ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/AnyEvent/lib/AnyEvent.pm

Comparing AnyEvent/lib/AnyEvent.pm (file contents):
Revision 1.160 by root, Sun Jun 22 12:17:47 2008 UTC vs.
Revision 1.312 by root, Mon Feb 15 18:02:35 2010 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	AnyEvent - provide framework for multiple event loops	3	AnyEvent - the DBI of event loop programming
4		4
5	EV, Event, Glib, Tk, Perl, Event::Lib, Qt, POE - various supported event loops	5	EV, Event, Glib, Tk, Perl, Event::Lib, Irssi, rxvt-unicode, IO::Async, Qt
		6	and POE are various supported event loops/environments.
6		7
7	=head1 SYNOPSIS	8	=head1 SYNOPSIS
8		9
9	use AnyEvent;	10	use AnyEvent;
10		11
		12	# file descriptor readable
11	my $w = AnyEvent->io (fh => $fh, poll => "r|w", cb => sub {	13	my $w = AnyEvent->io (fh => $fh, poll => "r", cb => sub { ... });
		14
		15	# one-shot or repeating timers
		16	my $w = AnyEvent->timer (after => $seconds, cb => sub { ... });
		17	my $w = AnyEvent->timer (after => $seconds, interval => $seconds, cb => ...
		18
		19	print AnyEvent->now; # prints current event loop time
		20	print AnyEvent->time; # think Time::HiRes::time or simply CORE::time.
		21
		22	# POSIX signal
		23	my $w = AnyEvent->signal (signal => "TERM", cb => sub { ... });
		24
		25	# child process exit
		26	my $w = AnyEvent->child (pid => $pid, cb => sub {
		27	my ($pid, $status) = @_;
12	...	28	...
13	});	29	});
14		30
15	my $w = AnyEvent->timer (after => $seconds, cb => sub {	31	# called when event loop idle (if applicable)
16	...	32	my $w = AnyEvent->idle (cb => sub { ... });
17	});
18		33
19	my $w = AnyEvent->condvar; # stores whether a condition was flagged	34	my $w = AnyEvent->condvar; # stores whether a condition was flagged
20	$w->send; # wake up current and all future recv's	35	$w->send; # wake up current and all future recv's
21	$w->recv; # enters "main loop" till $condvar gets ->send	36	$w->recv; # enters "main loop" till $condvar gets ->send
		37	# use a condvar in callback mode:
		38	$w->cb (sub { $_[0]->recv });
22		39
23	=head1 INTRODUCTION/TUTORIAL	40	=head1 INTRODUCTION/TUTORIAL
24		41
25	This manpage is mainly a reference manual. If you are interested	42	This manpage is mainly a reference manual. If you are interested
26	in a tutorial or some gentle introduction, have a look at the	43	in a tutorial or some gentle introduction, have a look at the
27	L<AnyEvent::Intro> manpage.	44	L<AnyEvent::Intro> manpage.
28		45
		46	=head1 SUPPORT
		47
		48	There is a mailinglist for discussing all things AnyEvent, and an IRC
		49	channel, too.
		50
		51	See the AnyEvent project page at the B<Schmorpforge Ta-Sa Software
		52	Repository>, at L<http://anyevent.schmorp.de>, for more info.
		53
29	=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)	54	=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)
30		55
31	Glib, POE, IO::Async, Event... CPAN offers event models by the dozen	56	Glib, POE, IO::Async, Event... CPAN offers event models by the dozen
32	nowadays. So what is different about AnyEvent?	57	nowadays. So what is different about AnyEvent?
33		58
34	Executive Summary: AnyEvent is I<compatible>, AnyEvent is I<free of	59	Executive Summary: AnyEvent is I<compatible>, AnyEvent is I<free of
35	policy> and AnyEvent is I<small and efficient>.	60	policy> and AnyEvent is I<small and efficient>.
36		61
37	First and foremost, I<AnyEvent is not an event model> itself, it only	62	First and foremost, I<AnyEvent is not an event model> itself, it only
38	interfaces to whatever event model the main program happens to use in a	63	interfaces to whatever event model the main program happens to use, in a
39	pragmatic way. For event models and certain classes of immortals alike,	64	pragmatic way. For event models and certain classes of immortals alike,
40	the statement "there can only be one" is a bitter reality: In general,	65	the statement "there can only be one" is a bitter reality: In general,
41	only one event loop can be active at the same time in a process. AnyEvent	66	only one event loop can be active at the same time in a process. AnyEvent
42	helps hiding the differences between those event loops.	67	cannot change this, but it can hide the differences between those event
		68	loops.
43		69
44	The goal of AnyEvent is to offer module authors the ability to do event	70	The goal of AnyEvent is to offer module authors the ability to do event
45	programming (waiting for I/O or timer events) without subscribing to a	71	programming (waiting for I/O or timer events) without subscribing to a
46	religion, a way of living, and most importantly: without forcing your	72	religion, a way of living, and most importantly: without forcing your
47	module users into the same thing by forcing them to use the same event	73	module users into the same thing by forcing them to use the same event
48	model you use.	74	model you use.
49		75
50	For modules like POE or IO::Async (which is a total misnomer as it is	76	For modules like POE or IO::Async (which is a total misnomer as it is
51	actually doing all I/O I<synchronously>...), using them in your module is	77	actually doing all I/O I<synchronously>...), using them in your module is
52	like joining a cult: After you joined, you are dependent on them and you	78	like joining a cult: After you joined, you are dependent on them and you
53	cannot use anything else, as it is simply incompatible to everything that	79	cannot use anything else, as they are simply incompatible to everything
54	isn't itself. What's worse, all the potential users of your module are	80	that isn't them. What's worse, all the potential users of your
55	I<also> forced to use the same event loop you use.	81	module are I<also> forced to use the same event loop you use.
56		82
57	AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works	83	AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works
58	fine. AnyEvent + Tk works fine etc. etc. but none of these work together	84	fine. AnyEvent + Tk works fine etc. etc. but none of these work together
59	with the rest: POE + IO::Async? No go. Tk + Event? No go. Again: if	85	with the rest: POE + IO::Async? No go. Tk + Event? No go. Again: if
60	your module uses one of those, every user of your module has to use it,	86	your module uses one of those, every user of your module has to use it,
61	too. But if your module uses AnyEvent, it works transparently with all	87	too. But if your module uses AnyEvent, it works transparently with all
62	event models it supports (including stuff like POE and IO::Async, as long	88	event models it supports (including stuff like IO::Async, as long as those
63	as those use one of the supported event loops. It is trivial to add new	89	use one of the supported event loops. It is trivial to add new event loops
64	event loops to AnyEvent, too, so it is future-proof).	90	to AnyEvent, too, so it is future-proof).
65		91
66	In addition to being free of having to use I<the one and only true event	92	In addition to being free of having to use I<the one and only true event
67	model>, AnyEvent also is free of bloat and policy: with POE or similar	93	model>, AnyEvent also is free of bloat and policy: with POE or similar
68	modules, you get an enormous amount of code and strict rules you have to	94	modules, you get an enormous amount of code and strict rules you have to
69	follow. AnyEvent, on the other hand, is lean and up to the point, by only	95	follow. AnyEvent, on the other hand, is lean and up to the point, by only
…		…
127	These watchers are normal Perl objects with normal Perl lifetime. After	153	These watchers are normal Perl objects with normal Perl lifetime. After
128	creating a watcher it will immediately "watch" for events and invoke the	154	creating a watcher it will immediately "watch" for events and invoke the
129	callback when the event occurs (of course, only when the event model	155	callback when the event occurs (of course, only when the event model
130	is in control).	156	is in control).
131		157
		158	Note that B<callbacks must not permanently change global variables>
		159	potentially in use by the event loop (such as C<$_> or C<$[>) and that B<<
		160	callbacks must not C<die> >>. The former is good programming practise in
		161	Perl and the latter stems from the fact that exception handling differs
		162	widely between event loops.
		163
132	To disable the watcher you have to destroy it (e.g. by setting the	164	To disable the watcher you have to destroy it (e.g. by setting the
133	variable you store it in to C<undef> or otherwise deleting all references	165	variable you store it in to C<undef> or otherwise deleting all references
134	to it).	166	to it).
135		167
136	All watchers are created by calling a method on the C<AnyEvent> class.	168	All watchers are created by calling a method on the C<AnyEvent> class.
…		…
149	my variables are only visible after the statement in which they are	181	my variables are only visible after the statement in which they are
150	declared.	182	declared.
151		183
152	=head2 I/O WATCHERS	184	=head2 I/O WATCHERS
153		185
		186	$w = AnyEvent->io (
		187	fh => <filehandle_or_fileno>,
		188	poll => <"r" or "w">,
		189	cb => <callback>,
		190	);
		191
154	You can create an I/O watcher by calling the C<< AnyEvent->io >> method	192	You can create an I/O watcher by calling the C<< AnyEvent->io >> method
155	with the following mandatory key-value pairs as arguments:	193	with the following mandatory key-value pairs as arguments:
156		194
157	C<fh> the Perl I<file handle> (I<not> file descriptor) to watch	195	C<fh> is the Perl I<file handle> (or a naked file descriptor) to watch
		196	for events (AnyEvent might or might not keep a reference to this file
		197	handle). Note that only file handles pointing to things for which
		198	non-blocking operation makes sense are allowed. This includes sockets,
		199	most character devices, pipes, fifos and so on, but not for example files
		200	or block devices.
		201
158	for events. C<poll> must be a string that is either C<r> or C<w>,	202	C<poll> must be a string that is either C<r> or C<w>, which creates a
159	which creates a watcher waiting for "r"eadable or "w"ritable events,	203	watcher waiting for "r"eadable or "w"ritable events, respectively.
		204
160	respectively. C<cb> is the callback to invoke each time the file handle	205	C<cb> is the callback to invoke each time the file handle becomes ready.
161	becomes ready.
162		206
163	Although the callback might get passed parameters, their value and	207	Although the callback might get passed parameters, their value and
164	presence is undefined and you cannot rely on them. Portable AnyEvent	208	presence is undefined and you cannot rely on them. Portable AnyEvent
165	callbacks cannot use arguments passed to I/O watcher callbacks.	209	callbacks cannot use arguments passed to I/O watcher callbacks.
166		210
…		…
170		214
171	Some event loops issue spurious readyness notifications, so you should	215	Some event loops issue spurious readyness notifications, so you should
172	always use non-blocking calls when reading/writing from/to your file	216	always use non-blocking calls when reading/writing from/to your file
173	handles.	217	handles.
174		218
175	Example:
176
177	# wait for readability of STDIN, then read a line and disable the watcher	219	Example: wait for readability of STDIN, then read a line and disable the
		220	watcher.
		221
178	my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {	222	my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
179	chomp (my $input = <STDIN>);	223	chomp (my $input = <STDIN>);
180	warn "read: $input\n";	224	warn "read: $input\n";
181	undef $w;	225	undef $w;
182	});	226	});
183		227
184	=head2 TIME WATCHERS	228	=head2 TIME WATCHERS
185		229
		230	$w = AnyEvent->timer (after => <seconds>, cb => <callback>);
		231
		232	$w = AnyEvent->timer (
		233	after => <fractional_seconds>,
		234	interval => <fractional_seconds>,
		235	cb => <callback>,
		236	);
		237
186	You can create a time watcher by calling the C<< AnyEvent->timer >>	238	You can create a time watcher by calling the C<< AnyEvent->timer >>
187	method with the following mandatory arguments:	239	method with the following mandatory arguments:
188		240
189	C<after> specifies after how many seconds (fractional values are	241	C<after> specifies after how many seconds (fractional values are
190	supported) the callback should be invoked. C<cb> is the callback to invoke	242	supported) the callback should be invoked. C<cb> is the callback to invoke
…		…
192		244
193	Although the callback might get passed parameters, their value and	245	Although the callback might get passed parameters, their value and
194	presence is undefined and you cannot rely on them. Portable AnyEvent	246	presence is undefined and you cannot rely on them. Portable AnyEvent
195	callbacks cannot use arguments passed to time watcher callbacks.	247	callbacks cannot use arguments passed to time watcher callbacks.
196		248
197	The timer callback will be invoked at most once: if you want a repeating	249	The callback will normally be invoked once only. If you specify another
198	timer you have to create a new watcher (this is a limitation by both Tk	250	parameter, C<interval>, as a strictly positive number (> 0), then the
199	and Glib).	251	callback will be invoked regularly at that interval (in fractional
		252	seconds) after the first invocation. If C<interval> is specified with a
		253	false value, then it is treated as if it were missing.
200		254
201	Example:	255	The callback will be rescheduled before invoking the callback, but no
		256	attempt is done to avoid timer drift in most backends, so the interval is
		257	only approximate.
202		258
203	# fire an event after 7.7 seconds	259	Example: fire an event after 7.7 seconds.
		260
204	my $w = AnyEvent->timer (after => 7.7, cb => sub {	261	my $w = AnyEvent->timer (after => 7.7, cb => sub {
205	warn "timeout\n";	262	warn "timeout\n";
206	});	263	});
207		264
208	# to cancel the timer:	265	# to cancel the timer:
209	undef $w;	266	undef $w;
210		267
211	Example 2:
212
213	# fire an event after 0.5 seconds, then roughly every second	268	Example 2: fire an event after 0.5 seconds, then roughly every second.
214	my $w;
215		269
216	my $cb = sub {
217	# cancel the old timer while creating a new one
218	$w = AnyEvent->timer (after => 1, cb => $cb);	270	my $w = AnyEvent->timer (after => 0.5, interval => 1, cb => sub {
		271	warn "timeout\n";
219	};	272	};
220
221	# start the "loop" by creating the first watcher
222	$w = AnyEvent->timer (after => 0.5, cb => $cb);
223		273
224	=head3 TIMING ISSUES	274	=head3 TIMING ISSUES
225		275
226	There are two ways to handle timers: based on real time (relative, "fire	276	There are two ways to handle timers: based on real time (relative, "fire
227	in 10 seconds") and based on wallclock time (absolute, "fire at 12	277	in 10 seconds") and based on wallclock time (absolute, "fire at 12
…		…
300	In either case, if you care (and in most cases, you don't), then you	350	In either case, if you care (and in most cases, you don't), then you
301	can get whatever behaviour you want with any event loop, by taking the	351	can get whatever behaviour you want with any event loop, by taking the
302	difference between C<< AnyEvent->time >> and C<< AnyEvent->now >> into	352	difference between C<< AnyEvent->time >> and C<< AnyEvent->now >> into
303	account.	353	account.
304		354
		355	=item AnyEvent->now_update
		356
		357	Some event loops (such as L<EV> or L<AnyEvent::Impl::Perl>) cache
		358	the current time for each loop iteration (see the discussion of L<<
		359	AnyEvent->now >>, above).
		360
		361	When a callback runs for a long time (or when the process sleeps), then
		362	this "current" time will differ substantially from the real time, which
		363	might affect timers and time-outs.
