Windows Sockets: A Quick And Dirty Primerby Jim FrostLast modified December 31, 1999 |
establish()
function
call_socket()
function
Berkeley, which was already well-known for their work on UNIX, added a new interface to the operating system to support network communication. This interface is generally known as the Berkeley Sockets Interface and is the basis for almost all existing TCP/IP network protocol interfaces, including Windows Sockets (commonly referred to as WinSock).
A socket is very much like a telephone - it's the endpoint of a two-way communications channel. By connecting two sockets together you can pass data between processes, even processes running on different computers, just as you can talk over the telephone once you've made a phone call connecting your telephone with someone else's.
The telephone analogy is a very good one, and will be used
repeatedly to describe socket behavior, but unlike the telephone there
is a distinction in the terminology used for a programs which accept
incoming connections and those which make connections. A
server is a program which waits for incoming connections and
presumably provides some service to other programs. In contrast a
client is a program which connects to a server, usually to
ask it to do something. It is important to remember that it is not
the computer that distinguishes what's a client or a server,
but the way that the program uses the socket. Most MIS
managers believe that "server" means "mainframe" and "client" means
"PC." This is not necessarily the case, and has resulted in a lot of
confusion as desktop computers often work in both client and server
roles simultaneously.
Waiting For The Phone Guy (or: Before you begin)
At the start of every program that uses sockets you must call the
WinSock function WSAStartup()
:
The first argument is the version number of the WinSock library that you expect to be using; version 1.1 is the most popular, although 2.0 is increasingly becoming available. Since newer libraries must support applications which request WinSock 1.1, and very few programmers require the additional features in WinSock 2.0, specifying 1.1 will allow you to work with most of the WinSock libraries on the market.
If the initialization function (or most of the other WinSock
functions) fails you can get additional error information by calling
WSAGetLastError(),
which returns an error code indicating
the cause of the failure.
Similarly you must call the WSACleanup()
function
before your program exits to properly shut down the library. In Win32
applications this good form although not strictly necessary; it is
critical in Win16 or Win32s applications.
The Internet Phone Book (or: Address resolution)
As with the telephone, every socket has a unique address made up of a
couple of different parts.
The first part is the IP address, a four-digit number usually written as four decimal numbers separated by periods (e.g. 192.9.200.10), which specifies which computer you want to talk to. Every computer on the Internet has at least one IP address.
The second is the port number, allowing more than one conversation to take place per computer - much like extension numbers in an office. An application may either pick a port number to use (some are, of course, reserved) or request a random port when assigning an address to a socket.
Unfortunately numerics are difficult to remember, particularly when
you have to deal with a lot of them on a regular basis. As with the
telephone a lookup service exists so that you can remember a simple
name (e.g. world.std.com) rather than a set of digits
(192.74.137.5). The most commonly used interface to this service is
the gethostbyname()
function, which takes the name of a
computer and returns its IP address information. Similarly it's
possible to find the name of a computer if you have its number using the
gethostbyaddr()
function.
Back in the early days of the ARPAnet when there were only a few hundred computers on the entire net everyone kept around a list of all the computers, and these functions would simply search the file for the requested name. As the net grew into the tens of thousands of computers this rapidly became unworkable; changes happened too fast to keep a master file up-to-date, and so many people needed the file that the computer that stored it was always overloaded.
The solution to this problem was the Domain Name Service, or DNS. As with a postal address a DNS host name is broken up into several parts which have increasingly finer resolution, starting at the right with the top-level domain (e.g. .com, .edu) and working leftward to the organizational domain (e.g. std, harvard), then on into organizational sub-domains if they exist, and finally to the computer name.
The idea here is that nobody can remember all of the addresses on the Internet, but everyone can remember their own local addresses and a few top-level addresses. When you don't know the address of a computer you're interested in, you can ask the top-level domain and it'll forward the request on to the organizational domain and so on down the chain until someone knows the answer. This process is very similar to using the telephone company's 555-1212 service; dial 555-1212 in the area code you're interested in and provide the name and address information of the person you're looking for and they give you their telephone number.
The result is a huge distributed database which has proven able to support millions of different computers without long delays in looking up the name.
As well as needing to know the address of the computer you want to talk
to, you must also know the address of your own computer. Unfortunately
there is no way to say "give me my address," in part because it's
possible for a single computer to have more than one address. Instead
you can use the gethostname()
function to ask "what's my
name," and then use gethostbyname()
to look up your own
address. This process will be illustrated shortly.
Installing Your New Phone (or: How to listen for socket
connections)
In order to receive telephone calls, you must first have a telephone
installed. Likewise you must create a socket to listen for
connections, a process which involves several steps.
First, you must create a socket - which is much like getting a
telephone line installed from the phone company. The
socket()
function is used to do this.
Since sockets can have several types, you must specify what type of
socket you want when you create one. One option that you have is the
address family of a socket. Just as the mail service uses a
different scheme to deliver mail than the telephone company uses to
complete calls, so can sockets differ. The most common address
family (and the only one supported by WinSock 1.1) is the internet
format, which is specified by the name AF_INET
.
