I have the following code:
var User = {
get: function (options) {
var self = this;
$.ajax({
url: options.url,
success: function (data, response) {
self.nextPageUrl = data.pagination.next_page;
options.success(data, response);
}
});
},
nextPage: function (success) {
this.get({
url: this.nextPageUrl,
success: success
});
}
}
User.get({
url: 'https://cache.getchute.com/v2/albums/aus6kwrg/assets',
success: function (data, response) {
// Through `data.pagination.next_page` I can get the URL
// of the next page.
}
});
User.nextPage({
success: function (data, response) {
// Here I want to make the same request but using the next_page
// based on the next related to the previous' one.
}
});
The problem
Basically, I want to perform the nextPage() operation based on the antecessor request (User.get()), but due to its asynchronousity, the nextPage() method doesn't know the this.nextPageUrl property—it returns undefined as expected.
Finally, the question is: can someone think in a way to keep the current syntax flow but solving this approach? Actually, is there a way?
And no, I'm not available to make a synchronous request.
General knowledge
I thought to use an event mechanism to deal with this: when the request is made and .nextPage() is called, then try to listen to an event to be emitted for x seconds, then I expected the this.nextPageUrl property to be available in that event-based scope.
What do you guys think?
DISCLAIMER: The logic of next_page is preprocessed by the server and only then is sent to the client. I have no option to use an increment/decrement behavioral operation.
If you want to play with this problem, click here for the jsFiddle.
jsFiddle Demo
There are a couple of options. You could bind the setter of the property to also call the nextPage, you could poll from calling nextPage every n milliseconds until the nextPageUrl property was populated, you could use a promise, you could use a nextPageQueue.
I think that a queue may be the simplest form of completing this. I also think it may be useful to have User store some local variables in this situation, and that the use of a function object may be more inline with that.
It would look like this
var User = new function(){
var pageQueue = [];
var get = this.get = function (options) {
$.ajax({
url: options.url,
dataType: 'JSON',
success: function (data, response) {
options.success(data, response);
var nextPageUrl = data.pagination.next_page;
if( pageQueue.length > 0 ){
pageQueue[0](nextPageUrl);
pageQueue.splice(0,1);
}
}
});
};
var nextPage = this.nextPage = function (options) {
pageQueue.push(function(nextPageUrl){
get({
url: nextPageUrl,
success: options.success
});
});
};
};
and your calls would not change.
User.get({
url: 'https://cache.getchute.com/v2/albums/aus6kwrg/assets',
success: function (data, response) {
// Through `data.pagination.next_page` I can get the URL
// of the next page.
console.log('get');
console.log(data);
console.log(response);
}
});
User.nextPage({
success: function (data, response) {
// Here I want to make the same request but using the next_page
// based on the next related to the previous' one.
console.log('next');
console.log(data);
console.log(response);
}
});
You can grab a reference to your User object before making the asynchronous request.
var User = {
get: function (options) {
var self = this;
$.ajax({
url: options.url,
success: function (data, response) {
self.nextPageUrl = data.pagination.next_page;
options.success(data, response);
}
});
},
You can modify your get method and eliminate your nextPage method entirely:
var User = {
url: 'https://cache.getchute.com/v2/albums/aus6kwrg/assets',
get: function (options) {
var self = this;
self.pageXhr = $.ajax({
url: self.nextPageUrl ? self.nextPageUrl : self.url,
dataType: 'JSON',
success: function (data, response) {
self.nextPageUrl = data.pagination.next_page;
self.pageXhr = false;
options.success(data, response);
}
});
}
}
Then whenever you call User.get, it will either call the current page or the next page. I am not sure about the context of when you want to get the subsequent pages, but if you need the requests to be queued you can wait for the existing request to finish before triggering the next request. For example:
if (self.pageXhr) {
self.pageXhr = self.pageXhr.then(function() {
return $.ajax({
url: self.nextPageUrl ? self.nextPageUrl : self.url,
dataType: 'JSON',
success: function (data, response) {
self.nextPageUrl = data.pagination.next_page;
options.success(data, response);
}
});
});
}
Related
Here I'm using Angularjs1.x and here is my condition. If condition is success then show the table otherwise throw an error. I know some code if its Success.
