3 perlipc - Perl interprocess communication (signals, fifos, pipes, safe subprocesses, sockets, and semaphores)
7 The basic IPC facilities of Perl are built out of the good old Unix
8 signals, named pipes, pipe opens, the Berkeley socket routines, and SysV
9 IPC calls. Each is used in slightly different situations.
13 Perl uses a simple signal handling model: the %SIG hash contains names or
14 references of user-installed signal handlers. These handlers will be called
15 with an argument which is the name of the signal that triggered it. A
16 signal may be generated intentionally from a particular keyboard sequence like
17 control-C or control-Z, sent to you from another process, or
18 triggered automatically by the kernel when special events transpire, like
19 a child process exiting, your process running out of stack space, or
20 hitting file size limit.
22 For example, to trap an interrupt signal, set up a handler like this.
23 Do as little as you possibly can in your handler; notice how all we do is
24 set a global variable and then raise an exception. That's because on most
25 systems, libraries are not re-entrant; particularly, memory allocation and
26 I/O routines are not. That means that doing nearly I<anything> in your
27 handler could in theory trigger a memory fault and subsequent core dump.
32 die "Somebody sent me a SIG$signame";
34 $SIG{INT} = 'catch_zap'; # could fail in modules
35 $SIG{INT} = \&catch_zap; # best strategy
37 The names of the signals are the ones listed out by C<kill -l> on your
38 system, or you can retrieve them from the Config module. Set up an
39 @signame list indexed by number to get the name and a %signo table
40 indexed by name to get the number:
43 defined $Config{sig_name} || die "No sigs?";
44 foreach $name (split(' ', $Config{sig_name})) {
50 So to check whether signal 17 and SIGALRM were the same, do just this:
52 print "signal #17 = $signame[17]\n";
54 print "SIGALRM is $signo{ALRM}\n";
57 You may also choose to assign the strings C<'IGNORE'> or C<'DEFAULT'> as
58 the handler, in which case Perl will try to discard the signal or do the
59 default thing. Some signals can be neither trapped nor ignored, such as
60 the KILL and STOP (but not the TSTP) signals. One strategy for
61 temporarily ignoring signals is to use a local() statement, which will be
62 automatically restored once your block is exited. (Remember that local()
63 values are "inherited" by functions called from within that block.)
66 local $SIG{INT} = 'IGNORE';
70 # interrupts still ignored, for now...
73 Sending a signal to a negative process ID means that you send the signal
74 to the entire Unix process-group. This code sends a hang-up signal to all
75 processes in the current process group (and sets $SIG{HUP} to IGNORE so
76 it doesn't kill itself):
79 local $SIG{HUP} = 'IGNORE';
81 # snazzy writing of: kill('HUP', -$$)
84 Another interesting signal to send is signal number zero. This doesn't
85 actually affect another process, but instead checks whether it's alive
86 or has changed its UID.
88 unless (kill 0 => $kid_pid) {
89 warn "something wicked happened to $kid_pid";
92 You might also want to employ anonymous functions for simple signal
95 $SIG{INT} = sub { die "\nOutta here!\n" };
97 But that will be problematic for the more complicated handlers that need
98 to reinstall themselves. Because Perl's signal mechanism is currently
99 based on the signal(3) function from the C library, you may sometimes be so
100 misfortunate as to run on systems where that function is "broken", that
101 is, it behaves in the old unreliable SysV way rather than the newer, more
102 reasonable BSD and POSIX fashion. So you'll see defensive people writing
103 signal handlers like this:
107 # loathe sysV: it makes us not only reinstate
108 # the handler, but place it after the wait
109 $SIG{CHLD} = \&REAPER;
111 $SIG{CHLD} = \&REAPER;
112 # now do something that forks...
114 or even the more elaborate:
116 use POSIX ":sys_wait_h";
119 while ($child = waitpid(-1,WNOHANG)) {
120 $Kid_Status{$child} = $?;
122 $SIG{CHLD} = \&REAPER; # still loathe sysV
124 $SIG{CHLD} = \&REAPER;
125 # do something that forks...
127 Signal handling is also used for timeouts in Unix, While safely
128 protected within an C<eval{}> block, you set a signal handler to trap
129 alarm signals and then schedule to have one delivered to you in some
130 number of seconds. Then try your blocking operation, clearing the alarm
131 when it's done but not before you've exited your C<eval{}> block. If it
132 goes off, you'll use die() to jump out of the block, much as you might
133 using longjmp() or throw() in other languages.
138 local $SIG{ALRM} = sub { die "alarm clock restart" };
140 flock(FH, 2); # blocking write lock
143 if ($@ and $@ !~ /alarm clock restart/) { die }
145 For more complex signal handling, you might see the standard POSIX
146 module. Lamentably, this is almost entirely undocumented, but
147 the F<t/lib/posix.t> file from the Perl source distribution has some
152 A named pipe (often referred to as a FIFO) is an old Unix IPC
153 mechanism for processes communicating on the same machine. It works
154 just like a regular, connected anonymous pipes, except that the
155 processes rendezvous using a filename and don't have to be related.
157 To create a named pipe, use the Unix command mknod(1) or on some
158 systems, mkfifo(1). These may not be in your normal path.
