Good afternoon everyone. I was wondering if there is anyway to have the MVC framework automatically wire up the data-val* attributes on the controls or do we need to manually create and apply the attributes to dynamic content?
I have a view that initially calls a partial view passing in the main viewmodel. This partial view is bound to a complex property on my main viewmodel. The partial view simply contains a set of cascading dropdown lists. On initial load of the page I have a call to #Html.Partial("PartialName", Model), the two dropdown lists’ validation works perfectly if I try to submit without selecting proper values. I also have another button on the page that if clicked loads another instance of the partial view on the page. If I now try to submit the form these controls, although they are bound to the same model and although I have set the correct .ValidationMessageFor helpers, no validation appears for them since the dropdownlists do not appear to be generated with the data-val* attributes. Is there any way that I can get them to appear correctly? I also noticed that the associated <span /> tag associated to the .ValidationMessageFor is not generated either. Has anyone run into this problem as well, if so how did you resolved?
UPDATE
Here is the javascript function that I call to load the partial on the button's onClick event:
function AddNewVehicle() {
$.ajax({
type: 'GET',
url: '/ReservationWizard/AddVehicleToReservation',
data: $('#reservation-wizard-form').serialize(),
dataType: 'HTML',
async: true,
success: function (data) {
if (data != null) {
$('#vehicle-selection-container').append(data);
}
}
});
}
The problem is that if you are not inside a form context, the HTML helpers such as TextBoxFor do not output any client validation data-* attributes. The first time when the page loads you invoke your Html.RenderPartial inside an Html.BeginForm() but later when you use AJAX to append form elements there is no longer this form context and there won't be any data-* client validation attributes generated. One possible solution would be to put the form inside the partial and then update the entire form during the AJAX call and in the success callback re-parse the client validation rules using $.validator.unobtrusive.parse('#vehicle-selection-container').
But if you want to keep only a single element inside the partial you are pretty much on your own :-) Here's a blog post which covers your scenario that you might take a look at.
So what can I say: unobtrusive client validation is great on paper and Scott Gu's blog posts but at some stage of the development of real world applications people start to realize its limitations. That's one of the reasons why I directly use the jquery.validate plugin and no MS jquery.unobtrusive. And, yes I know that I repeat my server validation logic in the javascript and yes I don't care because I have total control. Oh, and on the server I use FluentValidation.NET instead of data annotations for pretty much the same reasons as the client side part :-)
So maybe some day in MVC 4 Microsoft will finally make validation right (imperative vs declarative) but until this day comes, we just need to be searching for workarounds.
Related
I am newbie to MVC and Web App.
Recently I have went through the article
http://www.c-sharpcorner.com/UploadFile/pmfawas/Asp-Net-mvc-how-to-post-a-collection/
It uses the Ajax Form, to do the partial update towards a particular region alone..
But I have a doubt in that example...
I have seen the partial Page inside the Div with Id "AllTweets"....
<div id="AllTweets">
#Html.Partial("_AllTweets", Model) ***** (XXX)
</div>
And also in the controller action,
try
{
viewModel.Tweets.Add(viewModel.Tweet);
return PartialView("_AllTweets", viewModel); **** (YYYYY)
}
Now my question is,
They are returning the partial view along with the data from the action in the controller.
Whatever the data returned from the controller, the engine will place that data, inside the target div with id "AllTweets"...
But still, why I have to have the statement, #Html.Partial("_AllTweets", Model) inside the Div, since already I am returning the data from the controller...
And also in some of the examples, i have seen the same kind of the code..
But, even if I have removed the code "#Html.Partial("_AllTweets", Model)" inside the div, the code still works fine, and without any problem and i can able to post the data to the action in the controller.
I got totally stuck at this point.
May I kindly know, what is the reason behind it and why so.... So I can understand it more better.
Thanks in advance...
But, even if I have removed the code #Html.Partial("_AllTweets",
Model) inside the div, the code still works fine, and without any
problem and i can able to post the data to the action in the
controller.
Yes it will work fine. The Html.Partial("_AllTweets",Model) renders the partial with the specified model on every page load. After page is loaded, then ajax is used to fill the div with id AllTweets.
Html.Partial("_AllTweets",Model) is usefull when you want to display, for example, already saved tweets from your database to user when the page first loads. And then ajax takes care of later updates.
I am a beginner in MVC. In my application, I list categories (as a button) on the left hand-side (inside a div). Just right to this div, there is another div which displays the items associated with the clicked category.
