AmSdp.cpp currently has this for printing sdp to 200 ok reply:
236 void AmSdp::print(string& body) const
237 {
238 string out_buf = "v="+int2str(version)+"\r\n"
239 "o="+origin.user+" "+int2str(origin.sessId)+" "+
240 int2str(origin.sessV)+" IN " + addr_t_2_str(conn.addrType) + " ";
now i noticed that conn.addrType is not set when sems gets this kind of
sdp from pjsua:
v=0.
o=- 3575798801 3575798801 IN IP4 188.67.175.126.
s=pjmedia.
b=AS:84.
t=0 0.
a=X-nat:0.
m=audio 4000 RTP/AVP 98 97 99 104 3 0 8 9 96.
c=IN IP4 188.67.175.126.
b=TIAS:64000.
a=rtcp:4001 IN IP4 188.67.175.126.
a=sendrecv.
...
i guess the reason is that there is no session level c= line in the
sdp.
i then tied to solve the problem by setting origin.conn.addrType in
parse_sdp_origin:
case ADDR:
{
next = parse_until(origin_line, ' ');
if(next > line_end){
DBG("parse_sdp_origin: ST_ADDR: Incorrect number of value in o=\n");
origin_st = UNICAST_ADDR;
break;
}
string addr_type(origin_line, int(next-origin_line)-1);
if (addr_type == "IP4")
origin.conn.addrType = 1;
else if (addr_type == "IP6")
origin.conn.addrType = 2;
else {
INFO("parse_sdp_origin: ST_ADDR: Unknown address type\n");
origin.conn.addrType = 0;
origin_st = UNICAST_ADDR;
break;
}
INFO("origin.conn.addrType = %d\n", origin.conn.addrType);
origin_line = next;
origin_st = UNICAST_ADDR;
break;
}
and then changed in print conn.addrType to origin.conn.addrType:
240 int2str(origin.sessV)+" IN " +
addr_t_2_str(origin.conn.addrType) + " ";
unfortunately that didn't have any effect. i got to syslog:
Apr 24 18:01:56 siika sems[32246]: [#7f3ba9c17700] [parse_sdp_origin,
AmSdp.cpp:1196] INFO: origin.conn.addrType = 1
i.e., origin.conn.addrType was set, but for some reason it is not
available in AmSdp::print because i still get this kind of o= line
in to 200 ok:
o=sems 2018467236 250804390 IN <unknown address type 0> 192.98.102.10.
i'm obviously missing something here. what is the proper way to print
sdp o= line address type in answer so that it matches address type of o=
line in offer when session level c= line is missing?
-- juha
_______________________________________________
Semsdev mailing list
[email protected]
http://lists.iptel.org/mailman/listinfo/semsdev
