I am a newbie in JSF.I am creating a simple page with a checkbox and a readonly field.When I deploy to weblogic server ,I get what is expected output.
Now I have put autosubmit property on checkbox and partialtrigger propery on the other readonly field.My readonly field changes as expected on changing the state
of checkbox.I was curious to find out what Ajax code has been put in finally rendered page when i declare auto submit property to true.Basically I want to know
what is the html and ajax(javascript) code difference between the case when auto submit property is enabled and disabled.Is there any tool which can compare two source codes?
Thanks in advance.
Being able to see the exact difference in code may be difficult as the associated Javascript files for your JSF component toolkit have probably been minified, however you should at least be able to see the difference in the Javascript event declarations on the generated input element.
A tool like Firebug is the best choice as it gives you the ability to highlight DOM elements and view their corresponding styles, attributes, and events. It doubles as an excellent Javascript debugger as well, allowing you to place breakpoints in JS code so that you can walk through the execution of what is happening on each click event.
When autoSubmit is false, there is likely no Javascript event being triggered. When it is true however, there is likely an onclick event being triggered that is formulating an Ajax request. You might have a hard time figuring out what is happening because it is minified, however it is more than likely making such a call.
http://docs.oracle.com/cd/E17802_01/j2ee/javaee/javaserverfaces/2.0/docs/js-api/symbols/jsf.ajax.html
Related
the page markup has
<div wicket:id="stepPanel" />
tag in it and when the page is first loaded it works great that is
add(new MyFirstPanel("stepPanel"));
works fine. But then when I trigger an Ajax event and request redrawing
addOrReplace(new MySecondPanel("stepPanel"));
target.add(MyPage.this);
i get the following error
Last cause: Failed to find markup file associated. MyFirstPanel: [MyFirstPanel [Component id = stepPanel]]
please note that it tries to find the wrong markup (should look for markup for MySecondPanel) and it fails regardless it succedded to do so before!
I instantiate panels using reflection, but could it be a problem here? No exceptions thrown.
Anwser:
Actually it was something else - I have noticed that one of my AjaxSubmitLinks had reference to a form that was no longer placed in a markup... so whatever you do just remember not to leave that reference.
You are adding MyPage after replacing the Panel causing MyPage to re-render.
There is a good example on how to replace panels here.
Yes you can call panels via reflection. I don't clearly know what you are trying to do with event here but if you want you can attach your panel with AjaxSelfUpdatingTimerBehavior and define the duration which will update this component in the given time period.
Hope its useful.
I am using Sencha ExtJs grid 4.2 . I am using a Expander plugins for my grid and try to load data under expanded region from Ajax. Right now I am using this code to show data on expanding.
plugins: [{
ptype: 'rowexpander',
rowBodyTpl: new Ext.XTemplate(
'<br><img height="31" width="32" src="../upload/patient/thumb/{patient_image}">',
' <p><b>{fname}, {lname}</b></p>',
'<br> {accordian_view}'
)
}],
Here you can see that data is pre populated, but my requirement is to load data on expanding. I am trying hard to find the event or process to do it. But still no luck. If anyone have any idea please share.
Thanks in Advance
You might check out the expandbody event on the RowExpander plugin: http://docs.sencha.com/extjs/4.2.1/#!/api/Ext.grid.plugin.RowExpander-event-expandbody
This event passes not only the row's bound record, but also the expanded element, so you could:
Request data via Ext.Ajax.request({...})
Handle response
Add content to expanded row
One thing to keep in mind, however, is the async nature of this approach. That is, the row is going to expand immediately, regardless of how long the subsequent request takes to come back. So it would probably be a good idea to do something like this instead when handling the expandbody event:
Add "loading" text/icon/whatever into expanded row area
Make Ajax request
Handle response
Replace loading text/icon/whatever with the data returned from the Ajax request
It's likely someone has already done so, but you could also (and I would highly suggest it) wrap this process into a custom plugin of your own that extends the RowPlugin. That way you could use it elsewhere in your app for any grid. If you end up creating a custom plugin, please share it with the community!
EDIT: A quick Google revealed a number of custom plugins that do precisely this. For example: https://github.com/nickbretz/Ext.ux.AsyncRowExpander/blob/master/AsyncRowExpander.js
Please see jsfiddle for example, blank out First Name field to have validation tooltip show. In a normal form the validation tooltip positions correctly to the right of each element. But in the popup editor for the grid it still trying to position the tooltip below the element as if it where editing inline. I have tried <span class="k-invalid-msg" data-for="FirstName"></span>but it doesn't change anything. Is there a setting I am missing to get this working in popupeditor? I guess I could manually modify the .k-tooltip but I am hoping for something more built in that handles the positioning correctly, because I am not very good at css.
