I'm working on the front-end part of a project.
And the specified UI contains a set of dialogs which have same appearance on it.
Functionally speaking, two main of the dialogs behaves just like a MessageBox and ConfirmBox, the same with window.alert and window.confirm.
So, I'm trying to hijack the two functions, that when I call window.alert(msg), it shows my custom dialog:
<div id="dialog_alert">
<p class="dialog_content">Message</p>
<a class="btn-confirm">OK</a>
</div>
This behaves enough similar to the native window.alert.
Now comes the problem.
When I'm trying to hijack the window.confirm, I found it seems not possible:
window.confirm = function(message) {
var $dialog = $('#dialog_confirm');
$dialog.find('.dialog_content').text(message);
$dialog.show();
$dialog.find('.btn-yes').click(function() {
return true;
});
$dialog.find('.btn-no').click(function() {
return false;
});
// My GOD, I must return true or false immediately here.
};
As above, when I try to return the confirm result, because of the single thread/event loop/callback property of javascript, I just cannot work it out.
So, is it possible to hijack the confirm function as it was? I'm curious about an opinion, please help.
I read this article (Using dojo.behavior), and want to use the behavior module in my project as event handling module.
But I have a problem that, for DOM nodes, it works wonderful, but how can I use it on the Dojo widgets?
If I use dojo/on module, I can do it like this:
var buttonNode = dijit.byId("myButton");
on(buttonNode, "onClick", buttonHandler);
or
dijit.byId("myButton").onClick = buttonHandler;
But, if I use behavior module,
behavior.add({
"#myButton": {
onClick: buttonHandler
}
});
it doesn't work. (Of course I called behavior.apply() after I finished page render.)
The code below doesn't work either.
behavior.add({
"[widgetid='myButton']": {
onClick: buttonHandler
}
});
After some investigation, I found the reason the code above not work is because a button widget is composed by many s and an inner . And if I use the id specified by data-dojo-id, it will point to a instead of the that I hope the event bind with.
I found a solution which can walk out this situation,
behavior.add({
"[widgetid='myButton'] input": {
onclick: buttonHandler
}
}
but the css selector is too complex and it depends on what type the widget is.
Is there a good solution to apply dojo/behavior on widgets just like on dom nodes?
It looks like what you really what is to hook up an event on the widget object, but behavior is designed to access the DOM instead. So, I think you're stuck with your workaround.
I can set the onclick handler using jQuery by calling
$('#id').click(function(){
console.log('click!');
});
Also using jQuery, how can I get a reference to the function which is currently handling the click() event?
The reason is that I have another object and want to set its click handler to the same one as #id.
Update
Thank you for all the suggestions. The problem is that I do not know which function is currently handling the clicks. Keeping track of it would add state to an already complicated template-editing system.
jQuery's .click(function) method adds the function to a queue that is executed on the click event~
So actually pulling out a reference to the given function would probably be hairy-er than you expect.
As noted by others, it would be better to pass in a reference to the function; and then you already have the reference you need.
var clicky = function () { /* do stuff */ };
$('#id').click(clicky);
// Do other stuff with clicky
Update
If you really really need to get it out, try this:
jQuery._data(document.getElementById('id')).events.click[0].handler
Depending on your version of jQuery that may or may not work~ Try playing around with
jQuery._data(document.getElementById('id'))
and see what you get.
Got the idea from this section of the source:
https://github.com/jquery/jquery/blob/master/src/event.js#LC36
if you dont know the name of the function you can use
args.callee
https://developer.mozilla.org/en/JavaScript/Reference/Functions_and_function_scope/arguments/callee
function clickHandle(e){
if($(e.target) == $('#id')) {
$(newTarget).bind('click', clickHandle);
}
}
$('#id').bind('click',clickHandle);
I think this would be the most symantic way of going about it
I'd like to make it possible to use an onComplete event in my MooTools class by implementing Events and using this.fireEvent("complete"). By default I just want to let the class do nothing. So how would you realize the default "do nothing" thing the cleanest and safest way?
Should I just omit the option onComplete in the default class options object and still use this.fireEvent("complete") in the context? Or is this causing errors or something? I just did it that way all the time, but I'm not sure if this is the best way...
Would you use something like if(this.options.onComplete && instanceOf(this.options.onComplete, Function)) { this.fireEvent("complete") } instead? Or would you specify an empty function as default like options: { onComplete: function(){} }?
I don't care about the shortest code at this point but I was wondering how to do it the most conform and memory saving way.
Thanks for your recommendations! :)
Good question. Have a look at the source code of MooTools:
if (!events || !events[type]) return this;
As you can see, the fireEvent will just return the current instance when an event is not found. Omitting onComplete and keeping this.fireEvent('complete') would therefore not give any errors or weird behaviour.
