The server should be called as follows:

# ./ port [root]

Where port is a mandatory argument to specify the port number the server should listen on, and root
is an optional argument to specify the web server's root directory.


● If the root directory is not specified, use the current directory as root
● Use the netcat (sometimes called nc) command to listen on incoming connections
● The web server should at least understand the HTTP GET command
● The web server should correctly return the following status codes: 200 OK, 403
● The web server should correctly return a Content-Type header, at least for text/plain,
text/html and application/octet-stream (for all files which do not match any of the
other content types)
● If the web server receives a request for a directory, it should return a directory listing (does
not need to have clickable links). Optionally, you can return an index.html file instead, if
one exists.

any guide is appreciated

Hi again,
I don't need the advanced script, just I need the sample one, to get and have the idea ...

Hey There,

Bash doesn't natively support sockets programming, like in Perl. Are you looking to write a bash script that runs a separate "http server" program?

Also, if you've already gotten started, can you post that. This way anyone who checks this post will know where you're at and can help you more :)

Best wishes,


commented: Seems like a much better idea. +17

Ah, yes,

That's running a shell on a port, technically, but it is a clever solution. I don't think it's possible to write a standalone bash http server that doesn't depend on inetd running the bash shell.

I may be wrong. If anyone has any more examples like this one, I definitely would be interested. Very cool.

I wrote this perl shell runner that could be combined with that to emulate httpd, but it would easier to just write the web server in Perl straight up.

Cool link :)