File download / naming

Anon

Hi,

Thanx for your interest in my problem !

I've got a pretty straight forward PHP4.0 + MySQL + Apache setup which works fine. I've got a large number of files stored in the database which I'ld would like to download via PHP.

I use a .php as a script that sends the Header information etc., but my browser always asks me if I'ld like to download file .php? and after I click 'OK' it asks me the "default" question if I'ld like to save or open my file (Word file or whatever).

Has anybody a clue how to solve this (PHP or even in Apache) ?

Thank you,

Theo

Anon

OK, this is actually a header issue. What you need to do is before outputting the actual file name, add the following line:

Header("Content-Disposition: attachment; filename='myfile.doc'");

This will tell the browser to handle it as an attachment (download it) and what the filename should be.

Hope that helps!

Chris King

Anon

Hi Chris,

Thank you for your answer, but I already use this header entry. I'm thinking it is in the URL itself, but I don't know how to handle this. Anyone any ideas ?

Thanx,

Theo

Chris King wrote:

OK, this is actually a header issue. What you need to do is before outputting the actual file name, add the following line:

Header("Content-Disposition: attachment; filename='myfile.doc'");

This will tell the browser to handle it as an attachment (download it) and what the filename should be.

Hope that helps!

Chris King

Anon

The only problem is this doesn't seem to work. I have tried this with IE 5.5 and the browser still uses the file name of the script. Any suggestions?

Anon

This is just kicking my butt too. It's so incredibly frustrating.

I'm trying to make a semi-secure download script; The user clicks their file, and it gets fed to them via HTTP. There is no direct URL to the file it's self, they have to be "logged in".

All of the browsers seem to do it differently. Some support Content-Disposition, some don't (it's not a standard, so you get what you pay for I guess). I can get Netscape 4.x and IE 5.x to work, but Netscape 6 makes a mess of it.. if you go by the direct URL method, this works in NS6, but if you URL-encode the filename if it contains spaces, the user winds up getting a filename with the encoding in it (Some%20File%20Name.txt) whereas NS 4.7x will decode the name properly first.

Argh! There has to be some happy medium to make it do what you want all the time, short of having to look at the headers from the browser, figure out what kind it is, and then customize the download process for them. I'm not a huge web guy and not up on the RFCs, but is some sort of standard being worked on for the next version of HTTP??

Personal Opinion: Netscape 6 is trash, but unfortunately there are people out there who are going to download it and use it not knowing any better, so I can't just pretend it doesn't exist.