I would not even specify an empty function in the options. As you can see here, the events are simply left undefined and only mentioned using a comment so people know they are present and can be used:
options: {
//onComplete: $empty
}
I need to override onClick method in the Tree.js. Is there any common way to override dojo/dijit classes methods?
I'm a bit confused, since you were already doing this in the last question you posted.
You've got a few choices, depending on what you want to do.
Clobbering method stubs
In the case of true stubs like onClick, it's potentially as easy as clobbering that method on your widget instance.
Programmatically:
var myWidget = new dijit.Tree({
...,
onClick: function(item, node, evt) {
/* handler code here */
}
};
Or declaratively (this is exactly the same as what you were doing in your last question):
<div dojoType="dijit.Tree" ...>
<script type="dojo/method" event="onClick" args="item,node,evt">
/* handler code here */
</script>
</div>
Connecting to method invocations
In other cases, perhaps you need to do something whenever a given method gets called, in which case you could use the widget's connect method (which is a nicer version of dojo.connect that will automatically clear itself when the widget is destroyed). In that case you could do something like this:
Programmatically:
//execute the given function whenever myWidget's onClick method is called
myWidget.connect(myWidget, 'onClick', function(item, node, evt) {
/* handler code here */
});
Declaratively, this can be done very similarly to the above, except instead of type="dojo/method", make sure you use type="dojo/connect"
<div dojoType="dijit.Tree" ...>
<script type="dojo/connect" event="onClick" args="item,node,evt">
/* handler code here */
</script>
</div>
Note that when you connect like this, your code will execute after the method whose invocation you are connecting to. If you need to do something before, your best option is probably to dojo.declare your own extension to the widget. If you need to go that far, I'll elaborate, but I think you'll likely be set with one of the above options.
Edit #1: Connecting the dots (no pun intended...oh heck, yes it was)
Since it seems that my comment appended to my answer to the original question was somehow not clear enough, here's a code block modifying the original code in that question based on the simple two steps exactly as I explained in that comment. The only tiny wrinkle is that the arguments passed to _onClick are slightly different.
<div dojoType="dijit.Tree" ...>
<script type="dojo/connect" event="_onClick" args="node,evt">
/* handler code here. In this case, item is accessible via node.item */
</script>
</div>
This solution may feel a bit sub-optimal since it involves connecting to a method that's suggested to be private. However, it's a way that should work regardless of whether openOnClick is true or not. If you are certain you're going to have openOnClick set to true, you could potentially write a single function, then connect it to both onClick and onOpen instead (both get passed the item, then the node).
Edit #2: Common functions, connecting programmatically
To answer your follow-up questions, I'd like to actually address them in reverse order - since if you are interested in connecting programmatically, it will actually make the other question easier to answer.
So first, to answer your connect question: you definitely don't want to be using dojo.byId, as that's not giving you the Tree widget, that's giving you some DOM node (probably the topmost) associated with the widget. As a general rule, dojo methods don't know anything about dijit stuff.
What you do want to be doing, is what I suggested above. Here it is applied as per the code you attempted. Also note that onClick has a capital C - another general rule: widget events use camel case notation, as a way to distinguish them from plain DOM events which don't.
var mytree = dijit.byId("mytree");
mytree.connect(mytree, "onClick", function(item) {
/* ... */
});
Now, to take that a step further and resolve your other inquiry, if you want to bind some common functionality to onClick and onOpen and onClose, you could define a function first, then connect it to both. This is one of the many things that makes JavaScript awesome - the availability of functions as first-class objects that can be easily passed around.
function handleClick(item) {
/* do stuff here.
Inside this function you can assume 'this' is the tree,
since connect will ensure it runs in-context.
*/
}
var mytree = dijit.byId("mytree");
mytree.connect(mytree, "onClick", handleClick);
mytree.connect(mytree, "onOpen", handleClick);
mytree.connect(mytree, "onClose", handleClick);
Now there's one important remaining question: where should we do this? The best place is likely inside a function passed to dojo.ready (or dojo.addOnLoad, same thing, ready was added as a synonym in 1.4) so that it will run only after:
The DOM is parsed by the browser
All dojo.required modules have loaded
If you set parseOnLoad: true in djConfig, all widgets defined in the document's HTML will already be instantiated
So, sometime after your dojo.requires, in script, you'd do this:
dojo.ready(function() {
/* code from previous example goes here */
});
Give it a shot.
Also, if you're interested in a bit of reading, I've actually written about a couple of topics I touched on in this edit:
Of Dijits and DOM Nodes
dojo.required Reading
You could use:
dojo.connect(tree, 'onClick', function(item) {
/** Your Action **/
});