ViewVC Help

View File | Revision Log | Show Annotations | Download File

/cvs/AnyEvent/lib/AnyEvent.pm

Comparing AnyEvent/lib/AnyEvent.pm (file contents):
Revision 1.98 by root, Sun Apr 27 16:31:48 2008 UTC vs.
Revision 1.289 by root, Tue Sep 1 16:44:58 2009 UTC

1	=head1 NAME	1	=head1 NAME
2		2
3	AnyEvent - provide framework for multiple event loops	3	AnyEvent - the DBI of event loop programming
4		4
5	EV, Event, Coro::EV, Coro::Event, Glib, Tk, Perl, Event::Lib, Qt, POE - various supported event loops	5	EV, Event, Glib, Tk, Perl, Event::Lib, Irssi, rxvt-unicode, IO::Async, Qt
		6	and POE are various supported event loops/environments.
6		7
7	=head1 SYNOPSIS	8	=head1 SYNOPSIS
8		9
9	use AnyEvent;	10	use AnyEvent;
10		11
		12	# file descriptor readable
11	my $w = AnyEvent->io (fh => $fh, poll => "r|w", cb => sub {	13	my $w = AnyEvent->io (fh => $fh, poll => "r", cb => sub { ... });
		14
		15	# one-shot or repeating timers
		16	my $w = AnyEvent->timer (after => $seconds, cb => sub { ... });
		17	my $w = AnyEvent->timer (after => $seconds, interval => $seconds, cb => ...
		18
		19	print AnyEvent->now; # prints current event loop time
		20	print AnyEvent->time; # think Time::HiRes::time or simply CORE::time.
		21
		22	# POSIX signal
		23	my $w = AnyEvent->signal (signal => "TERM", cb => sub { ... });
		24
		25	# child process exit
		26	my $w = AnyEvent->child (pid => $pid, cb => sub {
		27	my ($pid, $status) = @_;
12	...	28	...
13	});	29	});
14		30
15	my $w = AnyEvent->timer (after => $seconds, cb => sub {	31	# called when event loop idle (if applicable)
16	...	32	my $w = AnyEvent->idle (cb => sub { ... });
17	});
18		33
19	my $w = AnyEvent->condvar; # stores whether a condition was flagged	34	my $w = AnyEvent->condvar; # stores whether a condition was flagged
		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.
		45
		46	=head1 SUPPORT
		47
		48	There is a mailinglist for discussing all things AnyEvent, and an IRC
		49	channel, too.
		50
		51	See the AnyEvent project page at the B<Schmorpforge Ta-Sa Software
		52	Repository>, at L<http://anyevent.schmorp.de>, for more info.
22		53
23	=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)	54	=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)
24		55
25	Glib, POE, IO::Async, Event... CPAN offers event models by the dozen	56	Glib, POE, IO::Async, Event... CPAN offers event models by the dozen
26	nowadays. So what is different about AnyEvent?	57	nowadays. So what is different about AnyEvent?
27		58
28	Executive Summary: AnyEvent is I<compatible>, AnyEvent is I<free of	59	Executive Summary: AnyEvent is I<compatible>, AnyEvent is I<free of
29	policy> and AnyEvent is I<small and efficient>.	60	policy> and AnyEvent is I<small and efficient>.
30		61
31	First and foremost, I<AnyEvent is not an event model> itself, it only	62	First and foremost, I<AnyEvent is not an event model> itself, it only
32	interfaces to whatever event model the main program happens to use in a	63	interfaces to whatever event model the main program happens to use, in a
33	pragmatic way. For event models and certain classes of immortals alike,	64	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,	65	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	66	only one event loop can be active at the same time in a process. AnyEvent
36	helps hiding the differences between those event loops.	67	cannot change this, but it can hide the differences between those event
		68	loops.
37		69
38	The goal of AnyEvent is to offer module authors the ability to do event	70	The goal of AnyEvent is to offer module authors the ability to do event
39	programming (waiting for I/O or timer events) without subscribing to a	71	programming (waiting for I/O or timer events) without subscribing to a
40	religion, a way of living, and most importantly: without forcing your	72	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	73	module users into the same thing by forcing them to use the same event
42	model you use.	74	model you use.
43		75
44	For modules like POE or IO::Async (which is a total misnomer as it is	76	For modules like POE or IO::Async (which is a total misnomer as it is
45	actually doing all I/O I<synchronously>...), using them in your module is	77	actually doing all I/O I<synchronously>...), using them in your module is
46	like joining a cult: After you joined, you are dependent on them and you	78	like joining a cult: After you joined, you are dependent on them and you
47	cannot use anything else, as it is simply incompatible to everything that	79	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	80	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.	81	module are I<also> forced to use the same event loop you use.
50		82
51	AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works	83	AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works
52	fine. AnyEvent + Tk works fine etc. etc. but none of these work together	84	fine. AnyEvent + Tk works fine etc. etc. but none of these work together
53	with the rest: POE + IO::Async? no go. Tk + Event? no go. Again: if	85	with the rest: POE + IO::Async? No go. Tk + Event? No go. Again: if
54	your module uses one of those, every user of your module has to use it,	86	your module uses one of those, every user of your module has to use it,
55	too. But if your module uses AnyEvent, it works transparently with all	87	too. But if your module uses AnyEvent, it works transparently with all
56	event models it supports (including stuff like POE and IO::Async, as long	88	event models it supports (including stuff like IO::Async, as long as those
57	as those use one of the supported event loops. It is trivial to add new	89	use one of the supported event loops. It is trivial to add new event loops
58	event loops to AnyEvent, too, so it is future-proof).	90	to AnyEvent, too, so it is future-proof).
59		91
60	In addition to being free of having to use I<the one and only true event	92	In addition to being free of having to use I<the one and only true event
61	model>, AnyEvent also is free of bloat and policy: with POE or similar	93	model>, AnyEvent also is free of bloat and policy: with POE or similar
62	modules, you get an enourmous amount of code and strict rules you have to	94	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	95	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	96	offering the functionality that is necessary, in as thin as a wrapper as
65	technically possible.	97	technically possible.
66		98
		99	Of course, AnyEvent comes with a big (and fully optional!) toolbox
		100	of useful functionality, such as an asynchronous DNS resolver, 100%
		101	non-blocking connects (even with TLS/SSL, IPv6 and on broken platforms
		102	such as Windows) and lots of real-world knowledge and workarounds for
		103	platform bugs and differences.
		104
67	Of course, if you want lots of policy (this can arguably be somewhat	105	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	106	useful) and you want to force your users to use the one and only event
69	model, you should I<not> use this module.	107	model, you should I<not> use this module.
70
71	#TODO#
72
73	Net::IRC3
74	AnyEvent::HTTPD
75	AnyEvent::DNS
76	IO::AnyEvent
77	Net::FPing
78	Net::XMPP2
79	Coro
80
81	AnyEvent::IRC
82	AnyEvent::HTTPD
83	AnyEvent::DNS
84	AnyEvent::Handle
85	AnyEvent::Socket
86	AnyEvent::FPing
87	AnyEvent::XMPP
88	AnyEvent::SNMP
89	Coro
90		108
91	=head1 DESCRIPTION	109	=head1 DESCRIPTION
92		110
93	L<AnyEvent> provides an identical interface to multiple event loops. This	111	L<AnyEvent> provides an identical interface to multiple event loops. This
94	allows module authors to utilise an event loop without forcing module	112	allows module authors to utilise an event loop without forcing module
…		…
98	The interface itself is vaguely similar, but not identical to the L<Event>	116	The interface itself is vaguely similar, but not identical to the L<Event>
99	module.	117	module.
100		118
101	During the first call of any watcher-creation method, the module tries	119	During the first call of any watcher-creation method, the module tries
102	to detect the currently loaded event loop by probing whether one of the	120	to detect the currently loaded event loop by probing whether one of the
103	following modules is already loaded: L<Coro::EV>, L<Coro::Event>, L<EV>,	121	following modules is already loaded: L<EV>,
104	L<Event>, L<Glib>, L<AnyEvent::Impl::Perl>, L<Tk>, L<Event::Lib>, L<Qt>,	122	L<Event>, L<Glib>, L<AnyEvent::Impl::Perl>, L<Tk>, L<Event::Lib>, L<Qt>,
105	L<POE>. The first one found is used. If none are found, the module tries	123	L<POE>. The first one found is used. If none are found, the module tries
106	to load these modules (excluding Tk, Event::Lib, Qt and POE as the pure perl	124	to load these modules (excluding Tk, Event::Lib, Qt and POE as the pure perl
107	adaptor should always succeed) in the order given. The first one that can	125	adaptor should always succeed) in the order given. The first one that can
108	be successfully loaded will be used. If, after this, still none could be	126	be successfully loaded will be used. If, after this, still none could be
…		…
122	starts using it, all bets are off. Maybe you should tell their authors to	140	starts using it, all bets are off. Maybe you should tell their authors to
123	use AnyEvent so their modules work together with others seamlessly...	141	use AnyEvent so their modules work together with others seamlessly...
124		142
125	The pure-perl implementation of AnyEvent is called	143	The pure-perl implementation of AnyEvent is called
126	C<AnyEvent::Impl::Perl>. Like other event modules you can load it	144	C<AnyEvent::Impl::Perl>. Like other event modules you can load it
127	explicitly.	145	explicitly and enjoy the high availability of that event loop :)
128		146
129	=head1 WATCHERS	147	=head1 WATCHERS
130		148
131	AnyEvent has the central concept of a I<watcher>, which is an object that	149	AnyEvent has the central concept of a I<watcher>, which is an object that
132	stores relevant data for each kind of event you are waiting for, such as	150	stores relevant data for each kind of event you are waiting for, such as
133	the callback to call, the filehandle to watch, etc.	151	the callback to call, the file handle to watch, etc.
134		152
135	These watchers are normal Perl objects with normal Perl lifetime. After	153	These watchers are normal Perl objects with normal Perl lifetime. After
136	creating a watcher it will immediately "watch" for events and invoke the	154	creating a watcher it will immediately "watch" for events and invoke the
137	callback when the event occurs (of course, only when the event model	155	callback when the event occurs (of course, only when the event model
138	is in control).	156	is in control).
139		157
		158	Note that B<callbacks must not permanently change global variables>
		159	potentially in use by the event loop (such as C<$_> or C<$[>) and that B<<
		160	callbacks must not C<die> >>. The former is good programming practise in
		161	Perl and the latter stems from the fact that exception handling differs
		162	widely between event loops.
		163
140	To disable the watcher you have to destroy it (e.g. by setting the	164	To disable the watcher you have to destroy it (e.g. by setting the
141	variable you store it in to C<undef> or otherwise deleting all references	165	variable you store it in to C<undef> or otherwise deleting all references
142	to it).	166	to it).
143		167
144	All watchers are created by calling a method on the C<AnyEvent> class.	168	All watchers are created by calling a method on the C<AnyEvent> class.
…		…
146	Many watchers either are used with "recursion" (repeating timers for	170	Many watchers either are used with "recursion" (repeating timers for
147	example), or need to refer to their watcher object in other ways.	171	example), or need to refer to their watcher object in other ways.
148		172
149	An any way to achieve that is this pattern:	173	An any way to achieve that is this pattern:
150		174
151	my $w; $w = AnyEvent->type (arg => value ..., cb => sub {	175	my $w; $w = AnyEvent->type (arg => value ..., cb => sub {
152	# you can use $w here, for example to undef it	176	# you can use $w here, for example to undef it
153	undef $w;	177	undef $w;
154	});	178	});
155		179
156	Note that C<my $w; $w => combination. This is necessary because in Perl,	180	Note that C<my $w; $w => combination. This is necessary because in Perl,
157	my variables are only visible after the statement in which they are	181	my variables are only visible after the statement in which they are
158	declared.	182	declared.
159		183
160	=head2 I/O WATCHERS	184	=head2 I/O WATCHERS
161		185
		186	$w = AnyEvent->io (
		187	fh => <filehandle_or_fileno>,
		188	poll => <"r" or "w">,
		189	cb => <callback>,
		190	);
		191
162	You can create an I/O watcher by calling the C<< AnyEvent->io >> method	192	You can create an I/O watcher by calling the C<< AnyEvent->io >> method
163	with the following mandatory key-value pairs as arguments:	193	with the following mandatory key-value pairs as arguments:
164		194
165	C<fh> the Perl I<file handle> (I<not> file descriptor) to watch	195	C<fh> is the Perl I<file handle> (or a naked file descriptor) to watch
		196	for events (AnyEvent might or might not keep a reference to this file
		197	handle). Note that only file handles pointing to things for which
		198	non-blocking operation makes sense are allowed. This includes sockets,
		199	most character devices, pipes, fifos and so on, but not for example files
		200	or block devices.
		201
166	for events. C<poll> must be a string that is either C<r> or C<w>,	202	C<poll> must be a string that is either C<r> or C<w>, which creates a
167	which creates a watcher waiting for "r"eadable or "w"ritable events,	203	watcher waiting for "r"eadable or "w"ritable events, respectively.
		204
168	respectively. C<cb> is the callback to invoke each time the file handle	205	C<cb> is the callback to invoke each time the file handle becomes ready.
169	becomes ready.
170		206
171	Although the callback might get passed parameters, their value and	207	Although the callback might get passed parameters, their value and
172	presence is undefined and you cannot rely on them. Portable AnyEvent	208	presence is undefined and you cannot rely on them. Portable AnyEvent
173	callbacks cannot use arguments passed to I/O watcher callbacks.	209	callbacks cannot use arguments passed to I/O watcher callbacks.
174		210
…		…
178		214
179	Some event loops issue spurious readyness notifications, so you should	215	Some event loops issue spurious readyness notifications, so you should
180	always use non-blocking calls when reading/writing from/to your file	216	always use non-blocking calls when reading/writing from/to your file
181	handles.	217	handles.
182		218
183	Example:
184
185	# wait for readability of STDIN, then read a line and disable the watcher	219	Example: wait for readability of STDIN, then read a line and disable the
		220	watcher.
		221
186	my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {	222	my $w; $w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
187	chomp (my $input = <STDIN>);	223	chomp (my $input = <STDIN>);
188	warn "read: $input\n";	224	warn "read: $input\n";
189	undef $w;	225	undef $w;
190	});	226	});
191		227
192	=head2 TIME WATCHERS	228	=head2 TIME WATCHERS
193		229
		230	$w = AnyEvent->timer (after => <seconds>, cb => <callback>);
		231
		232	$w = AnyEvent->timer (
		233	after => <fractional_seconds>,
		234	interval => <fractional_seconds>,
		235	cb => <callback>,
		236	);
		237
194	You can create a time watcher by calling the C<< AnyEvent->timer >>	238	You can create a time watcher by calling the C<< AnyEvent->timer >>
195	method with the following mandatory arguments:	239	method with the following mandatory arguments:
196		240
197	C<after> specifies after how many seconds (fractional values are	241	C<after> specifies after how many seconds (fractional values are
198	supported) the callback should be invoked. C<cb> is the callback to invoke	242	supported) the callback should be invoked. C<cb> is the callback to invoke
…		…
200		244
201	Although the callback might get passed parameters, their value and	245	Although the callback might get passed parameters, their value and
202	presence is undefined and you cannot rely on them. Portable AnyEvent	246	presence is undefined and you cannot rely on them. Portable AnyEvent
203	callbacks cannot use arguments passed to time watcher callbacks.	247	callbacks cannot use arguments passed to time watcher callbacks.
204		248
205	The timer callback will be invoked at most once: if you want a repeating	249	The callback will normally be invoked once only. If you specify another
206	timer you have to create a new watcher (this is a limitation by both Tk	250	parameter, C<interval>, as a strictly positive number (> 0), then the
207	and Glib).	251	callback will be invoked regularly at that interval (in fractional
		252	seconds) after the first invocation. If C<interval> is specified with a
		253	false value, then it is treated as if it were missing.
208		254
209	Example:	255	The callback will be rescheduled before invoking the callback, but no
		256	attempt is done to avoid timer drift in most backends, so the interval is
		257	only approximate.
210		258
211	# fire an event after 7.7 seconds	259	Example: fire an event after 7.7 seconds.
