Tiny HTTP Server

This project implements a small, iterative HTTP server in C, designed primarily as an educational exercise to deepen understanding of network programming concepts. It's built upon the principles discussed in chapters 11: Network Programming and 12: Concurrent Programming of the book Computer Systems: A Programmer's Perspective (CS:APP) .

It's a foundational server capable of handling basic HTTP requests and serving both static and dynamic content.

Features

• HTTP/1.1 Support: Handles GET, HEAD, and POST HTTP methods.
• Static Content Serving: Serves HTML, images (JPG, PNG, GIF, MP4), and plain text files directly from a specified document root (wwwroot).
• Dynamic Content (CGI): Executes CGI (Common Gateway Interface) programs for dynamic content generation. Includes example CGI scripts (hello.c and adder.c) to demonstrate functionality like query string parsing and POST data handling.
• Error Handling: Provides basic error responses (e.g., 400 Bad Request, 403 Forbidden, 404 Not Found, 500 Internal Server Error) and handles common network issues.

Experimental

While master branch is iterative (handling one client request at a time), this project also explored various concurrency models in dedicated branches:

• concurrency-with-fork: Process-based concurrency using fork().
• concurrency-with-threads: Thread-based concurrency using pthreads.
• concurrency-with-select: Concurrency using I/O multiplexing with select().
• pre-threading: Pre-threaded (thread pool) model with a simple auto-scaling heuristic.

Project Structure

• tiny.c: The main server implementation. It accepts client connections, parses HTTP requests, and dispatches to appropriate handlers for static or dynamic content.
• mynetlib.c / mynetlib.h: A custom utility library providing wrappers around socket functions (clientConnect, serverListen, sendBytes, receiveBytes) and a buffered reader (buffered_reader_t, bufReadLine, bufReadBytes) for efficient network I/O.
• cgi/hello.c and cgi/adder.c: Simple example CGI programs.
• Makefile: Used to compile the server and the CGI programs.
• wwwroot/: (Created by Makefile) The web document root directory for static files.
• wwwroot/cgi-bin/: (Created by Makefile) The directory where compiled CGI executables are placed.

---

Requirements

• Developed and tested on Linux
• make utility
• gcc C compiler

---

Building and Running

$ cd src
$ make clean
$ make
$ ./tiny 8080   # or another port number

Usage Examples

Once the server is running, you can access it using a web browser or curl.

curl http://localhost:8000/
curl http://localhost:8000/cgi-bin/hello
curl "http://localhost:8000/cgi-bin/adder?a=100&b=200"
curl -X POST -H "Content-Type: application/x-www-form-urlencoded" -d "a=50&b=75" http://localhost:8000/cgi-bin/adder

Limitations

This project is an educational effort and, as such, has several intentional simplifications and limitations!!!

• Poor security
• Lacks robustness
• Inefficient large file serving
• Poor CGI input parsing
• No Persistent Connections
• No HTTPS/SSL/TLS
• No Advanced Features
• Etc.

Other projects

For other HTTP server implementations with a different set of features, please check:

• Build Your Own HTTP Server in C#
• Build Your Own HTTP Server in Javascript
• Build Your Own HTTP Server in Go