I am trying to figure out the best way to handle pages in our application like the dashboard, where there are a number of different panes with various bits of data in it.
The main issue is that the controller action becomes unwieldy when the page needs so much data. The client side can be broken up into partial views to make it more manageable, but all the data still needs to passed into the View to be distributed down to the partials. Or does it?
Obviously some of the panels could be loaded dynamically or something like that, but I was looking for the best approach besides loading individual piece of the page from the browser.
Have you considered using Html.Action in your view? You would pass enough data to the main view to enable you to give the required data to each of the child actions. The main action would render the main view which would call actions for each pane. Each action would be responsible for that pane, rendering its own partial view. Additionally, you could call back to each of the child actions directly from the client to update that pane dynamically via AJAX.
Here's an example with some mocked up actions of what your main view might look like.
<div class="left-pane">
#Html.Action("Summary", new { id = Model.ID } )
</div>
<div class="middle-pane">
#Html.Action("PendingItems", new { id = Model.ID, timestamp = DateTime.Now } )
</div>
<div class="right-pane">
#Html.Action("News")
</div>
Related
I am looking for the way to refresh a template inside a view rendered from another controller than the template's controller, I mean:
I got two controllers AdminController & UserController. And two gsps /admin/listUsers & /user/_searchResult.
Then a want to render view listUsers who have inside the template _searchResult and all right.
Now, i want to refresh the template _searchResult, but cant find how. I tryed calling render(view:"/admin/listUsers", template:"/user/_searchResult", model:[searchResult:result])
AdminController.groovy
#Secured(['ROLE_ADMIN'])
def listUsers(){
//...
}
UserController.groovy
#Secured(['ROLE_ADMIN'])
def search(){
//search users for the givven params and send result by chain if there's an action or update a template if it's needed
//in my case this method need to update the template _searchResult
}
#Secured(['ROLE_ADMIN'])
def searchResult(){
//...
[searchResult:result]
}
listUsers.gsp
//...
<formRemote name="searchForm" url="[action:"search", controller:"user"]">
//Some fields for the search
//I need to place here some hidden inputs to send which
//template i want to update or action to redirect
</formRemote>
<g:render template="/user/_searchResult"/>
//...
_searchResult.gsp
//Just itterate and print the search result in a table
I hope I have explained the problem correctly, thanks!
I don't think I entirely understand your question, but I think the source of your confusion is that the way you are naming things doesn't follow regular conventions and you're not using the right tools for the job. Let me explain...
The methods on Controllers are called Actions. They send some data (the Model) to a View to be rendered into HTML. Views can be composed from smaller, reusable fragments called Templates. (sorry if I sound like I'm being condescending here, but I'm just trying to make sure we're all on the same page).
Now, what you've called templateA is actually a View, not a Template. You're correct that templateA (your View) can call templateB to render some markup, but then having the templateB try to call a method on another Controller doesn't make sense. That's not how things flow.
If you have some logic that needs to be executed after you've sent your Model to the View, you want to use a Tag Library (http://grails.org/doc/latest/guide/theWebLayer.html#taglibs).
To summarise, here's a quick recap.
A request should only call one Action, which sends the model to only one view.
If you need to reuse logic between Controllers, move that code to a Service.
If you need to reuse markup between Views, move that markup to a Template.
If you have logic that you want to have executed after you've sent the Model to the View, use a Tag Library.
Hopefully this will point you in the right direction.
--- UPDATE ---
OK, with the real code I can see better what you're trying to achieve. Firstly, as you're using the <g:formRemote> tag, you should have a good read of the docs at http://grails.org/doc/latest/ref/Tags/formRemote.html to understand what it does.
What you will have here is 2 separate requests. The first will be a regular page load by your browser, which is handled by the listUsers() action. Once the page is then finished loading, the user will enter a search term and hit the submit button. This will fire off a second ajax request, which will be handled by the search() action. This action could use the _searchResult.gsp template to render a HTML table to display the search results. When the browser get this, it will insert it into the DOM where you've told it to put it using the "update" attribute of the <g:formRemote> tag.
The important thing here is that from the server's perspective, these are 2 separate requests that are completely independent. They both first call an action, then send a model (a Map containing some data) to a view, which renders/merges the data with HTML and sends it back to the browser.