AngCtrl.Js
$scope.BtnCall = function () {
ServiceCntrl.CallData().then(function (d) {
$scope.EmpData = d.data;
});
}
AngService.Js
eApp.service("ServiceCntrl", function ($http) {
var xx = '';
xx= $http({
data: formData,
method: 'post',
url: Url,
datatype: "json"
}).success(function (rsp) {
RspData = rsp;
return RspData;
}).error(function (rsp) {
console.log('Error');
});
return xx;
};
Your x.then receives two functions x.then(function(){}, function(){}); first function is called when promise is successfully resolved and second function is called if promise is rejected(failed).
If your service function is return $http promise then your first function can have a parameter named(anything you like) and it will have response data that you can use. Second function can receive error parameters if any.
you should look at angular $http service documentation.
If your service is returning the promise of the get request, then you can write
$scope.BtnCall = function () {
var x = ServiceCntrl.CallData();
x.then(function(response) {
//Success callback
//code on success goes here
//response.data
}, function(response) {
//error callback
//code on error goes here
// server returns response with an error status.
});
you can use the ng-show/ng-hide to show and hide the contents on the html page.
You can write your success/fail code as the following:
$scope.BtnCall = function() {
var x = ServiceCntrl.CallData();
x.then(function(result) {
// Success code here
// Do something and resolved the promise
}, function(reason) {
// Error code here
// You may handle/reject the reason in params
});
});
See also the Angular docs for $q.
The AngularJS $http service makes a request to the server, and returns a response
The example above executes the $http service with an object as an argument. The object is specifying the HTTP method, the url, what to do on success, and what to do on failure.
$scope.BtnCall = function () {
ServiceCntrl.CallData().then(function (d) {
$scope.EmpData = d.data;
});
}
AngService.Js :
eApp.service("ServiceCntrl", function ($http) {
var xx = '';
xx= $http({
data: formData,
method: 'post',
url: Url,
datatype: "json"
}).success(function (rsp) {
RspData = rsp;
return RspData;
}).error(function (rsp) {
console.log('Error');
});
return xx;
};
Pardon as I am new to AngularJS.
I have the following nested $http call,
$http({
method: 'GET',
url: host1,
params: {
'format': 'json',
}
}).then(function successCallback(response) {
for (var i = 0; i < response.data.length; i++) {
$http({
method: 'GET',
url: response.data[i]['url'] + "packet-trace/base",
params: {
'format': 'json',
}
}).then(function successCallback(response2) {
//Retrieve some information from first $http call
var doSomething = response.data[i]['information'];
var doSomething2 = doSomething + response2.data['information'];
}, function errorCallback(response2) {
//Error
});
}
}, function errorCallback(response) {
//Error
});
I need to retrieve data from the first $http call, and then retrieve data from the $http, and use both of these data as part of my logic. However, I am unable to access the data from the first $http call. The loop counter 'i' always equal the length of response.data.
How can I access the first the data of $http call?
Additionally, are there any specific coding conventions or specific API I can use to call $http sequentially? Nesting $http calls gets messy and difficult to maintain.
Thanks.
The reason why this happens is because the loop probably has finished iterating before the AJAX call succeeds. At this stage i will obviously equal response.data.length.
You could capture the value outside the AJAX call and then use it in the success callback so that you don't depend on the i variable inside the callback:
for (var i = 0; i < response.data.length; i++) {
// Capture the value you need here
var someValue = response.data[i]['information'];
$http({
method: 'GET',
url: response.data[i]['url'] + "packet-trace/base",
params: {
'format': 'json',
}
}).then(function successCallback(response2) {
// Use the captured value from the first call here
var doSomething = someValue;
var doSomething2 = doSomething + response2.data['information'];
}, function errorCallback(response2) {
//Error
});
}
Well that is because of asynchronous nature of javascript. I suggest using async for your use case. May look a bit complex, but once you understand async concepts, it is breeze to use. This what you can do:
async.waterfall(
[
function(callback){
$http({
method: 'GET',
url: host1,
params: {
'format': 'json',
}
}).then(function successCallback(response) {
callback(null, response);
});
},
function(callback, response){
async.times(response.data.length, function(i,next){
$http({
method: 'GET',
url: response.data[i]['url'] + "packet-trace/base",
params: {
'format': 'json',
}
}).then(function successCallback(response2) {
//Retrieve some information from first $http call
var doSomething = response.data[i]['information'];
var doSomething2 = doSomething + response2.data['information'];
next(null,doSomething, doSomething2);
}, function errorCallback(err) {
next(err, null);
});
});
}
],
function(err,doSomethingArr, doSomething2Arr){
// get your arrays here
});
I have two buttons that both performs AJAX call:
$("#save").click(function() {
$.ajax({
type: "POST",
url: saveEntryURL,
data: { id: $("#id").val() },
success: function(r) {
...