Another option which you must supply when creating a socket is the
type of socket. The two most common types are
SOCK_STREAM
and
SOCK_DGRAM
. SOCK_STREAM
indicates that data
will come across the socket as a stream of characters, while
SOCK_DGRAM
indicates that data will come in bunches
(called datagrams). We will be dealing with SOCK_STREAM
sockets, which are the most common and easiest to use.
After creating a socket, we must give the socket an address to
listen to, just as you get a telephone number so that you can receive
calls. The bind()
function is used to do this (it binds
a socket to an address, hence the name). An internet socket address
is specified using the sockaddr_in
structure, which
contains fields that specify the address family, address, and port
number for a socket. A pointer to this structure is passed to
functions like bind()
which need an address. Because
sockets are intended to support more than one address family it's
necessary to cast the sockaddr_in
pointer to a
sockaddr
structure pointer to avoid compiler warnings.
SOCK_STREAM
type sockets have the ability to queue
incoming connection requests, which is a lot like having "call
waiting" for your telephone. If you are busy handling a connection,
the connection request will wait until you can deal with it. The
listen()
function is used to set the maximum number of
requests (up to a maximum of five, usually) that will be queued before
requests start being denied.
Figure 1 shows how to use the
socket()
, gethostname()
,
gethostbyname()
, bind()
, and
listen()
functions to establish a socket which can accept
incoming connections.
After you create a socket to get calls, you must wait for calls to
that socket. The accept()
function is used to do this.
Calling accept()
is analogous to picking up the telephone
if it's ringing. Accept()
returns a new socket which is
connected to the caller.
A program which listens for incoming socket connections usually
calls accept()
in a loop and handles each incoming
connection. A skeleton framework for a server program is shown in
figure 2.
Dialing (or: How to call a socket)
You now know how to create a socket that will accept incoming calls.
So how do you call it? As with the telephone, you must first have the
phone before using it to call. You use the socket()
function to do this, exactly as you establish a socket to receive
connections.
After creating a socket to make the call with, and giving it an
address, you use the connect() function to try to connect to a
listening socket. Figure 3 illustrates a
function that creates the socket, sets it up, and calls a particular
port number on a particular computer, returning a connected socket through
which data can flow.
Conversation (or: How to talk between sockets)
Now that you have a connection between sockets you want to send data
between them. The send() and recv()
functions are used to do this.
Unlike when you're reading and writing a file, the network can only send or receive a limited amount of data at a time. As a result you can ask for many characters but you'll often get back fewer than you asked for. One way to handle this is to loop until you have received the number of characters that you want. A simple function to read a given number of characters into a buffer is: |
A very similar function should be used to send data; we leave that
function as an exercise to the reader.
Hanging Up (or: What to do when you're done with a socket)
Just as you hang up when you're through speaking to someone over the
telephone, so must you close a connection between sockets. The
closesocket()
function is used to close each end of a
socket connection. If one end of a socket is closed and you try to
send()
to the other end, send()
will return
an error. A recv()
which is waiting when the other end
of a socket connection is closed will return zero bytes.
Speaking The Language (or: Byte order is important)
Now that you can talk between computers, you have to be careful what
you say. Some computers use differing dialects, such as ASCII versus
(yech) EBCDIC, although this has become increasingly unusual. More
commonly there are byte-order problems; unless you always pass text,
you'll run up against the byte-order problem eventually. Luckily
people have already figured out what to do about it.
Once upon a time in the dark ages someone decided which byte order was
"right." Now there exist functions that convert one to the other if
necessary. Some of these functions are:
htons()
(host to network short integer)
ntohs()
(network to host short integer)
htonl()
(host to network long integer)
ntohl()
(network to host long integer)
For these functions, a "short integer" is a 16-bit entity, and a
"long integer" is a 32-bit entity. Before sending an integer through
a socket, you should first massage it with the htonl()
function:
As you probably expect, it's possible to get a message delivered to
your program whenever there's data waiting on a socket. The function
WSAAsyncSelect()
is used to make Windows send a message
to a window whenever a socket changes its status. Its use is
illustrated by the application framework seen in figure 4, which sets up the socket to send a
message to its main window whenever data is available to be read.
Additional flags to WSAAsyncSelect()
cause Windows to
notify you of many other conditions other than "data has arrived,"
including the status of the connect()
call.
In addition, there are several asynchronous versions of socket
functions which normally must wait for something to happen, including
WSAAsyncGetHostByName()
, a function which usually goes
out across the network to find information about the indicated host.
You should consult the online documentation for more information about
these functions.
The Future Is In Your Hands (or: What to do now)
Using just what's been discussed here, you should be able to build
your own programs that communicate with sockets. As with all new
things, however, it would be a good idea to look at what's already
been done. Luckily there exists a large body of WinSock examples and
documentation; simply search for the word socket on the MSDN
documentation disk and you'll find more than you'll ever read,
including information on creating and using datagram sockets and the
multiprotocol enhancements found in WinSock 2.0.
Beware that the examples given here leave out a lot of error checking which should be used in a real application. You should check the documentation for each of the functions discussed here for further information.
establish()
function.
call_socket()
function.