I've built a web application that makes heavy use of the jQuery.ajax() function on dozens of different pages. Naturally, each use of jQuery.ajax() is passed a configuration object with a function to serve as the callback for the 'success' property.
I've now run into a situation where I would like to globally perform a certain action on multiple pages if the server returns a certain error code for any given ajax request. I'm thinking about somehow overriding a part of the jQuery library by including javascript in the header of my pages which would perform a conditional check operation, and then continue to execute whatever callback function was passed in the configuration object. Is this possible? I'm currently reviewing the source for the jQuery library to see if I can do this.
I think you need to look at the Global Ajax Event Handlers.
http://api.jquery.com/category/ajax/global-ajax-event-handlers/
In jQuery 1.5+ , you can accomplish this via the converters option for $.ajax. Normally, the converters are just supposed to convert data received form the web server, and pass it off to the success callback, but you can run other code in there if you want.
Drop this code into a js file that is embedded on all of your pages.
$.ajaxSetup({
// the converter option is a list of dataType-to-dataType conversion functions
// example: converters['text json'] is a function that accepts text and returns an object
// note: to redefine any single converter, you must redefine all of them
converters: {
"text json": function (text) {
// NOTE: this converter must return a parsed version of the json
var result = jQuery.parseJSON(text);
if (result.errorMessage) {
// catch the error message and alert
alert(result.errorMessage)
}
return result;
},
"text html": true,
"* text": window.String,
"text xml": jQuery.parseXML
}
});
Full example:
http://jsfiddle.net/waltbosz/8a8fZ/
I ended up overriding the success callback in a global manner using $.ajaxSetup, like this:
$.ajaxSetup({
beforeSend: function(xhr, options) {
var originalSuccess = options.success
options.success = function(data, textStatus, jqXhr) {
// trigger some sort of event here for interested listeners
originalSuccess.apply(this, arguments)
}
}
})
Related
I am making an ajaxSubmit call to a web service, that will return XML.
When I make a call to the same service using XMLHttpRequest.send, the response is correct.
However if I use:
$(form).ajaxSubmit({
error: function _(response) {
$(iframeEl).remove();
config.error.call(scope, Thunderhead.util.JSON.decode(response));
},
success: function _(response) {
console.log(response);
$(iframeEl).remove();
var result = response;
config.success.call(scope, result);
},
iframe: true
});
This returns the correct XML response, but all tags have been transformed to lowercase.
I've checked the call in the Network tab in the developer console, and the case is correct in there, but when it is returned by the ajaxSubmit, it is lowercase.
Does anyone know what is causing this or how to rectify it?
Are you using Malsups jQuery form plugin
This plugin does a lot of toLowerCase transforms, I've not looked too closely but it does seem to lowercase the tag names of something, so this is probably your culprit.
I'd recommend refactoring to using a simple jQuery.ajax() call instead
$(form).on('submit', function(e) {
var url = $(form).attr('action');
e.preventDefault();
$.ajax( url, {
error: function _(jqXHResponse) {
// your code
},
success: function _(response) {
console.log(response);
// your code
}
});
This might be happening, because js is assuming xml as an answer. There is no difference for most xml-parsers which case is used in xml tag names.
I suggest trying to change response data type.
For example there is such option in jQuery.ajax method: http://api.jquery.com/jquery.ajax/ (named dataType). I would try using "text" dataType if case is really important.
Some further issues arose from this in the end, so just posting my eventual solution in case anyone else has this problem. I'm fairly new to javascript, so this might have been obvious to most, but it might help someone else out.
The success callback can actually take in 3 parameters, the third of which (arg2) is the actual response from the request, without any changes from the Malsups form plugin.
So in the end, the solution was simply to use this third parameter instead of the response parameter.
$(form).ajaxSubmit({
error: function _(response) {
$(iframeEl).remove();
config.error.call(scope, Thunderhead.util.JSON.decode(response));
},
success: function _(response, arg1, arg2) {
console.log(response);
$(iframeEl).remove();
var result = response;
config.success.call(scope, arg2.responseXML);
},
iframe: true
});
I want to use John Resig's pretty date for replacing my ugly time stamps with some nice-to-read time specification.
So I thought about using the following unobtrusive html markup:
<span data-type="prettyDate">25.04.2012 10:16:37</span>
Acording to that I use following Javascript/jQuery to prettify the date:
$(function() {
$('[data-type="prettyDate"]').prettyDate();
}
My problem is that I don't know how to deal with markup that is loaded using ajax because that would not be caught since it does not yet exist when the DOM ready event fires. Reacting to events on "ajaxed" elements is pretty easy using the on handler. But this is not an event.
You have to call .prettyDate() after each Ajax response is added to the DOM. A simple way to do that is to set a global complete handler with ajaxComplete.
You can use jQuery to target dynamic content before it's actually been inserted into the document, something like:
success: function(html) {
var $html = $(html);
$html.find('[data-type="prettyDate"]').prettyDate();
$(somewhere in document).append($html);
}
What you want to do to get the best performance out of this is have a function which get called on the data as it gets returned from the ajax callback. That way you can prettify your date before adding them to the DOM.
You don't want to call pretty date on element in the DOM every time as you will process date already done too.
So, something like this.
$.ajax({
url:'someurl',
success: function(data) {
var $content = $(data).find('[data-type="prettyDate"]').prettyDate();
$('#mycontainer').append($content);
}
});
or have an helper function which you call
function prettify(data) {
return $(data).find('[data-type="prettyDate"]').prettyDate();
}
or even better hook into the ajax call so that it is done for all html content
There have been a number of cases where I needed certain code to execute after every AJAX call. I'm not sure if it's considered the "correct" solution but I simply decided to create my own wrapper method and use that whenever I needed to make an AJAX request. It typically looks something like this:
AJAXLoadData: function (url, data, successCallBack) {
$.ajax({
type: "GET",
data: data,
url: url,
contentType: "application/json; charset=utf-8",
dataType: "json",
success: function (msg) {
// Code I want to execute with every AJAX call
// goes here.
// Then trigger the callback function.
if (successCallBack) successCallBack(msg);
},
error: function (msg) {
alert("Server error.");
}
});
}
In my case this made it particularly convenient to create a javascript caching system for static HTML files.
You could incorporate this code into your ajax success callback function. When the ajax is done and you update your page, also run the code to prettify the dates.
This is one of the things .on() is for. (In the olden days, .live() would have been used.)
According to documentation:
If html is specified, any embedded JavaScript inside the retrieved
data is executed before the HTML is returned as a string. Similarly,
script will execute the JavaScript that is pulled back from the
server, then return nothing.
How to prevent this?
I have js that shall modify the content that is obtained through ajax. Executing it before the html is returned makes no sense as it does not have content to work on (at least in my case).
my code:
function do_ajax(url) {
$.ajax({
cache: false,
url : url,
success: function(response, status, xhr) {
var ct = xhr.getResponseHeader("content-type") || "";
if (ct.indexOf('script') > -1) {
try {
eval(response);
}
catch(error) {}
} else {
var edit_dialog = $('<div class="edit_dialog" style="display:hidden"></div>').appendTo('body');
edit_dialog.html(response);
edit_dialog.dialog({ modal:true, close: function(event, ui) { $(this).dialog('destroy').remove(); } });
}
},
error: function (xhr, ajaxOptions, thrownError) {
alert(xhr.status);
alert(thrownError);
}
});
}
the script received by ajax is executed twice. First by me in the eval(response), then jquery execute it again (as described in the documentation)
Lee's answer already adequately addresses the case of HTML responses - scripts embedded in these are not in fact executed automatically unless you add the HTML to the DOM, contrary to the erroneous documentation you quoted.
That leaves the other case asked about in your question title - preventing script responses from being executed automatically when received. You can do this easily using the dataType setting.
$.ajax('myscript.js', {
dataType: 'text',
success: function (response) {
// Do something with the response
}
})
Setting dataType to 'text' will cause jQuery to disregard the Content-Type header returned from the server and treat the response like plain text, thus preventing the default behaviour for JavaScript responses (which is to execute them). From the (recently corrected) docs:
The type of pre-processing depends by default upon the Content-Type of the response, but can be set explicitly using the dataType option. If the dataType option is provided, the Content-Type header of the response will be disregarded.
...
If text or html is specified, no pre-processing occurs. The data is simply passed on to the success handler, and made available through the responseText property of the jqXHR object.
jQuery.ajax does not evaluate scripts on return when requesting HTML. The passage you quoted in the question was in fact a long-standing error in the documentation, fixed as of April 2014. The new docs have this to say (emphasis mine):
"html": Returns HTML as plain text; included script tags are evaluated when inserted in the DOM.
...
If text or html is specified, no pre-processing occurs. The data is simply passed on to the success handler, and made available through the responseText property of the jqXHR object.
The scripts are evaluated in this case when you call
edit_dialog.html(response);
If you don't want to evaluate the scripts before inserting your response in to the DOM, you should be able to do something like:
edit_dialog.html($($.parseHTML(response)));
parseHTML is the key in that by default it removes script tags. However, be aware that parseHTML is NOT XXS safe and if your source is unknown this is still a security concern.
The documentation states that any embedded Javascript inside the retrieved data will be executed before the HTML is returned as a string. If you want to then alter whatever you have retrieved using your ajax call, you can do so within the succes property:
$.ajax({
url: "example.html",
type: "GET",
dataType: "html",
succes: function(data){
// Example: alert(data);
// Do whatever you want with the returned data using JS or jQuery methods
}
});
That's one of the really annoying things about jQuery that it executes javascript on response.
Other frameworks like MooTools disable script execution from responses unless you specifically set that you want them executed which is a much better approach.
The only way I could figure to prevent scripts being executed is to add a custom dataFilter
Its easy enough but I think it should be default and an ajax option to enable script execution if you want it (I've never had a use for it and other frameworks disable by default for security etc.)
Example
$.ajax('uri',{
dataFilter: function(data, type)
{
type = type || 'text';
if(type=='html'||type=='text'){
/*return data.replace(/<script.*>.*?<\/script>/gi, '');*/
return data.replace(/<script.*?>([\w\W\d\D\s\S\0\n\f\r\t\v\b\B]*?)<\/script>/gi, '');
}
return data;
}
, success: function(data)
{
// whatever
}
});
** UPDATED **
Needs that crazy regex to cover more script tag instances
NOTE
if dataType hasnt been set in options it will be undefined in dataFilter so I just default it to text for the filter - if you remove that line then it will only work if dataType is explicitly set.
I have a web application that gets data with a get call. When it is called, it sets a static variable. After the call is complete I want to get the value, but the complete function that I call is not working, its calling it to fast because its null.
$.get('../Controls/myfullcontrol.aspx', {}, function (data) {
$('#SiteFullControl').html(data);
}).complete(function () {
$('#FullControlCount').html('Full Control (<%=GlobalClass.GlobalVariables.User.Count %>)');
$('#LoadingAjax').dialog('close');
});
Using Firebug or Chrome's Inspector, look at the XHR and ensure that the script is actually returning something.
Are you looking for data returned to the complete function?
Then you need a parameter in the function:
}).complete(function (data) {
You don't need the $.complete() method because the $.get() contains allready an success callback.
http://api.jquery.com/jQuery.get/
You could work with jQuery the Promise interface (introduced in 1.5).So you can chain multiple callbacks on a single request...
Take look: http://api.jquery.com/jQuery.get/#jqxhr-object
jQuery.get( url, [data,] [success(data, textStatus, jqXHR),] [dataType] )
$.get('ajax/test.html', function(data) {
//STUFF AT SUCCESS
});
Why does this not work? anybody:
In my code I have:
$.getJSON("http://isp123.co.uk/cw/NorthWales/Test.txt?jsoncallback=?",
function(data){
//This never gets executed
alert('here');
});
The text file can be viewed here:
http://isp123.co.uk/cw/NorthWales/Test.txt
This is not a JSONP response:
({"name" : "hello world"});
If you had a proper JSONP response, then your code should work.
The question mark in the "callback=?" part of the URL is changed by jQuery before making the request, your JSONP server needs to be able to dynamically create the JSONP "function" in response to the unique jQuery request. If you can't dynamically create your JSONP, perhaps you could use YQL/Yahoo pipes to turn it into JSONP?
This pipe should do the trick, to see if it works, use this URL instead in your getJSON function: http://pipes.yahoo.com/pipes/pipe.run?u=http%3A%2F%2Fisp123.co.uk%2Fcw%2FNorthWales%2FTest.txt&_id=332d9216d8910ba39e6c2577fd321a6a&_render=json&_callback=?
I just tried this and it worked:
$.getJSON("http://pipes.yahoo.com/pipes/pipe.run?u=http%3A%2F%2Fisp123.co.uk%2Fcw%2FNorthWales%2FTest.txt&_id=332d9216d8910ba39e6c2577fd321a6a&_render=json&_callback=?", function(data){
//This always gets executed!!!
alert('here');
});
I don't know if you know enough about JSONP but this is not JSONP
?({"name" : "hello world"});
It really should be something like this http://isp123.co.uk/cw/NorthWales/Test.txt?jsoncallback=foo
foo({"name" : "hello world"});
From the jQuery.getJson manual page:
Important: As of jQuery 1.4, if the JSON file contains a syntax error, the request will usually fail silently. Avoid frequent hand-editing of JSON data for this reason. JSON is a data-interchange format with syntax rules that are stricter than those of JavaScript's object literal notation. For example, all strings represented in JSON, whether they are properties or values, must be enclosed in double-quotes. For details on the JSON format, see http://json.org/.
Your JSON is invalid according to http://jsonlint.com/
Here Clearly mentioned
As of jQuery 1.5, setting the jsonp
option to false prevents jQuery from
adding the "?callback" string to the
URL or attempting to use "=?" for
transformation. In this case, you
should also explicitly set the
jsonpCallback setting
and read jsonpCallback section
jsonpCallback,
Specify the callback function name for
a JSONP request. This value will be
used instead of the random name
automatically generated by jQuery. It
is preferable to let jQuery generate a
unique name as it'll make it easier to
manage the requests and provide
callbacks and error handling. You may
want to specify the callback when you
want to enable better browser caching
of GET requests. As of jQuery 1.5, you
can also use a function for this
setting, in which case the value of
jsonpCallback is set to the return
value of that function
Probably worth using jQuery.ajax() - http://api.jquery.com/jQuery.ajax/
You can pass in the dataType as "jsonp" and then jQuery takes care of all the callback business, but more importantly you can specify a function to run when there's an error, which may help you:
$.ajax({
dataType: "jsonp",
success: function(d) {console.log(d);},
error: function() { console.log("error") } //do your debugging in here
//add other parameters such as URL, etc
});
The error function you define can be passed 3 variables, read up on it on the ajax() page on the jQuery docs (linked at the beginning of my post) to find out more about that and how to use them.
Your problem lies with how your server is outputting the information. In the link you've supplied, the assumption is that any name placed in the ?jsonpcallback should result in wrapping the JSONP code in a function with that same name. It, however, is not the case.
So the next option is this: use a static function name in your server file and wrap the code. (e.g. use foo(<jsonp>) and stick with it) Then, you have to explicitly tell jQuery that we are going to use a specific function name (leave jQuery with the assumption it's supplying (and thus receiving) that name back, when in-fact you're just supplying it server side and filling in the blanks).
Once you have your file setup, use something like the following:
$.ajax({
// setup the request
url: 'http://isp123.co.uk/cw/NorthWales/Test.txt',
crossDomain: true,
dataType: 'jsonp',
jsonp: false,
jsonpCallback: 'foo', // "supply" the jsonp function (pseudo-defined)
// function to call when completed
complete: function(data){
alert(data);
}
// just in case, catch the error
error: function(j,t,e){
alert('AJAX Error');
}
});
So now when jQuery makes the call and it thinks it's supplying the callback, it's really just getting the server-defined callback in return. So, for the above to work, your text file should look something like this:
foo({name:"Hello, World!"});
Also, if you can, change your header to application/javascript, though this is some-what optional.