[ViewVC] Diff of: cvs/EV/EV.pm

Comparing EV/EV.pm (file contents):
Revision 1.1 by root, Fri Oct 26 16:50:05 2007 UTC vs.
Revision 1.79 by root, Sat Dec 22 16:37:07 2007 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	EV - perl interface to libevent, monkey.org/~provos/libevent/	3	EV - perl interface to libev, a high performance full-featured event loop
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use EV;	7	use EV;
		8
		9	# TIMERS
		10
		11	my $w = EV::timer 2, 0, sub {
		12	warn "is called after 2s";
		13	};
		14
		15	my $w = EV::timer 2, 2, sub {
		16	warn "is called roughly every 2s (repeat = 2)";
		17	};
		18
		19	undef $w; # destroy event watcher again
		20
		21	my $w = EV::periodic 0, 60, 0, sub {
		22	warn "is called every minute, on the minute, exactly";
		23	};
		24
		25	# IO
		26
		27	my $w = EV::io *STDIN, EV::READ, sub {
		28	my ($w, $revents) = @_; # all callbacks receive the watcher and event mask
		29	warn "stdin is readable, you entered: ", <STDIN>;
		30	};
		31
		32	# SIGNALS
		33
		34	my $w = EV::signal 'QUIT', sub {
		35	warn "sigquit received\n";
		36	};
		37
		38	# CHILD/PID STATUS CHANGES
		39
		40	my $w = EV::child 666, sub {
		41	my ($w, $revents) = @_;
		42	my $status = $w->rstatus;
		43	};
		44
		45	# STAT CHANGES
		46	my $w = EV::stat "/etc/passwd", 10, sub {
		47	my ($w, $revents) = @_;
		48	warn $w->path, " has changed somehow.\n";
		49	};
		50
		51	# MAINLOOP
		52	EV::loop; # loop until EV::unloop is called or all watchers stop
		53	EV::loop EV::LOOP_ONESHOT; # block until at least one event could be handled
		54	EV::loop EV::LOOP_NONBLOCK; # try to handle same events, but do not block
8		55
9	=head1 DESCRIPTION	56	=head1 DESCRIPTION
10		57
11	This module provides an interface to libevent	58	This module provides an interface to libev
12	(L<http://monkey.org/~provos/libevent/>).	59	(L<http://software.schmorp.de/pkg/libev.html>). While the documentation
		60	below is comprehensive, one might also consult the documentation of libev
		61	itself (L<http://cvs.schmorp.de/libev/ev.html>) for more subtle details on
		62	watcher semantics or some discussion on the available backends, or how to
		63	force a specific backend with C<LIBEV_FLAGS>, or just about in any case
		64	because it has much more detailed information.
13		65
14	=cut	66	=cut
15		67
16	package EV;	68	package EV;
17		69
18	use strict;	70	use strict;
19		71
20	BEGIN {	72	BEGIN {
21	our $VERSION = '0.01';	73	our $VERSION = '2.0';
22	use XSLoader;	74	use XSLoader;
23	XSLoader::load "EV", $VERSION;	75	XSLoader::load "EV", $VERSION;
24	}	76	}
25		77
26	=head1 FUNCTIONAL INTERFACE	78	@EV::IO::ISA =
		79	@EV::Timer::ISA =
		80	@EV::Periodic::ISA =
		81	@EV::Signal::ISA =
		82	@EV::Child::ISA =
		83	@EV::Stat::ISA =
		84	@EV::Idle::ISA =
		85	@EV::Prepare::ISA =
		86	@EV::Check::ISA =
		87	@EV::Embed::ISA =
		88	@EV::Fork::ISA =
		89	"EV::Watcher";
27		90
28	=over 4	91	@EV::Loop::Default::ISA = "EV::Loop";
29		92
30	=back	93	=head1 EVENT LOOPS
31		94
		95	EV supports multiple event loops: There is a single "default event loop"
		96	that can handle everything including signals and child watchers, and any
		97	number of "dynamic event loops" that can use different backends (with
		98	various limitations), but no child and signal watchers.
32		99
33	=head1 OBJECT-ORIENTED INTERFACE	100	You do not have to do anything to create the default event loop: When
		101	the module is loaded a suitable backend is selected on the premise of
		102	selecting a working backend (which for example rules out kqueue on most
		103	BSDs). Modules should, unless they have "special needs" always use the
		104	default loop as this is fastest (perl-wise), best supported by other
		105	modules (e.g. AnyEvent or Coro) and most portable event loop.
34		106
35	The object oriented interface lets you configure your own encoding or	107	For specific programs you can create additional event loops dynamically.
36	decoding style, within the limits of supported formats.
37		108
38	=over 4	109	=over 4
39		110
40	=back	111	=item $loop = new EV::loop [$flags]
41		112
42	=head1 BUGS	113	Create a new event loop as per the specified flags. Please refer to the
		114	C<ev_loop_new ()> function description in the libev documentation
		115	(L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#GLOBAL_FUNCTIONS>)
		116	for more info.