		364
		365	When this is the case, you can call this method, which will update the
		366	event loop's idea of "current time".
		367
		368	A typical example would be a script in a web server (e.g. C<mod_perl>) -
		369	when mod_perl executes the script, then the event loop will have the wrong
		370	idea about the "current time" (being potentially far in the past, when the
		371	script ran the last time). In that case you should arrange a call to C<<
		372	AnyEvent->now_update >> each time the web server process wakes up again
		373	(e.g. at the start of your script, or in a handler).
		374
		375	Note that updating the time I<might> cause some events to be handled.
		376
305	=back	377	=back
306		378
307	=head2 SIGNAL WATCHERS	379	=head2 SIGNAL WATCHERS
308		380
		381	$w = AnyEvent->signal (signal => <uppercase_signal_name>, cb => <callback>);
		382
309	You can watch for signals using a signal watcher, C<signal> is the signal	383	You can watch for signals using a signal watcher, C<signal> is the signal
310	I<name> without any C<SIG> prefix, C<cb> is the Perl callback to	384	I<name> in uppercase and without any C<SIG> prefix, C<cb> is the Perl
311	be invoked whenever a signal occurs.	385	callback to be invoked whenever a signal occurs.
312		386
313	Although the callback might get passed parameters, their value and	387	Although the callback might get passed parameters, their value and
314	presence is undefined and you cannot rely on them. Portable AnyEvent	388	presence is undefined and you cannot rely on them. Portable AnyEvent
315	callbacks cannot use arguments passed to signal watcher callbacks.	389	callbacks cannot use arguments passed to signal watcher callbacks.
316		390
…		…
318	invocation, and callback invocation will be synchronous. Synchronous means	392	invocation, and callback invocation will be synchronous. Synchronous means
319	that it might take a while until the signal gets handled by the process,	393	that it might take a while until the signal gets handled by the process,
320	but it is guaranteed not to interrupt any other callbacks.	394	but it is guaranteed not to interrupt any other callbacks.
321		395
322	The main advantage of using these watchers is that you can share a signal	396	The main advantage of using these watchers is that you can share a signal
323	between multiple watchers.	397	between multiple watchers, and AnyEvent will ensure that signals will not
		398	interrupt your program at bad times.
324		399
325	This watcher might use C<%SIG>, so programs overwriting those signals	400	This watcher might use C<%SIG> (depending on the event loop used),
326	directly will likely not work correctly.	401	so programs overwriting those signals directly will likely not work
		402	correctly.
327		403
328	Example: exit on SIGINT	404	Example: exit on SIGINT
329		405
330	my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 });	406	my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 });
331		407
		408	=head3 Restart Behaviour
		409
		410	While restart behaviour is up to the event loop implementation, most will
		411	not restart syscalls (that includes L<Async::Interrupt> and AnyEvent's
		412	pure perl implementation).
		413
		414	=head3 Safe/Unsafe Signals
		415
		416	Perl signals can be either "safe" (synchronous to opcode handling) or
		417	"unsafe" (asynchronous) - the former might get delayed indefinitely, the
		418	latter might corrupt your memory.
		419
		420	AnyEvent signal handlers are, in addition, synchronous to the event loop,
		421	i.e. they will not interrupt your running perl program but will only be
		422	called as part of the normal event handling (just like timer, I/O etc.
		423	callbacks, too).
		424
		425	=head3 Signal Races, Delays and Workarounds
		426
		427	Many event loops (e.g. Glib, Tk, Qt, IO::Async) do not support attaching
		428	callbacks to signals in a generic way, which is a pity, as you cannot
		429	do race-free signal handling in perl, requiring C libraries for
		430	this. AnyEvent will try to do it's best, which means in some cases,
		431	signals will be delayed. The maximum time a signal might be delayed is
		432	specified in C<$AnyEvent::MAX_SIGNAL_LATENCY> (default: 10 seconds). This
		433	variable can be changed only before the first signal watcher is created,
		434	and should be left alone otherwise. This variable determines how often
		435	AnyEvent polls for signals (in case a wake-up was missed). Higher values
		436	will cause fewer spurious wake-ups, which is better for power and CPU
		437	saving.
		438
		439	All these problems can be avoided by installing the optional
		440	L<Async::Interrupt> module, which works with most event loops. It will not
		441	work with inherently broken event loops such as L<Event> or L<Event::Lib>
		442	(and not with L<POE> currently, as POE does it's own workaround with
		443	one-second latency). For those, you just have to suffer the delays.
		444
332	=head2 CHILD PROCESS WATCHERS	445	=head2 CHILD PROCESS WATCHERS
333		446
		447	$w = AnyEvent->child (pid => <process id>, cb => <callback>);
		448
334	You can also watch on a child process exit and catch its exit status.	449	You can also watch on a child process exit and catch its exit status.
335		450
336	The child process is specified by the C<pid> argument (if set to C<0>, it	451	The child process is specified by the C<pid> argument (one some backends,
337	watches for any child process exit). The watcher will trigger as often	452	using C<0> watches for any child process exit, on others this will
338	as status change for the child are received. This works by installing a	453	croak). The watcher will be triggered only when the child process has
339	signal handler for C<SIGCHLD>. The callback will be called with the pid	454	finished and an exit status is available, not on any trace events
340	and exit status (as returned by waitpid), so unlike other watcher types,	455	(stopped/continued).
341	you I<can> rely on child watcher callback arguments.	456
		457	The callback will be called with the pid and exit status (as returned by
		458	waitpid), so unlike other watcher types, you I<can> rely on child watcher
		459	callback arguments.
		460
		461	This watcher type works by installing a signal handler for C<SIGCHLD>,
		462	and since it cannot be shared, nothing else should use SIGCHLD or reap
		463	random child processes (waiting for specific child processes, e.g. inside
		464	C<system>, is just fine).
342		465
343	There is a slight catch to child watchers, however: you usually start them	466	There is a slight catch to child watchers, however: you usually start them
344	I<after> the child process was created, and this means the process could	467	I<after> the child process was created, and this means the process could
345	have exited already (and no SIGCHLD will be sent anymore).	468	have exited already (and no SIGCHLD will be sent anymore).
346		469
347	Not all event models handle this correctly (POE doesn't), but even for	470	Not all event models handle this correctly (neither POE nor IO::Async do,
		471	see their AnyEvent::Impl manpages for details), but even for event models
348	event models that I<do> handle this correctly, they usually need to be	472	that I<do> handle this correctly, they usually need to be loaded before
349	loaded before the process exits (i.e. before you fork in the first place).	473	the process exits (i.e. before you fork in the first place). AnyEvent's
		474	pure perl event loop handles all cases correctly regardless of when you
		475	start the watcher.
350		476
351	This means you cannot create a child watcher as the very first thing in an	477	This means you cannot create a child watcher as the very first
352	AnyEvent program, you I<have> to create at least one watcher before you	478	thing in an AnyEvent program, you I<have> to create at least one
353	C<fork> the child (alternatively, you can call C<AnyEvent::detect>).	479	watcher before you C<fork> the child (alternatively, you can call
		480	C<AnyEvent::detect>).
		481
		482	As most event loops do not support waiting for child events, they will be
		483	emulated by AnyEvent in most cases, in which the latency and race problems
		484	mentioned in the description of signal watchers apply.
354		485
355	Example: fork a process and wait for it	486	Example: fork a process and wait for it
356		487
357	my $done = AnyEvent->condvar;	488	my $done = AnyEvent->condvar;
358		489
…		…
368	);	499	);
369		500
370	# do something else, then wait for process exit	501	# do something else, then wait for process exit
371	$done->recv;	502	$done->recv;
372		503
		504	=head2 IDLE WATCHERS
		505
		506	$w = AnyEvent->idle (cb => <callback>);
		507
		508	Repeatedly invoke the callback after the process becomes idle, until
		509	either the watcher is destroyed or new events have been detected.
		510
		511	Idle watchers are useful when there is a need to do something, but it
		512	is not so important (or wise) to do it instantly. The callback will be
		513	invoked only when there is "nothing better to do", which is usually
		514	defined as "all outstanding events have been handled and no new events
		515	have been detected". That means that idle watchers ideally get invoked
		516	when the event loop has just polled for new events but none have been
		517	detected. Instead of blocking to wait for more events, the idle watchers
		518	will be invoked.
		519
		520	Unfortunately, most event loops do not really support idle watchers (only
		521	EV, Event and Glib do it in a usable fashion) - for the rest, AnyEvent
		522	will simply call the callback "from time to time".
		523
		524	Example: read lines from STDIN, but only process them when the
		525	program is otherwise idle:
		526
		527	my @lines; # read data
		528	my $idle_w;
		529	my $io_w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
		530	push @lines, scalar <STDIN>;
		531
		532	# start an idle watcher, if not already done
		533	$idle_w ||= AnyEvent->idle (cb => sub {
		534	# handle only one line, when there are lines left
		535	if (my $line = shift @lines) {
		536	print "handled when idle: $line";
		537	} else {
		538	# otherwise disable the idle watcher again
		539	undef $idle_w;
		540	}
		541	});
		542	});
		543
373	=head2 CONDITION VARIABLES	544	=head2 CONDITION VARIABLES
		545
		546	$cv = AnyEvent->condvar;
		547
		548	$cv->send (<list>);
		549	my @res = $cv->recv;
374		550
375	If you are familiar with some event loops you will know that all of them	551	If you are familiar with some event loops you will know that all of them
376	require you to run some blocking "loop", "run" or similar function that	552	require you to run some blocking "loop", "run" or similar function that
377	will actively watch for new events and call your callbacks.	553	will actively watch for new events and call your callbacks.
378		554
379	AnyEvent is different, it expects somebody else to run the event loop and	555	AnyEvent is slightly different: it expects somebody else to run the event
380	will only block when necessary (usually when told by the user).	556	loop and will only block when necessary (usually when told by the user).
381		557
382	The instrument to do that is called a "condition variable", so called	558	The instrument to do that is called a "condition variable", so called
383	because they represent a condition that must become true.	559	because they represent a condition that must become true.
		560
		561	Now is probably a good time to look at the examples further below.
384		562
385	Condition variables can be created by calling the C<< AnyEvent->condvar	563	Condition variables can be created by calling the C<< AnyEvent->condvar
386	>> method, usually without arguments. The only argument pair allowed is	564	>> method, usually without arguments. The only argument pair allowed is
387	C<cb>, which specifies a callback to be called when the condition variable	565	C<cb>, which specifies a callback to be called when the condition variable
388	becomes true.	566	becomes true, with the condition variable as the first argument (but not
		567	the results).
389		568
390	After creation, the condition variable is "false" until it becomes "true"	569	After creation, the condition variable is "false" until it becomes "true"
391	by calling the C<send> method (or calling the condition variable as if it	570	by calling the C<send> method (or calling the condition variable as if it
392	were a callback, read about the caveats in the description for the C<<	571	were a callback, read about the caveats in the description for the C<<
393	->send >> method).	572	->send >> method).
…		…
395	Condition variables are similar to callbacks, except that you can	574	Condition variables are similar to callbacks, except that you can
396	optionally wait for them. They can also be called merge points - points	575	optionally wait for them. They can also be called merge points - points
397	in time where multiple outstanding events have been processed. And yet	576	in time where multiple outstanding events have been processed. And yet
398	another way to call them is transactions - each condition variable can be	577	another way to call them is transactions - each condition variable can be
399	used to represent a transaction, which finishes at some point and delivers	578	used to represent a transaction, which finishes at some point and delivers
400	a result.	579	a result. And yet some people know them as "futures" - a promise to
		580	compute/deliver something that you can wait for.
401		581
402	Condition variables are very useful to signal that something has finished,	582	Condition variables are very useful to signal that something has finished,
403	for example, if you write a module that does asynchronous http requests,	583	for example, if you write a module that does asynchronous http requests,
404	then a condition variable would be the ideal candidate to signal the	584	then a condition variable would be the ideal candidate to signal the
405	availability of results. The user can either act when the callback is	585	availability of results. The user can either act when the callback is
…		…
439	after => 1,	619	after => 1,
440	cb => sub { $result_ready->send },	620	cb => sub { $result_ready->send },
441	);	621	);
442		622
443	# this "blocks" (while handling events) till the callback	623	# this "blocks" (while handling events) till the callback
444	# calls send	624	# calls ->send
445	$result_ready->recv;	625	$result_ready->recv;
446		626
447	Example: wait for a timer, but take advantage of the fact that	627	Example: wait for a timer, but take advantage of the fact that condition
448	condition variables are also code references.	628	variables are also callable directly.
449		629
450	my $done = AnyEvent->condvar;	630	my $done = AnyEvent->condvar;
451	my $delay = AnyEvent->timer (after => 5, cb => $done);	631	my $delay = AnyEvent->timer (after => 5, cb => $done);
452	$done->recv;	632	$done->recv;
		633
		634	Example: Imagine an API that returns a condvar and doesn't support
		635	callbacks. This is how you make a synchronous call, for example from
		636	the main program:
		637
		638	use AnyEvent::CouchDB;
		639
		640	...
		641
		642	my @info = $couchdb->info->recv;
		643
		644	And this is how you would just set a callback to be called whenever the
		645	results are available:
		646
		647	$couchdb->info->cb (sub {
		648	my @info = $_[0]->recv;
		649	});
453		650
454	=head3 METHODS FOR PRODUCERS	651	=head3 METHODS FOR PRODUCERS
455		652
456	These methods should only be used by the producing side, i.e. the	653	These methods should only be used by the producing side, i.e. the
457	code/module that eventually sends the signal. Note that it is also	654	code/module that eventually sends the signal. Note that it is also
…		…
470	immediately from within send.	667	immediately from within send.
471		668
472	Any arguments passed to the C<send> call will be returned by all	669	Any arguments passed to the C<send> call will be returned by all
473	future C<< ->recv >> calls.	670	future C<< ->recv >> calls.
474		671
475	Condition variables are overloaded so one can call them directly	672	Condition variables are overloaded so one can call them directly (as if
476	(as a code reference). Calling them directly is the same as calling	673	they were a code reference). Calling them directly is the same as calling
477	C<send>. Note, however, that many C-based event loops do not handle	674	C<send>.
478	overloading, so as tempting as it may be, passing a condition variable
479	instead of a callback does not work. Both the pure perl and EV loops
480	support overloading, however, as well as all functions that use perl to
481	invoke a callback (as in L<AnyEvent::Socket> and L<AnyEvent::DNS> for
482	example).
483		675
484	=item $cv->croak ($error)	676	=item $cv->croak ($error)
485		677
486	Similar to send, but causes all call's to C<< ->recv >> to invoke	678	Similar to send, but causes all call's to C<< ->recv >> to invoke
487	C<Carp::croak> with the given error message/object/scalar.	679	C<Carp::croak> with the given error message/object/scalar.