When a category button is clicked, I am planning to call the action of the controller that will retrieve the Items from the database and add these Items to the ViewModel (e.g., model.Items = db.Items...), and then call the View with the updated Model and display the Items.
However, I am curios if it is better to make an Ajax call here and use a partial view for displaying the Items of the clicked category.
If feel like between these two approaches in my scenario only difference will be the page-refresh, they should work the same in terms of speed since both of them require the same database call.
Can anyone share good practices in MVC for such scenarios?
Yes AJAX is faster and good way to update detail in same page without refresh.
For that you have to create JsonResult method in controller. It will give you result in Json.
Try JQuery Template plugin for repeated code.
<script id="trTemplate" type="text/x-jquery-tmpl">
<tr>
{{each $data}}
<td>${Col}</td>
{{/each}}
</tr>
</script>
<table id="containerTable">
</table>
AJAX Call
$.ajax({
url: 'Your JsonResult Method URL',
type: "GET",
data: data,
beforeSend: function () {
},
success: function (data) {
// It will pass data to template and template will bind parsing json
$('#trTemplate').tmpl(data).appendTo('#containerTable');
//Business logic
},
complete: function () {
// Your Code
}
});
Your JsonResult Method
[HttpPost]
public JsonResult GetData(ViewModel model)
{
// Your Code here
}
using ajax or not when displaying new records in asp.net mvc
Using jquery would be the best approach for this scenario. As you don't have to load the layout page which will have to render the scripts and stuff all over again. Stick on to Ajax calls in MVC as much as possible, The technologies are being improved and a lot of single page applications are out there, And if we still use a page load for every new request then there is no point.
Coming to comparison between passing back Partial View And Json Data. Which is better to use in the application design?
Both partial Views and Json data hold the same weight depending on the scenario.
When to use partial view: Lets say you have a Model and you have to build the view HTML by lot of if checks and loops and possibly some c# code ( in rare scenarios), etc, in such scenario using Partial view would be the better choice, Because if we try to build the same thing in Jquery using json data the complexity of the code required would be high compared to what can be done in Partial views, But still achieving it is possible but wont be that easy and we might make errors during development.
When to use Json Data: If the requirement is like updating a grid, generating dynamic drop down or dealing with some Jquery plugins in the page I think Json data would be better, as many plugins play with json data as the core requirement.
A Small Example Of Deciding Between Partial View And Json Data - interested folks and read through
Lets take a scenario where we have to display a grid of data. This is our initial requirement. So we can happily build our viewModel with data and pass it to our partial view and render the table using for loops. All set, Now the requirement changes and we are asked to build sorting, filtering and paging stuff in our table. So at present we look for a plugin that can be easily integrated with current code and yes the easy one to use at this scenario is Datatables. Ok, we wrote a small Jquery to apply the plugin to the table and all set we have the fancy stuff ready.
Now here is the tricky part, we are asked to add functionalities like add, edit, delete record from the table. Yes its possible but is little tricky to get it done in the best possible way with the current code which we have. What we tend to do is, when ever there is a change in the table we plan to recall the partial view. Which works fine but still asking to ourselves just to delete one record from the table is it good to reload the partial view again?? Definitely NOT,
What can we do? When ever there is any add, edit, delete operation we hit the controller to update the database and we can make the controller return a JSON data and just pass this Json data to the plugin API and refresh the table, This will be more neat and faster. So here you see JSON data would be the better choice. Also some might even want to make it more cleaner by just playing with that one record of data and writing up some jquery code to manipulate the table, which is absolutely fine, But it requires us to pass the Json data itself back from controller.
So having this done, we can go back and refactor our code to make partial view to use json data for the grid initially too or leave it as it is saying the initial load will be a partial view, but following operations would be a json result, which is fine but I feel let all the data related stuff come from one point.
So that explains how a simple module can change from being a partial view to then use Json data. There are scenarios where the story is the other way around, You have to pick the right one for the right work.
Ok, this idea might seem quite a bit crazy and it kindo' is (at least for me at my level).
I have a fairly standarad rails app (some content pages, a blog, a news block, some authentication). And I want to make it into a single page app.
What I want to accomplish is:
All the pages are fetched through AJAX like when using turbolinks, except that the AJAX returns only the view part (the yield part in the layout) withought the layout itself, which stays the same (less data in the responces, quicker render and load time).
The pages are mostly just static html with AngularJS markup so not much to process.
All the actual data is loaded separately through JSON and populated in the view.
Also the url and the page title get changed accordingly.
