On Tue, Dec 16, 2003 at 12:01:20PM +0100, Dominik Stadler wrote:
>
> Hi,
>
> We have a perl-program that uses DBD::Oracle for accessing tables in an
> Oracle 9-Database. Some of the tables have a column with type TIMESTAMP.
>
> We are using DBD::Oracle-1.14 and DBI 1.38.
>
> When we execute $sth->fetchrow_array(), we get segfaults when executing
> the script on Windows 2000.
>
> I searched on the net and found in a thread at
> http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=UTF-8&threadm=20030514214011.7170D29873%40chronozon.artofdns.com&rnum=4&prev=/&frame=on
> that this is a problem in DBD::Oracle because new types of Oracle 9 are
> not handled currently.
>
> I looked at the beta for 1.15 of DBD::Oracle, but it doesn't contain the
> fix for this.
I hadn't got round to it then - and getting round to it is one of
the things that's held up the release.
Here's a quick fix...
Edit oci8.c, around line 1083, add these lines:
case 187: /* TIMESTAMP */
case 188: /* TIMESTAMP WITH TIME ZONE */
case 232: /* TIMESTAMP WITH LOCAL TIME ZONE*/
immediately below the line that says:
case 12: /* DATE */
Tim.
p.s. I'm expecting myself to make a release over the holiday period.