		260
212	my $w = AnyEvent->timer (after => 7.7, cb => sub {	261	my $w = AnyEvent->timer (after => 7.7, cb => sub {
213	warn "timeout\n";	262	warn "timeout\n";
214	});	263	});
215		264
216	# to cancel the timer:	265	# to cancel the timer:
217	undef $w;	266	undef $w;
218		267
219	Example 2:
220
221	# fire an event after 0.5 seconds, then roughly every second	268	Example 2: fire an event after 0.5 seconds, then roughly every second.
222	my $w;
223		269
224	my $cb = sub {
225	# cancel the old timer while creating a new one
226	$w = AnyEvent->timer (after => 1, cb => $cb);	270	my $w = AnyEvent->timer (after => 0.5, interval => 1, cb => sub {
		271	warn "timeout\n";
227	};	272	};
228
229	# start the "loop" by creating the first watcher
230	$w = AnyEvent->timer (after => 0.5, cb => $cb);
231		273
232	=head3 TIMING ISSUES	274	=head3 TIMING ISSUES
233		275
234	There are two ways to handle timers: based on real time (relative, "fire	276	There are two ways to handle timers: based on real time (relative, "fire
235	in 10 seconds") and based on wallclock time (absolute, "fire at 12	277	in 10 seconds") and based on wallclock time (absolute, "fire at 12
…		…
247	timers.	289	timers.
248		290
249	AnyEvent always prefers relative timers, if available, matching the	291	AnyEvent always prefers relative timers, if available, matching the
250	AnyEvent API.	292	AnyEvent API.
251		293
		294	AnyEvent has two additional methods that return the "current time":
		295
		296	=over 4
		297
		298	=item AnyEvent->time
		299
		300	This returns the "current wallclock time" as a fractional number of
		301	seconds since the Epoch (the same thing as C<time> or C<Time::HiRes::time>
		302	return, and the result is guaranteed to be compatible with those).
		303
		304	It progresses independently of any event loop processing, i.e. each call
		305	will check the system clock, which usually gets updated frequently.
		306
		307	=item AnyEvent->now
		308
		309	This also returns the "current wallclock time", but unlike C<time>, above,
		310	this value might change only once per event loop iteration, depending on
		311	the event loop (most return the same time as C<time>, above). This is the
		312	time that AnyEvent's timers get scheduled against.
		313
		314	I<In almost all cases (in all cases if you don't care), this is the
		315	function to call when you want to know the current time.>
		316
		317	This function is also often faster then C<< AnyEvent->time >>, and
		318	thus the preferred method if you want some timestamp (for example,
		319	L<AnyEvent::Handle> uses this to update it's activity timeouts).
		320
		321	The rest of this section is only of relevance if you try to be very exact
		322	with your timing, you can skip it without bad conscience.
		323
		324	For a practical example of when these times differ, consider L<Event::Lib>
		325	and L<EV> and the following set-up:
		326
		327	The event loop is running and has just invoked one of your callback at
		328	time=500 (assume no other callbacks delay processing). In your callback,
		329	you wait a second by executing C<sleep 1> (blocking the process for a
		330	second) and then (at time=501) you create a relative timer that fires
		331	after three seconds.
		332
		333	With L<Event::Lib>, C<< AnyEvent->time >> and C<< AnyEvent->now >> will
		334	both return C<501>, because that is the current time, and the timer will
		335	be scheduled to fire at time=504 (C<501> + C<3>).
		336
		337	With L<EV>, C<< AnyEvent->time >> returns C<501> (as that is the current
		338	time), but C<< AnyEvent->now >> returns C<500>, as that is the time the
		339	last event processing phase started. With L<EV>, your timer gets scheduled
		340	to run at time=503 (C<500> + C<3>).
		341
		342	In one sense, L<Event::Lib> is more exact, as it uses the current time
		343	regardless of any delays introduced by event processing. However, most
		344	callbacks do not expect large delays in processing, so this causes a
		345	higher drift (and a lot more system calls to get the current time).
		346
		347	In another sense, L<EV> is more exact, as your timer will be scheduled at
		348	the same time, regardless of how long event processing actually took.
		349
		350	In either case, if you care (and in most cases, you don't), then you
		351	can get whatever behaviour you want with any event loop, by taking the
		352	difference between C<< AnyEvent->time >> and C<< AnyEvent->now >> into
		353	account.
		354
		355	=item AnyEvent->now_update
		356
		357	Some event loops (such as L<EV> or L<AnyEvent::Impl::Perl>) cache
		358	the current time for each loop iteration (see the discussion of L<<
		359	AnyEvent->now >>, above).
		360
		361	When a callback runs for a long time (or when the process sleeps), then
		362	this "current" time will differ substantially from the real time, which
		363	might affect timers and time-outs.
		364
		365	When this is the case, you can call this method, which will update the
		366	event loop's idea of "current time".
		367
		368	Note that updating the time I<might> cause some events to be handled.
		369
		370	=back
		371
252	=head2 SIGNAL WATCHERS	372	=head2 SIGNAL WATCHERS
253		373
		374	$w = AnyEvent->signal (signal => <uppercase_signal_name>, cb => <callback>);
		375
254	You can watch for signals using a signal watcher, C<signal> is the signal	376	You can watch for signals using a signal watcher, C<signal> is the signal
255	I<name> without any C<SIG> prefix, C<cb> is the Perl callback to	377	I<name> in uppercase and without any C<SIG> prefix, C<cb> is the Perl
256	be invoked whenever a signal occurs.	378	callback to be invoked whenever a signal occurs.
257		379
258	Although the callback might get passed parameters, their value and	380	Although the callback might get passed parameters, their value and
259	presence is undefined and you cannot rely on them. Portable AnyEvent	381	presence is undefined and you cannot rely on them. Portable AnyEvent
260	callbacks cannot use arguments passed to signal watcher callbacks.	382	callbacks cannot use arguments passed to signal watcher callbacks.
261		383
262	Multiple signal occurances can be clumped together into one callback	384	Multiple signal occurrences can be clumped together into one callback
263	invocation, and callback invocation will be synchronous. synchronous means	385	invocation, and callback invocation will be synchronous. Synchronous means
264	that it might take a while until the signal gets handled by the process,	386	that it might take a while until the signal gets handled by the process,
265	but it is guarenteed not to interrupt any other callbacks.	387	but it is guaranteed not to interrupt any other callbacks.
266		388
267	The main advantage of using these watchers is that you can share a signal	389	The main advantage of using these watchers is that you can share a signal
268	between multiple watchers.	390	between multiple watchers, and AnyEvent will ensure that signals will not
		391	interrupt your program at bad times.
269		392
270	This watcher might use C<%SIG>, so programs overwriting those signals	393	This watcher might use C<%SIG> (depending on the event loop used),
271	directly will likely not work correctly.	394	so programs overwriting those signals directly will likely not work
		395	correctly.
272		396
273	Example: exit on SIGINT	397	Example: exit on SIGINT
274		398
275	my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 });	399	my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 });
276		400
		401	=head3 Signal Races, Delays and Workarounds
		402
		403	Many event loops (e.g. Glib, Tk, Qt, IO::Async) do not support attaching
		404	callbacks to signals in a generic way, which is a pity, as you cannot
		405	do race-free signal handling in perl, requiring C libraries for
		406	this. AnyEvent will try to do it's best, which means in some cases,
		407	signals will be delayed. The maximum time a signal might be delayed is
		408	specified in C<$AnyEvent::MAX_SIGNAL_LATENCY> (default: 10 seconds). This
		409	variable can be changed only before the first signal watcher is created,
		410	and should be left alone otherwise. This variable determines how often
		411	AnyEvent polls for signals (in case a wake-up was missed). Higher values
		412	will cause fewer spurious wake-ups, which is better for power and CPU
		413	saving.
		414
		415	All these problems can be avoided by installing the optional
		416	L<Async::Interrupt> module, which works with most event loops. It will not
		417	work with inherently broken event loops such as L<Event> or L<Event::Lib>
		418	(and not with L<POE> currently, as POE does it's own workaround with
		419	one-second latency). For those, you just have to suffer the delays.
		420
277	=head2 CHILD PROCESS WATCHERS	421	=head2 CHILD PROCESS WATCHERS
278		422
		423	$w = AnyEvent->child (pid => <process id>, cb => <callback>);
		424
279	You can also watch on a child process exit and catch its exit status.	425	You can also watch on a child process exit and catch its exit status.
280		426
281	The child process is specified by the C<pid> argument (if set to C<0>, it	427	The child process is specified by the C<pid> argument (one some backends,
282	watches for any child process exit). The watcher will trigger as often	428	using C<0> watches for any child process exit, on others this will
283	as status change for the child are received. This works by installing a	429	croak). The watcher will be triggered only when the child process has
284	signal handler for C<SIGCHLD>. The callback will be called with the pid	430	finished and an exit status is available, not on any trace events
285	and exit status (as returned by waitpid), so unlike other watcher types,	431	(stopped/continued).
286	you I<can> rely on child watcher callback arguments.	432
		433	The callback will be called with the pid and exit status (as returned by
		434	waitpid), so unlike other watcher types, you I<can> rely on child watcher
		435	callback arguments.
		436
		437	This watcher type works by installing a signal handler for C<SIGCHLD>,
		438	and since it cannot be shared, nothing else should use SIGCHLD or reap
		439	random child processes (waiting for specific child processes, e.g. inside
		440	C<system>, is just fine).
287		441
288	There is a slight catch to child watchers, however: you usually start them	442	There is a slight catch to child watchers, however: you usually start them
289	I<after> the child process was created, and this means the process could	443	I<after> the child process was created, and this means the process could
290	have exited already (and no SIGCHLD will be sent anymore).	444	have exited already (and no SIGCHLD will be sent anymore).
291		445
292	Not all event models handle this correctly (POE doesn't), but even for	446	Not all event models handle this correctly (neither POE nor IO::Async do,
		447	see their AnyEvent::Impl manpages for details), but even for event models
293	event models that I<do> handle this correctly, they usually need to be	448	that I<do> handle this correctly, they usually need to be loaded before
294	loaded before the process exits (i.e. before you fork in the first place).	449	the process exits (i.e. before you fork in the first place). AnyEvent's
		450	pure perl event loop handles all cases correctly regardless of when you
		451	start the watcher.
295		452
296	This means you cannot create a child watcher as the very first thing in an	453	This means you cannot create a child watcher as the very first
297	AnyEvent program, you I<have> to create at least one watcher before you	454	thing in an AnyEvent program, you I<have> to create at least one
298	C<fork> the child (alternatively, you can call C<AnyEvent::detect>).	455	watcher before you C<fork> the child (alternatively, you can call
		456	C<AnyEvent::detect>).
		457
		458	As most event loops do not support waiting for child events, they will be
		459	emulated by AnyEvent in most cases, in which the latency and race problems
		460	mentioned in the description of signal watchers apply.
299		461
300	Example: fork a process and wait for it	462	Example: fork a process and wait for it
301		463
302	my $done = AnyEvent->condvar;	464	my $done = AnyEvent->condvar;
303		465
304	AnyEvent::detect; # force event module to be initialised
305
306	my $pid = fork or exit 5;	466	my $pid = fork or exit 5;
307		467
308	my $w = AnyEvent->child (	468	my $w = AnyEvent->child (
309	pid => $pid,	469	pid => $pid,
310	cb => sub {	470	cb => sub {
311	my ($pid, $status) = @_;	471	my ($pid, $status) = @_;
312	warn "pid $pid exited with status $status";	472	warn "pid $pid exited with status $status";
313	$done->broadcast;	473	$done->send;
314	},	474	},
315	);	475	);
316		476
317	# do something else, then wait for process exit	477	# do something else, then wait for process exit
318	$done->wait;	478	$done->recv;
		479
		480	=head2 IDLE WATCHERS
		481
		482	$w = AnyEvent->idle (cb => <callback>);
		483
		484	Sometimes there is a need to do something, but it is not so important
		485	to do it instantly, but only when there is nothing better to do. This
		486	"nothing better to do" is usually defined to be "no other events need
		487	attention by the event loop".
		488
		489	Idle watchers ideally get invoked when the event loop has nothing
		490	better to do, just before it would block the process to wait for new
		491	events. Instead of blocking, the idle watcher is invoked.
		492
		493	Most event loops unfortunately do not really support idle watchers (only
		494	EV, Event and Glib do it in a usable fashion) - for the rest, AnyEvent
		495	will simply call the callback "from time to time".
		496
		497	Example: read lines from STDIN, but only process them when the
		498	program is otherwise idle:
		499
		500	my @lines; # read data
		501	my $idle_w;
		502	my $io_w = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
		503	push @lines, scalar <STDIN>;
		504
		505	# start an idle watcher, if not already done
		506	$idle_w ||= AnyEvent->idle (cb => sub {
		507	# handle only one line, when there are lines left
		508	if (my $line = shift @lines) {
		509	print "handled when idle: $line";
		510	} else {
		511	# otherwise disable the idle watcher again
		512	undef $idle_w;
		513	}
		514	});
		515	});
319		516
320	=head2 CONDITION VARIABLES	517	=head2 CONDITION VARIABLES
321		518
		519	$cv = AnyEvent->condvar;
		520
		521	$cv->send (<list>);
		522	my @res = $cv->recv;
		523
		524	If you are familiar with some event loops you will know that all of them
		525	require you to run some blocking "loop", "run" or similar function that
		526	will actively watch for new events and call your callbacks.
		527
		528	AnyEvent is slightly different: it expects somebody else to run the event
		529	loop and will only block when necessary (usually when told by the user).
		530
		531	The instrument to do that is called a "condition variable", so called
		532	because they represent a condition that must become true.
		533
		534	Now is probably a good time to look at the examples further below.
		535
322	Condition variables can be created by calling the C<< AnyEvent->condvar >>	536	Condition variables can be created by calling the C<< AnyEvent->condvar
323	method without any arguments.	537	>> method, usually without arguments. The only argument pair allowed is
		538	C<cb>, which specifies a callback to be called when the condition variable
		539	becomes true, with the condition variable as the first argument (but not
		540	the results).
324		541
325	A condition variable waits for a condition - precisely that the C<<	542	After creation, the condition variable is "false" until it becomes "true"
326	->broadcast >> method has been called.	543	by calling the C<send> method (or calling the condition variable as if it
		544	were a callback, read about the caveats in the description for the C<<
		545	->send >> method).
327		546
328	They are very useful to signal that a condition has been fulfilled, for	547	Condition variables are similar to callbacks, except that you can
		548	optionally wait for them. They can also be called merge points - points
		549	in time where multiple outstanding events have been processed. And yet
		550	another way to call them is transactions - each condition variable can be
		551	used to represent a transaction, which finishes at some point and delivers
		552	a result. And yet some people know them as "futures" - a promise to
		553	compute/deliver something that you can wait for.
		554
		555	Condition variables are very useful to signal that something has finished,
329	example, if you write a module that does asynchronous http requests,	556	for example, if you write a module that does asynchronous http requests,
330	then a condition variable would be the ideal candidate to signal the	557	then a condition variable would be the ideal candidate to signal the
331	availability of results.	558	availability of results. The user can either act when the callback is
		559	called or can synchronously C<< ->recv >> for the results.
332		560
333	You can also use condition variables to block your main program until	561	You can also use them to simulate traditional event loops - for example,
334	an event occurs - for example, you could C<< ->wait >> in your main	562	you can block your main program until an event occurs - for example, you
335	program until the user clicks the Quit button in your app, which would C<<	563	could C<< ->recv >> in your main program until the user clicks the Quit
336	->broadcast >> the "quit" event.	564	button of your app, which would C<< ->send >> the "quit" event.
337		565
338	Note that condition variables recurse into the event loop - if you have	566	Note that condition variables recurse into the event loop - if you have
339	two pirces of code that call C<< ->wait >> in a round-robbin fashion, you	567	two pieces of code that call C<< ->recv >> in a round-robin fashion, you
340	lose. Therefore, condition variables are good to export to your caller, but	568	lose. Therefore, condition variables are good to export to your caller, but
341	you should avoid making a blocking wait yourself, at least in callbacks,	569	you should avoid making a blocking wait yourself, at least in callbacks,
342	as this asks for trouble.	570	as this asks for trouble.
343		571
344	This object has two methods:	572	Condition variables are represented by hash refs in perl, and the keys
		573	used by AnyEvent itself are all named C<_ae_XXX> to make subclassing
		574	easy (it is often useful to build your own transaction class on top of
		575	AnyEvent). To subclass, use C<AnyEvent::CondVar> as base class and call
		576	it's C<new> method in your own C<new> method.
