Reg: MDSALDynamicAuthorization


Venkatrangan
 

Hi aaa-devs,
   Can you please review and merge the below patches -[0] and [1].
    The init was never getting invoked, causing the datalistener to never getting registered. If lazy initialization is preferred, please let us know the failing scenario, so that we can search for suitable solution. This feature has been failing ever since the lazy initialization was implemented.




--
Regards,
Venkatrangan G
(The Mind is Everything...
 What you think...you Become!!!)


Robert Varga
 

On 06/02/2020 08:20, Venkatrangan wrote:
Hi aaa-devs,
Hey Venkat,

   Can you please review and merge the below patches -[0] and [1].
    The init was never getting invoked, causing the datalistener to
never getting registered. If lazy initialization is preferred, please
let us know the failing scenario, so that we can search for suitable
solution. This feature has been failing ever since the lazy
initialization was implemented.
Okay, memory is sketchy, but there was a reason for this. We certainly
we do not run init in the constructor.

As per
https://shiro.apache.org/static/1.3.2/apidocs/org/apache/shiro/web/servlet/AbstractShiroFilter.html
should be the correct method, but perhaps it actually needs to be moved
to init().

Can you try that, please/

Thanks,
Robert


Venkatrangan
 

Hi Robert,

 Thanks for the suggestion. It looks like the initfilter(shiro) method never gets invoked for this filter.
 I have identified other method that is invoked fof this authorizationfilter which can be used for lazy initialization.
 Can you please review and merge if it okay

Regards,
Venkat G
 
 


On Fri, Feb 7, 2020 at 6:07 PM Robert Varga <nite@...> wrote:
On 06/02/2020 08:20, Venkatrangan wrote:
> Hi aaa-devs,

Hey Venkat,

>    Can you please review and merge the below patches -[0] and [1].
>     The init was never getting invoked, causing the datalistener to
> never getting registered. If lazy initialization is preferred, please
> let us know the failing scenario, so that we can search for suitable
> solution. This feature has been failing ever since the lazy
> initialization was implemented.

Okay, memory is sketchy, but there was a reason for this. We certainly
we do not run init in the constructor.

As per
https://shiro.apache.org/static/1.3.2/apidocs/org/apache/shiro/web/servlet/AbstractShiroFilter.html
should be the correct method, but perhaps it actually needs to be moved
to init().

Can you try that, please/

Thanks,
Robert



--
Regards,
Venkatrangan G
(The Mind is Everything...
 What you think...you Become!!!)