},
error: function(r) {
...
}
})
})
$("#tag-as-final").click(function() {
$.ajax({
type: "POST",
url: finalizeEntryURL,
data: { id: $("#id").val() },
success: function(r) {
...
},
error: function(r) {
...
}
})
})
The requirement is that when the user click the finalize button, the system will first perform a save before actually tagging it as final. To reuse the code attached to the save button, I call the onclick listener of the save button before the actual AJAX call like this:
$("#tag-as-final").click(function() {
$("#save").click()
^^^^^^^^^^^^^^^^^^
$.ajax({
type: "POST",
url: finalizeEntryURL,
But it will not do "save-and-finalize-after" behavior since both AJAX calls are asynchronous. I need to run one after another, but cannot afford to make the AJAX call of the save button synchronous (I'm doing also a lot of other things while the tagging occurs occurs). I know this would be silly but I'm thinking something similar to...
$("#tag-as-final").click(function() {
$("#save").click().peformAsyc()
^^^^^^^^^^^^
$.ajax({
type: "POST",
url: finalizeEntryURL,
...that will force it to finish performing first the chained function before continuing, but I know that is not available. Is there any way to do this? My current work-around is placing the same save AJAX function inside the finalize AJAX function, though it doesn't allow me to code DRY (Don't Repeat Yourself):
$("#tag-as-final").click(function() {
$.ajax({
type: "POST",
url: saveEntryURL,
data: { id: $("#id").val() },
success: function(r) {
...
$.ajax({
type: "POST",
url: finalizeEntryURL,
data: { id: $("#id").val() },
success: function(r) {
...
},
error: function(r) {
...
}
})
},
error: function(r) {
...
}
})
})
It's pretty simple, you are better using jquery "promises". Like so:
var generalSettings = { }; //Settings for AJAX call.
var jqXHR = $.ajax(generalSettings); //Do AJAX call.
generalSettings.data = 'newdata'; //update generalSettings
jqXHR.done(function(data){
$.ajax(generalSettings); //New Petition with updated settings.
});
This is using ES6 promises and jQuery promises:
function doAjaxAsPromise(settings){
return new Promise(function(resolve){
var jqXHR = $.ajax(settings);
jqXHR.done(function(data){
resolve(data);
});
});
}
var settings = { };
var petition = doAjaxAsPromise(settings);
var secondpetition = petition.then(function(data){
//work with data
//new settings
var settings = { };
return doAjaxAsPromise(settings);
});
var thirdpetition = secondpetition.then(function(data){
//work with data
//new settings
var settings = { };
return doAjaxAsPromise(settings);
});
//If needed to reuse settings object outside promise scope:
//var settings = Object.create(settings);
Some other nice thing you can do for code reuse:
function save(settings) {
var prom = doAjaxAsPromise(settings);
return prom.then(function(data){
//do something with your data.
});
}
function tagAsFinal(savedPromise, settings){
return savedPromised.then(function(){
var prom = doAjaxAsPromise(settings);
return prom.then(function(data){
//work with data;
});
});
}
$('save').on('click', function(){
save(settings); //settings = $.ajax settings.
});
$('tagAsFinal').on('click', function(){
var generalSettings = { };
var settingsone = Object.create(generalSettings);
var settingstwo = Object.create(generalSettings);
var saved = save(settingsone); //$.ajax settings.
tagAsFinal(saved, settingstwo);
});
//Can still be reduced.
I am using $.when and .done to make sure that the close window happens after the data is saved. But, this doesn't seem to work as expected.