160 # system return val is backwards, so && not ||
162 $ENV{PATH} .= ":/etc:/usr/etc";
163 if ( system('mknod', $path, 'p')
164 && system('mkfifo', $path) )
166 die "mk{nod,fifo} $path failed;
170 A fifo is convenient when you want to connect a process to an unrelated
171 one. When you open a fifo, the program will block until there's something
174 For example, let's say you'd like to have your F<.signature> file be a
175 named pipe that has a Perl program on the other end. Now every time any
176 program (like a mailer, news reader, finger program, etc.) tries to read
177 from that file, the reading program will block and your program will
178 supply the new signature. We'll use the pipe-checking file test B<-p>
179 to find out whether anyone (or anything) has accidentally removed our fifo.
182 $FIFO = '.signature';
183 $ENV{PATH} .= ":/etc:/usr/games";
188 system('mknod', $FIFO, 'p')
189 && die "can't mknod $FIFO: $!";
192 # next line blocks until there's a reader
193 open (FIFO, "> $FIFO") || die "can't write $FIFO: $!";
194 print FIFO "John Smith (smith\@host.org)\n", `fortune -s`;
196 sleep 2; # to avoid dup signals
200 =head1 Using open() for IPC
202 Perl's basic open() statement can also be used for unidirectional interprocess
203 communication by either appending or prepending a pipe symbol to the second
204 argument to open(). Here's how to start something up in a child process you
207 open(SPOOLER, "| cat -v | lpr -h 2>/dev/null")
208 || die "can't fork: $!";
209 local $SIG{PIPE} = sub { die "spooler pipe broke" };
210 print SPOOLER "stuff\n";
211 close SPOOLER || die "bad spool: $! $?";
213 And here's how to start up a child process you intend to read from:
215 open(STATUS, "netstat -an 2>&1 |")
216 || die "can't fork: $!";
218 next if /^(tcp|udp)/;
221 close STATUS || die "bad netstat: $! $?";
223 If one can be sure that a particular program is a Perl script that is
224 expecting filenames in @ARGV, the clever programmer can write something
227 $ program f1 "cmd1|" - f2 "cmd2|" f3 < tmpfile
229 and irrespective of which shell it's called from, the Perl program will
230 read from the file F<f1>, the process F<cmd1>, standard input (F<tmpfile>
231 in this case), the F<f2> file, the F<cmd2> command, and finally the F<f3>
232 file. Pretty nifty, eh?
234 You might notice that you could use backticks for much the
235 same effect as opening a pipe for reading:
237 print grep { !/^(tcp|udp)/ } `netstat -an 2>&1`;
238 die "bad netstat" if $?;
240 While this is true on the surface, it's much more efficient to process the
241 file one line or record at a time because then you don't have to read the
242 whole thing into memory at once. It also gives you finer control of the
243 whole process, letting you to kill off the child process early if you'd
246 Be careful to check both the open() and the close() return values. If
247 you're I<writing> to a pipe, you should also trap SIGPIPE. Otherwise,
248 think of what happens when you start up a pipe to a command that doesn't
249 exist: the open() will in all likelihood succeed (it only reflects the
250 fork()'s success), but then your output will fail--spectacularly. Perl
251 can't know whether the command worked because your command is actually
252 running in a separate process whose exec() might have failed. Therefore,
253 while readers of bogus commands return just a quick end of file, writers
254 to bogus command will trigger a signal they'd better be prepared to
263 Both the main process and the child process share the same STDIN,
264 STDOUT and STDERR filehandles. If both processes try to access them
265 at once, strange things can happen. You may want to close or reopen
266 the filehandles for the child. You can get around this by opening
267 your pipe with open(), but on some systems this means that the child
268 process cannot outlive the parent.
270 =head2 Background Processes
272 You can run a command in the background with:
276 The command's STDOUT and STDERR (and possibly STDIN, depending on your
277 shell) will be the same as the parent's. You won't need to catch
278 SIGCHLD because of the double-fork taking place (see below for more
281 =head2 Complete Dissociation of Child from Parent
283 In some cases (starting server processes, for instance) you'll want to
284 complete dissociate the child process from the parent. The following
285 process is reported to work on most Unixish systems. Non-Unix users
286 should check their Your_OS::Process module for other solutions.
292 Open /dev/tty and use the TIOCNOTTY ioctl on it. See L<tty(4)>
297 Change directory to /
301 Reopen STDIN, STDOUT, and STDERR so they're not connected to the old
306 Background yourself like this:
312 =head2 Safe Pipe Opens
314 Another interesting approach to IPC is making your single program go
315 multiprocess and communicate between (or even amongst) yourselves. The
316 open() function will accept a file argument of either C<"-|"> or C<"|-">
317 to do a very interesting thing: it forks a child connected to the
318 filehandle you've opened. The child is running the same program as the
319 parent. This is useful for safely opening a file when running under an
320 assumed UID or GID, for example. If you open a pipe I<to> minus, you can
321 write to the filehandle you opened and your kid will find it in his
322 STDIN. If you open a pipe I<from> minus, you can read from the filehandle
323 you opened whatever your kid writes to his STDOUT.
329 $pid = open(KID_TO_WRITE, "|-");
330 unless (defined $pid) {
331 warn "cannot fork: $!";
332 die "bailing out" if $sleep_count++ > 6;
335 } until defined $pid;
338 print KID_TO_WRITE @some_data;
339 close(KID_TO_WRITE) || warn "kid exited $?";
341 ($EUID, $EGID) = ($UID, $GID); # suid progs only
342 open (FILE, "> /safe/file")
343 || die "can't open /safe/file: $!";
345 print FILE; # child's STDIN is parent's KID
347 exit; # don't forget this
350 Another common use for this construct is when you need to execute
351 something without the shell's interference. With system(), it's
352 straightforward, but you can't use a pipe open or backticks safely.
353 That's because there's no way to stop the shell from getting its hands on
354 your arguments. Instead, use lower-level control to call exec() directly.
356 Here's a safe backtick or pipe open for read:
358 # add error processing as above
359 $pid = open(KID_TO_READ, "-|");
362 while (<KID_TO_READ>) {
363 # do something interesting
365 close(KID_TO_READ) || warn "kid exited $?";
368 ($EUID, $EGID) = ($UID, $GID); # suid only
369 exec($program, @options, @args)
370 || die "can't exec program: $!";
375 And here's a safe pipe open for writing:
377 # add error processing as above
378 $pid = open(KID_TO_WRITE, "|-");
379 $SIG{ALRM} = sub { die "whoops, $program pipe broke" };
385 close(KID_TO_WRITE) || warn "kid exited $?";
388 ($EUID, $EGID) = ($UID, $GID);
389 exec($program, @options, @args)
390 || die "can't exec program: $!";
394 Note that these operations are full Unix forks, which means they may not be
395 correctly implemented on alien systems. Additionally, these are not true
396 multithreading. If you'd like to learn more about threading, see the
397 F<modules> file mentioned below in the SEE ALSO section.
399 =head2 Bidirectional Communication with Another Process
401 While this works reasonably well for unidirectional communication, what
402 about bidirectional communication? The obvious thing you'd like to do
403 doesn't actually work:
405 open(PROG_FOR_READING_AND_WRITING, "| some program |")
407 and if you forget to use the B<-w> flag, then you'll miss out
408 entirely on the diagnostic message:
410 Can't do bidirectional pipe at -e line 1.
412 If you really want to, you can use the standard open2() library function
413 to catch both ends. There's also an open3() for tridirectional I/O so you
414 can also catch your child's STDERR, but doing so would then require an
415 awkward select() loop and wouldn't allow you to use normal Perl input
418 If you look at its source, you'll see that open2() uses low-level
419 primitives like Unix pipe() and exec() to create all the connections.
420 While it might have been slightly more efficient by using socketpair(), it
421 would have then been even less portable than it already is. The open2()
422 and open3() functions are unlikely to work anywhere except on a Unix
423 system or some other one purporting to be POSIX compliant.
425 Here's an example of using open2():
429 $pid = open2( \*Reader, \*Writer, "cat -u -n" );
430 Writer->autoflush(); # default here, actually
431 print Writer "stuff\n";
434 The problem with this is that Unix buffering is really going to
435 ruin your day. Even though your C<Writer> filehandle is auto-flushed,
436 and the process on the other end will get your data in a timely manner,
437 you can't usually do anything to force it to give it back to you
438 in a similarly quick fashion. In this case, we could, because we
439 gave I<cat> a B<-u> flag to make it unbuffered. But very few Unix
440 commands are designed to operate over pipes, so this seldom works
441 unless you yourself wrote the program on the other end of the
444 A solution to this is the nonstandard F<Comm.pl> library. It uses
445 pseudo-ttys to make your program behave more reasonably:
448 $ph = open_proc('cat -n');
450 print $ph "a line\n";
451 print "got back ", scalar <$ph>;
454 This way you don't have to have control over the source code of the
455 program you're using. The F<Comm> library also has expect()
456 and interact() functions. Find the library (and we hope its
457 successor F<IPC::Chat>) at your nearest CPAN archive as detailed
458 in the SEE ALSO section below.
460 =head1 Sockets: Client/Server Communication
462 While not limited to Unix-derived operating systems (e.g., WinSock on PCs
463 provides socket support, as do some VMS libraries), you may not have
464 sockets on your system, in which case this section probably isn't going to do
465 you much good. With sockets, you can do both virtual circuits (i.e., TCP
466 streams) and datagrams (i.e., UDP packets). You may be able to do even more
467 depending on your system.
469 The Perl function calls for dealing with sockets have the same names as
470 the corresponding system calls in C, but their arguments tend to differ
471 for two reasons: first, Perl filehandles work differently than C file
472 descriptors. Second, Perl already knows the length of its strings, so you
473 don't need to pass that information.
475 One of the major problems with old socket code in Perl was that it used
476 hard-coded values for some of the constants, which severely hurt
477 portability. If you ever see code that does anything like explicitly
478 setting C<$AF_INET = 2>, you know you're in for big trouble: An
479 immeasurably superior approach is to use the C<Socket> module, which more
480 reliably grants access to various constants and functions you'll need.