I've been thinking about this concept for quite a while and I just can't seem to come up with a solution. At this point I've got to some ideas on how this actualy might be done along with some problems I can't pass. Any ideas or solutions are greatly appreciated. Or might be I've just gone crazy and 3 small requests to load a page are worse then I big that needs all the rendering done on server side.
So, here's my idea and known problems.
When user first visits the app, the view template with angular markup is rendered regularly and the second request comes from the Angular Resource.
Then on ngClick on any link that adress is sent to ngInclude of the content wrapper.
How do I bind that onClick on any link and how can I exclude certain links from that bind (e.g. links to external authentication services)?
How do I tell the server not to render the layout if the request is comming from Angular? I though about adding a parameter to the request, but there might be a better idea.
When ngInclude gets the requested template, it fires the ngInit functions of the controllers (usually a single one) in that template and gets the data from the server as JSON (along with the proper page title).
Angular populates the template with the received data, sets the browser url to the url of the link and sets the page title to what it just got.
How do I change the page title and the page url? The title can be changed using jQuery, but is there a way through Angular itself?
Again, I keep thinking about some kind of animation to make this change more fancy.
Profit!
So. What do you guys think?
OK, in case enyone ever finds this idea worth thinking about.
The key can be solved as follows.
Server-side decision of whether to render the view or not.
Use a param in the ngInclude and set the layout: false in the controller if that param is present.
Have not found an easier way.
Client-side binding all links except those that have a particular class no-ajax
Here's a directive that does it.
App.directive('allClicks', function($parse) {
return {
restrict: 'A',
transclude: true,
replace: true,
link: function(scope, element, attrs) {
var $a = element.find('a').not($('a.no-ajax')),
fn = $parse(attrs['allLinks']);
$a.on('click', function(event) {
event.preventDefault();
scope.$apply(function() {
var $this = angular.element(event.target);
fn(scope, {
$event: event,
$href: $this.attr('href'),
$link: $this
});
});
});
}
};
})
And then use it on some wrapper div or body tag like <body ng-controller="WrapperCtrl" all-links="ajaxLink($href)"> and then in your content div do <div id="content" ng-include="current_page_template">
In your angular controller set the current_page template to the document.URL and implement that ajaxLink function.
$scope.ajaxLink = function(path) {
$scope.current_page_template = path+"?nolayout=true";
}
And then when you get your JSON with your data from the server don't forget to use history.pushState to set the url line and document.title = to setr the title.
I'm confused as to how to accomplish this. I have a page which, has a popup filter, which has some input elements and an "Apply" button (not a submit). When the button is clicked, two jquery .get() calls are made, which load a graph, a DataTables grid, photos, and miscellaneous info into four separate tabs. Inside the graph, if one clicks on a particular element, the user is taken to another page where the data is drilled down to a finer level. All this works well.
The problem is if the user decides to go back to the original page, but with the ajax generated graph/grid/photos etc. Originally I thought that I would store a session variable with the filter variables used to form the original query, and on returning to the page, if the session var was found, the original ajax call would be made again, re-populating the tabs.
The problem that I find with this method is that Coldfusion doesn't recognize that the session variable has been set when returning to the page using the browser's back button. If I dump out the session var at both the original and the second page, I can see the newly set var at the second page, and I can see it if I go to the original page through the navigation menu, but NOT if I use the back button.
SO.... from reading posts on here about ajax browser history plugins, it seems that there are various jquery plugins which help with this, including BBQ. The problem that I see with this approach is that it requires the use of anchor elements to trigger it, and then modifies the query string using the anchors' href attributes. I suppose that I could modify the page to include a hidden anchor.
My question, at long last is: is an ajax history plugin like BBQ the best way to accomplish this, or is there a way to make Coldfusion see the newly created session var when returning to the page via the back button? Or, should I consider re-architecting the page so that the ajax calls are replaced by a form submission back to the page instead?
Thanks in advance, as always.
EDIT: some code to help clarify things:
Here's the button that makes the original ajax calls:
<button id="applyFilter">APPLY</button>
and part of the js called on #applyFilter, wrapped in $(document).ready():
$('#applyFilter').click(function(){
// fill in the Photos tab
$.get('tracking/listPhotos.cfm',
{
id: id,
randParam: Math.random()
},
function(response){
$('#tabs-photos').html(response);
}
);
});
Finally, when the user calls the drill-down on the ajax generated graph, it uses the MaintAction form which has been populated with the needed variables:
function DrillDown() {
//get the necessary variables and populate the form inputs
document.MaintAction.action = "index.cfm?file=somepage.cfm&Config=someConfig";
document.MaintAction.submit();
}
and that takes us to the new page, from which we'd like to return to the first page but with the ajax-loaded photos.