The difference between the 2 is that the first is a complete page load by the browser, whereas for the second request, the browser only loads a small chunk of HTML (the search results table) and updates the page content without reloading it.
So your code would look more like this...
AdminController.groovy
#Secured(['ROLE_ADMIN'])
def listUsers() {
render(view:"/admin/listUsers")
}
listUsers.gsp
<g:formRemote name="searchForm" update="insertSearchResultsHere"
url="[controller: 'user', action:'search']">
<input name="searchTerm" type="text" />
</g:formRemote>
<div id="insertSearchResultsHere"></div>
UserController.groovy
#Secured(['ROLE_ADMIN'])
def search() {
// use the search term to get a List<User>
render(template: "/user/searchResult", model: [users: users])
}
_searchResult.gsp
<table>
<g:each var="user" in="${users}">
%{-- Iterate through your search results --}%
</g:each>
</table>
I solved it by placing the attribute update and rendering the template alone:
gsp:
<formRemote name="searchForm" url="[action:"search", controller:"user"]" update="divToUpdate">
//Some fields for the search
</formRemote>
<div id="divToUpdate">
<g:render template="/user/_searchResult"/>
</div>
Controller:
def search(){
render(template:"/user/_searchResult", model:[searchResult:result])
}
When i asked this question, i was new on Grails community and i was confused with the use of remoteFunction and tags that use it like remoteForm. But i had this confusion because of i had not read the documentation. So in my case, the solution was search for documentation about how to use remote tags and render. Thanks to #AndrewW for show me the way.
The page that I'm working with has a couple tabs and the content of each tab is loaded in via ajax by requesting a partial view from the controller. The problem is that the partial view uses knockoutjs, so it is bound to a view model. In this particular scenario, the page is loaded up in its entirety first time through, so all of the bindings work fine. When you switch tabs, it requests a partial view and replaces the tab content area with the new page. When you switch back to the first tab, it'll successfully loads the partial, except it would appear that all of the knockout bindings have been lost so there is a lot of missing data.
I can't place the viewmodel declaration and model bind in the partial because jquery hasn't been loaded by that point. Or so it would seem ($ is not defined).
The view model is declared and bound on the main page that calls the partial view(s), not the partial view itself, so I thought the model would still be available and bind successfully, but it does not. I know I'm doing this wrong, and partial view are super wonky when it comes to javscript so I'm hoping to steal a bit of insight from you guys.
Here's the basic setup:
If you are able to bind to specific non-overlapping areas of the page, then you could choose to call ko.applyBindings(someViewModel, someDomElement) like in this answer: Can you call ko.applyBindings to bind a partial view?
However, if you have an overall view model bound to the page and then "islands" of content that are loaded via a partial that you want to bind later, then one option would be to go for something like this: http://www.knockmeout.net/2012/05/quick-tip-skip-binding.html. So, you would set up a binding on the container of where your partial goes that tells Knockout to keep its hands off of that area. Then when you load the partial, you can safely call ko.applyBindings(someViewModel, innerContainer).
The binding might look like:
ko.bindingHandlers.stopBinding = {
init: function() {
return { controlsDescendantBindings: true };
}
};
and you would use it like:
<div id="outerContainer" data-bind="stopBinding: true">
<div id="innerContainer">
...load your partial here
</div>
</div>
Then, ko.applyBindings(someViewModel, document.getElementById("innerContainer"));
I am having a kendoUI TabStrip widget control. In it, I have two tabs : products and support.
The two tabs are two different views. I have defined the tabstrip like this :
<div id="tabs">
<div>
<iframe src="../Products/Index"></iframe>
</div>
<div>
<iframe src="../Support/Index"></iframe>
</div>
</div>
And in script :
$("#tabs").kendoTabstrip({});
Then tabs with their respective view pages are coming. But the products and support pages will not open directly. I mean in the url bar when I enter localhost:4567/Product/Index, it will not open the page, it will only open when we select the tab.
I have defined for both controllers such actions :
[ChildActionOnly]
public ActionResult Index()
{
return View();
}
Then when I run the main page, it is getting an exception it will be called only by child action only. Where when I select the tab it should become child action.
