Hi Milinda, thanks for the fix but I see that you just replaced
attr_value = axiom_attribute_get_localname(om_attr, env);
by
attr_value = axiom_attribute_get_value(om_attr, env);

I think it may be safer to first check the attribute's localname
(IsReferenceParameter) and then check the attribute's value (true)
like :

attr_value = axiom_attribute_get_localname(om_attr, env);
if (axutil_strcmp(attr_value, AXIS2_WSA_IS_REFERENCE_PARAMETER_ATTRIBUTE) == 0)
 {
       attr_value = axiom_attribute_get_value(om_attr, env);
       if(axutil_strcmp(attr_value, AXIS2_WSA_TYPE_ATTRIBUTE_VALUE) == 0)
       axis2_msg_info_headers_add_ref_param(msg_info_headers, env,
header_block_node);
}

Moreover, I think that there is still a problem with the wsa_qname
variable (same function) as it is constructed with a NULL namespace
prefix instead of "wsa" or something like that. As a result,
axiom_element_get_attribute() always returns a NULL pointer as it
looks for IsReferenceParameter in the om_element->attributes hash
table without taking care of the namespace prefix.

Rgds,

julien

On Wed, Jun 25, 2008 at 9:50 AM, Milinda Lakmal Pathirage (JIRA)
<[EMAIL PROTECTED]> wrote:
>
>     [ 
> https://issues.apache.org/jira/browse/AXIS2C-1165?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
>  ]
>
> Milinda Lakmal Pathirage resolved AXIS2C-1165.
> ----------------------------------------------
>
>       Resolution: Fixed
>    Fix Version/s: Current (Nightly)
>
> Fixed in current SVN Head.
>
>> ReferenceParameter in ws-addressing module: problems with namespace and 
>> attribute verification
>> ----------------------------------------------------------------------------------------------
>>
>>                 Key: AXIS2C-1165
>>                 URL: https://issues.apache.org/jira/browse/AXIS2C-1165
>>             Project: Axis2-C
>>          Issue Type: Bug
>>          Components: core/addressing
>>    Affects Versions: 1.3.0, 1.4.0
>>         Environment: Windows XP SP2
>>            Reporter: Julien Billon
>>            Assignee: Milinda Lakmal Pathirage
>>             Fix For: Current (Nightly)
>>
>>
>> In axis2/c ws-addressing module, I've noticed something strange with 
>> Reference Parameters in EPR:
>> I have a message with a <soapenv:Header> like
>> <soapenv:Header xmlns:wsa="http://www.w3.org/2005/08/addressing"; 
>> xmlns:test="http://example.com/test";>
>>    <wsa:To>http://example.com/services/testSvc</wsa:To>
>>    <test:CorrID wsa:IsReferenceParameter="true">123456789</test:CorrID>
>>    <wsa:Action>http://example.com/actionTest</wsa:Action>
>>    <wsa:MessageID>6dc6e535-1a70-4544-9715-26f06cdcf7bb</wsa:MessageID>
>> </soapenv:Header>
>> In the axis2_addr_in_extract_ref_params() function of the ws-addressing 
>> module, the wsa_qname variable is defined with axutil_qname_create(env, 
>> "IsReferenceParameter", "http://www.w3.org/2005/08/addressing";, NULL) and as 
>> you can see, there is no prefix specified whereas the wsa namespace must be 
>> used according to the w3c recommendation (for ws-addressing). As a result, 
>> axiom_element_get_attribute( ..., wsa_qname) returns a NULL pointer. I've 
>> replaced the NULL prefix with "wsa" in wsa_qname and 
>> axiom_element_get_attribute( ..., wsa_qname) returns the correct attribute 
>> (IsReferenceParameter). The next function axiom_attribute_get_localname() 
>> returns "IsReferenceParameter" but this value is then compared with "true" 
>> (axutil_strcmp(attr_value, "true")). I think there's a little 
>> misunderstanding here between axiom_attribute_get_localname() and 
>> axiom_attribute_get_value(), we must first check if 
>> axiom_attribute_get_localname() returns "IsReferenceParameter" and then if 
>> axiom_attribute_get_value() returns "true".
>> Regards,
>> Julien
>
> --
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to