ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent/lib/AnyEvent.pm
(Generate patch)

Comparing AnyEvent/lib/AnyEvent.pm (file contents):
Revision 1.215 by root, Tue Jun 23 12:19:33 2009 UTC vs.
Revision 1.367 by root, Wed Aug 17 02:14:17 2011 UTC

1=head1 NAME 1=head1 NAME
2 2
3AnyEvent - provide framework for multiple event loops 3AnyEvent - the DBI of event loop programming
4 4
5EV, Event, Glib, Tk, Perl, Event::Lib, Qt and POE are various supported 5EV, Event, Glib, Tk, Perl, Event::Lib, Irssi, rxvt-unicode, IO::Async, Qt
6event loops. 6and POE are various supported event loops/environments.
7 7
8=head1 SYNOPSIS 8=head1 SYNOPSIS
9 9
10 use AnyEvent; 10 use AnyEvent;
11 11
12 # if you prefer function calls, look at the AE manpage for
13 # an alternative API.
14
12 # file descriptor readable 15 # file handle or descriptor readable
13 my $w = AnyEvent->io (fh => $fh, poll => "r", cb => sub { ... }); 16 my $w = AnyEvent->io (fh => $fh, poll => "r", cb => sub { ... });
14 17
15 # one-shot or repeating timers 18 # one-shot or repeating timers
16 my $w = AnyEvent->timer (after => $seconds, cb => sub { ... }); 19 my $w = AnyEvent->timer (after => $seconds, cb => sub { ... });
17 my $w = AnyEvent->timer (after => $seconds, interval => $seconds, cb => ... 20 my $w = AnyEvent->timer (after => $seconds, interval => $seconds, cb => ...);
18 21
19 print AnyEvent->now; # prints current event loop time 22 print AnyEvent->now; # prints current event loop time
20 print AnyEvent->time; # think Time::HiRes::time or simply CORE::time. 23 print AnyEvent->time; # think Time::HiRes::time or simply CORE::time.
21 24
22 # POSIX signal 25 # POSIX signal
40=head1 INTRODUCTION/TUTORIAL 43=head1 INTRODUCTION/TUTORIAL
41 44
42This manpage is mainly a reference manual. If you are interested 45This manpage is mainly a reference manual. If you are interested
43in a tutorial or some gentle introduction, have a look at the 46in a tutorial or some gentle introduction, have a look at the
44L<AnyEvent::Intro> manpage. 47L<AnyEvent::Intro> manpage.
48
49=head1 SUPPORT
50
51An FAQ document is available as L<AnyEvent::FAQ>.
52
53There also is a mailinglist for discussing all things AnyEvent, and an IRC
54channel, too.
55
56See the AnyEvent project page at the B<Schmorpforge Ta-Sa Software
57Repository>, at L<http://anyevent.schmorp.de>, for more info.
45 58
46=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT) 59=head1 WHY YOU SHOULD USE THIS MODULE (OR NOT)
47 60
48Glib, POE, IO::Async, Event... CPAN offers event models by the dozen 61Glib, POE, IO::Async, Event... CPAN offers event models by the dozen
49nowadays. So what is different about AnyEvent? 62nowadays. So what is different about AnyEvent?
65module users into the same thing by forcing them to use the same event 78module users into the same thing by forcing them to use the same event
66model you use. 79model you use.
67 80
68For modules like POE or IO::Async (which is a total misnomer as it is 81For modules like POE or IO::Async (which is a total misnomer as it is
69actually doing all I/O I<synchronously>...), using them in your module is 82actually doing all I/O I<synchronously>...), using them in your module is
70like joining a cult: After you joined, you are dependent on them and you 83like joining a cult: After you join, you are dependent on them and you
71cannot use anything else, as they are simply incompatible to everything 84cannot use anything else, as they are simply incompatible to everything
72that isn't them. What's worse, all the potential users of your 85that isn't them. What's worse, all the potential users of your
73module are I<also> forced to use the same event loop you use. 86module are I<also> forced to use the same event loop you use.
74 87
75AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works 88AnyEvent is different: AnyEvent + POE works fine. AnyEvent + Glib works
76fine. AnyEvent + Tk works fine etc. etc. but none of these work together 89fine. AnyEvent + Tk works fine etc. etc. but none of these work together
77with the rest: POE + IO::Async? No go. Tk + Event? No go. Again: if 90with the rest: POE + EV? No go. Tk + Event? No go. Again: if your module
78your module uses one of those, every user of your module has to use it, 91uses one of those, every user of your module has to use it, too. But if
79too. But if your module uses AnyEvent, it works transparently with all 92your module uses AnyEvent, it works transparently with all event models it
80event models it supports (including stuff like IO::Async, as long as those 93supports (including stuff like IO::Async, as long as those use one of the
81use one of the supported event loops. It is trivial to add new event loops 94supported event loops. It is easy to add new event loops to AnyEvent, too,
82to AnyEvent, too, so it is future-proof). 95so it is future-proof).
83 96
84In addition to being free of having to use I<the one and only true event 97In addition to being free of having to use I<the one and only true event
85model>, AnyEvent also is free of bloat and policy: with POE or similar 98model>, AnyEvent also is free of bloat and policy: with POE or similar
86modules, you get an enormous amount of code and strict rules you have to 99modules, you get an enormous amount of code and strict rules you have to
87follow. AnyEvent, on the other hand, is lean and up to the point, by only 100follow. AnyEvent, on the other hand, is lean and to the point, by only
88offering the functionality that is necessary, in as thin as a wrapper as 101offering the functionality that is necessary, in as thin as a wrapper as
89technically possible. 102technically possible.
90 103
91Of course, AnyEvent comes with a big (and fully optional!) toolbox 104Of course, AnyEvent comes with a big (and fully optional!) toolbox
92of useful functionality, such as an asynchronous DNS resolver, 100% 105of useful functionality, such as an asynchronous DNS resolver, 100%
98useful) and you want to force your users to use the one and only event 111useful) and you want to force your users to use the one and only event
99model, you should I<not> use this module. 112model, you should I<not> use this module.
