Getting the extension right on a download [resolved]

[Acidic]

I'm using this code to download a file:

<?
//Download a file
$file = $_GET['file'];
header("Cache-Control: no-cache, must-revalidate");
header("Pragma: no-cache");
header("Content-type: application/x-zip");
header("Content-disposition: attachment; filename=".$file);
readfile($file);
?>

I have to use this so that .html files can be downloaded and not simply viewed. I've noticed thought that in IE there is no extension to the file. in FireFox it gives the corrrect file xtension, but not in IE. I've even tried reading the user agent and manually entering an extension for IE. Still didn't work.

Actually, the file even has different names in the two browsers, neither of which are the name of the real file, they still download correctly though.

Can someone please tell me why IE doesn't give an extension and how to give it one.

[Electroman]

I use this:

$FName = substr($File,strrchr($File,"/"));

header("Expires: Mon, 26 Nov 1962 00:00:00 GMT");
header("Last-Modified: " . gmdate("D,d M Y H:i:s") . " GMT");
header("Cache-Control: no-cache, must-revalidate");
header("Pragma: no-cache");
header("Content-type: application/x-zip");
header("Content-disposition: inline; filename=$FName");
header("Location: $File"); 

This works in IE haven't tried it in anything else but nobodies complained about it not working in FireFox that I know people have used with it.$File is the Full URL of the file to be downloaded. I notice you are using Attachemnet and I am using inline though. You might notice though the Full URL in mine is shaved down to only the filename for the filename.
Thing is I'm using the Location thing instead of the ReadFile(). There is meant to be some problems when using the ReadFile() with IE mind.

[Acidic]

What have I done wrong?
here's the URL I go to:
'download.php?File=www.horspool.org.uk/sid/JavaScript/".$pre.$a.".html'

it says it cannot fomd the document. and the URL becomes:
http://www.horspool.org.uk/sid/www.h...ript/0002.html

[Acidic]

ok. now it just loads up the page, in both browsers.

I used EXACTLY the code you gave me, pasted that into download.php

then here's the link which goes to it:

print "<input type='button' value='Get Script' onClick=\"goto_URL('download.php?File=http://www.horspool.org.uk/sid/JavaScript/".$pre.$a.".html')\">"; 

[Electroman]

Posted by Acidic
What have I done wrong?
here's the URL I go to:
'download.php?File=www.horspool.org.uk/sid/JavaScript/".$pre.$a.".html'

it says it cannot fomd the document. and the URL becomes:
http://www.horspool.org.uk/sid/www.h...ript/0002.html

What does $pre & $a hold?

Also You realise you have single quotes then are cloing them using double quotes and reopening it with double and closing with single.

I would have though you would have ended up with:
http://www.horspool.org.uk/sid/www.h...id/JavaScript/"0002".html

Are you using the header("Location: $File"); thing now? Otherwise its not going to amke a differeance.

The file its looking for is: http://www.horspool.org.uk/sid/www.h...ript/0002.html
Which woundn't exist. I don't know if its making it a relitive path unless you include the http:// in it as well. I do in mine. Mind I dont pass mine as a parameter in the URL I pass a file ID in the URL and then look it up in a database, that way I can log dowloads and make sure only files I want people to download can be downloaded.

[Acidic]

Originally posted by Electroman
What does $pre & $a hold?

$pre = "000"
$a = 2
... at least in this case. They are not what's wrong.

Originally posted by Electroman

Also You realise you have single quotes then are cloing them using double quotes and reopening it with double and closing with single.

I think they are working correctly. If you're referring to:

'download.php?File=http://www.horspool.org.uk/sid/JavaScript/".$pre.$a.".html'

The single qoute is there so the JavaScript function goto_URL knows it's a string, not a variable. The double qoutes are there to open and close the PHP bit. It's very difficult to explain, but I think it's right.

Originally posted by Electroman

I would have though you would have ended up with:
http://www.horspool.org.uk/sid/www.h...id/JavaScript/"0002".html

that's exactly what I get

Originally posted by Electroman

Are you using the header("Location: $File"); thing now? Otherwise its not going to amke a differeance.

yes: header("Location: $File");

Originally posted by Electroman

The file its looking for is: http://www.horspool.org.uk/sid/www.h...ript/0002.html
Which woundn't exist. I don't know if its making it a relitive path unless you include the http:// in it as well. I do in mine.

When I do include the "http://". It goes to the file, when I don't include it, it makes a relative path, to a file which doesn't exist.

[Electroman]

Try using Content-disposition: attachment when using the http:// so that it was going to the page when you were using Content-disposition: inline. Also I think you'll need to change the Content-Type, but dont worry about that just yet.

[Acidic]

it's now:

<?
$FName = substr($File,strrchr($File,"/"));

header("Expires: Mon, 26 Nov 1962 00:00:00 GMT");
header("Last-Modified: " . gmdate("D,d M Y H:i:s") . " GMT");
header("Cache-Control: no-cache, must-revalidate");
header("Pragma: no-cache");
header("Content-type: application/x-zip");
header("Content-disposition: attachment ; filename=$FName");
header("Location: $File");
?>

and I'm using "http://" in my link. and it goes to the file, doesn't download it.

[Electroman]

Try:

<? 
$FName = substr($File,strrchr($File,"/")); 

header("Expires: Mon, 26 Nov 1962 00:00:00 GMT"); 
header("Last-Modified: " . gmdate("D,d M Y H:i:s") . " GMT"); 
header("Cache-Control: no-cache, must-revalidate"); 
header("Pragma: no-cache"); 
header("Content-type: application/octet-stream"); 
header("Content-disposition: attachment ; filename=$FName"); 
header("Location: $File"); 
?>

I'm not sure if it will work but I'm nearly sure the problem will lie with the Mine Type.

[Electroman]

You might want to check this out too: http://ppewww.ph.gla.ac.uk/~flavell/...tent-type.html
At the very bottom it mentions that if you set the Content-type to text/plain when its the results of a cgi script then IE will decide to download the file instead of display it. that could help you when the user is using IE. Dont know if php would count as cgi in this case though, might just be perl or such.

[Acidic]

It still goes there.

[Acidic]

whoaa. I changed it back to readfile($File) and it worked in both browsers. You're right, MIME type did it.

IE names the file download and FX gives it the name of the URL. But that's something for me to wrry about later. Thanks.

Edit: thanks for the link, I'll have a look.

[Electroman]

How about other browsers, do they act as expected?

[Electroman]

Originally posted by Acidic
whoaa. I changed it back to readfile($File) and it worked in both browsers. You're right, MIME type did it.

IE names the file download and FX gives it the name of the URL. But that's something for me to wrry about later. Thanks.

Edit: thanks for the link, I'll have a look.

Cool , which Mime type worked in the end?

[Acidic]

application/octet-stream

[Electroman]

Posted by Acidic
application/octet-stream

Did you check the downloaded result looked ok in notepad? Just becuase the NewLine Charaters may be messed up becuase were downloading it in binary, hopefuly not though.

[Acidic]

Originally posted by Electroman
Did you check the downloaded result looked ok in notepad? Just becuase the NewLine Charaters may be messed up becuase were downloading it in binary, hopefuly not though.

No it doesn't do NewLine, it places one of those thingies there (same thing you got when downloading NoteMe's CSS file). But any editor other than Notepad, ie. one that understands ASCII will display it correctly. I'll leave it as it is. at least for now.

[CornedBee]

"Any other editor" is one that understands Unix line endings. Nothing to do with ASCII.

[Acidic]

OK, note taken. Thanks.