345		577
346	=over 4	578	There are two "sides" to a condition variable - the "producer side" which
		579	eventually calls C<< -> send >>, and the "consumer side", which waits
		580	for the send to occur.
347		581
348	=item $cv->wait	582	Example: wait for a timer.
349
350	Wait (blocking if necessary) until the C<< ->broadcast >> method has been
351	called on c<$cv>, while servicing other watchers normally.
352
353	You can only wait once on a condition - additional calls will return
354	immediately.
355
356	Not all event models support a blocking wait - some die in that case
357	(programs might want to do that to stay interactive), so I<if you are
358	using this from a module, never require a blocking wait>, but let the
359	caller decide whether the call will block or not (for example, by coupling
360	condition variables with some kind of request results and supporting
361	callbacks so the caller knows that getting the result will not block,
362	while still suppporting blocking waits if the caller so desires).
363
364	Another reason I<never> to C<< ->wait >> in a module is that you cannot
365	sensibly have two C<< ->wait >>'s in parallel, as that would require
366	multiple interpreters or coroutines/threads, none of which C<AnyEvent>
367	can supply (the coroutine-aware backends L<AnyEvent::Impl::CoroEV> and
368	L<AnyEvent::Impl::CoroEvent> explicitly support concurrent C<< ->wait >>'s
369	from different coroutines, however).
370
371	=item $cv->broadcast
372
373	Flag the condition as ready - a running C<< ->wait >> and all further
374	calls to C<wait> will (eventually) return after this method has been
375	called. If nobody is waiting the broadcast will be remembered..
376
377	=back
378
379	Example:
380		583
381	# wait till the result is ready	584	# wait till the result is ready
382	my $result_ready = AnyEvent->condvar;	585	my $result_ready = AnyEvent->condvar;
383		586
384	# do something such as adding a timer	587	# do something such as adding a timer
385	# or socket watcher the calls $result_ready->broadcast	588	# or socket watcher the calls $result_ready->send
386	# when the "result" is ready.	589	# when the "result" is ready.
387	# in this case, we simply use a timer:	590	# in this case, we simply use a timer:
388	my $w = AnyEvent->timer (	591	my $w = AnyEvent->timer (
389	after => 1,	592	after => 1,
390	cb => sub { $result_ready->broadcast },	593	cb => sub { $result_ready->send },
391	);	594	);
392		595
393	# this "blocks" (while handling events) till the watcher	596	# this "blocks" (while handling events) till the callback
394	# calls broadcast	597	# calls ->send
395	$result_ready->wait;	598	$result_ready->recv;
		599
		600	Example: wait for a timer, but take advantage of the fact that condition
		601	variables are also callable directly.
		602
		603	my $done = AnyEvent->condvar;
		604	my $delay = AnyEvent->timer (after => 5, cb => $done);
		605	$done->recv;
		606
		607	Example: Imagine an API that returns a condvar and doesn't support
		608	callbacks. This is how you make a synchronous call, for example from
		609	the main program:
		610
		611	use AnyEvent::CouchDB;
		612
		613	...
		614
		615	my @info = $couchdb->info->recv;
		616
		617	And this is how you would just set a callback to be called whenever the
		618	results are available:
		619
		620	$couchdb->info->cb (sub {
		621	my @info = $_[0]->recv;
		622	});
		623
		624	=head3 METHODS FOR PRODUCERS
		625
		626	These methods should only be used by the producing side, i.e. the
		627	code/module that eventually sends the signal. Note that it is also
		628	the producer side which creates the condvar in most cases, but it isn't
		629	uncommon for the consumer to create it as well.
		630
		631	=over 4
		632
		633	=item $cv->send (...)
		634
		635	Flag the condition as ready - a running C<< ->recv >> and all further
		636	calls to C<recv> will (eventually) return after this method has been
		637	called. If nobody is waiting the send will be remembered.
		638
		639	If a callback has been set on the condition variable, it is called
		640	immediately from within send.
		641
		642	Any arguments passed to the C<send> call will be returned by all
		643	future C<< ->recv >> calls.
		644
		645	Condition variables are overloaded so one can call them directly (as if
		646	they were a code reference). Calling them directly is the same as calling
		647	C<send>.
		648
		649	=item $cv->croak ($error)
		650
		651	Similar to send, but causes all call's to C<< ->recv >> to invoke
		652	C<Carp::croak> with the given error message/object/scalar.
		653
		654	This can be used to signal any errors to the condition variable
		655	user/consumer. Doing it this way instead of calling C<croak> directly
		656	delays the error detetcion, but has the overwhelmign advantage that it
		657	diagnoses the error at the place where the result is expected, and not
		658	deep in some event clalback without connection to the actual code causing
		659	the problem.
		660
		661	=item $cv->begin ([group callback])
		662
		663	=item $cv->end
		664
		665	These two methods can be used to combine many transactions/events into
		666	one. For example, a function that pings many hosts in parallel might want
		667	to use a condition variable for the whole process.
		668
		669	Every call to C<< ->begin >> will increment a counter, and every call to
		670	C<< ->end >> will decrement it. If the counter reaches C<0> in C<< ->end
		671	>>, the (last) callback passed to C<begin> will be executed, passing the
		672	condvar as first argument. That callback is I<supposed> to call C<< ->send
		673	>>, but that is not required. If no group callback was set, C<send> will
		674	be called without any arguments.
		675
		676	You can think of C<< $cv->send >> giving you an OR condition (one call
		677	sends), while C<< $cv->begin >> and C<< $cv->end >> giving you an AND
		678	condition (all C<begin> calls must be C<end>'ed before the condvar sends).
		679
		680	Let's start with a simple example: you have two I/O watchers (for example,
		681	STDOUT and STDERR for a program), and you want to wait for both streams to
		682	close before activating a condvar:
		683
		684	my $cv = AnyEvent->condvar;
		685
		686	$cv->begin; # first watcher
		687	my $w1 = AnyEvent->io (fh => $fh1, cb => sub {
		688	defined sysread $fh1, my $buf, 4096
		689	or $cv->end;
		690	});
		691
		692	$cv->begin; # second watcher
		693	my $w2 = AnyEvent->io (fh => $fh2, cb => sub {
		694	defined sysread $fh2, my $buf, 4096
		695	or $cv->end;
		696	});
		697
		698	$cv->recv;
		699
		700	This works because for every event source (EOF on file handle), there is
		701	one call to C<begin>, so the condvar waits for all calls to C<end> before
		702	sending.
		703
		704	The ping example mentioned above is slightly more complicated, as the
		705	there are results to be passwd back, and the number of tasks that are
		706	begung can potentially be zero:
		707
		708	my $cv = AnyEvent->condvar;
		709
		710	my %result;
		711	$cv->begin (sub { shift->send (\%result) });
		712
		713	for my $host (@list_of_hosts) {
		714	$cv->begin;
		715	ping_host_then_call_callback $host, sub {
		716	$result{$host} = ...;
		717	$cv->end;
		718	};
		719	}
		720
		721	$cv->end;
		722
		723	This code fragment supposedly pings a number of hosts and calls
		724	C<send> after results for all then have have been gathered - in any
		725	order. To achieve this, the code issues a call to C<begin> when it starts
		726	each ping request and calls C<end> when it has received some result for
		727	it. Since C<begin> and C<end> only maintain a counter, the order in which
		728	results arrive is not relevant.
		729
		730	There is an additional bracketing call to C<begin> and C<end> outside the
		731	loop, which serves two important purposes: first, it sets the callback
		732	to be called once the counter reaches C<0>, and second, it ensures that
		733	C<send> is called even when C<no> hosts are being pinged (the loop
		734	doesn't execute once).
		735
		736	This is the general pattern when you "fan out" into multiple (but
		737	potentially none) subrequests: use an outer C<begin>/C<end> pair to set
		738	the callback and ensure C<end> is called at least once, and then, for each
		739	subrequest you start, call C<begin> and for each subrequest you finish,
		740	call C<end>.
		741
		742	=back
		743
		744	=head3 METHODS FOR CONSUMERS
		745
		746	These methods should only be used by the consuming side, i.e. the
		747	code awaits the condition.
		748
		749	=over 4
		750
		751	=item $cv->recv
		752
		753	Wait (blocking if necessary) until the C<< ->send >> or C<< ->croak
		754	>> methods have been called on c<$cv>, while servicing other watchers
		755	normally.
		756
		757	You can only wait once on a condition - additional calls are valid but
		758	will return immediately.
		759
		760	If an error condition has been set by calling C<< ->croak >>, then this
		761	function will call C<croak>.
		762
		763	In list context, all parameters passed to C<send> will be returned,
		764	in scalar context only the first one will be returned.
		765
		766	Note that doing a blocking wait in a callback is not supported by any
		767	event loop, that is, recursive invocation of a blocking C<< ->recv
		768	>> is not allowed, and the C<recv> call will C<croak> if such a
		769	condition is detected. This condition can be slightly loosened by using
		770	L<Coro::AnyEvent>, which allows you to do a blocking C<< ->recv >> from
		771	any thread that doesn't run the event loop itself.
		772
		773	Not all event models support a blocking wait - some die in that case
		774	(programs might want to do that to stay interactive), so I<if you are
		775	using this from a module, never require a blocking wait>. Instead, let the
		776	caller decide whether the call will block or not (for example, by coupling
		777	condition variables with some kind of request results and supporting
		778	callbacks so the caller knows that getting the result will not block,
		779	while still supporting blocking waits if the caller so desires).
		780
		781	You can ensure that C<< -recv >> never blocks by setting a callback and
		782	only calling C<< ->recv >> from within that callback (or at a later
		783	time). This will work even when the event loop does not support blocking
		784	waits otherwise.
		785
		786	=item $bool = $cv->ready
		787
		788	Returns true when the condition is "true", i.e. whether C<send> or
		789	C<croak> have been called.
		790
		791	=item $cb = $cv->cb ($cb->($cv))
		792
		793	This is a mutator function that returns the callback set and optionally
		794	replaces it before doing so.
		795
		796	The callback will be called when the condition becomes (or already was)
		797	"true", i.e. when C<send> or C<croak> are called (or were called), with
		798	the only argument being the condition variable itself. Calling C<recv>
		799	inside the callback or at any later time is guaranteed not to block.
		800
		801	=back
		802
		803	=head1 SUPPORTED EVENT LOOPS/BACKENDS
		804
		805	The available backend classes are (every class has its own manpage):
		806
		807	=over 4
		808
		809	=item Backends that are autoprobed when no other event loop can be found.
		810
		811	EV is the preferred backend when no other event loop seems to be in
		812	use. If EV is not installed, then AnyEvent will fall back to its own
		813	pure-perl implementation, which is available everywhere as it comes with
		814	AnyEvent itself.
		815
		816	AnyEvent::Impl::EV based on EV (interface to libev, best choice).
		817	AnyEvent::Impl::Perl pure-perl implementation, fast and portable.
		818
		819	=item Backends that are transparently being picked up when they are used.
		820
		821	These will be used when they are currently loaded when the first watcher
		822	is created, in which case it is assumed that the application is using
		823	them. This means that AnyEvent will automatically pick the right backend
		824	when the main program loads an event module before anything starts to
		825	create watchers. Nothing special needs to be done by the main program.
		826
		827	AnyEvent::Impl::Event based on Event, very stable, few glitches.
		828	AnyEvent::Impl::Glib based on Glib, slow but very stable.
		829	AnyEvent::Impl::Tk based on Tk, very broken.
		830	AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
		831	AnyEvent::Impl::POE based on POE, very slow, some limitations.
		832	AnyEvent::Impl::Irssi used when running within irssi.
		833
		834	=item Backends with special needs.
		835
		836	Qt requires the Qt::Application to be instantiated first, but will
		837	otherwise be picked up automatically. As long as the main program
		838	instantiates the application before any AnyEvent watchers are created,
		839	everything should just work.
		840
		841	AnyEvent::Impl::Qt based on Qt.
		842
		843	Support for IO::Async can only be partial, as it is too broken and
		844	architecturally limited to even support the AnyEvent API. It also
		845	is the only event loop that needs the loop to be set explicitly, so
		846	it can only be used by a main program knowing about AnyEvent. See
		847	L<AnyEvent::Impl::Async> for the gory details.
		848
		849	AnyEvent::Impl::IOAsync based on IO::Async, cannot be autoprobed.
		850
		851	=item Event loops that are indirectly supported via other backends.
		852
		853	Some event loops can be supported via other modules:
		854
		855	There is no direct support for WxWidgets (L<Wx>) or L<Prima>.
		856
		857	B<WxWidgets> has no support for watching file handles. However, you can
		858	use WxWidgets through the POE adaptor, as POE has a Wx backend that simply
		859	polls 20 times per second, which was considered to be too horrible to even
		860	consider for AnyEvent.
		861
		862	B<Prima> is not supported as nobody seems to be using it, but it has a POE
		863	backend, so it can be supported through POE.
		864
		865	AnyEvent knows about both L<Prima> and L<Wx>, however, and will try to
		866	load L<POE> when detecting them, in the hope that POE will pick them up,
		867	in which case everything will be automatic.
		868
		869	=back
396		870
397	=head1 GLOBAL VARIABLES AND FUNCTIONS	871	=head1 GLOBAL VARIABLES AND FUNCTIONS
398		872
		873	These are not normally required to use AnyEvent, but can be useful to
		874	write AnyEvent extension modules.
		875
399	=over 4	876	=over 4
400		877
401	=item $AnyEvent::MODEL	878	=item $AnyEvent::MODEL
402		879
403	Contains C<undef> until the first watcher is being created. Then it	880	Contains C<undef> until the first watcher is being created, before the
		881	backend has been autodetected.
		882
404	contains the event model that is being used, which is the name of the	883	Afterwards it contains the event model that is being used, which is the
405	Perl class implementing the model. This class is usually one of the	884	name of the Perl class implementing the model. This class is usually one
406	C<AnyEvent::Impl:xxx> modules, but can be any other class in the case	885	of the C<AnyEvent::Impl:xxx> modules, but can be any other class in the
407	AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode>).	886	case AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode> it
408		887	will be C<urxvt::anyevent>).
409	The known classes so far are:
410
411	AnyEvent::Impl::CoroEV based on Coro::EV, best choice.
412	AnyEvent::Impl::CoroEvent based on Coro::Event, second best choice.
413	AnyEvent::Impl::EV based on EV (an interface to libev, best choice).
414	AnyEvent::Impl::Event based on Event, second best choice.
415	AnyEvent::Impl::Glib based on Glib, third-best choice.
416	AnyEvent::Impl::Perl pure-perl implementation, inefficient but portable.
417	AnyEvent::Impl::Tk based on Tk, very bad choice.
418	AnyEvent::Impl::Qt based on Qt, cannot be autoprobed (see its docs).
419	AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
420	AnyEvent::Impl::POE based on POE, not generic enough for full support.
421
422	There is no support for WxWidgets, as WxWidgets has no support for
423	watching file handles. However, you can use WxWidgets through the
424	POE Adaptor, as POE has a Wx backend that simply polls 20 times per
425	second, which was considered to be too horrible to even consider for
426	AnyEvent. Likewise, other POE backends can be used by AnyEvent by using
427	it's adaptor.
428
429	AnyEvent knows about L<Prima> and L<Wx> and will try to use L<POE> when
430	autodetecting them.
431		888
432	=item AnyEvent::detect	889	=item AnyEvent::detect
433		890
434	Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model	891	Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model
435	if necessary. You should only call this function right before you would	892	if necessary. You should only call this function right before you would
436	have created an AnyEvent watcher anyway, that is, as late as possible at	893	have created an AnyEvent watcher anyway, that is, as late as possible at
437	runtime.	894	runtime, and not e.g. while initialising of your module.
		895
		896	If you need to do some initialisation before AnyEvent watchers are
		897	created, use C<post_detect>.
		898
		899	=item $guard = AnyEvent::post_detect { BLOCK }
		900
		901	Arranges for the code block to be executed as soon as the event model is
		902	autodetected (or immediately if this has already happened).
