Using XMLHttpRequest

Types of requests

A request made via XMLHttpRequest can fetch the data in one of two ways, asynchronously or synchronously. The type of request is dictated by the optional async argument (the third argument) that is set on the XMLHttpRequest.open() method. If this argument is true or not specified, the XMLHttpRequest is processed asynchronously, otherwise the process is handled synchronously. A detailed discussion and demonstrations of these two types of requests can be found on the synchronous and asynchronous requests page. You can't use synchronous requests outside web workers as it freezes the main interface.

Handling responses

There are several types of response attributes defined for the XMLHttpRequest() constructor. These tell the client making the XMLHttpRequest important information about the status of the response. Some cases where dealing with non-text response types may involve some manipulation and analysis are outlined in the following sections.

Analyzing and manipulating the responseXML property

If you use XMLHttpRequest to get the content of a remote XML document, the responseXML property will be a DOM object containing a parsed XML document. This could prove difficult to manipulate and analyze. There are four primary ways of analyzing this XML document:

1. Using XPath to address (or point to) parts of it.
2. Manually Parsing and serializing XML to strings or objects.
3. Using XMLSerializer to serialize DOM trees to strings or to files.
4. RegExp can be used if you always know the content of the XML document beforehand. You might want to remove line breaks, if you use RegExp to scan with regard to line breaks. However, this method is a "last resort" since if the XML code changes slightly, the method will likely fail.

Processing a responseText property containing an HTML document

If you use XMLHttpRequest to get the content of a remote HTML webpage, the responseText property is a string containing the raw HTML. This could prove difficult to manipulate and analyze. There are three primary ways to analyze and parse this raw HTML string:

1. Use the XMLHttpRequest.responseXML property as covered in the article HTML in XMLHttpRequest.
2. Inject the content into the body of a document fragment via fragment.body.innerHTML and traverse the DOM of the fragment.
3. RegExp can be used if you always know the content of the HTML responseText beforehand. You might want to remove line breaks, if you use RegExp to scan with regard to line breaks. However, this method is a "last resort" since if the HTML code changes slightly, the method will likely fail.

Handling binary data

Although XMLHttpRequest is most commonly used to send and receive textual data, it can be used to send and receive binary content. There are several well tested methods for coercing the response of an XMLHttpRequest into sending binary data. These involve utilizing the overrideMimeType() method on the XMLHttpRequest object and is a workable solution.

const req = new XMLHttpRequest();
req.open("GET", url);
// retrieve data unprocessed as a binary string
req.overrideMimeType("text/plain; charset=x-user-defined");
/* … */

However, more modern techniques are available, since the responseType attribute now supports a number of additional content types, which makes sending and receiving binary data much easier.

For example, consider this snippet, which uses the responseType of "arraybuffer" to fetch the remote content into an ArrayBuffer object, which stores the raw binary data.

const req = new XMLHttpRequest();

req.onload = (e) => {
  const arraybuffer = req.response; // not responseText
  /* … */
};
req.open("GET", url);
req.responseType = "arraybuffer";
req.send();

For more examples check out the Sending and Receiving Binary Data page.

Monitoring progress

XMLHttpRequest provides the ability to listen to various events that can occur while the request is being processed. This includes periodic progress notifications, error notifications, and so forth.

Support for DOM progress event monitoring of XMLHttpRequest transfers follows the specification for progress events: these events implement the ProgressEvent interface. The actual events you can monitor to determine the state of an ongoing transfer are:

progress

The amount of data that has been retrieved has changed.

load

The transfer is complete; all data is now in the response

const req = new XMLHttpRequest();

req.addEventListener("progress", updateProgress);
req.addEventListener("load", transferComplete);
req.addEventListener("error", transferFailed);
req.addEventListener("abort", transferCanceled);

req.open();

// …

// progress on transfers from the server to the client (downloads)
function updateProgress(event) {
  if (event.lengthComputable) {
    const percentComplete = (event.loaded / event.total) * 100;
    // …
  } else {
    // Unable to compute progress information since the total size is unknown
  }
}

function transferComplete(evt) {
  console.log("The transfer is complete.");
}

function transferFailed(evt) {
  console.log("An error occurred while transferring the file.");
}

function transferCanceled(evt) {
  console.log("The transfer has been canceled by the user.");
}

We add event listeners for the various events that are sent while performing a data transfer using XMLHttpRequest.

