CSCI201-Networking Programming

Introduction 
A web server is a very simple application that receives requests from a client (typically a web browser), and sends back out the contents of a file, if it exists. A web server differs from an application server in that it only sends the content of a file back but does not do any server-side processing. This just requires accepting a connection from a client, parsing the messages that is passed to it, and responding with an appropriate file.

Description 
We have been using Tomcat throughout the semester. Essentially, Tomcat handles all incoming HTTP requests, and finds and returns the requested files if they exist. However, in this lab, you will be writing a native server, which does Tomcat’s job (exciting!).

Assuming you run your web server on port 6789, to request the page test.html, you will load the page http://localhost:6789/test.html in a browser. Your web server will then try to find the file test.html. If it finds it, it will send the content of that file back to the browser. If that file contains HTML, CSS, and JavaScript, the browser will interpret the code and display it just like a regular web site. If the file cannot be found, send back an HTTP 404 message in the header of the HTTP response. The browser will then display the default 404 message.

You will need to allow as many clients to connect to your web server as possible (requiring multi- threading). Each connection should be handled by a separate thread. Refer back to the server example (ChatRoom.java and ServerThread.java) we did in the multithreaded lecture to give you a better idea of what we are doing in this lab. The code you will be writing today will resemble the two Java classes that we wrote in class.

Please download the skeleton code under the lab assignment, and add it to an Eclipse project. Some of the steps in this lab instruction have already been completed for you in the skeleton project. There is no need to redo those steps or to create duplicate files.

In your backend src folder, add the Server class. In the Server class constructor, you will need to instantiate a new ServerSocket, and you may bind the socket to any unused port that you would like (most likely one above 1023). Since the server will need to be constantly and repeatedly


accepting new connections, do the following steps in an infinite while loop. The ServerSocket will be accepting client connections using the accept() method. Upon receiving a new connection, the server will need to create a new ServerThread (which we will be writing in a moment) to handle that request. Remember to write create a main method in your Server class, where you will create a new instance of the server.

Now, it's time for us to write the ServerThread to handle the request. Modify the ServerThread that extends from Thread. The thread will need to keep track of the client socket that connected to the server, so make sure you pass the socket into the thread's constructor, and create a Socket variable to keep track of it. You will also need to create instance variables for PrintStream and BufferedReader, which we will be using later to determine the content of the request and write response back. In the ServerThread constructor, instantiate the two variables. You can get a new instance of PrintStream from the socket OutputStream by calling

new PrintStream(new BufferedOutputStream(socket.getOutputStream())). Similarly, instantiate the BufferedReader from the socket InputStream by using new BufferedReader(new InputStreamReader(socket.getInputStream())). Make sure you start your thread in the constructor.

In your ServerThread run method, you will need to parse the request and send the appropriate response back to the clients. Clients will be connecting to your web server using a web browser. That means that you need to communicate using HTTP. You can look up online what the header of an HTTP request looks like. Use the BufferedReader to read the request header. Once you parse the incoming request, you will try to find the file that was requested. Before writing the response, create a response header based on the file type that was requested. Print the response header to the PrintStream that we created earlier in the constructor. If the file requested is there, read in the contents of the file through an InputStream, then pass the streams of the file from the InputStream to the PrintStream byte by byte. If the file requested is invalid, respond with a 404 page. Make sure you close the PrintStream before exiting the run method.

You can put a few test HTML files into a directory on your computer where your web server will check. Once your web server is working, try adding images into the directory, and modify your HTML files to include img tags. The browser will automatically request the image files, and you will just need to change the content type of the HTTP header to represent an image being returned, followed by the contents of the image file.