16:332:568            Software Engineering of Web Applications

Homework Assignment #4:  Network Programming

This assignment is to be done by each student individually. The submission due date is given here.

Copy the MS Windows code for server, Listing B-5 in Appendix B:  Network Programming (starting on page 279). Modify the code to get it running and implement the client side, as well. The client application must be running on MS Windows and for the server application you can choose between Windows and Linux platforms. Both client and server should be written in the C language, and should be single-threaded, so the server can serve only one client at a time.
Your task is to implement the following simple communication protocol. The client should be allowed to send two different "commands" to the server:

  1. GET   <filename>
  2. BOUNCE   <text-to-bounce>
The server responds as requested, or sends a message back containing the requested file or the bounced text message. The user should be informed about the outcome (a simple text message would suffice). In the case of the GET command, the response may be an error message in case the specified filename does not exist:
      "ERROR:  no such file"

It is a good idea to prepare several test files copntaining plain text and store them on the server side, so the client can request some of those. When the client receives the file from the server, it can just print its contents. (This would be inconvenient for binary files.) The client should be able to recognize ERROR messages and display them to the user.

The client should be continuously running and taking user's input. The client should be sufficiently "smart" to check and filter the user's input and accept only GET and BOUNCE commands with appropriate syntax and arguments. All other input should be discarded and the user warned about it.

Submit the source code for the client and the server by email on the due date, no later than 3:00 PM.. Attach also a brief README.txt document that explains how to install and run your software.

Important note: Your code must be based on the example code given in Appendix B:  Network Programming. Submitting just any code downloaded from the Web and modified to do the job will not get you any credit for the assignment.

Some Notes on Running the Book Examples

If you want to test the code in Appendix B:  Network Programming, here are some clarifications.
You should be able to run the Linux/Unix client and server without any modifications to the code. Just copy the code in Listing B-3 and Listing B-4 and compile to get the executables.

First run the server program. You can run the server on one machine and the client on a different machine, or on the same machine as the server.
To run the server, you must provide one argument: the "port number." The number can be any number up to 65535, but the low numbers are reserved for the system, so I usually put some number larger than 5000, say 5500. You shouldn't be able to run the server unless you specify this number, and the specific number is up to you to choose.

When running the client, you must give the server's host name and the number that you chose when you run the server.
You should know what's the name of the machine on which the server is running, such as "ece.rutgers.edu". This may be difficult if you're connected from home over DHCP because the DHCP will assign you the host name dynamically. To find it out, open a Linux shell and type "hostname".
When running, the client is waiting for your input, so you just type some text message and hit "ENTER".

More Notes on Modifying the Book Examples

For converting a network address in struct in_addr to and from dotted-decimal address of the host, please check: inet_ntoa(), inet_aton().

The following two functions could be used to modify the book examples as follows:

#include 		/* standard I/O, NULL, etc.	   */
#include 		/* string manipulation functions   */

#if !defined(WIN16) && !defined(WIN32)
#  define	FAR
   typedef struct in_addr IN_ADDR;
   typedef struct in_addr FAR* LPIN_ADDR;
   typedef struct hostent FAR* LPHOSTENT;
#  define MAXHOSTNAMELEN	256

/* ------------------------------------------------------------------------- *
 *	Function name:  parse_IPaddr
 *	Description: Checks whether given address string is proper
 *		     (in dotted-decimal notation) and if so,
 *		     returns the address in network byte order.
 *	Arguments: str	- (INPUT) Internet address string
 *		   addr	- (OUTPUT) pointer to the network byte order address
 *	Returns:   0, if the input address is in dotted-decimal notation
 *		  -1, if not
 *	Functions called: gethostbyname(), inet_ntoa(), htonl(), inet_addr()
 * _________________________________________________________________________ */

#if defined(__STDC__) || defined(__cplusplus) || defined(WIN16)||defined(WIN32)
int parse_IPaddr( char FAR* str, unsigned long FAR* addr )
int parse_IPaddr( str, addr )
     char*		str;
     unsigned long*	addr;
	int	b1, b2, b3, b4;

	**  First try to convert the host name as a dotted-decimal number.
	**  Only if that fails do we call gethostbyname().
	if ( sscanf( str, "%d.%d.%d.%d", &b1, &b2, &b3, &b4 ) != 4 )
	  {			/* 4 is the length of an INET address */
	    LPHOSTENT		hp;
	    LPIN_ADDR		ptr;

	    if ( (hp = gethostbyname( str )) == NULL )
		return -1;	/* Bad address error: Not an INET host! */

	    /* Check the address type for an Internet host. */
	    if (hp->h_addrtype != AF_INET)
		return -1;	/* Bad address error: Not an INET host! */

	    /* extract dotted-decimal address as the 1st from the list */
	    ptr = (LPIN_ADDR) hp->h_addr;
	    str = inet_ntoa( *ptr );
	    if ( sscanf( str, "%d.%d.%d.%d", &b1, &b2, &b3, &b4 ) != 4 )
		return -1;	/* Bad address error. */
	*addr = inet_addr( str );
	return 0;

/* ------------------------------------------------------------------------- *
 *	Function Name:  get_hostIPaddr
 *	Description:  Finds and returns the IP address of the machine
 *			we are running on.
 *	Arguments:  ipaddrstr - (INPUT) pointer to the host address in
 *				dotted-decimal notation
 *	Returns:  On success,	host IP address in network byte order
 *		  On error,	0
 *	Functions called:  gethostname(), gethostbyname(), inet_ntoa(),
 *			   inet_addr()
 *	Notes:  Error is returned as 0, since an Internet address of 0
 *		is not possible for any host ( 0 refers to `this' host
 *		in the INET context ).
 * _________________________________________________________________________ */

#if defined(__STDC__) || defined(__cplusplus) || defined(WIN16)||defined(WIN32)
unsigned long get_hostIPaddr( char FAR* ipaddrstr )
unsigned long get_hostIPaddr( ipaddrstr )
     char*	ipaddrstr;
	char		hostname[MAXHOSTNAMELEN];
	LPHOSTENT	host_ptr;

	/* get host name */

	if ( gethostname( hostname, sizeof hostname ) < 0 )
	    return 0;	/* Error: That hostname is not found/bad */

	/* lookup host's address by name */

	if ( ( host_ptr = gethostbyname( hostname ) ) == NULL )
	    return 0;	/* Error: Cannot get the host! */

	/* Check the address type for an Internet host. */
	if ( host_ptr->h_addrtype != AF_INET )
	    return 0;	/* Error: Not an INET host! */

	/* extract dotted-decimal address as the 1st from the list */
	  IN_ADDR	inaddr;

	  inaddr = *( (LPIN_ADDR)( host_ptr->h_addr_list[ 0 ]) );
	  strcpy( ipaddrstr, inet_ntoa( inaddr ));

	return  inet_addr( ipaddrstr  );  

Ivan Marsic
Sun Feb 17 15:19:42 EST 2008