Thread: Re: BUG #2599: AM/PM doesn't work in to_timestamp in
bruce wrote:
> Josh Tolley wrote:
> >
> > The following bug has been logged online:
> >
> > Bug reference: 2599
> > Logged by: Josh Tolley
> > Email address: eggyknap@gmail.com
> > PostgreSQL version: 8.1.4
> > Operating system: Fedora Core 5
> > Description: AM/PM doesn't work in to_timestamp in the middle of a
> > string
> > Details:
> >
> > eggyknap=# select to_timestamp('30 Aug 06:01:03.223 PM 2006', 'DD Mon
> > HH:MI:SS.MS AM YYYY'), to_timestamp('30 Aug 2006 06:01:03.223 PM', 'DD Mon
> > YYYY HH:MI:SS.MS AM');
> > to_timestamp | to_timestamp
> > ---------------------------+----------------------------
> > 0001-08-30 18:01:03-08 BC | 2006-08-30 18:01:03.223-06
> >
> > This appears to happen when AM/PM isn't the last element in the string.
>
> Nice report. The attached patch fixes it, and will be in 8.2. I am not
> backpatching because someone might be relying on this behavior.
Sorry, I attached the wrong patch to the email. This is the right one.
--
Bruce Momjian bruce@momjian.us
EnterpriseDB http://www.enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +
Index: src/backend/utils/adt/formatting.c
===================================================================
RCS file: /cvsroot/pgsql/src/backend/utils/adt/formatting.c,v
retrieving revision 1.110
retrieving revision 1.111
diff -c -r1.110 -r1.111
*** src/backend/utils/adt/formatting.c 19 Apr 2006 18:49:09 -0000 1.110
--- src/backend/utils/adt/formatting.c 3 Sep 2006 01:22:56 -0000 1.111
***************
*** 1762,1768 ****
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(p_inout);
}
break;
case DCH_AM:
--- 1762,1768 ----
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(P_M_STR);
}
break;
case DCH_AM:
***************
*** 1781,1787 ****
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(p_inout);
}
break;
case DCH_a_m:
--- 1781,1787 ----
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(PM_STR);
}
break;
case DCH_a_m:
***************
*** 1800,1806 ****
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(p_inout);
}
break;
case DCH_am:
--- 1800,1806 ----
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(p_m_STR);
}
break;
case DCH_am:
***************
*** 1819,1825 ****
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(p_inout);
}
break;
case DCH_HH:
--- 1819,1825 ----
tmfc->am = TRUE;
else
AMPM_ERROR;
! return strlen(pm_STR);
}
break;
case DCH_HH:
On 2006-09-03, Bruce Momjian <bruce@momjian.us> wrote: >> Nice report. The attached patch fixes it, and will be in 8.2. I am not >> backpatching because someone might be relying on this behavior. The bug appears to have been introduced just before the 8.1 betas; since it causes the AM/PM indicator simply to never work except at the end of the string, I see no possible justification for not backpatching the fix. -- Andrew, Supernews http://www.supernews.com - individual and corporate NNTP services
Backpatched to 8.1.X. --------------------------------------------------------------------------- Andrew - Supernews wrote: > On 2006-09-03, Bruce Momjian <bruce@momjian.us> wrote: > >> Nice report. The attached patch fixes it, and will be in 8.2. I am not > >> backpatching because someone might be relying on this behavior. > > The bug appears to have been introduced just before the 8.1 betas; since > it causes the AM/PM indicator simply to never work except at the end of > the string, I see no possible justification for not backpatching the fix. > > -- > Andrew, Supernews > http://www.supernews.com - individual and corporate NNTP services > > ---------------------------(end of broadcast)--------------------------- > TIP 3: Have you checked our extensive FAQ? > > http://www.postgresql.org/docs/faq -- Bruce Momjian bruce@momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +