We Keep Coding

Internet explorer passes undefined to success callback when fetching 30MB of data or more

I have a problem with fetching JSON data from server in Internet explorer. I am using standard jquery ajax call to the server to retrieve binary data representing file, which is then opened in FlowPaper's PDF viewer. Everything is OK for files of sizes up to 20MB. When I try to get data for files with size 30MB and up, the call comes back as successfull, but data sent to success callback is undefined. This behavior happens in Internet explorer only, but in Chrome or Opera everything works for all file sizes.
JS code is straightforward and simple
var id = "documentID";
var url = 'API/GetDocumentURL';
var data = JSON.stringify({
"DocumentID": id
});
var type = 'POST';
var async = true;
var success = function (data) {
debugger; // in IE data is undefined, other browsers have results stored in this variable
};
$.ajax({
url: url,
type: type,
async: async,
data: data,
success: success
});
Jquery version is 1-8-2, but I have a feeling that this is more a problem of Internet explorer, and not Jquery. Internet explorer version is 11.
Does anyone have any idea why Internet explorer is firing success callback, but not passing in the data retrieved with AJAX call?

Returning the MIMEType of application/json; charset=utf8 caused this same behavior for me in IE8. Changing it to application/json; made IE8 magically start functioning. Check what your server is returning for a MIMEType and see if fiddling with that is causing an issue.
Edit: Actually what was causing the real problem is that it should be charset=utf-8 and not as shown above without the hyphen.
For more infor jquery Ajax response "undefined" with Internet Explorer

Get jQuery $.ajax to send data in body for GET [duplicate]

The service API I am consuming has a given GET method that requires the data be sent in the body of the request.
The data required in the body is a list of id's separated by hypen and could potentially be very large and thus it must be sent in the body otherwise it will likely foobar somewhere in the browsers/proxies/webservers etc chain. Note I don't have control over the service or API so please don't make suggestions to change it.
I am using the following jQuery code however observing the request/response in fiddler I can see that the "data" I am sending is ALWAYS converted and appended to the query string despite me setting the "processData" option to false...
$.ajax({
url: "htttp://api.com/entity/list($body)",
type: "GET",
data: "id1-id2-id3",
contentType: "text/plain",
dataType: "json",
processData: false, // avoid the data being parsed to query string params
success: onSuccess,
error: onError
});
Anyone know how I can force the "data" value to be sent in the body of the request?

In general, that's not how systems use GET requests. So, it will be hard to get your libraries to play along. In fact, the spec says that "If the request method is a case-sensitive match for GET or HEAD act as if data is null." So, I think you are out of luck unless the browser you are using doesn't respect that part of the spec.
You can probably setup an endpoint on your own server for a POST ajax request, then redirect that in your server code to a GET request with a body.
If you aren't absolutely tied to GET requests with the body being the data, you have two options.
POST with data: This is probably what you want. If you are passing data along, that probably means you are modifying some model or performing some action on the server. These types of actions are typically done with POST requests.
GET with query string data: You can convert your data to query string parameters and pass them along to the server that way.
url: 'somesite.com/models/thing?ids=1,2,3'

we all know generally that for sending the data according to the http standards we generally use POST request.
But if you really want to use Get for sending the data in your scenario
I would suggest you to use the query-string or query-parameters.
1.GET use of Query string as.
{{url}}admin/recordings/some_id
here the some_id is mendatory parameter to send and can be used and req.params.some_id at server side.
2.GET use of query string as{{url}}admin/recordings?durationExact=34&isFavourite=true
here the durationExact ,isFavourite is optional strings to send and can be used and req.query.durationExact and req.query.isFavourite at server side.
3.GET Sending arrays
{{url}}admin/recordings/sessions/?os["Windows","Linux","Macintosh"]
and you can access those array values at server side like this
let osValues = JSON.parse(req.query.os);
if(osValues.length > 0)
{
for (let i=0; i<osValues.length; i++)
{
console.log(osValues[i])
//do whatever you want to do here
}
}