The workflow is that, user clicks on a button "Save and Close", which should save the data first, trigger print and close the window. But the save data and close window happens at the same time which makes the print fail.
I have read about when..then and deferred object. Tried to implement it here the following code, sometimes it work but most of the time it would break.
$("#btnSaveAndClose").click(function (event) {
$.when(zSaveSomeData()).done(function (value) {
zCloseMyWindow();
});
});
function zSaveSomeData() {
return zSaveMasterData(masterdata, function () {
return zSaveDetailData();
});
};
function zSaveMasterData(masterdata, fnAfterSave) {
return $.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/masterdata/',
data: JSON.stringify(masterdata),
success: function (data) {
fnAfterSave();
}
});
};
function zSaveDetailData() {
var selectedDataGroups;
// some logic here
zSaveDetails(selectedDataGroups);
};
function zSaveDetails(selectedDataGroups) {
var deferred = $.Deferred();
$.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/detaildata/',
data: JSON.stringify(selectedDataGroups),
success: function (data) {
var printableGroupIDs = [];
$.each(data, function () {
if (this.IsPrintable)
printableGroupIDs.push(this.ID);
});
if (printableGroupIDs.length > 0) {
zPrintGroups(printableGroupIDs);
}
deferred.resolve('done');
}
});
zAuditSave();
return deferred.promise();
};
function zPrintGroups(newGroupIDs) {
// calls external program to print groups
};
function zCloseWindow() {
window.close();
};
function zAuditSave() {
$.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/audit'
success: function (data) {
}
});
};
Only thing is that the save calls other methods inside to same master and details data. There are couple of ajax calls too. An unusual thing is that after the data is saved, there is a call to VB code that actually triggers a Print. I am so confused on why would close window fire before the other methods are executed. Any help would be appreciated.
For me the code is overly divided into functions, with some doing little more than fronting for others.
I would prefer to see the click handler as a comprehensive master routine which sequences three promise-returning functions zSaveMasterData(), zSaveDetails() and zAuditSave(), then closes the window. Thus, some of the current functions will be subsumed by the click handler.
$("#btnSaveAndClose").click(function(event) {
zSaveMasterData(masterdata).then(function() {
var selectedDataGroups;
/* some logic here */
var detailsSaved = zSaveDetails(selectedDataGroups).then(function(data) {
var printableGroupIDs = $.map(data, function (obj) {
return obj.IsPrintable ? obj.ID : null;
});
if (printableGroupIDs.length > 0) {
// calls external program to print groups
}
});
// Here, it is assumed that zSaveDetails() and zAuditSave() can be performed in parallel.
// If the calls need to be sequential, then the code will be slightly different.
return $.when(detailsSaved, zAuditSave());
}).then(function() {
window.close();
});
});
function zSaveMasterData(masterdata) {
return $.ajax({
type: 'POST',
url: '/api/masterdata/',
contentType: 'application/json',
data: JSON.stringify(masterdata),
});
};
function zSaveDetails(selectedDataGroups) {
return $.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/detaildata/',
data: JSON.stringify(selectedDataGroups)
});
};
function zAuditSave() {
return $.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/audit'
});
};
Note the returns in the three functions with ajax calls. These returns are vital to the sequencing process.
A potentially bigger issue, not addressed in the question (nor in this answer) is how to recover from errors. Presumably, the database will be inconsistent if the sequence of saves was to fail part way through. It may well be better to ditch this client-side sequencing approach in favour of a server-side transaction that the client sees as a single operation.
The problem here is your code doesn't depend on when fnAfterSave() has completed.
Short answer: don't mix success methods, callbacks, and promises - use one pattern and stick to it - and the easiest pattern to use is promises.
$("#btnSaveAndClose").click(function (event) {
zSaveSomeData().then(function() { zCloseMyWindow(); });
});
function zSaveSomeData() {
return zSaveMasterData(masterdata).then(function(data) { zSaveDetailData() });
};
function zSaveMasterData(masterdata) {
return $.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/masterdata/',
data: JSON.stringify(masterdata)
});
//remove success callback here as it breaks the chaining
};
It seems like your problem is that you are doing asynchronous things inside an ajax success callback. The promise returned by $.ajax still resolves immediately after the response is received - and executes your done callback before the asynchronous zSaveDetailData() has finished.