482 If you're not writing a server/client for an existing protocol like
483 NNTP or SMTP, you should give some thought to how your server will
484 know when the client has finished talking, and vice-versa. Most
485 protocols are based on one-line messages and responses (so one party
486 knows the other has finished when a "\n" is received) or multi-line
487 messages and responses that end with a period on an empty line
488 ("\n.\n" terminates a message/response).
490 =head2 Internet TCP Clients and Servers
492 Use Internet-domain sockets when you want to do client-server
493 communication that might extend to machines outside of your own system.
495 Here's a sample TCP client using Internet-domain sockets:
501 my ($remote,$port, $iaddr, $paddr, $proto, $line);
503 $remote = shift || 'localhost';
504 $port = shift || 2345; # random port
505 if ($port =~ /\D/) { $port = getservbyname($port, 'tcp') }
506 die "No port" unless $port;
507 $iaddr = inet_aton($remote) || die "no host: $remote";
508 $paddr = sockaddr_in($port, $iaddr);
510 $proto = getprotobyname('tcp');
511 socket(SOCK, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
512 connect(SOCK, $paddr) || die "connect: $!";
513 while (defined($line = <SOCK>)) {
517 close (SOCK) || die "close: $!";
520 And here's a corresponding server to go along with it. We'll
521 leave the address as INADDR_ANY so that the kernel can choose
522 the appropriate interface on multihomed hosts. If you want sit
523 on a particular interface (like the external side of a gateway
524 or firewall machine), you should fill this in with your real address
530 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
534 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
536 my $port = shift || 2345;
537 my $proto = getprotobyname('tcp');
538 $port = $1 if $port =~ /(\d+)/; # untaint port number
540 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
541 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
542 pack("l", 1)) || die "setsockopt: $!";
543 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
544 listen(Server,SOMAXCONN) || die "listen: $!";
546 logmsg "server started on port $port";
550 $SIG{CHLD} = \&REAPER;
552 for ( ; $paddr = accept(Client,Server); close Client) {
553 my($port,$iaddr) = sockaddr_in($paddr);
554 my $name = gethostbyaddr($iaddr,AF_INET);
556 logmsg "connection from $name [",
557 inet_ntoa($iaddr), "]
560 print Client "Hello there, $name, it's now ",
561 scalar localtime, "\n";
564 And here's a multithreaded version. It's multithreaded in that
565 like most typical servers, it spawns (forks) a slave server to
566 handle the client request so that the master server can quickly
567 go back to service a new client.
572 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
576 sub spawn; # forward declaration
577 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
579 my $port = shift || 2345;
580 my $proto = getprotobyname('tcp');
581 $port = $1 if $port =~ /(\d+)/; # untaint port number
583 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
584 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
585 pack("l", 1)) || die "setsockopt: $!";
586 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
587 listen(Server,SOMAXCONN) || die "listen: $!";
589 logmsg "server started on port $port";
596 $SIG{CHLD} = \&REAPER; # loathe sysV
597 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
600 $SIG{CHLD} = \&REAPER;
602 for ( $waitedpid = 0;
603 ($paddr = accept(Client,Server)) || $waitedpid;
604 $waitedpid = 0, close Client)
606 next if $waitedpid and not $paddr;
607 my($port,$iaddr) = sockaddr_in($paddr);
608 my $name = gethostbyaddr($iaddr,AF_INET);
610 logmsg "connection from $name [",
611 inet_ntoa($iaddr), "]
615 print "Hello there, $name, it's now ", scalar localtime, "\n";
616 exec '/usr/games/fortune'
617 or confess "can't exec fortune: $!";
625 unless (@_ == 0 && $coderef && ref($coderef) eq 'CODE') {
626 confess "usage: spawn CODEREF";
630 if (!defined($pid = fork)) {
631 logmsg "cannot fork: $!";
635 return; # I'm the parent
637 # else I'm the child -- go spawn
639 open(STDIN, "<&Client") || die "can't dup client to stdin";
640 open(STDOUT, ">&Client") || die "can't dup client to stdout";
641 ## open(STDERR, ">&STDOUT") || die "can't dup stdout to stderr";
645 This server takes the trouble to clone off a child version via fork() for
646 each incoming request. That way it can handle many requests at once,
647 which you might not always want. Even if you don't fork(), the listen()
648 will allow that many pending connections. Forking servers have to be
649 particularly careful about cleaning up their dead children (called
650 "zombies" in Unix parlance), because otherwise you'll quickly fill up your
653 We suggest that you use the B<-T> flag to use taint checking (see L<perlsec>)
654 even if we aren't running setuid or setgid. This is always a good idea
655 for servers and other programs run on behalf of someone else (like CGI
656 scripts), because it lessens the chances that people from the outside will
657 be able to compromise your system.