Just in case somebody ist still coming along this question:
There is a body query object in any request. You do not need to parse it yourself.
E.g. if you want to send an accessToken from a client with GET, you could do it like this:
const request = require('superagent');
request.get(`http://localhost:3000/download?accessToken=${accessToken}`).end((err, res) => {
if (err) throw new Error(err);
console.log(res);
});
The server request object then looks like {request: { ... query: { accessToken: abcfed } ... } }

You know, I have a not so standard way around this. I typically use nextjs. I like to make things restful if at all possible. If I need to make a get request I instead use post and in the body I add a submethod parameter which is GET. At which point my server side handles it. I know it's still a post method technically but this makes the intention clear and I don't need to add any query parameters. Then the get method handles a get request using the data provided in the post method. Hopefully this helps. It's a bit of a side step around proper protocol but it does mean there's no crazy work around and the code on the server side can handle it without any problems. The first thing present in the server side is if(subMethod === "GET"){|DO WHATEVER YOU NEED|}

Jquery not appending timestamp for cache

I want to send an ajax request using sortable from jquery-ui. When I re-sort a list, for some reason, the ajax call isn't appending timestamp onto the querystring to avoid caching. Why won't this code work?
$('#categorylist').sortable({
placeholder: "ui-state-highlight",
update:function(e, ui){
var categorylist = $('#categorylist').sortable('serialize', {attribute:'data-id'});
$.ajax({
url:'actions_category.php?action=rank',
method:'post',
data:categorylist,
cache:false
});
}
});

It will only append the time stamp for GET and HEAD requests. It should not be needed for other types of requests except in IE8 for POST requests where you have already issued a GET request against the URL.
http://api.jquery.com/jQuery.ajax/

Note the supported HTTP Methods
http://api.jquery.com/jQuery.ajax/
cache (default: true, false for dataType 'script' and 'jsonp')
Type: Boolean
If set to false, it will force requested pages not to be cached by the browser. Note: Setting cache to false will only work correctly with HEAD and GET requests. It works by appending "_={timestamp}" to the GET parameters. The parameter is not needed for other types of requests, except in IE8 when a POST is made to a URL that has already been requested by a GET.

Mootools Request() works only once per session IE9

Issue:
I have a Mootools Request object which sends a GET request to a PHP script. The script returns an array of IDs. In every browser except IE each time a new instantiation of the Request object is made it hits the PHP backend script - no problem. BUT with IE it only every hits the PHP script once per browser session. If I clear the cache it flows through to the backend again OK, but after a browser restart or cache refresh.
Question:
Is IE blocking my requests with caching somehow?
If so I there something I can change to prevent IE from blocking the Requests?
Note:
The issue is present on IE9 and on
IE9 in compatibility mode. I have not
tested on older IE versions.
Using
page-wide no-cache is not an option.
I used a error_log() statement in my script to check whether is was getting reached or not.
Here's my code:
requestIDs : function() {
new Request({
url : "/identity/idGenerator.php?generate_ids",
method : "get",
onSuccess : function(response) {
this.container.set('html', '');
var idList = response.split(",");
console.log(idList.join(","));
}.bind(this)
}).send();
}
});

You need to prevent IE from caching the request, here is one solution.
And the right solution which is built in mootools is initializing the Request with the config param noCache:true
var myRequest = new Request({...
...
noCache:true,
...
...);

Use method: 'post' to prevent IE from caching.

Is there a Request.IsMvcAjaxRequest() equivalent for jQuery?

I prefer to use jQuery with my ASP.NET MVC apps than the Microsoft Ajax library. I have been adding a parameter called "mode" to my actions, which I set in my ajax calls. If it is provided, I return a JsonViewResult. If it isn't supplied, I assume it was a standard Http post and I return a ViewResult.
I'd like to be able to use something similar to the IsMvcAjaxRequest in my controllers when using jQuery so I could eliminate the extra parameter in my Actions.
Is there anything out there that would provide this capability within my controllers or some simple way to accomplish it? I don't want to go crazy writing code since adding a single parameter works, it just isn't ideal.

