jsr344-experts@javaserverfaces-spec-public.java.net

[jsr344-experts] Re: [1111-PassThruElements] Proposal second version (was: First commit complete)

From: Edward Burns <edward.burns_at_oracle.com>
Date: Thu, 25 Oct 2012 14:45:05 -0700

>>>>> On Wed, 26 Sep 2012 19:56:56 +0200, Frank Caputo <frank_at_frankcaputo.de> said:

FC> Hi,
FC> Am 17.09.2012 um 10:23 schrieb Frank Caputo:

EB> The <jsf:element> tag causes a UIInput with a
EB> javax.faces.passthrough.Input Renderer to be placed in the tree. Here
EB> is the spec for this renderer.
EB>
EB> Decode Behavior
EB>
EB> Define a data structure that represents the following table.

EB> localName submittedValueAttributeName renderCurrentValue
EB> ========= =========================== ==================
EB>
EB> keygen name false

EB> <other elements to follow>

FC> we should add "submittedValueAttributeName" and "renderCurrentValue"
FC> as component properties. This way, we can support new input elements
FC> out of the box. E.g.:

FC> <newinputelement jsf:submittedValueAttributeName="name"
FC> jsf:renderCurrentValue="false" jsf:value="#{bean.newInputValue}"/>

FC> If there is no existing mapping use the component properties. They
FC> should also supersede existing mappings.

I'm going to push back on this one because I don't think it's justified
for such a corner case.

Ed

-- 
| edward.burns_at_oracle.com | office: +1 407 458 0017
| homepage:               | http://ridingthecrest.com/