ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/AnyEvent/lib/AnyEvent.pm

Comparing AnyEvent/lib/AnyEvent.pm (file contents):
Revision 1.75 by root, Fri Apr 25 07:49:39 2008 UTC vs.
Revision 1.240 by root, Fri Jul 17 14:57:03 2009 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	AnyEvent - provide framework for multiple event loops	3	AnyEvent - provide framework for multiple event loops
4		4
5	EV, Event, Coro::EV, Coro::Event, Glib, Tk, Perl, Event::Lib, Qt, POE - various supported event loops	5	EV, Event, Glib, Tk, Perl, Event::Lib, Qt and POE are various supported
		6	event loops.
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
		35	$w->send; # wake up current and all future recv's
20	$w->wait; # enters "main loop" till $condvar gets ->broadcast	36	$w->recv; # enters "main loop" till $condvar gets ->send
21	$w->broadcast; # wake up current and all future wait's	37	# use a condvar in callback mode:
		38	$w->cb (sub { $_[0]->recv });
		39
		40	=head1 INTRODUCTION/TUTORIAL
		41
		42	This manpage is mainly a reference manual. If you are interested
		43	in a tutorial or some gentle introduction, have a look at the
		44	L<AnyEvent::Intro> manpage.
22		45
23	=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)	46	=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)
24		47
25	Glib, POE, IO::Async, Event... CPAN offers event models by the dozen	48	Glib, POE, IO::Async, Event... CPAN offers event models by the dozen
26	nowadays. So what is different about AnyEvent?	49	nowadays. So what is different about AnyEvent?
27		50
28	Executive Summary: AnyEvent is I<compatible>, AnyEvent is I<free of	51	Executive Summary: AnyEvent is I<compatible>, AnyEvent is I<free of
29	policy> and AnyEvent is I<small and efficient>.	52	policy> and AnyEvent is I<small and efficient>.
30		53
31	First and foremost, I<AnyEvent is not an event model> itself, it only	54	First and foremost, I<AnyEvent is not an event model> itself, it only
32	interfaces to whatever event model the main program happens to use in a	55	interfaces to whatever event model the main program happens to use, in a
33	pragmatic way. For event models and certain classes of immortals alike,	56	pragmatic way. For event models and certain classes of immortals alike,
34	the statement "there can only be one" is a bitter reality: In general,	57	the statement "there can only be one" is a bitter reality: In general,
35	only one event loop can be active at the same time in a process. AnyEvent	58	only one event loop can be active at the same time in a process. AnyEvent
36	helps hiding the differences between those event loops.	59	cannot change this, but it can hide the differences between those event
		60	loops.
37		61
38	The goal of AnyEvent is to offer module authors the ability to do event	62	The goal of AnyEvent is to offer module authors the ability to do event
39	programming (waiting for I/O or timer events) without subscribing to a	63	programming (waiting for I/O or timer events) without subscribing to a
40	religion, a way of living, and most importantly: without forcing your	64	religion, a way of living, and most importantly: without forcing your
41	module users into the same thing by forcing them to use the same event	65	module users into the same thing by forcing them to use the same event
42	model you use.	66	model you use.
43		67
44	For modules like POE or IO::Async (which is a total misnomer as it is	68	For modules like POE or IO::Async (which is a total misnomer as it is
45	actually doing all I/O I<synchronously>...), using them in your module is	69	actually doing all I/O I<synchronously>...), using them in your module is
46	like joining a cult: After you joined, you are dependent on them and you	70	like joining a cult: After you joined, you are dependent on them and you
47	cannot use anything else, as it is simply incompatible to everything that	71	cannot use anything else, as they are simply incompatible to everything
48	isn't itself. What's worse, all the potential users of your module are	72	that isn't them. What's worse, all the potential users of your
49	I<also> forced to use the same event loop you use.	73	module are I<also> forced to use the same event loop you use.
50		74
51	AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works	75	AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works
52	fine. AnyEvent + Tk works fine etc. etc. but none of these work together	76	fine. AnyEvent + Tk works fine etc. etc. but none of these work together
53	with the rest: POE + IO::Async? no go. Tk + Event? no go. Again: if	77	with the rest: POE + IO::Async? No go. Tk + Event? No go. Again: if
54	your module uses one of those, every user of your module has to use it,	78	your module uses one of those, every user of your module has to use it,
55	too. But if your module uses AnyEvent, it works transparently with all	79	too. But if your module uses AnyEvent, it works transparently with all
56	event models it supports (including stuff like POE and IO::Async, as long	80	event models it supports (including stuff like IO::Async, as long as those
57	as those use one of the supported event loops. It is trivial to add new	81	use one of the supported event loops. It is trivial to add new event loops
58	event loops to AnyEvent, too, so it is future-proof).	82	to AnyEvent, too, so it is future-proof).
59		83
60	In addition to being free of having to use I<the one and only true event	84	In addition to being free of having to use I<the one and only true event
61	model>, AnyEvent also is free of bloat and policy: with POE or similar	85	model>, AnyEvent also is free of bloat and policy: with POE or similar
62	modules, you get an enourmous amount of code and strict rules you have to	86	modules, you get an enormous amount of code and strict rules you have to
63	follow. AnyEvent, on the other hand, is lean and up to the point, by only	87	follow. AnyEvent, on the other hand, is lean and up to the point, by only
64	offering the functionality that is necessary, in as thin as a wrapper as	88	offering the functionality that is necessary, in as thin as a wrapper as
65	technically possible.	89	technically possible.
66		90
		91	Of course, AnyEvent comes with a big (and fully optional!) toolbox
		92	of useful functionality, such as an asynchronous DNS resolver, 100%
		93	non-blocking connects (even with TLS/SSL, IPv6 and on broken platforms
		94	such as Windows) and lots of real-world knowledge and workarounds for
		95	platform bugs and differences.
		96
67	Of course, if you want lots of policy (this can arguably be somewhat	97	Now, if you I<do want> lots of policy (this can arguably be somewhat
68	useful) and you want to force your users to use the one and only event	98	useful) and you want to force your users to use the one and only event
69	model, you should I<not> use this module.	99	model, you should I<not> use this module.
70
71		100
72	=head1 DESCRIPTION	101	=head1 DESCRIPTION
73		102
74	L<AnyEvent> provides an identical interface to multiple event loops. This	103	L<AnyEvent> provides an identical interface to multiple event loops. This
75	allows module authors to utilise an event loop without forcing module	104	allows module authors to utilise an event loop without forcing module
…		…
79	The interface itself is vaguely similar, but not identical to the L<Event>	108	The interface itself is vaguely similar, but not identical to the L<Event>
80	module.	109	module.
81		110
82	During the first call of any watcher-creation method, the module tries	111	During the first call of any watcher-creation method, the module tries
83	to detect the currently loaded event loop by probing whether one of the	112	to detect the currently loaded event loop by probing whether one of the
84	following modules is already loaded: L<Coro::EV>, L<Coro::Event>, L<EV>,	113	following modules is already loaded: L<EV>,
85	L<Event>, L<Glib>, L<Tk>, L<AnyEvent::Impl::Perl>, L<Event::Lib>, L<Qt>,	114	L<Event>, L<Glib>, L<AnyEvent::Impl::Perl>, L<Tk>, L<Event::Lib>, L<Qt>,
86	L<POE>. The first one found is used. If none are found, the module tries	115	L<POE>. The first one found is used. If none are found, the module tries
87	to load these modules (excluding Event::Lib, Qt and POE as the pure perl	116	to load these modules (excluding Tk, Event::Lib, Qt and POE as the pure perl
88	adaptor should always succeed) in the order given. The first one that can	117	adaptor should always succeed) in the order given. The first one that can
89	be successfully loaded will be used. If, after this, still none could be	118	be successfully loaded will be used. If, after this, still none could be
90	found, AnyEvent will fall back to a pure-perl event loop, which is not	119	found, AnyEvent will fall back to a pure-perl event loop, which is not
91	very efficient, but should work everywhere.	120	very efficient, but should work everywhere.
92		121
…		…
103	starts using it, all bets are off. Maybe you should tell their authors to	132	starts using it, all bets are off. Maybe you should tell their authors to
104	use AnyEvent so their modules work together with others seamlessly...	133	use AnyEvent so their modules work together with others seamlessly...
105		134
106	The pure-perl implementation of AnyEvent is called	135	The pure-perl implementation of AnyEvent is called
107	C<AnyEvent::Impl::Perl>. Like other event modules you can load it	136	C<AnyEvent::Impl::Perl>. Like other event modules you can load it
108	explicitly.	137	explicitly and enjoy the high availability of that event loop :)
109		138
110	=head1 WATCHERS	139	=head1 WATCHERS
111		140
112	AnyEvent has the central concept of a I<watcher>, which is an object that	141	AnyEvent has the central concept of a I<watcher>, which is an object that
113	stores relevant data for each kind of event you are waiting for, such as	142	stores relevant data for each kind of event you are waiting for, such as
114	the callback to call, the filehandle to watch, etc.	143	the callback to call, the file handle to watch, etc.
115		144
116	These watchers are normal Perl objects with normal Perl lifetime. After	145	These watchers are normal Perl objects with normal Perl lifetime. After
117	creating a watcher it will immediately "watch" for events and invoke the	146	creating a watcher it will immediately "watch" for events and invoke the
118	callback when the event occurs (of course, only when the event model	147	callback when the event occurs (of course, only when the event model
119	is in control).	148	is in control).
120		149
		150	Note that B<callbacks must not permanently change global variables>
		151	potentially in use by the event loop (such as C<$_> or C<$[>) and that B<<
		152	callbacks must not C<die> >>. The former is good programming practise in
		153	Perl and the latter stems from the fact that exception handling differs
		154	widely between event loops.
		155
121	To disable the watcher you have to destroy it (e.g. by setting the	156	To disable the watcher you have to destroy it (e.g. by setting the
122	variable you store it in to C<undef> or otherwise deleting all references	157	variable you store it in to C<undef> or otherwise deleting all references
123	to it).	158	to it).
124		159
125	All watchers are created by calling a method on the C<AnyEvent> class.	160	All watchers are created by calling a method on the C<AnyEvent> class.
…		…
127	Many watchers either are used with "recursion" (repeating timers for	162	Many watchers either are used with "recursion" (repeating timers for
128	example), or need to refer to their watcher object in other ways.	163	example), or need to refer to their watcher object in other ways.
129		164
130	An any way to achieve that is this pattern:	165	An any way to achieve that is this pattern:
131		166
132	my $w; $w = AnyEvent->type (arg => value ..., cb => sub {	167	my $w; $w = AnyEvent->type (arg => value ..., cb => sub {
133	# you can use $w here, for example to undef it	168	# you can use $w here, for example to undef it
134	undef $w;	169	undef $w;
135	});	170	});
136		171
137	Note that C<my $w; $w => combination. This is necessary because in Perl,	172	Note that C<my $w; $w => combination. This is necessary because in Perl,
138	my variables are only visible after the statement in which they are	173	my variables are only visible after the statement in which they are
139	declared.	174	declared.
140		175
141	=head2 IO WATCHERS	176	=head2 I/O WATCHERS
142		177
143	You can create an I/O watcher by calling the C<< AnyEvent->io >> method	178	You can create an I/O watcher by calling the C<< AnyEvent->io >> method
144	with the following mandatory key-value pairs as arguments:	179	with the following mandatory key-value pairs as arguments:
145		180
146	C<fh> the Perl I<file handle> (I<not> file descriptor) to watch for	181	C<fh> is the Perl I<file handle> (or a naked file descriptor) to watch
		182	for events (AnyEvent might or might not keep a reference to this file
		183	handle). Note that only file handles pointing to things for which
		184	non-blocking operation makes sense are allowed. This includes sockets,
		185	most character devices, pipes, fifos and so on, but not for example files
		186	or block devices.
		187
147	events. C<poll> must be a string that is either C<r> or C<w>, which	188	C<poll> must be a string that is either C<r> or C<w>, which creates a
148	creates a watcher waiting for "r"eadable or "w"ritable events,	189	watcher waiting for "r"eadable or "w"ritable events, respectively.
		190
149	respectively. C<cb> is the callback to invoke each time the file handle	191	C<cb> is the callback to invoke each time the file handle becomes ready.
150	becomes ready.
151		192
152	As long as the I/O watcher exists it will keep the file descriptor or a	193	Although the callback might get passed parameters, their value and
153	copy of it alive/open.	194	presence is undefined and you cannot rely on them. Portable AnyEvent
		195	callbacks cannot use arguments passed to I/O watcher callbacks.
154		196
		197	The I/O watcher might use the underlying file descriptor or a copy of it.
155	It is not allowed to close a file handle as long as any watcher is active	198	You must not close a file handle as long as any watcher is active on the
156	on the underlying file descriptor.	199	underlying file descriptor.
157		200
158	Some event loops issue spurious readyness notifications, so you should	201	Some event loops issue spurious readyness notifications, so you should
159	always use non-blocking calls when reading/writing from/to your file	202	always use non-blocking calls when reading/writing from/to your file
160	handles.	203	handles.
161		204
162	Example:
163
164	# wait for readability of STDIN, then read a line and disable the watcher	205	Example: wait for readability of STDIN, then read a line and disable the
		206	watcher.
		207
165	my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {	208	my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
166	chomp (my $input = <STDIN>);	209	chomp (my $input = <STDIN>);
167	warn "read: $input\n";	210	warn "read: $input\n";
168	undef $w;	211	undef $w;
169	});	212	});
…		…
172		215
173	You can create a time watcher by calling the C<< AnyEvent->timer >>	216	You can create a time watcher by calling the C<< AnyEvent->timer >>
174	method with the following mandatory arguments:	217	method with the following mandatory arguments:
175		218
176	C<after> specifies after how many seconds (fractional values are	219	C<after> specifies after how many seconds (fractional values are
177	supported) should the timer activate. C<cb> the callback to invoke in that	220	supported) the callback should be invoked. C<cb> is the callback to invoke
178	case.	221	in that case.
