From looking at lump creation and processing code, it seems to me that

this happens because both the lumps related to the route block and to the failure route block are applied, and because they are stored in the lump list according to the offsets in such a way that results in the 'del' lumps for remove_hf() calls being applied before the 'add' lumps for append_hf() calls.

Can anyone confirm this? Shouldn't the route block lumps be cleared in the faked_req when calling fake_req() in modules/tm/t_reply.c ? This way in the failure route only the lumps created there will be apllied...?

JF

On 8/16/06, Jose Silva joesilvas@gmail.com wrote:

...

Message: 1 Date: Tue, 01 Aug 2006 15:27:20 +0300 From: OpenSer Users openser-users@list.coretech.ro Subject: [Users] heders in route & failure_route blocks To: users@openser.org Message-ID: 44CF48A8.4020503@list.coretech.ro Content-Type: text/plain; charset=ISO-8859-1; format=flowed

...

hello,

...

I want to add Remote-Party-ID header in route block and change it in failure_route block. the problem is that after the failure_route block the message sent out has two Remote-Party-ID headers.

...

in failure_route I am using remove_hf("Remote-Party-ID")

...

scenario: original invite has no Remote-Party-ID header in route block I use: remove_hf("Remote-Party-ID"); append_rpid_hf("<", ">;party=calling;privacy=full") in failure_route block I use: remove_hf("Remote-Party-ID"); append_rpid_hf("<", ">;party=calling;privacy=off")

...

the invite sent out after failure_route block has two Remote-Party-ID headers, one with privacy=full and another with privacy=off.

...

is there a way to correct the problem ? from what I understand the failure_route block works on a copy of the original message, but the original message has no Remote-Party-ID header, so why the header from the route block is still visible ? on the other hand if the header added from the route block is visible why is it no seen by the remove_hf function ?

...

thanks, Razvan Radu

Hi I all so have a similar problem! Is this behavior normal? Is it possible to have this type of manipulation of the message on the failure-route?

Thanks in advance, Jose Silva

---

Users mailing list Users@openser.org http://openser.org/cgi-bin/mailman/listinfo/users