Skip Menu |

This queue is for tickets about the DBD-Pg CPAN distribution.

Report information
The Basics
Id: 12503
Status: resolved
Priority: 0/
Queue: DBD-Pg

People
Owner: Nobody in particular
Requestors: alexw [...] benon.com
Cc:
AdminCc:

Bug Information
Severity: Normal
Broken in: 1.40
Fixed in: 1.42



From: Alexandra Walford <alexw [...] benon.com>
To: bug-dbd-pg [...] rt.cpan.org
Subject: DBD::Pg 1.41: last_insert_id method calls die
Date: Tue, 3 May 2005 14:09:45 +1000
Hi, Since the last_insert_id method was added to DBD::Pg 1.40 (Bug # 3803), we've noted that calls to die occur on lines 229, 246, 267, 281, 293 and 299 of DBD/Pg.pm (revision 1.153, although the situation is unchanged in latest CVS - revision 1.155). We use DBD::Pg in a production environment, and this use of die causes a few problems for us (which we can work around using $SIG{__DIE__} handlers and/or eval) - but IMHO die does not belong in a publically released package such as DBD::Pg. In fact, apart from the data_sources method, last_insert_id is the only method which uses die in DBD::Pg. This leads me to believe that you (the developers) also take this view and usually minimise the use of die in this package. I have created a patch (attached) which simply replaces the calls to die with 'return undef'; this may or may not be the correct approach in your eyes but it Works For Me (tm) and seems somewhat consistent with the rest of the package. Anyway, I hope that either this patch is of use or that someone better qualified :) can patch DBD::Pg to resolve this issue. [By the way, my apologies if there is already a bug tracking this issue; I am unused to rt.cpan.org and cannot find a way to quickly search through bug reports. I did find a thread on dbdpg-general touching on the use of die but am unaware of any further activity: http://gborg.postgresql.org/pipermail/dbdpg-general/2004-November/001305.html ] Regards, Alexandra Walford

Message body is not shown because sender requested not to inline it.