179		222
180	The timer callback will be invoked at most once: if you want a repeating	223	Although the callback might get passed parameters, their value and
181	timer you have to create a new watcher (this is a limitation by both Tk	224	presence is undefined and you cannot rely on them. Portable AnyEvent
182	and Glib).	225	callbacks cannot use arguments passed to time watcher callbacks.
183		226
184	Example:	227	The callback will normally be invoked once only. If you specify another
		228	parameter, C<interval>, as a strictly positive number (> 0), then the
		229	callback will be invoked regularly at that interval (in fractional
		230	seconds) after the first invocation. If C<interval> is specified with a
		231	false value, then it is treated as if it were missing.
185		232
		233	The callback will be rescheduled before invoking the callback, but no
		234	attempt is done to avoid timer drift in most backends, so the interval is
		235	only approximate.
		236
186	# fire an event after 7.7 seconds	237	Example: fire an event after 7.7 seconds.
		238
187	my $w = AnyEvent->timer (after => 7.7, cb => sub {	239	my $w = AnyEvent->timer (after => 7.7, cb => sub {
188	warn "timeout\n";	240	warn "timeout\n";
189	});	241	});
190		242
191	# to cancel the timer:	243	# to cancel the timer:
192	undef $w;	244	undef $w;
193		245
194	Example 2:
195
196	# fire an event after 0.5 seconds, then roughly every second	246	Example 2: fire an event after 0.5 seconds, then roughly every second.
197	my $w;
198		247
199	my $cb = sub {
200	# cancel the old timer while creating a new one
201	$w = AnyEvent->timer (after => 1, cb => $cb);	248	my $w = AnyEvent->timer (after => 0.5, interval => 1, cb => sub {
		249	warn "timeout\n";
202	};	250	};
203
204	# start the "loop" by creating the first watcher
205	$w = AnyEvent->timer (after => 0.5, cb => $cb);
206		251
207	=head3 TIMING ISSUES	252	=head3 TIMING ISSUES
208		253
209	There are two ways to handle timers: based on real time (relative, "fire	254	There are two ways to handle timers: based on real time (relative, "fire
210	in 10 seconds") and based on wallclock time (absolute, "fire at 12	255	in 10 seconds") and based on wallclock time (absolute, "fire at 12
…		…
222	timers.	267	timers.
223		268
224	AnyEvent always prefers relative timers, if available, matching the	269	AnyEvent always prefers relative timers, if available, matching the
225	AnyEvent API.	270	AnyEvent API.
226		271
		272	AnyEvent has two additional methods that return the "current time":
		273
		274	=over 4
		275
		276	=item AnyEvent->time
		277
		278	This returns the "current wallclock time" as a fractional number of
		279	seconds since the Epoch (the same thing as C<time> or C<Time::HiRes::time>
		280	return, and the result is guaranteed to be compatible with those).
		281
		282	It progresses independently of any event loop processing, i.e. each call
		283	will check the system clock, which usually gets updated frequently.
		284
		285	=item AnyEvent->now
		286
		287	This also returns the "current wallclock time", but unlike C<time>, above,
		288	this value might change only once per event loop iteration, depending on
		289	the event loop (most return the same time as C<time>, above). This is the
		290	time that AnyEvent's timers get scheduled against.
		291
		292	I<In almost all cases (in all cases if you don't care), this is the
		293	function to call when you want to know the current time.>
		294
		295	This function is also often faster then C<< AnyEvent->time >>, and
		296	thus the preferred method if you want some timestamp (for example,
		297	L<AnyEvent::Handle> uses this to update it's activity timeouts).
		298
		299	The rest of this section is only of relevance if you try to be very exact
		300	with your timing, you can skip it without bad conscience.
		301
		302	For a practical example of when these times differ, consider L<Event::Lib>
		303	and L<EV> and the following set-up:
		304
		305	The event loop is running and has just invoked one of your callback at
		306	time=500 (assume no other callbacks delay processing). In your callback,
		307	you wait a second by executing C<sleep 1> (blocking the process for a
		308	second) and then (at time=501) you create a relative timer that fires
		309	after three seconds.
		310
		311	With L<Event::Lib>, C<< AnyEvent->time >> and C<< AnyEvent->now >> will
		312	both return C<501>, because that is the current time, and the timer will
		313	be scheduled to fire at time=504 (C<501> + C<3>).
		314
		315	With L<EV>, C<< AnyEvent->time >> returns C<501> (as that is the current
		316	time), but C<< AnyEvent->now >> returns C<500>, as that is the time the
		317	last event processing phase started. With L<EV>, your timer gets scheduled
		318	to run at time=503 (C<500> + C<3>).
		319
		320	In one sense, L<Event::Lib> is more exact, as it uses the current time
		321	regardless of any delays introduced by event processing. However, most
		322	callbacks do not expect large delays in processing, so this causes a
		323	higher drift (and a lot more system calls to get the current time).
		324
		325	In another sense, L<EV> is more exact, as your timer will be scheduled at
		326	the same time, regardless of how long event processing actually took.
		327
		328	In either case, if you care (and in most cases, you don't), then you
		329	can get whatever behaviour you want with any event loop, by taking the
		330	difference between C<< AnyEvent->time >> and C<< AnyEvent->now >> into
		331	account.
		332
		333	=item AnyEvent->now_update
		334
		335	Some event loops (such as L<EV> or L<AnyEvent::Impl::Perl>) cache
		336	the current time for each loop iteration (see the discussion of L<<
		337	AnyEvent->now >>, above).
		338
		339	When a callback runs for a long time (or when the process sleeps), then
		340	this "current" time will differ substantially from the real time, which
		341	might affect timers and time-outs.
		342
		343	When this is the case, you can call this method, which will update the
		344	event loop's idea of "current time".
		345
		346	Note that updating the time I<might> cause some events to be handled.
		347
		348	=back
		349
227	=head2 SIGNAL WATCHERS	350	=head2 SIGNAL WATCHERS
228		351
229	You can watch for signals using a signal watcher, C<signal> is the signal	352	You can watch for signals using a signal watcher, C<signal> is the signal
230	I<name> without any C<SIG> prefix, C<cb> is the Perl callback to	353	I<name> in uppercase and without any C<SIG> prefix, C<cb> is the Perl
231	be invoked whenever a signal occurs.	354	callback to be invoked whenever a signal occurs.
232		355
		356	Although the callback might get passed parameters, their value and
		357	presence is undefined and you cannot rely on them. Portable AnyEvent
		358	callbacks cannot use arguments passed to signal watcher callbacks.
		359
233	Multiple signal occurances can be clumped together into one callback	360	Multiple signal occurrences can be clumped together into one callback
234	invocation, and callback invocation will be synchronous. synchronous means	361	invocation, and callback invocation will be synchronous. Synchronous means
235	that it might take a while until the signal gets handled by the process,	362	that it might take a while until the signal gets handled by the process,
236	but it is guarenteed not to interrupt any other callbacks.	363	but it is guaranteed not to interrupt any other callbacks.
237		364
238	The main advantage of using these watchers is that you can share a signal	365	The main advantage of using these watchers is that you can share a signal
239	between multiple watchers.	366	between multiple watchers.
240		367
241	This watcher might use C<%SIG>, so programs overwriting those signals	368	This watcher might use C<%SIG>, so programs overwriting those signals
…		…
248	=head2 CHILD PROCESS WATCHERS	375	=head2 CHILD PROCESS WATCHERS
249		376
250	You can also watch on a child process exit and catch its exit status.	377	You can also watch on a child process exit and catch its exit status.
251		378
252	The child process is specified by the C<pid> argument (if set to C<0>, it	379	The child process is specified by the C<pid> argument (if set to C<0>, it
253	watches for any child process exit). The watcher will trigger as often	380	watches for any child process exit). The watcher will triggered only when
254	as status change for the child are received. This works by installing a	381	the child process has finished and an exit status is available, not on
255	signal handler for C<SIGCHLD>. The callback will be called with the pid	382	any trace events (stopped/continued).
256	and exit status (as returned by waitpid).
257		383
258	Example: wait for pid 1333	384	The callback will be called with the pid and exit status (as returned by
		385	waitpid), so unlike other watcher types, you I<can> rely on child watcher
		386	callback arguments.
259		387
		388	This watcher type works by installing a signal handler for C<SIGCHLD>,
		389	and since it cannot be shared, nothing else should use SIGCHLD or reap
		390	random child processes (waiting for specific child processes, e.g. inside
		391	C<system>, is just fine).
		392
		393	There is a slight catch to child watchers, however: you usually start them
		394	I<after> the child process was created, and this means the process could
		395	have exited already (and no SIGCHLD will be sent anymore).
		396
		397	Not all event models handle this correctly (neither POE nor IO::Async do,
		398	see their AnyEvent::Impl manpages for details), but even for event models
		399	that I<do> handle this correctly, they usually need to be loaded before
		400	the process exits (i.e. before you fork in the first place). AnyEvent's
		401	pure perl event loop handles all cases correctly regardless of when you
		402	start the watcher.
		403
		404	This means you cannot create a child watcher as the very first
		405	thing in an AnyEvent program, you I<have> to create at least one
		406	watcher before you C<fork> the child (alternatively, you can call
		407	C<AnyEvent::detect>).
		408
		409	Example: fork a process and wait for it
		410
		411	my $done = AnyEvent->condvar;
		412
		413	my $pid = fork or exit 5;
		414
260	my $w = AnyEvent->child (	415	my $w = AnyEvent->child (
261	pid => 1333,	416	pid => $pid,
262	cb => sub {	417	cb => sub {
263	my ($pid, $status) = @_;	418	my ($pid, $status) = @_;
264	warn "pid $pid exited with status $status";	419	warn "pid $pid exited with status $status";
		420	$done->send;
265	},	421	},
266	);	422	);
		423
		424	# do something else, then wait for process exit
		425	$done->recv;
		426
		427	=head2 IDLE WATCHERS
		428
		429	Sometimes there is a need to do something, but it is not so important
		430	to do it instantly, but only when there is nothing better to do. This
		431	"nothing better to do" is usually defined to be "no other events need
		432	attention by the event loop".
		433
		434	Idle watchers ideally get invoked when the event loop has nothing
		435	better to do, just before it would block the process to wait for new
		436	events. Instead of blocking, the idle watcher is invoked.
		437
		438	Most event loops unfortunately do not really support idle watchers (only
		439	EV, Event and Glib do it in a usable fashion) - for the rest, AnyEvent
		440	will simply call the callback "from time to time".
		441
		442	Example: read lines from STDIN, but only process them when the
		443	program is otherwise idle:
		444
		445	my @lines; # read data
		446	my $idle_w;
		447	my $io_w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
		448	push @lines, scalar <STDIN>;
		449
		450	# start an idle watcher, if not already done
		451	$idle_w ||= AnyEvent->idle (cb => sub {
		452	# handle only one line, when there are lines left
		453	if (my $line = shift @lines) {
		454	print "handled when idle: $line";
		455	} else {
		456	# otherwise disable the idle watcher again
		457	undef $idle_w;
		458	}
		459	});
		460	});
267		461
268	=head2 CONDITION VARIABLES	462	=head2 CONDITION VARIABLES
269		463
		464	If you are familiar with some event loops you will know that all of them
		465	require you to run some blocking "loop", "run" or similar function that
		466	will actively watch for new events and call your callbacks.
		467
		468	AnyEvent is slightly different: it expects somebody else to run the event
		469	loop and will only block when necessary (usually when told by the user).
		470
		471	The instrument to do that is called a "condition variable", so called
		472	because they represent a condition that must become true.
		473
		474	Now is probably a good time to look at the examples further below.
		475
270	Condition variables can be created by calling the C<< AnyEvent->condvar >>	476	Condition variables can be created by calling the C<< AnyEvent->condvar
271	method without any arguments.	477	>> method, usually without arguments. The only argument pair allowed is
		478	C<cb>, which specifies a callback to be called when the condition variable
		479	becomes true, with the condition variable as the first argument (but not
		480	the results).
272		481
273	A condition variable waits for a condition - precisely that the C<<	482	After creation, the condition variable is "false" until it becomes "true"
274	->broadcast >> method has been called.	483	by calling the C<send> method (or calling the condition variable as if it
		484	were a callback, read about the caveats in the description for the C<<
		485	->send >> method).
275		486
276	They are very useful to signal that a condition has been fulfilled, for	487	Condition variables are similar to callbacks, except that you can
		488	optionally wait for them. They can also be called merge points - points
		489	in time where multiple outstanding events have been processed. And yet
		490	another way to call them is transactions - each condition variable can be
		491	used to represent a transaction, which finishes at some point and delivers
		492	a result.
		493
		494	Condition variables are very useful to signal that something has finished,
277	example, if you write a module that does asynchronous http requests,	495	for example, if you write a module that does asynchronous http requests,
278	then a condition variable would be the ideal candidate to signal the	496	then a condition variable would be the ideal candidate to signal the
279	availability of results.	497	availability of results. The user can either act when the callback is
		498	called or can synchronously C<< ->recv >> for the results.
280		499
281	You can also use condition variables to block your main program until	500	You can also use them to simulate traditional event loops - for example,
282	an event occurs - for example, you could C<< ->wait >> in your main	501	you can block your main program until an event occurs - for example, you
283	program until the user clicks the Quit button in your app, which would C<<	502	could C<< ->recv >> in your main program until the user clicks the Quit
284	->broadcast >> the "quit" event.	503	button of your app, which would C<< ->send >> the "quit" event.