		117
		118	The loop will automatically be destroyed when it is no longer referenced
		119	by any watcher and the loop object goes out of scope.
		120
		121	Using C<EV::FLAG_FORKCHECK> is recommended, as only the default event loop
		122	is protected by this module.
		123
		124	=item $loop->loop_fork
		125
		126	Must be called after a fork in the child, before entering or continuing
		127	the event loop. An alternative is to use C<EV::FLAG_FORKCHECK> which calls
		128	this fucntion automatically, at some performance loss (refer to the libev
		129	documentation).
		130
		131	=item $loop = EV::default_loop [$flags]
		132
		133	Return the default loop (which is a singleton object).
		134
		135	=back
		136
		137
		138	=head1 BASIC INTERFACE
		139
		140	=over 4
		141
		142	=item $EV::DIED
		143
		144	Must contain a reference to a function that is called when a callback
		145	throws an exception (with $@ containing the error). The default prints an
		146	informative message and continues.
		147
		148	If this callback throws an exception it will be silently ignored.
		149
		150	=item $flags = EV::supported_backends
		151
		152	=item $flags = EV::recommended_backends
		153
		154	=item $flags = EV::embeddable_backends
		155
		156	Returns the set (see C<EV::BACKEND_*> flags) of backends supported by this
		157	instance of EV, the set of recommended backends (supposed to be good) for
		158	this platform and the set of embeddable backends (see EMBED WATCHERS).
		159
		160	=item $time = EV::time
		161
		162	Returns the current time in (fractional) seconds since the epoch.
		163
		164	=item $time = EV::now
		165
		166	=item $time = $loop->now
		167
		168	Returns the time the last event loop iteration has been started. This
		169	is the time that (relative) timers are based on, and refering to it is
		170	usually faster then calling EV::time.
		171
		172	=item $backend = EV::backend
		173
		174	=item $backend = $loop->backend
		175
		176	Returns an integer describing the backend used by libev (EV::METHOD_SELECT
		177	or EV::METHOD_EPOLL).
		178
		179	=item EV::loop [$flags]
		180
		181	=item $loop->loop ([$flags])
		182
		183	Begin checking for events and calling callbacks. It returns when a
		184	callback calls EV::unloop.
		185
		186	The $flags argument can be one of the following:
		187
		188	0 as above
		189	EV::LOOP_ONESHOT block at most once (wait, but do not loop)
		190	EV::LOOP_NONBLOCK do not block at all (fetch/handle events but do not wait)
		191
		192	=item EV::unloop [$how]
		193
		194	=item $loop->unloop ([$how])
		195
		196	When called with no arguments or an argument of EV::UNLOOP_ONE, makes the
		197	innermost call to EV::loop return.
		198
		199	When called with an argument of EV::UNLOOP_ALL, all calls to EV::loop will return as
		200	fast as possible.
		201
		202	=item $count = EV::loop_count
		203
		204	=item $count = $loop->loop_count
		205
		206	Return the number of times the event loop has polled for new
		207	events. Sometiems useful as a generation counter.
		208
		209	=item EV::once $fh_or_undef, $events, $timeout, $cb->($revents)
		210
		211	=item $loop->once ($fh_or_undef, $events, $timeout, $cb->($revents))
		212
		213	This function rolls together an I/O and a timer watcher for a single
		214	one-shot event without the need for managing a watcher object.
		215
		216	If C<$fh_or_undef> is a filehandle or file descriptor, then C<$events>
		217	must be a bitset containing either C<EV::READ>, C<EV::WRITE> or C<EV::READ
		218	\| EV::WRITE>, indicating the type of I/O event you want to wait for. If
		219	you do not want to wait for some I/O event, specify C<undef> for
		220	C<$fh_or_undef> and C<0> for C<$events>).
		221
		222	If timeout is C<undef> or negative, then there will be no
		223	timeout. Otherwise a EV::timer with this value will be started.
		224
		225	When an error occurs or either the timeout or I/O watcher triggers, then
		226	the callback will be called with the received event set (in general
		227	you can expect it to be a combination of C<EV::ERROR>, C<EV::READ>,
		228	C<EV::WRITE> and C<EV::TIMEOUT>).
		229
		230	EV::once doesn't return anything: the watchers stay active till either
		231	of them triggers, then they will be stopped and freed, and the callback
		232	invoked.
		233
		234	=item EV::feed_fd_event ($fd, $revents)
		235
		236	=item $loop->feed_fd_event ($fd, $revents)
		237
		238	Feed an event on a file descriptor into EV. EV will react to this call as
		239	if the readyness notifications specified by C<$revents> (a combination of
		240	C<EV::READ> and C<EV::WRITE>) happened on the file descriptor C<$fd>.
		241
		242	=item EV::feed_signal_event ($signal)
		243
		244	Feed a signal event into EV. EV will react to this call as if the signal
		245	specified by C<$signal> had occured.
		246
		247	=item EV::set_io_collect_interval $time
		248
		249	=item $loop->set_io_collect_interval ($time)
		250
		251	=item EV::set_timeout_collect_interval $time
		252
		253	=item $loop->set_timeout_collect_interval ($time)
		254
		255	These advanced functions set the minimum block interval when polling for I/O events and the minimum
		256	wait interval for timer events. See the libev documentation at
		257	L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#FUNCTIONS_CONTROLLING_THE_EVENT_LOOP> for
		258	a more detailed discussion.
		259
		260	=back
		261
		262
		263	=head1 WATCHER OBJECTS
		264
		265	A watcher is an object that gets created to record your interest in some
		266	event. For instance, if you want to wait for STDIN to become readable, you
		267	would create an EV::io watcher for that:
		268
		269	my $watcher = EV::io *STDIN, EV::READ, sub {
		270	my ($watcher, $revents) = @_;
		271	warn "yeah, STDIN should now be readable without blocking!\n"
		272	};
		273
		274	All watchers can be active (waiting for events) or inactive (paused). Only
		275	active watchers will have their callbacks invoked. All callbacks will be
		276	called with at least two arguments: the watcher and a bitmask of received
		277	events.
		278
		279	Each watcher type has its associated bit in revents, so you can use the
		280	same callback for multiple watchers. The event mask is named after the
		281	type, i..e. EV::child sets EV::CHILD, EV::prepare sets EV::PREPARE,
		282	EV::periodic sets EV::PERIODIC and so on, with the exception of I/O events
		283	(which can set both EV::READ and EV::WRITE bits), and EV::timer (which
		284	uses EV::TIMEOUT).
		285
		286	In the rare case where one wants to create a watcher but not start it at
		287	the same time, each constructor has a variant with a trailing C<_ns> in
		288	its name, e.g. EV::io has a non-starting variant EV::io_ns and so on.
		289
		290	Please note that a watcher will automatically be stopped when the watcher
		291	object is destroyed, so you I<need> to keep the watcher objects returned by
		292	the constructors.
		293
		294	Also, all methods changing some aspect of a watcher (->set, ->priority,
		295	->fh and so on) automatically stop and start it again if it is active,
		296	which means pending events get lost.
		297
		298	=head2 COMMON WATCHER METHODS
		299
		300	This section lists methods common to all watchers.
		301
		302	=over 4
		303
		304	=item $w->start
		305
		306	Starts a watcher if it isn't active already. Does nothing to an already
		307	active watcher. By default, all watchers start out in the active state
		308	(see the description of the C<_ns> variants if you need stopped watchers).
		309
		310	=item $w->stop
		311
		312	Stop a watcher if it is active. Also clear any pending events (events that
		313	have been received but that didn't yet result in a callback invocation),
		314	regardless of whether the watcher was active or not.
		315
		316	=item $bool = $w->is_active
		317
		318	Returns true if the watcher is active, false otherwise.
		319
		320	=item $current_data = $w->data
		321
		322	=item $old_data = $w->data ($new_data)
		323
		324	Queries a freely usable data scalar on the watcher and optionally changes
		325	it. This is a way to associate custom data with a watcher:
		326
		327	my $w = EV::timer 60, 0, sub {
		328	warn $_[0]->data;
		329	};
		330	$w->data ("print me!");
		331
		332	=item $current_cb = $w->cb
		333
		334	=item $old_cb = $w->cb ($new_cb)
		335
		336	Queries the callback on the watcher and optionally changes it. You can do
		337	this at any time without the watcher restarting.
		338
		339	=item $current_priority = $w->priority
		340
		341	=item $old_priority = $w->priority ($new_priority)
		342
		343	Queries the priority on the watcher and optionally changes it. Pending
		344	watchers with higher priority will be invoked first. The valid range of
		345	priorities lies between EV::MAXPRI (default 2) and EV::MINPRI (default
		346	-2). If the priority is outside this range it will automatically be
		347	normalised to the nearest valid priority.
		348
		349	The default priority of any newly-created watcher is 0.
		350
		351	Note that the priority semantics have not yet been fleshed out and are
		352	subject to almost certain change.
		353
		354	=item $w->invoke ($revents)
		355
		356	Call the callback now with the given event mask.
		357
		358	=item $w->feed_event ($revents)
		359
		360	Feed some events on this watcher into EV. EV will react to this call as if
		361	the watcher had received the given C<$revents> mask.
		362
		363	=item $revents = $w->clear_pending
		364
		365	If the watcher is pending, this function clears its pending status and
		366	returns its C<$revents> bitset (as if its callback was invoked). If the
		367	watcher isn't pending it does nothing and returns C<0>.
		368
		369	=item $previous_state = $w->keepalive ($bool)
		370
		371	Normally, C<EV::loop> will return when there are no active watchers
		372	(which is a "deadlock" because no progress can be made anymore). This is
		373	convinient because it allows you to start your watchers (and your jobs),
		374	call C<EV::loop> once and when it returns you know that all your jobs are
		375	finished (or they forgot to register some watchers for their task :).
		376
		377	Sometimes, however, this gets in your way, for example when the module
		378	that calls C<EV::loop> (usually the main program) is not the same module
		379	as a long-living watcher (for example a DNS client module written by
		380	somebody else even). Then you might want any outstanding requests to be
		381	handled, but you would not want to keep C<EV::loop> from returning just
		382	because you happen to have this long-running UDP port watcher.
		383
		384	In this case you can clear the keepalive status, which means that even
		385	though your watcher is active, it won't keep C<EV::loop> from returning.
		386
		387	The initial value for keepalive is true (enabled), and you cna change it
		388	any time.
		389
		390	Example: Register an I/O watcher for some UDP socket but do not keep the
		391	event loop from running just because of that watcher.
		392
		393	my $udp_socket = ...
		394	my $udp_watcher = EV::io $udp_socket, EV::READ, sub { ... };
		395	$1000udp_watcher->keepalive (0);
		396
		397	=item $loop = $w->loop
		398
		399	Return the loop that this watcher is attached to.
		400
		401	=back
		402
		403
		404	=head1 WATCHER TYPES
		405
		406	Each of the following subsections describes a single watcher type.
		407
		408	=head3 I/O WATCHERS - is this file descriptor readable or writable?
		409
		410	=over 4
		411
		412	=item $w = EV::io $fileno_or_fh, $eventmask, $callback
		413
		414	=item $w = EV::io_ns $fileno_or_fh, $eventmask, $callback
		415
		416	=item $w = $loop->io ($fileno_or_fh, $eventmask, $callback)
		417
		418	=item $w = $loop->io_ns ($fileno_or_fh, $eventmask, $callback)
		419
		420	As long as the returned watcher object is alive, call the C<$callback>
		421	when at least one of events specified in C<$eventmask> occurs.
		422
		423	The $eventmask can be one or more of these constants ORed together:
		424
		425	EV::READ wait until read() wouldn't block anymore
		426	EV::WRITE wait until write() wouldn't block anymore
		427
		428	The C<io_ns> variant doesn't start (activate) the newly created watcher.
		429
		430	=item $w->set ($fileno_or_fh, $eventmask)
		431
		432	Reconfigures the watcher, see the constructor above for details. Can be
		433	called at any time.
		434
		435	=item $current_fh = $w->fh
		436
		437	=item $old_fh = $w->fh ($new_fh)
		438
		439	Returns the previously set filehandle and optionally set a new one.
		440
		441	=item $current_eventmask = $w->events
		442
		443	=item $old_eventmask = $w->events ($new_eventmask)
		444
		445	Returns the previously set event mask and optionally set a new one.
		446
		447	=back
		448
		449
		450	=head3 TIMER WATCHERS - relative and optionally repeating timeouts
		451
		452	=over 4
		453
		454	=item $w = EV::timer $after, $repeat, $callback
		455
		456	=item $w = EV::timer_ns $after, $repeat, $callback
		457
		458	=item $w = $loop->timer ($after, $repeat, $callback)
		459
		460	=item $w = $loop->timer_ns ($after, $repeat, $callback)
		461
		462	Calls the callback after C<$after> seconds (which may be fractional). If
		463	C<$repeat> is non-zero, the timer will be restarted (with the $repeat
		464	value as $after) after the callback returns.
		465
		466	This means that the callback would be called roughly after C<$after>
		467	seconds, and then every C<$repeat> seconds. The timer does his best not
		468	to drift, but it will not invoke the timer more often then once per event
		469	loop iteration, and might drift in other cases. If that isn't acceptable,
		470	look at EV::periodic, which can provide long-term stable timers.
		471
		472	The timer is based on a monotonic clock, that is, if somebody is sitting
		473	in front of the machine while the timer is running and changes the system
		474	clock, the timer will nevertheless run (roughly) the same time.
		475
		476	The C<timer_ns> variant doesn't start (activate) the newly created watcher.
		477
		478	=item $w->set ($after, $repeat)
		479
		480	Reconfigures the watcher, see the constructor above for details. Can be called at
		481	any time.
		482
		483	=item $w->again
		484
		485	Similar to the C<start> method, but has special semantics for repeating timers:
		486
		487	If the timer is active and non-repeating, it will be stopped.
		488
		489	If the timer is active and repeating, reset the timeout to occur
		490	C<$repeat> seconds after now.
		491
		492	If the timer is inactive and repeating, start it using the repeat value.
		493
		494	Otherwise do nothing.
		495
		496	This behaviour is useful when you have a timeout for some IO
		497	operation. You create a timer object with the same value for C<$after> and
		498	C<$repeat>, and then, in the read/write watcher, run the C<again> method
		499	on the timeout.
		500
		501	=back
		502
		503
		504	=head3 PERIODIC WATCHERS - to cron or not to cron?
		505
		506	=over 4
		507
		508	=item $w = EV::periodic $at, $interval, $reschedule_cb, $callback
		509
		510	=item $w = EV::periodic_ns $at, $interval, $reschedule_cb, $callback
		511
		512	=item $w = $loop->periodic ($at, $interval, $reschedule_cb, $callback)
		513
		514	=item $w = $loop->periodic_ns ($at, $interval, $reschedule_cb, $callback)
		515
		516	Similar to EV::timer, but is not based on relative timeouts but on
		517	absolute times. Apart from creating "simple" timers that trigger "at" the
		518	specified time, it can also be used for non-drifting absolute timers and
		519	more complex, cron-like, setups that are not adversely affected by time
		520	jumps (i.e. when the system clock is changed by explicit date -s or other
		521	means such as ntpd). It is also the most complex watcher type in EV.
		522
		523	It has three distinct "modes":
		524
		525	=over 4
		526
		527	=item * absolute timer ($interval = $reschedule_cb = 0)
		528
		529	This time simply fires at the wallclock time C<$at> and doesn't repeat. It
		530	will not adjust when a time jump occurs, that is, if it is to be run
		531	at January 1st 2011 then it will run when the system time reaches or
		532	surpasses this time.
		533
		534	=item * non-repeating interval timer ($interval > 0, $reschedule_cb = 0)
		535
		536	In this mode the watcher will always be scheduled to time out at the
		537	next C<$at + N * $interval> time (for some integer N) and then repeat,
		538	regardless of any time jumps.
		539
		540	This can be used to create timers that do not drift with respect to system
		541	time:
		542
		543	my $hourly = EV::periodic 0, 3600, 0, sub { print "once/hour\n" };
		544
		545	That doesn't mean there will always be 3600 seconds in between triggers,
		546	but only that the the clalback will be called when the system time shows a
		547	full hour (UTC).
		548
		549	Another way to think about it (for the mathematically inclined) is that
		550	EV::periodic will try to run the callback in this mode at the next
		551	possible time where C<$time = $at (mod $interval)>, regardless of any time
		552	jumps.
		553
		554	=item * manual reschedule mode ($reschedule_cb = coderef)
		555
		556	In this mode $interval and $at are both being ignored. Instead, each
		557	time the periodic watcher gets scheduled, the reschedule callback
		558	($reschedule_cb) will be called with the watcher as first, and the current
		559	time as second argument.
		560
		561	I<This callback MUST NOT stop or destroy this or any other periodic
		562	watcher, ever>. If you need to stop it, return 1e30 and stop it
		563	afterwards.
		564
		565	It must return the next time to trigger, based on the passed time value
		566	(that is, the lowest time value larger than to the second argument). It
		567	will usually be called just before the callback will be triggered, but
		568	might be called at other times, too.
		569
		570	This can be used to create very complex timers, such as a timer that
		571	triggers on each midnight, local time (actually 24 hours after the last
		572	midnight, to keep the example simple. If you know a way to do it correctly
		573	in about the same space (without requiring elaborate modules), drop me a
		574	note :):
		575
		576	my $daily = EV::periodic 0, 0, sub {
		577	my ($w, $now) = @_;
		578
		579	use Time::Local ();
		580	my (undef, undef, undef, $d, $m, $y) = localtime $now;
		581	86400 + Time::Local::timelocal 0, 0, 0, $d, $m, $y
		582	}, sub {
		583	print "it's midnight or likely shortly after, now\n";
		584	};
		585
		586	=back
		587
		588	The C<periodic_ns> variant doesn't start (activate) the newly created watcher.
		589
		590	=item $w->set ($at, $interval, $reschedule_cb)
		591
		592	Reconfigures the watcher, see the constructor above for details. Can be called at
		593	any time.
		594
		595	=item $w->again
		596
		597	Simply stops and starts the watcher again.
		598
		599	=item $time = $w->at
		600
		601	Return the time that the watcher is expected to trigger next.
		602
		603	=back
		604
		605
		606	=head3 SIGNAL WATCHERS - signal me when a signal gets signalled!
		607
		608	=over 4
		609
		610	=item $w = EV::signal $signal, $callback
		611
		612	=item $w = EV::signal_ns $signal, $callback
		613
		614	Call the callback when $signal is received (the signal can be specified by
		615	number or by name, just as with C<kill> or C<%SIG>).
		616
		617	EV will grab the signal for the process (the kernel only allows one
		618	component to receive a signal at a time) when you start a signal watcher,
		619	and removes it again when you stop it. Perl does the same when you
		620	add/remove callbacks to C<%SIG>, so watch out.
		621
		622	You can have as many signal watchers per signal as you want.
		623
		624	The C<signal_ns> variant doesn't start (activate) the newly created watcher.
		625
		626	=item $w->set ($signal)
		627
		628	Reconfigures the watcher, see the constructor above for details. Can be
		629	called at any time.
		630
		631	=item $current_signum = $w->signal
		632
		633	=item $old_signum = $w->signal ($new_signal)
		634
		635	Returns the previously set signal (always as a number not name) and
		636	optionally set a new one.
		637
		638	=back
		639
		640
		641	=head3 CHILD WATCHERS - watch out for process status changes
		642
		643	=over 4
		644
		645	=item $w = EV::child $pid, $callback
		646
		647	=item $w = EV::child_ns $pid, $callback
		648
		649	=item $w = $loop->child ($pid, $callback)
		650
		651	=item $w = $loop->child_ns ($pid, $callback)
		652
		653	Call the callback when a status change for pid C<$pid> (or any pid if
		654	C<$pid> is 0) has been received. More precisely: when the process receives
		655	a C<SIGCHLD>, EV will fetch the outstanding exit/wait status for all
		656	changed/zombie children and call the callback.
		657
		658	It is valid (and fully supported) to install a child watcher after a child
		659	has exited but before the event loop has started its next iteration (for
		660	example, first you C<fork>, then the new child process might exit, and
		661	only then do you install a child watcher in the parent for the new pid).
		662
		663	You can access both exit (or tracing) status and pid by using the
		664	C<rstatus> and C<rpid> methods on the watcher object.
		665
		666	You can have as many pid watchers per pid as you want, they will all be
		667	called.
		668
		669	The C<child_ns> variant doesn't start (activate) the newly created watcher.
		670
		671	=item $w->set ($pid)
		672
		673	Reconfigures the watcher, see the constructor above for details. Can be called at
		674	any time.
		675
		676	=item $current_pid = $w->pid
		677
		678	=item $old_pid = $w->pid ($new_pid)
		679
		680	Returns the previously set process id and optionally set a new one.
		681
		682	=item $exit_status = $w->rstatus
		683
		684	Return the exit/wait status (as returned by waitpid, see the waitpid entry
		685	in perlfunc).
		686
		687	=item $pid = $w->rpid
		688
		689	Return the pid of the awaited child (useful when you have installed a
		690	watcher for all pids).
		691
		692	=back
		693
		694
		695	=head3 STAT WATCHERS - did the file attributes just change?
		696
		697	=over 4
		698
		699	=item $w = EV::stat $path, $interval, $callback
		700
		701	=item $w = EV::stat_ns $path, $interval, $callback
		702
		703	=item $w = $loop->stat ($path, $interval, $callback)
		704
		705	=item $w = $loop->stat_ns ($path, $interval, $callback)
		706
		707	Call the callback when a file status change has been detected on
		708	C<$path>. The C<$path> does not need to exist, changing from "path exists"
		709	to "path does not exist" is a status change like any other.
		710
		711	The C<$interval> is a recommended polling interval for systems where
		712	OS-supported change notifications don't exist or are not supported. If
		713	you use C<0> then an unspecified default is used (which is highly
		714	recommended!), which is to be expected to be around five seconds usually.
		715
		716	This watcher type is not meant for massive numbers of stat watchers,
		717	as even with OS-supported change notifications, this can be
		718	resource-intensive.
		719
		720	The C<stat_ns> variant doesn't start (activate) the newly created watcher.
		721
		722	=item ... = $w->stat
		723
		724	This call is very similar to the perl C<stat> built-in: It stats (using
		725	C<lstat>) the path specified in the watcher and sets perls stat cache (as
		726	well as EV's idea of the current stat values) to the values found.
		727
		728	In scalar context, a boolean is return indicating success or failure of
		729	the stat. In list context, the same 13-value list as with stat is returned
		730	(except that the blksize and blocks fields are not reliable).
		731
		732	In the case of an error, errno is set to C<ENOENT> (regardless of the
		733	actual error value) and the C<nlink> value is forced to zero (if the stat
		734	was successful then nlink is guaranteed to be non-zero).
		735
		736	See also the next two entries for more info.
		737
		738	=item ... = $w->attr
		739
		740	Just like C<< $w->stat >>, but without the initial stat'ing: this returns
		741	the values most recently detected by EV. See the next entry for more info.
		742
		743	=item ... = $w->prev
		744
		745	Just like C<< $w->stat >>, but without the initial stat'ing: this returns
		746	the previous set of values, before the change.
		747
		748	That is, when the watcher callback is invoked, C<< $w->prev >> will be set
		749	to the values found I<before> a change was detected, while C<< $w->attr >>
		750	returns the values found leading to the change detection. The difference (if any)
		751	between C<prev> and C<attr> is what triggered the callback.
		752
		753	If you did something to the filesystem object and do not want to trigger
		754	yet another change, you can call C<stat> to update EV's idea of what the
		755	current attributes are.
		756
		757	=item $w->set ($path, $interval)
		758
		759	Reconfigures the watcher, see the constructor above for details. Can be
		760	called at any time.
		761
		762	=item $current_path = $w->path
		763
		764	=item $old_path = $w->path ($new_path)
		765
		766	Returns the previously set path and optionally set a new one.
		767
		768	=item $current_interval = $w->interval
		769
		770	=item $old_interval = $w->interval ($new_interval)
		771
		772	Returns the previously set interval and optionally set a new one. Can be
		773	used to query the actual interval used.
		774
		775	=back
		776
		777
		778	=head3 IDLE WATCHERS - when you've got nothing better to do...