659 Let's look at another TCP client. This one connects to the TCP "time"
660 service on a number of different machines and shows how far their clocks
661 differ from the system on which it's being run:
668 my $SECS_of_70_YEARS = 2208988800;
669 sub ctime { scalar localtime(shift) }
671 my $iaddr = gethostbyname('localhost');
672 my $proto = getprotobyname('tcp');
673 my $port = getservbyname('time', 'tcp');
674 my $paddr = sockaddr_in(0, $iaddr);
678 printf "%-24s %8s %s\n", "localhost", 0, ctime(time());
680 foreach $host (@ARGV) {
681 printf "%-24s ", $host;
682 my $hisiaddr = inet_aton($host) || die "unknown host";
683 my $hispaddr = sockaddr_in($port, $hisiaddr);
684 socket(SOCKET, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
685 connect(SOCKET, $hispaddr) || die "bind: $!";
687 read(SOCKET, $rtime, 4);
689 my $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
690 printf "%8d %s\n", $histime - time, ctime($histime);
693 =head2 Unix-Domain TCP Clients and Servers
695 That's fine for Internet-domain clients and servers, but what about local
696 communications? While you can use the same setup, sometimes you don't
697 want to. Unix-domain sockets are local to the current host, and are often
698 used internally to implement pipes. Unlike Internet domain sockets, Unix
699 domain sockets can show up in the file system with an ls(1) listing.
702 srw-rw-rw- 1 root 0 Oct 31 07:23 /dev/log
704 You can test for these with Perl's B<-S> file test:
706 unless ( -S '/dev/log' ) {
707 die "something's wicked with the print system";
710 Here's a sample Unix-domain client:
716 my ($rendezvous, $line);
718 $rendezvous = shift || '/tmp/catsock';
719 socket(SOCK, PF_UNIX, SOCK_STREAM, 0) || die "socket: $!";
720 connect(SOCK, sockaddr_un($rendezvous)) || die "connect: $!";
721 while (defined($line = <SOCK>)) {
726 And here's a corresponding server.
734 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
736 my $NAME = '/tmp/catsock';
737 my $uaddr = sockaddr_un($NAME);
738 my $proto = getprotobyname('tcp');
740 socket(Server,PF_UNIX,SOCK_STREAM,0) || die "socket: $!";
742 bind (Server, $uaddr) || die "bind: $!";
743 listen(Server,SOMAXCONN) || die "listen: $!";
745 logmsg "server started on $NAME";
747 $SIG{CHLD} = \&REAPER;
749 for ( $waitedpid = 0;
750 accept(Client,Server) || $waitedpid;
751 $waitedpid = 0, close Client)
754 logmsg "connection on $NAME";
756 print "Hello there, it's now ", scalar localtime, "\n";
757 exec '/usr/games/fortune' or die "can't exec fortune: $!";
761 As you see, it's remarkably similar to the Internet domain TCP server, so
762 much so, in fact, that we've omitted several duplicate functions--spawn(),
763 logmsg(), ctime(), and REAPER()--which are exactly the same as in the
766 So why would you ever want to use a Unix domain socket instead of a
767 simpler named pipe? Because a named pipe doesn't give you sessions. You
768 can't tell one process's data from another's. With socket programming,
769 you get a separate session for each client: that's why accept() takes two
772 For example, let's say that you have a long running database server daemon
773 that you want folks from the World Wide Web to be able to access, but only
774 if they go through a CGI interface. You'd have a small, simple CGI
775 program that does whatever checks and logging you feel like, and then acts
776 as a Unix-domain client and connects to your private server.
778 =head1 TCP Clients with IO::Socket
780 For those preferring a higher-level interface to socket programming, the
781 IO::Socket module provides an object-oriented approach. IO::Socket is
782 included as part of the standard Perl distribution as of the 5.004
783 release. If you're running an earlier version of Perl, just fetch
784 IO::Socket from CPAN, where you'll also find find modules providing easy
785 interfaces to the following systems: DNS, FTP, Ident (RFC 931), NIS and
786 NISPlus, NNTP, Ping, POP3, SMTP, SNMP, SSLeay, Telnet, and Time--just
789 =head2 A Simple Client
791 Here's a client that creates a TCP connection to the "daytime"
792 service at port 13 of the host name "localhost" and prints out everything
793 that the server there cares to provide.
797 $remote = IO::Socket::INET->new(
799 PeerAddr => "localhost",
800 PeerPort => "daytime(13)",
802 or die "cannot connect to daytime port at localhost";
803 while ( <$remote> ) { print }
805 When you run this program, you should get something back that
808 Wed May 14 08:40:46 MDT 1997
810 Here are what those parameters to the C<new> constructor mean:
816 This is which protocol to use. In this case, the socket handle returned
817 will be connected to a TCP socket, because we want a stream-oriented
818 connection, that is, one that acts pretty much like a plain old file.
819 Not all sockets are this of this type. For example, the UDP protocol
820 can be used to make a datagram socket, used for message-passing.
824 This is the name or Internet address of the remote host the server is
825 running on. We could have specified a longer name like C<"www.perl.com">,
826 or an address like C<"204.148.40.9">. For demonstration purposes, we've
827 used the special hostname C<"localhost">, which should always mean the
828 current machine you're running on. The corresponding Internet address
829 for localhost is C<"127.1">, if you'd rather use that.
833 This is the service name or port number we'd like to connect to.
834 We could have gotten away with using just C<"daytime"> on systems with a
835 well-configured system services file,[FOOTNOTE: The system services file
836 is in I</etc/services> under Unix] but just in case, we've specified the
837 port number (13) in parentheses. Using just the number would also have
838 worked, but constant numbers make careful programmers nervous.