		903
		904	The block will be executed I<after> the actual backend has been detected
		905	(C<$AnyEvent::MODEL> is set), but I<before> any watchers have been
		906	created, so it is possible to e.g. patch C<@AnyEvent::ISA> or do
		907	other initialisations - see the sources of L<AnyEvent::Strict> or
		908	L<AnyEvent::AIO> to see how this is used.
		909
		910	The most common usage is to create some global watchers, without forcing
		911	event module detection too early, for example, L<AnyEvent::AIO> creates
		912	and installs the global L<IO::AIO> watcher in a C<post_detect> block to
		913	avoid autodetecting the event module at load time.
		914
		915	If called in scalar or list context, then it creates and returns an object
		916	that automatically removes the callback again when it is destroyed (or
		917	C<undef> when the hook was immediately executed). See L<AnyEvent::AIO> for
		918	a case where this is useful.
		919
		920	Example: Create a watcher for the IO::AIO module and store it in
		921	C<$WATCHER>. Only do so after the event loop is initialised, though.
		922
		923	our WATCHER;
		924
		925	my $guard = AnyEvent::post_detect {
		926	$WATCHER = AnyEvent->io (fh => IO::AIO::poll_fileno, poll => 'r', cb => \&IO::AIO::poll_cb);
		927	};
		928
		929	# the ||= is important in case post_detect immediately runs the block,
		930	# as to not clobber the newly-created watcher. assigning both watcher and
		931	# post_detect guard to the same variable has the advantage of users being
		932	# able to just C<undef $WATCHER> if the watcher causes them grief.
		933
		934	$WATCHER ||= $guard;
		935
		936	=item @AnyEvent::post_detect
		937
		938	If there are any code references in this array (you can C<push> to it
		939	before or after loading AnyEvent), then they will called directly after
		940	the event loop has been chosen.
		941
		942	You should check C<$AnyEvent::MODEL> before adding to this array, though:
		943	if it is defined then the event loop has already been detected, and the
		944	array will be ignored.
		945
		946	Best use C<AnyEvent::post_detect { BLOCK }> when your application allows
		947	it,as it takes care of these details.
		948
		949	This variable is mainly useful for modules that can do something useful
		950	when AnyEvent is used and thus want to know when it is initialised, but do
		951	not need to even load it by default. This array provides the means to hook
		952	into AnyEvent passively, without loading it.
438		953
439	=back	954	=back
440		955
441	=head1 WHAT TO DO IN A MODULE	956	=head1 WHAT TO DO IN A MODULE
442		957
…		…
446	Be careful when you create watchers in the module body - AnyEvent will	961	Be careful when you create watchers in the module body - AnyEvent will
447	decide which event module to use as soon as the first method is called, so	962	decide which event module to use as soon as the first method is called, so
448	by calling AnyEvent in your module body you force the user of your module	963	by calling AnyEvent in your module body you force the user of your module
449	to load the event module first.	964	to load the event module first.
450		965
451	Never call C<< ->wait >> on a condition variable unless you I<know> that	966	Never call C<< ->recv >> on a condition variable unless you I<know> that
452	the C<< ->broadcast >> method has been called on it already. This is	967	the C<< ->send >> method has been called on it already. This is
453	because it will stall the whole program, and the whole point of using	968	because it will stall the whole program, and the whole point of using
454	events is to stay interactive.	969	events is to stay interactive.
455		970
456	It is fine, however, to call C<< ->wait >> when the user of your module	971	It is fine, however, to call C<< ->recv >> when the user of your module
457	requests it (i.e. if you create a http request object ad have a method	972	requests it (i.e. if you create a http request object ad have a method
458	called C<results> that returns the results, it should call C<< ->wait >>	973	called C<results> that returns the results, it should call C<< ->recv >>
459	freely, as the user of your module knows what she is doing. always).	974	freely, as the user of your module knows what she is doing. always).
460		975
461	=head1 WHAT TO DO IN THE MAIN PROGRAM	976	=head1 WHAT TO DO IN THE MAIN PROGRAM
462		977
463	There will always be a single main program - the only place that should	978	There will always be a single main program - the only place that should
…		…
465		980
466	If it doesn't care, it can just "use AnyEvent" and use it itself, or not	981	If it doesn't care, it can just "use AnyEvent" and use it itself, or not
467	do anything special (it does not need to be event-based) and let AnyEvent	982	do anything special (it does not need to be event-based) and let AnyEvent
468	decide which implementation to chose if some module relies on it.	983	decide which implementation to chose if some module relies on it.
469		984
470	If the main program relies on a specific event model. For example, in	985	If the main program relies on a specific event model - for example, in
471	Gtk2 programs you have to rely on the Glib module. You should load the	986	Gtk2 programs you have to rely on the Glib module - you should load the
472	event module before loading AnyEvent or any module that uses it: generally	987	event module before loading AnyEvent or any module that uses it: generally
473	speaking, you should load it as early as possible. The reason is that	988	speaking, you should load it as early as possible. The reason is that
474	modules might create watchers when they are loaded, and AnyEvent will	989	modules might create watchers when they are loaded, and AnyEvent will
475	decide on the event model to use as soon as it creates watchers, and it	990	decide on the event model to use as soon as it creates watchers, and it
476	might chose the wrong one unless you load the correct one yourself.	991	might chose the wrong one unless you load the correct one yourself.
477		992
478	You can chose to use a rather inefficient pure-perl implementation by	993	You can chose to use a pure-perl implementation by loading the
479	loading the C<AnyEvent::Impl::Perl> module, which gives you similar	994	C<AnyEvent::Impl::Perl> module, which gives you similar behaviour
480	behaviour everywhere, but letting AnyEvent chose is generally better.	995	everywhere, but letting AnyEvent chose the model is generally better.
		996
		997	=head2 MAINLOOP EMULATION
		998
		999	Sometimes (often for short test scripts, or even standalone programs who
		1000	only want to use AnyEvent), you do not want to run a specific event loop.
		1001
		1002	In that case, you can use a condition variable like this:
		1003
		1004	AnyEvent->condvar->recv;
		1005
		1006	This has the effect of entering the event loop and looping forever.
		1007
		1008	Note that usually your program has some exit condition, in which case
		1009	it is better to use the "traditional" approach of storing a condition
		1010	variable somewhere, waiting for it, and sending it when the program should
		1011	exit cleanly.
		1012
		1013
		1014	=head1 OTHER MODULES
		1015
		1016	The following is a non-exhaustive list of additional modules that use
		1017	AnyEvent as a client and can therefore be mixed easily with other AnyEvent
		1018	modules and other event loops in the same program. Some of the modules
		1019	come with AnyEvent, most are available via CPAN.
		1020
		1021	=over 4
		1022
		1023	=item L<AnyEvent::Util>
		1024
		1025	Contains various utility functions that replace often-used but blocking
		1026	functions such as C<inet_aton> by event-/callback-based versions.
		1027
		1028	=item L<AnyEvent::Socket>
		1029
		1030	Provides various utility functions for (internet protocol) sockets,
		1031	addresses and name resolution. Also functions to create non-blocking tcp
		1032	connections or tcp servers, with IPv6 and SRV record support and more.
		1033
		1034	=item L<AnyEvent::Handle>
		1035
		1036	Provide read and write buffers, manages watchers for reads and writes,
		1037	supports raw and formatted I/O, I/O queued and fully transparent and
		1038	non-blocking SSL/TLS (via L<AnyEvent::TLS>.
		1039
		1040	=item L<AnyEvent::DNS>
		1041
		1042	Provides rich asynchronous DNS resolver capabilities.
		1043
		1044	=item L<AnyEvent::HTTP>
		1045
		1046	A simple-to-use HTTP library that is capable of making a lot of concurrent
		1047	HTTP requests.
		1048
		1049	=item L<AnyEvent::HTTPD>
		1050
		1051	Provides a simple web application server framework.
		1052
		1053	=item L<AnyEvent::FastPing>
		1054
		1055	The fastest ping in the west.
		1056
		1057	=item L<AnyEvent::DBI>
		1058
		1059	Executes L<DBI> requests asynchronously in a proxy process.
		1060
		1061	=item L<AnyEvent::AIO>
		1062
		1063	Truly asynchronous I/O, should be in the toolbox of every event
		1064	programmer. AnyEvent::AIO transparently fuses L<IO::AIO> and AnyEvent
		1065	together.
		1066
		1067	=item L<AnyEvent::BDB>
		1068
		1069	Truly asynchronous Berkeley DB access. AnyEvent::BDB transparently fuses
		1070	L<BDB> and AnyEvent together.
		1071
		1072	=item L<AnyEvent::GPSD>
		1073
		1074	A non-blocking interface to gpsd, a daemon delivering GPS information.
		1075
		1076	=item L<AnyEvent::IRC>
		1077
		1078	AnyEvent based IRC client module family (replacing the older Net::IRC3).
		1079
		1080	=item L<AnyEvent::XMPP>
		1081
		1082	AnyEvent based XMPP (Jabber protocol) module family (replacing the older
		1083	Net::XMPP2>.
		1084
		1085	=item L<AnyEvent::IGS>
		1086
		1087	A non-blocking interface to the Internet Go Server protocol (used by
		1088	L<App::IGS>).
		1089
		1090	=item L<Net::FCP>
		1091
		1092	AnyEvent-based implementation of the Freenet Client Protocol, birthplace
		1093	of AnyEvent.
		1094
		1095	=item L<Event::ExecFlow>
		1096
		1097	High level API for event-based execution flow control.
		1098
		1099	=item L<Coro>
		1100
		1101	Has special support for AnyEvent via L<Coro::AnyEvent>.
		1102
		1103	=back
481		1104
482	=cut	1105	=cut
483		1106
484	package AnyEvent;	1107	package AnyEvent;
485		1108
486	no warnings;	1109	# basically a tuned-down version of common::sense
487	use strict;	1110	sub common_sense {
		1111	# from common:.sense 1.0
		1112	${^WARNING_BITS} = "\xfc\x3f\xf3\x00\x0f\xf3\xcf\xc0\xf3\xfc\x33\x03";
		1113	# use strict vars subs
		1114	$^H |= 0x00000600;
		1115	}
488		1116
		1117	BEGIN { AnyEvent::common_sense }
		1118
489	use Carp;	1119	use Carp ();
490		1120
491	our $VERSION = '3.3';	1121	our $VERSION = '5.112';
492	our $MODEL;	1122	our $MODEL;
493		1123
494	our $AUTOLOAD;	1124	our $AUTOLOAD;
495	our @ISA;	1125	our @ISA;
496		1126
497	our $verbose = $ENV{PERL_ANYEVENT_VERBOSE}*1;
498
499	our @REGISTRY;	1127	our @REGISTRY;
500		1128
		1129	our $WIN32;
		1130
		1131	our $VERBOSE;
		1132
		1133	BEGIN {
		1134	eval "sub WIN32(){ " . (($^O =~ /mswin32/i)*1) ." }";
		1135	eval "sub TAINT(){ " . (${^TAINT}*1) . " }";
		1136
		1137	delete @ENV{grep /^PERL_ANYEVENT_/, keys %ENV}
		1138	if ${^TAINT};
		1139
		1140	$VERBOSE = $ENV{PERL_ANYEVENT_VERBOSE}*1;
		1141
		1142	}
		1143
		1144	our $MAX_SIGNAL_LATENCY = 10;
		1145
		1146	our %PROTOCOL; # (ipv4|ipv6) => (1|2), higher numbers are preferred
		1147
		1148	{
		1149	my $idx;
		1150	$PROTOCOL{$_} = ++$idx
		1151	for reverse split /\s*,\s*/,
		1152	$ENV{PERL_ANYEVENT_PROTOCOLS} || "ipv4,ipv6";
		1153	}
		1154
501	my @models = (	1155	my @models = (
502	[Coro::EV:: => AnyEvent::Impl::CoroEV::],
503	[Coro::Event:: => AnyEvent::Impl::CoroEvent::],
504	[EV:: => AnyEvent::Impl::EV::],	1156	[EV:: => AnyEvent::Impl::EV:: , 1],
		1157	[AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl:: , 1],
		1158	# everything below here will not (normally) be autoprobed
		1159	# as the pureperl backend should work everywhere
		1160	# and is usually faster
505	[Event:: => AnyEvent::Impl::Event::],	1161	[Event:: => AnyEvent::Impl::Event::, 1],
506	[Glib:: => AnyEvent::Impl::Glib::],	1162	[Glib:: => AnyEvent::Impl::Glib:: , 1], # becomes extremely slow with many watchers
		1163	[Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy
		1164	[Irssi:: => AnyEvent::Impl::Irssi::], # Irssi has a bogus "Event" package
507	[Tk:: => AnyEvent::Impl::Tk::],	1165	[Tk:: => AnyEvent::Impl::Tk::], # crashes with many handles
		1166	[Qt:: => AnyEvent::Impl::Qt::], # requires special main program
		1167	[POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza
508	[Wx:: => AnyEvent::Impl::POE::],	1168	[Wx:: => AnyEvent::Impl::POE::],
509	[Prima:: => AnyEvent::Impl::POE::],	1169	[Prima:: => AnyEvent::Impl::POE::],
510	[AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl::],	1170	# IO::Async is just too broken - we would need workarounds for its
511	# everything below here will not be autoprobed as the pureperl backend should work everywhere	1171	# byzantine signal and broken child handling, among others.
512	[Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy	1172	# IO::Async is rather hard to detect, as it doesn't have any
		1173	# obvious default class.
513	[Qt:: => AnyEvent::Impl::Qt::], # requires special main program	1174	[IO::Async:: => AnyEvent::Impl::IOAsync::], # requires special main program
514	[POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza	1175	[IO::Async::Loop:: => AnyEvent::Impl::IOAsync::], # requires special main program
		1176	[IO::Async::Notifier:: => AnyEvent::Impl::IOAsync::], # requires special main program
		1177	[AnyEvent::Impl::IOAsync:: => AnyEvent::Impl::IOAsync::], # requires special main program
515	);	1178	);
516		1179
517	our %method = map +($_ => 1), qw(io timer signal child condvar broadcast wait one_event DESTROY);	1180	our %method = map +($_ => 1),
		1181	qw(io timer time now now_update signal child idle condvar one_event DESTROY);
		1182
		1183	our @post_detect;
		1184
		1185	sub post_detect(&) {
		1186	my ($cb) = @_;
		1187
		1188	if ($MODEL) {
		1189	$cb->();
		1190
		1191	undef
		1192	} else {
		1193	push @post_detect, $cb;
		1194
		1195	defined wantarray
		1196	? bless \$cb, "AnyEvent::Util::postdetect"
		1197	: ()
		1198	}
		1199	}
		1200
		1201	sub AnyEvent::Util::postdetect::DESTROY {
		1202	@post_detect = grep $_ != ${$_[0]}, @post_detect;
		1203	}
518		1204
519	sub detect() {	1205	sub detect() {
520	unless ($MODEL) {	1206	unless ($MODEL) {
521	no strict 'refs';	1207	local $SIG{__DIE__};
522		1208
523	if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {	1209	if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {
524	my $model = "AnyEvent::Impl::$1";	1210	my $model = "AnyEvent::Impl::$1";
525	if (eval "require $model") {	1211	if (eval "require $model") {
526	$MODEL = $model;	1212	$MODEL = $model;
527	warn "AnyEvent: loaded model '$model' (forced by \$PERL_ANYEVENT_MODEL), using it.\n" if $verbose > 1;	1213	warn "AnyEvent: loaded model '$model' (forced by \$ENV{PERL_ANYEVENT_MODEL}), using it.\n" if $VERBOSE >= 2;
528	} else {	1214	} else {
529	warn "AnyEvent: unable to load model '$model' (from \$PERL_ANYEVENT_MODEL):\n$@" if $verbose;	1215	warn "AnyEvent: unable to load model '$model' (from \$ENV{PERL_ANYEVENT_MODEL}):\n$@" if $VERBOSE;
530	}	1216	}
531	}	1217	}
532		1218
533	# check for already loaded models	1219	# check for already loaded models
534	unless ($MODEL) {	1220	unless ($MODEL) {
535	for (@REGISTRY, @models) {	1221	for (@REGISTRY, @models) {
536	my ($package, $model) = @$_;	1222	my ($package, $model) = @$_;
537	if (${"$package\::VERSION"} > 0) {	1223	if (${"$package\::VERSION"} > 0) {
538	if (eval "require $model") {	1224	if (eval "require $model") {
539	$MODEL = $model;	1225	$MODEL = $model;
540	warn "AnyEvent: autodetected model '$model', using it.\n" if $verbose > 1;	1226	warn "AnyEvent: autodetected model '$model', using it.\n" if $VERBOSE >= 2;
541	last;	1227	last;
542	}	1228	}
543	}	1229	}
544	}	1230	}
545		1231
546	unless ($MODEL) {	1232	unless ($MODEL) {
547	# try to load a model	1233	# try to autoload a model
548
549	for (@REGISTRY, @models) {	1234	for (@REGISTRY, @models) {
550	my ($package, $model) = @$_;	1235	my ($package, $model, $autoload) = @$_;
		1236	if (
		1237	$autoload
551	if (eval "require $package"	1238	and eval "require $package"
552	and ${"$package\::VERSION"} > 0	1239	and ${"$package\::VERSION"} > 0
553	and eval "require $model") {	1240	and eval "require $model"
		1241	) {
554	$MODEL = $model;	1242	$MODEL = $model;
555	warn "AnyEvent: autoprobed model '$model', using it.\n" if $verbose > 1;	1243	warn "AnyEvent: autoloaded model '$model', using it.\n" if $VERBOSE >= 2;
556	last;	1244	last;
557	}	1245	}
558	}	1246	}
559		1247
560	$MODEL	1248	$MODEL
561	or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV (or Coro+EV), Event (or Coro+Event) or Glib.";	1249	or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV, Event or Glib.\n";
562	}	1250	}
563	}	1251	}
564		1252
		1253	push @{"$MODEL\::ISA"}, "AnyEvent::Base";
		1254
565	unshift @ISA, $MODEL;	1255	unshift @ISA, $MODEL;
566	push @{"$MODEL\::ISA"}, "AnyEvent::Base";	1256
		1257	require AnyEvent::Strict if $ENV{PERL_ANYEVENT_STRICT};
		1258
		1259	(shift @post_detect)->() while @post_detect;
567	}	1260	}
568		1261
569	$MODEL	1262	$MODEL
570	}	1263	}
571		1264
572	sub AUTOLOAD {	1265	sub AUTOLOAD {
573	(my $func = $AUTOLOAD) =~ s/.*://;	1266	(my $func = $AUTOLOAD) =~ s/.*://;
574		1267
575	$method{$func}	1268	$method{$func}
576	or croak "$func: not a valid method for AnyEvent objects";	1269	or Carp::croak "$func: not a valid method for AnyEvent objects";
577		1270
578	detect unless $MODEL;	1271	detect unless $MODEL;
579		1272
580	my $class = shift;	1273	my $class = shift;
581	$class->$func (@_);	1274	$class->$func (@_);
582	}	1275	}
583		1276
		1277	# utility function to dup a filehandle. this is used by many backends
		1278	# to support binding more than one watcher per filehandle (they usually
		1279	# allow only one watcher per fd, so we dup it to get a different one).