488		680
489	This can be used to signal any errors to the condition variable	681	This can be used to signal any errors to the condition variable
490	user/consumer.	682	user/consumer. Doing it this way instead of calling C<croak> directly
		683	delays the error detetcion, but has the overwhelmign advantage that it
		684	diagnoses the error at the place where the result is expected, and not
		685	deep in some event clalback without connection to the actual code causing
		686	the problem.
491		687
492	=item $cv->begin ([group callback])	688	=item $cv->begin ([group callback])
493		689
494	=item $cv->end	690	=item $cv->end
495
496	These two methods are EXPERIMENTAL and MIGHT CHANGE.
497		691
498	These two methods can be used to combine many transactions/events into	692	These two methods can be used to combine many transactions/events into
499	one. For example, a function that pings many hosts in parallel might want	693	one. For example, a function that pings many hosts in parallel might want
500	to use a condition variable for the whole process.	694	to use a condition variable for the whole process.
501		695
502	Every call to C<< ->begin >> will increment a counter, and every call to	696	Every call to C<< ->begin >> will increment a counter, and every call to
503	C<< ->end >> will decrement it. If the counter reaches C<0> in C<< ->end	697	C<< ->end >> will decrement it. If the counter reaches C<0> in C<< ->end
504	>>, the (last) callback passed to C<begin> will be executed. That callback	698	>>, the (last) callback passed to C<begin> will be executed, passing the
505	is I<supposed> to call C<< ->send >>, but that is not required. If no	699	condvar as first argument. That callback is I<supposed> to call C<< ->send
506	callback was set, C<send> will be called without any arguments.	700	>>, but that is not required. If no group callback was set, C<send> will
		701	be called without any arguments.
507		702
508	Let's clarify this with the ping example:	703	You can think of C<< $cv->send >> giving you an OR condition (one call
		704	sends), while C<< $cv->begin >> and C<< $cv->end >> giving you an AND
		705	condition (all C<begin> calls must be C<end>'ed before the condvar sends).
		706
		707	Let's start with a simple example: you have two I/O watchers (for example,
		708	STDOUT and STDERR for a program), and you want to wait for both streams to
		709	close before activating a condvar:
509		710
510	my $cv = AnyEvent->condvar;	711	my $cv = AnyEvent->condvar;
511		712
		713	$cv->begin; # first watcher
		714	my $w1 = AnyEvent->io (fh => $fh1, cb => sub {
		715	defined sysread $fh1, my $buf, 4096
		716	or $cv->end;
		717	});
		718
		719	$cv->begin; # second watcher
		720	my $w2 = AnyEvent->io (fh => $fh2, cb => sub {
		721	defined sysread $fh2, my $buf, 4096
		722	or $cv->end;
		723	});
		724
		725	$cv->recv;
		726
		727	This works because for every event source (EOF on file handle), there is
		728	one call to C<begin>, so the condvar waits for all calls to C<end> before
		729	sending.
		730
		731	The ping example mentioned above is slightly more complicated, as the
		732	there are results to be passwd back, and the number of tasks that are
		733	begung can potentially be zero:
		734
		735	my $cv = AnyEvent->condvar;
		736
512	my %result;	737	my %result;
513	$cv->begin (sub { $cv->send (\%result) });	738	$cv->begin (sub { shift->send (\%result) });
514		739
515	for my $host (@list_of_hosts) {	740	for my $host (@list_of_hosts) {
516	$cv->begin;	741	$cv->begin;
517	ping_host_then_call_callback $host, sub {	742	ping_host_then_call_callback $host, sub {
518	$result{$host} = ...;	743	$result{$host} = ...;
…		…
533	loop, which serves two important purposes: first, it sets the callback	758	loop, which serves two important purposes: first, it sets the callback
534	to be called once the counter reaches C<0>, and second, it ensures that	759	to be called once the counter reaches C<0>, and second, it ensures that
535	C<send> is called even when C<no> hosts are being pinged (the loop	760	C<send> is called even when C<no> hosts are being pinged (the loop
536	doesn't execute once).	761	doesn't execute once).
537		762
538	This is the general pattern when you "fan out" into multiple subrequests:	763	This is the general pattern when you "fan out" into multiple (but
539	use an outer C<begin>/C<end> pair to set the callback and ensure C<end>	764	potentially none) subrequests: use an outer C<begin>/C<end> pair to set
540	is called at least once, and then, for each subrequest you start, call	765	the callback and ensure C<end> is called at least once, and then, for each
541	C<begin> and for each subrequest you finish, call C<end>.	766	subrequest you start, call C<begin> and for each subrequest you finish,
		767	call C<end>.
542		768
543	=back	769	=back
544		770
545	=head3 METHODS FOR CONSUMERS	771	=head3 METHODS FOR CONSUMERS
546		772
…		…
562	function will call C<croak>.	788	function will call C<croak>.
563		789
564	In list context, all parameters passed to C<send> will be returned,	790	In list context, all parameters passed to C<send> will be returned,
565	in scalar context only the first one will be returned.	791	in scalar context only the first one will be returned.
566		792
		793	Note that doing a blocking wait in a callback is not supported by any
		794	event loop, that is, recursive invocation of a blocking C<< ->recv
		795	>> is not allowed, and the C<recv> call will C<croak> if such a
		796	condition is detected. This condition can be slightly loosened by using
		797	L<Coro::AnyEvent>, which allows you to do a blocking C<< ->recv >> from
		798	any thread that doesn't run the event loop itself.
		799
567	Not all event models support a blocking wait - some die in that case	800	Not all event models support a blocking wait - some die in that case
568	(programs might want to do that to stay interactive), so I<if you are	801	(programs might want to do that to stay interactive), so I<if you are
569	using this from a module, never require a blocking wait>, but let the	802	using this from a module, never require a blocking wait>. Instead, let the
570	caller decide whether the call will block or not (for example, by coupling	803	caller decide whether the call will block or not (for example, by coupling
571	condition variables with some kind of request results and supporting	804	condition variables with some kind of request results and supporting
572	callbacks so the caller knows that getting the result will not block,	805	callbacks so the caller knows that getting the result will not block,
573	while still supporting blocking waits if the caller so desires).	806	while still supporting blocking waits if the caller so desires).
574		807
575	Another reason I<never> to C<< ->recv >> in a module is that you cannot
576	sensibly have two C<< ->recv >>'s in parallel, as that would require
577	multiple interpreters or coroutines/threads, none of which C<AnyEvent>
578	can supply.
579
580	The L<Coro> module, however, I<can> and I<does> supply coroutines and, in
581	fact, L<Coro::AnyEvent> replaces AnyEvent's condvars by coroutine-safe
582	versions and also integrates coroutines into AnyEvent, making blocking
583	C<< ->recv >> calls perfectly safe as long as they are done from another
584	coroutine (one that doesn't run the event loop).
585
586	You can ensure that C<< -recv >> never blocks by setting a callback and	808	You can ensure that C<< -recv >> never blocks by setting a callback and
587	only calling C<< ->recv >> from within that callback (or at a later	809	only calling C<< ->recv >> from within that callback (or at a later
588	time). This will work even when the event loop does not support blocking	810	time). This will work even when the event loop does not support blocking
589	waits otherwise.	811	waits otherwise.
590		812
591	=item $bool = $cv->ready	813	=item $bool = $cv->ready
592		814
593	Returns true when the condition is "true", i.e. whether C<send> or	815	Returns true when the condition is "true", i.e. whether C<send> or
594	C<croak> have been called.	816	C<croak> have been called.
595		817
596	=item $cb = $cv->cb ([new callback])	818	=item $cb = $cv->cb ($cb->($cv))
597		819
598	This is a mutator function that returns the callback set and optionally	820	This is a mutator function that returns the callback set and optionally
599	replaces it before doing so.	821	replaces it before doing so.
600		822
601	The callback will be called when the condition becomes "true", i.e. when	823	The callback will be called when the condition becomes (or already was)
602	C<send> or C<croak> are called, with the only argument being the condition	824	"true", i.e. when C<send> or C<croak> are called (or were called), with
603	variable itself. Calling C<recv> inside the callback or at any later time	825	the only argument being the condition variable itself. Calling C<recv>
604	is guaranteed not to block.	826	inside the callback or at any later time is guaranteed not to block.
605		827
606	=back	828	=back
607		829
		830	=head1 SUPPORTED EVENT LOOPS/BACKENDS
		831
		832	The available backend classes are (every class has its own manpage):
		833
		834	=over 4
		835
		836	=item Backends that are autoprobed when no other event loop can be found.
		837
		838	EV is the preferred backend when no other event loop seems to be in
		839	use. If EV is not installed, then AnyEvent will fall back to its own
		840	pure-perl implementation, which is available everywhere as it comes with
		841	AnyEvent itself.
		842
		843	AnyEvent::Impl::EV based on EV (interface to libev, best choice).
		844	AnyEvent::Impl::Perl pure-perl implementation, fast and portable.
		845
		846	=item Backends that are transparently being picked up when they are used.
		847
		848	These will be used when they are currently loaded when the first watcher
		849	is created, in which case it is assumed that the application is using
		850	them. This means that AnyEvent will automatically pick the right backend
		851	when the main program loads an event module before anything starts to
		852	create watchers. Nothing special needs to be done by the main program.
		853
		854	AnyEvent::Impl::Event based on Event, very stable, few glitches.
		855	AnyEvent::Impl::Glib based on Glib, slow but very stable.
		856	AnyEvent::Impl::Tk based on Tk, very broken.
		857	AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
		858	AnyEvent::Impl::POE based on POE, very slow, some limitations.
		859	AnyEvent::Impl::Irssi used when running within irssi.
		860
		861	=item Backends with special needs.
		862
		863	Qt requires the Qt::Application to be instantiated first, but will
		864	otherwise be picked up automatically. As long as the main program
		865	instantiates the application before any AnyEvent watchers are created,
		866	everything should just work.
		867
		868	AnyEvent::Impl::Qt based on Qt.
		869
		870	Support for IO::Async can only be partial, as it is too broken and
		871	architecturally limited to even support the AnyEvent API. It also
		872	is the only event loop that needs the loop to be set explicitly, so
		873	it can only be used by a main program knowing about AnyEvent. See
		874	L<AnyEvent::Impl::Async> for the gory details.
		875
		876	AnyEvent::Impl::IOAsync based on IO::Async, cannot be autoprobed.
		877
		878	=item Event loops that are indirectly supported via other backends.
		879
		880	Some event loops can be supported via other modules:
		881
		882	There is no direct support for WxWidgets (L<Wx>) or L<Prima>.
		883
		884	B<WxWidgets> has no support for watching file handles. However, you can
		885	use WxWidgets through the POE adaptor, as POE has a Wx backend that simply
		886	polls 20 times per second, which was considered to be too horrible to even
		887	consider for AnyEvent.
		888
		889	B<Prima> is not supported as nobody seems to be using it, but it has a POE
		890	backend, so it can be supported through POE.
		891
		892	AnyEvent knows about both L<Prima> and L<Wx>, however, and will try to
		893	load L<POE> when detecting them, in the hope that POE will pick them up,
		894	in which case everything will be automatic.
		895
		896	=back
		897
608	=head1 GLOBAL VARIABLES AND FUNCTIONS	898	=head1 GLOBAL VARIABLES AND FUNCTIONS
609		899
		900	These are not normally required to use AnyEvent, but can be useful to
		901	write AnyEvent extension modules.
		902
610	=over 4	903	=over 4
611		904
612	=item $AnyEvent::MODEL	905	=item $AnyEvent::MODEL
613		906
614	Contains C<undef> until the first watcher is being created. Then it	907	Contains C<undef> until the first watcher is being created, before the
		908	backend has been autodetected.
		909
615	contains the event model that is being used, which is the name of the	910	Afterwards it contains the event model that is being used, which is the
616	Perl class implementing the model. This class is usually one of the	911	name of the Perl class implementing the model. This class is usually one
617	C<AnyEvent::Impl:xxx> modules, but can be any other class in the case	912	of the C<AnyEvent::Impl:xxx> modules, but can be any other class in the
618	AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode>).	913	case AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode> it
619		914	will be C<urxvt::anyevent>).
620	The known classes so far are:
621
622	AnyEvent::Impl::EV based on EV (an interface to libev, best choice).
623	AnyEvent::Impl::Event based on Event, second best choice.
624	AnyEvent::Impl::Perl pure-perl implementation, fast and portable.
625	AnyEvent::Impl::Glib based on Glib, third-best choice.
626	AnyEvent::Impl::Tk based on Tk, very bad choice.
627	AnyEvent::Impl::Qt based on Qt, cannot be autoprobed (see its docs).
628	AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
629	AnyEvent::Impl::POE based on POE, not generic enough for full support.
630
631	There is no support for WxWidgets, as WxWidgets has no support for
632	watching file handles. However, you can use WxWidgets through the
633	POE Adaptor, as POE has a Wx backend that simply polls 20 times per
634	second, which was considered to be too horrible to even consider for
635	AnyEvent. Likewise, other POE backends can be used by AnyEvent by using
636	it's adaptor.
637
638	AnyEvent knows about L<Prima> and L<Wx> and will try to use L<POE> when
639	autodetecting them.
640		915
641	=item AnyEvent::detect	916	=item AnyEvent::detect
642		917
643	Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model	918	Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model
644	if necessary. You should only call this function right before you would	919	if necessary. You should only call this function right before you would
645	have created an AnyEvent watcher anyway, that is, as late as possible at	920	have created an AnyEvent watcher anyway, that is, as late as possible at
646	runtime.	921	runtime, and not e.g. while initialising of your module.
		922
		923	If you need to do some initialisation before AnyEvent watchers are
		924	created, use C<post_detect>.
647		925
648	=item $guard = AnyEvent::post_detect { BLOCK }	926	=item $guard = AnyEvent::post_detect { BLOCK }
649		927
650	Arranges for the code block to be executed as soon as the event model is	928	Arranges for the code block to be executed as soon as the event model is
651	autodetected (or immediately if this has already happened).	929	autodetected (or immediately if this has already happened).
652		930
		931	The block will be executed I<after> the actual backend has been detected
		932	(C<$AnyEvent::MODEL> is set), but I<before> any watchers have been
		933	created, so it is possible to e.g. patch C<@AnyEvent::ISA> or do
		934	other initialisations - see the sources of L<AnyEvent::Strict> or
		935	L<AnyEvent::AIO> to see how this is used.
		936
		937	The most common usage is to create some global watchers, without forcing
		938	event module detection too early, for example, L<AnyEvent::AIO> creates
		939	and installs the global L<IO::AIO> watcher in a C<post_detect> block to
		940	avoid autodetecting the event module at load time.
