Checking file has downloaded - IE problem

optik

I have a download script I've written that allows a file to be downloaded a certain number of times before the link expires, the script also removes the download link from the users account when the file has been fully downloaded.

Now this all works fine, except IE's security settings are messing things up and I can't ask each user to start fiddling with these settings even though that fixes the problem.

In IE7+ 8 when the user clicks the download link IE pops up its info bar and the user has to select to download the file.

What's happening though is that somehow IE has managed to already download the file from the server without actually saving it anywhere accessible to the user, this prompts the PHP script to expire the link so when IE goes for it's supposed first attempt and error is thrown up because the link has expired.

Does anyone have any suggestions about what could be done here, I need to to preserve the functionality of the script and I will definitely not resort to instructing user to adjust their security settings.

bradgrafelman

optik wrote:
What's happening though is that somehow IE has managed to already download the file from the server without actually saving it anywhere accessible to the user, this prompts the PHP script to expire the link so when IE goes for it's supposed first attempt and error is thrown up because the link has expired.

Actually, no it hasn't.

What's most likely happening is that IE sends a normal GET request just like it would for any other link, except when your server tells IE that it's a binary file IE suddenly throws up red flags and aborts the connection, prompting the user to confirm that the link they clicked on is a binary file they want to download to their computer.

optik wrote:
Does anyone have any suggestions about what could be done here, I need to to preserve the functionality of the script

If you right click on the link and choose Save Target As (or whatever IE calls it), are you able to download the file normally? If so, I would suggest including those instructions near the download link.

optik

The link is not a direct link so can't be saved as target, for security purposes I'm using PHP's header function to force the download.

In the script I'm using the bytes read information from the fread function to check whether the all the bytes have been read, and even when the download is blocked all the bytes are read, maybe this is just PHP reading the bytes even though they are not being served?

If that's the case there doesn't seem to be anyway to check from PHP whether the user has either started or finished the download in question.

bradgrafelman

optik wrote:
The link is not a direct link so can't be saved as target, for security purposes I'm using PHP's header function to force the download.

That shouldn't matter. Have you tried the Save Target As option already?

optik wrote:
even when the download is blocked all the bytes are read

Are you reading the file in small chunks, or all at once?

optik wrote:
If that's the case there doesn't seem to be anyway to check from PHP whether the user has either started or finished the download in question.

[man]connection_aborted/man

optik

I pretty sure it will attempt to download the php file and fail, in any case I don't want to confuse users with browser specific instructions.

I'm reading the file in one block however the file I'm testing is very small, which I'm sure isn't helping but some of the downloads are for small files so I have to make allowance for that.

I'll try out the connection aborted function and see if that helps.

Thanks

bradgrafelman

optik wrote:
I pretty sure it will attempt to download the php file and fail

No, it won't. IE doesn't know what a PHP file is, nor does a URL have any connection to what "file" the webserver uses to handle the request. Try the Save Target As option and find out.

optik wrote:
in any case I don't want to confuse users with browser specific instructions.

So you want to solve a browser-specific problem but you don't want to hear about any browser-specific fixes?

You can't have your cake and eat it too, y'know.

optik

Ok I've tried save target and that didn't work, I'm using an iFrame in the page the link goes to initiate the download so it's just downloading the container html page where the link goes to.

I've also tried the connection aborted function and it's made no difference.