		1280	sub _dupfh($$;$$) {
		1281	my ($poll, $fh, $r, $w) = @_;
		1282
		1283	# cygwin requires the fh mode to be matching, unix doesn't
		1284	my ($rw, $mode) = $poll eq "r" ? ($r, "<&") : ($w, ">&");
		1285
		1286	open my $fh2, $mode, $fh
		1287	or die "AnyEvent->io: cannot dup() filehandle in mode '$poll': $!,";
		1288
		1289	# we assume CLOEXEC is already set by perl in all important cases
		1290
		1291	($fh2, $rw)
		1292	}
		1293
		1294	=head1 SIMPLIFIED AE API
		1295
		1296	Starting with version 5.0, AnyEvent officially supports a second, much
		1297	simpler, API that is designed to reduce the calling, typing and memory
		1298	overhead.
		1299
		1300	See the L<AE> manpage for details.
		1301
		1302	=cut
		1303
		1304	package AE;
		1305
		1306	our $VERSION = $AnyEvent::VERSION;
		1307
		1308	sub io($$$) {
		1309	AnyEvent->io (fh => $_[0], poll => $_[1] ? "w" : "r", cb => $_[2])
		1310	}
		1311
		1312	sub timer($$$) {
		1313	AnyEvent->timer (after => $_[0], interval => $_[1], cb => $_[2])
		1314	}
		1315
		1316	sub signal($$) {
		1317	AnyEvent->signal (signal => $_[0], cb => $_[1])
		1318	}
		1319
		1320	sub child($$) {
		1321	AnyEvent->child (pid => $_[0], cb => $_[1])
		1322	}
		1323
		1324	sub idle($) {
		1325	AnyEvent->idle (cb => $_[0])
		1326	}
		1327
		1328	sub cv(;&) {
		1329	AnyEvent->condvar (@_ ? (cb => $_[0]) : ())
		1330	}
		1331
		1332	sub now() {
		1333	AnyEvent->now
		1334	}
		1335
		1336	sub now_update() {
		1337	AnyEvent->now_update
		1338	}
		1339
		1340	sub time() {
		1341	AnyEvent->time
		1342	}
		1343
584	package AnyEvent::Base;	1344	package AnyEvent::Base;
585		1345
		1346	# default implementations for many methods
		1347
		1348	sub _time() {
		1349	# probe for availability of Time::HiRes
		1350	if (eval "use Time::HiRes (); Time::HiRes::time (); 1") {
		1351	warn "AnyEvent: using Time::HiRes for sub-second timing accuracy.\n" if $VERBOSE >= 8;
		1352	*_time = \&Time::HiRes::time;
		1353	# if (eval "use POSIX (); (POSIX::times())...
		1354	} else {
		1355	warn "AnyEvent: using built-in time(), WARNING, no sub-second resolution!\n" if $VERBOSE;
		1356	*_time = sub { time }; # epic fail
		1357	}
		1358
		1359	&_time
		1360	}
		1361
		1362	sub time { _time }
		1363	sub now { _time }
		1364	sub now_update { }
		1365
586	# default implementation for ->condvar, ->wait, ->broadcast	1366	# default implementation for ->condvar
587		1367
588	sub condvar {	1368	sub condvar {
589	bless \my $flag, "AnyEvent::Base::CondVar"	1369	bless { @_ == 3 ? (_ae_cb => $_[2]) : () }, "AnyEvent::CondVar"
590	}
591
592	sub AnyEvent::Base::CondVar::broadcast {
593	${$_[0]}++;
594	}
595
596	sub AnyEvent::Base::CondVar::wait {
597	AnyEvent->one_event while !${$_[0]};
598	}	1370	}
599		1371
600	# default implementation for ->signal	1372	# default implementation for ->signal
601		1373
602	our %SIG_CB;	1374	our $HAVE_ASYNC_INTERRUPT;
		1375
		1376	sub _have_async_interrupt() {
		1377	$HAVE_ASYNC_INTERRUPT = 1*(!$ENV{PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT}
		1378	&& eval "use Async::Interrupt 1.02 (); 1")
		1379	unless defined $HAVE_ASYNC_INTERRUPT;
		1380
		1381	$HAVE_ASYNC_INTERRUPT
		1382	}
		1383
		1384	our ($SIGPIPE_R, $SIGPIPE_W, %SIG_CB, %SIG_EV, $SIG_IO);
		1385	our (%SIG_ASY, %SIG_ASY_W);
		1386	our ($SIG_COUNT, $SIG_TW);
		1387
		1388	sub _signal_exec {
		1389	$HAVE_ASYNC_INTERRUPT
		1390	? $SIGPIPE_R->drain
		1391	: sysread $SIGPIPE_R, my $dummy, 9;
		1392
		1393	while (%SIG_EV) {
		1394	for (keys %SIG_EV) {
		1395	delete $SIG_EV{$_};
		1396	$_->() for values %{ $SIG_CB{$_} || {} };
		1397	}
		1398	}
		1399	}
		1400
		1401	# install a dummy wakeup watcher to reduce signal catching latency
		1402	sub _sig_add() {
		1403	unless ($SIG_COUNT++) {
		1404	# try to align timer on a full-second boundary, if possible
		1405	my $NOW = AE::now;
		1406
		1407	$SIG_TW = AE::timer
		1408	$MAX_SIGNAL_LATENCY - ($NOW - int $NOW),
		1409	$MAX_SIGNAL_LATENCY,
		1410	sub { } # just for the PERL_ASYNC_CHECK
		1411	;
		1412	}
		1413	}
		1414
		1415	sub _sig_del {
		1416	undef $SIG_TW
		1417	unless --$SIG_COUNT;
		1418	}
		1419
		1420	our $_sig_name_init; $_sig_name_init = sub {
		1421	eval q{ # poor man's autoloading
		1422	undef $_sig_name_init;
		1423
		1424	if (_have_async_interrupt) {
		1425	*sig2num = \&Async::Interrupt::sig2num;
		1426	*sig2name = \&Async::Interrupt::sig2name;
		1427	} else {
		1428	require Config;
		1429
		1430	my %signame2num;
		1431	@signame2num{ split ' ', $Config::Config{sig_name} }
		1432	= split ' ', $Config::Config{sig_num};
		1433
		1434	my @signum2name;
		1435	@signum2name[values %signame2num] = keys %signame2num;
		1436
		1437	*sig2num = sub($) {
		1438	$_[0] > 0 ? shift : $signame2num{+shift}
		1439	};
		1440	*sig2name = sub ($) {
		1441	$_[0] > 0 ? $signum2name[+shift] : shift
		1442	};
		1443	}
		1444	};
		1445	die if $@;
		1446	};
		1447
		1448	sub sig2num ($) { &$_sig_name_init; &sig2num }
		1449	sub sig2name($) { &$_sig_name_init; &sig2name }
603		1450
604	sub signal {	1451	sub signal {
		1452	eval q{ # poor man's autoloading {}
		1453	# probe for availability of Async::Interrupt
		1454	if (_have_async_interrupt) {
		1455	warn "AnyEvent: using Async::Interrupt for race-free signal handling.\n" if $VERBOSE >= 8;
		1456
		1457	$SIGPIPE_R = new Async::Interrupt::EventPipe;
		1458	$SIG_IO = AE::io $SIGPIPE_R->fileno, 0, \&_signal_exec;
		1459
		1460	} else {
		1461	warn "AnyEvent: using emulated perl signal handling with latency timer.\n" if $VERBOSE >= 8;
		1462
		1463	require Fcntl;
		1464
		1465	if (AnyEvent::WIN32) {
		1466	require AnyEvent::Util;
		1467
		1468	($SIGPIPE_R, $SIGPIPE_W) = AnyEvent::Util::portable_pipe ();
		1469	AnyEvent::Util::fh_nonblocking ($SIGPIPE_R, 1) if $SIGPIPE_R;
		1470	AnyEvent::Util::fh_nonblocking ($SIGPIPE_W, 1) if $SIGPIPE_W; # just in case
		1471	} else {
		1472	pipe $SIGPIPE_R, $SIGPIPE_W;
		1473	fcntl $SIGPIPE_R, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_R;
		1474	fcntl $SIGPIPE_W, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_W; # just in case
		1475
		1476	# not strictly required, as $^F is normally 2, but let's make sure...
		1477	fcntl $SIGPIPE_R, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC;
		1478	fcntl $SIGPIPE_W, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC;
		1479	}
		1480
		1481	$SIGPIPE_R
		1482	or Carp::croak "AnyEvent: unable to create a signal reporting pipe: $!\n";
		1483
		1484	$SIG_IO = AE::io $SIGPIPE_R, 0, \&_signal_exec;
		1485	}
		1486
		1487	*signal = sub {
605	my (undef, %arg) = @_;	1488	my (undef, %arg) = @_;
606		1489
607	my $signal = uc $arg{signal}	1490	my $signal = uc $arg{signal}
608	or Carp::croak "required option 'signal' is missing";	1491	or Carp::croak "required option 'signal' is missing";
609		1492
		1493	if ($HAVE_ASYNC_INTERRUPT) {
		1494	# async::interrupt
		1495
		1496	$signal = sig2num $signal;
610	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};	1497	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};
		1498
		1499	$SIG_ASY{$signal} ||= new Async::Interrupt
		1500	cb => sub { undef $SIG_EV{$signal} },
		1501	signal => $signal,
		1502	pipe => [$SIGPIPE_R->filenos],
		1503	pipe_autodrain => 0,
		1504	;
		1505
		1506	} else {
		1507	# pure perl
		1508
		1509	# AE::Util has been loaded in signal
		1510	$signal = sig2name $signal;
		1511	$SIG_CB{$signal}{$arg{cb}} = $arg{cb};
		1512
611	$SIG{$signal} ||= sub {	1513	$SIG{$signal} ||= sub {
612	$_->() for values %{ $SIG_CB{$signal} || {} };	1514	local $!;
		1515	syswrite $SIGPIPE_W, "\x00", 1 unless %SIG_EV;
		1516	undef $SIG_EV{$signal};
		1517	};
		1518
		1519	# can't do signal processing without introducing races in pure perl,
		1520	# so limit the signal latency.
		1521	_sig_add;
		1522	}
		1523
		1524	bless [$signal, $arg{cb}], "AnyEvent::Base::signal"
		1525	};
		1526
		1527	*AnyEvent::Base::signal::DESTROY = sub {
		1528	my ($signal, $cb) = @{$_[0]};
		1529
		1530	_sig_del;
		1531
		1532	delete $SIG_CB{$signal}{$cb};
		1533
		1534	$HAVE_ASYNC_INTERRUPT
		1535	? delete $SIG_ASY{$signal}
		1536	: # delete doesn't work with older perls - they then
		1537	# print weird messages, or just unconditionally exit
		1538	# instead of getting the default action.
		1539	undef $SIG{$signal}
		1540	unless keys %{ $SIG_CB{$signal} };
		1541	};
613	};	1542	};
614		1543	die if $@;
615	bless [$signal, $arg{cb}], "AnyEvent::Base::Signal"	1544	&signal
616	}
617
618	sub AnyEvent::Base::Signal::DESTROY {
619	my ($signal, $cb) = @{$_[0]};
620
621	delete $SIG_CB{$signal}{$cb};
622
623	$SIG{$signal} = 'DEFAULT' unless keys %{ $SIG_CB{$signal} };
624	}	1545	}
625		1546
626	# default implementation for ->child	1547	# default implementation for ->child
627		1548
628	our %PID_CB;	1549	our %PID_CB;
629	our $CHLD_W;	1550	our $CHLD_W;
630	our $CHLD_DELAY_W;	1551	our $CHLD_DELAY_W;
631	our $PID_IDLE;
632	our $WNOHANG;	1552	our $WNOHANG;
633		1553
634	sub _child_wait {	1554	sub _emit_childstatus($$) {
635	while (0 < (my $pid = waitpid -1, $WNOHANG)) {	1555	my (undef, $rpid, $rstatus) = @_;
		1556
		1557	$_->($rpid, $rstatus)
636	$_->($pid, $?) for (values %{ $PID_CB{$pid} || {} }),	1558	for values %{ $PID_CB{$rpid} || {} },
637	(values %{ $PID_CB{0} || {} });	1559	values %{ $PID_CB{0} || {} };
638	}
639
640	undef $PID_IDLE;
641	}	1560	}
642		1561
643	sub _sigchld {	1562	sub _sigchld {
644	# make sure we deliver these changes "synchronous" with the event loop.	1563	my $pid;
645	$CHLD_DELAY_W ||= AnyEvent->timer (after => 0, cb => sub {	1564
646	undef $CHLD_DELAY_W;	1565	AnyEvent->_emit_childstatus ($pid, $?)