842 Notice how the return value from the C<new> constructor is used as
843 a filehandle in the C<while> loop? That's what's called an indirect
844 filehandle, a scalar variable containing a filehandle. You can use
845 it the same way you would a normal filehandle. For example, you
846 can read one line from it this way:
850 all remaining lines from is this way:
854 and send a line of data to it this way:
856 print $handle "some data\n";
858 =head2 A Webget Client
860 Here's a simple client that takes a remote host to fetch a document
861 from, and then a list of documents to get from that host. This is a
862 more interesting client than the previous one because it first sends
863 something to the server before fetching the server's response.
867 unless (@ARGV > 1) { die "usage: $0 host document ..." }
868 $host = shift(@ARGV);
869 foreach $document ( @ARGV ) {
870 $remote = IO::Socket::INET->new( Proto => "tcp",
872 PeerPort => "http(80)",
874 unless ($remote) { die "cannot connect to http daemon on $host" }
875 $remote->autoflush(1);
876 print $remote "GET $document HTTP/1.0\n\n";
877 while ( <$remote> ) { print }
881 The web server handing the "http" service, which is assumed to be at
882 its standard port, number 80. If your the web server you're trying to
883 connect to is at a different port (like 1080 or 8080), you should specify
884 as the named-parameter pair, C<PeerPort =E<gt> 8080>. The C<autoflush>
885 method is used on the socket because otherwise the system would buffer
886 up the output we sent it. (If you're on a Mac, you'll also need to
887 change every C<"\n"> in your code that sends data over the network to
888 be a C<"\015\012"> instead.)
890 Connecting to the server is only the first part of the process: once you
891 have the connection, you have to use the server's language. Each server
892 on the network has its own little command language that it expects as
893 input. The string that we send to the server starting with "GET" is in
894 HTTP syntax. In this case, we simply request each specified document.
895 Yes, we really are making a new connection for each document, even though
896 it's the same host. That's the way you always used to have to speak HTTP.
897 Recent versions of web browsers may request that the remote server leave
898 the connection open a little while, but the server doesn't have to honor
901 Here's an example of running that program, which we'll call I<webget>:
903 shell_prompt$ webget www.perl.com /guanaco.html
904 HTTP/1.1 404 File Not Found
905 Date: Thu, 08 May 1997 18:02:32 GMT
908 Content-type: text/html
910 <HEAD><TITLE>404 File Not Found</TITLE></HEAD>
911 <BODY><H1>File Not Found</H1>
912 The requested URL /guanaco.html was not found on this server.<P>
915 Ok, so that's not very interesting, because it didn't find that
916 particular document. But a long response wouldn't have fit on this page.
918 For a more fully-featured version of this program, you should look to
919 the I<lwp-request> program included with the LWP modules from CPAN.
921 =head2 Interactive Client with IO::Socket
923 Well, that's all fine if you want to send one command and get one answer,
924 but what about setting up something fully interactive, somewhat like
925 the way I<telnet> works? That way you can type a line, get the answer,
926 type a line, get the answer, etc.
928 This client is more complicated than the two we've done so far, but if
929 you're on a system that supports the powerful C<fork> call, the solution
930 isn't that rough. Once you've made the connection to whatever service
931 you'd like to chat with, call C<fork> to clone your process. Each of
932 these two identical process has a very simple job to do: the parent
933 copies everything from the socket to standard output, while the child
934 simultaneously copies everything from standard input to the socket.
935 To accomplish the same thing using just one process would be I<much>
936 harder, because it's easier to code two processes to do one thing than it
937 is to code one process to do two things. (This keep-it-simple principle
938 is one of the cornerstones of the Unix philosophy, and good software
939 engineering as well, which is probably why it's spread to other systems
947 my ($host, $port, $kidpid, $handle, $line);
949 unless (@ARGV == 2) { die "usage: $0 host port" }
950 ($host, $port) = @ARGV;
952 # create a tcp connection to the specified host and port
953 $handle = IO::Socket::INET->new(Proto => "tcp",
956 or die "can't connect to port $port on $host: $!";
958 $handle->autoflush(1); # so output gets there right away
959 print STDERR "[Connected to $host:$port]\n";
961 # split the program into two processes, identical twins
962 die "can't fork: $!" unless defined($kidpid = fork());
964 # the if{} block runs only in the parent process
966 # copy the socket to standard output
967 while (defined ($line = <$handle>)) {
970 kill("TERM", $kidpid); # send SIGTERM to child
972 # the else{} block runs only in the child process
974 # copy standard input to the socket
975 while (defined ($line = <STDIN>)) {
980 The C<kill> function in the parent's C<if> block is there to send a
981 signal to our child process (current running in the C<else> block)
982 as soon as the remote server has closed its end of the connection.
984 The C<kill> at the end of the parent's block is there to eliminate the
985 child process as soon as the server we connect to closes its end.
987 If the remote server sends data a byte at time, and you need that
988 data immediately without waiting for a newline (which might not happen),
989 you may wish to replace the C<while> loop in the parent with the
993 while (sysread($handle, $byte, 1) == 1) {
997 Making a system call for each byte you want to read is not very efficient
998 (to put it mildly) but is the simplest to explain and works reasonably
1001 =head1 TCP Servers with IO::Socket
1003 Setting up server is little bit more involved than running a client.
1004 The model is that the server creates a special kind of socket that
1005 does nothing but listen on a particular port for incoming connections.
