Hi list.

I have this problem, when sending binary data (with UDH) to kannel, for example  : 
ringtones. under some conditions (not sure which, except that maybe it has something 
to do with DCS), kannel converts bytes containing '\x0C' to '\x20'. the weird thing, 
is that it doesn't always do that - on a single message, some characters where 
converted but other not. for example, here's a snippet of the logs :

kannel-access.log
<snip>
2002-08-25 20:07:04 Receive SMS [SMSC:7] [SVC:] [ACT:] [from:8181] [to:7] 
[flags:0:2:0:0:0] 
[msg:128:024A3A612DA5CDCD2DA5CDCC08043724D30C35021034C2AC30C35021034C20C30C34C38C511824E514468938C35021034C34C0027A8C30D40840D30AB0C30D408408112E34834834C69A61A41A61A6288A08B08B12718A22C49C6288B12710698698610618618710718718698718698718698418698698698698690A22822C22]
 [udh:11:05041581158100032C0201]
2002-08-25 20:07:11 Receive SMS [SMSC:7] [SVC:] [ACT:] [from:8181] [to:7] 
[flags:0:2:0:0:0] [msg:29:C49C6288B12718A22C49C41A61A61841861861C41C410438E20E20E200] 
[udh:11:05041581158100032C0202]
</snip>

smsbox-access.log
<snip>
2002-08-25 20:07:04 SMS HTTP-request sender:8181 request: '' url: 
'http://localhost/interface/mo.php?text=%02J%3Aa-%A5%CD%CD-%A5%CD%CC%08%047%24%D3+5%02%104%C2%AC0%C3P!%03L+%C3+4%C3%8CQ%18%24%E5%14F%898%C3P!%03L4%C0%02z%8C0%D4%08%40%D3+%B0%C3+%40%84%08%11.4%83H4%C6%9Aa%A4%1Aa%A6(%8A%08%B0%8B%12q%8A%22%C4%9Cb%88%B1'%10i%86%98a%06%18a%87%10q%87%18i%87%18i%87%18i%84%18i%86%98i%86%98i+%22%82%2C%22&from=8181&to=7&time=1975-08-03+22:09:51&coding=2&smsc=7&udh=%05%04%15%81%15%81%00%03%2C%02%01'
 reply: 200 '<< successful >>'
2002-08-25 20:07:11 SMS HTTP-request sender:8181 request: '' url: 
'http://localhost/interface/mo.php?text=%C4%9Cb%88%B1'%18%A2%2CI%C4%1Aa%A6%18A%86%18a%C4%1CA%048%E2%0E+%E2%00&from=8181&to=7&time=1975-08-03+22:09:52&coding=2&smsc=7&udh=%05%04%15%81%15%81%00%03%2C%02%02'
 reply: 200 '<< successful>>'
</snip>

Careful examination would reveil that somewhere between these two logs, some 
characters have been changed. for example, take a look at the HTTP call URL, about 15 
characters down the text parameter :
<snip>
%047%24%D3+5%02
</snip>
and the same part from the first log
<snip>
043724D30C3502
</snip>
as you can see, '\x0C' has been converted to the url encoded character '+', which 
means <space> or '\x20'.

Would someone happen to know why or where it does that ?

--
Oded Arbel
m-Wise mobile solutions
[EMAIL PROTECTED]

+972-9-9581711 (116)
+972-67-340014

::..
-- To code or not to code, hmm, easy question.


Reply via email to