So, to chain asynchronous actions, always use then. Use it even for synchronous actions, it makes the sequence clear.
Don't use success callbacks when you're working with promises. You also don't need deferreds. You might want to have a look at these generic rules as well, especially that you never must forget to return promises from async functions that you want to await.
$("#btnSaveAndClose").click(function (event) {
zSaveSomeData().then(zCloseMyWindow);
});
function zSaveSomeData() {
return zSaveMasterData(masterdata).then(zSaveDetailData);
}
function zSaveMasterData(masterdata) {
return $.ajax({
type: 'POST',
contentType: 'application/json',
url: '/api/masterdata/',
data: JSON.stringify(masterdata),
});
}
function zSaveDetailData() {
var selectedDataGroups;
// some logic here
return zSaveDetails(selectedDataGroups);
// ^^^^^^
}
function zSaveOrderGroups(selectedDataGroups) {
return $.ajax({
// ^^^^^^
type: 'POST',
contentType: 'application/json',
url: '/api/detaildata/',
data: JSON.stringify(selectedDataGroups)
}).then(function(data) {
// ^^^^^^^^^^^^^^^^^^^^^^
var printableGroupIDs = [];
$.each(data, function () {
if (this.IsPrintable)
printableGroupIDs.push(this.ID);
});
if (printableGroupIDs.length > 0) {
return zPrintGroups(printableGroupIDs);
// ^^^^^^
}
}).then(zAuditSave);
// ^^^^^^^^^^^^^^^^^
}
function zPrintGroups(newGroupIDs) {
// calls external program to print groups
}
function zCloseWindow() {
window.close();
}
function zAuditSave() {
return $.ajax({
// ^^^^^^
type: 'POST',
contentType: 'application/json',
url: '/api/audit'
});
}
Can we make a callback in a chain like this?
Widget.update(...).onUpdate(function(data){
console.log('updated');
});
current code,
var Gateway = {};
Gateway.put = function(url, data, callback) {
$.ajax({
type: "POST",
dataType: "xml",
url: url,
data: data,
async: true,
success: function (returndata,textStatus, jqXHR) {
callback(returndata);
}
});
};
var Plugin = function() {};
Plugin.prototype = {
update : function(options, callback) {
/// other stuff
Gateway.put($url, $data, function(data){
callback(data);
}
return this;
}
}
usage,
var Widget = new Plugin();
Widget.put({
button: options.button
}, function(data){
console.log('updated');
});
but ideally,
Widget.update(...).onUpdate(function(data){
console.log('updated');
});
EDIT:
at jsfiddle.
What you are trying to do will work however you need to pass your callback to update
Widget.update(yourOptions, function(data){
console.log('updated');
});
You could also return your ajax request directly and chain onto it
var Gateway = {};
Gateway.put = function(url, data) {
return $.ajax({
type: "POST",
dataType: "xml",
url: url,
data: data,
async: true
});
};
var Plugin = function() {};
Plugin.prototype = {
update : function(options) {
/// other stuff
return Gateway.put($url, $data);
}
}
var Widget = new Plugin();
Widget.update(yourOptions).done(function() {
console.log('updated');
});
I really like the callback hell coding style, but sometimes it hurts. As suggested by other users, have you already heard about promises?
The core idea behind promises is that a promise represents the result of an asynchronous operation.
As suggested in the link above - that proposed a standard for them - once polyfill'd the browser using
<script src="https://www.promisejs.org/polyfills/promise-done-6.1.0.min.js"></script>
you will be able to create new Promise's, hence to use their nice done() attribute.
You will end up with
Plugin.prototype.update = function (options) {
return new Promise(function (fullfill, reject) {
Gateway.put($url, $data, function (data) {
fullfill(data);
});
});
};
That is, Plugin.prototype.update returns a promise.
Widget.update(...).done(function(data){
console.log('updated');
});
I have not tested the code, but the spirit is that. :)
EDIT: Using promises is awesome. I simply don't like when people discover them, use them in newer parts of the codebase but finally do not refactor the rest of it.