Can we get someone to take a look at this patch? Without it, all the
clientIds for editable fields in a table are the same for each row, when
using the Sun RI with the Tomahawk dataTable.
thanks!
Eric
On 26 Oct 2005 10:23, Senthil Chelladurai (JIRA) wrote:
[ http://issues.apache.org/jira/browse/MYFACES-627?page=all ]
Senthil Chelladurai updated MYFACES-627:
----------------------------------------
Attachment: HtmlDataTableHack.patch
I have identified the bug and have a solution.
----------------------Start from SUN RI
UIData----------------------------------------------------
public String getClientId(FacesContext context) {
if (context == null) {
throw new NullPointerException();
}
String baseClientId = super.getClientId(context);
if (rowIndex >= 0) {
return (baseClientId + NamingContainer.SEPARATOR_CHAR + rowIndex);
} else {
return (baseClientId);
}
}
-----------------------End SUN RI
UIData------------------------------------------------------------
-----------------------Start Myfaces
UIData----------------------------------------------------------
public String getClientId(FacesContext context)
{
String clientId = super.getClientId(context);
int rowIndex = getRowIndex();
if (rowIndex == -1)
{
return clientId;
}
return clientId + "_" + rowIndex;
}
-----------------------End Myfaces
UIData------------------------------------------------------------
Carefully look at the above code fragment and you will notice that SUN RI uses a class private variable rowIndex to check for current row and creates the clientId accordingly, whereas Myfaces uses the method getRowIndex() method to get the current row number.
Why is this a problem?
Myfaces(tomahawk) has a class called HtmlDataTableHack which duplicates and
overrides most of the methods and field variables, so the rowIndex variable is
not getting updated in Sun RI UIData which uses field variable instead of the
getRowIndex() method. This ends up creating the same id for all rows so the
Decode phase sees only one value and so updates only the first value, this has
a slightly different effect on command components where the action is fired for
all rows even though only a specific action was triggered.
Solution: the getClientId method should also be overridden in HtmlDataTableHack
to solve this problem.
I have attached the patch file with the fix.
Hope to see the fix incorporated.
Senthil
valueChangeListener not called for elements in HtmlDataTable
------------------------------------------------------------
Key: MYFACES-627
URL: http://issues.apache.org/jira/browse/MYFACES-627
Project: MyFaces
Type: Bug
Components: Tomahawk
Versions: 1.1.1
Environment: JSF RI 1.1.01
Reporter: Keith Lynch
Priority: Critical
Attachments: HtmlDataTableHack.patch
The processValueChange method is never called in some cases and called only for
the first element in the list for others.
If the value of the first row inside the inputText component is changed then the valueChangeListener is called. If the inputText component in any other row is changed the valueChangeListener is not called.
The valueChangeListener is never called for the selectBooleanCheckbox.
I have not tested with any other sub-components.
This is possibly an RI compatibility issue however I cannot test with the
myfaces impl as I am using components which are not compatible with it.
x:dataTable value="#{ilrTestsMgr.details.currentPageContent}"
var="test"
rendered="true"
rowClasses="even-row,odd-row"
styleClass="bluelist"
width="100%">
<h:column>
<f:facet name="header">
<h:outputText value="#{bundle.generic_enabled}"/>
</f:facet>
<h:selectBooleanCheckbox value="#{test.enabled}"
valueChangeListener="#{ilrTestsMgr.details.processValueChange}"/>
</h:column>
<h:column>
<f:facet name="header">
<h:outputText value="#{bundle.testcase_value}"/>
</f:facet>
<h:inputText value="#{test.expected}" valueChangeListener="#{ilrTestsMgr.details.processValueChange}"/>
</h:column>
</x:dataTable>
