I wasn't proposing changing the validation model at all - but with the
advent of Chain we could deprecate the validate(mapping, request) method in
favour of a validate(Context) method in ActionForm. This would provide more
flexibility and cause less confusion because the method name has stayed the
same. I would also suggest that the new validate method shouldn't return
anything, with the ActionForm being responsible for sticking the messages in
the Context. That way if/when thing move to commons resources we won't face
the same problem again.
Niall
----- Original Message -----
From: "Joe Germuska" <[EMAIL PROTECTED]>
To: "Niall Pemberton" <[EMAIL PROTECTED]>; "Struts Developers
List" <[EMAIL PROTECTED]>
Sent: Friday, December 03, 2004 9:09 PM
Subject: Re: ActionForm.validateForm(...) to replace
ActionForm.validate(...)
> At 8:41 PM +0000 12/3/04, Niall Pemberton wrote:
> >Wouldn't it be better to get rid of this in 1.3 with the move to Chain?
> >Doesn't everything get thrown up in the air and re-defined at that point
> >including Action's being deprecated in favour of objects that just
implement
> >the Command interface?
>
> I guess I had figured on 1.3 being more transitional than that. But,
> even if one were to use a command instead of an action, we haven't
> talked (yet) about changing the validation model.
>
> I haven't heard anyone propose a major change to the model of "Struts
> populates an ActionForm and calls a method on it which tests its
> validity and is able to return a bundle of messages explaining
> validation errors if there are any." My preference would be to defer
> any changes that dramatic until 1.4, although with the chain, it
> would be a little easier for people to prototype those kinds in the
> sandbox or on SourceForge (or whereever...)
>
> Perhaps it is worth trying to come up with a more future proof
> implementation, though. This isn't a burning issue -- clearly people
> are confused about the ActionMessages/ActionErrors situation, but I
> think that's "under control." By future-proof, I mean something that
> passes in the resources-equivalent of ActionMessages, and possibly
> which passes in something like a "ValidationContext" which would
> eliminate the explicit dependency on HttpServletRequest.
>
> Joe
>
>
>
> >Niall
> >
> >----- Original Message -----
> >From: "Martin Cooper" <[EMAIL PROTECTED]>
> >To: "Struts Developers List" <[EMAIL PROTECTED]>
> >Sent: Friday, December 03, 2004 7:59 PM
> >Subject: Re: ActionForm.validateForm(...) to replace
> >ActionForm.validate(...)
> >
> >
> >> We did just get Commons Resources promoted out of the sandbox, and I'm
> >> hopeful that we'll get that puppy released soon. Finally!
> >>
> >> --
> >> Martin Cooper
> >>
> >>
> >> On Fri, 3 Dec 2004 11:37:38 -0800 (PST), David Graham
> >> <[EMAIL PROTECTED]> wrote:
> >> > We didn't do it earlier because we wanted to use commons-resources
for
> >> > message passing. That hasn't happened so we may as well add the
> >> > validateForm() method and deprecate validate().
> >> >
> >> > David
> >> >
> >> >
> >> >
> >> > --- Joe Germuska <[EMAIL PROTECTED]> wrote:
> >> >
> >> > > In order to push forward on full deprecation of ActionErrors, I
> >> > > propose adding the following method to ActionForm:
> >> > >
> >> > > public ActionMessages validateForm(ActionMapping mapping,
> >> > > HttpServletRequest
> >request) {
> >> > >
> >> > > return validate(mapping, request);
> >> > >
> >> > > }
> >> > >
> >> > > and then changing one line in the Request Processing chain:
> >> > >
> >> > > ActionMessages errors = form.validate(mapping, request);
> >> > > to
> >> > > ActionMessages errors = form.validateForm(mapping,
request);
> >> > >
> >> > > I'm not sure now why we haven't done this earlier. Someone
suggested
> >> > > it on one of the lists a while ago and it seemed clear once I saw
it,
> >> > > but I haven't had time to do anything about it.
> >> > >
> >> > > My inclination is to do this only on the 1.3 (HEAD) branch, and to
> >> > > make the change in RequestProcessor.java even though it is slated
for
> >> > > obsolescence, and then also to make the change in
> >> > > o.a.s.chain.AbstractValidateActionForm (which actually still uses
> >> > > ActionErrors, actually.)
> >> > >
> >> > > If it didn't seem strange that it hasn't been done already, I
might
> >> > > have just gone ahead and done it without raising the question --
so
> >> > > I'm wondering if I'm missing something?
> >> > >
> >> > > Joe
> >> > >
> >> > > --
> >> > > Joe Germuska
> >> > > [EMAIL PROTECTED]
> >> > > http://blog.germuska.com
> >> > > "Narrow minds are weapons made for mass destruction" -The Ex
> >> > >
> >> >
> ---------------------------------------------------------------------
> >> > > To unsubscribe, e-mail: [EMAIL PROTECTED]
> >> > > For additional commands, e-mail: [EMAIL PROTECTED]
> > > > >
> >> > >
> >> >
> >> >
> >> > __________________________________________________
> >> > Do You Yahoo!?
> >> > Tired of spam? Yahoo! Mail has the best spam protection around
> >> > http://mail.yahoo.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]
> >>
> >>
> >>
> >
> >
> >
> >---------------------------------------------------------------------
> >To unsubscribe, e-mail: [EMAIL PROTECTED]
> >For additional commands, e-mail: [EMAIL PROTECTED]
>
>
> --
> Joe Germuska
> [EMAIL PROTECTED]
> http://blog.germuska.com
> "Narrow minds are weapons made for mass destruction" -The Ex
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
