Export to File

barb

I have a database driven web site and I want to allow epxorting data from the database to a file. I want tb be able to browse to the directory where the file should be saved on my PC and also choose the name of the file if possible.

Could some kind person tell me I could achieve this if possible.

jazz_snob

So many questions!
1. how do you want to see/use the data? just the raw values? or pull something directly to excel, for example
2. do you want to code php or avoid doing that?
3. if you want to avoid coding, you can use phpmyadmin if its installed on the server to extract the data in a variety of formats
4. if the database allows connections from remote resources (not typical in shared hosting environments) you could use a desktop application like sqlyog or mysql query browser to access and export the data.
5. if you want to code php to write the data, that is very simple: have php run the necessary sql statements to extract the data and write the results row by row to a file on the server...then do a header() redirect to the file on the server (for example, there are other ways to download a file).

barb

Thanks for responding. I don't need any help creating the export.

Just was asking about the ability to browse to where the file is saved and name the file.

Weedpacket

That's done by sending suitable HTTP headers in front of the file, in PHP by using the [man]header[/man] function; see example 1 on that page.

To just allow the user to download the file without the client trying to be clever with it, the appropriate Content-Type is "application/octet-stream".

barb

Thanks weedpacket, that worked great, I now get a save dialog and my export file saves to my pc.

Now I have a new problem. I had to move my php code above the head of my web page because I was getting an error saying the headers had already been sent. That's fine, but I'm getting all the HTML for my page added at the bottom of my exported file.

Any ideas.

Weedpacket

For a given response you have a choice of one (the file) or the other (the page), but not both.

There are generally two ways around this; one is to have a <meta> refresh tag, and the URL in that is for the file. The other (this is what Sourceforge does) is have a vanishingly small <iframe> in the page that uses the URL to the file as its src. Either way, that extra URL is the one that has the Content-Type:application/octet-stream header and the actual contents of the file.

barb

Thanks again for the reply.

Where do I have to put the meta tag? I would assume it would have to go just before the header statements?

Weedpacket

It would go where the other <meta> tags go: between <head> and </head> in the HTML page. (The <iframe>, if you go with that alternative, on the other hand, would go in the <body>.)

barb

Thanks again.

What if I don't actually want to redirect to the file?

Weedpacket

I haven't experimented with the <meta> tag lately, but if I recall correctly the browser wouldn't stop displaying the page it's already displaying if it's asked to follow a link to a generic binary file: instead it would display the download dialogue. That's my recollection of browser behaviour in this case.
With an <iframe> there isn't any redirection going on.

One question this raises:

barb wrote:
That's fine, but I'm getting all the HTML for my page added at the bottom of my exported file.

If you don't want to display another page when you download a file, what's all that HTML for?