285		504
286	Note that condition variables recurse into the event loop - if you have	505	Note that condition variables recurse into the event loop - if you have
287	two pirces of code that call C<< ->wait >> in a round-robbin fashion, you	506	two pieces of code that call C<< ->recv >> in a round-robin fashion, you
288	lose. Therefore, condition variables are good to export to your caller, but	507	lose. Therefore, condition variables are good to export to your caller, but
289	you should avoid making a blocking wait yourself, at least in callbacks,	508	you should avoid making a blocking wait yourself, at least in callbacks,
290	as this asks for trouble.	509	as this asks for trouble.
291		510
292	This object has two methods:	511	Condition variables are represented by hash refs in perl, and the keys
		512	used by AnyEvent itself are all named C<_ae_XXX> to make subclassing
		513	easy (it is often useful to build your own transaction class on top of
		514	AnyEvent). To subclass, use C<AnyEvent::CondVar> as base class and call
		515	it's C<new> method in your own C<new> method.
293		516
294	=over 4	517	There are two "sides" to a condition variable - the "producer side" which
		518	eventually calls C<< -> send >>, and the "consumer side", which waits
		519	for the send to occur.
295		520
296	=item $cv->wait	521	Example: wait for a timer.
297
298	Wait (blocking if necessary) until the C<< ->broadcast >> method has been
299	called on c<$cv>, while servicing other watchers normally.
300
301	You can only wait once on a condition - additional calls will return
302	immediately.
303
304	Not all event models support a blocking wait - some die in that case
305	(programs might want to do that to stay interactive), so I<if you are
306	using this from a module, never require a blocking wait>, but let the
307	caller decide whether the call will block or not (for example, by coupling
308	condition variables with some kind of request results and supporting
309	callbacks so the caller knows that getting the result will not block,
310	while still suppporting blocking waits if the caller so desires).
311
312	Another reason I<never> to C<< ->wait >> in a module is that you cannot
313	sensibly have two C<< ->wait >>'s in parallel, as that would require
314	multiple interpreters or coroutines/threads, none of which C<AnyEvent>
315	can supply (the coroutine-aware backends L<AnyEvent::Impl::CoroEV> and
316	L<AnyEvent::Impl::CoroEvent> explicitly support concurrent C<< ->wait >>'s
317	from different coroutines, however).
318
319	=item $cv->broadcast
320
321	Flag the condition as ready - a running C<< ->wait >> and all further
322	calls to C<wait> will (eventually) return after this method has been
323	called. If nobody is waiting the broadcast will be remembered..
324
325	=back
326
327	Example:
328		522
329	# wait till the result is ready	523	# wait till the result is ready
330	my $result_ready = AnyEvent->condvar;	524	my $result_ready = AnyEvent->condvar;
331		525
332	# do something such as adding a timer	526	# do something such as adding a timer
333	# or socket watcher the calls $result_ready->broadcast	527	# or socket watcher the calls $result_ready->send
334	# when the "result" is ready.	528	# when the "result" is ready.
335	# in this case, we simply use a timer:	529	# in this case, we simply use a timer:
336	my $w = AnyEvent->timer (	530	my $w = AnyEvent->timer (
337	after => 1,	531	after => 1,
338	cb => sub { $result_ready->broadcast },	532	cb => sub { $result_ready->send },
339	);	533	);
340		534
341	# this "blocks" (while handling events) till the watcher	535	# this "blocks" (while handling events) till the callback
342	# calls broadcast	536	# calls -<send
343	$result_ready->wait;	537	$result_ready->recv;
		538
		539	Example: wait for a timer, but take advantage of the fact that condition
		540	variables are also callable directly.
		541
		542	my $done = AnyEvent->condvar;
		543	my $delay = AnyEvent->timer (after => 5, cb => $done);
		544	$done->recv;
		545
		546	Example: Imagine an API that returns a condvar and doesn't support
		547	callbacks. This is how you make a synchronous call, for example from
		548	the main program:
		549
		550	use AnyEvent::CouchDB;
		551
		552	...
		553
		554	my @info = $couchdb->info->recv;
		555
		556	And this is how you would just set a callback to be called whenever the
		557	results are available:
		558
		559	$couchdb->info->cb (sub {
		560	my @info = $_[0]->recv;
		561	});
		562
		563	=head3 METHODS FOR PRODUCERS
		564
		565	These methods should only be used by the producing side, i.e. the
		566	code/module that eventually sends the signal. Note that it is also
		567	the producer side which creates the condvar in most cases, but it isn't
		568	uncommon for the consumer to create it as well.
		569
		570	=over 4
		571
		572	=item $cv->send (...)
		573
		574	Flag the condition as ready - a running C<< ->recv >> and all further
		575	calls to C<recv> will (eventually) return after this method has been
		576	called. If nobody is waiting the send will be remembered.
		577
		578	If a callback has been set on the condition variable, it is called
		579	immediately from within send.
		580
		581	Any arguments passed to the C<send> call will be returned by all
		582	future C<< ->recv >> calls.
		583
		584	Condition variables are overloaded so one can call them directly (as if
		585	they were a code reference). Calling them directly is the same as calling
		586	C<send>.
		587
		588	=item $cv->croak ($error)
		589
		590	Similar to send, but causes all call's to C<< ->recv >> to invoke
		591	C<Carp::croak> with the given error message/object/scalar.
		592
		593	This can be used to signal any errors to the condition variable
		594	user/consumer. Doing it this way instead of calling C<croak> directly
		595	delays the error detetcion, but has the overwhelmign advantage that it
		596	diagnoses the error at the place where the result is expected, and not
		597	deep in some event clalback without connection to the actual code causing
		598	the problem.
		599
		600	=item $cv->begin ([group callback])
		601
		602	=item $cv->end
		603
		604	These two methods can be used to combine many transactions/events into
		605	one. For example, a function that pings many hosts in parallel might want
		606	to use a condition variable for the whole process.
		607
		608	Every call to C<< ->begin >> will increment a counter, and every call to
		609	C<< ->end >> will decrement it. If the counter reaches C<0> in C<< ->end
		610	>>, the (last) callback passed to C<begin> will be executed. That callback
		611	is I<supposed> to call C<< ->send >>, but that is not required. If no
		612	callback was set, C<send> will be called without any arguments.
		613
		614	You can think of C<< $cv->send >> giving you an OR condition (one call
		615	sends), while C<< $cv->begin >> and C<< $cv->end >> giving you an AND
		616	condition (all C<begin> calls must be C<end>'ed before the condvar sends).
		617
		618	Let's start with a simple example: you have two I/O watchers (for example,
		619	STDOUT and STDERR for a program), and you want to wait for both streams to
		620	close before activating a condvar:
		621
		622	my $cv = AnyEvent->condvar;
		623
		624	$cv->begin; # first watcher
		625	my $w1 = AnyEvent->io (fh => $fh1, cb => sub {
		626	defined sysread $fh1, my $buf, 4096
		627	or $cv->end;
		628	});
		629
		630	$cv->begin; # second watcher
		631	my $w2 = AnyEvent->io (fh => $fh2, cb => sub {
		632	defined sysread $fh2, my $buf, 4096
		633	or $cv->end;
		634	});
		635
		636	$cv->recv;
		637
		638	This works because for every event source (EOF on file handle), there is
		639	one call to C<begin>, so the condvar waits for all calls to C<end> before
		640	sending.
		641
		642	The ping example mentioned above is slightly more complicated, as the
		643	there are results to be passwd back, and the number of tasks that are
		644	begung can potentially be zero:
		645
		646	my $cv = AnyEvent->condvar;
		647
		648	my %result;
		649	$cv->begin (sub { $cv->send (\%result) });
		650
		651	for my $host (@list_of_hosts) {
		652	$cv->begin;
		653	ping_host_then_call_callback $host, sub {
		654	$result{$host} = ...;
		655	$cv->end;
		656	};
		657	}
		658
		659	$cv->end;
		660
		661	This code fragment supposedly pings a number of hosts and calls
		662	C<send> after results for all then have have been gathered - in any
		663	order. To achieve this, the code issues a call to C<begin> when it starts
		664	each ping request and calls C<end> when it has received some result for
		665	it. Since C<begin> and C<end> only maintain a counter, the order in which
		666	results arrive is not relevant.
		667
		668	There is an additional bracketing call to C<begin> and C<end> outside the
		669	loop, which serves two important purposes: first, it sets the callback
		670	to be called once the counter reaches C<0>, and second, it ensures that
		671	C<send> is called even when C<no> hosts are being pinged (the loop
		672	doesn't execute once).
		673
		674	This is the general pattern when you "fan out" into multiple (but
		675	potentially none) subrequests: use an outer C<begin>/C<end> pair to set
		676	the callback and ensure C<end> is called at least once, and then, for each
		677	subrequest you start, call C<begin> and for each subrequest you finish,
		678	call C<end>.
		679
		680	=back
		681
		682	=head3 METHODS FOR CONSUMERS
		683
		684	These methods should only be used by the consuming side, i.e. the
		685	code awaits the condition.
		686
		687	=over 4
		688
		689	=item $cv->recv
		690
		691	Wait (blocking if necessary) until the C<< ->send >> or C<< ->croak
		692	>> methods have been called on c<$cv>, while servicing other watchers
		693	normally.
		694
		695	You can only wait once on a condition - additional calls are valid but
		696	will return immediately.
		697
		698	If an error condition has been set by calling C<< ->croak >>, then this
		699	function will call C<croak>.
		700
		701	In list context, all parameters passed to C<send> will be returned,
		702	in scalar context only the first one will be returned.
		703
		704	Note that doing a blocking wait in a callback is not supported by any
		705	event loop, that is, recursive invocation of a blocking C<< ->recv
		706	>> is not allowed, and the C<recv> call will C<croak> if such a
		707	condition is detected. This condition can be slightly loosened by using
		708	L<Coro::AnyEvent>, which allows you to do a blocking C<< ->recv >> from
		709	any thread that doesn't run the event loop itself.
		710
		711	Not all event models support a blocking wait - some die in that case
		712	(programs might want to do that to stay interactive), so I<if you are
		713	using this from a module, never require a blocking wait>. Instead, let the
		714	caller decide whether the call will block or not (for example, by coupling
		715	condition variables with some kind of request results and supporting
		716	callbacks so the caller knows that getting the result will not block,
		717	while still supporting blocking waits if the caller so desires).
		718
		719	You can ensure that C<< -recv >> never blocks by setting a callback and
		720	only calling C<< ->recv >> from within that callback (or at a later
		721	time). This will work even when the event loop does not support blocking
		722	waits otherwise.
		723
		724	=item $bool = $cv->ready
		725
		726	Returns true when the condition is "true", i.e. whether C<send> or
		727	C<croak> have been called.
		728
		729	=item $cb = $cv->cb ($cb->($cv))
		730
		731	This is a mutator function that returns the callback set and optionally
		732	replaces it before doing so.
		733
		734	The callback will be called when the condition becomes "true", i.e. when
		735	C<send> or C<croak> are called, with the only argument being the condition
		736	variable itself. Calling C<recv> inside the callback or at any later time
		737	is guaranteed not to block.
		738
		739	=back
		740
		741	=head1 SUPPORTED EVENT LOOPS/BACKENDS
		742
		743	The available backend classes are (every class has its own manpage):
		744
		745	=over 4
		746
		747	=item Backends that are autoprobed when no other event loop can be found.
		748
		749	EV is the preferred backend when no other event loop seems to be in
		750	use. If EV is not installed, then AnyEvent will try Event, and, failing
		751	that, will fall back to its own pure-perl implementation, which is
		752	available everywhere as it comes with AnyEvent itself.
		753
		754	AnyEvent::Impl::EV based on EV (interface to libev, best choice).
		755	AnyEvent::Impl::Event based on Event, very stable, few glitches.
		756	AnyEvent::Impl::Perl pure-perl implementation, fast and portable.
		757
		758	=item Backends that are transparently being picked up when they are used.
		759
		760	These will be used when they are currently loaded when the first watcher
		761	is created, in which case it is assumed that the application is using
		762	them. This means that AnyEvent will automatically pick the right backend
		763	when the main program loads an event module before anything starts to
		764	create watchers. Nothing special needs to be done by the main program.
		765
		766	AnyEvent::Impl::Glib based on Glib, slow but very stable.
		767	AnyEvent::Impl::Tk based on Tk, very broken.
		768	AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
		769	AnyEvent::Impl::POE based on POE, very slow, some limitations.
		770
		771	=item Backends with special needs.
		772
		773	Qt requires the Qt::Application to be instantiated first, but will
		774	otherwise be picked up automatically. As long as the main program
		775	instantiates the application before any AnyEvent watchers are created,
		776	everything should just work.
		777
		778	AnyEvent::Impl::Qt based on Qt.
		779
		780	Support for IO::Async can only be partial, as it is too broken and
		781	architecturally limited to even support the AnyEvent API. It also
		782	is the only event loop that needs the loop to be set explicitly, so
		783	it can only be used by a main program knowing about AnyEvent. See
		784	L<AnyEvent::Impl::Async> for the gory details.
		785
		786	AnyEvent::Impl::IOAsync based on IO::Async, cannot be autoprobed.
		787
		788	=item Event loops that are indirectly supported via other backends.
		789
		790	Some event loops can be supported via other modules:
		791
		792	There is no direct support for WxWidgets (L<Wx>) or L<Prima>.
		793
		794	B<WxWidgets> has no support for watching file handles. However, you can
		795	use WxWidgets through the POE adaptor, as POE has a Wx backend that simply
		796	polls 20 times per second, which was considered to be too horrible to even
		797	consider for AnyEvent.
		798
		799	B<Prima> is not supported as nobody seems to be using it, but it has a POE
		800	backend, so it can be supported through POE.
		801
		802	AnyEvent knows about both L<Prima> and L<Wx>, however, and will try to
		803	load L<POE> when detecting them, in the hope that POE will pick them up,
		804	in which case everything will be automatic.