How can I rectify that problem? I have not included :
#Html.action("Index","Products")
If I need to include that, where should I add that?
Hope you understand my question...
The ChildActionOnly attribute ensures that an action method can be called only as a child method from within a view. An action method doesn’t need to have this attribute to be used as a child action, but we tend to use this attribute to prevent the action methods from being invoked as a result of a user request. Having defined an action method, we need to create what will be rendered when the action is invoked. Child actions are typically associated with partial views, although this is not compulsory
So instead
#Html.action("Index","Products").
you should try
#Html.Partial("Products").
So I have a problem...I am separating concerns in my web app: HTML in razor pages and JS in js files. My problem lies in the fact that I want to use data from my controller passed down in the view model from the server as the options of a select list. The problem lies in the fact that I have separated my js from my HTML and I can't access Razor inside the js files.
I have a list of items coming down into my view model...
public List Stuffs { get; set; }
I json encode it server side and make sure to take care of circular refrences, so it looks like this
[{"id":1,"name":"blah"},{"id":2,"name":"blah2"},{"id":3,"name":"blah3"},{"id":4,"name":"blah4"}]
The problem is, I want to keep my separation of concerns, so how do I get that array into my js file so I can bind it to a select list using knockout? I definitely don't want to do another round trip to the server.
So lets say you have some view that looks like this:
<div> (Some stuff) </div>
<script type="text/javascript">
var initialData = //Your JSON
<script>
And then in your javascript (which you have linked below your view, so that it loads after it), you have something like this:
var data = initialData || {} //Or some other default value, in case initialData wasn't set
This will load in the initialData if it was set, or use nothing (or a default your define) if it wasn't. Loosely coupled.
Of course, more complex data will require you to standardize the strucure, but this is essentially one method that allows you to pull in data from the razor generated View without tightly coupling it to your javascript file.
Does your Javascript file contain a view model in the form of an instantiable function, e.g. something like:
function MasterVM(data)
or an object literal, e.g.
var masterVM = { ...
? I tend to go for instantiable view models (partly because they chain down through a hierarchy of view models using the mapping plugin - the top level one builds its children) and that being the case I think it's nice for the Razor view to instantiate the view model with the JSON rendered from the MVC view model, e.g.:
var masterVM = new MasterVM(#(Html.Raw(JsonConvert.SerializeObject(Model.ProductViewModel))));
Or even have the Knockout view model "owned" by a revealing module and initialise it like:
productModule.init(#(Html.Raw(JsonConvert.SerializeObject(Model.ProductViewModel))));
Then productModule is also responsible for other things like mediating between your view model(s) and the server, watching for dirty state, etc.
Also, doing another round trip to the server is not necessarily bad if you are sourcing a massive set of reusable options which you'd like the browser to cache. If you are, you might want to have an MVC controller which emits a big object literal containing all of your commonly used options, which you can then use as the "options" property across multiple selects. But if your options are specific to the particular view model, then I guess they have to be part of your view model.
I've been using Codeigniter in order to get accustomed to the Model-View-Controller architecture, and to try and speed up the process of making and implementing sites.
I keep seeing references to "Partial Views" but can't find a definition for the term.
Can anyone tell me what a partial view is, and where it is used?
A partial view is just a sub-view that you can include in a parent view. Let's take a look at a common example:
// Controller:
$data['myvar'] = array('element1', 'element2', 'element3');
$this->load->view('myview', $data);
// Myview:
<ul>
foreach ($myvar as $var) {
$this->load->view('partialview', array('var', $var));
}
</ul>
// Partialview:
<li><?= $var ?></li>
This is useful to repeat content according to a list.
Note that nothing differs between a view and a partialview, it's just the way you include it that defines the term.
The best way to describe a "partial view" is to think of it as a template, it displays a chunk of html with Model data passed to it.
Good examples of where to use one would be where you plan on displaying the same html over and over, like a menu or a page header or even better yet use them to display content requested using ajax.
Basically you call an action on the controller that returns the partial view from lets say jQuery and then put the returned markup into a select or div tag. Here is an example of doing that from my blog easy ajax with aspnet mvc and jquery, yes I know it asp.net mvc not php and codeigniter, but the principal is the same.