Re: Possible BUG in dlr.c (sdb_callbackk_add() CVS)

Stipe Tolj wrote:

>>After recompiling CVS tree with sdb support (oracle), I got error from
>>sdb_callback_add function: Row has wrong length 1.
>>
>>For some reasons it only looks for p[0] value, but it only contains
>>dlrmask field, p[1] is service type and p[2] is url.
>>In my case:
>>p[0] -- 3
>>p[1] -- 1
>>p[2] -- http://localhost/dlr.php?type=%d&id=11
>>
>>Please tell me why we expect that all 3 values are in p[0]?
>>    
>>
>
>because p[0] contains the whole row 1, p[1] contains the whole row 2,
>etc.
>
>Hence in this DLR usage, we select for one row and everything should
>be in p[0]. Then we have to split that into the single attributes.
>
>At least this worked for me using mysql. I can not image that sdb is
>acting differently for the others dbs, because that would break the
>whole abstraction concept.
>
>Can you check using sdb_client, doing the same SQL statement and see
>how many lines are actualy printf'ed?!
>
>Stipe
>
>[email protected]
>-------------------------------------------------------------------
>Wapme Systems AG
>
>Vogelsanger Weg 80
>40470 Düsseldorf
>
>Tel: +49-211-74845-0
>Fax: +49-211-74845-299
>
>E-Mail: [email protected]
>Internet: http://www.wapme-systems.de
>-------------------------------------------------------------------
>wapme.net - wherever you are
>
>  
>
sdb_client returned 1 line, but actually there is 3 lines.

from sdb_client.c:

static int db_callback(int n, char **p, void *closure)
{
	int i;

	if (!n) return 0;
	for (i = 0; i < n; i++) {
		printf("%s ", p[i]);
	}
	printf("\n");
	return 0;
}

As you see they printf-ed p[i] without "\n", that's why we see 1 line in 
sdb_client output.





-- 
David Chkhartishvili
Tel: 995 99 182418