		805
		806	=back
344		807
345	=head1 GLOBAL VARIABLES AND FUNCTIONS	808	=head1 GLOBAL VARIABLES AND FUNCTIONS
346		809
		810	These are not normally required to use AnyEvent, but can be useful to
		811	write AnyEvent extension modules.
		812
347	=over 4	813	=over 4
348		814
349	=item $AnyEvent::MODEL	815	=item $AnyEvent::MODEL
350		816
351	Contains C<undef> until the first watcher is being created. Then it	817	Contains C<undef> until the first watcher is being created, before the
		818	backend has been autodetected.
		819
352	contains the event model that is being used, which is the name of the	820	Afterwards it contains the event model that is being used, which is the
353	Perl class implementing the model. This class is usually one of the	821	name of the Perl class implementing the model. This class is usually one
354	C<AnyEvent::Impl:xxx> modules, but can be any other class in the case	822	of the C<AnyEvent::Impl:xxx> modules, but can be any other class in the
355	AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode>).	823	case AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode> it
356		824	will be C<urxvt::anyevent>).
357	The known classes so far are:
358
359	AnyEvent::Impl::CoroEV based on Coro::EV, best choice.
360	AnyEvent::Impl::CoroEvent based on Coro::Event, second best choice.
361	AnyEvent::Impl::EV based on EV (an interface to libev, best choice).
362	AnyEvent::Impl::Event based on Event, second best choice.
363	AnyEvent::Impl::Glib based on Glib, third-best choice.
364	AnyEvent::Impl::Tk based on Tk, very bad choice.
365	AnyEvent::Impl::Perl pure-perl implementation, inefficient but portable.
366	AnyEvent::Impl::Qt based on Qt, cannot be autoprobed (see its docs).
367	AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
368	AnyEvent::Impl::POE based on POE, not generic enough for full support.
369
370	There is no support for WxWidgets, as WxWidgets has no support for
371	watching file handles. However, you can use WxWidgets through the
372	POE Adaptor, as POE has a Wx backend that simply polls 20 times per
373	second, which was considered to be too horrible to even consider for
374	AnyEvent. Likewise, other POE backends can be used by AnyEvent by using
375	it's adaptor.
376
377	AnyEvent knows about L<Prima> and L<Wx> and will try to use L<POE> when
378	autodetecting them.
379		825
380	=item AnyEvent::detect	826	=item AnyEvent::detect
381		827
382	Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model	828	Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model
383	if necessary. You should only call this function right before you would	829	if necessary. You should only call this function right before you would
384	have created an AnyEvent watcher anyway, that is, as late as possible at	830	have created an AnyEvent watcher anyway, that is, as late as possible at
385	runtime.	831	runtime, and not e.g. while initialising of your module.
		832
		833	If you need to do some initialisation before AnyEvent watchers are
		834	created, use C<post_detect>.
		835
		836	=item $guard = AnyEvent::post_detect { BLOCK }
		837
		838	Arranges for the code block to be executed as soon as the event model is
		839	autodetected (or immediately if this has already happened).
		840
		841	The block will be executed I<after> the actual backend has been detected
		842	(C<$AnyEvent::MODEL> is set), but I<before> any watchers have been
		843	created, so it is possible to e.g. patch C<@AnyEvent::ISA> or do
		844	other initialisations - see the sources of L<AnyEvent::Strict> or
		845	L<AnyEvent::AIO> to see how this is used.
		846
		847	The most common usage is to create some global watchers, without forcing
		848	event module detection too early, for example, L<AnyEvent::AIO> creates
		849	and installs the global L<IO::AIO> watcher in a C<post_detect> block to
		850	avoid autodetecting the event module at load time.
		851
		852	If called in scalar or list context, then it creates and returns an object
		853	that automatically removes the callback again when it is destroyed. See
		854	L<Coro::BDB> for a case where this is useful.
		855
		856	=item @AnyEvent::post_detect
		857
		858	If there are any code references in this array (you can C<push> to it
		859	before or after loading AnyEvent), then they will called directly after
		860	the event loop has been chosen.
		861
		862	You should check C<$AnyEvent::MODEL> before adding to this array, though:
		863	if it is defined then the event loop has already been detected, and the
		864	array will be ignored.
		865
		866	Best use C<AnyEvent::post_detect { BLOCK }> when your application allows
		867	it,as it takes care of these details.
		868
		869	This variable is mainly useful for modules that can do something useful
		870	when AnyEvent is used and thus want to know when it is initialised, but do
		871	not need to even load it by default. This array provides the means to hook
		872	into AnyEvent passively, without loading it.
386		873
387	=back	874	=back
388		875
389	=head1 WHAT TO DO IN A MODULE	876	=head1 WHAT TO DO IN A MODULE
390		877
…		…
394	Be careful when you create watchers in the module body - AnyEvent will	881	Be careful when you create watchers in the module body - AnyEvent will
395	decide which event module to use as soon as the first method is called, so	882	decide which event module to use as soon as the first method is called, so
396	by calling AnyEvent in your module body you force the user of your module	883	by calling AnyEvent in your module body you force the user of your module
397	to load the event module first.	884	to load the event module first.
398		885
399	Never call C<< ->wait >> on a condition variable unless you I<know> that	886	Never call C<< ->recv >> on a condition variable unless you I<know> that
400	the C<< ->broadcast >> method has been called on it already. This is	887	the C<< ->send >> method has been called on it already. This is
401	because it will stall the whole program, and the whole point of using	888	because it will stall the whole program, and the whole point of using
402	events is to stay interactive.	889	events is to stay interactive.
403		890
404	It is fine, however, to call C<< ->wait >> when the user of your module	891	It is fine, however, to call C<< ->recv >> when the user of your module
405	requests it (i.e. if you create a http request object ad have a method	892	requests it (i.e. if you create a http request object ad have a method
406	called C<results> that returns the results, it should call C<< ->wait >>	893	called C<results> that returns the results, it should call C<< ->recv >>
407	freely, as the user of your module knows what she is doing. always).	894	freely, as the user of your module knows what she is doing. always).
408		895
409	=head1 WHAT TO DO IN THE MAIN PROGRAM	896	=head1 WHAT TO DO IN THE MAIN PROGRAM
410		897
411	There will always be a single main program - the only place that should	898	There will always be a single main program - the only place that should
…		…
413		900
414	If it doesn't care, it can just "use AnyEvent" and use it itself, or not	901	If it doesn't care, it can just "use AnyEvent" and use it itself, or not
415	do anything special (it does not need to be event-based) and let AnyEvent	902	do anything special (it does not need to be event-based) and let AnyEvent
416	decide which implementation to chose if some module relies on it.	903	decide which implementation to chose if some module relies on it.
417		904
418	If the main program relies on a specific event model. For example, in	905	If the main program relies on a specific event model - for example, in
419	Gtk2 programs you have to rely on the Glib module. You should load the	906	Gtk2 programs you have to rely on the Glib module - you should load the
420	event module before loading AnyEvent or any module that uses it: generally	907	event module before loading AnyEvent or any module that uses it: generally
421	speaking, you should load it as early as possible. The reason is that	908	speaking, you should load it as early as possible. The reason is that
422	modules might create watchers when they are loaded, and AnyEvent will	909	modules might create watchers when they are loaded, and AnyEvent will
423	decide on the event model to use as soon as it creates watchers, and it	910	decide on the event model to use as soon as it creates watchers, and it
424	might chose the wrong one unless you load the correct one yourself.	911	might chose the wrong one unless you load the correct one yourself.
425		912
426	You can chose to use a rather inefficient pure-perl implementation by	913	You can chose to use a pure-perl implementation by loading the
427	loading the C<AnyEvent::Impl::Perl> module, which gives you similar	914	C<AnyEvent::Impl::Perl> module, which gives you similar behaviour
428	behaviour everywhere, but letting AnyEvent chose is generally better.	915	everywhere, but letting AnyEvent chose the model is generally better.
		916
		917	=head2 MAINLOOP EMULATION
		918
		919	Sometimes (often for short test scripts, or even standalone programs who
		920	only want to use AnyEvent), you do not want to run a specific event loop.
		921
		922	In that case, you can use a condition variable like this:
		923
		924	AnyEvent->condvar->recv;
		925
		926	This has the effect of entering the event loop and looping forever.
		927
		928	Note that usually your program has some exit condition, in which case
		929	it is better to use the "traditional" approach of storing a condition
		930	variable somewhere, waiting for it, and sending it when the program should
		931	exit cleanly.
		932
		933
		934	=head1 OTHER MODULES
		935
		936	The following is a non-exhaustive list of additional modules that use
		937	AnyEvent as a client and can therefore be mixed easily with other AnyEvent
		938	modules and other event loops in the same program. Some of the modules
		939	come with AnyEvent, most are available via CPAN.
		940
		941	=over 4
		942
		943	=item L<AnyEvent::Util>
		944
		945	Contains various utility functions that replace often-used but blocking
		946	functions such as C<inet_aton> by event-/callback-based versions.
		947
		948	=item L<AnyEvent::Socket>
		949
		950	Provides various utility functions for (internet protocol) sockets,
		951	addresses and name resolution. Also functions to create non-blocking tcp
		952	connections or tcp servers, with IPv6 and SRV record support and more.
		953
		954	=item L<AnyEvent::Handle>
		955
		956	Provide read and write buffers, manages watchers for reads and writes,
		957	supports raw and formatted I/O, I/O queued and fully transparent and
		958	non-blocking SSL/TLS (via L<AnyEvent::TLS>.
		959
		960	=item L<AnyEvent::DNS>
		961
		962	Provides rich asynchronous DNS resolver capabilities.
		963
		964	=item L<AnyEvent::HTTP>
		965
		966	A simple-to-use HTTP library that is capable of making a lot of concurrent
		967	HTTP requests.
		968
		969	=item L<AnyEvent::HTTPD>
		970
		971	Provides a simple web application server framework.
		972
		973	=item L<AnyEvent::FastPing>
		974
		975	The fastest ping in the west.
		976
		977	=item L<AnyEvent::DBI>
		978
		979	Executes L<DBI> requests asynchronously in a proxy process.
		980
		981	=item L<AnyEvent::AIO>
		982
		983	Truly asynchronous I/O, should be in the toolbox of every event
		984	programmer. AnyEvent::AIO transparently fuses L<IO::AIO> and AnyEvent
		985	together.
		986
		987	=item L<AnyEvent::BDB>
		988
		989	Truly asynchronous Berkeley DB access. AnyEvent::BDB transparently fuses
		990	L<BDB> and AnyEvent together.
		991
		992	=item L<AnyEvent::GPSD>
		993
		994	A non-blocking interface to gpsd, a daemon delivering GPS information.
		995
		996	=item L<AnyEvent::IRC>
		997
		998	AnyEvent based IRC client module family (replacing the older Net::IRC3).
		999
		1000	=item L<AnyEvent::XMPP>
		1001
		1002	AnyEvent based XMPP (Jabber protocol) module family (replacing the older
		1003	Net::XMPP2>.
		1004
		1005	=item L<AnyEvent::IGS>
		1006
		1007	A non-blocking interface to the Internet Go Server protocol (used by
		1008	L<App::IGS>).
		1009
		1010	=item L<Net::FCP>
		1011
		1012	AnyEvent-based implementation of the Freenet Client Protocol, birthplace
		1013	of AnyEvent.
		1014
		1015	=item L<Event::ExecFlow>
		1016
		1017	High level API for event-based execution flow control.
		1018
		1019	=item L<Coro>
		1020
		1021	Has special support for AnyEvent via L<Coro::AnyEvent>.
		1022
		1023	=back
429		1024
430	=cut	1025	=cut
431		1026
432	package AnyEvent;	1027	package AnyEvent;
433		1028
434	no warnings;	1029	no warnings;
435	use strict;	1030	use strict qw(vars subs);
436		1031
437	use Carp;	1032	use Carp ();
438		1033
439	our $VERSION = '3.3';	1034	our $VERSION = 4.83;
440	our $MODEL;	1035	our $MODEL;
441		1036
442	our $AUTOLOAD;	1037	our $AUTOLOAD;
443	our @ISA;	1038	our @ISA;
444		1039
		1040	our @REGISTRY;
		1041
		1042	our $WIN32;
		1043
		1044	BEGIN {
		1045	eval "sub WIN32(){ " . (($^O =~ /mswin32/i)*1) ." }";
		1046	eval "sub TAINT(){ " . (${^TAINT}*1) . " }";
		1047
		1048	delete @ENV{grep /^PERL_ANYEVENT_/, keys %ENV}
		1049	if ${^TAINT};
		1050	}
		1051
445	our $verbose = $ENV{PERL_ANYEVENT_VERBOSE}*1;	1052	our $verbose = $ENV{PERL_ANYEVENT_VERBOSE}*1;
446		1053
447	our @REGISTRY;	1054	our %PROTOCOL; # (ipv4|ipv6) => (1|2), higher numbers are preferred
		1055
		1056	{
		1057	my $idx;
		1058	$PROTOCOL{$_} = ++$idx
		1059	for reverse split /\s*,\s*/,
		1060	$ENV{PERL_ANYEVENT_PROTOCOLS} || "ipv4,ipv6";
		1061	}
448		1062
449	my @models = (	1063	my @models = (
450	[Coro::EV:: => AnyEvent::Impl::CoroEV::],
451	[Coro::Event:: => AnyEvent::Impl::CoroEvent::],
452	[EV:: => AnyEvent::Impl::EV::],	1064	[EV:: => AnyEvent::Impl::EV::],
453	[Event:: => AnyEvent::Impl::Event::],	1065	[Event:: => AnyEvent::Impl::Event::],
454	[Glib:: => AnyEvent::Impl::Glib::],	1066	[AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl::],
		1067	# everything below here will not be autoprobed
		1068	# as the pureperl backend should work everywhere
		1069	# and is usually faster
		1070	[Glib:: => AnyEvent::Impl::Glib::], # becomes extremely slow with many watchers
		1071	[Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy
455	[Tk:: => AnyEvent::Impl::Tk::],	1072	[Tk:: => AnyEvent::Impl::Tk::], # crashes with many handles
		1073	[Qt:: => AnyEvent::Impl::Qt::], # requires special main program
		1074	[POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza
456	[Wx:: => AnyEvent::Impl::POE::],	1075	[Wx:: => AnyEvent::Impl::POE::],
457	[Prima:: => AnyEvent::Impl::POE::],	1076	[Prima:: => AnyEvent::Impl::POE::],
458	[AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl::],	1077	# IO::Async is just too broken - we would need workarounds for its
459	# everything below here will not be autoprobed as the pureperl backend should work everywhere	1078	# byzantine signal and broken child handling, among others.