		941
653	If called in scalar or list context, then it creates and returns an object	942	If called in scalar or list context, then it creates and returns an object
654	that automatically removes the callback again when it is destroyed. See	943	that automatically removes the callback again when it is destroyed (or
		944	C<undef> when the hook was immediately executed). See L<AnyEvent::AIO> for
655	L<Coro::BDB> for a case where this is useful.	945	a case where this is useful.
		946
		947	Example: Create a watcher for the IO::AIO module and store it in
		948	C<$WATCHER>. Only do so after the event loop is initialised, though.
		949
		950	our WATCHER;
		951
		952	my $guard = AnyEvent::post_detect {
		953	$WATCHER = AnyEvent->io (fh => IO::AIO::poll_fileno, poll => 'r', cb => \&IO::AIO::poll_cb);
		954	};
		955
		956	# the ||= is important in case post_detect immediately runs the block,
		957	# as to not clobber the newly-created watcher. assigning both watcher and
		958	# post_detect guard to the same variable has the advantage of users being
		959	# able to just C<undef $WATCHER> if the watcher causes them grief.
		960
		961	$WATCHER ||= $guard;
656		962
657	=item @AnyEvent::post_detect	963	=item @AnyEvent::post_detect
658		964
659	If there are any code references in this array (you can C<push> to it	965	If there are any code references in this array (you can C<push> to it
660	before or after loading AnyEvent), then they will called directly after	966	before or after loading AnyEvent), then they will called directly after
661	the event loop has been chosen.	967	the event loop has been chosen.
662		968
663	You should check C<$AnyEvent::MODEL> before adding to this array, though:	969	You should check C<$AnyEvent::MODEL> before adding to this array, though:
664	if it contains a true value then the event loop has already been detected,	970	if it is defined then the event loop has already been detected, and the
665	and the array will be ignored.	971	array will be ignored.
666		972
667	Best use C<AnyEvent::post_detect { BLOCK }> instead.	973	Best use C<AnyEvent::post_detect { BLOCK }> when your application allows
		974	it, as it takes care of these details.
		975
		976	This variable is mainly useful for modules that can do something useful
		977	when AnyEvent is used and thus want to know when it is initialised, but do
		978	not need to even load it by default. This array provides the means to hook
		979	into AnyEvent passively, without loading it.
		980
		981	Example: To load Coro::AnyEvent whenever Coro and AnyEvent are used
		982	together, you could put this into Coro (this is the actual code used by
		983	Coro to accomplish this):
		984
		985	if (defined $AnyEvent::MODEL) {
		986	# AnyEvent already initialised, so load Coro::AnyEvent
		987	require Coro::AnyEvent;
		988	} else {
		989	# AnyEvent not yet initialised, so make sure to load Coro::AnyEvent
		990	# as soon as it is
		991	push @AnyEvent::post_detect, sub { require Coro::AnyEvent };
		992	}
668		993
669	=back	994	=back
670		995
671	=head1 WHAT TO DO IN A MODULE	996	=head1 WHAT TO DO IN A MODULE
672		997
…		…
727		1052
728		1053
729	=head1 OTHER MODULES	1054	=head1 OTHER MODULES
730		1055
731	The following is a non-exhaustive list of additional modules that use	1056	The following is a non-exhaustive list of additional modules that use
732	AnyEvent and can therefore be mixed easily with other AnyEvent modules	1057	AnyEvent as a client and can therefore be mixed easily with other AnyEvent
733	in the same program. Some of the modules come with AnyEvent, some are	1058	modules and other event loops in the same program. Some of the modules
734	available via CPAN.	1059	come with AnyEvent, most are available via CPAN.
735		1060
736	=over 4	1061	=over 4
737		1062
738	=item L<AnyEvent::Util>	1063	=item L<AnyEvent::Util>
739		1064
740	Contains various utility functions that replace often-used but blocking	1065	Contains various utility functions that replace often-used but blocking
741	functions such as C<inet_aton> by event-/callback-based versions.	1066	functions such as C<inet_aton> by event-/callback-based versions.
742
743	=item L<AnyEvent::Handle>
744
745	Provide read and write buffers and manages watchers for reads and writes.
746		1067
747	=item L<AnyEvent::Socket>	1068	=item L<AnyEvent::Socket>
748		1069
749	Provides various utility functions for (internet protocol) sockets,	1070	Provides various utility functions for (internet protocol) sockets,
750	addresses and name resolution. Also functions to create non-blocking tcp	1071	addresses and name resolution. Also functions to create non-blocking tcp
751	connections or tcp servers, with IPv6 and SRV record support and more.	1072	connections or tcp servers, with IPv6 and SRV record support and more.
752		1073
		1074	=item L<AnyEvent::Handle>
		1075
		1076	Provide read and write buffers, manages watchers for reads and writes,
		1077	supports raw and formatted I/O, I/O queued and fully transparent and
		1078	non-blocking SSL/TLS (via L<AnyEvent::TLS>.
		1079
753	=item L<AnyEvent::DNS>	1080	=item L<AnyEvent::DNS>
754		1081
755	Provides rich asynchronous DNS resolver capabilities.	1082	Provides rich asynchronous DNS resolver capabilities.
756		1083
757	=item L<AnyEvent::HTTP>	1084	=item L<AnyEvent::HTTP>
…		…
767		1094
768	The fastest ping in the west.	1095	The fastest ping in the west.
769		1096
770	=item L<AnyEvent::DBI>	1097	=item L<AnyEvent::DBI>
771		1098
772	Executes DBI requests asynchronously in a proxy process.	1099	Executes L<DBI> requests asynchronously in a proxy process.
773		1100
		1101	=item L<AnyEvent::AIO>
		1102
		1103	Truly asynchronous I/O, should be in the toolbox of every event
		1104	programmer. AnyEvent::AIO transparently fuses L<IO::AIO> and AnyEvent
		1105	together.
		1106
		1107	=item L<AnyEvent::BDB>
		1108
		1109	Truly asynchronous Berkeley DB access. AnyEvent::BDB transparently fuses
		1110	L<BDB> and AnyEvent together.
		1111
		1112	=item L<AnyEvent::GPSD>
		1113
		1114	A non-blocking interface to gpsd, a daemon delivering GPS information.
		1115
774	=item L<Net::IRC3>	1116	=item L<AnyEvent::IRC>
775		1117
776	AnyEvent based IRC client module family.	1118	AnyEvent based IRC client module family (replacing the older Net::IRC3).
777		1119
778	=item L<Net::XMPP2>	1120	=item L<AnyEvent::XMPP>
779		1121
780	AnyEvent based XMPP (Jabber protocol) module family.	1122	AnyEvent based XMPP (Jabber protocol) module family (replacing the older
		1123	Net::XMPP2>.
		1124
		1125	=item L<AnyEvent::IGS>
		1126
		1127	A non-blocking interface to the Internet Go Server protocol (used by
		1128	L<App::IGS>).
781		1129
782	=item L<Net::FCP>	1130	=item L<Net::FCP>
783		1131
784	AnyEvent-based implementation of the Freenet Client Protocol, birthplace	1132	AnyEvent-based implementation of the Freenet Client Protocol, birthplace
785	of AnyEvent.	1133	of AnyEvent.
…		…
790		1138
791	=item L<Coro>	1139	=item L<Coro>
792		1140
793	Has special support for AnyEvent via L<Coro::AnyEvent>.	1141	Has special support for AnyEvent via L<Coro::AnyEvent>.
794		1142
795	=item L<AnyEvent::AIO>, L<IO::AIO>
796
797	Truly asynchronous I/O, should be in the toolbox of every event
798	programmer. AnyEvent::AIO transparently fuses IO::AIO and AnyEvent
799	together.
800
801	=item L<AnyEvent::BDB>, L<BDB>
802
803	Truly asynchronous Berkeley DB access. AnyEvent::AIO transparently fuses
804	IO::AIO and AnyEvent together.
805
806	=item L<IO::Lambda>
807
808	The lambda approach to I/O - don't ask, look there. Can use AnyEvent.
809
810	=back	1143	=back
811		1144
812	=cut	1145	=cut
813		1146
814	package AnyEvent;	1147	package AnyEvent;
815		1148
816	no warnings;	1149	# basically a tuned-down version of common::sense
817	use strict;	1150	sub common_sense {
		1151	# from common:.sense 1.0
		1152	${^WARNING_BITS} = "\xfc\x3f\x33\x00\x0f\xf3\xcf\xc0\xf3\xfc\x33\x00";
		1153	# use strict vars subs - NO UTF-8, as Util.pm doesn't like this atm. (uts46data.pl)
		1154	$^H |= 0x00000600;
		1155	}
818		1156
		1157	BEGIN { AnyEvent::common_sense }
		1158
819	use Carp;	1159	use Carp ();
820		1160
821	our $VERSION = 4.152;	1161	our $VERSION = '5.24';
822	our $MODEL;	1162	our $MODEL;
823		1163
824	our $AUTOLOAD;	1164	our $AUTOLOAD;
825	our @ISA;	1165	our @ISA;
826		1166
827	our @REGISTRY;	1167	our @REGISTRY;
828		1168
829	our $WIN32;	1169	our $VERBOSE;
830		1170
831	BEGIN {	1171	BEGIN {
832	my $win32 = ! ! ($^O =~ /mswin32/i);	1172	eval "sub CYGWIN(){" . (($^O =~ /cygwin/i) *1) . "}";
833	eval "sub WIN32(){ $win32 }";	1173	eval "sub WIN32 (){" . (($^O =~ /mswin32/i)*1) . "}";
834	}	1174	eval "sub TAINT (){" . (${^TAINT} *1) . "}";
835		1175
		1176	delete @ENV{grep /^PERL_ANYEVENT_/, keys %ENV}
		1177	if ${^TAINT};
		1178
836	our $verbose = $ENV{PERL_ANYEVENT_VERBOSE}*1;	1179	$VERBOSE = $ENV{PERL_ANYEVENT_VERBOSE}*1;
		1180
		1181	}
		1182
		1183	our $MAX_SIGNAL_LATENCY = 10;
837		1184
838	our %PROTOCOL; # (ipv4|ipv6) => (1|2), higher numbers are preferred	1185	our %PROTOCOL; # (ipv4|ipv6) => (1|2), higher numbers are preferred
839		1186
840	{	1187	{
841	my $idx;	1188	my $idx;
…		…
843	for reverse split /\s*,\s*/,	1190	for reverse split /\s*,\s*/,
844	$ENV{PERL_ANYEVENT_PROTOCOLS} || "ipv4,ipv6";	1191	$ENV{PERL_ANYEVENT_PROTOCOLS} || "ipv4,ipv6";
845	}	1192	}
846		1193
847	my @models = (	1194	my @models = (
848	[EV:: => AnyEvent::Impl::EV::],	1195	[EV:: => AnyEvent::Impl::EV:: , 1],
849	[Event:: => AnyEvent::Impl::Event::],
850	[AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl::],	1196	[AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl:: , 1],
851	# everything below here will not be autoprobed	1197	# everything below here will not (normally) be autoprobed
852	# as the pureperl backend should work everywhere	1198	# as the pureperl backend should work everywhere
853	# and is usually faster	1199	# and is usually faster
		1200	[Event:: => AnyEvent::Impl::Event::, 1],
		1201	[Glib:: => AnyEvent::Impl::Glib:: , 1], # becomes extremely slow with many watchers
		1202	[Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy
		1203	[Irssi:: => AnyEvent::Impl::Irssi::], # Irssi has a bogus "Event" package
854	[Tk:: => AnyEvent::Impl::Tk::], # crashes with many handles	1204	[Tk:: => AnyEvent::Impl::Tk::], # crashes with many handles
855	[Glib:: => AnyEvent::Impl::Glib::], # becomes extremely slow with many watchers
856	[Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy
857	[Qt:: => AnyEvent::Impl::Qt::], # requires special main program	1205	[Qt:: => AnyEvent::Impl::Qt::], # requires special main program
858	[POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza	1206	[POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza
859	[Wx:: => AnyEvent::Impl::POE::],	1207	[Wx:: => AnyEvent::Impl::POE::],
860	[Prima:: => AnyEvent::Impl::POE::],	1208	[Prima:: => AnyEvent::Impl::POE::],
		1209	# IO::Async is just too broken - we would need workarounds for its
		1210	# byzantine signal and broken child handling, among others.
		1211	# IO::Async is rather hard to detect, as it doesn't have any
		1212	# obvious default class.
