How to start automatic download of a file in Internet Explorer?

Question

How do I initialize an automatic download of a file in Internet Explorer?

For example, in the download page, I want the download link to appear and a message: "If you download doesn't start automatically .... etc". The download should begin shortly after the page loads.

In Firefox this is easy, you just need to include a meta tag in the header, <meta http-equiv="Refresh" content="n;url"> where n is the number of seconds and url is the download URL. This does not work in Internet Explorer. How do I make this work in Internet Explorer browsers?

Solution 1

SourceForge uses an <iframe> element with the src="" attribute pointing to the file to download.

<iframe width="1" height="1" frameborder="0" src="[File location]"></iframe>

(Side effect: no redirect, no JavaScript, original URL remains unchanged.)

Solution 2

I hate when sites complicate download so much and use hacks instead of a good old link.

Dead simple version:

<a href="file.zip">Start automatic download!</a>

It works! In every browser!

---

If you want to download a file that is usually displayed inline (such as an image) then HTML5 has a download attribute that forces download of the file. It also allows you to override filename (although there is a better way to do it):

<a href="report-generator.php" download="result.xls">Download</a>

Version with a "thanks" page:

If you want to display "thanks" after download, then use:

<a href="file.zip" 
   onclick="if (event.button==0) 
     setTimeout(function(){document.body.innerHTML='thanks!'},500)">
 Start automatic download!
</a>

Function in that setTimeout might be more advanced and e.g. download full page via AJAX (but don't navigate away from the page — don't touch window.location or activate other links).

The point is that link to download is real, can be copied, dragged, intercepted by download accelerators, gets :visited color, doesn't re-download if page is left open after browser restart, etc.

That's what I use for ImageOptim

Solution 3

I recently solved it by placing the following script on the page.

setTimeout(function () { window.location = 'my download url'; }, 5000)

I agree that a meta-refresh would be nicer but if it doesn't work what do you do...

Solution 4

I had a similar issue and none of the above solutions worked for me. Here's my try (requires jquery):

$(function() {
  $('a[data-auto-download]').each(function(){
    var $this = $(this);
    setTimeout(function() {
      window.location = $this.attr('href');
    }, 2000);
  });
});

Usage: Just add an attribute called data-auto-download to the link pointing to the download in question:

<p>The download should start shortly. If it doesn't, click
<a data-auto-download href="/your/file/url">here</a>.</p>

It should work in all cases.

Solution 5

A simple bit of jQuery solved this problem for me.

$(function() {
   $(window).bind('load', function() {
      $("div.downloadProject").delay(1500).append('<iframe width="0" height="0" frameborder="0" src="[YOUR FILE SRC]"></iframe>'); 
   });
});

In my HTML, I simply have

<div class="downloadProject"></div>

All this does is wait a second and a half, then append the div with the iframe referring to the file that you want to download. When the iframe is updated onto the page, your browser downloads the file. Simple as that. :D

View more solutions

Comments

[David Robbins]

I like the simplicity of your answer.

[Budda]

The lack of such approach is that browser waits for all banners loading... Sometime that takes some time and user is unable to get file due to stupid banners...

[Budda]

Could somebody explain how that should work? I don't see delay for 'n' seconds... Thank you in advance.

[sharat87]

should be setTimeout(function () { window.location = 'my download url'; }, 5000) (no strings to setTimeout please)

[Vishwanath Dalvi]

@DavidRobbins and i like the simplicity of your comment. :D

[Tim Post]

Regarding your flag, that would be something to raise on our meta site. I believe this has come up before, however (there are quite a few similar complaints, but I can't find one directly matching this)

[Kohsuke Kawaguchi]

One word of caution. At least in my Chrome 21, the attempt to follow @href gets cancelled if the script from setTimeout tries to move to another page. So this really only works when you can show the thank you page inline.

[daGUY]

I like that. Nice and clean and you can add it to any link you want. I love finding cool new uses for the data-* attribute!

[Ralph]

@Cerberus: because of the general idea.

[raydowe]

I totally agree with using a good old link, but there is a problem with this: clients. Sometimes they want what they want no matter what you try and tell them. They've seen it on other sites, and that's what they want on their site.

[Kornel]

@raydowe if you make link work well, then the client won't know. Unless you have clients that view page source and make nonsensical complaints…

[raydowe]

@porneL I don't understand. I thought you were suggesting just using a link to the file INSTEAD of the automatic download after X seconds.

[Kornel]

@raydowe yes, but you can display post-download screen that looks the same as if it was implemented using the messy meta/iframe hack — load the 'thank you' page using AJAX from onclick handler. So if client's requirement is to have such page, you can use the link and still fulfill the requirement.

[3on]

This is also how audible.com does it apparently.

[Cosine]

Ooh that is clever. You could make it display:none, too. Also, @Budda, some simple Javascript (Oh, no, Javascript?!?!) can add that HTML in after n seconds: HTML: <iframe id="download" width="1" height="1" style="display:none"></iframe> Javascript: function startDownload () { document.getElementById("download").src="[File Location]"; } setTimeout (startDownload, n * 1000);

[Dmitriy Konovalov]

Brilliant! Thanks for the tip!

