[freeside-devel] [freeside-commits] freeside/FS/FS cust_credit.pm, 1.27, 1.28

[Ivan Kohler]

On Thu, Dec 06, 2007 at 08:27:48PM -0500, Jeff Finucane wrote:
> "Ivan,,," <ivan at wavetail.420.am>:
> 
>    Thank you for your message of Fri, 07 Dec 2007 01:05:05 +0000....
> 
> +----------
> | From: "Ivan,,," <ivan at wavetail.420.am>
> | To: ivan-freeside-commits at sisd.com
> | Message-Id: <20071207010505.D409EFC8269 at wavetail.420.am>
> | Subject: [freeside-commits] freeside/FS/FS cust_credit.pm,1.27,1.28
> | 
> | Modified Files:
> | 	cust_credit.pm 
> | Log Message:
> | some old databases may have 'hanging' cust_credit records; we don't want to abort the upgrade because of that, just warn
> | 
> | Index: cust_credit.pm
> | ===================================================================
> | RCS file: /home/cvs/cvsroot/freeside/FS/FS/cust_credit.pm,v
> | retrieving revision 1.27
> | retrieving revision 1.28
> | diff -u -d -r1.27 -r1.28
> | --- cust_credit.pm	4 Dec 2007 18:20:56 -0000	1.27
> | +++ cust_credit.pm	7 Dec 2007 01:04:59 -0000	1.28
> | @@ -458,7 +458,8 @@
> |          $cust_credit->setfield('reason', '');
> |          my $error = $cust_credit->replace;
> |  
> | -        die "error inserting $self into database: $error\n"
> | +        warn "*** WARNING: error replacing reason in $self ".
> | +             $self->crednum. ": $error ***\n"
> |            if $error;
> |        }
> |      }
> +----------
> 
>   Sure about this?  I am not certain what you mean by "hanging"... do
> you mean it won't pass ::check() for lack of find a cust_main?

Right.  That shouldn't happen (and hasn't in years AFAIK), but I 
need to upgrade at least one old database with records like this.

What alternative would you suggest?  Modify cust_credit::check to ignore 
missing custnums ("local($ignore_custnum_check) = 1" or something) so 
the reasons can migrate?

>   As a practical matter, your change means upgrade completes, the reasons
> disappear from the UI,

Well, just for the old credits that aren't attached to a customer 
record.  :)

> and the warnings are issued again on a subsequent upgrade run.

Well, they should probably be cleaned up at some point anyway...

-- 
_ivan