As you've discovered, the error template for the grid is different to that provided by the kendo validator when applied to standard inputs.
Unfortunately, the validator that is created internally by the grid does not pass along any errorTemplate that you might define in the options object and by the time the "edit" event fires, the validator has already been created and the error template compiled, hence why setting the errorTemplate in the manner you describe does not work. Really, I think the Kendo grid should respect any user defined errorTemplate option but until it does we have to hack a little bit.
The key is to define a custom template and to apply it in the edit event, but instead of using the options object, set the private instance directly. Not ideal, but it works:
edit: function (e) {
e.sender.editable.validatable._errorTemplate =
kendo.template($('#tooltip-template').html());
}
See this updated fiddle for an example of what I think you might be looking to achieve.
http://jsfiddle.net/nukefusion/eQ2j7/10/
(I would post this as a comment but not enough reputation yet...)
I'm successfully using nukefusion's solution. I, too, fought with the syntax error from jQuery for a long time and discovered through debugging that how you define the template is important. In particular, it appears that the template has to be written on a single line without any formatting, as in:
<script id="tooltip-template" type="text/x-kendo-template"><span class="k-widget k-tooltip k-tooltip-validation"><span class="k-icon k-warning"></span>#=message#</span></script>
If you try to make it "pretty" and format the html in the template, you get the syntax error. I don't know where the real bug is, as this sort of thing shouldn't cause an error. But it does and I stopped worrying about it once I got it to work correctly.
is it possible to delay loading of some controls on an xpage?
This is the problem: let's say you have a control that does a fultextsearch and displays the result in a repeat control. this ft search might take a long time and will hold the webpage loading in a waiting state until the search result is ready.
I want my page to load most of the data initally, and some "time consuming" controls should be loaded in to the page as a sperate request after the inital load.
this way the user will immediatly see the webpage, but some of the data on the page will load a little bit later without holding the webpage in a waiting state from the server.
possible?
The downside to using rendered is that all the value bindings will still evaluate, even if the corresponding markup isn't sent to the page. So the trick here is making sure the components don't even exist until you want them to.
Every component has a getChildren() method. This returns a mutable List of components, which has a add() method. This allows you to add components to the page on the fly, either while the page is loading, or later during an event. For the purposes of what you're trying to do, you would want to defer adding the "expensive" components until a subsequent event.
Create an event handler attached directly to the view root (), give it a unique ID (e.g. "loadExpensiveComponentsEvent", set its refresh mode to partial, set a refresh ID to whatever div or panel will contain the search results, and set its event name to an arbitrary event (e.g. "loadExpensiveComponents"). This prevents your event from being triggered by actual user behavior. Set the event's code to SSJS that will inject your components.
Then add a script block () to trigger the event after the page has loaded:
XSP.addOnLoad(function(){
XSP.firePartial(null, "#{id:loadExpensiveComponentsEvent}");
});
Your page will load without the search result components. Once the page has fully loaded, it will trigger the component injection event automatically.
For guidance on how to code the injection event, open the Java file that has been generated from your existing page to see what components need to be injected and what to set their values to.
You can pack them into a panel and set their rendered status to rendered=#{viewScope.pageFullyLoaded}. Then in the onLoad event have a XSP. partialRefresh request where you set viewScope.pageFullyLoaded=true
A little ugly but doable. Now you can wrap that code into your own custom control, so you could have a "lazyGrid", "lazyPanel" etc.
Not sure why I did not think of this before. the dynamic content control in extlib actually solves this problem. the dcc can be triggered onClientLoad both using javascript and ssjs afer the page has loaded.
one problem I am facing now is that I am already using the dcc on my site so I need to put another dcc within my dcc. and this seem to be a bit buggy. I have reported it to the extlib team on openNTF.
After I load a page through a WebBrowser, and I click a link that fires an AJAX script, I need to detect when the AJAX java script finishes loading HTML changes into a div. Since no DocumentCompleted event is fired when the AJAX script is run, I don't know when it finish running. Is there a way I can attach an event to be raised after I know 100% that the javascript finished changing the div?
The project is in C#.
Thanks
I did something similar recently (using jQuery):
$('#mydiv').change(function(){
// do stuff
}
);
Granted, I also use jQuery to set the HTML of that div. I suppose one non-jQuery approach for you could be to set HTML through your own function, which in turn can fire an onchange event.
#New in town: From my experience that is not correct. I use this on multiple DIVs that never get focus in the first place, and it works well and consistently. The normal behavior is as you describe, and normally only applies to the INPUT and SELECT elements, but this is not the case with jQuery.
There is no event. You must patch the JavaScript callback that the browser runs when the reply for the AJAX request comes in. This will contains code like "div.innerHTML = ...". Just put your code below that.