460	[Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy	1079	# IO::Async is rather hard to detect, as it doesn't have any
		1080	# obvious default class.
461	[Qt:: => AnyEvent::Impl::Qt::], # requires special main program	1081	# [IO::Async:: => AnyEvent::Impl::IOAsync::], # requires special main program
462	[POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza	1082	# [IO::Async::Loop:: => AnyEvent::Impl::IOAsync::], # requires special main program
		1083	# [IO::Async::Notifier:: => AnyEvent::Impl::IOAsync::], # requires special main program
463	);	1084	);
464		1085
465	our %method = map +($_ => 1), qw(io timer signal child condvar broadcast wait one_event DESTROY);	1086	our %method = map +($_ => 1),
		1087	qw(io timer time now now_update signal child idle condvar one_event DESTROY);
		1088
		1089	our @post_detect;
		1090
		1091	sub post_detect(&) {
		1092	my ($cb) = @_;
		1093
		1094	if ($MODEL) {
		1095	$cb->();
		1096
		1097	1
		1098	} else {
		1099	push @post_detect, $cb;
		1100
		1101	defined wantarray
		1102	? bless \$cb, "AnyEvent::Util::postdetect"
		1103	: ()
		1104	}
		1105	}
		1106
		1107	sub AnyEvent::Util::postdetect::DESTROY {
		1108	@post_detect = grep $_ != ${$_[0]}, @post_detect;
		1109	}
466		1110
467	sub detect() {	1111	sub detect() {
468	unless ($MODEL) {	1112	unless ($MODEL) {
469	no strict 'refs';	1113	no strict 'refs';
		1114	local $SIG{__DIE__};
470		1115
471	if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {	1116	if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {
472	my $model = "AnyEvent::Impl::$1";	1117	my $model = "AnyEvent::Impl::$1";
473	if (eval "require $model") {	1118	if (eval "require $model") {
474	$MODEL = $model;	1119	$MODEL = $model;
475	warn "AnyEvent: loaded model '$model' (forced by \$PERL_ANYEVENT_MODEL), using it.\n" if $verbose > 1;	1120	warn "AnyEvent: loaded model '$model' (forced by \$ENV{PERL_ANYEVENT_MODEL}), using it.\n" if $verbose > 1;
476	} else {	1121	} else {
477	warn "AnyEvent: unable to load model '$model' (from \$PERL_ANYEVENT_MODEL):\n$@" if $verbose;	1122	warn "AnyEvent: unable to load model '$model' (from \$ENV{PERL_ANYEVENT_MODEL}):\n$@" if $verbose;
478	}	1123	}
479	}	1124	}
480		1125
481	# check for already loaded models	1126	# check for already loaded models
482	unless ($MODEL) {	1127	unless ($MODEL) {
…		…
504	last;	1149	last;
505	}	1150	}
506	}	1151	}
507		1152
508	$MODEL	1153	$MODEL
509	or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV (or Coro+EV), Event (or Coro+Event) or Glib.";	1154	or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV, Event or Glib.\n";
510	}	1155	}
511	}	1156	}
512		1157
		1158	push @{"$MODEL\::ISA"}, "AnyEvent::Base";
		1159
513	unshift @ISA, $MODEL;	1160	unshift @ISA, $MODEL;
514	push @{"$MODEL\::ISA"}, "AnyEvent::Base";	1161
		1162	require AnyEvent::Strict if $ENV{PERL_ANYEVENT_STRICT};
		1163
		1164	(shift @post_detect)->() while @post_detect;
515	}	1165	}
516		1166
517	$MODEL	1167	$MODEL
518	}	1168	}
519		1169
520	sub AUTOLOAD {	1170	sub AUTOLOAD {
521	(my $func = $AUTOLOAD) =~ s/.*://;	1171	(my $func = $AUTOLOAD) =~ s/.*://;
522		1172
523	$method{$func}	1173	$method{$func}
524	or croak "$func: not a valid method for AnyEvent objects";	1174	or Carp::croak "$func: not a valid method for AnyEvent objects";
525		1175
526	detect unless $MODEL;	1176	detect unless $MODEL;
527		1177
528	my $class = shift;	1178	my $class = shift;
529	$class->$func (@_);	1179	$class->$func (@_);
530	}	1180	}
531		1181
		1182	# utility function to dup a filehandle. this is used by many backends
		1183	# to support binding more than one watcher per filehandle (they usually
		1184	# allow only one watcher per fd, so we dup it to get a different one).
		1185	sub _dupfh($$;$$) {
		1186	my ($poll, $fh, $r, $w) = @_;
		1187
		1188	# cygwin requires the fh mode to be matching, unix doesn't
		1189	my ($rw, $mode) = $poll eq "r" ? ($r, "<") : ($w, ">");
		1190
		1191	open my $fh2, "$mode&", $fh
		1192	or die "AnyEvent->io: cannot dup() filehandle in mode '$poll': $!,";
		1193
		1194	# we assume CLOEXEC is already set by perl in all important cases
		1195
		1196	($fh2, $rw)
		1197	}
		1198
532	package AnyEvent::Base;	1199	package AnyEvent::Base;
533		1200
		1201	# default implementations for many methods
		1202
		1203	BEGIN {
		1204	if (eval "use Time::HiRes (); Time::HiRes::time (); 1") {
		1205	*_time = \&Time::HiRes::time;
		1206	# if (eval "use POSIX (); (POSIX::times())...
		1207	} else {
		1208	*_time = sub { time }; # epic fail
		1209	}
		1210	}
		1211
		1212	sub time { _time }
		1213	sub now { _time }
		1214	sub now_update { }
		1215
534	# default implementation for ->condvar, ->wait, ->broadcast	1216	# default implementation for ->condvar
535		1217
536	sub condvar {	1218	sub condvar {
537	bless \my $flag, "AnyEvent::Base::CondVar"	1219	bless { @_ == 3 ? (_ae_cb => $_[2]) : () }, "AnyEvent::CondVar"
538	}
539
540	sub AnyEvent::Base::CondVar::broadcast {
541	${$_[0]}++;
542	}
543
544	sub AnyEvent::Base::CondVar::wait {
545	AnyEvent->one_event while !${$_[0]};
546	}	1220	}
547		1221
548	# default implementation for ->signal	1222	# default implementation for ->signal
549		1223
550	our %SIG_CB;	1224	our ($SIGPIPE_R, $SIGPIPE_W, %SIG_CB, %SIG_EV, $SIG_IO);
		1225
		1226	sub _signal_exec {
		1227	sysread $SIGPIPE_R, my $dummy, 4;
		1228
		1229	while (%SIG_EV) {
		1230	for (keys %SIG_EV) {
		1231	delete $SIG_EV{$_};
		1232	$_->() for values %{ $SIG_CB{$_} || {} };
		1233	}
		1234	}
		1235	}
551		1236
552	sub signal {	1237	sub signal {
553	my (undef, %arg) = @_;	1238	my (undef, %arg) = @_;
554		1239
		1240	unless ($SIGPIPE_R) {
		1241	require Fcntl;
		1242
		1243	if (AnyEvent::WIN32) {
		1244	require AnyEvent::Util;
		1245
		1246	($SIGPIPE_R, $SIGPIPE_W) = AnyEvent::Util::portable_pipe ();
		1247	AnyEvent::Util::fh_nonblocking ($SIGPIPE_R) if $SIGPIPE_R;
		1248	AnyEvent::Util::fh_nonblocking ($SIGPIPE_W) if $SIGPIPE_W; # just in case
		1249	} else {
		1250	pipe $SIGPIPE_R, $SIGPIPE_W;
		1251	fcntl $SIGPIPE_R, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_R;
		1252	fcntl $SIGPIPE_W, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_W; # just in case
		1253
		1254	# not strictly required, as $^F is normally 2, but let's make sure...
		1255	fcntl $SIGPIPE_R, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC;
		1256	fcntl $SIGPIPE_W, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC;
		1257	}
		1258
		1259	$SIGPIPE_R
		1260	or Carp::croak "AnyEvent: unable to create a signal reporting pipe: $!\n";
		1261
		1262	$SIG_IO = AnyEvent->io (fh => $SIGPIPE_R, poll => "r", cb => \&_signal_exec);
		1263	}
		1264
555	my $signal = uc $arg{signal}	1265	my $signal = uc $arg{signal}
556	or Carp::croak "required option 'signal' is missing";	1266	or Carp::croak "required option 'signal' is missing";
557		1267
558	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};	1268	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};
559	$SIG{$signal} ||= sub {	1269	$SIG{$signal} ||= sub {
560	$_->() for values %{ $SIG_CB{$signal} || {} };	1270	local $!;
		1271	syswrite $SIGPIPE_W, "\x00", 1 unless %SIG_EV;
		1272	undef $SIG_EV{$signal};
561	};	1273	};
562		1274
563	bless [$signal, $arg{cb}], "AnyEvent::Base::Signal"	1275	bless [$signal, $arg{cb}], "AnyEvent::Base::signal"
564	}	1276	}
565		1277
566	sub AnyEvent::Base::Signal::DESTROY {	1278	sub AnyEvent::Base::signal::DESTROY {
567	my ($signal, $cb) = @{$_[0]};	1279	my ($signal, $cb) = @{$_[0]};
568		1280
569	delete $SIG_CB{$signal}{$cb};	1281	delete $SIG_CB{$signal}{$cb};
570		1282
		1283	# delete doesn't work with older perls - they then
		1284	# print weird messages, or just unconditionally exit
		1285	# instead of getting the default action.
571	$SIG{$signal} = 'DEFAULT' unless keys %{ $SIG_CB{$signal} };	1286	undef $SIG{$signal} unless keys %{ $SIG_CB{$signal} };
572	}	1287	}
573		1288
574	# default implementation for ->child	1289	# default implementation for ->child
575		1290
576	our %PID_CB;	1291	our %PID_CB;
577	our $CHLD_W;	1292	our $CHLD_W;
578	our $CHLD_DELAY_W;	1293	our $CHLD_DELAY_W;
579	our $PID_IDLE;
580	our $WNOHANG;	1294	our $WNOHANG;
581		1295
582	sub _child_wait {	1296	sub _sigchld {
583	while (0 < (my $pid = waitpid -1, $WNOHANG)) {	1297	while (0 < (my $pid = waitpid -1, $WNOHANG)) {
584	$_->($pid, $?) for (values %{ $PID_CB{$pid} || {} }),	1298	$_->($pid, $?) for (values %{ $PID_CB{$pid} || {} }),
585	(values %{ $PID_CB{0} || {} });	1299	(values %{ $PID_CB{0} || {} });
586	}	1300	}
587
588	undef $PID_IDLE;
589	}
590
591	sub _sigchld {
592	# make sure we deliver these changes "synchronous" with the event loop.
593	$CHLD_DELAY_W ||= AnyEvent->timer (after => 0, cb => sub {
594	undef $CHLD_DELAY_W;
595	&_child_wait;
596	});
597	}	1301	}
598		1302
599	sub child {	1303	sub child {
600	my (undef, %arg) = @_;	1304	my (undef, %arg) = @_;
601		1305
602	defined (my $pid = $arg{pid} + 0)	1306	defined (my $pid = $arg{pid} + 0)
603	or Carp::croak "required option 'pid' is missing";	1307	or Carp::croak "required option 'pid' is missing";
604		1308
605	$PID_CB{$pid}{$arg{cb}} = $arg{cb};	1309	$PID_CB{$pid}{$arg{cb}} = $arg{cb};
606		1310
607	unless ($WNOHANG) {
608	$WNOHANG = eval { require POSIX; &POSIX::WNOHANG } || 1;	1311	$WNOHANG ||= eval { local $SIG{__DIE__}; require POSIX; &POSIX::WNOHANG } || 1;
609	}
610		1312
611	unless ($CHLD_W) {	1313	unless ($CHLD_W) {
612	$CHLD_W = AnyEvent->signal (signal => 'CHLD', cb => \&_sigchld);	1314	$CHLD_W = AnyEvent->signal (signal => 'CHLD', cb => \&_sigchld);
613	# child could be a zombie already, so make at least one round	1315	# child could be a zombie already, so make at least one round
614	&_sigchld;	1316	&_sigchld;
615	}	1317	}
616		1318
617	bless [$pid, $arg{cb}], "AnyEvent::Base::Child"	1319	bless [$pid, $arg{cb}], "AnyEvent::Base::child"
618	}	1320	}
619		1321
620	sub AnyEvent::Base::Child::DESTROY {	1322	sub AnyEvent::Base::child::DESTROY {
621	my ($pid, $cb) = @{$_[0]};	1323	my ($pid, $cb) = @{$_[0]};
622		1324
623	delete $PID_CB{$pid}{$cb};	1325	delete $PID_CB{$pid}{$cb};
624	delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };	1326	delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };
625		1327
626	undef $CHLD_W unless keys %PID_CB;	1328	undef $CHLD_W unless keys %PID_CB;
627	}	1329	}
		1330
		1331	# idle emulation is done by simply using a timer, regardless
		1332	# of whether the process is idle or not, and not letting
		1333	# the callback use more than 50% of the time.
