Re: reloptions with a "namespace"

Alvaro Herrera escreveu:
> New patch attached, with pg_dump support (thanks to Tom for the SQL
> heads-up).
>
Great! We're close. Just two minor gripes:

+ char *validnsps[] = { "toast" };

Surely, you forgot to add a NULL at the end. Patch is attached.

IIRC, my last patch includes a partial validation code for RESET cases. For
example, the last SQL will not be atomic (invalid reloption silently ignored).
So, why not apply the namespace validation code to RESET case too? Patch is
attached too. It does not handle the reloptions validation because the relOpts
initialization code is at parseRelOptions(); i leave it for a future refactor.

euler=# create table foo (a text) with (fillfactor=10);
CREATE TABLE
euler=# \d+ foo
Tabela "public.foo"
Coluna | Tipo | Modificadores | Storage | Descrição
--------+------+---------------+----------+-----------
a | text | | extended |
Têm OIDs: não
Options: fillfactor=10

euler=# alter table foo reset (fillfactor,foo.fillfactor);
ALTER TABLE
euler=# \d+ foo
Tabela "public.foo"
Coluna | Tipo | Modificadores | Storage | Descrição
--------+------+---------------+----------+-----------
a | text | | extended |
Têm OIDs: não

--
Euler Taveira de Oliveira
http://www.timbira.com/