		1213	[IO::Async:: => AnyEvent::Impl::IOAsync::], # requires special main program
		1214	[IO::Async::Loop:: => AnyEvent::Impl::IOAsync::], # requires special main program
		1215	[IO::Async::Notifier:: => AnyEvent::Impl::IOAsync::], # requires special main program
		1216	[AnyEvent::Impl::IOAsync:: => AnyEvent::Impl::IOAsync::], # requires special main program
861	);	1217	);
862		1218
863	our %method = map +($_ => 1), qw(io timer time now signal child condvar one_event DESTROY);	1219	our %method = map +($_ => 1),
		1220	qw(io timer time now now_update signal child idle condvar one_event DESTROY);
864		1221
865	our @post_detect;	1222	our @post_detect;
866		1223
867	sub post_detect(&) {	1224	sub post_detect(&) {
868	my ($cb) = @_;	1225	my ($cb) = @_;
869		1226
870	if ($MODEL) {	1227	if ($MODEL) {
871	$cb->();	1228	$cb->();
872		1229
873	1	1230	undef
874	} else {	1231	} else {
875	push @post_detect, $cb;	1232	push @post_detect, $cb;
876		1233
877	defined wantarray	1234	defined wantarray
878	? bless \$cb, "AnyEvent::Util::PostDetect"	1235	? bless \$cb, "AnyEvent::Util::postdetect"
879	: ()	1236	: ()
880	}	1237	}
881	}	1238	}
882		1239
883	sub AnyEvent::Util::PostDetect::DESTROY {	1240	sub AnyEvent::Util::postdetect::DESTROY {
884	@post_detect = grep $_ != ${$_[0]}, @post_detect;	1241	@post_detect = grep $_ != ${$_[0]}, @post_detect;
885	}	1242	}
886		1243
887	sub detect() {	1244	sub detect() {
		1245	# free some memory
		1246	*detect = sub () { $MODEL };
		1247
		1248	local $!; # for good measure
		1249	local $SIG{__DIE__};
		1250
		1251	if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {
		1252	my $model = "AnyEvent::Impl::$1";
		1253	if (eval "require $model") {
		1254	$MODEL = $model;
		1255	warn "AnyEvent: loaded model '$model' (forced by \$ENV{PERL_ANYEVENT_MODEL}), using it.\n" if $VERBOSE >= 2;
		1256	} else {
		1257	warn "AnyEvent: unable to load model '$model' (from \$ENV{PERL_ANYEVENT_MODEL}):\n$@" if $VERBOSE;
		1258	}
		1259	}
		1260
		1261	# check for already loaded models
888	unless ($MODEL) {	1262	unless ($MODEL) {
889	no strict 'refs';	1263	for (@REGISTRY, @models) {
890	local $SIG{__DIE__};	1264	my ($package, $model) = @$_;
891		1265	if (${"$package\::VERSION"} > 0) {
892	if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {
893	my $model = "AnyEvent::Impl::$1";
894	if (eval "require $model") {	1266	if (eval "require $model") {
895	$MODEL = $model;	1267	$MODEL = $model;
896	warn "AnyEvent: loaded model '$model' (forced by \$PERL_ANYEVENT_MODEL), using it.\n" if $verbose > 1;	1268	warn "AnyEvent: autodetected model '$model', using it.\n" if $VERBOSE >= 2;
897	} else {	1269	last;
898	warn "AnyEvent: unable to load model '$model' (from \$PERL_ANYEVENT_MODEL):\n$@" if $verbose;	1270	}
899	}	1271	}
900	}	1272	}
901		1273
902	# check for already loaded models
903	unless ($MODEL) {	1274	unless ($MODEL) {
		1275	# try to autoload a model
904	for (@REGISTRY, @models) {	1276	for (@REGISTRY, @models) {
905	my ($package, $model) = @$_;	1277	my ($package, $model, $autoload) = @$_;
		1278	if (
		1279	$autoload
		1280	and eval "require $package"
906	if (${"$package\::VERSION"} > 0) {	1281	and ${"$package\::VERSION"} > 0
907	if (eval "require $model") {	1282	and eval "require $model"
		1283	) {
908	$MODEL = $model;	1284	$MODEL = $model;
909	warn "AnyEvent: autodetected model '$model', using it.\n" if $verbose > 1;	1285	warn "AnyEvent: autoloaded model '$model', using it.\n" if $VERBOSE >= 2;
910	last;	1286	last;
911	}
912	}	1287	}
913	}	1288	}
914		1289
915	unless ($MODEL) {
916	# try to load a model
917
918	for (@REGISTRY, @models) {
919	my ($package, $model) = @$_;
920	if (eval "require $package"
921	and ${"$package\::VERSION"} > 0
922	and eval "require $model") {
923	$MODEL = $model;
924	warn "AnyEvent: autoprobed model '$model', using it.\n" if $verbose > 1;
925	last;
926	}
927	}
928
929	$MODEL	1290	$MODEL
930	or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV, Event or Glib.";	1291	or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV, Event or Glib.\n";
931	}
932	}	1292	}
933
934	unshift @ISA, $MODEL;
935	push @{"$MODEL\::ISA"}, "AnyEvent::Base";
936
937	(shift @post_detect)->() while @post_detect;
938	}	1293	}
		1294
		1295	@models = (); # free probe data
		1296
		1297	push @{"$MODEL\::ISA"}, "AnyEvent::Base";
		1298	unshift @ISA, $MODEL;
		1299
		1300	require AnyEvent::Strict if $ENV{PERL_ANYEVENT_STRICT};
		1301
		1302	(shift @post_detect)->() while @post_detect;
939		1303
940	$MODEL	1304	$MODEL
941	}	1305	}
942		1306
943	sub AUTOLOAD {	1307	sub AUTOLOAD {
944	(my $func = $AUTOLOAD) =~ s/.*://;	1308	(my $func = $AUTOLOAD) =~ s/.*://;
945		1309
946	$method{$func}	1310	$method{$func}
947	or croak "$func: not a valid method for AnyEvent objects";	1311	or Carp::croak "$func: not a valid AnyEvent class method";
948		1312
949	detect unless $MODEL;	1313	detect;
950		1314
951	my $class = shift;	1315	my $class = shift;
952	$class->$func (@_);	1316	$class->$func (@_);
953	}	1317	}
954		1318
		1319	# utility function to dup a filehandle. this is used by many backends
		1320	# to support binding more than one watcher per filehandle (they usually
		1321	# allow only one watcher per fd, so we dup it to get a different one).
		1322	sub _dupfh($$;$$) {
		1323	my ($poll, $fh, $r, $w) = @_;
		1324
		1325	# cygwin requires the fh mode to be matching, unix doesn't
		1326	my ($rw, $mode) = $poll eq "r" ? ($r, "<&") : ($w, ">&");
		1327
		1328	open my $fh2, $mode, $fh
		1329	or die "AnyEvent->io: cannot dup() filehandle in mode '$poll': $!,";
		1330
		1331	# we assume CLOEXEC is already set by perl in all important cases
		1332
		1333	($fh2, $rw)
		1334	}
		1335
		1336	=head1 SIMPLIFIED AE API
		1337
		1338	Starting with version 5.0, AnyEvent officially supports a second, much
		1339	simpler, API that is designed to reduce the calling, typing and memory
		1340	overhead.
		1341
		1342	See the L<AE> manpage for details.
		1343
		1344	=cut
		1345
		1346	package AE;
		1347
		1348	our $VERSION = $AnyEvent::VERSION;
		1349
		1350	sub io($$$) {
		1351	AnyEvent->io (fh => $_[0], poll => $_[1] ? "w" : "r", cb => $_[2])
		1352	}
		1353
		1354	sub timer($$$) {
		1355	AnyEvent->timer (after => $_[0], interval => $_[1], cb => $_[2])
		1356	}
		1357
		1358	sub signal($$) {
		1359	AnyEvent->signal (signal => $_[0], cb => $_[1])
		1360	}
		1361
		1362	sub child($$) {
		1363	AnyEvent->child (pid => $_[0], cb => $_[1])
		1364	}
		1365
		1366	sub idle($) {
		1367	AnyEvent->idle (cb => $_[0])
		1368	}
		1369
		1370	sub cv(;&) {
		1371	AnyEvent->condvar (@_ ? (cb => $_[0]) : ())
		1372	}
		1373
		1374	sub now() {
		1375	AnyEvent->now
		1376	}
		1377
		1378	sub now_update() {
		1379	AnyEvent->now_update
		1380	}
		1381
		1382	sub time() {
		1383	AnyEvent->time
		1384	}
		1385
955	package AnyEvent::Base;	1386	package AnyEvent::Base;
956		1387
957	# default implementation for now and time	1388	# default implementations for many methods
958		1389
959	use Time::HiRes ();	1390	sub _time() {
		1391	eval q{ # poor man's autoloading
		1392	# probe for availability of Time::HiRes
		1393	if (eval "use Time::HiRes (); Time::HiRes::time (); 1") {
		1394	warn "AnyEvent: using Time::HiRes for sub-second timing accuracy.\n" if $VERBOSE >= 8;
		1395	*_time = \&Time::HiRes::time;
		1396	# if (eval "use POSIX (); (POSIX::times())...
		1397	} else {
		1398	warn "AnyEvent: using built-in time(), WARNING, no sub-second resolution!\n" if $VERBOSE;
		1399	*_time = sub (){ time }; # epic fail
		1400	}
		1401	};
		1402	die if $@;
960		1403
961	sub time { Time::HiRes::time }	1404	&_time
962	sub now { Time::HiRes::time }	1405	}
		1406
		1407	sub time { _time }
		1408	sub now { _time }
		1409	sub now_update { }
963		1410
964	# default implementation for ->condvar	1411	# default implementation for ->condvar
965		1412
966	sub condvar {	1413	sub condvar {
967	bless { @_ == 3 ? (_ae_cb => $_[2]) : () }, AnyEvent::CondVar::	1414	bless { @_ == 3 ? (_ae_cb => $_[2]) : () }, "AnyEvent::CondVar"
968	}	1415	}
969		1416
970	# default implementation for ->signal	1417	# default implementation for ->signal
971		1418
972	our %SIG_CB;	1419	our $HAVE_ASYNC_INTERRUPT;
		1420
		1421	sub _have_async_interrupt() {
		1422	$HAVE_ASYNC_INTERRUPT = 1*(!$ENV{PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT}
		1423	&& eval "use Async::Interrupt 1.02 (); 1")
		1424	unless defined $HAVE_ASYNC_INTERRUPT;
		1425
		1426	$HAVE_ASYNC_INTERRUPT
		1427	}
		1428
		1429	our ($SIGPIPE_R, $SIGPIPE_W, %SIG_CB, %SIG_EV, $SIG_IO);
		1430	our (%SIG_ASY, %SIG_ASY_W);
		1431	our ($SIG_COUNT, $SIG_TW);
		1432
		1433	# install a dummy wakeup watcher to reduce signal catching latency
		1434	# used by Impls
		1435	sub _sig_add() {
		1436	unless ($SIG_COUNT++) {
		1437	# try to align timer on a full-second boundary, if possible
		1438	my $NOW = AE::now;
		1439
		1440	$SIG_TW = AE::timer
		1441	$MAX_SIGNAL_LATENCY - ($NOW - int $NOW),
		1442	$MAX_SIGNAL_LATENCY,
		1443	sub { } # just for the PERL_ASYNC_CHECK
		1444	;
		1445	}
		1446	}
		1447
		1448	sub _sig_del {
		1449	undef $SIG_TW
		1450	unless --$SIG_COUNT;
		1451	}
		1452
		1453	our $_sig_name_init; $_sig_name_init = sub {
		1454	eval q{ # poor man's autoloading
		1455	undef $_sig_name_init;
		1456
		1457	if (_have_async_interrupt) {
		1458	*sig2num = \&Async::Interrupt::sig2num;
		1459	*sig2name = \&Async::Interrupt::sig2name;
		1460	} else {
		1461	require Config;
		1462
		1463	my %signame2num;
		1464	@signame2num{ split ' ', $Config::Config{sig_name} }
		1465	= split ' ', $Config::Config{sig_num};
		1466
		1467	my @signum2name;
		1468	@signum2name[values %signame2num] = keys %signame2num;
		1469
		1470	*sig2num = sub($) {
		1471	$_[0] > 0 ? shift : $signame2num{+shift}
		1472	};
		1473	*sig2name = sub ($) {
		1474	$_[0] > 0 ? $signum2name[+shift] : shift
		1475	};
		1476	}
		1477	};
		1478	die if $@;
		1479	};
		1480
		1481	sub sig2num ($) { &$_sig_name_init; &sig2num }
		1482	sub sig2name($) { &$_sig_name_init; &sig2name }
973		1483
974	sub signal {	1484	sub signal {
		1485	eval q{ # poor man's autoloading {}
		1486	# probe for availability of Async::Interrupt
		1487	if (_have_async_interrupt) {
		1488	warn "AnyEvent: using Async::Interrupt for race-free signal handling.\n" if $VERBOSE >= 8;
		1489
		1490	$SIGPIPE_R = new Async::Interrupt::EventPipe;
		1491	$SIG_IO = AE::io $SIGPIPE_R->fileno, 0, \&_signal_exec;
		1492
		1493	} else {
		1494	warn "AnyEvent: using emulated perl signal handling with latency timer.\n" if $VERBOSE >= 8;
		1495
		1496	require Fcntl;
		1497
		1498	if (AnyEvent::WIN32) {
		1499	require AnyEvent::Util;
		1500
		1501	($SIGPIPE_R, $SIGPIPE_W) = AnyEvent::Util::portable_pipe ();
		1502	AnyEvent::Util::fh_nonblocking ($SIGPIPE_R, 1) if $SIGPIPE_R;
		1503	AnyEvent::Util::fh_nonblocking ($SIGPIPE_W, 1) if $SIGPIPE_W; # just in case
		1504	} else {
		1505	pipe $SIGPIPE_R, $SIGPIPE_W;
		1506	fcntl $SIGPIPE_R, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_R;
		1507	fcntl $SIGPIPE_W, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_W; # just in case
		1508
		1509	# not strictly required, as $^F is normally 2, but let's make sure...
		1510	fcntl $SIGPIPE_R, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC;
		1511	fcntl $SIGPIPE_W, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC;
		1512	}
		1513
		1514	$SIGPIPE_R
		1515	or Carp::croak "AnyEvent: unable to create a signal reporting pipe: $!\n";
		1516
		1517	$SIG_IO = AE::io $SIGPIPE_R, 0, \&_signal_exec;
		1518	}
		1519
		1520	*signal = sub {
975	my (undef, %arg) = @_;	1521	my (undef, %arg) = @_;
976		1522
977	my $signal = uc $arg{signal}	1523	my $signal = uc $arg{signal}
978	or Carp::croak "required option 'signal' is missing";	1524	or Carp::croak "required option 'signal' is missing";
979		1525
		1526	if ($HAVE_ASYNC_INTERRUPT) {
		1527	# async::interrupt
		1528
		1529	$signal = sig2num $signal;
980	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};	1530	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};
		1531
		1532	$SIG_ASY{$signal} ||= new Async::Interrupt
		1533	cb => sub { undef $SIG_EV{$signal} },
		1534	signal => $signal,
		1535	pipe => [$SIGPIPE_R->filenos],
		1536	pipe_autodrain => 0,
		1537	;
		1538
		1539	} else {
		1540	# pure perl
		1541
		1542	# AE::Util has been loaded in signal
		1543	$signal = sig2name $signal;
		1544	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};
		1545
981	$SIG{$signal} ||= sub {	1546	$SIG{$signal} ||= sub {
		1547	local $!;
		1548	syswrite $SIGPIPE_W, "\x00", 1 unless %SIG_EV;
		1549	undef $SIG_EV{$signal};
		1550	};
		1551
		1552	# can't do signal processing without introducing races in pure perl,
		1553	# so limit the signal latency.
		1554	_sig_add;
		1555	}
		1556
		1557	bless [$signal, $arg{cb}], "AnyEvent::Base::signal"
		1558	};
		1559
		1560	*AnyEvent::Base::signal::DESTROY = sub {
		1561	my ($signal, $cb) = @{$_[0]};
		1562
		1563	_sig_del;
		1564
		1565	delete $SIG_CB{$signal}{$cb};
		1566
		1567	$HAVE_ASYNC_INTERRUPT
		1568	? delete $SIG_ASY{$signal}
		1569	: # delete doesn't work with older perls - they then
		1570	# print weird messages, or just unconditionally exit
		1571	# instead of getting the default action.