100 113
101=head1 DESCRIPTION 114=head1 DESCRIPTION
102 115
103L<AnyEvent> provides an identical interface to multiple event loops. This 116L<AnyEvent> provides a uniform interface to various event loops. This
104allows module authors to utilise an event loop without forcing module 117allows module authors to use event loop functionality without forcing
105users to use the same event loop (as only a single event loop can coexist 118module users to use a specific event loop implementation (since more
106peacefully at any one time). 119than one event loop cannot coexist peacefully).
107 120
108The interface itself is vaguely similar, but not identical to the L<Event> 121The interface itself is vaguely similar, but not identical to the L<Event>
109module. 122module.
110 123
111During the first call of any watcher-creation method, the module tries 124During the first call of any watcher-creation method, the module tries
112to detect the currently loaded event loop by probing whether one of the 125to detect the currently loaded event loop by probing whether one of the
113following modules is already loaded: L<EV>, 126following modules is already loaded: L<EV>, L<AnyEvent::Loop>,
114L<Event>, L<Glib>, L<AnyEvent::Impl::Perl>, L<Tk>, L<Event::Lib>, L<Qt>, 127L<Event>, L<Glib>, L<Tk>, L<Event::Lib>, L<Qt>, L<POE>. The first one
115L<POE>. The first one found is used. If none are found, the module tries 128found is used. If none are detected, the module tries to load the first
116to load these modules (excluding Tk, Event::Lib, Qt and POE as the pure perl 129four modules in the order given; but note that if L<EV> is not
117adaptor should always succeed) in the order given. The first one that can 130available, the pure-perl L<AnyEvent::Loop> should always work, so
118be successfully loaded will be used. If, after this, still none could be 131the other two are not normally tried.
119found, AnyEvent will fall back to a pure-perl event loop, which is not
120very efficient, but should work everywhere.
121 132
122Because AnyEvent first checks for modules that are already loaded, loading 133Because AnyEvent first checks for modules that are already loaded, loading
123an event model explicitly before first using AnyEvent will likely make 134an event model explicitly before first using AnyEvent will likely make
124that model the default. For example: 135that model the default. For example:
125 136
127 use AnyEvent; 138 use AnyEvent;
128 139
129 # .. AnyEvent will likely default to Tk 140 # .. AnyEvent will likely default to Tk
130 141
131The I<likely> means that, if any module loads another event model and 142The I<likely> means that, if any module loads another event model and
132starts using it, all bets are off. Maybe you should tell their authors to 143starts using it, all bets are off - this case should be very rare though,
133use AnyEvent so their modules work together with others seamlessly... 144as very few modules hardcode event loops without announcing this very
145loudly.
134 146
135The pure-perl implementation of AnyEvent is called 147The pure-perl implementation of AnyEvent is called C<AnyEvent::Loop>. Like
136C<AnyEvent::Impl::Perl>. Like other event modules you can load it 148other event modules you can load it explicitly and enjoy the high
137explicitly and enjoy the high availability of that event loop :) 149availability of that event loop :)
138 150
139=head1 WATCHERS 151=head1 WATCHERS
140 152
141AnyEvent has the central concept of a I<watcher>, which is an object that 153AnyEvent has the central concept of a I<watcher>, which is an object that
142stores relevant data for each kind of event you are waiting for, such as 154stores relevant data for each kind of event you are waiting for, such as
147callback when the event occurs (of course, only when the event model 159callback when the event occurs (of course, only when the event model
148is in control). 160is in control).
149 161
150Note that B<callbacks must not permanently change global variables> 162Note that B<callbacks must not permanently change global variables>
151potentially in use by the event loop (such as C<$_> or C<$[>) and that B<< 163potentially in use by the event loop (such as C<$_> or C<$[>) and that B<<
152callbacks must not C<die> >>. The former is good programming practise in 164callbacks must not C<die> >>. The former is good programming practice in
153Perl and the latter stems from the fact that exception handling differs 165Perl and the latter stems from the fact that exception handling differs
154widely between event loops. 166widely between event loops.
155 167
156To disable the watcher you have to destroy it (e.g. by setting the 168To disable a watcher you have to destroy it (e.g. by setting the
157variable you store it in to C<undef> or otherwise deleting all references 169variable you store it in to C<undef> or otherwise deleting all references
158to it). 170to it).
159 171
160All watchers are created by calling a method on the C<AnyEvent> class. 172All watchers are created by calling a method on the C<AnyEvent> class.
161 173
162Many watchers either are used with "recursion" (repeating timers for 174Many watchers either are used with "recursion" (repeating timers for
163example), or need to refer to their watcher object in other ways. 175example), or need to refer to their watcher object in other ways.
164 176
165An any way to achieve that is this pattern: 177One way to achieve that is this pattern:
166 178
167 my $w; $w = AnyEvent->type (arg => value ..., cb => sub { 179 my $w; $w = AnyEvent->type (arg => value ..., cb => sub {
168 # you can use $w here, for example to undef it 180 # you can use $w here, for example to undef it
169 undef $w; 181 undef $w;
170 }); 182 });
173my variables are only visible after the statement in which they are 185my variables are only visible after the statement in which they are
174declared. 186declared.
175 187
176=head2 I/O WATCHERS 188=head2 I/O WATCHERS
177 189
190 $w = AnyEvent->io (
191 fh => <filehandle_or_fileno>,
192 poll => <"r" or "w">,
193 cb => <callback>,
194 );
195
178You can create an I/O watcher by calling the C<< AnyEvent->io >> method 196You can create an I/O watcher by calling the C<< AnyEvent->io >> method
179with the following mandatory key-value pairs as arguments: 197with the following mandatory key-value pairs as arguments:
180 198
181C<fh> is the Perl I<file handle> (I<not> file descriptor) to watch 199C<fh> is the Perl I<file handle> (or a naked file descriptor) to watch
182for events (AnyEvent might or might not keep a reference to this file 200for events (AnyEvent might or might not keep a reference to this file
183handle). Note that only file handles pointing to things for which 201handle). Note that only file handles pointing to things for which
184non-blocking operation makes sense are allowed. This includes sockets, 202non-blocking operation makes sense are allowed. This includes sockets,
185most character devices, pipes, fifos and so on, but not for example files 203most character devices, pipes, fifos and so on, but not for example files
186or block devices. 204or block devices.