		779
		780	=over 4
		781
		782	=item $w = EV::idle $callback
		783
		784	=item $w = EV::idle_ns $callback
		785
		786	=item $w = $loop->idle ($callback)
		787
		788	=item $w = $loop->idle_ns ($callback)
		789
		790	Call the callback when there are no other pending watchers of the same or
		791	higher priority (excluding check, prepare and other idle watchers of the
		792	same or lower priority, of course). They are called idle watchers because
		793	when the watcher is the highest priority pending event in the process, the
		794	process is considered to be idle at that priority.
		795
		796	If you want a watcher that is only ever called when I<no> other events are
		797	outstanding you have to set the priority to C<EV::MINPRI>.
		798
		799	The process will not block as long as any idle watchers are active, and
		800	they will be called repeatedly until stopped.
		801
		802	For example, if you have idle watchers at priority C<0> and C<1>, and
		803	an I/O watcher at priority C<0>, then the idle watcher at priority C<1>
		804	and the I/O watcher will always run when ready. Only when the idle watcher
		805	at priority C<1> is stopped and the I/O watcher at priority C<0> is not
		806	pending with the C<0>-priority idle watcher be invoked.
		807
		808	The C<idle_ns> variant doesn't start (activate) the newly created watcher.
		809
		810	=back
		811
		812
		813	=head3 PREPARE WATCHERS - customise your event loop!
		814
		815	=over 4
		816
		817	=item $w = EV::prepare $callback
		818
		819	=item $w = EV::prepare_ns $callback
		820
		821	=item $w = $loop->prepare ($callback)
		822
		823	=item $w = $loop->prepare_ns ($callback)
		824
		825	Call the callback just before the process would block. You can still
		826	create/modify any watchers at this point.
		827
		828	See the EV::check watcher, below, for explanations and an example.
		829
		830	The C<prepare_ns> variant doesn't start (activate) the newly created watcher.
		831
		832	=back
		833
		834
		835	=head3 CHECK WATCHERS - customise your event loop even more!
		836
		837	=over 4
		838
		839	=item $w = EV::check $callback
		840
		841	=item $w = EV::check_ns $callback
		842
		843	=item $w = $loop->check ($callback)
		844
		845	=item $w = $loop->check_ns ($callback)
		846
		847	Call the callback just after the process wakes up again (after it has
		848	gathered events), but before any other callbacks have been invoked.
		849
		850	This is used to integrate other event-based software into the EV
		851	mainloop: You register a prepare callback and in there, you create io and
		852	timer watchers as required by the other software. Here is a real-world
		853	example of integrating Net::SNMP (with some details left out):
		854
		855	our @snmp_watcher;
		856
		857	our $snmp_prepare = EV::prepare sub {
		858	# do nothing unless active
		859	$dispatcher->{_event_queue_h}
		860	or return;
		861
		862	# make the dispatcher handle any outstanding stuff
		863	... not shown
		864
		865	# create an I/O watcher for each and every socket
		866	@snmp_watcher = (
		867	(map { EV::io $_, EV::READ, sub { } }
		868	keys %{ $dispatcher->{_descriptors} }),
		869
		870	EV::timer +($event->[Net::SNMP::Dispatcher::_ACTIVE]
		871	? $event->[Net::SNMP::Dispatcher::_TIME] - EV::now : 0),
		872	0, sub { },
		873	);
		874	};
		875
		876	The callbacks are irrelevant (and are not even being called), the
		877	only purpose of those watchers is to wake up the process as soon as
		878	one of those events occurs (socket readable, or timer timed out). The
		879	corresponding EV::check watcher will then clean up:
		880
		881	our $snmp_check = EV::check sub {
		882	# destroy all watchers
		883	@snmp_watcher = ();
		884
		885	# make the dispatcher handle any new stuff
		886	... not shown
		887	};
		888
		889	The callbacks of the created watchers will not be called as the watchers
		890	are destroyed before this cna happen (remember EV::check gets called
		891	first).
		892
		893	The C<check_ns> variant doesn't start (activate) the newly created watcher.
		894
		895	=back
		896
		897
		898	=head3 FORK WATCHERS - the audacity to resume the event loop after a fork
		899
		900	Fork watchers are called when a C<fork ()> was detected. The invocation
		901	is done before the event loop blocks next and before C<check> watchers
		902	are being called, and only in the child after the fork.
		903
		904	=over 4
		905
		906	=item $w = EV::fork $callback
		907
		908	=item $w = EV::fork_ns $callback
		909
		910	=item $w = $loop->fork ($callback)
		911
		912	=item $w = $loop->fork_ns ($callback)
		913
		914	Call the callback before the event loop is resumed in the child process
		915	after a fork.
		916
		917	The C<fork_ns> variant doesn't start (activate) the newly created watcher.
		918
		919	=back
		920
		921
		922	=head3 EMBED WATCHERS - when one backend isn't enough...
		923
		924	This is a rather advanced watcher type that lets you embed one event loop
		925	into another (currently only IO events are supported in the embedded
		926	loop, other types of watchers might be handled in a delayed or incorrect
		927	fashion and must not be used).