		1572	undef $SIG{$signal}
		1573	unless keys %{ $SIG_CB{$signal} };
		1574	};
		1575
		1576	*_signal_exec = sub {
		1577	$HAVE_ASYNC_INTERRUPT
		1578	? $SIGPIPE_R->drain
		1579	: sysread $SIGPIPE_R, (my $dummy), 9;
		1580
		1581	while (%SIG_EV) {
		1582	for (keys %SIG_EV) {
		1583	delete $SIG_EV{$_};
982	$_->() for values %{ $SIG_CB{$signal} || {} };	1584	$_->() for values %{ $SIG_CB{$_} || {} };
		1585	}
		1586	}
		1587	};
983	};	1588	};
		1589	die if $@;
984		1590
985	bless [$signal, $arg{cb}], "AnyEvent::Base::Signal"	1591	&signal
986	}
987
988	sub AnyEvent::Base::Signal::DESTROY {
989	my ($signal, $cb) = @{$_[0]};
990
991	delete $SIG_CB{$signal}{$cb};
992
993	$SIG{$signal} = 'DEFAULT' unless keys %{ $SIG_CB{$signal} };
994	}	1592	}
995		1593
996	# default implementation for ->child	1594	# default implementation for ->child
997		1595
998	our %PID_CB;	1596	our %PID_CB;
999	our $CHLD_W;	1597	our $CHLD_W;
1000	our $CHLD_DELAY_W;	1598	our $CHLD_DELAY_W;
1001	our $PID_IDLE;
1002	our $WNOHANG;	1599	our $WNOHANG;
1003		1600
1004	sub _child_wait {	1601	# used by many Impl's
1005	while (0 < (my $pid = waitpid -1, $WNOHANG)) {	1602	sub _emit_childstatus($$) {
		1603	my (undef, $rpid, $rstatus) = @_;
		1604
		1605	$_->($rpid, $rstatus)
1006	$_->($pid, $?) for (values %{ $PID_CB{$pid} || {} }),	1606	for values %{ $PID_CB{$rpid} || {} },
1007	(values %{ $PID_CB{0} || {} });	1607	values %{ $PID_CB{0} || {} };
1008	}
1009
1010	undef $PID_IDLE;
1011	}
1012
1013	sub _sigchld {
1014	# make sure we deliver these changes "synchronous" with the event loop.
1015	$CHLD_DELAY_W ||= AnyEvent->timer (after => 0, cb => sub {
1016	undef $CHLD_DELAY_W;
1017	&_child_wait;
1018	});
1019	}	1608	}
1020		1609
1021	sub child {	1610	sub child {
		1611	eval q{ # poor man's autoloading {}
		1612	*_sigchld = sub {
		1613	my $pid;
		1614
		1615	AnyEvent->_emit_childstatus ($pid, $?)
		1616	while ($pid = waitpid -1, $WNOHANG) > 0;
		1617	};
		1618
		1619	*child = sub {
1022	my (undef, %arg) = @_;	1620	my (undef, %arg) = @_;
1023		1621
1024	defined (my $pid = $arg{pid} + 0)	1622	defined (my $pid = $arg{pid} + 0)
1025	or Carp::croak "required option 'pid' is missing";	1623	or Carp::croak "required option 'pid' is missing";
1026		1624
1027	$PID_CB{$pid}{$arg{cb}} = $arg{cb};	1625	$PID_CB{$pid}{$arg{cb}} = $arg{cb};
1028		1626
1029	unless ($WNOHANG) {	1627	# WNOHANG is almost cetrainly 1 everywhere
		1628	$WNOHANG ||= $^O =~ /^(?:openbsd|netbsd|linux|freebsd|cygwin|MSWin32)$/
		1629	? 1
1030	$WNOHANG = eval { local $SIG{__DIE__}; require POSIX; &POSIX::WNOHANG } || 1;	1630	: eval { local $SIG{__DIE__}; require POSIX; &POSIX::WNOHANG } || 1;
1031	}
1032		1631
1033	unless ($CHLD_W) {	1632	unless ($CHLD_W) {
1034	$CHLD_W = AnyEvent->signal (signal => 'CHLD', cb => \&_sigchld);	1633	$CHLD_W = AE::signal CHLD => \&_sigchld;
1035	# child could be a zombie already, so make at least one round	1634	# child could be a zombie already, so make at least one round
1036	&_sigchld;	1635	&_sigchld;
1037	}	1636	}
1038		1637
1039	bless [$pid, $arg{cb}], "AnyEvent::Base::Child"	1638	bless [$pid, $arg{cb}], "AnyEvent::Base::child"
1040	}	1639	};
1041		1640
1042	sub AnyEvent::Base::Child::DESTROY {	1641	*AnyEvent::Base::child::DESTROY = sub {
1043	my ($pid, $cb) = @{$_[0]};	1642	my ($pid, $cb) = @{$_[0]};
1044		1643
1045	delete $PID_CB{$pid}{$cb};	1644	delete $PID_CB{$pid}{$cb};
1046	delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };	1645	delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };
1047		1646
1048	undef $CHLD_W unless keys %PID_CB;	1647	undef $CHLD_W unless keys %PID_CB;
		1648	};
		1649	};
		1650	die if $@;
		1651
		1652	&child
		1653	}
		1654
		1655	# idle emulation is done by simply using a timer, regardless
		1656	# of whether the process is idle or not, and not letting
		1657	# the callback use more than 50% of the time.
		1658	sub idle {
		1659	eval q{ # poor man's autoloading {}
		1660	*idle = sub {
		1661	my (undef, %arg) = @_;
		1662
		1663	my ($cb, $w, $rcb) = $arg{cb};
		1664
		1665	$rcb = sub {
		1666	if ($cb) {
		1667	$w = _time;
		1668	&$cb;
		1669	$w = _time - $w;
		1670
		1671	# never use more then 50% of the time for the idle watcher,
		1672	# within some limits
		1673	$w = 0.0001 if $w < 0.0001;
		1674	$w = 5 if $w > 5;
		1675
		1676	$w = AE::timer $w, 0, $rcb;
		1677	} else {
		1678	# clean up...
		1679	undef $w;
		1680	undef $rcb;
		1681	}
		1682	};
		1683
		1684	$w = AE::timer 0.05, 0, $rcb;
		1685
		1686	bless \\$cb, "AnyEvent::Base::idle"
		1687	};
		1688
		1689	*AnyEvent::Base::idle::DESTROY = sub {
		1690	undef $${$_[0]};
		1691	};
		1692	};
		1693	die if $@;
		1694
		1695	&idle
1049	}	1696	}
1050		1697
1051	package AnyEvent::CondVar;	1698	package AnyEvent::CondVar;
1052		1699
1053	our @ISA = AnyEvent::CondVar::Base::;	1700	our @ISA = AnyEvent::CondVar::Base::;
1054		1701
1055	package AnyEvent::CondVar::Base;	1702	package AnyEvent::CondVar::Base;
1056		1703
1057	use overload	1704	#use overload
1058	'&{}' => sub { my $self = shift; sub { $self->send (@_) } },	1705	# '&{}' => sub { my $self = shift; sub { $self->send (@_) } },
1059	fallback => 1;	1706	# fallback => 1;
		1707
		1708	# save 300+ kilobytes by dirtily hardcoding overloading
		1709	${"AnyEvent::CondVar::Base::OVERLOAD"}{dummy}++; # Register with magic by touching.
		1710	*{'AnyEvent::CondVar::Base::()'} = sub { }; # "Make it findable via fetchmethod."
		1711	*{'AnyEvent::CondVar::Base::(&{}'} = sub { my $self = shift; sub { $self->send (@_) } }; # &{}
		1712	${'AnyEvent::CondVar::Base::()'} = 1; # fallback
		1713
		1714	our $WAITING;
1060		1715
1061	sub _send {	1716	sub _send {
1062	# nop	1717	# nop
1063	}	1718	}
1064		1719
…		…
1077	sub ready {	1732	sub ready {
1078	$_[0]{_ae_sent}	1733	$_[0]{_ae_sent}
1079	}	1734	}
1080		1735
1081	sub _wait {	1736	sub _wait {
		1737	$WAITING
		1738	and !$_[0]{_ae_sent}
		1739	and Carp::croak "AnyEvent::CondVar: recursive blocking wait detected";
		1740
		1741	local $WAITING = 1;
1082	AnyEvent->one_event while !$_[0]{_ae_sent};	1742	AnyEvent->one_event while !$_[0]{_ae_sent};
1083	}	1743	}
1084		1744
1085	sub recv {	1745	sub recv {
1086	$_[0]->_wait;	1746	$_[0]->_wait;
…		…
1088	Carp::croak $_[0]{_ae_croak} if $_[0]{_ae_croak};	1748	Carp::croak $_[0]{_ae_croak} if $_[0]{_ae_croak};
1089	wantarray ? @{ $_[0]{_ae_sent} } : $_[0]{_ae_sent}[0]	1749	wantarray ? @{ $_[0]{_ae_sent} } : $_[0]{_ae_sent}[0]
1090	}	1750	}
1091		1751
1092	sub cb {	1752	sub cb {
1093	$_[0]{_ae_cb} = $_[1] if @_ > 1;	1753	my $cv = shift;
		1754
		1755	@_
		1756	and $cv->{_ae_cb} = shift
		1757	and $cv->{_ae_sent}
		1758	and (delete $cv->{_ae_cb})->($cv);
		1759
1094	$_[0]{_ae_cb}	1760	$cv->{_ae_cb}
1095	}	1761	}
1096		1762
1097	sub begin {	1763	sub begin {
1098	++$_[0]{_ae_counter};	1764	++$_[0]{_ae_counter};
1099	$_[0]{_ae_end_cb} = $_[1] if @_ > 1;	1765	$_[0]{_ae_end_cb} = $_[1] if @_ > 1;
…		…
1105	}	1771	}
1106		1772
1107	# undocumented/compatibility with pre-3.4	1773	# undocumented/compatibility with pre-3.4
1108	*broadcast = \&send;	1774	*broadcast = \&send;
1109	*wait = \&_wait;	1775	*wait = \&_wait;
		1776
		1777	=head1 ERROR AND EXCEPTION HANDLING
		1778
		1779	In general, AnyEvent does not do any error handling - it relies on the
		1780	caller to do that if required. The L<AnyEvent::Strict> module (see also
		1781	the C<PERL_ANYEVENT_STRICT> environment variable, below) provides strict
		1782	checking of all AnyEvent methods, however, which is highly useful during
		1783	development.
		1784
		1785	As for exception handling (i.e. runtime errors and exceptions thrown while
		1786	executing a callback), this is not only highly event-loop specific, but
		1787	also not in any way wrapped by this module, as this is the job of the main
		1788	program.
		1789
		1790	The pure perl event loop simply re-throws the exception (usually
		1791	within C<< condvar->recv >>), the L<Event> and L<EV> modules call C<<
		1792	$Event/EV::DIED->() >>, L<Glib> uses C<< install_exception_handler >> and
		1793	so on.
		1794
		1795	=head1 ENVIRONMENT VARIABLES
		1796
		1797	The following environment variables are used by this module or its
		1798	submodules.
		1799
		1800	Note that AnyEvent will remove I<all> environment variables starting with
		1801	C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
		1802	enabled.
		1803
		1804	=over 4
		1805
		1806	=item C<PERL_ANYEVENT_VERBOSE>
		1807
		1808	By default, AnyEvent will be completely silent except in fatal
		1809	conditions. You can set this environment variable to make AnyEvent more
		1810	talkative.
		1811
		1812	When set to C<1> or higher, causes AnyEvent to warn about unexpected
		1813	conditions, such as not being able to load the event model specified by
		1814	C<PERL_ANYEVENT_MODEL>.
		1815
		1816	When set to C<2> or higher, cause AnyEvent to report to STDERR which event
		1817	model it chooses.
		1818
		1819	When set to C<8> or higher, then AnyEvent will report extra information on
		1820	which optional modules it loads and how it implements certain features.
		1821
		1822	=item C<PERL_ANYEVENT_STRICT>
		1823
		1824	AnyEvent does not do much argument checking by default, as thorough
		1825	argument checking is very costly. Setting this variable to a true value
		1826	will cause AnyEvent to load C<AnyEvent::Strict> and then to thoroughly
		1827	check the arguments passed to most method calls. If it finds any problems,
		1828	it will croak.
		1829
		1830	In other words, enables "strict" mode.
		1831
		1832	Unlike C<use strict> (or it's modern cousin, C<< use L<common::sense>
		1833	>>, it is definitely recommended to keep it off in production. Keeping
		1834	C<PERL_ANYEVENT_STRICT=1> in your environment while developing programs
		1835	can be very useful, however.
		1836
		1837	=item C<PERL_ANYEVENT_MODEL>
		1838
		1839	This can be used to specify the event model to be used by AnyEvent, before
		1840	auto detection and -probing kicks in. It must be a string consisting
		1841	entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended
		1842	and the resulting module name is loaded and if the load was successful,
		1843	used as event model. If it fails to load AnyEvent will proceed with
		1844	auto detection and -probing.
		1845
		1846	This functionality might change in future versions.
		1847
		1848	For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you
		1849	could start your program like this:
		1850
		1851	PERL_ANYEVENT_MODEL=Perl perl ...
		1852
		1853	=item C<PERL_ANYEVENT_PROTOCOLS>
		1854
		1855	Used by both L<AnyEvent::DNS> and L<AnyEvent::Socket> to determine preferences
		1856	for IPv4 or IPv6. The default is unspecified (and might change, or be the result
		1857	of auto probing).
		1858
		1859	Must be set to a comma-separated list of protocols or address families,
		1860	current supported: C<ipv4> and C<ipv6>. Only protocols mentioned will be
		1861	used, and preference will be given to protocols mentioned earlier in the
		1862	list.
		1863
		1864	This variable can effectively be used for denial-of-service attacks
		1865	against local programs (e.g. when setuid), although the impact is likely
		1866	small, as the program has to handle conenction and other failures anyways.
		1867
		1868	Examples: C<PERL_ANYEVENT_PROTOCOLS=ipv4,ipv6> - prefer IPv4 over IPv6,
		1869	but support both and try to use both. C<PERL_ANYEVENT_PROTOCOLS=ipv4>
		1870	- only support IPv4, never try to resolve or contact IPv6
		1871	addresses. C<PERL_ANYEVENT_PROTOCOLS=ipv6,ipv4> support either IPv4 or
		1872	IPv6, but prefer IPv6 over IPv4.
		1873
		1874	=item C<PERL_ANYEVENT_EDNS0>
		1875
		1876	Used by L<AnyEvent::DNS> to decide whether to use the EDNS0 extension
		1877	for DNS. This extension is generally useful to reduce DNS traffic, but
		1878	some (broken) firewalls drop such DNS packets, which is why it is off by
		1879	default.
		1880
		1881	Setting this variable to C<1> will cause L<AnyEvent::DNS> to announce
		1882	EDNS0 in its DNS requests.
		1883
		1884	=item C<PERL_ANYEVENT_MAX_FORKS>
		1885
		1886	The maximum number of child processes that C<AnyEvent::Util::fork_call>
		1887	will create in parallel.
		1888
		1889	=item C<PERL_ANYEVENT_MAX_OUTSTANDING_DNS>
		1890
		1891	The default value for the C<max_outstanding> parameter for the default DNS
		1892	resolver - this is the maximum number of parallel DNS requests that are
		1893	sent to the DNS server.
		1894
		1895	=item C<PERL_ANYEVENT_RESOLV_CONF>
		1896
		1897	The file to use instead of F</etc/resolv.conf> (or OS-specific
		1898	configuration) in the default resolver. When set to the empty string, no
		1899	default config will be used.