196 214
197The I/O watcher might use the underlying file descriptor or a copy of it. 215The I/O watcher might use the underlying file descriptor or a copy of it.
198You must not close a file handle as long as any watcher is active on the 216You must not close a file handle as long as any watcher is active on the
199underlying file descriptor. 217underlying file descriptor.
200 218
201Some event loops issue spurious readyness notifications, so you should 219Some event loops issue spurious readiness notifications, so you should
202always use non-blocking calls when reading/writing from/to your file 220always use non-blocking calls when reading/writing from/to your file
203handles. 221handles.
204 222
205Example: wait for readability of STDIN, then read a line and disable the 223Example: wait for readability of STDIN, then read a line and disable the
206watcher. 224watcher.
211 undef $w; 229 undef $w;
212 }); 230 });
213 231
214=head2 TIME WATCHERS 232=head2 TIME WATCHERS
215 233
234 $w = AnyEvent->timer (after => <seconds>, cb => <callback>);
235
236 $w = AnyEvent->timer (
237 after => <fractional_seconds>,
238 interval => <fractional_seconds>,
239 cb => <callback>,
240 );
241
216You can create a time watcher by calling the C<< AnyEvent->timer >> 242You can create a time watcher by calling the C<< AnyEvent->timer >>
217method with the following mandatory arguments: 243method with the following mandatory arguments:
218 244
219C<after> specifies after how many seconds (fractional values are 245C<after> specifies after how many seconds (fractional values are
220supported) the callback should be invoked. C<cb> is the callback to invoke 246supported) the callback should be invoked. C<cb> is the callback to invoke
222 248
223Although the callback might get passed parameters, their value and 249Although the callback might get passed parameters, their value and
224presence is undefined and you cannot rely on them. Portable AnyEvent 250presence is undefined and you cannot rely on them. Portable AnyEvent
225callbacks cannot use arguments passed to time watcher callbacks. 251callbacks cannot use arguments passed to time watcher callbacks.
226 252
227The callback will normally be invoked once only. If you specify another 253The callback will normally be invoked only once. If you specify another
228parameter, C<interval>, as a strictly positive number (> 0), then the 254parameter, C<interval>, as a strictly positive number (> 0), then the
229callback will be invoked regularly at that interval (in fractional 255callback will be invoked regularly at that interval (in fractional
230seconds) after the first invocation. If C<interval> is specified with a 256seconds) after the first invocation. If C<interval> is specified with a
231false value, then it is treated as if it were missing. 257false value, then it is treated as if it were not specified at all.
232 258
233The callback will be rescheduled before invoking the callback, but no 259The callback will be rescheduled before invoking the callback, but no
234attempt is done to avoid timer drift in most backends, so the interval is 260attempt is made to avoid timer drift in most backends, so the interval is
235only approximate. 261only approximate.
236 262
237Example: fire an event after 7.7 seconds. 263Example: fire an event after 7.7 seconds.
238 264
239 my $w = AnyEvent->timer (after => 7.7, cb => sub { 265 my $w = AnyEvent->timer (after => 7.7, cb => sub {
257 283
258While most event loops expect timers to specified in a relative way, they 284While most event loops expect timers to specified in a relative way, they
259use absolute time internally. This makes a difference when your clock 285use absolute time internally. This makes a difference when your clock
260"jumps", for example, when ntp decides to set your clock backwards from 286"jumps", for example, when ntp decides to set your clock backwards from
261the wrong date of 2014-01-01 to 2008-01-01, a watcher that is supposed to 287the wrong date of 2014-01-01 to 2008-01-01, a watcher that is supposed to
262fire "after" a second might actually take six years to finally fire. 288fire "after a second" might actually take six years to finally fire.
263 289
264AnyEvent cannot compensate for this. The only event loop that is conscious 290AnyEvent cannot compensate for this. The only event loop that is conscious
265about these issues is L<EV>, which offers both relative (ev_timer, based 291of these issues is L<EV>, which offers both relative (ev_timer, based
266on true relative time) and absolute (ev_periodic, based on wallclock time) 292on true relative time) and absolute (ev_periodic, based on wallclock time)
267timers. 293timers.
268 294
269AnyEvent always prefers relative timers, if available, matching the 295AnyEvent always prefers relative timers, if available, matching the
270AnyEvent API. 296AnyEvent API.
292I<In almost all cases (in all cases if you don't care), this is the 318I<In almost all cases (in all cases if you don't care), this is the
293function to call when you want to know the current time.> 319function to call when you want to know the current time.>
294 320
295This function is also often faster then C<< AnyEvent->time >>, and 321This function is also often faster then C<< AnyEvent->time >>, and
296thus the preferred method if you want some timestamp (for example, 322thus the preferred method if you want some timestamp (for example,
297L<AnyEvent::Handle> uses this to update it's activity timeouts). 323L<AnyEvent::Handle> uses this to update its activity timeouts).
298 324
299The rest of this section is only of relevance if you try to be very exact 325The rest of this section is only of relevance if you try to be very exact
300with your timing, you can skip it without bad conscience. 326with your timing; you can skip it without a bad conscience.
301 327
302For a practical example of when these times differ, consider L<Event::Lib> 328For a practical example of when these times differ, consider L<Event::Lib>
303and L<EV> and the following set-up: 329and L<EV> and the following set-up:
304 330
305The event loop is running and has just invoked one of your callback at 331The event loop is running and has just invoked one of your callbacks at
306time=500 (assume no other callbacks delay processing). In your callback, 332time=500 (assume no other callbacks delay processing). In your callback,
307you wait a second by executing C<sleep 1> (blocking the process for a 333you wait a second by executing C<sleep 1> (blocking the process for a
308second) and then (at time=501) you create a relative timer that fires 334second) and then (at time=501) you create a relative timer that fires
309after three seconds. 335after three seconds.