647	&_child_wait;	1566	while ($pid = waitpid -1, $WNOHANG) > 0;
648	});
649	}	1567	}
650		1568
651	sub child {	1569	sub child {
652	my (undef, %arg) = @_;	1570	my (undef, %arg) = @_;
653		1571
654	defined (my $pid = $arg{pid} + 0)	1572	defined (my $pid = $arg{pid} + 0)
655	or Carp::croak "required option 'pid' is missing";	1573	or Carp::croak "required option 'pid' is missing";
656		1574
657	$PID_CB{$pid}{$arg{cb}} = $arg{cb};	1575	$PID_CB{$pid}{$arg{cb}} = $arg{cb};
658		1576
659	unless ($WNOHANG) {	1577	# WNOHANG is almost cetrainly 1 everywhere
660	$WNOHANG = eval { require POSIX; &POSIX::WNOHANG } || 1;	1578	$WNOHANG ||= $^O =~ /^(?:openbsd|netbsd|linux|freebsd|cygwin|MSWin32)$/
661	}	1579	? 1
		1580	: eval { local $SIG{__DIE__}; require POSIX; &POSIX::WNOHANG } || 1;
662		1581
663	unless ($CHLD_W) {	1582	unless ($CHLD_W) {
664	$CHLD_W = AnyEvent->signal (signal => 'CHLD', cb => \&_sigchld);	1583	$CHLD_W = AE::signal CHLD => \&_sigchld;
665	# child could be a zombie already, so make at least one round	1584	# child could be a zombie already, so make at least one round
666	&_sigchld;	1585	&_sigchld;
667	}	1586	}
668		1587
669	bless [$pid, $arg{cb}], "AnyEvent::Base::Child"	1588	bless [$pid, $arg{cb}], "AnyEvent::Base::child"
670	}	1589	}
671		1590
672	sub AnyEvent::Base::Child::DESTROY {	1591	sub AnyEvent::Base::child::DESTROY {
673	my ($pid, $cb) = @{$_[0]};	1592	my ($pid, $cb) = @{$_[0]};
674		1593
675	delete $PID_CB{$pid}{$cb};	1594	delete $PID_CB{$pid}{$cb};
676	delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };	1595	delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };
677		1596
678	undef $CHLD_W unless keys %PID_CB;	1597	undef $CHLD_W unless keys %PID_CB;
679	}	1598	}
		1599
		1600	# idle emulation is done by simply using a timer, regardless
		1601	# of whether the process is idle or not, and not letting
		1602	# the callback use more than 50% of the time.
		1603	sub idle {
		1604	my (undef, %arg) = @_;
		1605
		1606	my ($cb, $w, $rcb) = $arg{cb};
		1607
		1608	$rcb = sub {
		1609	if ($cb) {
		1610	$w = _time;
		1611	&$cb;
		1612	$w = _time - $w;
		1613
		1614	# never use more then 50% of the time for the idle watcher,
		1615	# within some limits
		1616	$w = 0.0001 if $w < 0.0001;
		1617	$w = 5 if $w > 5;
		1618
		1619	$w = AE::timer $w, 0, $rcb;
		1620	} else {
		1621	# clean up...
		1622	undef $w;
		1623	undef $rcb;
		1624	}
		1625	};
		1626
		1627	$w = AE::timer 0.05, 0, $rcb;
		1628
		1629	bless \\$cb, "AnyEvent::Base::idle"
		1630	}
		1631
		1632	sub AnyEvent::Base::idle::DESTROY {
		1633	undef $${$_[0]};
		1634	}
		1635
		1636	package AnyEvent::CondVar;
		1637
		1638	our @ISA = AnyEvent::CondVar::Base::;
		1639
		1640	package AnyEvent::CondVar::Base;
		1641
		1642	#use overload
		1643	# '&{}' => sub { my $self = shift; sub { $self->send (@_) } },
		1644	# fallback => 1;
		1645
		1646	# save 300+ kilobytes by dirtily hardcoding overloading
		1647	${"AnyEvent::CondVar::Base::OVERLOAD"}{dummy}++; # Register with magic by touching.
		1648	*{'AnyEvent::CondVar::Base::()'} = sub { }; # "Make it findable via fetchmethod."
		1649	*{'AnyEvent::CondVar::Base::(&{}'} = sub { my $self = shift; sub { $self->send (@_) } }; # &{}
		1650	${'AnyEvent::CondVar::Base::()'} = 1; # fallback
		1651
		1652	our $WAITING;
		1653
		1654	sub _send {
		1655	# nop
		1656	}
		1657
		1658	sub send {
		1659	my $cv = shift;
		1660	$cv->{_ae_sent} = [@_];
		1661	(delete $cv->{_ae_cb})->($cv) if $cv->{_ae_cb};
		1662	$cv->_send;
		1663	}
		1664
		1665	sub croak {
		1666	$_[0]{_ae_croak} = $_[1];
		1667	$_[0]->send;
		1668	}
		1669
		1670	sub ready {
		1671	$_[0]{_ae_sent}
		1672	}
		1673
		1674	sub _wait {
		1675	$WAITING
		1676	and !$_[0]{_ae_sent}
		1677	and Carp::croak "AnyEvent::CondVar: recursive blocking wait detected";
		1678
		1679	local $WAITING = 1;
		1680	AnyEvent->one_event while !$_[0]{_ae_sent};
		1681	}
		1682
		1683	sub recv {
		1684	$_[0]->_wait;
		1685
		1686	Carp::croak $_[0]{_ae_croak} if $_[0]{_ae_croak};
		1687	wantarray ? @{ $_[0]{_ae_sent} } : $_[0]{_ae_sent}[0]
		1688	}
		1689
		1690	sub cb {
		1691	my $cv = shift;
		1692
		1693	@_
		1694	and $cv->{_ae_cb} = shift
		1695	and $cv->{_ae_sent}
		1696	and (delete $cv->{_ae_cb})->($cv);
		1697
		1698	$cv->{_ae_cb}
		1699	}
		1700
		1701	sub begin {
		1702	++$_[0]{_ae_counter};
		1703	$_[0]{_ae_end_cb} = $_[1] if @_ > 1;
		1704	}
		1705
		1706	sub end {
		1707	return if --$_[0]{_ae_counter};
		1708	&{ $_[0]{_ae_end_cb} || sub { $_[0]->send } };
		1709	}
		1710
		1711	# undocumented/compatibility with pre-3.4
		1712	*broadcast = \&send;
		1713	*wait = \&_wait;
		1714
		1715	=head1 ERROR AND EXCEPTION HANDLING
		1716
		1717	In general, AnyEvent does not do any error handling - it relies on the
		1718	caller to do that if required. The L<AnyEvent::Strict> module (see also
		1719	the C<PERL_ANYEVENT_STRICT> environment variable, below) provides strict
		1720	checking of all AnyEvent methods, however, which is highly useful during
		1721	development.
		1722
		1723	As for exception handling (i.e. runtime errors and exceptions thrown while
		1724	executing a callback), this is not only highly event-loop specific, but
		1725	also not in any way wrapped by this module, as this is the job of the main
		1726	program.
		1727
		1728	The pure perl event loop simply re-throws the exception (usually
		1729	within C<< condvar->recv >>), the L<Event> and L<EV> modules call C<<
		1730	$Event/EV::DIED->() >>, L<Glib> uses C<< install_exception_handler >> and
		1731	so on.
		1732
		1733	=head1 ENVIRONMENT VARIABLES
		1734
		1735	The following environment variables are used by this module or its
		1736	submodules.
		1737
		1738	Note that AnyEvent will remove I<all> environment variables starting with
		1739	C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
		1740	enabled.
		1741
		1742	=over 4
		1743
		1744	=item C<PERL_ANYEVENT_VERBOSE>
		1745
		1746	By default, AnyEvent will be completely silent except in fatal
		1747	conditions. You can set this environment variable to make AnyEvent more
		1748	talkative.
		1749
		1750	When set to C<1> or higher, causes AnyEvent to warn about unexpected
		1751	conditions, such as not being able to load the event model specified by
		1752	C<PERL_ANYEVENT_MODEL>.
		1753
		1754	When set to C<2> or higher, cause AnyEvent to report to STDERR which event
		1755	model it chooses.
		1756
		1757	When set to C<8> or higher, then AnyEvent will report extra information on
		1758	which optional modules it loads and how it implements certain features.
		1759
		1760	=item C<PERL_ANYEVENT_STRICT>
		1761
		1762	AnyEvent does not do much argument checking by default, as thorough
		1763	argument checking is very costly. Setting this variable to a true value
		1764	will cause AnyEvent to load C<AnyEvent::Strict> and then to thoroughly
		1765	check the arguments passed to most method calls. If it finds any problems,
		1766	it will croak.
		1767
		1768	In other words, enables "strict" mode.
		1769
		1770	Unlike C<use strict> (or it's modern cousin, C<< use L<common::sense>
		1771	>>, it is definitely recommended to keep it off in production. Keeping
		1772	C<PERL_ANYEVENT_STRICT=1> in your environment while developing programs
		1773	can be very useful, however.
		1774
		1775	=item C<PERL_ANYEVENT_MODEL>
		1776
		1777	This can be used to specify the event model to be used by AnyEvent, before
		1778	auto detection and -probing kicks in. It must be a string consisting
		1779	entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended
		1780	and the resulting module name is loaded and if the load was successful,
		1781	used as event model. If it fails to load AnyEvent will proceed with
		1782	auto detection and -probing.
		1783
		1784	This functionality might change in future versions.
		1785
		1786	For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you
		1787	could start your program like this:
		1788
		1789	PERL_ANYEVENT_MODEL=Perl perl ...
		1790
		1791	=item C<PERL_ANYEVENT_PROTOCOLS>
		1792
		1793	Used by both L<AnyEvent::DNS> and L<AnyEvent::Socket> to determine preferences
		1794	for IPv4 or IPv6. The default is unspecified (and might change, or be the result
		1795	of auto probing).
		1796
		1797	Must be set to a comma-separated list of protocols or address families,
		1798	current supported: C<ipv4> and C<ipv6>. Only protocols mentioned will be
		1799	used, and preference will be given to protocols mentioned earlier in the
		1800	list.
		1801
		1802	This variable can effectively be used for denial-of-service attacks
		1803	against local programs (e.g. when setuid), although the impact is likely
		1804	small, as the program has to handle conenction and other failures anyways.
		1805
		1806	Examples: C<PERL_ANYEVENT_PROTOCOLS=ipv4,ipv6> - prefer IPv4 over IPv6,
		1807	but support both and try to use both. C<PERL_ANYEVENT_PROTOCOLS=ipv4>
		1808	- only support IPv4, never try to resolve or contact IPv6
		1809	addresses. C<PERL_ANYEVENT_PROTOCOLS=ipv6,ipv4> support either IPv4 or
		1810	IPv6, but prefer IPv6 over IPv4.
		1811
		1812	=item C<PERL_ANYEVENT_EDNS0>
		1813
		1814	Used by L<AnyEvent::DNS> to decide whether to use the EDNS0 extension
		1815	for DNS. This extension is generally useful to reduce DNS traffic, but
		1816	some (broken) firewalls drop such DNS packets, which is why it is off by
		1817	default.
		1818
		1819	Setting this variable to C<1> will cause L<AnyEvent::DNS> to announce
		1820	EDNS0 in its DNS requests.
		1821
		1822	=item C<PERL_ANYEVENT_MAX_FORKS>
		1823
		1824	The maximum number of child processes that C<AnyEvent::Util::fork_call>
		1825	will create in parallel.
		1826
		1827	=item C<PERL_ANYEVENT_MAX_OUTSTANDING_DNS>
		1828
		1829	The default value for the C<max_outstanding> parameter for the default DNS
		1830	resolver - this is the maximum number of parallel DNS requests that are
		1831	sent to the DNS server.
		1832
		1833	=item C<PERL_ANYEVENT_RESOLV_CONF>
		1834
		1835	The file to use instead of F</etc/resolv.conf> (or OS-specific
		1836	configuration) in the default resolver. When set to the empty string, no
		1837	default config will be used.
		1838
		1839	=item C<PERL_ANYEVENT_CA_FILE>, C<PERL_ANYEVENT_CA_PATH>.
		1840
		1841	When neither C<ca_file> nor C<ca_path> was specified during
		1842	L<AnyEvent::TLS> context creation, and either of these environment
		1843	variables exist, they will be used to specify CA certificate locations
		1844	instead of a system-dependent default.
		1845
		1846	=item C<PERL_ANYEVENT_AVOID_GUARD> and C<PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT>
		1847
		1848	When these are set to C<1>, then the respective modules are not
		1849	loaded. Mostly good for testing AnyEvent itself.
		1850
		1851	=back
680		1852
681	=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE	1853	=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE
682		1854
683	This is an advanced topic that you do not normally need to use AnyEvent in	1855	This is an advanced topic that you do not normally need to use AnyEvent in
684	a module. This section is only of use to event loop authors who want to	1856	a module. This section is only of use to event loop authors who want to
…		…
718		1890
719	I<rxvt-unicode> also cheats a bit by not providing blocking access to	1891	I<rxvt-unicode> also cheats a bit by not providing blocking access to
720	condition variables: code blocking while waiting for a condition will	1892	condition variables: code blocking while waiting for a condition will
721	C<die>. This still works with most modules/usages, and blocking calls must	1893	C<die>. This still works with most modules/usages, and blocking calls must
722	not be done in an interactive application, so it makes sense.	1894	not be done in an interactive application, so it makes sense.
723
724	=head1 ENVIRONMENT VARIABLES
725
726	The following environment variables are used by this module:
727
728	=over 4
729
730	=item C<PERL_ANYEVENT_VERBOSE>
731
732	By default, AnyEvent will be completely silent except in fatal
733	conditions. You can set this environment variable to make AnyEvent more
734	talkative.
735
736	When set to C<1> or higher, causes AnyEvent to warn about unexpected
737	conditions, such as not being able to load the event model specified by
738	C<PERL_ANYEVENT_MODEL>.
739
740	When set to C<2> or higher, cause AnyEvent to report to STDERR which event
741	model it chooses.
742
743	=item C<PERL_ANYEVENT_MODEL>
744
745	This can be used to specify the event model to be used by AnyEvent, before
746	autodetection and -probing kicks in. It must be a string consisting
747	entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended
748	and the resulting module name is loaded and if the load was successful,
749	used as event model. If it fails to load AnyEvent will proceed with
750	autodetection and -probing.
751
752	This functionality might change in future versions.
753
754	For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you
755	could start your program like this:
756
757	PERL_ANYEVENT_MODEL=Perl perl ...
758
759	=back
760		1895
761	=head1 EXAMPLE PROGRAM	1896	=head1 EXAMPLE PROGRAM
762		1897
763	The following program uses an I/O watcher to read data from STDIN, a timer	1898	The following program uses an I/O watcher to read data from STDIN, a timer
764	to display a message once per second, and a condition variable to quit the	1899	to display a message once per second, and a condition variable to quit the
…		…
773	poll => 'r',	1908	poll => 'r',
774	cb => sub {	1909	cb => sub {
775	warn "io event <$_[0]>\n"; # will always output <r>	1910	warn "io event <$_[0]>\n"; # will always output <r>
776	chomp (my $input = <STDIN>); # read a line	1911	chomp (my $input = <STDIN>); # read a line
777	warn "read: $input\n"; # output what has been read	1912	warn "read: $input\n"; # output what has been read
778	$cv->broadcast if $input =~ /^q/i; # quit program if /^q/i	1913	$cv->send if $input =~ /^q/i; # quit program if /^q/i
779	},	1914	},
780	);	1915	);
781		1916
782	my $time_watcher; # can only be used once
783
784	sub new_timer {
785	$timer = AnyEvent->timer (after => 1, cb => sub {	1917	my $time_watcher = AnyEvent->timer (after => 1, interval => 1, cb => sub {
786	warn "timeout\n"; # print 'timeout' about every second	1918	warn "timeout\n"; # print 'timeout' at most every second
787	&new_timer; # and restart the time
788	});	1919	});
789	}
790		1920
791	new_timer; # create first timer
792
793	$cv->wait; # wait until user enters /^q/i	1921	$cv->recv; # wait until user enters /^q/i
794		1922
795	=head1 REAL-WORLD EXAMPLE	1923	=head1 REAL-WORLD EXAMPLE
796		1924
797	Consider the L<Net::FCP> module. It features (among others) the following	1925	Consider the L<Net::FCP> module. It features (among others) the following
798	API calls, which are to freenet what HTTP GET requests are to http:	1926	API calls, which are to freenet what HTTP GET requests are to http:
…		…
848	syswrite $txn->{fh}, $txn->{request}	1976	syswrite $txn->{fh}, $txn->{request}
849	or die "connection or write error";	1977	or die "connection or write error";
850	$txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });	1978	$txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });
851		1979
852	Again, C<fh_ready_r> waits till all data has arrived, and then stores the	1980	Again, C<fh_ready_r> waits till all data has arrived, and then stores the
853	result and signals any possible waiters that the request ahs finished:	1981	result and signals any possible waiters that the request has finished:
854		1982
855	sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};	1983	sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};
856		1984
857	if (end-of-file or data complete) {	1985	if (end-of-file or data complete) {
858	$txn->{result} = $txn->{buf};	1986	$txn->{result} = $txn->{buf};
859	$txn->{finished}->broadcast;	1987	$txn->{finished}->send;
860	$txb->{cb}->($txn) of $txn->{cb}; # also call callback	1988	$txb->{cb}->($txn) of $txn->{cb}; # also call callback
861	}	1989	}
862		1990
863	The C<result> method, finally, just waits for the finished signal (if the	1991	The C<result> method, finally, just waits for the finished signal (if the
864	request was already finished, it doesn't wait, of course, and returns the	1992	request was already finished, it doesn't wait, of course, and returns the
865	data:	1993	data:
866		1994
867	$txn->{finished}->wait;	1995	$txn->{finished}->recv;
868	return $txn->{result};	1996	return $txn->{result};
869		1997
870	The actual code goes further and collects all errors (C<die>s, exceptions)	1998	The actual code goes further and collects all errors (C<die>s, exceptions)
871	that occured during request processing. The C<result> method detects	1999	that occurred during request processing. The C<result> method detects
872	whether an exception as thrown (it is stored inside the $txn object)	2000	whether an exception as thrown (it is stored inside the $txn object)
873	and just throws the exception, which means connection errors and other	2001	and just throws the exception, which means connection errors and other
874	problems get reported tot he code that tries to use the result, not in a	2002	problems get reported tot he code that tries to use the result, not in a
875	random callback.	2003	random callback.
