Re: [Wicket-user] Preserve form state

[Laurent PETIT]

Whoops, the missing test files in attachment ... :-)

On 11/18/05, Laurent PETIT <[EMAIL PROTECTED]> wrote:
> OK, now I have a first attempt to submit.
>
> It is what "could" be considered as a cleaner way to address Matej's
> initial problem.
> But it is a hack to the core, so it's normal that I've come up with
> something cleaner I think.
>
> And as it's a modification to the core, it's also more error-prone.
>
> It is too late tonight for me to explain what I did, so please for the
> moment take a look at the 2 new classes versions :
>
> wicket.markup.html.form.Form
> wicket.markup.html.form.FormComponent
>
> or the diffs files if you prefer.
>
> And I've joined 2 very basic test files (markup & class)
>
> In short : in the new version, on formSubmitted(), child form
> components are told that a new user input is available (method
> FormComponent.newUserInputAvailable()).
> In this method, the FormComponent stores the user input in a rawInput field.
> This rawInput is then available for re-rendering for both use cases :
> when data is invalid and when model is not up to date (for example
> when returning from another page as in Matej's use case, the form will
> be rendered again with the rawInput, and not the model).
>
> You can see that it also frees the FormComponent code from the weird
> NO_INVALID_INPUT constant (by using a modelUpToDate boolean field)
>
> Fiuu, time to sleep now ! :-)
>
> regards,
>
> --
> laurent
>
>
>

Form.java.diff
Description: Binary data

Form.java

Description: Binary data


FormComponent.java

Description: Binary data


FormComponent.java.diff

Description: Binary data
Title: Wicket SmartForms Test Page




 

 
 

 
 

 
 
 ...
 
 





TestPage.java

Description: Binary data


TestPage.properties

Description: Binary data