310 336
330difference between C<< AnyEvent->time >> and C<< AnyEvent->now >> into 356difference between C<< AnyEvent->time >> and C<< AnyEvent->now >> into
331account. 357account.
332 358
333=item AnyEvent->now_update 359=item AnyEvent->now_update
334 360
335Some event loops (such as L<EV> or L<AnyEvent::Impl::Perl>) cache 361Some event loops (such as L<EV> or L<AnyEvent::Loop>) cache the current
336the current time for each loop iteration (see the discussion of L<< 362time for each loop iteration (see the discussion of L<< AnyEvent->now >>,
337AnyEvent->now >>, above). 363above).
338 364
339When a callback runs for a long time (or when the process sleeps), then 365When a callback runs for a long time (or when the process sleeps), then
340this "current" time will differ substantially from the real time, which 366this "current" time will differ substantially from the real time, which
341might affect timers and time-outs. 367might affect timers and time-outs.
342 368
343When this is the case, you can call this method, which will update the 369When this is the case, you can call this method, which will update the
344event loop's idea of "current time". 370event loop's idea of "current time".
345 371
372A typical example would be a script in a web server (e.g. C<mod_perl>) -
373when mod_perl executes the script, then the event loop will have the wrong
374idea about the "current time" (being potentially far in the past, when the
375script ran the last time). In that case you should arrange a call to C<<
376AnyEvent->now_update >> each time the web server process wakes up again
377(e.g. at the start of your script, or in a handler).
378
346Note that updating the time I<might> cause some events to be handled. 379Note that updating the time I<might> cause some events to be handled.
347 380
348=back 381=back
349 382
350=head2 SIGNAL WATCHERS 383=head2 SIGNAL WATCHERS
384
385 $w = AnyEvent->signal (signal => <uppercase_signal_name>, cb => <callback>);
351 386
352You can watch for signals using a signal watcher, C<signal> is the signal 387You can watch for signals using a signal watcher, C<signal> is the signal
353I<name> in uppercase and without any C<SIG> prefix, C<cb> is the Perl 388I<name> in uppercase and without any C<SIG> prefix, C<cb> is the Perl
354callback to be invoked whenever a signal occurs. 389callback to be invoked whenever a signal occurs.
355 390
361invocation, and callback invocation will be synchronous. Synchronous means 396invocation, and callback invocation will be synchronous. Synchronous means
362that it might take a while until the signal gets handled by the process, 397that it might take a while until the signal gets handled by the process,
363but it is guaranteed not to interrupt any other callbacks. 398but it is guaranteed not to interrupt any other callbacks.
364 399
365The main advantage of using these watchers is that you can share a signal 400The main advantage of using these watchers is that you can share a signal
366between multiple watchers. 401between multiple watchers, and AnyEvent will ensure that signals will not
402interrupt your program at bad times.
367 403
368This watcher might use C<%SIG>, so programs overwriting those signals 404This watcher might use C<%SIG> (depending on the event loop used),
369directly will likely not work correctly. 405so programs overwriting those signals directly will likely not work
406correctly.
370 407
371Example: exit on SIGINT 408Example: exit on SIGINT
372 409
373 my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 }); 410 my $w = AnyEvent->signal (signal => "INT", cb => sub { exit 1 });
374 411
412=head3 Restart Behaviour
413
414While restart behaviour is up to the event loop implementation, most will
415not restart syscalls (that includes L<Async::Interrupt> and AnyEvent's
416pure perl implementation).
417
418=head3 Safe/Unsafe Signals
419
420Perl signals can be either "safe" (synchronous to opcode handling) or
421"unsafe" (asynchronous) - the former might get delayed indefinitely, the
422latter might corrupt your memory.
423
424AnyEvent signal handlers are, in addition, synchronous to the event loop,
425i.e. they will not interrupt your running perl program but will only be
426called as part of the normal event handling (just like timer, I/O etc.
427callbacks, too).
428
429=head3 Signal Races, Delays and Workarounds
430
431Many event loops (e.g. Glib, Tk, Qt, IO::Async) do not support attaching
432callbacks to signals in a generic way, which is a pity, as you cannot
433do race-free signal handling in perl, requiring C libraries for
434this. AnyEvent will try to do its best, which means in some cases,
435signals will be delayed. The maximum time a signal might be delayed is
436specified in C<$AnyEvent::MAX_SIGNAL_LATENCY> (default: 10 seconds). This
437variable can be changed only before the first signal watcher is created,
438and should be left alone otherwise. This variable determines how often
439AnyEvent polls for signals (in case a wake-up was missed). Higher values
440will cause fewer spurious wake-ups, which is better for power and CPU
441saving.
442
443All these problems can be avoided by installing the optional
444L<Async::Interrupt> module, which works with most event loops. It will not
445work with inherently broken event loops such as L<Event> or L<Event::Lib>
446(and not with L<POE> currently, as POE does its own workaround with
447one-second latency). For those, you just have to suffer the delays.
448
375=head2 CHILD PROCESS WATCHERS 449=head2 CHILD PROCESS WATCHERS
376 450
451 $w = AnyEvent->child (pid => <process id>, cb => <callback>);
452
377You can also watch on a child process exit and catch its exit status. 453You can also watch for a child process exit and catch its exit status.
378 454
379The child process is specified by the C<pid> argument (if set to C<0>, it 455The child process is specified by the C<pid> argument (on some backends,
380watches for any child process exit). The watcher will triggered only when 456using C<0> watches for any child process exit, on others this will
381the child process has finished and an exit status is available, not on 457croak). The watcher will be triggered only when the child process has
382any trace events (stopped/continued). 458finished and an exit status is available, not on any trace events
459(stopped/continued).
383 460
384The callback will be called with the pid and exit status (as returned by 461The callback will be called with the pid and exit status (as returned by
385waitpid), so unlike other watcher types, you I<can> rely on child watcher 462waitpid), so unlike other watcher types, you I<can> rely on child watcher
386callback arguments. 463callback arguments.
