I am beginning to develop an application, destined to run especially on web servers. One of the aspects would involve allowing access to data stored on server(s) by all means, like FTP, SMB, NFS etc., which is my main black hole in the project.
Can anybody suggest some starting points on this (documentation, approach etc...). I know it's a big programming chapter, in which I don't have much experience, but any help would be appreciated.
The project will be (almost exclusively) C/C++ based
Thank you!

Nosferatu
We are what we eat...
There's no such thing as free meal...
once stated: methane@personal.ro

Hey Nos, I'm curious about this project:
is this a web application you're writing, or a service to be run on the web server?

If you're wanting something where the user interfaces with various things like FTP, etc., it might be better to use the existing technologies, and just use web development technologies on the client side to determine how they access the services on the web server.

let me know a bit more about it.

Jack

Thanks for your interest, Jack.
The application is well... some kind of a document management app, almost entirely web based.

However, the user interface does not go into my primary interest target and probably such an interface won't be provided at all in the final product.
What I want to do is a module (part of the bigger project) which will be able to listen to all kinds of requests that can be made over a network and of course, provide the needed answers.
This module is going to "sit" on a server, probably listening to ports andI just don't know how to do that.

The main idea of the project is that it should be as independent as possible so no MS/UNIX/MAC solutions are acceptable. At least, I'm trying to minimize the OS impact, I know there are points where you cannot obscure the platform you're on.

Moreover, I get the hunch that the code written would/should be placed at a low level, as far as communications are concerned.

I am sure that I'm a bit blury with this but... that's how I am and I want to clear the f(r)og on this chapter.

 

Nosferatu
We are what we eat...
There's no such thing as free meal...
once stated: methane@personal.ro

Hmmm...interesting.

So just so I'm clear:
Your app doesn't care how the requests come in, it just knows that it may get ftp, http, smb, etc. requests, and if it does that all the files are in one place and its only the means to those files that need to be figured out?

jack

Yes. The actual location of the files is resolved at a lower level thorugh a mapping.
The client application will be informed by the documents' storage structure in a "protocol friendly" way. What I mean is that there is a difference betweed the actual location of the files and the things an FTP (or other) client sees.

What I wonder is if it is possible to hook the requests in the application and then gather the data independently from that protocol; just use the protocol to send the data away from the application to the requesting client.

Hoping the answer to my hope is yes, I am interested in finding out as more as I can about dealing with these data trasfer protocols and find the simplest way to do it. The ideal case is, of course: listen to the port XXX for XXXType connection and use that port to roll data.
I'm sure that's more than that and that's what I want to know about.

Nosferatu
We are what we eat...
There's no such thing as free meal...
once stated: methane@personal.ro