1006 It does this by calling the C<IO::Socket::INET-E<gt>new()> method with
1007 slightly different arguments than the client did.
1013 This is which protocol to use. Like our clients, we'll
1014 still specify C<"tcp"> here.
1019 port in the C<LocalPort> argument, which we didn't do for the client.
1020 This is service name or port number for which you want to be the
1021 server. (Under Unix, ports under 1024 are restricted to the
1022 superuser.) In our sample, we'll use port 9000, but you can use
1023 any port that's not currently in use on your system. If you try
1024 to use one already in used, you'll get an "Address already in use"
1025 message. Under Unix, the C<netstat -a> command will show
1026 which services current have servers.
1030 The C<Listen> parameter is set to the maximum number of
1031 pending connections we can accept until we turn away incoming clients.
1032 Think of it as a call-waiting queue for your telephone.
1033 The low-level Socket module has a special symbol for the system maximum, which
1038 The C<Reuse> parameter is needed so that we restart our server
1039 manually without waiting a few minutes to allow system buffers to
1044 Once the generic server socket has been created using the parameters
1045 listed above, the server then waits for a new client to connect
1046 to it. The server blocks in the C<accept> method, which eventually an
1047 bidirectional connection to the remote client. (Make sure to autoflush
1048 this handle to circumvent buffering.)
1050 To add to user-friendliness, our server prompts the user for commands.
1051 Most servers don't do this. Because of the prompt without a newline,
1052 you'll have to use the C<sysread> variant of the interactive client above.
1054 This server accepts one of five different commands, sending output
1055 back to the client. Note that unlike most network servers, this one
1056 only handles one incoming client at a time. Multithreaded servers are
1057 covered in Chapter 6 of the Camel or in the perlipc(1) manpage.
1059 Here's the code. We'll
1063 use Net::hostent; # for OO version of gethostbyaddr
1065 $PORT = 9000; # pick something not in use
1067 $server = IO::Socket::INET->new( Proto => 'tcp',
1069 Listen => SOMAXCONN,
1072 die "can't setup server" unless $server;
1073 print "[Server $0 accepting clients]\n";
1075 while ($client = $server->accept()) {
1076 $client->autoflush(1);
1077 print $client "Welcome to $0; type help for command list.\n";
1078 $hostinfo = gethostbyaddr($client->peeraddr);
1079 printf "[Connect from %s]\n", $hostinfo->name || $client->peerhost;
1080 print $client "Command? ";
1081 while ( <$client>) {
1082 next unless /\S/; # blank line
1083 if (/quit|exit/i) { last; }
1084 elsif (/date|time/i) { printf $client "%s\n", scalar localtime; }
1085 elsif (/who/i ) { print $client `who 2>&1`; }
1086 elsif (/cookie/i ) { print $client `/usr/games/fortune 2>&1`; }
1087 elsif (/motd/i ) { print $client `cat /etc/motd 2>&1`; }
1089 print $client "Commands: quit date who cookie motd\n";
1092 print $client "Command? ";
1097 =head1 UDP: Message Passing
1099 Another kind of client-server setup is one that uses not connections, but
1100 messages. UDP communications involve much lower overhead but also provide
1101 less reliability, as there are no promises that messages will arrive at
1102 all, let alone in order and unmangled. Still, UDP offers some advantages
1103 over TCP, including being able to "broadcast" or "multicast" to a whole
1104 bunch of destination hosts at once (usually on your local subnet). If you
1105 find yourself overly concerned about reliability and start building checks
1106 into your message system, then you probably should use just TCP to start
1109 Here's a UDP program similar to the sample Internet TCP client given
1110 earlier. However, instead of checking one host at a time, the UDP version
1111 will check many of them asynchronously by simulating a multicast and then
1112 using select() to do a timed-out wait for I/O. To do something similar
1113 with TCP, you'd have to use a different socket handle for each host.
1121 my ( $count, $hisiaddr, $hispaddr, $histime,
1122 $host, $iaddr, $paddr, $port, $proto,
1123 $rin, $rout, $rtime, $SECS_of_70_YEARS);
1125 $SECS_of_70_YEARS = 2208988800;
1127 $iaddr = gethostbyname(hostname());
1128 $proto = getprotobyname('udp');
1129 $port = getservbyname('time', 'udp');
1130 $paddr = sockaddr_in(0, $iaddr); # 0 means let kernel pick
1132 socket(SOCKET, PF_INET, SOCK_DGRAM, $proto) || die "socket: $!";
1133 bind(SOCKET, $paddr) || die "bind: $!";
1136 printf "%-12s %8s %s\n", "localhost", 0, scalar localtime time;
1140 $hisiaddr = inet_aton($host) || die "unknown host";
1141 $hispaddr = sockaddr_in($port, $hisiaddr);
1142 defined(send(SOCKET, 0, 0, $hispaddr)) || die "send $host: $!";
1146 vec($rin, fileno(SOCKET), 1) = 1;
1148 # timeout after 10.0 seconds
1149 while ($count && select($rout = $rin, undef, undef, 10.0)) {
1151 ($hispaddr = recv(SOCKET, $rtime, 4, 0)) || die "recv: $!";
1152 ($port, $hisiaddr) = sockaddr_in($hispaddr);
1153 $host = gethostbyaddr($hisiaddr, AF_INET);
1154 $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
1155 printf "%-12s ", $host;
1156 printf "%8d %s\n", $histime - time, scalar localtime($histime);
1162 While System V IPC isn't so widely used as sockets, it still has some
1163 interesting uses. You can't, however, effectively use SysV IPC or
1164 Berkeley mmap() to have shared memory so as to share a variable amongst
1165 several processes. That's because Perl would reallocate your string when
1166 you weren't wanting it to.