387 464
392 469
393There is a slight catch to child watchers, however: you usually start them 470There is a slight catch to child watchers, however: you usually start them
394I<after> the child process was created, and this means the process could 471I<after> the child process was created, and this means the process could
395have exited already (and no SIGCHLD will be sent anymore). 472have exited already (and no SIGCHLD will be sent anymore).
396 473
397Not all event models handle this correctly (POE doesn't), but even for 474Not all event models handle this correctly (neither POE nor IO::Async do,
475see their AnyEvent::Impl manpages for details), but even for event models
398event models that I<do> handle this correctly, they usually need to be 476that I<do> handle this correctly, they usually need to be loaded before
399loaded before the process exits (i.e. before you fork in the first place). 477the process exits (i.e. before you fork in the first place). AnyEvent's
478pure perl event loop handles all cases correctly regardless of when you
479start the watcher.
400 480
401This means you cannot create a child watcher as the very first thing in an 481This means you cannot create a child watcher as the very first
402AnyEvent program, you I<have> to create at least one watcher before you 482thing in an AnyEvent program, you I<have> to create at least one
403C<fork> the child (alternatively, you can call C<AnyEvent::detect>). 483watcher before you C<fork> the child (alternatively, you can call
484C<AnyEvent::detect>).
485
486As most event loops do not support waiting for child events, they will be
487emulated by AnyEvent in most cases, in which case the latency and race
488problems mentioned in the description of signal watchers apply.
404 489
405Example: fork a process and wait for it 490Example: fork a process and wait for it
406 491
407 my $done = AnyEvent->condvar; 492 my $done = AnyEvent->condvar;
408 493
420 # do something else, then wait for process exit 505 # do something else, then wait for process exit
421 $done->recv; 506 $done->recv;
422 507
423=head2 IDLE WATCHERS 508=head2 IDLE WATCHERS
424 509
425Sometimes there is a need to do something, but it is not so important 510 $w = AnyEvent->idle (cb => <callback>);
426to do it instantly, but only when there is nothing better to do. This
427"nothing better to do" is usually defined to be "no other events need
428attention by the event loop".
429 511
430Idle watchers ideally get invoked when the event loop has nothing 512This will repeatedly invoke the callback after the process becomes idle,
431better to do, just before it would block the process to wait for new 513until either the watcher is destroyed or new events have been detected.
432events. Instead of blocking, the idle watcher is invoked.
433 514
434Most event loops unfortunately do not really support idle watchers (only 515Idle watchers are useful when there is a need to do something, but it
516is not so important (or wise) to do it instantly. The callback will be
517invoked only when there is "nothing better to do", which is usually
518defined as "all outstanding events have been handled and no new events
519have been detected". That means that idle watchers ideally get invoked
520when the event loop has just polled for new events but none have been
521detected. Instead of blocking to wait for more events, the idle watchers
522will be invoked.
523
524Unfortunately, most event loops do not really support idle watchers (only
435EV, Event and Glib do it in a usable fashion) - for the rest, AnyEvent 525EV, Event and Glib do it in a usable fashion) - for the rest, AnyEvent
436will simply call the callback "from time to time". 526will simply call the callback "from time to time".
437 527
438Example: read lines from STDIN, but only process them when the 528Example: read lines from STDIN, but only process them when the
439program is otherwise idle: 529program is otherwise idle:
455 }); 545 });
456 }); 546 });
457 547
458=head2 CONDITION VARIABLES 548=head2 CONDITION VARIABLES
459 549
550 $cv = AnyEvent->condvar;
551
552 $cv->send (<list>);
553 my @res = $cv->recv;
554
460If you are familiar with some event loops you will know that all of them 555If you are familiar with some event loops you will know that all of them
461require you to run some blocking "loop", "run" or similar function that 556require you to run some blocking "loop", "run" or similar function that
462will actively watch for new events and call your callbacks. 557will actively watch for new events and call your callbacks.
463 558
464AnyEvent is different, it expects somebody else to run the event loop and 559AnyEvent is slightly different: it expects somebody else to run the event
465will only block when necessary (usually when told by the user). 560loop and will only block when necessary (usually when told by the user).
466 561
467The instrument to do that is called a "condition variable", so called 562The tool to do that is called a "condition variable", so called because
468because they represent a condition that must become true. 563they represent a condition that must become true.
564
565Now is probably a good time to look at the examples further below.
469 566
470Condition variables can be created by calling the C<< AnyEvent->condvar 567Condition variables can be created by calling the C<< AnyEvent->condvar
471>> method, usually without arguments. The only argument pair allowed is 568>> method, usually without arguments. The only argument pair allowed is
472
473C<cb>, which specifies a callback to be called when the condition variable 569C<cb>, which specifies a callback to be called when the condition variable
474becomes true, with the condition variable as the first argument (but not 570becomes true, with the condition variable as the first argument (but not
475the results). 571the results).
476 572
477After creation, the condition variable is "false" until it becomes "true" 573After creation, the condition variable is "false" until it becomes "true"
478by calling the C<send> method (or calling the condition variable as if it 574by calling the C<send> method (or calling the condition variable as if it
479were a callback, read about the caveats in the description for the C<< 575were a callback, read about the caveats in the description for the C<<
480->send >> method). 576->send >> method).
481 577
482Condition variables are similar to callbacks, except that you can 578Since condition variables are the most complex part of the AnyEvent API, here are
483optionally wait for them. They can also be called merge points - points 579some different mental models of what they are - pick the ones you can connect to:
484in time where multiple outstanding events have been processed. And yet 580
485another way to call them is transactions - each condition variable can be 581=over 4
486used to represent a transaction, which finishes at some point and delivers 582
487a result. 583=item * Condition variables are like callbacks - you can call them (and pass them instead
584of callbacks). Unlike callbacks however, you can also wait for them to be called.
585
586=item * Condition variables are signals - one side can emit or send them,
587the other side can wait for them, or install a handler that is called when
588the signal fires.