876		2004
…		…
907		2035
908	my $quit = AnyEvent->condvar;	2036	my $quit = AnyEvent->condvar;
909		2037
910	$fcp->txn_client_get ($url)->cb (sub {	2038	$fcp->txn_client_get ($url)->cb (sub {
911	...	2039	...
912	$quit->broadcast;	2040	$quit->send;
913	});	2041	});
914		2042
915	$quit->wait;	2043	$quit->recv;
916		2044
917		2045
918	=head1 BENCHMARKS	2046	=head1 BENCHMARKS
919		2047
920	To give you an idea of the performance and overheads that AnyEvent adds	2048	To give you an idea of the performance and overheads that AnyEvent adds
…		…
922	of various event loops I prepared some benchmarks.	2050	of various event loops I prepared some benchmarks.
923		2051
924	=head2 BENCHMARKING ANYEVENT OVERHEAD	2052	=head2 BENCHMARKING ANYEVENT OVERHEAD
925		2053
926	Here is a benchmark of various supported event models used natively and	2054	Here is a benchmark of various supported event models used natively and
927	through anyevent. The benchmark creates a lot of timers (with a zero	2055	through AnyEvent. The benchmark creates a lot of timers (with a zero
928	timeout) and I/O watchers (watching STDOUT, a pty, to become writable,	2056	timeout) and I/O watchers (watching STDOUT, a pty, to become writable,
929	which it is), lets them fire exactly once and destroys them again.	2057	which it is), lets them fire exactly once and destroys them again.
930		2058
931	Source code for this benchmark is found as F<eg/bench> in the AnyEvent	2059	Source code for this benchmark is found as F<eg/bench> in the AnyEvent
932	distribution.	2060	distribution. It uses the L<AE> interface, which makes a real difference
		2061	for the EV and Perl backends only.
933		2062
934	=head3 Explanation of the columns	2063	=head3 Explanation of the columns
935		2064
936	I<watcher> is the number of event watchers created/destroyed. Since	2065	I<watcher> is the number of event watchers created/destroyed. Since
937	different event models feature vastly different performances, each event	2066	different event models feature vastly different performances, each event
…		…
949	all watchers, to avoid adding memory overhead. That means closure creation	2078	all watchers, to avoid adding memory overhead. That means closure creation
950	and memory usage is not included in the figures.	2079	and memory usage is not included in the figures.
951		2080
952	I<invoke> is the time, in microseconds, used to invoke a simple	2081	I<invoke> is the time, in microseconds, used to invoke a simple
953	callback. The callback simply counts down a Perl variable and after it was	2082	callback. The callback simply counts down a Perl variable and after it was
954	invoked "watcher" times, it would C<< ->broadcast >> a condvar once to	2083	invoked "watcher" times, it would C<< ->send >> a condvar once to
955	signal the end of this phase.	2084	signal the end of this phase.
956		2085
957	I<destroy> is the time, in microseconds, that it takes to destroy a single	2086	I<destroy> is the time, in microseconds, that it takes to destroy a single
958	watcher.	2087	watcher.
959		2088
960	=head3 Results	2089	=head3 Results
961		2090
962	name watchers bytes create invoke destroy comment	2091	name watchers bytes create invoke destroy comment
963	EV/EV 400000 244 0.56 0.46 0.31 EV native interface	2092	EV/EV 100000 223 0.47 0.43 0.27 EV native interface
964	EV/Any 100000 244 2.50 0.46 0.29 EV + AnyEvent watchers	2093	EV/Any 100000 223 0.48 0.42 0.26 EV + AnyEvent watchers
965	CoroEV/Any 100000 244 2.49 0.44 0.29 coroutines + Coro::Signal	2094	Coro::EV/Any 100000 223 0.47 0.42 0.26 coroutines + Coro::Signal
966	Perl/Any 100000 513 4.92 0.87 1.12 pure perl implementation	2095	Perl/Any 100000 431 2.70 0.74 0.92 pure perl implementation
967	Event/Event 16000 516 31.88 31.30 0.85 Event native interface	2096	Event/Event 16000 516 31.16 31.84 0.82 Event native interface
968	Event/Any 16000 590 35.75 31.42 1.08 Event + AnyEvent watchers	2097	Event/Any 16000 1203 42.61 34.79 1.80 Event + AnyEvent watchers
		2098	IOAsync/Any 16000 1911 41.92 27.45 16.81 via IO::Async::Loop::IO_Poll
		2099	IOAsync/Any 16000 1726 40.69 26.37 15.25 via IO::Async::Loop::Epoll
969	Glib/Any 16000 1357 98.22 12.41 54.00 quadratic behaviour	2100	Glib/Any 16000 1118 89.00 12.57 51.17 quadratic behaviour
970	Tk/Any 2000 1860 26.97 67.98 14.00 SEGV with >> 2000 watchers	2101	Tk/Any 2000 1346 20.96 10.75 8.00 SEGV with >> 2000 watchers
971	POE/Event 2000 6644 108.64 736.02 14.73 via POE::Loop::Event	2102	POE/Any 2000 6951 108.97 795.32 14.24 via POE::Loop::Event
972	POE/Select 2000 6343 94.13 809.12 565.96 via POE::Loop::Select	2103	POE/Any 2000 6648 94.79 774.40 575.51 via POE::Loop::Select
973		2104
974	=head3 Discussion	2105	=head3 Discussion
975		2106
976	The benchmark does I<not> measure scalability of the event loop very	2107	The benchmark does I<not> measure scalability of the event loop very
977	well. For example, a select-based event loop (such as the pure perl one)	2108	well. For example, a select-based event loop (such as the pure perl one)
…		…
989	benchmark machine, handling an event takes roughly 1600 CPU cycles with	2120	benchmark machine, handling an event takes roughly 1600 CPU cycles with
990	EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU	2121	EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU
991	cycles with POE.	2122	cycles with POE.
992		2123
993	C<EV> is the sole leader regarding speed and memory use, which are both	2124	C<EV> is the sole leader regarding speed and memory use, which are both
994	maximal/minimal, respectively. Even when going through AnyEvent, it uses	2125	maximal/minimal, respectively. When using the L<AE> API there is zero
		2126	overhead (when going through the AnyEvent API create is about 5-6 times
		2127	slower, with other times being equal, so still uses far less memory than
995	far less memory than any other event loop and is still faster than Event	2128	any other event loop and is still faster than Event natively).
996	natively.
997		2129
998	The pure perl implementation is hit in a few sweet spots (both the	2130	The pure perl implementation is hit in a few sweet spots (both the
999	constant timeout and the use of a single fd hit optimisations in the perl	2131	constant timeout and the use of a single fd hit optimisations in the perl
1000	interpreter and the backend itself). Nevertheless this shows that it	2132	interpreter and the backend itself). Nevertheless this shows that it
1001	adds very little overhead in itself. Like any select-based backend its	2133	adds very little overhead in itself. Like any select-based backend its
1002	performance becomes really bad with lots of file descriptors (and few of	2134	performance becomes really bad with lots of file descriptors (and few of
1003	them active), of course, but this was not subject of this benchmark.	2135	them active), of course, but this was not subject of this benchmark.
1004		2136
1005	The C<Event> module has a relatively high setup and callback invocation	2137	The C<Event> module has a relatively high setup and callback invocation
1006	cost, but overall scores in on the third place.	2138	cost, but overall scores in on the third place.
		2139
		2140	C<IO::Async> performs admirably well, about on par with C<Event>, even
		2141	when using its pure perl backend.
1007		2142
1008	C<Glib>'s memory usage is quite a bit higher, but it features a	2143	C<Glib>'s memory usage is quite a bit higher, but it features a
1009	faster callback invocation and overall ends up in the same class as	2144	faster callback invocation and overall ends up in the same class as
1010	C<Event>. However, Glib scales extremely badly, doubling the number of	2145	C<Event>. However, Glib scales extremely badly, doubling the number of
1011	watchers increases the processing time by more than a factor of four,	2146	watchers increases the processing time by more than a factor of four,
…		…
1019	file descriptor is dup()ed for each watcher. This shows that the dup()	2154	file descriptor is dup()ed for each watcher. This shows that the dup()
1020	employed by some adaptors is not a big performance issue (it does incur a	2155	employed by some adaptors is not a big performance issue (it does incur a
1021	hidden memory cost inside the kernel which is not reflected in the figures	2156	hidden memory cost inside the kernel which is not reflected in the figures
1022	above).	2157	above).
1023		2158
1024	C<POE>, regardless of underlying event loop (whether using its pure	2159	C<POE>, regardless of underlying event loop (whether using its pure perl
1025	perl select-based backend or the Event module, the POE-EV backend	2160	select-based backend or the Event module, the POE-EV backend couldn't
1026	couldn't be tested because it wasn't working) shows abysmal performance	2161	be tested because it wasn't working) shows abysmal performance and
1027	and memory usage: Watchers use almost 30 times as much memory as	2162	memory usage with AnyEvent: Watchers use almost 30 times as much memory
1028	EV watchers, and 10 times as much memory as Event (the high memory	2163	as EV watchers, and 10 times as much memory as Event (the high memory
1029	requirements are caused by requiring a session for each watcher). Watcher	2164	requirements are caused by requiring a session for each watcher). Watcher
1030	invocation speed is almost 900 times slower than with AnyEvent's pure perl	2165	invocation speed is almost 900 times slower than with AnyEvent's pure perl
		2166	implementation.
		2167
1031	implementation. The design of the POE adaptor class in AnyEvent can not	2168	The design of the POE adaptor class in AnyEvent can not really account
1032	really account for this, as session creation overhead is small compared	2169	for the performance issues, though, as session creation overhead is
1033	to execution of the state machine, which is coded pretty optimally within	2170	small compared to execution of the state machine, which is coded pretty
1034	L<AnyEvent::Impl::POE>. POE simply seems to be abysmally slow.	2171	optimally within L<AnyEvent::Impl::POE> (and while everybody agrees that
		2172	using multiple sessions is not a good approach, especially regarding
		2173	memory usage, even the author of POE could not come up with a faster
		2174	design).
1035		2175
1036	=head3 Summary	2176	=head3 Summary
1037		2177
1038	=over 4	2178	=over 4
1039		2179
…		…
1050		2190
1051	=back	2191	=back
1052		2192
1053	=head2 BENCHMARKING THE LARGE SERVER CASE	2193	=head2 BENCHMARKING THE LARGE SERVER CASE
1054		2194
1055	This benchmark atcually benchmarks the event loop itself. It works by	2195	This benchmark actually benchmarks the event loop itself. It works by
1056	creating a number of "servers": each server consists of a socketpair, a	2196	creating a number of "servers": each server consists of a socket pair, a
1057	timeout watcher that gets reset on activity (but never fires), and an I/O	2197	timeout watcher that gets reset on activity (but never fires), and an I/O
1058	watcher waiting for input on one side of the socket. Each time the socket	2198	watcher waiting for input on one side of the socket. Each time the socket
1059	watcher reads a byte it will write that byte to a random other "server".	2199	watcher reads a byte it will write that byte to a random other "server".
1060		2200
1061	The effect is that there will be a lot of I/O watchers, only part of which	2201	The effect is that there will be a lot of I/O watchers, only part of which
1062	are active at any one point (so there is a constant number of active	2202	are active at any one point (so there is a constant number of active
1063	fds for each loop iterstaion, but which fds these are is random). The	2203	fds for each loop iteration, but which fds these are is random). The
1064	timeout is reset each time something is read because that reflects how	2204	timeout is reset each time something is read because that reflects how
1065	most timeouts work (and puts extra pressure on the event loops).	2205	most timeouts work (and puts extra pressure on the event loops).
1066		2206
1067	In this benchmark, we use 10000 socketpairs (20000 sockets), of which 100	2207	In this benchmark, we use 10000 socket pairs (20000 sockets), of which 100
1068	(1%) are active. This mirrors the activity of large servers with many	2208	(1%) are active. This mirrors the activity of large servers with many
1069	connections, most of which are idle at any one point in time.	2209	connections, most of which are idle at any one point in time.
1070		2210
1071	Source code for this benchmark is found as F<eg/bench2> in the AnyEvent	2211	Source code for this benchmark is found as F<eg/bench2> in the AnyEvent
1072	distribution.	2212	distribution. It uses the L<AE> interface, which makes a real difference
		2213	for the EV and Perl backends only.
1073		2214
1074	=head3 Explanation of the columns	2215	=head3 Explanation of the columns
1075		2216
1076	I<sockets> is the number of sockets, and twice the number of "servers" (as	2217	I<sockets> is the number of sockets, and twice the number of "servers" (as
1077	each server has a read and write socket end).	2218	each server has a read and write socket end).
1078		2219
1079	I<create> is the time it takes to create a socketpair (which is	2220	I<create> is the time it takes to create a socket pair (which is
1080	nontrivial) and two watchers: an I/O watcher and a timeout watcher.	2221	nontrivial) and two watchers: an I/O watcher and a timeout watcher.
1081		2222
1082	I<request>, the most important value, is the time it takes to handle a	2223	I<request>, the most important value, is the time it takes to handle a
1083	single "request", that is, reading the token from the pipe and forwarding	2224	single "request", that is, reading the token from the pipe and forwarding
1084	it to another server. This includes deleting the old timeout and creating	2225	it to another server. This includes deleting the old timeout and creating
1085	a new one that moves the timeout into the future.	2226	a new one that moves the timeout into the future.
1086		2227
1087	=head3 Results	2228	=head3 Results
1088		2229
1089	name sockets create request	2230	name sockets create request
1090	EV 20000 69.01 11.16	2231	EV 20000 62.66 7.99
1091	Perl 20000 75.28 112.76	2232	Perl 20000 68.32 32.64
1092	Event 20000 212.62 257.32	2233	IOAsync 20000 174.06 101.15 epoll
1093	Glib 20000 651.16 1896.30	2234	IOAsync 20000 174.67 610.84 poll
		2235	Event 20000 202.69 242.91
		2236	Glib 20000 557.01 1689.52
1094	POE 20000 349.67 12317.24 uses POE::Loop::Event	2237	POE 20000 341.54 12086.32 uses POE::Loop::Event
1095		2238
1096	=head3 Discussion	2239	=head3 Discussion
1097		2240
1098	This benchmark I<does> measure scalability and overall performance of the	2241	This benchmark I<does> measure scalability and overall performance of the
1099	particular event loop.	2242	particular event loop.
…		…
1101	EV is again fastest. Since it is using epoll on my system, the setup time	2244	EV is again fastest. Since it is using epoll on my system, the setup time
1102	is relatively high, though.	2245	is relatively high, though.
1103		2246
1104	Perl surprisingly comes second. It is much faster than the C-based event	2247	Perl surprisingly comes second. It is much faster than the C-based event
1105	loops Event and Glib.	2248	loops Event and Glib.
		2249
		2250	IO::Async performs very well when using its epoll backend, and still quite
		2251	good compared to Glib when using its pure perl backend.