Here's an except from MVC RC1 release notes - Jan 2009
IsMvcAjaxRequest Renamed to IsAjaxRequest
The IsMvcAjaxRequest method been
renamed to IsAjaxRequest. As part of
this change, the IsAjaxRequest method
was updated to recognize the
X-Requested-With HTTP header. This is
a well known header sent by the major
JavaScript libraries such as
Prototype.js, jQuery, and Dojo.
The ASP.NET AJAX helpers were updated to send this header in
requests. However, they continue to
also send it in the body of the form
post in order to work around the issue
of firewalls that strip unknown
headers.
In other words - it was specifically renamed to be more 'compatible' with other libraries.
In addition, for anyone who hasnt read the full release notes but has been using previous versions - even as recent as the beta - I STRONGLY recommend you read them in full. It will save you time in future and most likely excite you with some of the new features. Its quite surprising how much new stuff is in there.
Important note: You will need to make sure you upgrade the .js file for MicrosoftAjax.MVC (not the exact name) if upgrading to RC1 from the Beta - otherwise this method won't work. It isn't listed in the release notes as a required task for upgrading so don't forget to.

See Simons answer below. The method I describe here is no longer needed in the latest version of ASP.NET MVC.
The way the IsMvcAjaxRequest extension method currently works is that it checks Request["__MVCASYNCPOST"] == "true", and it only works when the method is a HTTP POST request.
If you are making HTTP POST requests throug jQuery you could dynamically insert the __MVCASYNCPOST value into your request and then you could take advantage of the IsMvcAjaxRequest extension method.
Here is a link to the source of the IsMvcAjaxRequest extension method for your convenience.
Alternatively, you could create a clone of the IsMvcAjaxRequest extension method called
IsjQueryAjaxRequest that checks Request["__JQUERYASYNCPOST"] == "true" and you could dynamically insert that value into the HTTP POST.
Update
I decided to go ahead and give this a shot here is what I came up with.
Extension Method
public static class HttpRequestBaseExtensions
{
public static bool IsjQueryAjaxRequest(this HttpRequestBase request)
{
if (request == null)
throw new ArgumentNullException("request");
return request["__JQUERYASYNCPOST"] == "true";
}
}
Checking from an action if a method is a jQuery $.ajax() request:
if (Request.IsjQueryAjaxRequest())
//some code here
JavaScript
$('form input[type=submit]').click(function(evt) {
//intercept submit button and use AJAX instead
evt.preventDefault();
$.ajax(
{
type: "POST",
url: "<%= Url.Action("Create") %>",
dataType: "json",
data: { "__JQUERYASYNCPOST": "true" },
success: function(data) {alert(':)');},
error: function(res, textStatus, errorThrown) {alert(':(');}
}
);
});

Why don't you simply check the "X-Requested-With" HTTP header sent automatically by most Javascript libraries (like jQuery) ?
It has the value 'XMLHttpRequest' when a GET or POST request is sent.
In order to test it you should just need to check the "Request.Headers" NameValueCollection in your action, that is :
if (Request.Headers["X-Requested-With"] == "XMLHttpRequest")
return Json(...);
else
return View();
This way, you can simply differentiate regular browser requests from Ajax requests.

Ok, I have taken this one step farther and modified my jQuery file to load the additional parameter into the post data, so I don't have to repeat the "__JQUERYASYNCPOST: true" for every call to post. For anybody that's interested, here's what my new definition for $.post looks like:
post: function(url, data, callback, type) {
var postIdentifier = {};
if (jQuery.isFunction(data)) {
callback = data;
data = {};
}
else {
postIdentifier = { __JQUERYASYNCPOST: true };
jQuery.extend(data, postIdentifier);
}
return jQuery.ajax({
type: "POST",
url: url,
data: data,
success: callback,
dataType: type
});
}
I added the "postIdentifier" variable as well as the call to jQuery.extend. Now the Helper explained in spoon16's response works without having to add any thing special to my page-level jQuery code.