The best bet is to use the BBQ method. For this, you don't have to actually include the anchor tags in your page; in fact, doing so would cause problems. This page: http://ajaxpatterns.org/Unique_URLs explains how the underlying process works. I'm sure a jQuery plugin would make the actual implementation much easier.
Regarding your other question, about how this could be done with session variables - I've actually done something similar to that, prior to learning about the BBQ method. This was specifically to save the state of a jqGrid component, but it could be easily changed to support any particular Ajax state. Basically, what I did was keep a session variable around for each instance of each component that stored the last parameters passed to the server via AJAX requests. Then, on the client side, the first thing I did was run a synchronous XHR request back to the server to fetch the state from that session variable. Using the callback method for that synchronous request, I then set up the components on my page using those saved parameters. This worked for me, but if I had to do it again I would definitely go with the BBQ method because it is much simpler to deal with and also allows more than one level of history.
Some example code based on your update:
$('#applyFilter').click(function(){
var id = $("#filterid").val(); // assumes the below id value is stored in some input on the page with the id "filterid"
// fill in the Photos tab
$.get('tracking/listPhotos.cfm',
{
id: id // I'm assuming this is what you need to remember when the page is returned to via a back-button...
//randParam: Math.random() - I assume this is to prevent caching? See below
},
function(response){
$('#tabs-photos').html(response);
}
);
});
/* fixes stupid caching behavior, primarily in IE */
$.ajaxSetup({ cache: false });
$.ajax({
async: false,
url: 'tracking/listPhotosSessionKeeper.cfm',
success: function (data, textStatus, XMLHttpRequest)
{
if (data.length)
{
$("#filterid").val(data);
$('#applyFilter').trigger('click');
}
}
});
This is what you need on the client-side to fetch the state of the photo list. On the server side, you'll need to add this modification to tracking/listPhotos.cfm:
<cfset session.lastUsedPhotoFilterID = URL.id>
And add this new one-line file, tracking/listPhotosSessionKeeper.cfm:
<cfif IsDefined("session.lastUsedPhotoFilterID")><cfoutput>#session.lastUsedPhotoFilterID#</cfoutput></cfif>
Together these changes will keep track of the last ID used by the user, and will load it up each time the page is rendered (whether via a back button, or simply by the user revisiting the page).
I'm requesting an ASP.net MVC view into a live box and the view contains form fields that have been marked up with attributes to be used by JQuery's unobtrusive validators plug-in.
The client script is not however working and my theory is that its because the validation framework is only being triggered on page load which has long since passed by the time the MVC view has been loaded into the live box.
Thus how can I let the validation framework know that it has new form fields to fix up?
Cheers, Ian.
var $form = $("form");
$form.unbind();
$form.data("validator", null);
$.validator.unobtrusive.parse(document);
// Re add validation with changes
$form.validate($form.data("unobtrusiveValidation").options);
You may take a look at the following blog post. And here's another one.
Another option, rather trick, which worked for me. Just add following line in the beginning of the partial view which is being returned by ajax call
this.ViewContext.FormContext = new FormContext();
Reference
For some reason I had to combine bjan and dfortun's answers...
So I put this in my view:
#{
this.ViewContext.FormContext = new FormContext();
}
And this execute this after the ajax call finishes:
var form = $("#EnrollmentForm");
form.unbind();
form.data("validator", null);
$.validator.unobtrusive.parse(document);
form.validate(form.data("unobtrusiveValidation").options);
I had a similar issue. I had a form that was using Ajax requests to re-display a part of the form with different form fields. I used unobtrusive validation by manually doing it on the client side using the
#Html.TextBoxFor
for my text boxes. For some reason the validation works when attempting to submit with invalid fields (i.e., the text boxes get outlined in red and the appropriate error messages display with the content I put in the
data_val_required
attribute, for example.
However, after I click a button that makes an Ajax request to modify the form with different fields and then submit again, only the red outline on the invalid fields display, but no error messages are rendered.
bjan's trick worked for me, but I still can't see what was causing the issue. All the HTML necessary to carry out the client-side validation was there I just can't figure out why the error message attribute values wouldn't display.
All I can think of is that the jQuery validation code doesn't make a second attempt to check the form fields after a submit was made.