19 dec 2012 kl. 23:14 skrev Daniel-Constantin Mierla <mico...@gmail.com>:
>
> On 12/19/12 10:28 PM, Olle E. Johansson wrote:
>> 19 dec 2012 kl. 19:35 skrev Daniel-Constantin Mierla <mico...@gmail.com>:
>>
>>> On 12/19/12 7:20 PM, Andreas Granig wrote:
>>>> Hi Olle,
>>>>
>>>> On 12/19/2012 02:43 PM, Olle E. Johansson wrote:
>>>>> Yeah, I already have Kamailio sending all kinds of crazy stuff while
>>>>> testing a new platform,
>>>>> using the UAC module. But I wanted to be lazy and change on the fly...
>>>>>
>>>>> Evil stuff happening here. Kamailio is a good test-tool.
>>>>>
>>>>> Now I fail to change Contact: headers. The docs for textops say that
>>>>> remove_hf can remove
>>>>> contact and Remove_hf(Contact) returns true - but the old contact is
>>>>> still there!
>>> Is this about generic operation of removing the contact header, or in the
>>> context of taking it from old request and adding it to the new one
>>> generated by uac_req()? For the second, Andreas pointed to the right
>>> direction.
>> No, changing one and send it away. Like changing a Contact: in a register
>> that's forwarded.
> So, you do: remove_hf("Contact"); t_relay(); -- and the Contact header is
> still there? Can you get the ngrep and logs printed by debugger module with
> cfgtrace parameter set to 1?
Will try testing a bit more later.
>
>>> I wanted to make the remark about changing the method on the fly directly
>>> to the incoming request, that is not straight as it seems, because it has
>>> impact on transaction management, the method being part of CSeq header and
>>> has to be restored for the reply.
>> Ouch. But if I'm totally and utterly stateless, would it work?
>
> You'd still have to revert back for replies, otherwise the previous hop will
> not match its transactions.
Match which transaction if I'm stateless ?? :-)
>
> Anyhow, there is no function in config that can change the method in first
> line of request, should not be hard to add one though (it is a function that
> can change the reply code). The cseq can be changed with header related
> functions.
Ok.
Thanks for all the feedback!
/O
>
> Cheers,
> Daniel
>
>>
>> /O
>>> With uac you can create a new transaction, but it will be in a new dialog.
>>>
>>> Cheers,
>>> Daniel
>>>> You could try using msg_apply_changes() from
>>>> http://kamailio.org/docs/modules/stable/modules/textopsx.html#textopsx.msg_apply_changes
>>>> after remove_hf and before actually building your new request with
>>>> $uac_req, but this is really dangerous.
>>>>
>>>> One thing I know for sure is that you better not call record_route()
>>>> anywhere before msg_apply_changes(), as record_route() only adds a hint to
>>>> the message and the value is filled in once the sending socket is known
>>>> (manually setting $fs doesn't seem to help). If you call
>>>> msg_apply_changes() after record_route(), you'll get errors and will end
>>>> up with a broken Record-Route header.
>>>>
>>>> Hope this helps,
>>>> Andreas
>>>>
>>>> _______________________________________________
>>>> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list
>>>> sr-users@lists.sip-router.org
>>>> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
>>> --
>>> Daniel-Constantin Mierla - http://www.asipto.com
>>> http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
>>>
>>>
>>> _______________________________________________
>>> SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list
>>> sr-users@lists.sip-router.org
>>> http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
>
> --
> Daniel-Constantin Mierla - http://www.asipto.com
> http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda
_______________________________________________
SIP Express Router (SER) and Kamailio (OpenSER) - sr-users mailing list
sr-users@lists.sip-router.org
http://lists.sip-router.org/cgi-bin/mailman/listinfo/sr-users
