Some pages trigger the onBeginRequest() method in my RequestCycle around 3-4
times, others even 10 times. The number is constant for each page, i.e. for
PageA, the number of calls is always 5, and for PageB the number of calls is
always 8 (though, I haven't monitored the number of executions closely - the
number may be different between the first time a page is invoked, and
subsequent times).
The way I monitor the number of calls is a simple system.out.println call
here:
public class MyRequestCycle extends WebRequestCycle {
public MyRequestCycle(MyApp application, WebRequest request,
Response response) {
super(application, request, response);
}
@Override
protected void onBeginRequest() {
System.out.println("On begin request called!");
}
}
I wonder if the complexity of a page should have any affect on the number of
calls to RequestCycle.onBeginRequest()? if a page has many panels and those
panels in turn have even more nested panels....?
thanks,
Benny.
igor.vaynberg wrote:
>
> by numerous you mean two?
>
> by default wicket uses something similar to redirect-after-post so
> after you click a link there is a http redirect to a view-url that
> will no longer execute the link. so that is two http requests/two
> request cycles.
>
> -igor
>
> On Wed, Aug 13, 2008 at 7:18 AM, Benny Weingarten <[EMAIL PROTECTED]>
> wrote:
>>
>> Thanks Igor.
>>
>> I have done that, and it is exactly what I was looking for.
>>
>> However, I have noticed something that strikes me odd:
>> on every request to render a page in my application (every time I click a
>> link in my application to another page in my application), the method
>> newRequestCycle() in WebApplication is called numerous times, and
>> subsequently, my pre-render actions are performed many times. Is this
>> expected? shouldn't one page require only one RequestCycle?
>>
>> thanks,
>> Benny.
>>
>>
>>
>>
>> igor.vaynberg wrote:
>>>
>>> create a custom request cycle and put this stuff in
>>> requestcycle#onbeginrequest
>>>
>>> -igor
>>>
>>> On Sun, Aug 10, 2008 at 11:15 PM, Benny Weingarten
>>> <[EMAIL PROTECTED]> wrote:
>>>>
>>>> Hello.
>>>>
>>>> I'm writing a wicket application for facebook, but I think my problem
>>>> may
>>>> be
>>>> common to many wicket applications.
>>>>
>>>> In my case, I'm using the wicket-facebook authentication strategy that
>>>> is
>>>> described in a different thread here on nabble. That works great - when
>>>> a
>>>> page is rendered, the AuthenticationStrategy makes sure that I have a
>>>> valid
>>>> FacebookRestClient object in my session.
>>>> In my application I need to do some other "init" stuff in a service
>>>> class,
>>>> and I do that in the onUnauthorizedInstantiation method. This "init"
>>>> information is stored in a ThreadLocal variable in my BACKEND-Service.
>>>> This
>>>> is an important point, because wicket uses many different threads for
>>>> user
>>>> requests, so I need to make sure that in the beginning of each render
>>>> request, the init information is stored on THAT REQUEST's thread.
>>>>
>>>> ALL my components in all my pages and panels try to fetch information
>>>> from
>>>> the BACKEND-Service. If the BACKEND-Service hasn't been initialized
>>>> with
>>>> the
>>>> "init" information on the ThreadLocal variable on the same thread that
>>>> the
>>>> render request is issued on, an exception would be thrown and the page
>>>> would
>>>> fail to render.
>>>>
>>>> This is all great when a user is browsing through PAGES. However, in
>>>> one
>>>> page I have a tabbedpanel. When first the user loads the page, the
>>>> authorization strategy kicks in, and the init actions are performed,
>>>> the
>>>> page renders correctly.
>>>>
>>>> *** and here is my problem:
>>>> when the user clicks on the second tab in the page, the request can be
>>>> made
>>>> from a different thread (wicket controls that), but the "init" stuff is
>>>> not
>>>> performed because it is not a new page that needs to be rendered. This
>>>> problem also arises when I use ajax components.
>>>>
>>>> I thought of putting the init stuff in the "onBeforeRender()" method of
>>>> each
>>>> separate component that requires the "init" information.
>>>>
>>>> Would this be the preferred way of doing "init" actions for components?
>>>>
>>>> Thanks,
>>>> Benny.
>>>> --
>>>> View this message in context:
>>>> http://www.nabble.com/Pre-render-action-common-to-all-pages-tp18920691p18920691.html
>>>> Sent from the Wicket - User mailing list archive at Nabble.com.
>>>>
>>>>
>>>> ---------------------------------------------------------------------
>>>> 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]
>>>
>>>
>>>
>>
>> --
>> View this message in context:
>> http://www.nabble.com/Pre-render-action-common-to-all-pages-tp18920691p18963966.html
>> Sent from the Wicket - User mailing list archive at Nabble.com.
>>
>>
>> ---------------------------------------------------------------------
>> 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]
>
>
>
--
View this message in context:
http://www.nabble.com/Pre-render-action-common-to-all-pages-tp18920691p18976848.html
Sent from the Wicket - User mailing list archive at Nabble.com.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
