On Thu, Jan 25, 2007 at 14:04:28 +0300, Konstantin Osipov wrote:
> * Vasil Dimov <vd@stripped> [07/01/22 19:12]:
> > > My personal preference would be to have this feature integrated
> > > with the library.
> > [...]
> > Of-course. Should the patch be created against 5.2 since it will lead to
> > ABI changes?
>
> What ABI changes do you foresee?
Well, "uint field_idx" should be added to MYSQL_FIELD in order to drop
the introduced MYSQL_FIELD_WITH_IDX (see my post "Get/Fetch field by
name: sample implementation" from Nov 30, 2006 [1]). And similarly
"HASH *hash" should be added to MYSQL_RES in order to drop the
introduced MYSQL_FIELDS_HASH. This part of the code, I sent (see [1]),
should make it clear if it is not already (sorry if I am redundant):
--- cut ---
/* This struct can be dropped and MYSQL_FIELD used instead if
* MYSQL_FIELD has index field */
typedef struct st_mysql_field_with_idx
{
MYSQL_FIELD *field;
uint field_idx;
} MYSQL_FIELD_WITH_IDX;
/* This struct can be dropped and MYSQL_RES used instead if:
* - MYSQL_FIELD has index field
* - MYSQL_RES includes the following `hash' */
typedef struct st_mysql_fields_hash
{
MYSQL_FIELD_WITH_IDX *fields_with_idx;
size_t fields_count;
HASH *hash;
} MYSQL_FIELDS_HASH;
--- cut ---
> Yes, 5.2 should be okay.
Great, I will submit the patch here in a few days.
[1] http://lists.mysql.com/internals/34152
--
Vasil Dimov
gro.DSBeerF@dv
%
If it's green or wiggles, it's biology.
If it stinks, it's chemistry.
If it doesn't work, it's physics.
Attachment: [application/pgp-signature]