589
590=item * Condition variables are like "Merge Points" - points in your program
591where you merge multiple independent results/control flows into one.
592
593=item * Condition variables represent a transaction - functions that start
594some kind of transaction can return them, leaving the caller the choice
595between waiting in a blocking fashion, or setting a callback.
596
597=item * Condition variables represent future values, or promises to deliver
598some result, long before the result is available.
599
600=back
488 601
489Condition variables are very useful to signal that something has finished, 602Condition variables are very useful to signal that something has finished,
490for example, if you write a module that does asynchronous http requests, 603for example, if you write a module that does asynchronous http requests,
491then a condition variable would be the ideal candidate to signal the 604then a condition variable would be the ideal candidate to signal the
492availability of results. The user can either act when the callback is 605availability of results. The user can either act when the callback is
505 618
506Condition variables are represented by hash refs in perl, and the keys 619Condition variables are represented by hash refs in perl, and the keys
507used by AnyEvent itself are all named C<_ae_XXX> to make subclassing 620used by AnyEvent itself are all named C<_ae_XXX> to make subclassing
508easy (it is often useful to build your own transaction class on top of 621easy (it is often useful to build your own transaction class on top of
509AnyEvent). To subclass, use C<AnyEvent::CondVar> as base class and call 622AnyEvent). To subclass, use C<AnyEvent::CondVar> as base class and call
510it's C<new> method in your own C<new> method. 623its C<new> method in your own C<new> method.
511 624
512There are two "sides" to a condition variable - the "producer side" which 625There are two "sides" to a condition variable - the "producer side" which
513eventually calls C<< -> send >>, and the "consumer side", which waits 626eventually calls C<< -> send >>, and the "consumer side", which waits
514for the send to occur. 627for the send to occur.
515 628
516Example: wait for a timer. 629Example: wait for a timer.
517 630
518 # wait till the result is ready 631 # condition: "wait till the timer is fired"
519 my $result_ready = AnyEvent->condvar; 632 my $timer_fired = AnyEvent->condvar;
520 633
521 # do something such as adding a timer 634 # create the timer - we could wait for, say
522 # or socket watcher the calls $result_ready->send 635 # a handle becomign ready, or even an
523 # when the "result" is ready. 636 # AnyEvent::HTTP request to finish, but
524 # in this case, we simply use a timer: 637 # in this case, we simply use a timer:
525 my $w = AnyEvent->timer ( 638 my $w = AnyEvent->timer (
526 after => 1, 639 after => 1,
527 cb => sub { $result_ready->send }, 640 cb => sub { $timer_fired->send },
528 ); 641 );
529 642
530 # this "blocks" (while handling events) till the callback 643 # this "blocks" (while handling events) till the callback
531 # calls send 644 # calls ->send
532 $result_ready->recv; 645 $timer_fired->recv;
533 646
534Example: wait for a timer, but take advantage of the fact that 647Example: wait for a timer, but take advantage of the fact that condition
535condition variables are also code references. 648variables are also callable directly.
536 649
537 my $done = AnyEvent->condvar; 650 my $done = AnyEvent->condvar;
538 my $delay = AnyEvent->timer (after => 5, cb => $done); 651 my $delay = AnyEvent->timer (after => 5, cb => $done);
539 $done->recv; 652 $done->recv;
540 653
546 659
547 ... 660 ...
548 661
549 my @info = $couchdb->info->recv; 662 my @info = $couchdb->info->recv;
550 663
551And this is how you would just ste a callback to be called whenever the 664And this is how you would just set a callback to be called whenever the
552results are available: 665results are available:
553 666
554 $couchdb->info->cb (sub { 667 $couchdb->info->cb (sub {
555 my @info = $_[0]->recv; 668 my @info = $_[0]->recv;
556 }); 669 });
574immediately from within send. 687immediately from within send.
575 688
576Any arguments passed to the C<send> call will be returned by all 689Any arguments passed to the C<send> call will be returned by all
577future C<< ->recv >> calls. 690future C<< ->recv >> calls.
578 691
579Condition variables are overloaded so one can call them directly 692Condition variables are overloaded so one can call them directly (as if
580(as a code reference). Calling them directly is the same as calling 693they were a code reference). Calling them directly is the same as calling
581C<send>. Note, however, that many C-based event loops do not handle 694C<send>.
582overloading, so as tempting as it may be, passing a condition variable
583instead of a callback does not work. Both the pure perl and EV loops
584support overloading, however, as well as all functions that use perl to
585invoke a callback (as in L<AnyEvent::Socket> and L<AnyEvent::DNS> for
586example).
587 695
588=item $cv->croak ($error) 696=item $cv->croak ($error)
589 697
590Similar to send, but causes all call's to C<< ->recv >> to invoke 698Similar to send, but causes all calls to C<< ->recv >> to invoke
591C<Carp::croak> with the given error message/object/scalar. 699C<Carp::croak> with the given error message/object/scalar.
592 700
593This can be used to signal any errors to the condition variable 701This can be used to signal any errors to the condition variable
594user/consumer. 702user/consumer. Doing it this way instead of calling C<croak> directly
703delays the error detection, but has the overwhelming advantage that it
704diagnoses the error at the place where the result is expected, and not
705deep in some event callback with no connection to the actual code causing
706the problem.
595 707
596=item $cv->begin ([group callback]) 708=item $cv->begin ([group callback])
597 709
598=item $cv->end 710=item $cv->end
599
600These two methods are EXPERIMENTAL and MIGHT CHANGE.
601 711
602These two methods can be used to combine many transactions/events into 712These two methods can be used to combine many transactions/events into
603one. For example, a function that pings many hosts in parallel might want 713one. For example, a function that pings many hosts in parallel might want
604to use a condition variable for the whole process. 714to use a condition variable for the whole process.
