From: | Marek Lewczuk <newsy(at)lewczuk(dot)com> |
---|---|
To: | Kris Jurka <books(at)ejurka(dot)com> |
Cc: | pgsql-jdbc <pgsql-jdbc(at)postgresql(dot)org> |
Subject: | Re: AbstractJdbc2Array - another patch |
Date: | 2007-10-26 18:08:17 |
Message-ID: | 47222D11.5010006@lewczuk.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-jdbc |
Kris Jurka pisze:
>
>
> That doesn't sound right to me because we won't be able to put every
> possible type (think about user defined) into the Oid class. Perhaps
I was thinking only about base types (they are already in Oid class),
because it will work much faster if they will be statically written in
Oid class.
> getResultSet should convert getBaseTypeName() to oid instead of
> getBaseType? Then you just need to know if your output is an array or
> not (by checking isMultiDimensional) to know whether you want the oid
> for type or _type.
I still think that we should add base type arrays into Oid class - it
will work much faster and will be appropriate cause Oid class already
contains base types, so it is logical to put there base type arrays too.
For user defined types I would provide a way to fetch oid from pg_type -
but for now user defined types are not supported. However, if you really
think that we should fetch oid for every array type then I'm able to do
it but in my opinion we should stick with Oid class for now (only for
base types).
Regards,
ML
From | Date | Subject | |
---|---|---|---|
Next Message | Thomas Kellerer | 2007-10-26 22:55:35 | Question about refcursors |
Previous Message | Kris Jurka | 2007-10-26 16:13:06 | Re: AbstractJdbc2Array - another patch |