[Rt-devel] Where is $RT::TicketBaseURI defined?

Ruslan Zakirov ruslan.zakirov at gmail.com
Wed Sep 27 16:13:44 EDT 2006


This is deprecated option and shouldn't be used. I believe warning
doesn't appear with code from the repository.

On 8/25/06, Duncan McEwan <duncan at mcs.vuw.ac.nz> wrote:
> We are running RT 3.4.5 abd have a locally developed system that uses
> the RT API to create RT tickets.
>
> I just noticed that when our system runs I get the following message in
> my rt logs.
>
>   RT: Use of uninitialized value in concatenation (.) or string at \
>   /usr/pkg/lib/rt3/RT/Ticket_Overlay.pm line 243. \
>   (/usr/pkg/lib/rt3/RT.pm:287)
>
> Line 243 of Ticket_Overlay.pm is:
>
>     #If it's a local URI, turn it into a ticket id
>     if ( $id =~ /^$RT::TicketBaseURI(\d+)$/ ) {         # <----- line 243
>         $id = $1;
>     }
>
> This is part of the Ticket->Load() method.  I was pretty sure that we were
> passing a valid $id value in so that suggested that the undefined value
> was $RT::TicketBaseURI.  And indeed some debugging I added just before the
> above seemed to indicate that was the case.  Furthermore a 'grep -R' of
> all the likely places I could think of didn't reveal anywhere that this
> variable was being set.
>
> The other odd aspect is that we only see the warning when our own system
> calls the RT API.  We don't see it when the RT web interface loads tickets,
> even though my debugging showed that it was still undefined in that case.
> It's almost like warnings for uninitialised variables have been turned off,
> but I couldn't see where or why that would have been done.
>
> I've currently hacked around the problem by doing a
>
>    Set($TicketBaseURI, "");
>
> in my RT_SiteConfig.pm.  But I'd really rather prefer to understand what
> was going on here so any suggestions would be appreciated.
>
> Thanks,
>
> Duncan
>
>
> _______________________________________________
> List info: http://lists.bestpractical.com/cgi-bin/mailman/listinfo/rt-devel
>


-- 
Best regards, Ruslan.


More information about the Rt-devel mailing list