		1334	sub idle {
		1335	my (undef, %arg) = @_;
		1336
		1337	my ($cb, $w, $rcb) = $arg{cb};
		1338
		1339	$rcb = sub {
		1340	if ($cb) {
		1341	$w = _time;
		1342	&$cb;
		1343	$w = _time - $w;
		1344
		1345	# never use more then 50% of the time for the idle watcher,
		1346	# within some limits
		1347	$w = 0.0001 if $w < 0.0001;
		1348	$w = 5 if $w > 5;
		1349
		1350	$w = AnyEvent->timer (after => $w, cb => $rcb);
		1351	} else {
		1352	# clean up...
		1353	undef $w;
		1354	undef $rcb;
		1355	}
		1356	};
		1357
		1358	$w = AnyEvent->timer (after => 0.05, cb => $rcb);
		1359
		1360	bless \\$cb, "AnyEvent::Base::idle"
		1361	}
		1362
		1363	sub AnyEvent::Base::idle::DESTROY {
		1364	undef $${$_[0]};
		1365	}
		1366
		1367	package AnyEvent::CondVar;
		1368
		1369	our @ISA = AnyEvent::CondVar::Base::;
		1370
		1371	package AnyEvent::CondVar::Base;
		1372
		1373	use overload
		1374	'&{}' => sub { my $self = shift; sub { $self->send (@_) } },
		1375	fallback => 1;
		1376
		1377	our $WAITING;
		1378
		1379	sub _send {
		1380	# nop
		1381	}
		1382
		1383	sub send {
		1384	my $cv = shift;
		1385	$cv->{_ae_sent} = [@_];
		1386	(delete $cv->{_ae_cb})->($cv) if $cv->{_ae_cb};
		1387	$cv->_send;
		1388	}
		1389
		1390	sub croak {
		1391	$_[0]{_ae_croak} = $_[1];
		1392	$_[0]->send;
		1393	}
		1394
		1395	sub ready {
		1396	$_[0]{_ae_sent}
		1397	}
		1398
		1399	sub _wait {
		1400	$WAITING
		1401	and !$_[0]{_ae_sent}
		1402	and Carp::croak "AnyEvent::CondVar: recursive blocking wait detected";
		1403
		1404	local $WAITING = 1;
		1405	AnyEvent->one_event while !$_[0]{_ae_sent};
		1406	}
		1407
		1408	sub recv {
		1409	$_[0]->_wait;
		1410
		1411	Carp::croak $_[0]{_ae_croak} if $_[0]{_ae_croak};
		1412	wantarray ? @{ $_[0]{_ae_sent} } : $_[0]{_ae_sent}[0]
		1413	}
		1414
		1415	sub cb {
		1416	$_[0]{_ae_cb} = $_[1] if @_ > 1;
		1417	$_[0]{_ae_cb}
		1418	}
		1419
		1420	sub begin {
		1421	++$_[0]{_ae_counter};
		1422	$_[0]{_ae_end_cb} = $_[1] if @_ > 1;
		1423	}
		1424
		1425	sub end {
		1426	return if --$_[0]{_ae_counter};
		1427	&{ $_[0]{_ae_end_cb} || sub { $_[0]->send } };
		1428	}
		1429
		1430	# undocumented/compatibility with pre-3.4
		1431	*broadcast = \&send;
		1432	*wait = \&_wait;
		1433
		1434	=head1 ERROR AND EXCEPTION HANDLING
		1435
		1436	In general, AnyEvent does not do any error handling - it relies on the
		1437	caller to do that if required. The L<AnyEvent::Strict> module (see also
		1438	the C<PERL_ANYEVENT_STRICT> environment variable, below) provides strict
		1439	checking of all AnyEvent methods, however, which is highly useful during
		1440	development.
		1441
		1442	As for exception handling (i.e. runtime errors and exceptions thrown while
		1443	executing a callback), this is not only highly event-loop specific, but
		1444	also not in any way wrapped by this module, as this is the job of the main
		1445	program.
		1446
		1447	The pure perl event loop simply re-throws the exception (usually
		1448	within C<< condvar->recv >>), the L<Event> and L<EV> modules call C<<
		1449	$Event/EV::DIED->() >>, L<Glib> uses C<< install_exception_handler >> and
		1450	so on.
		1451
		1452	=head1 ENVIRONMENT VARIABLES
		1453
		1454	The following environment variables are used by this module or its
		1455	submodules.
		1456
		1457	Note that AnyEvent will remove I<all> environment variables starting with
		1458	C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
		1459	enabled.
		1460
		1461	=over 4
		1462
		1463	=item C<PERL_ANYEVENT_VERBOSE>
		1464
		1465	By default, AnyEvent will be completely silent except in fatal
		1466	conditions. You can set this environment variable to make AnyEvent more
		1467	talkative.
		1468
		1469	When set to C<1> or higher, causes AnyEvent to warn about unexpected
		1470	conditions, such as not being able to load the event model specified by
		1471	C<PERL_ANYEVENT_MODEL>.
		1472
		1473	When set to C<2> or higher, cause AnyEvent to report to STDERR which event
		1474	model it chooses.
		1475
		1476	=item C<PERL_ANYEVENT_STRICT>
		1477
		1478	AnyEvent does not do much argument checking by default, as thorough
		1479	argument checking is very costly. Setting this variable to a true value
		1480	will cause AnyEvent to load C<AnyEvent::Strict> and then to thoroughly
		1481	check the arguments passed to most method calls. If it finds any problems,
		1482	it will croak.
		1483
		1484	In other words, enables "strict" mode.
		1485
		1486	Unlike C<use strict>, it is definitely recommended to keep it off in
		1487	production. Keeping C<PERL_ANYEVENT_STRICT=1> in your environment while
		1488	developing programs can be very useful, however.
		1489
		1490	=item C<PERL_ANYEVENT_MODEL>
		1491
		1492	This can be used to specify the event model to be used by AnyEvent, before
		1493	auto detection and -probing kicks in. It must be a string consisting
		1494	entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended
		1495	and the resulting module name is loaded and if the load was successful,
		1496	used as event model. If it fails to load AnyEvent will proceed with
		1497	auto detection and -probing.
		1498
		1499	This functionality might change in future versions.
		1500
		1501	For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you
		1502	could start your program like this:
		1503
		1504	PERL_ANYEVENT_MODEL=Perl perl ...
		1505
		1506	=item C<PERL_ANYEVENT_PROTOCOLS>
		1507
		1508	Used by both L<AnyEvent::DNS> and L<AnyEvent::Socket> to determine preferences
		1509	for IPv4 or IPv6. The default is unspecified (and might change, or be the result
		1510	of auto probing).
		1511
		1512	Must be set to a comma-separated list of protocols or address families,
		1513	current supported: C<ipv4> and C<ipv6>. Only protocols mentioned will be
		1514	used, and preference will be given to protocols mentioned earlier in the
		1515	list.
		1516
		1517	This variable can effectively be used for denial-of-service attacks
		1518	against local programs (e.g. when setuid), although the impact is likely
		1519	small, as the program has to handle conenction and other failures anyways.
		1520
		1521	Examples: C<PERL_ANYEVENT_PROTOCOLS=ipv4,ipv6> - prefer IPv4 over IPv6,
		1522	but support both and try to use both. C<PERL_ANYEVENT_PROTOCOLS=ipv4>
		1523	- only support IPv4, never try to resolve or contact IPv6
		1524	addresses. C<PERL_ANYEVENT_PROTOCOLS=ipv6,ipv4> support either IPv4 or
		1525	IPv6, but prefer IPv6 over IPv4.
		1526
		1527	=item C<PERL_ANYEVENT_EDNS0>
		1528
		1529	Used by L<AnyEvent::DNS> to decide whether to use the EDNS0 extension
		1530	for DNS. This extension is generally useful to reduce DNS traffic, but
		1531	some (broken) firewalls drop such DNS packets, which is why it is off by
		1532	default.
		1533
		1534	Setting this variable to C<1> will cause L<AnyEvent::DNS> to announce
		1535	EDNS0 in its DNS requests.
		1536
		1537	=item C<PERL_ANYEVENT_MAX_FORKS>
		1538
		1539	The maximum number of child processes that C<AnyEvent::Util::fork_call>
		1540	will create in parallel.
		1541
		1542	=item C<PERL_ANYEVENT_MAX_OUTSTANDING_DNS>
		1543
		1544	The default value for the C<max_outstanding> parameter for the default DNS
		1545	resolver - this is the maximum number of parallel DNS requests that are
		1546	sent to the DNS server.
		1547
		1548	=item C<PERL_ANYEVENT_RESOLV_CONF>
		1549
		1550	The file to use instead of F</etc/resolv.conf> (or OS-specific
		1551	configuration) in the default resolver. When set to the empty string, no
		1552	default config will be used.
		1553
		1554	=item C<PERL_ANYEVENT_CA_FILE>, C<PERL_ANYEVENT_CA_PATH>.
		1555
		1556	When neither C<ca_file> nor C<ca_path> was specified during
		1557	L<AnyEvent::TLS> context creation, and either of these environment
		1558	variables exist, they will be used to specify CA certificate locations
		1559	instead of a system-dependent default.
		1560
		1561	=back
628		1562
629	=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE	1563	=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE
630		1564
631	This is an advanced topic that you do not normally need to use AnyEvent in	1565	This is an advanced topic that you do not normally need to use AnyEvent in
632	a module. This section is only of use to event loop authors who want to	1566	a module. This section is only of use to event loop authors who want to
…		…
667	I<rxvt-unicode> also cheats a bit by not providing blocking access to	1601	I<rxvt-unicode> also cheats a bit by not providing blocking access to
668	condition variables: code blocking while waiting for a condition will	1602	condition variables: code blocking while waiting for a condition will
669	C<die>. This still works with most modules/usages, and blocking calls must	1603	C<die>. This still works with most modules/usages, and blocking calls must
670	not be done in an interactive application, so it makes sense.	1604	not be done in an interactive application, so it makes sense.
671		1605
672	=head1 ENVIRONMENT VARIABLES
673
674	The following environment variables are used by this module:
675
676	=over 4
677
678	=item C<PERL_ANYEVENT_VERBOSE>
679
680	By default, AnyEvent will be completely silent except in fatal
681	conditions. You can set this environment variable to make AnyEvent more
682	talkative.
683
684	When set to C<1> or higher, causes AnyEvent to warn about unexpected
685	conditions, such as not being able to load the event model specified by
686	C<PERL_ANYEVENT_MODEL>.
687
688	When set to C<2> or higher, cause AnyEvent to report to STDERR which event
689	model it chooses.
690
691	=item C<PERL_ANYEVENT_MODEL>
692
693	This can be used to specify the event model to be used by AnyEvent, before
694	autodetection and -probing kicks in. It must be a string consisting
695	entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended
696	and the resulting module name is loaded and if the load was successful,
697	used as event model. If it fails to load AnyEvent will proceed with
698	autodetection and -probing.
699
700	This functionality might change in future versions.
701
702	For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you
703	could start your program like this:
704
705	PERL_ANYEVENT_MODEL=Perl perl ...
706
707	=back
708
709	=head1 EXAMPLE PROGRAM	1606	=head1 EXAMPLE PROGRAM
710		1607
711	The following program uses an IO watcher to read data from STDIN, a timer	1608	The following program uses an I/O watcher to read data from STDIN, a timer
712	to display a message once per second, and a condition variable to quit the	1609	to display a message once per second, and a condition variable to quit the
713	program when the user enters quit:	1610	program when the user enters quit:
714		1611
715	use AnyEvent;	1612	use AnyEvent;
716		1613
…		…
721	poll => 'r',	1618	poll => 'r',
722	cb => sub {	1619	cb => sub {
723	warn "io event <$_[0]>\n"; # will always output <r>	1620	warn "io event <$_[0]>\n"; # will always output <r>
724	chomp (my $input = <STDIN>); # read a line	1621	chomp (my $input = <STDIN>); # read a line
725	warn "read: $input\n"; # output what has been read	1622	warn "read: $input\n"; # output what has been read
726	$cv->broadcast if $input =~ /^q/i; # quit program if /^q/i	1623	$cv->send if $input =~ /^q/i; # quit program if /^q/i
727	},	1624	},
728	);	1625	);
729		1626
730	my $time_watcher; # can only be used once	1627	my $time_watcher; # can only be used once
731		1628
…		…
736	});	1633	});
737	}	1634	}
738		1635
739	new_timer; # create first timer	1636	new_timer; # create first timer
740		1637
741	$cv->wait; # wait until user enters /^q/i	1638	$cv->recv; # wait until user enters /^q/i
742		1639
743	=head1 REAL-WORLD EXAMPLE	1640	=head1 REAL-WORLD EXAMPLE
744		1641
745	Consider the L<Net::FCP> module. It features (among others) the following	1642	Consider the L<Net::FCP> module. It features (among others) the following
746	API calls, which are to freenet what HTTP GET requests are to http:	1643	API calls, which are to freenet what HTTP GET requests are to http:
…		…
796	syswrite $txn->{fh}, $txn->{request}	1693	syswrite $txn->{fh}, $txn->{request}
797	or die "connection or write error";	1694	or die "connection or write error";
798	$txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });	1695	$txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });
799		1696
800	Again, C<fh_ready_r> waits till all data has arrived, and then stores the	1697	Again, C<fh_ready_r> waits till all data has arrived, and then stores the
801	result and signals any possible waiters that the request ahs finished:	1698	result and signals any possible waiters that the request has finished:
802		1699
803	sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};	1700	sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};
804		1701
805	if (end-of-file or data complete) {	1702	if (end-of-file or data complete) {
806	$txn->{result} = $txn->{buf};	1703	$txn->{result} = $txn->{buf};
807	$txn->{finished}->broadcast;	1704	$txn->{finished}->send;
808	$txb->{cb}->($txn) of $txn->{cb}; # also call callback	1705	$txb->{cb}->($txn) of $txn->{cb}; # also call callback
809	}	1706	}
810		1707
811	The C<result> method, finally, just waits for the finished signal (if the	1708	The C<result> method, finally, just waits for the finished signal (if the
812	request was already finished, it doesn't wait, of course, and returns the	1709	request was already finished, it doesn't wait, of course, and returns the
813	data:	1710	data:
814		1711
815	$txn->{finished}->wait;	1712	$txn->{finished}->recv;
816	return $txn->{result};	1713	return $txn->{result};
817		1714
818	The actual code goes further and collects all errors (C<die>s, exceptions)	1715	The actual code goes further and collects all errors (C<die>s, exceptions)
819	that occured during request processing. The C<result> method detects	1716	that occurred during request processing. The C<result> method detects
820	whether an exception as thrown (it is stored inside the $txn object)	1717	whether an exception as thrown (it is stored inside the $txn object)
821	and just throws the exception, which means connection errors and other	1718	and just throws the exception, which means connection errors and other
822	problems get reported tot he code that tries to use the result, not in a	1719	problems get reported tot he code that tries to use the result, not in a
823	random callback.	1720	random callback.