605 715
606Every call to C<< ->begin >> will increment a counter, and every call to 716Every call to C<< ->begin >> will increment a counter, and every call to
607C<< ->end >> will decrement it. If the counter reaches C<0> in C<< ->end 717C<< ->end >> will decrement it. If the counter reaches C<0> in C<< ->end
608>>, the (last) callback passed to C<begin> will be executed. That callback 718>>, the (last) callback passed to C<begin> will be executed, passing the
609is I<supposed> to call C<< ->send >>, but that is not required. If no 719condvar as first argument. That callback is I<supposed> to call C<< ->send
610callback was set, C<send> will be called without any arguments. 720>>, but that is not required. If no group callback was set, C<send> will
721be called without any arguments.
611 722
612Let's clarify this with the ping example: 723You can think of C<< $cv->send >> giving you an OR condition (one call
724sends), while C<< $cv->begin >> and C<< $cv->end >> giving you an AND
725condition (all C<begin> calls must be C<end>'ed before the condvar sends).
726
727Let's start with a simple example: you have two I/O watchers (for example,
728STDOUT and STDERR for a program), and you want to wait for both streams to
729close before activating a condvar:
613 730
614 my $cv = AnyEvent->condvar; 731 my $cv = AnyEvent->condvar;
615 732
733 $cv->begin; # first watcher
734 my $w1 = AnyEvent->io (fh => $fh1, cb => sub {
735 defined sysread $fh1, my $buf, 4096
736 or $cv->end;
737 });
738
739 $cv->begin; # second watcher
740 my $w2 = AnyEvent->io (fh => $fh2, cb => sub {
741 defined sysread $fh2, my $buf, 4096
742 or $cv->end;
743 });
744
745 $cv->recv;
746
747This works because for every event source (EOF on file handle), there is
748one call to C<begin>, so the condvar waits for all calls to C<end> before
749sending.
750
751The ping example mentioned above is slightly more complicated, as the
752there are results to be passwd back, and the number of tasks that are
753begun can potentially be zero:
754
755 my $cv = AnyEvent->condvar;
756
616 my %result; 757 my %result;
617 $cv->begin (sub { $cv->send (\%result) }); 758 $cv->begin (sub { shift->send (\%result) });
618 759
619 for my $host (@list_of_hosts) { 760 for my $host (@list_of_hosts) {
620 $cv->begin; 761 $cv->begin;
621 ping_host_then_call_callback $host, sub { 762 ping_host_then_call_callback $host, sub {
622 $result{$host} = ...; 763 $result{$host} = ...;
637loop, which serves two important purposes: first, it sets the callback 778loop, which serves two important purposes: first, it sets the callback
638to be called once the counter reaches C<0>, and second, it ensures that 779to be called once the counter reaches C<0>, and second, it ensures that
639C<send> is called even when C<no> hosts are being pinged (the loop 780C<send> is called even when C<no> hosts are being pinged (the loop
640doesn't execute once). 781doesn't execute once).
641 782
642This is the general pattern when you "fan out" into multiple subrequests: 783This is the general pattern when you "fan out" into multiple (but
643use an outer C<begin>/C<end> pair to set the callback and ensure C<end> 784potentially zero) subrequests: use an outer C<begin>/C<end> pair to set
644is called at least once, and then, for each subrequest you start, call 785the callback and ensure C<end> is called at least once, and then, for each
645C<begin> and for each subrequest you finish, call C<end>. 786subrequest you start, call C<begin> and for each subrequest you finish,
787call C<end>.
646 788
647=back 789=back
648 790
649=head3 METHODS FOR CONSUMERS 791=head3 METHODS FOR CONSUMERS
650 792
654=over 4 796=over 4
655 797
656=item $cv->recv 798=item $cv->recv
657 799
658Wait (blocking if necessary) until the C<< ->send >> or C<< ->croak 800Wait (blocking if necessary) until the C<< ->send >> or C<< ->croak
659>> methods have been called on c<$cv>, while servicing other watchers 801>> methods have been called on C<$cv>, while servicing other watchers
660normally. 802normally.
661 803
662You can only wait once on a condition - additional calls are valid but 804You can only wait once on a condition - additional calls are valid but
663will return immediately. 805will return immediately.
664 806
666function will call C<croak>. 808function will call C<croak>.
667 809
668In list context, all parameters passed to C<send> will be returned, 810In list context, all parameters passed to C<send> will be returned,
669in scalar context only the first one will be returned. 811in scalar context only the first one will be returned.
670 812
813Note that doing a blocking wait in a callback is not supported by any
814event loop, that is, recursive invocation of a blocking C<< ->recv
815>> is not allowed, and the C<recv> call will C<croak> if such a
816condition is detected. This condition can be slightly loosened by using
817L<Coro::AnyEvent>, which allows you to do a blocking C<< ->recv >> from
818any thread that doesn't run the event loop itself.
819
671Not all event models support a blocking wait - some die in that case 820Not all event models support a blocking wait - some die in that case
672(programs might want to do that to stay interactive), so I<if you are 821(programs might want to do that to stay interactive), so I<if you are
673using this from a module, never require a blocking wait>, but let the 822using this from a module, never require a blocking wait>. Instead, let the
674caller decide whether the call will block or not (for example, by coupling 823caller decide whether the call will block or not (for example, by coupling
675condition variables with some kind of request results and supporting 824condition variables with some kind of request results and supporting
676callbacks so the caller knows that getting the result will not block, 825callbacks so the caller knows that getting the result will not block,
677while still supporting blocking waits if the caller so desires). 826while still supporting blocking waits if the caller so desires).
678 827
679Another reason I<never> to C<< ->recv >> in a module is that you cannot
680sensibly have two C<< ->recv >>'s in parallel, as that would require
681multiple interpreters or coroutines/threads, none of which C<AnyEvent>
682can supply.
683
684The L<Coro> module, however, I<can> and I<does> supply coroutines and, in
685fact, L<Coro::AnyEvent> replaces AnyEvent's condvars by coroutine-safe
686versions and also integrates coroutines into AnyEvent, making blocking
687C<< ->recv >> calls perfectly safe as long as they are done from another
688coroutine (one that doesn't run the event loop).
