Re: incorrect comparison of ppg_service_name and msg->sms.service in smsbox.c:obey_request_thread()

Davy Chan wrote:

> I've been looking at this off and on since Jan of this year.  It was
> sparked by Shyam Kumar's problem with dlr-url
> ("Subject: dlr-url and dlr-mask").
> 
> Basically, if you have a sendsms-user group with a dlr-url defined _AND_
> a PPG group defined within the same config file, the dlr-url will
> not be instantiated since the call to urltrans_find_service()
> is never called.
> 
> I'm looking at the following code in
> gateway/gw/smsbox.c:obey_request_thread():
> 
> /*
>  * no smsbox services when we are doing ppg dlr - so trans would be
>  * NULL in this case.
>  */
> if (dreport) {
>     if (msg->sms.service == NULL || (msg->sms.service != NULL &&
>          ppg_service_name != NULL &&
>          octstr_compare(msg->sms.service, ppg_service_name) != 0)) {
>         trans = NULL;
>     } else {
>         trans = urltrans_find_service(translations, msg);
>     }
> 
>     info(0, "Starting delivery report <%s> from <%s>",
> 	octstr_get_cstr(msg->sms.service),
> 	octstr_get_cstr(msg->sms.sender));
> 
> } else {
> 
> In my opinion, the comparison to determine if the DLR is for a PPG
> initiated WAP Push is incorrect. The current logic is:
>   IF the sms.service is not defined OR
>     (when sms.service is defined and ppg_service_name is defined
>      AND sms.service _NOT_EQUAL_ to ppg_service_name)
>   THEN
>     Don't find the translation structure associated with the service
>   ELSE
>     Do find the translation structure associated with the service
> 
> I assume that the _NOT_EQUAL_ should actually be EQUAL and the 'if'
> comparision should be:
>     if (msg->sms.service == NULL || (msg->sms.service != NULL &&
>          ppg_service_name != NULL &&
>          octstr_compare(msg->sms.service, ppg_service_name) == 0)) {
> 
> which implies:
>   Don't find the translation structure associated with the service
>    IF the sms.service is not defined OR
>      (when both sms.service and ppg_service_name exists
>       AND sms.service EQUALS ppg_service_name)
> 
> Am I correct in my assumptions that this is a bug?  I'm not sure since
> I'm not too familar with the PPG side of Kannel just yet.  I'm not
> sure there won't be a time when a PPG service name is the same as
> the value inside the sms.service (since it seems that sms.service is
> set by either the 'name =' or 'username =' declarations).
> 
> See ya...

ok, Aarno should state on this? Aarno?

Stipe

mailto:stolj_{at}_wapme.de
-------------------------------------------------------------------
Wapme Systems AG

Vogelsanger Weg 80
40470 Düsseldorf, NRW, Germany

phone: +49.211.74845.0
fax: +49.211.74845.299

mailto:info_{at}_wapme-systems.de
http://www.wapme-systems.de/
-------------------------------------------------------------------