[George Sovetov]

Probably, the most simple solution.

[Cheruvian]

The problem is currently safari, notice user wants to automatically start the download and the only way to do that with a link is to snag the element and invoke .click(). safari doesn't support click().

[Kornel]

@Cheruvian think what action user takes before you "automatically" start the download. I'll bet that action is a click that goes to your "automatic" page. So move the download one step earlier!

[Cheruvian]

True but the issue is when you need to generate download content based on something. The current solution is 1 click for generating content and 1 click to download which can be pretty dirty

[Nathan Hornby]

I like the idea of using this, but it will have issues with certain file types. For example if you wish to provide a download link for a PDF or image (or any content that a browser can display), the browser would just try and display it silently in the hidden iframe.

[Bogdan Varlamov]

Often times you need to do additional work before the download actually starts--like submitting various parameters from JS to an API so that you can prepare the file for downloading. You can't really do that if it's just an href that the user is clicking which links directly to the file content.

[Kornel]

@BogdanVarlamov: You could include the data in the URL (dynamically modify href if necessary) and have that URL redirect to actual file, e.g. <a href="/api?param=meters&redirect=/download/file">. You can even change href from onclick handler and return true, so you can update it the instant user clicks.

[zealotous]

@NathanHornby, use Content-Disposition: attachment; filename=manual.pdf header on server side and everything will be ok.

[Bogdan Varlamov]

@porneL: Then one is sending potentially sensitive data in an insecure manner (via URL) instead of being able to send it in the body over HTTPS.

[Kornel]

@BogdanVarlamov HTTPS is a tunnel hiding all of HTTP protocol, including URLs, so there's no difference. And if filename contains sensitive information, maybe don't put that in the filename anyway? (it'll leak later when saved to disk)

[Bogdan Varlamov]

There are some issues specifically related to URLs which are not typically encountered for body params. Check out stackoverflow.com/questions/323200/… and blog.httpwatch.com/2009/02/20/… for reasons

[Kornel]

@BogdanVarlamov 1. GET via click vs GET via javascript/iframe hack is still GET (and issues like cross-site referrer leak don't apply to download links you create). 2. sensitive information should go to a cookie with secure flag, and then you won't need to worry about hiding potentially-insecure URL hacks with wonky JS hacks.

[Bogdan Varlamov]

@porneL, Nobody forces you to use a GET method in JS. You can (and should) use a POST/PUT to submit the sensitive parameters in JS in the BODY (see my first comments)... with a url-click you are limited only to the GET the browser sends.

[Kornel]

@BogdanVarlamov You can still use <form method="post"><button> for JS-hack-free equivalent of <a href>. Other answers to this question use frames or location.href and don't account for POST anyway.

[Bogdan Varlamov]

@porneL, I think what I imagine is a POST/PUT that submits sensitive info which returns a URL based on that criteria which you sent, that the JS client then redirects to for executing the file download. You can avoid this by sticking the sensitive info into the URL as params, sure, but that has security problems as I pointed out. Are you saying that if you had a form submit the params, you could have the server redirect to the right file based on the params?

[Kornel]

@BogdanVarlamov POST can send the file to download directly, so there's never need to perform any redirect or put any parameters in any URL at any point, e.g. see imageoptim.com/mozjpeg

[user890332]

or put into code behind like this: Page.ClientScript.RegisterStartupScript(Me.GetType, "filedownload", "setTimeout(function () { window.location = 'http://" & Page.Request.Url.Authority & filename & "'; }, 0)", True)

[ash123]

works in firefox only if its from the same origin download link

[EL missaoui habib]

@ash123 try this : function downloadURI(uri, name) { var link = document.createElement("a"); link.download = name; link.href = uri; link.click(); }

[Shi Jie Tio]

How to apply in IE 11, I really no idea on it can anyone share this to me?

[Teson]

IE: pointing the iframe to a src of stackoverflow.com/questions/24529910/… (Cendak) solved this for me.

[Robo Robok]

Why is Content-Disposition better than download attribute for setting file name?

[mix3d]

Except for some browsers + the content is one that could be viewed directly in the browser (like video files), does not actually force the download in all circumstances.

[mix3d]

Not all browsers let you click an anchor tag with JS. Safari, for example, especially if the url is cross-domain

[johny why]

@zealotous Where on the server do you put Content-Disposition: attachment; filename=manual.pdf header on server side?

[halfmoonhalf]

This answer saves my day. I have a follow up question: if I change "auto-download.zip" to a url like "stackoverflow.com/questions/156686/…", it does not work. My purpose is to download the stackoverflow page as html.

[JDQ]

I really like the href=<handler> download=<output> usage: I have only ever seen download used as an attribute without the optional value.

[Chris - Jr]

@johnywhy, you place in the header of the server response. So in pseudocode - myResponse.setHeader('Content-Disposition','attachment; filename=manual.pdf'); send(myResponse);