myfaces-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Shane Petroff <sh...@mayet.ca>
Subject Re: [Trinidad] PPR and inputDate problems
Date Fri, 08 Feb 2008 23:26:41 GMT
Andrew Robinson wrote:
> I think I know your problem. It is a phase issue 

I'll take your word for it, but I guess the only thing I could do to try 
and trace it through was to set a breakpoint in the decode method and 
step through everything.

> and why value change listeners (VCL) suck in JSF (really bad architecture)

Is there a better option? Even combos fire value changes.


> The VCL architecture sucks as it is not transaction aware, and the 
> events fired in the validation doesn't mean that the update will ever 
> be done (so the new value may never become the actual value). Also, 
> when the VCL is invoked, the model is not updated, so you can't trust 
> values from the model layer (beans) since they may have pending changes.
>
> I haven't tried it, but add a <tr:subform> around each input. This 
> should only validate & update the value that has changed instead of 
> the entire form.
Didn't work. I'm using

  <h:column>
        <tr:subform>
            <tr:inputText

and similar for the inputDate

One curious thing about this layout is that it can no longer find 
partial triggers fields when a normal name is specified for partialTriggers.

Feb 8, 2008 5:12:16 PM 
org.apache.myfaces.trinidadinternal.context.RequestContextImpl 
addPartialTriggerListeners
WARNING: Could not find partial trigger finalMarkField from 
CoreInputDate[UIXEditableFacesBeanImpl, id=_idJsp14]

I tried the preceding colons as well (4 of them this time because of the 
additional nesting), but no luck there either.

Surely I'm not the first person to try to use PPR/ValueChange on an 
editable table am I?

>
> On Feb 8, 2008 1:14 PM, Shane Petroff <shane@mayet.ca 
> <mailto:shane@mayet.ca>> wrote:
>
>     Andrew Robinson wrote:
>>
>>     What trinidad build are you using?
>
>     Trinidad 1.0.5,
>     MyFaces Core 1.1.5
>
>
>
>>     colons should not be necessary for components in the same naming
>>     container.
>
>     Tried with and without. Using the colons clearly does not work
>     because it issues the warning:
>
>     WARNING: Could not find partial trigger :::finalMarkField from
>     CoreInputDate[UIXEditableFacesBeanImpl, id=_idJsp12]
>
>     With no preceding colons there is no warning generated, but the
>     field is never refreshed either. The value change method is below,
>     as is a stripped down jsp. I can see the value change being fired,
>     but where would I even set a breakpoint to tell if anything is
>     happening on the partialTriggers side?
>
>
>
>
>         public void finalMarkChanged( ValueChangeEvent event )
>         {
>             StudentSection ss = (StudentSection) m_Table.getRowData();
>             ss.setWithdrawalDate(new Date()); // I'm assuming that
>     this instance is still around to refresh the field from
>
>             System.out.println( "finalMarkChanged for " +
>     ss.getStudentName() + " " + ss.getWithdrawalDate() );
>         }
>
>
>     <f:view>
>
>     <tr:document title="#{bundle.ClassListHeader}">
>
>
>     <body class="page-background">
>
>     <h:form>
>
>       <h:panelGrid headerClass="page-header" styleClass="panel"
>                    columns="1" cellpadding="5">
>
>         <h:messages showDetail="true" showSummary="false"
>     styleClass="errors"/>
>
>         <h:dataTable styleClass="dataTable"
>                      rowClasses="list-row-even,list-row-odd"
>     cellpadding="4" border="0"
>                      headerClass="list-header"
>                      cellspacing="0"
>                      value="#{classListBean.studentSections}"
>                      var="studentSec"
>                      binding="#{classListBean.table}">
>
>           <h:column>
>             <f:facet name="header">
>               <h:commandLink styleClass="table-header" id="studentDesc"
>                              actionListener="#{classListBean.sort}">
>                 <h:outputText value="#{bundle.StudentColHeader}"/>
>               </h:commandLink>
>             </f:facet>
>             <h:outputText value="#{studentSec.studentName}"/>
>           </h:column>
>
>           <h:column>
>             <f:facet name="header">
>                 <h:outputText
>     value="#{bundle.ClassListFinalMarkHeader}" styleClass="label"/>
>             </f:facet>
>             <tr:inputText id="finalMarkField"
>                           columns="5"
>                           value="#{studentSec.finalMark}"
>                           autoSubmit="true"
>                          
>     valueChangeListener="#{classListBean.finalMarkChanged}"/>
>           </h:column>
>
>           <h:column>
>             <f:facet name="header">
>                 <h:outputText
>     value="#{bundle.ClassListWithdrawalDateHeader}" styleClass="label"/>
>             </f:facet>
>             <tr:inputDate columns="11"
>
>                           partialTriggers=":::finalMarkField"
>                           value="#{studentSec.withdrawalDate}"
>                           required="false" >
>                 <f:convertDateTime pattern="dd-MMM-yyyy"
>     timeZone="#{configBundle.TimeZone}"/>
>             </tr:inputDate>
>           </h:column>
>
>         </h:dataTable>
>
>       </h:panelGrid>
>
>     </h:form>
>
>     </body>
>
>     </tr:document>
>
>     </f:view>
>
>
>>
>>
>>     On Feb 8, 2008 10:31 AM, Shane Petroff <shane@mayet.ca
>>     <mailto:shane@mayet.ca>> wrote:
>>
>>         Matthias Wessendorf wrote:
>>         > ah,
>>         > I just noticed, that "finalMarkField" is inside a table
>>         (naming container).
>>         > haven't read your code carefully, sorry.
>>         >
>>         > can you *try* partialTriggers=":::finalMarkField" ?
>>         >
>>
>>         No joy... cleared the browser cache first, then dumped
>>         tomcat's cache,
>>         did a full build and restarted the server. I also tried ::
>>         and ::::
>>         variants (3 does appear to be the correct number though based
>>         on the
>>         generated html), but still no partial triggers refresh. Would
>>         it be more
>>         likely to work if I switched to a tr:table?
>>
>>         --
>>         Shane
>>
>>
>     -- 
>     Shane
>
>


-- 
Shane


Mime
View raw message