1106		2252
1107	Event suffers from high setup time as well (look at its code and you will	2253	Event suffers from high setup time as well (look at its code and you will
1108	understand why). Callback invocation also has a high overhead compared to	2254	understand why). Callback invocation also has a high overhead compared to
1109	the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event	2255	the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event
1110	uses select or poll in basically all documented configurations.	2256	uses select or poll in basically all documented configurations.
…		…
1118		2264
1119	=head3 Summary	2265	=head3 Summary
1120		2266
1121	=over 4	2267	=over 4
1122		2268
1123	=item * The pure perl implementation performs extremely well, considering	2269	=item * The pure perl implementation performs extremely well.
1124	that it uses select.
1125		2270
1126	=item * Avoid Glib or POE in large projects where performance matters.	2271	=item * Avoid Glib or POE in large projects where performance matters.
1127		2272
1128	=back	2273	=back
1129		2274
…		…
1142		2287
1143	=head3 Results	2288	=head3 Results
1144		2289
1145	name sockets create request	2290	name sockets create request
1146	EV 16 20.00 6.54	2291	EV 16 20.00 6.54
		2292	Perl 16 25.75 12.62
1147	Event 16 81.27 35.86	2293	Event 16 81.27 35.86
1148	Glib 16 32.63 15.48	2294	Glib 16 32.63 15.48
1149	Perl 16 24.62 162.37
1150	POE 16 261.87 276.28 uses POE::Loop::Event	2295	POE 16 261.87 276.28 uses POE::Loop::Event
1151		2296
1152	=head3 Discussion	2297	=head3 Discussion
1153		2298
1154	The benchmark tries to test the performance of a typical small	2299	The benchmark tries to test the performance of a typical small
…		…
1158	speed most when you have lots of watchers, not when you only have a few of	2303	speed most when you have lots of watchers, not when you only have a few of
1159	them).	2304	them).
1160		2305
1161	EV is again fastest.	2306	EV is again fastest.
1162		2307
1163	The C-based event loops Event and Glib come in second this time, as the	2308	Perl again comes second. It is noticeably faster than the C-based event
1164	overhead of running an iteration is much smaller in C than in Perl (little	2309	loops Event and Glib, although the difference is too small to really
1165	code to execute in the inner loop, and perl's function calling overhead is	2310	matter.
1166	high, and updating all the data structures is costly).
1167
1168	The pure perl event loop is much slower, but still competitive.
1169		2311
1170	POE also performs much better in this case, but is is still far behind the	2312	POE also performs much better in this case, but is is still far behind the
1171	others.	2313	others.
1172		2314
1173	=head3 Summary	2315	=head3 Summary
…		…
1177	=item * C-based event loops perform very well with small number of	2319	=item * C-based event loops perform very well with small number of
1178	watchers, as the management overhead dominates.	2320	watchers, as the management overhead dominates.
1179		2321
1180	=back	2322	=back
1181		2323
		2324	=head2 THE IO::Lambda BENCHMARK
		2325
		2326	Recently I was told about the benchmark in the IO::Lambda manpage, which
		2327	could be misinterpreted to make AnyEvent look bad. In fact, the benchmark
		2328	simply compares IO::Lambda with POE, and IO::Lambda looks better (which
		2329	shouldn't come as a surprise to anybody). As such, the benchmark is
		2330	fine, and mostly shows that the AnyEvent backend from IO::Lambda isn't
		2331	very optimal. But how would AnyEvent compare when used without the extra
		2332	baggage? To explore this, I wrote the equivalent benchmark for AnyEvent.
		2333
		2334	The benchmark itself creates an echo-server, and then, for 500 times,
		2335	connects to the echo server, sends a line, waits for the reply, and then
		2336	creates the next connection. This is a rather bad benchmark, as it doesn't
		2337	test the efficiency of the framework or much non-blocking I/O, but it is a
		2338	benchmark nevertheless.
		2339
		2340	name runtime
		2341	Lambda/select 0.330 sec
		2342	+ optimized 0.122 sec
		2343	Lambda/AnyEvent 0.327 sec
		2344	+ optimized 0.138 sec
		2345	Raw sockets/select 0.077 sec
		2346	POE/select, components 0.662 sec
		2347	POE/select, raw sockets 0.226 sec
		2348	POE/select, optimized 0.404 sec
		2349
		2350	AnyEvent/select/nb 0.085 sec
		2351	AnyEvent/EV/nb 0.068 sec
		2352	+state machine 0.134 sec
		2353
		2354	The benchmark is also a bit unfair (my fault): the IO::Lambda/POE
		2355	benchmarks actually make blocking connects and use 100% blocking I/O,
		2356	defeating the purpose of an event-based solution. All of the newly
		2357	written AnyEvent benchmarks use 100% non-blocking connects (using
		2358	AnyEvent::Socket::tcp_connect and the asynchronous pure perl DNS
		2359	resolver), so AnyEvent is at a disadvantage here, as non-blocking connects
		2360	generally require a lot more bookkeeping and event handling than blocking
		2361	connects (which involve a single syscall only).
		2362
		2363	The last AnyEvent benchmark additionally uses L<AnyEvent::Handle>, which
		2364	offers similar expressive power as POE and IO::Lambda, using conventional
		2365	Perl syntax. This means that both the echo server and the client are 100%
		2366	non-blocking, further placing it at a disadvantage.
		2367
		2368	As you can see, the AnyEvent + EV combination even beats the
		2369	hand-optimised "raw sockets benchmark", while AnyEvent + its pure perl
		2370	backend easily beats IO::Lambda and POE.
		2371
		2372	And even the 100% non-blocking version written using the high-level (and
		2373	slow :) L<AnyEvent::Handle> abstraction beats both POE and IO::Lambda
		2374	higher level ("unoptimised") abstractions by a large margin, even though
		2375	it does all of DNS, tcp-connect and socket I/O in a non-blocking way.
		2376
		2377	The two AnyEvent benchmarks programs can be found as F<eg/ae0.pl> and
		2378	F<eg/ae2.pl> in the AnyEvent distribution, the remaining benchmarks are
		2379	part of the IO::Lambda distribution and were used without any changes.
		2380
		2381
		2382	=head1 SIGNALS
		2383
		2384	AnyEvent currently installs handlers for these signals:
		2385
		2386	=over 4
		2387
		2388	=item SIGCHLD
		2389
		2390	A handler for C<SIGCHLD> is installed by AnyEvent's child watcher
		2391	emulation for event loops that do not support them natively. Also, some
		2392	event loops install a similar handler.
		2393
		2394	Additionally, when AnyEvent is loaded and SIGCHLD is set to IGNORE, then
		2395	AnyEvent will reset it to default, to avoid losing child exit statuses.
		2396
		2397	=item SIGPIPE
		2398
		2399	A no-op handler is installed for C<SIGPIPE> when C<$SIG{PIPE}> is C<undef>
		2400	when AnyEvent gets loaded.
		2401
		2402	The rationale for this is that AnyEvent users usually do not really depend
		2403	on SIGPIPE delivery (which is purely an optimisation for shell use, or
		2404	badly-written programs), but C<SIGPIPE> can cause spurious and rare
		2405	program exits as a lot of people do not expect C<SIGPIPE> when writing to
		2406	some random socket.
		2407
		2408	The rationale for installing a no-op handler as opposed to ignoring it is
		2409	that this way, the handler will be restored to defaults on exec.
		2410
		2411	Feel free to install your own handler, or reset it to defaults.
		2412
		2413	=back
		2414
		2415	=cut
		2416
		2417	undef $SIG{CHLD}
		2418	if $SIG{CHLD} eq 'IGNORE';
		2419
		2420	$SIG{PIPE} = sub { }
		2421	unless defined $SIG{PIPE};
		2422
		2423	=head1 RECOMMENDED/OPTIONAL MODULES
		2424
		2425	One of AnyEvent's main goals is to be 100% Pure-Perl(tm): only perl (and
		2426	it's built-in modules) are required to use it.
		2427
		2428	That does not mean that AnyEvent won't take advantage of some additional
		2429	modules if they are installed.
		2430
		2431	This section epxlains which additional modules will be used, and how they
		2432	affect AnyEvent's operetion.
		2433
		2434	=over 4
		2435
		2436	=item L<Async::Interrupt>
		2437
		2438	This slightly arcane module is used to implement fast signal handling: To
		2439	my knowledge, there is no way to do completely race-free and quick
		2440	signal handling in pure perl. To ensure that signals still get
		2441	delivered, AnyEvent will start an interval timer to wake up perl (and
		2442	catch the signals) with some delay (default is 10 seconds, look for
		2443	C<$AnyEvent::MAX_SIGNAL_LATENCY>).
		2444
		2445	If this module is available, then it will be used to implement signal
		2446	catching, which means that signals will not be delayed, and the event loop
		2447	will not be interrupted regularly, which is more efficient (And good for
		2448	battery life on laptops).
		2449
		2450	This affects not just the pure-perl event loop, but also other event loops
		2451	that have no signal handling on their own (e.g. Glib, Tk, Qt).
		2452
		2453	Some event loops (POE, Event, Event::Lib) offer signal watchers natively,
		2454	and either employ their own workarounds (POE) or use AnyEvent's workaround
		2455	(using C<$AnyEvent::MAX_SIGNAL_LATENCY>). Installing L<Async::Interrupt>
		2456	does nothing for those backends.
		2457
		2458	=item L<EV>
		2459
		2460	This module isn't really "optional", as it is simply one of the backend
		2461	event loops that AnyEvent can use. However, it is simply the best event
		2462	loop available in terms of features, speed and stability: It supports
		2463	the AnyEvent API optimally, implements all the watcher types in XS, does
		2464	automatic timer adjustments even when no monotonic clock is available,
		2465	can take avdantage of advanced kernel interfaces such as C<epoll> and
		2466	C<kqueue>, and is the fastest backend I<by far>. You can even embed
		2467	L<Glib>/L<Gtk2> in it (or vice versa, see L<EV::Glib> and L<Glib::EV>).
		2468
		2469	=item L<Guard>
		2470
		2471	The guard module, when used, will be used to implement
		2472	C<AnyEvent::Util::guard>. This speeds up guards considerably (and uses a
		2473	lot less memory), but otherwise doesn't affect guard operation much. It is
		2474	purely used for performance.
		2475
		2476	=item L<JSON> and L<JSON::XS>
		2477
		2478	This module is required when you want to read or write JSON data via
		2479	L<AnyEvent::Handle>. It is also written in pure-perl, but can take
		2480	advantage of the ultra-high-speed L<JSON::XS> module when it is installed.
		2481
		2482	In fact, L<AnyEvent::Handle> will use L<JSON::XS> by default if it is
		2483	installed.
		2484
		2485	=item L<Net::SSLeay>
		2486
		2487	Implementing TLS/SSL in Perl is certainly interesting, but not very
		2488	worthwhile: If this module is installed, then L<AnyEvent::Handle> (with
		2489	the help of L<AnyEvent::TLS>), gains the ability to do TLS/SSL.
		2490
		2491	=item L<Time::HiRes>
		2492
		2493	This module is part of perl since release 5.008. It will be used when the
		2494	chosen event library does not come with a timing source on it's own. The
		2495	pure-perl event loop (L<AnyEvent::Impl::Perl>) will additionally use it to
		2496	try to use a monotonic clock for timing stability.
		2497
		2498	=back
		2499
1182		2500
1183	=head1 FORK	2501	=head1 FORK
1184		2502
1185	Most event libraries are not fork-safe. The ones who are usually are	2503	Most event libraries are not fork-safe. The ones who are usually are
1186	because they are so inefficient. Only L<EV> is fully fork-aware.	2504	because they rely on inefficient but fork-safe C<select> or C<poll>
		2505	calls. Only L<EV> is fully fork-aware.
1187		2506
1188	If you have to fork, you must either do so I<before> creating your first	2507	If you have to fork, you must either do so I<before> creating your first
1189	watcher OR you must not use AnyEvent at all in the child.	2508	watcher OR you must not use AnyEvent at all in the child OR you must do
		2509	something completely out of the scope of AnyEvent.
1190		2510
1191		2511
1192	=head1 SECURITY CONSIDERATIONS	2512	=head1 SECURITY CONSIDERATIONS
1193		2513
1194	AnyEvent can be forced to load any event model via	2514	AnyEvent can be forced to load any event model via
…		…
1199	specified in the variable.	2519	specified in the variable.
1200		2520
1201	You can make AnyEvent completely ignore this variable by deleting it	2521	You can make AnyEvent completely ignore this variable by deleting it
1202	before the first watcher gets created, e.g. with a C<BEGIN> block:	2522	before the first watcher gets created, e.g. with a C<BEGIN> block:
1203		2523
1204	BEGIN { delete $ENV{PERL_ANYEVENT_MODEL} }	2524	BEGIN { delete $ENV{PERL_ANYEVENT_MODEL} }
1205		2525
1206	use AnyEvent;	2526	use AnyEvent;
		2527
		2528	Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can
		2529	be used to probe what backend is used and gain other information (which is
		2530	probably even less useful to an attacker than PERL_ANYEVENT_MODEL), and
		2531	$ENV{PERL_ANYEVENT_STRICT}.
		2532
		2533	Note that AnyEvent will remove I<all> environment variables starting with
		2534	C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
		2535	enabled.
		2536
		2537
		2538	=head1 BUGS
		2539
		2540	Perl 5.8 has numerous memleaks that sometimes hit this module and are hard
		2541	to work around. If you suffer from memleaks, first upgrade to Perl 5.10
		2542	and check wether the leaks still show up. (Perl 5.10.0 has other annoying
		2543	memleaks, such as leaking on C<map> and C<grep> but it is usually not as
		2544	pronounced).
1207		2545
1208		2546
1209	=head1 SEE ALSO	2547	=head1 SEE ALSO
1210		2548
1211	Event modules: L<Coro::EV>, L<EV>, L<EV::Glib>, L<Glib::EV>,	2549	Utility functions: L<AnyEvent::Util>.
1212	L<Coro::Event>, L<Event>, L<Glib::Event>, L<Glib>, L<Coro>, L<Tk>,	2550
		2551	Event modules: L<EV>, L<EV::Glib>, L<Glib::EV>, L<Event>, L<Glib::Event>,
1213	L<Event::Lib>, L<Qt>, L<POE>.	2552	L<Glib>, L<Tk>, L<Event::Lib>, L<Qt>, L<POE>.
1214		2553
1215	Implementations: L<AnyEvent::Impl::CoroEV>, L<AnyEvent::Impl::EV>,	2554	Implementations: L<AnyEvent::Impl::EV>, L<AnyEvent::Impl::Event>,
		2555	L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>,
		2556	L<AnyEvent::Impl::EventLib>, L<AnyEvent::Impl::Qt>,
1216	L<AnyEvent::Impl::CoroEvent>, L<AnyEvent::Impl::Event>, L<AnyEvent::Impl::Glib>,	2557	L<AnyEvent::Impl::POE>, L<AnyEvent::Impl::IOAsync>, L<Anyevent::Impl::Irssi>.
1217	L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>, L<AnyEvent::Impl::EventLib>,
1218	L<AnyEvent::Impl::Qt>, L<AnyEvent::Impl::POE>.
1219		2558
		2559	Non-blocking file handles, sockets, TCP clients and
		2560	servers: L<AnyEvent::Handle>, L<AnyEvent::Socket>, L<AnyEvent::TLS>.
		2561
		2562	Asynchronous DNS: L<AnyEvent::DNS>.
		2563
		2564	Coroutine support: L<Coro>, L<Coro::AnyEvent>, L<Coro::EV>,
		2565	L<Coro::Event>,
		2566
1220	Nontrivial usage examples: L<Net::FCP>, L<Net::XMPP2>.	2567	Nontrivial usage examples: L<AnyEvent::GPSD>, L<AnyEvent::XMPP>,
		2568	L<AnyEvent::HTTP>.
1221		2569
1222		2570
1223	=head1 AUTHOR	2571	=head1 AUTHOR
1224		2572
1225	Marc Lehmann <schmorp@schmorp.de>	2573	Marc Lehmann <schmorp@schmorp.de>
1226	http://home.schmorp.de/	2574	http://home.schmorp.de/
1227		2575
1228	=cut	2576	=cut
1229		2577
1230	1	2578	1
1231		2579

Diff Legend

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