		928
		929	See the libev documentation at
		930	L<http://pod.tst.eu/http://cvs.schmorp.de/libev/ev.pod#code_ev_embed_code_when_one_backend_>
		931	for more details.
		932
		933	In short, this watcher is most useful on BSD systems without working
		934	kqueue to still be able to handle a large number of sockets:
		935
		936	my $socket_loop;
		937
		938	# check wether we use SELECT or POLL _and_ KQUEUE is supported
		939	if (
		940	(EV::backend & (EV::BACKEND_POLL \| EV::BACKEND_SELECT))
		941	&& (EV::supported_backends & EV::embeddable_backends & EV::BACKEND_KQUEUE)
		942	) {
		943	# use kqueue for sockets
		944	$socket_loop = new EV::Loop EV::BACKEND_KQUEUE \| EV::FLAG_NOENV;
		945	}
		946
		947	# use the default loop otherwise
		948	$socket_loop \|\|= EV::default_loop;
		949
		950	=over 4
		951
		952	=item $w = EV::embed $otherloop, $callback
		953
		954	=item $w = EV::embed_ns $otherloop, $callback
		955
		956	=item $w = $loop->embed ($otherloop, $callback)
		957
		958	=item $w = $loop->embed_ns ($otherloop, $callback)
		959
		960	Call the callback when the embedded event loop (C<$otherloop>) has any
		961	I/O activity. The C<$callback> should alwas be specified as C<undef> in
		962	this version of EV, which means the embedded event loop will be managed
		963	automatically.
		964
		965	The C<embed_ns> variant doesn't start (activate) the newly created watcher.
		966
		967	=back
		968
		969
		970	=head1 PERL SIGNALS
		971
		972	While Perl signal handling (C<%SIG>) is not affected by EV, the behaviour
		973	with EV is as the same as any other C library: Perl-signals will only be
		974	handled when Perl runs, which means your signal handler might be invoked
		975	only the next time an event callback is invoked.
		976
		977	The solution is to use EV signal watchers (see C<EV::signal>), which will
		978	ensure proper operations with regards to other event watchers.
		979
		980	If you cannot do this for whatever reason, you can also force a watcher
		981	to be called on every event loop iteration by installing a C<EV::check>
		982	watcher:
		983
		984	my $async_check = EV::check sub { };
		985
		986	This ensures that perl gets into control for a short time to handle any
		987	pending signals, and also ensures (slightly) slower overall operation.
		988
		989	=head1 THREADS
		990
		991	Threads are not supported by this module in any way. Perl pseudo-threads
		992	is evil stuff and must die. As soon as Perl gains real threads I will work
		993	on thread support for it.
		994
		995	=head1 FORK
		996
		997	Most of the "improved" event delivering mechanisms of modern operating
		998	systems have quite a few problems with fork(2) (to put it bluntly: it is
		999	not supported and usually destructive). Libev makes it possible to work
		1000	around this by having a function that recreates the kernel state after
		1001	fork in the child.
		1002
		1003	On non-win32 platforms, this module requires the pthread_atfork
		1004	functionality to do this automatically for you. This function is quite
		1005	buggy on most BSDs, though, so YMMV. The overhead for this is quite
		1006	negligible, because everything the function currently does is set a flag
		1007	that is checked only when the event loop gets used the next time, so when
		1008	you do fork but not use EV, the overhead is minimal.
		1009
		1010	On win32, there is no notion of fork so all this doesn't apply, of course.
43		1011
44	=cut	1012	=cut
45		1013
46	our $NPRI = 4;	1014	our $DIED = sub {
47	our $BASE = init;	1015	warn "EV: error in callback (ignoring): $@";
48	priority_init $NPRI;	1016	};
		1017
		1018	default_loop
		1019	or die 'EV: cannot initialise libev backend. bad $ENV{LIBEV_FLAGS}?';
49		1020
50	1;	1021	1;
		1022
		1023	=head1 SEE ALSO
		1024
		1025	L<EV::ADNS> (asynchronous DNS), L<Glib::EV> (makes Glib/Gtk2 use EV as
		1026	event loop), L<EV::Glib> (embed Glib into EV), L<Coro::EV> (efficient
		1027	coroutines with EV), L<Net::SNMP::EV> (asynchronous SNMP).
51		1028
52	=head1 AUTHOR	1029	=head1 AUTHOR
53		1030
54	Marc Lehmann <schmorp@schmorp.de>	1031	Marc Lehmann <schmorp@schmorp.de>
55	http://home.schmorp.de/	1032	http://home.schmorp.de/

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing EV/EV.pm (file contents): Revision 1.1 by root, Fri Oct 26 16:50:05 2007 UTC vs. Revision 1.79 by root, Sat Dec 22 16:37:07 2007 UTC

Diff Legend

Comparing EV/EV.pm (file contents):
Revision 1.1 by root, Fri Oct 26 16:50:05 2007 UTC vs.
Revision 1.79 by root, Sat Dec 22 16:37:07 2007 UTC