		1900
		1901	=item C<PERL_ANYEVENT_CA_FILE>, C<PERL_ANYEVENT_CA_PATH>.
		1902
		1903	When neither C<ca_file> nor C<ca_path> was specified during
		1904	L<AnyEvent::TLS> context creation, and either of these environment
		1905	variables exist, they will be used to specify CA certificate locations
		1906	instead of a system-dependent default.
		1907
		1908	=item C<PERL_ANYEVENT_AVOID_GUARD> and C<PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT>
		1909
		1910	When these are set to C<1>, then the respective modules are not
		1911	loaded. Mostly good for testing AnyEvent itself.
		1912
		1913	=back
1110		1914
1111	=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE	1915	=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE
1112		1916
1113	This is an advanced topic that you do not normally need to use AnyEvent in	1917	This is an advanced topic that you do not normally need to use AnyEvent in
1114	a module. This section is only of use to event loop authors who want to	1918	a module. This section is only of use to event loop authors who want to
…		…
1148		1952
1149	I<rxvt-unicode> also cheats a bit by not providing blocking access to	1953	I<rxvt-unicode> also cheats a bit by not providing blocking access to
1150	condition variables: code blocking while waiting for a condition will	1954	condition variables: code blocking while waiting for a condition will
1151	C<die>. This still works with most modules/usages, and blocking calls must	1955	C<die>. This still works with most modules/usages, and blocking calls must
1152	not be done in an interactive application, so it makes sense.	1956	not be done in an interactive application, so it makes sense.
1153
1154	=head1 ENVIRONMENT VARIABLES
1155
1156	The following environment variables are used by this module:
1157
1158	=over 4
1159
1160	=item C<PERL_ANYEVENT_VERBOSE>
1161
1162	By default, AnyEvent will be completely silent except in fatal
1163	conditions. You can set this environment variable to make AnyEvent more
1164	talkative.
1165
1166	When set to C<1> or higher, causes AnyEvent to warn about unexpected
1167	conditions, such as not being able to load the event model specified by
1168	C<PERL_ANYEVENT_MODEL>.
1169
1170	When set to C<2> or higher, cause AnyEvent to report to STDERR which event
1171	model it chooses.
1172
1173	=item C<PERL_ANYEVENT_MODEL>
1174
1175	This can be used to specify the event model to be used by AnyEvent, before
1176	auto detection and -probing kicks in. It must be a string consisting
1177	entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended
1178	and the resulting module name is loaded and if the load was successful,
1179	used as event model. If it fails to load AnyEvent will proceed with
1180	auto detection and -probing.
1181
1182	This functionality might change in future versions.
1183
1184	For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you
1185	could start your program like this:
1186
1187	PERL_ANYEVENT_MODEL=Perl perl ...
1188
1189	=item C<PERL_ANYEVENT_PROTOCOLS>
1190
1191	Used by both L<AnyEvent::DNS> and L<AnyEvent::Socket> to determine preferences
1192	for IPv4 or IPv6. The default is unspecified (and might change, or be the result
1193	of auto probing).
1194
1195	Must be set to a comma-separated list of protocols or address families,
1196	current supported: C<ipv4> and C<ipv6>. Only protocols mentioned will be
1197	used, and preference will be given to protocols mentioned earlier in the
1198	list.
1199
1200	This variable can effectively be used for denial-of-service attacks
1201	against local programs (e.g. when setuid), although the impact is likely
1202	small, as the program has to handle connection errors already-
1203
1204	Examples: C<PERL_ANYEVENT_PROTOCOLS=ipv4,ipv6> - prefer IPv4 over IPv6,
1205	but support both and try to use both. C<PERL_ANYEVENT_PROTOCOLS=ipv4>
1206	- only support IPv4, never try to resolve or contact IPv6
1207	addresses. C<PERL_ANYEVENT_PROTOCOLS=ipv6,ipv4> support either IPv4 or
1208	IPv6, but prefer IPv6 over IPv4.
1209
1210	=item C<PERL_ANYEVENT_EDNS0>
1211
1212	Used by L<AnyEvent::DNS> to decide whether to use the EDNS0 extension
1213	for DNS. This extension is generally useful to reduce DNS traffic, but
1214	some (broken) firewalls drop such DNS packets, which is why it is off by
1215	default.
1216
1217	Setting this variable to C<1> will cause L<AnyEvent::DNS> to announce
1218	EDNS0 in its DNS requests.
1219
1220	=item C<PERL_ANYEVENT_MAX_FORKS>
1221
1222	The maximum number of child processes that C<AnyEvent::Util::fork_call>
1223	will create in parallel.
1224
1225	=back
1226		1957
1227	=head1 EXAMPLE PROGRAM	1958	=head1 EXAMPLE PROGRAM
1228		1959
1229	The following program uses an I/O watcher to read data from STDIN, a timer	1960	The following program uses an I/O watcher to read data from STDIN, a timer
1230	to display a message once per second, and a condition variable to quit the	1961	to display a message once per second, and a condition variable to quit the
…		…
1243	warn "read: $input\n"; # output what has been read	1974	warn "read: $input\n"; # output what has been read
1244	$cv->send if $input =~ /^q/i; # quit program if /^q/i	1975	$cv->send if $input =~ /^q/i; # quit program if /^q/i
1245	},	1976	},
1246	);	1977	);
1247		1978
1248	my $time_watcher; # can only be used once
1249
1250	sub new_timer {
1251	$timer = AnyEvent->timer (after => 1, cb => sub {	1979	my $time_watcher = AnyEvent->timer (after => 1, interval => 1, cb => sub {
1252	warn "timeout\n"; # print 'timeout' about every second	1980	warn "timeout\n"; # print 'timeout' at most every second
1253	&new_timer; # and restart the time
1254	});	1981	});
1255	}
1256
1257	new_timer; # create first timer
1258		1982
1259	$cv->recv; # wait until user enters /^q/i	1983	$cv->recv; # wait until user enters /^q/i
1260		1984
1261	=head1 REAL-WORLD EXAMPLE	1985	=head1 REAL-WORLD EXAMPLE
1262		1986
…		…
1393	through AnyEvent. The benchmark creates a lot of timers (with a zero	2117	through AnyEvent. The benchmark creates a lot of timers (with a zero
1394	timeout) and I/O watchers (watching STDOUT, a pty, to become writable,	2118	timeout) and I/O watchers (watching STDOUT, a pty, to become writable,
1395	which it is), lets them fire exactly once and destroys them again.	2119	which it is), lets them fire exactly once and destroys them again.
1396		2120
1397	Source code for this benchmark is found as F<eg/bench> in the AnyEvent	2121	Source code for this benchmark is found as F<eg/bench> in the AnyEvent
1398	distribution.	2122	distribution. It uses the L<AE> interface, which makes a real difference
		2123	for the EV and Perl backends only.
1399		2124
1400	=head3 Explanation of the columns	2125	=head3 Explanation of the columns
1401		2126
1402	I<watcher> is the number of event watchers created/destroyed. Since	2127	I<watcher> is the number of event watchers created/destroyed. Since
1403	different event models feature vastly different performances, each event	2128	different event models feature vastly different performances, each event
…		…
1424	watcher.	2149	watcher.
1425		2150
1426	=head3 Results	2151	=head3 Results
1427		2152
1428	name watchers bytes create invoke destroy comment	2153	name watchers bytes create invoke destroy comment
1429	EV/EV 400000 244 0.56 0.46 0.31 EV native interface	2154	EV/EV 100000 223 0.47 0.43 0.27 EV native interface
1430	EV/Any 100000 244 2.50 0.46 0.29 EV + AnyEvent watchers	2155	EV/Any 100000 223 0.48 0.42 0.26 EV + AnyEvent watchers
1431	CoroEV/Any 100000 244 2.49 0.44 0.29 coroutines + Coro::Signal	2156	Coro::EV/Any 100000 223 0.47 0.42 0.26 coroutines + Coro::Signal
1432	Perl/Any 100000 513 4.92 0.87 1.12 pure perl implementation	2157	Perl/Any 100000 431 2.70 0.74 0.92 pure perl implementation
1433	Event/Event 16000 516 31.88 31.30 0.85 Event native interface	2158	Event/Event 16000 516 31.16 31.84 0.82 Event native interface
1434	Event/Any 16000 590 35.75 31.42 1.08 Event + AnyEvent watchers	2159	Event/Any 16000 1203 42.61 34.79 1.80 Event + AnyEvent watchers
		2160	IOAsync/Any 16000 1911 41.92 27.45 16.81 via IO::Async::Loop::IO_Poll
		2161	IOAsync/Any 16000 1726 40.69 26.37 15.25 via IO::Async::Loop::Epoll
1435	Glib/Any 16000 1357 98.22 12.41 54.00 quadratic behaviour	2162	Glib/Any 16000 1118 89.00 12.57 51.17 quadratic behaviour
1436	Tk/Any 2000 1860 26.97 67.98 14.00 SEGV with >> 2000 watchers	2163	Tk/Any 2000 1346 20.96 10.75 8.00 SEGV with >> 2000 watchers
1437	POE/Event 2000 6644 108.64 736.02 14.73 via POE::Loop::Event	2164	POE/Any 2000 6951 108.97 795.32 14.24 via POE::Loop::Event
1438	POE/Select 2000 6343 94.13 809.12 565.96 via POE::Loop::Select	2165	POE/Any 2000 6648 94.79 774.40 575.51 via POE::Loop::Select
1439		2166
1440	=head3 Discussion	2167	=head3 Discussion
1441		2168
1442	The benchmark does I<not> measure scalability of the event loop very	2169	The benchmark does I<not> measure scalability of the event loop very
1443	well. For example, a select-based event loop (such as the pure perl one)	2170	well. For example, a select-based event loop (such as the pure perl one)
…		…
1455	benchmark machine, handling an event takes roughly 1600 CPU cycles with	2182	benchmark machine, handling an event takes roughly 1600 CPU cycles with
1456	EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU	2183	EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU
1457	cycles with POE.	2184	cycles with POE.
1458		2185
1459	C<EV> is the sole leader regarding speed and memory use, which are both	2186	C<EV> is the sole leader regarding speed and memory use, which are both
1460	maximal/minimal, respectively. Even when going through AnyEvent, it uses	2187	maximal/minimal, respectively. When using the L<AE> API there is zero
		2188	overhead (when going through the AnyEvent API create is about 5-6 times
		2189	slower, with other times being equal, so still uses far less memory than
1461	far less memory than any other event loop and is still faster than Event	2190	any other event loop and is still faster than Event natively).
1462	natively.
1463		2191
1464	The pure perl implementation is hit in a few sweet spots (both the	2192	The pure perl implementation is hit in a few sweet spots (both the
1465	constant timeout and the use of a single fd hit optimisations in the perl	2193	constant timeout and the use of a single fd hit optimisations in the perl
1466	interpreter and the backend itself). Nevertheless this shows that it	2194	interpreter and the backend itself). Nevertheless this shows that it
1467	adds very little overhead in itself. Like any select-based backend its	2195	adds very little overhead in itself. Like any select-based backend its
1468	performance becomes really bad with lots of file descriptors (and few of	2196	performance becomes really bad with lots of file descriptors (and few of
1469	them active), of course, but this was not subject of this benchmark.	2197	them active), of course, but this was not subject of this benchmark.
1470		2198
1471	The C<Event> module has a relatively high setup and callback invocation	2199	The C<Event> module has a relatively high setup and callback invocation
1472	cost, but overall scores in on the third place.	2200	cost, but overall scores in on the third place.
		2201
		2202	C<IO::Async> performs admirably well, about on par with C<Event>, even
		2203	when using its pure perl backend.
1473		2204
1474	C<Glib>'s memory usage is quite a bit higher, but it features a	2205	C<Glib>'s memory usage is quite a bit higher, but it features a
1475	faster callback invocation and overall ends up in the same class as	2206	faster callback invocation and overall ends up in the same class as
1476	C<Event>. However, Glib scales extremely badly, doubling the number of	2207	C<Event>. However, Glib scales extremely badly, doubling the number of
1477	watchers increases the processing time by more than a factor of four,	2208	watchers increases the processing time by more than a factor of four,
…		…
1538	In this benchmark, we use 10000 socket pairs (20000 sockets), of which 100	2269	In this benchmark, we use 10000 socket pairs (20000 sockets), of which 100
1539	(1%) are active. This mirrors the activity of large servers with many	2270	(1%) are active. This mirrors the activity of large servers with many
1540	connections, most of which are idle at any one point in time.	2271	connections, most of which are idle at any one point in time.
1541		2272
1542	Source code for this benchmark is found as F<eg/bench2> in the AnyEvent	2273	Source code for this benchmark is found as F<eg/bench2> in the AnyEvent
1543	distribution.	2274	distribution. It uses the L<AE> interface, which makes a real difference
		2275	for the EV and Perl backends only.
1544		2276
1545	=head3 Explanation of the columns	2277	=head3 Explanation of the columns
1546		2278
1547	I<sockets> is the number of sockets, and twice the number of "servers" (as	2279	I<sockets> is the number of sockets, and twice the number of "servers" (as
1548	each server has a read and write socket end).	2280	each server has a read and write socket end).
…		…
1555	it to another server. This includes deleting the old timeout and creating	2287	it to another server. This includes deleting the old timeout and creating
1556	a new one that moves the timeout into the future.	2288	a new one that moves the timeout into the future.
1557		2289
1558	=head3 Results	2290	=head3 Results
1559		2291
1560	name sockets create request	2292	name sockets create request
1561	EV 20000 69.01 11.16	2293	EV 20000 62.66 7.99
1562	Perl 20000 73.32 35.87	2294	Perl 20000 68.32 32.64
1563	Event 20000 212.62 257.32	2295	IOAsync 20000 174.06 101.15 epoll
1564	Glib 20000 651.16 1896.30	2296	IOAsync 20000 174.67 610.84 poll
		2297	Event 20000 202.69 242.91
		2298	Glib 20000 557.01 1689.52
1565	POE 20000 349.67 12317.24 uses POE::Loop::Event	2299	POE 20000 341.54 12086.32 uses POE::Loop::Event
1566		2300
1567	=head3 Discussion	2301	=head3 Discussion
1568		2302
1569	This benchmark I<does> measure scalability and overall performance of the	2303	This benchmark I<does> measure scalability and overall performance of the
1570	particular event loop.	2304	particular event loop.
…		…
1572	EV is again fastest. Since it is using epoll on my system, the setup time	2306	EV is again fastest. Since it is using epoll on my system, the setup time
1573	is relatively high, though.	2307	is relatively high, though.
1574		2308
1575	Perl surprisingly comes second. It is much faster than the C-based event	2309	Perl surprisingly comes second. It is much faster than the C-based event
1576	loops Event and Glib.	2310	loops Event and Glib.