The progress event handler, specified by the updateProgress() function in this example, receives the total number of bytes to transfer as well as the number of bytes transferred so far in the event's total and loaded fields. However, if the lengthComputable field is false, the total length is not known and will be zero.

Progress events exist for both download and upload transfers. The download events are fired on the XMLHttpRequest object itself, as shown in the above sample. The upload events are fired on the XMLHttpRequest.upload object, as shown below:

const req = new XMLHttpRequest();

req.upload.addEventListener("progress", updateProgress);
req.upload.addEventListener("load", transferComplete);
req.upload.addEventListener("error", transferFailed);
req.upload.addEventListener("abort", transferCanceled);

req.open();

Progress events come in for every chunk of data received, including the last chunk in cases in which the last packet is received and the connection closed before the progress event is fired. In this case, the progress event is automatically fired when the load event occurs for that packet. This lets you now reliably monitor progress by only watching the "progress" event.

One can also detect all three load-ending conditions (abort, load, or error) using the loadend event:

req.addEventListener("loadend", loadEnd);

function loadEnd(e) {
  console.log(
    "The transfer finished (although we don't know if it succeeded or not).",
  );
}

Note there is no way to be certain, from the information received by the loadend event, as to which condition caused the operation to terminate; however, you can use this to handle tasks that need to be performed in all end-of-transfer scenarios.

Get last modified date

function getHeaderTime() {
  console.log(this.getResponseHeader("Last-Modified")); // A valid GMTString date or null
}

const req = new XMLHttpRequest();
req.open(
  "HEAD", // use HEAD when you only need the headers
  "your-page.html",
);
req.onload = getHeaderTime;
req.send();

Do something when last modified date changes

Let's create two functions:

function getHeaderTime() {
  const lastVisit = parseFloat(
    window.localStorage.getItem(`lm_${this.filepath}`),
  );
  const lastModified = Date.parse(this.getResponseHeader("Last-Modified"));

  if (isNaN(lastVisit) || lastModified > lastVisit) {
    window.localStorage.setItem(`lm_${this.filepath}`, Date.now());
    isFinite(lastVisit) && this.callback(lastModified, lastVisit);
  }
}

function ifHasChanged(URL, callback) {
  const req = new XMLHttpRequest();
  req.open("HEAD" /* use HEAD - we only need the headers! */, URL);
  req.callback = callback;
  req.filepath = URL;
  req.onload = getHeaderTime;
  req.send();
}

And to test:

// Let's test the file "your-page.html"
ifHasChanged("your-page.html", function (modified, visit) {
  console.log(
    `The page '${this.filepath}' has been changed on ${new Date(
      modified,
    ).toLocaleString()}!`,
  );
});

If you want to know if the current page has changed, refer to the article about document.lastModified.

Cross-site XMLHttpRequest

Modern browsers support cross-site requests by implementing the Cross-Origin Resource Sharing (CORS) standard. As long as the server is configured to allow requests from your web application's origin, XMLHttpRequest will work. Otherwise, an INVALID_ACCESS_ERR exception is thrown.

Bypassing the cache

A cross-browser compatible approach to bypassing the cache is appending a timestamp to the URL, being sure to include a "?" or "&" as appropriate. For example:

http://example.com/bar.html -> http://example.com/bar.html?12345
http://example.com/bar.html?foobar=baz -> http://example.com/bar.html?foobar=baz&12345

As the local cache is indexed by URL, this causes every request to be unique, thereby bypassing the cache.

You can automatically adjust URLs using the following code:

const req = new XMLHttpRequest();

req.open("GET", url + (/\?/.test(url) ? "&" : "?") + new Date().getTime());
req.send(null);

Security

The recommended way to enable cross-site scripting is to use the Access-Control-Allow-Origin HTTP header in the response to the XMLHttpRequest.

XMLHttpRequests being stopped

If you conclude with an XMLHttpRequest receiving status=0 and statusText=null, this means the request was not allowed to be performed. It was UNSENT. A likely cause for this is when the XMLHttpRequest origin (at the creation of the XMLHttpRequest) has changed when the XMLHttpRequest is subsequently open(). This case can happen, for example, when one has an XMLHttpRequest that gets fired on an onunload event for a window, the expected XMLHttpRequest is created when the window to be closed is still there, and finally sending the request (in other words, open()) when this window has lost its focus and another window gains focus. The most effective way to avoid this problem is to set a listener on the new window's DOMActivate event which is set once the terminated window has its unload event triggered.

Specifications

Browser compatibility