1168 Here's a small example showing shared memory usage.
1173 $key = shmget($IPC_PRIVATE, $size , 0777 );
1174 die unless defined $key;
1176 $message = "Message #1";
1177 shmwrite($key, $message, 0, 60 ) || die "$!";
1178 shmread($key,$buff,0,60) || die "$!";
1182 print "deleting $key\n";
1183 shmctl($key ,$IPC_RMID, 0) || die "$!";
1185 Here's an example of a semaphore:
1189 $IPC_CREATE = 0001000;
1190 $key = semget($IPC_KEY, $nsems , 0666 | $IPC_CREATE );
1191 die if !defined($key);
1194 Put this code in a separate file to be run in more than one process.
1195 Call the file F<take>:
1197 # create a semaphore
1200 $key = semget($IPC_KEY, 0 , 0 );
1201 die if !defined($key);
1207 # wait for semaphore to be zero
1209 $opstring1 = pack("sss", $semnum, $semop, $semflag);
1211 # Increment the semaphore count
1213 $opstring2 = pack("sss", $semnum, $semop, $semflag);
1214 $opstring = $opstring1 . $opstring2;
1216 semop($key,$opstring) || die "$!";
1218 Put this code in a separate file to be run in more than one process.
1219 Call this file F<give>:
1221 # 'give' the semaphore
1222 # run this in the original process and you will see
1223 # that the second process continues
1226 $key = semget($IPC_KEY, 0, 0);
1227 die if !defined($key);
1232 # Decrement the semaphore count
1234 $opstring = pack("sss", $semnum, $semop, $semflag);
1236 semop($key,$opstring) || die "$!";
1238 The SysV IPC code above was written long ago, and it's definitely
1239 clunky looking. It should at the very least be made to C<use strict>
1240 and C<require "sys/ipc.ph">. Better yet, check out the IPC::SysV modules
1245 If you are running under version 5.000 (dubious) or 5.001, you can still
1246 use most of the examples in this document. You may have to remove the
1247 C<use strict> and some of the my() statements for 5.000, and for both
1248 you'll have to load in version 1.2 or older of the F<Socket.pm> module, which
1249 is included in I<perl5.002>.
1251 Most of these routines quietly but politely return C<undef> when they fail
1252 instead of causing your program to die right then and there due to an
1253 uncaught exception. (Actually, some of the new I<Socket> conversion
1254 functions croak() on bad arguments.) It is therefore essential
1255 that you should check the return values of these functions. Always begin
1256 your socket programs this way for optimal success, and don't forget to add
1257 B<-T> taint checking flag to the pound-bang line for servers:
1267 All these routines create system-specific portability problems. As noted
1268 elsewhere, Perl is at the mercy of your C libraries for much of its system
1269 behaviour. It's probably safest to assume broken SysV semantics for
1270 signals and to stick with simple TCP and UDP socket operations; e.g., don't
1271 try to pass open file descriptors over a local UDP datagram socket if you
1272 want your code to stand a chance of being portable.
1274 Because few vendors provide C libraries that are safely re-entrant,
1275 the prudent programmer will do little else within a handler beyond
1276 setting a numeric variable that already exists; or, if locked into
1277 a slow (restarting) system call, using die() to raise an exception
1278 and longjmp(3) out. In fact, even these may in some cases cause a
1279 core dump. It's probably best to avoid signals except where they are
1280 absolutely inevitable. This perilous problems will be addressed in a
1281 future release of Perl.
1285 Tom Christiansen, with occasional vestiges of Larry Wall's original
1286 version and suggestions from the Perl Porters.
1290 There's a lot more to networking than this, but this should get you
1293 For intrepid programmers, the classic textbook I<Unix Network Programming>
1294 by Richard Stevens (published by Addison-Wesley). Note that most books
1295 on networking address networking from the perspective of a C programmer;
1296 translation to Perl is left as an exercise for the reader.
1298 The IO::Socket(3) manpage describes the object library, and the Socket(3)
1299 manpage describes the low-level interface to sockets. Besides the obvious
1300 functions in L<perlfunc>, you should also check out the F<modules> file
1301 at your nearest CPAN site. (See L<perlmodlib> or best yet, the F<Perl
1302 FAQ> for a description of what CPAN is and where to get it.)
1304 Section 5 of the F<modules> file is devoted to "Networking, Device Control
1305 (modems), and Interprocess Communication", and contains numerous unbundled
1306 modules numerous networking modules, Chat and Expect operations, CGI
1307 programming, DCE, FTP, IPC, NNTP, Proxy, Ptty, RPC, SNMP, SMTP, Telnet,
1308 Threads, and ToolTalk--just to name a few.