		2311
		2312	IO::Async performs very well when using its epoll backend, and still quite
		2313	good compared to Glib when using its pure perl backend.
1577		2314
1578	Event suffers from high setup time as well (look at its code and you will	2315	Event suffers from high setup time as well (look at its code and you will
1579	understand why). Callback invocation also has a high overhead compared to	2316	understand why). Callback invocation also has a high overhead compared to
1580	the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event	2317	the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event
1581	uses select or poll in basically all documented configurations.	2318	uses select or poll in basically all documented configurations.
…		…
1644	=item * C-based event loops perform very well with small number of	2381	=item * C-based event loops perform very well with small number of
1645	watchers, as the management overhead dominates.	2382	watchers, as the management overhead dominates.
1646		2383
1647	=back	2384	=back
1648		2385
		2386	=head2 THE IO::Lambda BENCHMARK
		2387
		2388	Recently I was told about the benchmark in the IO::Lambda manpage, which
		2389	could be misinterpreted to make AnyEvent look bad. In fact, the benchmark
		2390	simply compares IO::Lambda with POE, and IO::Lambda looks better (which
		2391	shouldn't come as a surprise to anybody). As such, the benchmark is
		2392	fine, and mostly shows that the AnyEvent backend from IO::Lambda isn't
		2393	very optimal. But how would AnyEvent compare when used without the extra
		2394	baggage? To explore this, I wrote the equivalent benchmark for AnyEvent.
		2395
		2396	The benchmark itself creates an echo-server, and then, for 500 times,
		2397	connects to the echo server, sends a line, waits for the reply, and then
		2398	creates the next connection. This is a rather bad benchmark, as it doesn't
		2399	test the efficiency of the framework or much non-blocking I/O, but it is a
		2400	benchmark nevertheless.
		2401
		2402	name runtime
		2403	Lambda/select 0.330 sec
		2404	+ optimized 0.122 sec
		2405	Lambda/AnyEvent 0.327 sec
		2406	+ optimized 0.138 sec
		2407	Raw sockets/select 0.077 sec
		2408	POE/select, components 0.662 sec
		2409	POE/select, raw sockets 0.226 sec
		2410	POE/select, optimized 0.404 sec
		2411
		2412	AnyEvent/select/nb 0.085 sec
		2413	AnyEvent/EV/nb 0.068 sec
		2414	+state machine 0.134 sec
		2415
		2416	The benchmark is also a bit unfair (my fault): the IO::Lambda/POE
		2417	benchmarks actually make blocking connects and use 100% blocking I/O,
		2418	defeating the purpose of an event-based solution. All of the newly
		2419	written AnyEvent benchmarks use 100% non-blocking connects (using
		2420	AnyEvent::Socket::tcp_connect and the asynchronous pure perl DNS
		2421	resolver), so AnyEvent is at a disadvantage here, as non-blocking connects
		2422	generally require a lot more bookkeeping and event handling than blocking
		2423	connects (which involve a single syscall only).
		2424
		2425	The last AnyEvent benchmark additionally uses L<AnyEvent::Handle>, which
		2426	offers similar expressive power as POE and IO::Lambda, using conventional
		2427	Perl syntax. This means that both the echo server and the client are 100%
		2428	non-blocking, further placing it at a disadvantage.
		2429
		2430	As you can see, the AnyEvent + EV combination even beats the
		2431	hand-optimised "raw sockets benchmark", while AnyEvent + its pure perl
		2432	backend easily beats IO::Lambda and POE.
		2433
		2434	And even the 100% non-blocking version written using the high-level (and
		2435	slow :) L<AnyEvent::Handle> abstraction beats both POE and IO::Lambda
		2436	higher level ("unoptimised") abstractions by a large margin, even though
		2437	it does all of DNS, tcp-connect and socket I/O in a non-blocking way.
		2438
		2439	The two AnyEvent benchmarks programs can be found as F<eg/ae0.pl> and
		2440	F<eg/ae2.pl> in the AnyEvent distribution, the remaining benchmarks are
		2441	part of the IO::Lambda distribution and were used without any changes.
		2442
		2443
		2444	=head1 SIGNALS
		2445
		2446	AnyEvent currently installs handlers for these signals:
		2447
		2448	=over 4
		2449
		2450	=item SIGCHLD
		2451
		2452	A handler for C<SIGCHLD> is installed by AnyEvent's child watcher
		2453	emulation for event loops that do not support them natively. Also, some
		2454	event loops install a similar handler.
		2455
		2456	Additionally, when AnyEvent is loaded and SIGCHLD is set to IGNORE, then
		2457	AnyEvent will reset it to default, to avoid losing child exit statuses.
		2458
		2459	=item SIGPIPE
		2460
		2461	A no-op handler is installed for C<SIGPIPE> when C<$SIG{PIPE}> is C<undef>
		2462	when AnyEvent gets loaded.
		2463
		2464	The rationale for this is that AnyEvent users usually do not really depend
		2465	on SIGPIPE delivery (which is purely an optimisation for shell use, or
		2466	badly-written programs), but C<SIGPIPE> can cause spurious and rare
		2467	program exits as a lot of people do not expect C<SIGPIPE> when writing to
		2468	some random socket.
		2469
		2470	The rationale for installing a no-op handler as opposed to ignoring it is
		2471	that this way, the handler will be restored to defaults on exec.
		2472
		2473	Feel free to install your own handler, or reset it to defaults.
		2474
		2475	=back
		2476
		2477	=cut
		2478
		2479	undef $SIG{CHLD}
		2480	if $SIG{CHLD} eq 'IGNORE';
		2481
		2482	$SIG{PIPE} = sub { }
		2483	unless defined $SIG{PIPE};
		2484
		2485	=head1 RECOMMENDED/OPTIONAL MODULES
		2486
		2487	One of AnyEvent's main goals is to be 100% Pure-Perl(tm): only perl (and
		2488	it's built-in modules) are required to use it.
		2489
		2490	That does not mean that AnyEvent won't take advantage of some additional
		2491	modules if they are installed.
		2492
		2493	This section explains which additional modules will be used, and how they
		2494	affect AnyEvent's operation.
		2495
		2496	=over 4
		2497
		2498	=item L<Async::Interrupt>
		2499
		2500	This slightly arcane module is used to implement fast signal handling: To
		2501	my knowledge, there is no way to do completely race-free and quick
		2502	signal handling in pure perl. To ensure that signals still get
		2503	delivered, AnyEvent will start an interval timer to wake up perl (and
		2504	catch the signals) with some delay (default is 10 seconds, look for
		2505	C<$AnyEvent::MAX_SIGNAL_LATENCY>).
		2506
		2507	If this module is available, then it will be used to implement signal
		2508	catching, which means that signals will not be delayed, and the event loop
		2509	will not be interrupted regularly, which is more efficient (and good for
		2510	battery life on laptops).
		2511
		2512	This affects not just the pure-perl event loop, but also other event loops
		2513	that have no signal handling on their own (e.g. Glib, Tk, Qt).
		2514
		2515	Some event loops (POE, Event, Event::Lib) offer signal watchers natively,
		2516	and either employ their own workarounds (POE) or use AnyEvent's workaround
		2517	(using C<$AnyEvent::MAX_SIGNAL_LATENCY>). Installing L<Async::Interrupt>
		2518	does nothing for those backends.
		2519
		2520	=item L<EV>
		2521
		2522	This module isn't really "optional", as it is simply one of the backend
		2523	event loops that AnyEvent can use. However, it is simply the best event
		2524	loop available in terms of features, speed and stability: It supports
		2525	the AnyEvent API optimally, implements all the watcher types in XS, does
		2526	automatic timer adjustments even when no monotonic clock is available,
		2527	can take avdantage of advanced kernel interfaces such as C<epoll> and
		2528	C<kqueue>, and is the fastest backend I<by far>. You can even embed
		2529	L<Glib>/L<Gtk2> in it (or vice versa, see L<EV::Glib> and L<Glib::EV>).
		2530
		2531	=item L<Guard>
		2532
		2533	The guard module, when used, will be used to implement
		2534	C<AnyEvent::Util::guard>. This speeds up guards considerably (and uses a
		2535	lot less memory), but otherwise doesn't affect guard operation much. It is
		2536	purely used for performance.
		2537
		2538	=item L<JSON> and L<JSON::XS>
		2539
		2540	One of these modules is required when you want to read or write JSON data
		2541	via L<AnyEvent::Handle>. It is also written in pure-perl, but can take
		2542	advantage of the ultra-high-speed L<JSON::XS> module when it is installed.
		2543
		2544	In fact, L<AnyEvent::Handle> will use L<JSON::XS> by default if it is
		2545	installed.
		2546
		2547	=item L<Net::SSLeay>
		2548
		2549	Implementing TLS/SSL in Perl is certainly interesting, but not very
		2550	worthwhile: If this module is installed, then L<AnyEvent::Handle> (with
		2551	the help of L<AnyEvent::TLS>), gains the ability to do TLS/SSL.
		2552
		2553	=item L<Time::HiRes>
		2554
		2555	This module is part of perl since release 5.008. It will be used when the
		2556	chosen event library does not come with a timing source on it's own. The
		2557	pure-perl event loop (L<AnyEvent::Impl::Perl>) will additionally use it to
		2558	try to use a monotonic clock for timing stability.
		2559
		2560	=back
		2561
1649		2562
1650	=head1 FORK	2563	=head1 FORK
1651		2564
1652	Most event libraries are not fork-safe. The ones who are usually are	2565	Most event libraries are not fork-safe. The ones who are usually are
1653	because they rely on inefficient but fork-safe C<select> or C<poll>	2566	because they rely on inefficient but fork-safe C<select> or C<poll> calls
1654	calls. Only L<EV> is fully fork-aware.	2567	- higher performance APIs such as BSD's kqueue or the dreaded Linux epoll
		2568	are usually badly thought-out hacks that are incompatible with fork in
		2569	one way or another. Only L<EV> is fully fork-aware and ensures that you
		2570	continue event-processing in both parent and child (or both, if you know
		2571	what you are doing).
		2572
		2573	This means that, in general, you cannot fork and do event processing in
		2574	the child if the event library was initialised before the fork (which
		2575	usually happens when the first AnyEvent watcher is created, or the library
		2576	is loaded).
1655		2577
1656	If you have to fork, you must either do so I<before> creating your first	2578	If you have to fork, you must either do so I<before> creating your first
1657	watcher OR you must not use AnyEvent at all in the child.	2579	watcher OR you must not use AnyEvent at all in the child OR you must do
		2580	something completely out of the scope of AnyEvent.
		2581
		2582	The problem of doing event processing in the parent I<and> the child
		2583	is much more complicated: even for backends that I<are> fork-aware or
		2584	fork-safe, their behaviour is not usually what you want: fork clones all
		2585	watchers, that means all timers, I/O watchers etc. are active in both
		2586	parent and child, which is almost never what you want. USing C<exec>
		2587	to start worker children from some kind of manage rprocess is usually
		2588	preferred, because it is much easier and cleaner, at the expense of having
		2589	to have another binary.
1658		2590
1659		2591
1660	=head1 SECURITY CONSIDERATIONS	2592	=head1 SECURITY CONSIDERATIONS
1661		2593
1662	AnyEvent can be forced to load any event model via	2594	AnyEvent can be forced to load any event model via
…		…
1673		2605
1674	use AnyEvent;	2606	use AnyEvent;
1675		2607
1676	Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can	2608	Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can
1677	be used to probe what backend is used and gain other information (which is	2609	be used to probe what backend is used and gain other information (which is
1678	probably even less useful to an attacker than PERL_ANYEVENT_MODEL).	2610	probably even less useful to an attacker than PERL_ANYEVENT_MODEL), and
		2611	$ENV{PERL_ANYEVENT_STRICT}.
		2612
		2613	Note that AnyEvent will remove I<all> environment variables starting with
		2614	C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
		2615	enabled.
1679		2616
1680		2617
1681	=head1 BUGS	2618	=head1 BUGS
1682		2619
1683	Perl 5.8 has numerous memleaks that sometimes hit this module and are hard	2620	Perl 5.8 has numerous memleaks that sometimes hit this module and are hard
1684	to work around. If you suffer from memleaks, first upgrade to Perl 5.10	2621	to work around. If you suffer from memleaks, first upgrade to Perl 5.10
1685	and check wether the leaks still show up. (Perl 5.10.0 has other annoying	2622	and check wether the leaks still show up. (Perl 5.10.0 has other annoying
1686	mamleaks, such as leaking on C<map> and C<grep> but it is usually not as	2623	memleaks, such as leaking on C<map> and C<grep> but it is usually not as
1687	pronounced).	2624	pronounced).
1688		2625
1689		2626
1690	=head1 SEE ALSO	2627	=head1 SEE ALSO
1691		2628
…		…
1695	L<Glib>, L<Tk>, L<Event::Lib>, L<Qt>, L<POE>.	2632	L<Glib>, L<Tk>, L<Event::Lib>, L<Qt>, L<POE>.
1696		2633
1697	Implementations: L<AnyEvent::Impl::EV>, L<AnyEvent::Impl::Event>,	2634	Implementations: L<AnyEvent::Impl::EV>, L<AnyEvent::Impl::Event>,
1698	L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>,	2635	L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>,
1699	L<AnyEvent::Impl::EventLib>, L<AnyEvent::Impl::Qt>,	2636	L<AnyEvent::Impl::EventLib>, L<AnyEvent::Impl::Qt>,
1700	L<AnyEvent::Impl::POE>.	2637	L<AnyEvent::Impl::POE>, L<AnyEvent::Impl::IOAsync>, L<Anyevent::Impl::Irssi>.
1701		2638
1702	Non-blocking file handles, sockets, TCP clients and	2639	Non-blocking file handles, sockets, TCP clients and
1703	servers: L<AnyEvent::Handle>, L<AnyEvent::Socket>.	2640	servers: L<AnyEvent::Handle>, L<AnyEvent::Socket>, L<AnyEvent::TLS>.
1704		2641
1705	Asynchronous DNS: L<AnyEvent::DNS>.	2642	Asynchronous DNS: L<AnyEvent::DNS>.
1706		2643
1707	Coroutine support: L<Coro>, L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>,	2644	Coroutine support: L<Coro>, L<Coro::AnyEvent>, L<Coro::EV>,
		2645	L<Coro::Event>,
1708		2646
1709	Nontrivial usage examples: L<Net::FCP>, L<Net::XMPP2>, L<AnyEvent::DNS>.	2647	Nontrivial usage examples: L<AnyEvent::GPSD>, L<AnyEvent::XMPP>,
		2648	L<AnyEvent::HTTP>.
1710		2649
1711		2650
1712	=head1 AUTHOR	2651	=head1 AUTHOR
1713		2652
1714	Marc Lehmann <schmorp@schmorp.de>	2653	Marc Lehmann <schmorp@schmorp.de>

Diff Legend

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