824		1721
…		…
855		1752
856	my $quit = AnyEvent->condvar;	1753	my $quit = AnyEvent->condvar;
857		1754
858	$fcp->txn_client_get ($url)->cb (sub {	1755	$fcp->txn_client_get ($url)->cb (sub {
859	...	1756	...
860	$quit->broadcast;	1757	$quit->send;
861	});	1758	});
862		1759
863	$quit->wait;	1760	$quit->recv;
864		1761
865		1762
866	=head1 BENCHMARK	1763	=head1 BENCHMARKS
867		1764
868	To give you an idea of the performance and overheads that AnyEvent adds	1765	To give you an idea of the performance and overheads that AnyEvent adds
869	over the event loops themselves (and to give you an impression of the	1766	over the event loops themselves and to give you an impression of the speed
870	speed of various event loops), here is a benchmark of various supported	1767	of various event loops I prepared some benchmarks.
871	event models natively and with anyevent. The benchmark creates a lot of	1768
872	timers (with a zero timeout) and io watchers (watching STDOUT, a pty, to	1769	=head2 BENCHMARKING ANYEVENT OVERHEAD
		1770
		1771	Here is a benchmark of various supported event models used natively and
		1772	through AnyEvent. The benchmark creates a lot of timers (with a zero
		1773	timeout) and I/O watchers (watching STDOUT, a pty, to become writable,
873	become writable, which it is), lets them fire exactly once and destroys	1774	which it is), lets them fire exactly once and destroys them again.
874	them again.
875		1775
		1776	Source code for this benchmark is found as F<eg/bench> in the AnyEvent
		1777	distribution.
		1778
876	=head2 Explanation of the columns	1779	=head3 Explanation of the columns
877		1780
878	I<watcher> is the number of event watchers created/destroyed. Since	1781	I<watcher> is the number of event watchers created/destroyed. Since
879	different event models feature vastly different performances, each event	1782	different event models feature vastly different performances, each event
880	loop was given a number of watchers so that overall runtime is acceptable	1783	loop was given a number of watchers so that overall runtime is acceptable
881	and similar between tested event loop (and keep them from crashing): Glib	1784	and similar between tested event loop (and keep them from crashing): Glib
…		…
891	all watchers, to avoid adding memory overhead. That means closure creation	1794	all watchers, to avoid adding memory overhead. That means closure creation
892	and memory usage is not included in the figures.	1795	and memory usage is not included in the figures.
893		1796
894	I<invoke> is the time, in microseconds, used to invoke a simple	1797	I<invoke> is the time, in microseconds, used to invoke a simple
895	callback. The callback simply counts down a Perl variable and after it was	1798	callback. The callback simply counts down a Perl variable and after it was
896	invoked "watcher" times, it would C<< ->broadcast >> a condvar once to	1799	invoked "watcher" times, it would C<< ->send >> a condvar once to
897	signal the end of this phase.	1800	signal the end of this phase.
898		1801
899	I<destroy> is the time, in microseconds, that it takes to destroy a single	1802	I<destroy> is the time, in microseconds, that it takes to destroy a single
900	watcher.	1803	watcher.
901		1804
902	=head2 Results	1805	=head3 Results
903		1806
904	name watchers bytes create invoke destroy comment	1807	name watchers bytes create invoke destroy comment
905	EV/EV 400000 244 0.56 0.46 0.31 EV native interface	1808	EV/EV 400000 224 0.47 0.35 0.27 EV native interface
906	EV/Any 100000 610 3.52 0.91 0.75 EV + AnyEvent watchers	1809	EV/Any 100000 224 2.88 0.34 0.27 EV + AnyEvent watchers
907	CoroEV/Any 100000 610 3.49 0.92 0.75 coroutines + Coro::Signal	1810	CoroEV/Any 100000 224 2.85 0.35 0.28 coroutines + Coro::Signal
908	Perl/Any 16000 654 4.64 1.22 0.77 pure perl implementation	1811	Perl/Any 100000 452 4.13 0.73 0.95 pure perl implementation
909	Event/Event 16000 523 28.05 21.38 0.86 Event native interface	1812	Event/Event 16000 517 32.20 31.80 0.81 Event native interface
910	Event/Any 16000 943 34.43 20.48 1.39 Event + AnyEvent watchers	1813	Event/Any 16000 590 35.85 31.55 1.06 Event + AnyEvent watchers
		1814	IOAsync/Any 16000 989 38.10 32.77 11.13 via IO::Async::Loop::IO_Poll
		1815	IOAsync/Any 16000 990 37.59 29.50 10.61 via IO::Async::Loop::Epoll
911	Glib/Any 16000 1357 96.99 12.55 55.51 quadratic behaviour	1816	Glib/Any 16000 1357 102.33 12.31 51.00 quadratic behaviour
912	Tk/Any 2000 1855 27.01 66.61 14.03 SEGV with >> 2000 watchers	1817	Tk/Any 2000 1860 27.20 66.31 14.00 SEGV with >> 2000 watchers
913	POE/Event 2000 6644 108.15 768.19 14.33 via POE::Loop::Event	1818	POE/Event 2000 6328 109.99 751.67 14.02 via POE::Loop::Event
914	POE/Select 2000 6343 94.69 807.65 562.69 via POE::Loop::Select	1819	POE/Select 2000 6027 94.54 809.13 579.80 via POE::Loop::Select
915		1820
916	=head2 Discussion	1821	=head3 Discussion
917		1822
918	The benchmark does I<not> measure scalability of the event loop very	1823	The benchmark does I<not> measure scalability of the event loop very
919	well. For example, a select-based event loop (such as the pure perl one)	1824	well. For example, a select-based event loop (such as the pure perl one)
920	can never compete with an event loop that uses epoll when the number of	1825	can never compete with an event loop that uses epoll when the number of
921	file descriptors grows high. In this benchmark, only a single filehandle	1826	file descriptors grows high. In this benchmark, all events become ready at
922	is used (although some of the AnyEvent adaptors dup() its file descriptor	1827	the same time, so select/poll-based implementations get an unnatural speed
923	to worka round bugs).	1828	boost.
		1829
		1830	Also, note that the number of watchers usually has a nonlinear effect on
		1831	overall speed, that is, creating twice as many watchers doesn't take twice
		1832	the time - usually it takes longer. This puts event loops tested with a
		1833	higher number of watchers at a disadvantage.
		1834
		1835	To put the range of results into perspective, consider that on the
		1836	benchmark machine, handling an event takes roughly 1600 CPU cycles with
		1837	EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU
		1838	cycles with POE.
924		1839
925	C<EV> is the sole leader regarding speed and memory use, which are both	1840	C<EV> is the sole leader regarding speed and memory use, which are both
926	maximal/minimal, respectively. Even when going through AnyEvent, there is	1841	maximal/minimal, respectively. Even when going through AnyEvent, it uses
927	only one event loop that uses less memory (the C<Event> module natively), and	1842	far less memory than any other event loop and is still faster than Event
928	no faster event model, not event C<Event> natively.	1843	natively.
929		1844
930	The pure perl implementation is hit in a few sweet spots (both the	1845	The pure perl implementation is hit in a few sweet spots (both the
931	zero timeout and the use of a single fd hit optimisations in the perl	1846	constant timeout and the use of a single fd hit optimisations in the perl
932	interpreter and the backend itself). Nevertheless tis shows that it	1847	interpreter and the backend itself). Nevertheless this shows that it
933	adds very little overhead in itself. Like any select-based backend its	1848	adds very little overhead in itself. Like any select-based backend its
934	performance becomes really bad with lots of file descriptors, of course,	1849	performance becomes really bad with lots of file descriptors (and few of
935	but this was not subjetc of this benchmark.	1850	them active), of course, but this was not subject of this benchmark.
936		1851
937	The C<Event> module has a relatively high setup and callback invocation cost,	1852	The C<Event> module has a relatively high setup and callback invocation
938	but overall scores on the third place.	1853	cost, but overall scores in on the third place.
939		1854
		1855	C<IO::Async> performs admirably well, about on par with C<Event>, even
		1856	when using its pure perl backend.
		1857
940	C<Glib>'s memory usage is quite a bit bit higher, but it features a	1858	C<Glib>'s memory usage is quite a bit higher, but it features a
941	faster callback invocation and overall ends up in the same class as	1859	faster callback invocation and overall ends up in the same class as
942	C<Event>. However, Glib scales extremely badly, doubling the number of	1860	C<Event>. However, Glib scales extremely badly, doubling the number of
943	watchers increases the processing time by more than a factor of four,	1861	watchers increases the processing time by more than a factor of four,
944	making it completely unusable when using larger numbers of watchers	1862	making it completely unusable when using larger numbers of watchers
945	(note that only a single file descriptor was used in the benchmark, so	1863	(note that only a single file descriptor was used in the benchmark, so
…		…
948	The C<Tk> adaptor works relatively well. The fact that it crashes with	1866	The C<Tk> adaptor works relatively well. The fact that it crashes with
949	more than 2000 watchers is a big setback, however, as correctness takes	1867	more than 2000 watchers is a big setback, however, as correctness takes
950	precedence over speed. Nevertheless, its performance is surprising, as the	1868	precedence over speed. Nevertheless, its performance is surprising, as the
951	file descriptor is dup()ed for each watcher. This shows that the dup()	1869	file descriptor is dup()ed for each watcher. This shows that the dup()
952	employed by some adaptors is not a big performance issue (it does incur a	1870	employed by some adaptors is not a big performance issue (it does incur a
953	hidden memory cost inside the kernel, though, that is not reflected in the	1871	hidden memory cost inside the kernel which is not reflected in the figures
954	figures above).	1872	above).
955		1873
956	C<POE>, regardless of underlying event loop (wether using its pure perl	1874	C<POE>, regardless of underlying event loop (whether using its pure perl
957	select-based backend or the Event module) shows abysmal performance and	1875	select-based backend or the Event module, the POE-EV backend couldn't
		1876	be tested because it wasn't working) shows abysmal performance and
958	memory usage: Watchers use almost 30 times as much memory as EV watchers,	1877	memory usage with AnyEvent: Watchers use almost 30 times as much memory
959	and 10 times as much memory as both Event or EV via AnyEvent. Watcher	1878	as EV watchers, and 10 times as much memory as Event (the high memory
		1879	requirements are caused by requiring a session for each watcher). Watcher
960	invocation is almost 700 times slower than with AnyEvent's pure perl	1880	invocation speed is almost 900 times slower than with AnyEvent's pure perl
		1881	implementation.
		1882
961	implementation. The design of the POE adaptor class in AnyEvent can not	1883	The design of the POE adaptor class in AnyEvent can not really account
962	really account for this, as session creation overhead is small compared	1884	for the performance issues, though, as session creation overhead is
963	to execution of the state machine, which is coded pretty optimally within	1885	small compared to execution of the state machine, which is coded pretty
964	L<AnyEvent::Impl::POE>. POE simply seems to be abysmally slow.	1886	optimally within L<AnyEvent::Impl::POE> (and while everybody agrees that
		1887	using multiple sessions is not a good approach, especially regarding
		1888	memory usage, even the author of POE could not come up with a faster
		1889	design).
965		1890
966	=head2 Summary	1891	=head3 Summary
967		1892
		1893	=over 4
		1894
968	Using EV through AnyEvent is faster than any other event loop, but most	1895	=item * Using EV through AnyEvent is faster than any other event loop
969	event loops have acceptable performance with or without AnyEvent.	1896	(even when used without AnyEvent), but most event loops have acceptable
		1897	performance with or without AnyEvent.
970		1898
971	The overhead AnyEvent adds is usually much smaller than the overhead of	1899	=item * The overhead AnyEvent adds is usually much smaller than the overhead of
972	the actual event loop, only with extremely fast event loops such as the EV	1900	the actual event loop, only with extremely fast event loops such as EV
973	adds AnyEvent significant overhead.	1901	adds AnyEvent significant overhead.
974		1902
975	And you should simply avoid POE like the plague if you want performance or	1903	=item * You should avoid POE like the plague if you want performance or
976	reasonable memory usage.	1904	reasonable memory usage.
977		1905
		1906	=back
		1907
		1908	=head2 BENCHMARKING THE LARGE SERVER CASE
		1909
		1910	This benchmark actually benchmarks the event loop itself. It works by
		1911	creating a number of "servers": each server consists of a socket pair, a
		1912	timeout watcher that gets reset on activity (but never fires), and an I/O
		1913	watcher waiting for input on one side of the socket. Each time the socket
		1914	watcher reads a byte it will write that byte to a random other "server".
		1915
		1916	The effect is that there will be a lot of I/O watchers, only part of which
		1917	are active at any one point (so there is a constant number of active
		1918	fds for each loop iteration, but which fds these are is random). The
		1919	timeout is reset each time something is read because that reflects how
		1920	most timeouts work (and puts extra pressure on the event loops).
