Обсуждение: [PATCH] Fix CREATE INDEX/CONSTRAINT syntax when tablespace and fillfactor are specified

Поиск
Список
Период
Сортировка

[PATCH] Fix CREATE INDEX/CONSTRAINT syntax when tablespace and fillfactor are specified

От
Marti Raudsepp
Дата:
Hi list,

Quite recently I stumbled upon this bug in pgAdminIII and today I was
reading Guillaume Lelarge's article about pgAdmin's git repository and
this is the result. :)

The patch is attached, or you can also get it at
http://github.com/intgr/pgadmin3 branch "mybugfix"

Regards,
Marti

----
Commit message:
Fix CREATE INDEX/CONSTRAINT syntax when tablespace and fillfactor are specified

Reverse the order of TABLESPACE and WITH() clauses for indexes.  Previously
pgAdminIII generated CREATE scripts like:

ALTER TABLE foo ADD CONSTRAINT foo_pkey PRIMARY KEY(foo_id)
  USING INDEX TABLESPACE bar WITH (FILLFACTOR=50);
CREATE INDEX foo_foo_id ON foo USING btree (foo_id)
  TABLESPACE bar WITH (FILLFACTOR=50);

However these are illegal PostgreSQL syntax, the WITH() clause must come before
TABLESPACE.

The statements generated by "new index" or "new primary key" dialogs are
already correct.

Вложения

Re: [PATCH] Fix CREATE INDEX/CONSTRAINT syntax when tablespace and fillfactor are specified

От
Guillaume Lelarge
Дата:
Hi Marti,

Le 05/10/2010 23:33, Marti Raudsepp a écrit :
> [...]
> Quite recently I stumbled upon this bug in pgAdminIII and today I was
> reading Guillaume Lelarge's article about pgAdmin's git repository and
> this is the result. :)
>

That's just great. I didn't expect that someone would use this blog post
so quickly. And you don't appear to be alone:
http://github.com/postgres/pgadmin3/network

> The patch is attached, or you can also get it at
> http://github.com/intgr/pgadmin3 branch "mybugfix"
>

OK, nice work. I'll check this tomorrow. Thanks for your work.


--
Guillaume
 http://www.postgresql.fr
 http://dalibo.com

Re: [PATCH] Fix CREATE INDEX/CONSTRAINT syntax when tablespace and fillfactor are specified

От
Guillaume Lelarge
Дата:
Le 06/10/2010 00:25, Guillaume Lelarge a écrit :
> Hi Marti,
>
> Le 05/10/2010 23:33, Marti Raudsepp a écrit :
>> [...]
>> Quite recently I stumbled upon this bug in pgAdminIII and today I was
>> reading Guillaume Lelarge's article about pgAdmin's git repository and
>> this is the result. :)
>>
>
> That's just great. I didn't expect that someone would use this blog post
> so quickly. And you don't appear to be alone:
> http://github.com/postgres/pgadmin3/network
>
>> The patch is attached, or you can also get it at
>> http://github.com/intgr/pgadmin3 branch "mybugfix"
>>
>
> OK, nice work. I'll check this tomorrow. Thanks for your work.
>

Done, and commited. I also commited a substract of your patch on the
1.12 branch (we didn't have the tablespace issue in the index constraint
back in 1.12).

Thanks again.


--
Guillaume
 http://www.postgresql.fr
 http://dalibo.com