var2;
     var2
     ------
      NULL
      (1 row)

    -- or set...
    =# LET var2 = 'test';

    -- doesn't change the result, it's still empty
    =# SELECT name, typname, can_select, can_update FROM pg_session_variables();
     name | typname | can_select | can_update
     ------+---------+------------+------------
     (0 rows)

Would it be a problem to make pg_session_variables inspect the catalog
or something similar if needed?