		1921
		1922	In this benchmark, we use 10000 socket pairs (20000 sockets), of which 100
		1923	(1%) are active. This mirrors the activity of large servers with many
		1924	connections, most of which are idle at any one point in time.
		1925
		1926	Source code for this benchmark is found as F<eg/bench2> in the AnyEvent
		1927	distribution.
		1928
		1929	=head3 Explanation of the columns
		1930
		1931	I<sockets> is the number of sockets, and twice the number of "servers" (as
		1932	each server has a read and write socket end).
		1933
		1934	I<create> is the time it takes to create a socket pair (which is
		1935	nontrivial) and two watchers: an I/O watcher and a timeout watcher.
		1936
		1937	I<request>, the most important value, is the time it takes to handle a
		1938	single "request", that is, reading the token from the pipe and forwarding
		1939	it to another server. This includes deleting the old timeout and creating
		1940	a new one that moves the timeout into the future.
		1941
		1942	=head3 Results
		1943
		1944	name sockets create request
		1945	EV 20000 69.01 11.16
		1946	Perl 20000 73.32 35.87
		1947	IOAsync 20000 157.00 98.14 epoll
		1948	IOAsync 20000 159.31 616.06 poll
		1949	Event 20000 212.62 257.32
		1950	Glib 20000 651.16 1896.30
		1951	POE 20000 349.67 12317.24 uses POE::Loop::Event
		1952
		1953	=head3 Discussion
		1954
		1955	This benchmark I<does> measure scalability and overall performance of the
		1956	particular event loop.
		1957
		1958	EV is again fastest. Since it is using epoll on my system, the setup time
		1959	is relatively high, though.
		1960
		1961	Perl surprisingly comes second. It is much faster than the C-based event
		1962	loops Event and Glib.
		1963
		1964	IO::Async performs very well when using its epoll backend, and still quite
		1965	good compared to Glib when using its pure perl backend.
		1966
		1967	Event suffers from high setup time as well (look at its code and you will
		1968	understand why). Callback invocation also has a high overhead compared to
		1969	the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event
		1970	uses select or poll in basically all documented configurations.
		1971
		1972	Glib is hit hard by its quadratic behaviour w.r.t. many watchers. It
		1973	clearly fails to perform with many filehandles or in busy servers.
		1974
		1975	POE is still completely out of the picture, taking over 1000 times as long
		1976	as EV, and over 100 times as long as the Perl implementation, even though
		1977	it uses a C-based event loop in this case.
		1978
		1979	=head3 Summary
		1980
		1981	=over 4
		1982
		1983	=item * The pure perl implementation performs extremely well.
		1984
		1985	=item * Avoid Glib or POE in large projects where performance matters.
		1986
		1987	=back
		1988
		1989	=head2 BENCHMARKING SMALL SERVERS
		1990
		1991	While event loops should scale (and select-based ones do not...) even to
		1992	large servers, most programs we (or I :) actually write have only a few
		1993	I/O watchers.
		1994
		1995	In this benchmark, I use the same benchmark program as in the large server
		1996	case, but it uses only eight "servers", of which three are active at any
		1997	one time. This should reflect performance for a small server relatively
		1998	well.
		1999
		2000	The columns are identical to the previous table.
		2001
		2002	=head3 Results
		2003
		2004	name sockets create request
		2005	EV 16 20.00 6.54
		2006	Perl 16 25.75 12.62
		2007	Event 16 81.27 35.86
		2008	Glib 16 32.63 15.48
		2009	POE 16 261.87 276.28 uses POE::Loop::Event
		2010
		2011	=head3 Discussion
		2012
		2013	The benchmark tries to test the performance of a typical small
		2014	server. While knowing how various event loops perform is interesting, keep
		2015	in mind that their overhead in this case is usually not as important, due
		2016	to the small absolute number of watchers (that is, you need efficiency and
		2017	speed most when you have lots of watchers, not when you only have a few of
		2018	them).
		2019
		2020	EV is again fastest.
		2021
		2022	Perl again comes second. It is noticeably faster than the C-based event
		2023	loops Event and Glib, although the difference is too small to really
		2024	matter.
		2025
		2026	POE also performs much better in this case, but is is still far behind the
		2027	others.
		2028
		2029	=head3 Summary
		2030
		2031	=over 4
		2032
		2033	=item * C-based event loops perform very well with small number of
		2034	watchers, as the management overhead dominates.
		2035
		2036	=back
		2037
		2038	=head2 THE IO::Lambda BENCHMARK
		2039
		2040	Recently I was told about the benchmark in the IO::Lambda manpage, which
		2041	could be misinterpreted to make AnyEvent look bad. In fact, the benchmark
		2042	simply compares IO::Lambda with POE, and IO::Lambda looks better (which
		2043	shouldn't come as a surprise to anybody). As such, the benchmark is
		2044	fine, and mostly shows that the AnyEvent backend from IO::Lambda isn't
		2045	very optimal. But how would AnyEvent compare when used without the extra
		2046	baggage? To explore this, I wrote the equivalent benchmark for AnyEvent.
		2047
		2048	The benchmark itself creates an echo-server, and then, for 500 times,
		2049	connects to the echo server, sends a line, waits for the reply, and then
		2050	creates the next connection. This is a rather bad benchmark, as it doesn't
		2051	test the efficiency of the framework or much non-blocking I/O, but it is a
		2052	benchmark nevertheless.
		2053
		2054	name runtime
		2055	Lambda/select 0.330 sec
		2056	+ optimized 0.122 sec
		2057	Lambda/AnyEvent 0.327 sec
		2058	+ optimized 0.138 sec
		2059	Raw sockets/select 0.077 sec
		2060	POE/select, components 0.662 sec
		2061	POE/select, raw sockets 0.226 sec
		2062	POE/select, optimized 0.404 sec
		2063
		2064	AnyEvent/select/nb 0.085 sec
		2065	AnyEvent/EV/nb 0.068 sec
		2066	+state machine 0.134 sec
		2067
		2068	The benchmark is also a bit unfair (my fault): the IO::Lambda/POE
		2069	benchmarks actually make blocking connects and use 100% blocking I/O,
		2070	defeating the purpose of an event-based solution. All of the newly
		2071	written AnyEvent benchmarks use 100% non-blocking connects (using
		2072	AnyEvent::Socket::tcp_connect and the asynchronous pure perl DNS
		2073	resolver), so AnyEvent is at a disadvantage here, as non-blocking connects
		2074	generally require a lot more bookkeeping and event handling than blocking
		2075	connects (which involve a single syscall only).
		2076
		2077	The last AnyEvent benchmark additionally uses L<AnyEvent::Handle>, which
		2078	offers similar expressive power as POE and IO::Lambda, using conventional
		2079	Perl syntax. This means that both the echo server and the client are 100%
		2080	non-blocking, further placing it at a disadvantage.
		2081
		2082	As you can see, the AnyEvent + EV combination even beats the
		2083	hand-optimised "raw sockets benchmark", while AnyEvent + its pure perl
		2084	backend easily beats IO::Lambda and POE.
		2085
		2086	And even the 100% non-blocking version written using the high-level (and
		2087	slow :) L<AnyEvent::Handle> abstraction beats both POE and IO::Lambda by a
		2088	large margin, even though it does all of DNS, tcp-connect and socket I/O
		2089	in a non-blocking way.
		2090
		2091	The two AnyEvent benchmarks programs can be found as F<eg/ae0.pl> and
		2092	F<eg/ae2.pl> in the AnyEvent distribution, the remaining benchmarks are
		2093	part of the IO::lambda distribution and were used without any changes.
		2094
		2095
		2096	=head1 SIGNALS
		2097
		2098	AnyEvent currently installs handlers for these signals:
		2099
		2100	=over 4
		2101
		2102	=item SIGCHLD
		2103
		2104	A handler for C<SIGCHLD> is installed by AnyEvent's child watcher
		2105	emulation for event loops that do not support them natively. Also, some
		2106	event loops install a similar handler.
		2107
		2108	Additionally, when AnyEvent is loaded and SIGCHLD is set to IGNORE, then
		2109	AnyEvent will reset it to default, to avoid losing child exit statuses.
		2110
		2111	=item SIGPIPE
		2112
		2113	A no-op handler is installed for C<SIGPIPE> when C<$SIG{PIPE}> is C<undef>
		2114	when AnyEvent gets loaded.
		2115
		2116	The rationale for this is that AnyEvent users usually do not really depend
		2117	on SIGPIPE delivery (which is purely an optimisation for shell use, or
		2118	badly-written programs), but C<SIGPIPE> can cause spurious and rare
		2119	program exits as a lot of people do not expect C<SIGPIPE> when writing to
		2120	some random socket.
		2121
		2122	The rationale for installing a no-op handler as opposed to ignoring it is
		2123	that this way, the handler will be restored to defaults on exec.
		2124
		2125	Feel free to install your own handler, or reset it to defaults.
		2126
		2127	=back
		2128
		2129	=cut
		2130
		2131	undef $SIG{CHLD}
		2132	if $SIG{CHLD} eq 'IGNORE';
		2133
		2134	$SIG{PIPE} = sub { }
		2135	unless defined $SIG{PIPE};
978		2136
979	=head1 FORK	2137	=head1 FORK
980		2138
981	Most event libraries are not fork-safe. The ones who are usually are	2139	Most event libraries are not fork-safe. The ones who are usually are
982	because they are so inefficient. Only L<EV> is fully fork-aware.	2140	because they rely on inefficient but fork-safe C<select> or C<poll>
		2141	calls. Only L<EV> is fully fork-aware.
983		2142
984	If you have to fork, you must either do so I<before> creating your first	2143	If you have to fork, you must either do so I<before> creating your first
985	watcher OR you must not use AnyEvent at all in the child.	2144	watcher OR you must not use AnyEvent at all in the child.
986		2145
987		2146
…		…
995	specified in the variable.	2154	specified in the variable.
996		2155
997	You can make AnyEvent completely ignore this variable by deleting it	2156	You can make AnyEvent completely ignore this variable by deleting it
998	before the first watcher gets created, e.g. with a C<BEGIN> block:	2157	before the first watcher gets created, e.g. with a C<BEGIN> block:
999		2158
1000	BEGIN { delete $ENV{PERL_ANYEVENT_MODEL} }	2159	BEGIN { delete $ENV{PERL_ANYEVENT_MODEL} }
1001		2160
1002	use AnyEvent;	2161	use AnyEvent;
		2162
		2163	Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can
		2164	be used to probe what backend is used and gain other information (which is
		2165	probably even less useful to an attacker than PERL_ANYEVENT_MODEL), and
		2166	$ENV{PERL_ANYEVENT_STRICT}.
		2167
		2168	Note that AnyEvent will remove I<all> environment variables starting with
		2169	C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
		2170	enabled.
		2171
		2172
		2173	=head1 BUGS
		2174
		2175	Perl 5.8 has numerous memleaks that sometimes hit this module and are hard
		2176	to work around. If you suffer from memleaks, first upgrade to Perl 5.10
		2177	and check wether the leaks still show up. (Perl 5.10.0 has other annoying
		2178	memleaks, such as leaking on C<map> and C<grep> but it is usually not as
		2179	pronounced).
1003		2180
1004		2181
1005	=head1 SEE ALSO	2182	=head1 SEE ALSO
1006		2183
1007	Event modules: L<Coro::EV>, L<EV>, L<EV::Glib>, L<Glib::EV>,	2184	Utility functions: L<AnyEvent::Util>.
1008	L<Coro::Event>, L<Event>, L<Glib::Event>, L<Glib>, L<Coro>, L<Tk>,	2185
		2186	Event modules: L<EV>, L<EV::Glib>, L<Glib::EV>, L<Event>, L<Glib::Event>,
1009	L<Event::Lib>, L<Qt>, L<POE>.	2187	L<Glib>, L<Tk>, L<Event::Lib>, L<Qt>, L<POE>.
1010		2188
1011	Implementations: L<AnyEvent::Impl::CoroEV>, L<AnyEvent::Impl::EV>,	2189	Implementations: L<AnyEvent::Impl::EV>, L<AnyEvent::Impl::Event>,
1012	L<AnyEvent::Impl::CoroEvent>, L<AnyEvent::Impl::Event>, L<AnyEvent::Impl::Glib>,	2190	L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>,
1013	L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>, L<AnyEvent::Impl::EventLib>,	2191	L<AnyEvent::Impl::EventLib>, L<AnyEvent::Impl::Qt>,
1014	L<AnyEvent::Impl::Qt>, L<AnyEvent::Impl::POE>.	2192	L<AnyEvent::Impl::POE>, L<AnyEvent::Impl::IOAsync>.
1015		2193
		2194	Non-blocking file handles, sockets, TCP clients and
		2195	servers: L<AnyEvent::Handle>, L<AnyEvent::Socket>, L<AnyEvent::TLS>.
		2196
		2197	Asynchronous DNS: L<AnyEvent::DNS>.
		2198
		2199	Coroutine support: L<Coro>, L<Coro::AnyEvent>, L<Coro::EV>,
		2200	L<Coro::Event>,
		2201
1016	Nontrivial usage examples: L<Net::FCP>, L<Net::XMPP2>.	2202	Nontrivial usage examples: L<AnyEvent::GPSD>, L<AnyEvent::XMPP>,
		2203	L<AnyEvent::HTTP>.
1017		2204
1018		2205
1019	=head1 AUTHOR	2206	=head1 AUTHOR
1020		2207
1021	Marc Lehmann <schmorp@schmorp.de>	2208	Marc Lehmann <schmorp@schmorp.de>
1022	http://home.schmorp.de/	2209	http://home.schmorp.de/
1023		2210
1024	=cut	2211	=cut
1025		2212
1026	1	2213	1
1027		2214

Diff Legend

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