[PATCH] Re: Fw: DLR sql select problem with message id containing space

Hi Alex,

would you please test attached patch?

On Friday 16 January 2004 15:20, Alex Kinch wrote:
> Resending as my mail's been playing up :-)
>
> ----- Original Message -----
> From: Alex Kinch
> To: [email protected]
> Sent: Friday, January 09, 2004 10:14 PM
> Subject: Re: DLR sql select problem with message id containing space
>
>
> Happy New Year to all :-)
>
> Don't suppose the problem with spaces in message id's and DLR's got fixed
> in CVS?
>
> Alex
>   ----- Original Message -----
>   From: Alex Kinch
>   To: Rene Kluwen ; [email protected]
>   Sent: Wednesday, December 24, 2003 12:59 AM
>   Subject: Re: DLR sql select problem with message id containing space
>
>
>   Hi Rene
>
>   In the SMPP v3.4 issue 1.2 (12th October 1999) docs that I have, it's
> been up'd from 9 to 33, and then to 65 octets. I guess this is to
> accomodate for the increasing use of GUID-style message ids?
>
>   It seems that Kannel handles the SUBMIT_SM_RESP ok with these long ids,
> but doesn't do it in the DELIVER_SM for the DLR.
>
>   Alex
>     ----- Original Message -----
>     From: Rene Kluwen
>     To: Alex Kinch ; [email protected]
>     Sent: Tuesday, December 23, 2003 9:24 PM
>     Subject: RE: DLR sql select problem with message id containing space
>
>
>     Fix would be not too hard to implement... BUT...
>     In the SMPP specs that I have (note: I am talking about v3.3) it says:
> Message ID, var max. 9. So taking off the terminating NUL this means that
> message ids can be 8 characters at most. I noticed this 'quirk' (message
> ids of more than 8 chars) also with a certain upstream provider that I
> have.
>
>     Question: Has this changed in successive SMPP versions? Or is it just
> plainly against specifications?
>
>     Regards,
>
>     Rene Kluwen
>     Chimit
>       -----Original Message-----
>       From: [email protected] [mailto:[email protected]]On Behalf
> Of Alex Kinch Sent: dinsdag 23 december 2003 18:47
>       To: [email protected]
>       Subject: DLR sql select problem with message id containing space
>
>
>       Hi,
>
>       Just hooked up another SMPP connection, and they use rather large
> message ids. For example:
>
>       00047274 20031223170148.364
>
>       Kannel is inserting this whole value (including characters after the
> space) into the ts field on the dlr table. However, when it comes to
> receiving the delivery report, it's truncating the message id after the
> first space, so it does:
>
>        DLR[mysql]: Looking for DLR smsc=XXXX, ts=00047274,
> dst=44xxxxxxxxxx, type=1
>
>       Which of course it won't find, as it's lost the rest of the message
> id.
>
>       My C coding skills aren't up to much, but I would imagine it requires
> something that tests the first character after the space to see if it's
> numeric.
>
>       Alex

-- 
Best regards / Mit besten Grüßen aus Düsseldorf

Dipl.-Ing.
Alexander Malysh
___________________________________________

Centrium GmbH
Vogelsanger Weg 80
40470 Düsseldorf

Fon: +49 (0211) 74 84 51 80
Fax: +49 (0211) 277 49 109

email: [email protected]
web: www.centrium.de
msn: [email protected]
icq: 98063111
___________________________________________

Please avoid sending me Word, Excel or PowerPoint attachments.
See http://www.fsf.org/philosophy/no-word-attachments.html

Index: gw/smsc/smsc_smpp.c
===================================================================
RCS file: /home/cvs/gateway/gw/smsc/smsc_smpp.c,v
retrieving revision 1.59
diff -a -u -r1.59 smsc_smpp.c
--- gw/smsc/smsc_smpp.c	4 Dec 2003 16:51:05 -0000	1.59
+++ gw/smsc/smsc_smpp.c	16 Jan 2004 15:23:28 -0000
@@ -966,33 +966,26 @@
 	     */
             if ((pdu->u.deliver_sm.esm_class & ~0xC3) == 0x04) {
                 Octstr *respstr;
-                Octstr *msgid = NULL;
-                Octstr *stat = NULL;
+                Octstr *msgid = NULL, *stat = NULL;
                 int dlrstat;
-                long curr = 0, vpos = 0;
+                char id_cstr[65], stat_cstr[11], sub_d_cstr[11], done_d_cstr[11];
+                int sub, dlrvrd, ret;
 
                 debug("bb.sms.smpp",0,"SMPP[%s] handle_pdu, got DLR",
                       octstr_get_cstr(smpp->conn->id));
 
                 respstr = pdu->u.deliver_sm.short_message;
 
-                /* get server message id */
-                if ((curr = octstr_search(respstr, octstr_imm("id:"), 0)) != -1) {
-                    vpos = octstr_search_char(respstr, ' ', curr);
-                    if ((vpos-curr >0) && (vpos != -1))
-                        msgid = octstr_copy(respstr, curr+3, vpos-curr-3);
-                } else {
-                    msgid = NULL;
-                }
-
-                /* get err & status code */
-                if ((curr = octstr_search(respstr, octstr_imm("stat:"), 0)) != -1) {
-                    vpos = octstr_search_char(respstr, ' ', curr);
-                    if ((vpos-curr >0) && (vpos != -1))
-                        stat = octstr_copy(respstr, curr+5, vpos-curr-5);
-                } else {
-                    stat = NULL;
+                ret = sscanf(octstr_get_cstr(respstr), "id:%64[^sub:] sub:%d dlvrd:%d submit date:%10s done date:%10s stat:%10[^err:]",
+                             id_cstr, &sub, &dlrvrd, sub_d_cstr, done_d_cstr, stat_cstr);
+                if (ret == 6) {
+                    msgid = octstr_create(id_cstr);
+                    octstr_strip_blanks(msgid);
+                    stat = octstr_create(stat_cstr);
+                    octstr_strip_blanks(stat);
                 }
+                else
+                    warning(0, "SMPP[%s]: Couldnot parse DLR string.", octstr_get_cstr(smpp->conn->id));
 
                 /*
                  * we get the following status:
@@ -1055,7 +1048,6 @@
                      * The DLR trigger URL is indicated by msg->sms.dlr_url.
                      */
                     dlrmsg->sms.msgdata = octstr_duplicate(respstr);
-                    dlrmsg->sms.sms_type = report;
 
                     reason = bb_smscconn_receive(smpp->conn, dlrmsg);
                 } else {