[I meant to post this one yesterday but it got lost for a while.] On Wed, Mar 12, 2003 at 11:59:33PM -0600, Stephen Clouse wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On Thu, Mar 13, 2003 at 12:25:06AM +0000, Tim Bunce wrote: > > With many thanks due to Jeff Urlwin, I've been able to get back into > > DBD::Oracle development after 18 months of, er, stability! > > > > I've made a snapshot of what'll be in 1.13 and put it here: > > > > http://homepage.eircom.net/~timbunce/DBD-Oracle-1.13-20030313.tar.gz > > > > I'd very much appreciate it if people could give it a whirl... > > Problems I had were fairly minor. > > 1. Warning: Aliases 'ORA_SYSDBA' and 'ORA_NUMBER' have identical values in > Oracle.xs, line 33 > > Don't know if this is really a problem or not, but it is bothersome to see.
It's not a problem, just xsubpp trying to be helpful. I've worked around it now. > 2. dbdimp.c: In function `ora_st_FETCH_attrib': > dbdimp.c:2032: warning: passing arg 2 of `Perl_newRV_noinc' from incompatible > pointer type > On line 2032, pvhv just needs to be cast to SV*. Done. > 3. cursor.t, line 46: Not error checked before statement is used. Needs an > || die. > > 4. cursor.t, line 46 again: The statement attemps to check v$parameter, which > by default is selectable only by sysdba-level login. I see that the selected > value is used by the rest of the tests, so it probably needs to be noted that > the tests need to be run using a sysdba login (and ora_session_mode needs > to be forced appropriately). Umm, t/cursor.t was contributed by Edwin Gustafson <[EMAIL PROTECTED]>. I've hacked on it a bit now so it'll default to just 1 cursor if it can't get the v$parameter value. I've also added tests for explicit closing of all the opened cursors. I've also made it run for OCI7, though I've had to disable the explicit closing as that's shown up a bug - it core dumps. > 5. t/ph_type....NOK 13expected 'trailing' but got 'trailing ' for VARCHAR2 > t/ph_type....FAILED test 13 > Failed 1/53 tests, 98.11% okay > > This seems to be an Oracle 9.2 issue, as 1.12 fails this test as well against > 9.2.0.2 on Linux. It would appear to be an Oracle bug, because the OCI > documentation still states that trailing space should be stripped. I've been > up and down Metalink for an answer, I'm probably going to open a TAR just to > get a definitive word on what the deal is here. That would be great! Thanks. Meanwhile, I've reworked and extended the ph_type test to make it easier to work with. I'd really like to get the STRING type working as I think that may be better for most people most of the time. Any volunteers to look into that would be most welcome. > > I'm aware that OCI7 support is broken - the tests just dump core. > > If someone could look into that I'd be grateful. > > Can you actually build an OCI7-only app against the v9 libs? If so I could try > to debug this. Jeff's fixed it, thanks. Could you double-check the updated metadata code? I ended up having quite a few versions of your patches in that area and I'd like to check that we're in sync now. Thanks. Tim.