689
690You can ensure that C<< -recv >> never blocks by setting a callback and 828You can ensure that C<< ->recv >> never blocks by setting a callback and
691only calling C<< ->recv >> from within that callback (or at a later 829only calling C<< ->recv >> from within that callback (or at a later
692time). This will work even when the event loop does not support blocking 830time). This will work even when the event loop does not support blocking
693waits otherwise. 831waits otherwise.
694 832
695=item $bool = $cv->ready 833=item $bool = $cv->ready
701 839
702This is a mutator function that returns the callback set and optionally 840This is a mutator function that returns the callback set and optionally
703replaces it before doing so. 841replaces it before doing so.
704 842
705The callback will be called when the condition becomes "true", i.e. when 843The callback will be called when the condition becomes "true", i.e. when
706C<send> or C<croak> are called, with the only argument being the condition 844C<send> or C<croak> are called, with the only argument being the
707variable itself. Calling C<recv> inside the callback or at any later time 845condition variable itself. If the condition is already true, the
708is guaranteed not to block. 846callback is called immediately when it is set. Calling C<recv> inside
847the callback or at any later time is guaranteed not to block.
709 848
710=back 849=back
711 850
851=head1 SUPPORTED EVENT LOOPS/BACKENDS
852
853The available backend classes are (every class has its own manpage):
854
855=over 4
856
857=item Backends that are autoprobed when no other event loop can be found.
858
859EV is the preferred backend when no other event loop seems to be in
860use. If EV is not installed, then AnyEvent will fall back to its own
861pure-perl implementation, which is available everywhere as it comes with
862AnyEvent itself.
863
864 AnyEvent::Impl::EV based on EV (interface to libev, best choice).
865 AnyEvent::Impl::Perl pure-perl AnyEvent::Loop, fast and portable.
866
867=item Backends that are transparently being picked up when they are used.
868
869These will be used if they are already loaded when the first watcher
870is created, in which case it is assumed that the application is using
871them. This means that AnyEvent will automatically pick the right backend
872when the main program loads an event module before anything starts to
873create watchers. Nothing special needs to be done by the main program.
874
875 AnyEvent::Impl::Event based on Event, very stable, few glitches.
876 AnyEvent::Impl::Glib based on Glib, slow but very stable.
877 AnyEvent::Impl::Tk based on Tk, very broken.
878 AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
879 AnyEvent::Impl::POE based on POE, very slow, some limitations.
880 AnyEvent::Impl::Irssi used when running within irssi.
881 AnyEvent::Impl::IOAsync based on IO::Async.
882 AnyEvent::Impl::Cocoa based on Cocoa::EventLoop.
883 AnyEvent::Impl::FLTK2 based on FLTK (fltk 2 binding).
884
885=item Backends with special needs.
886
887Qt requires the Qt::Application to be instantiated first, but will
888otherwise be picked up automatically. As long as the main program
889instantiates the application before any AnyEvent watchers are created,
890everything should just work.
891
892 AnyEvent::Impl::Qt based on Qt.
893
894=item Event loops that are indirectly supported via other backends.
895
896Some event loops can be supported via other modules:
897
898There is no direct support for WxWidgets (L<Wx>) or L<Prima>.
899
900B<WxWidgets> has no support for watching file handles. However, you can
901use WxWidgets through the POE adaptor, as POE has a Wx backend that simply
902polls 20 times per second, which was considered to be too horrible to even
903consider for AnyEvent.
904
905B<Prima> is not supported as nobody seems to be using it, but it has a POE
906backend, so it can be supported through POE.
907
908AnyEvent knows about both L<Prima> and L<Wx>, however, and will try to
909load L<POE> when detecting them, in the hope that POE will pick them up,
910in which case everything will be automatic.
911
912=back
913
712=head1 GLOBAL VARIABLES AND FUNCTIONS 914=head1 GLOBAL VARIABLES AND FUNCTIONS
713 915
916These are not normally required to use AnyEvent, but can be useful to
917write AnyEvent extension modules.
918
714=over 4 919=over 4
715 920
716=item $AnyEvent::MODEL 921=item $AnyEvent::MODEL
717 922
718Contains C<undef> until the first watcher is being created. Then it 923Contains C<undef> until the first watcher is being created, before the
924backend has been autodetected.
925
719contains the event model that is being used, which is the name of the 926Afterwards it contains the event model that is being used, which is the
720Perl class implementing the model. This class is usually one of the 927name of the Perl class implementing the model. This class is usually one
721C<AnyEvent::Impl:xxx> modules, but can be any other class in the case 928of the C<AnyEvent::Impl::xxx> modules, but can be any other class in the
722AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode>). 929case AnyEvent has been extended at runtime (e.g. in I<rxvt-unicode> it
723 930will be C<urxvt::anyevent>).
724The known classes so far are:
725
726 AnyEvent::Impl::EV based on EV (an interface to libev, best choice).
727 AnyEvent::Impl::Event based on Event, second best choice.
728 AnyEvent::Impl::Perl pure-perl implementation, fast and portable.
729 AnyEvent::Impl::Glib based on Glib, third-best choice.
730 AnyEvent::Impl::Tk based on Tk, very bad choice.
731 AnyEvent::Impl::Qt based on Qt, cannot be autoprobed (see its docs).
732 AnyEvent::Impl::EventLib based on Event::Lib, leaks memory and worse.
733 AnyEvent::Impl::POE based on POE, not generic enough for full support.
734
735There is no support for WxWidgets, as WxWidgets has no support for
736watching file handles. However, you can use WxWidgets through the
737POE Adaptor, as POE has a Wx backend that simply polls 20 times per
738second, which was considered to be too horrible to even consider for
739AnyEvent. Likewise, other POE backends can be used by AnyEvent by using
740it's adaptor.
741
742AnyEvent knows about L<Prima> and L<Wx> and will try to use L<POE> when
743autodetecting them.
744 931
745=item AnyEvent::detect 932=item AnyEvent::detect
746 933
747Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model 934Returns C<$AnyEvent::MODEL>, forcing autodetection of the event model
748if necessary. You should only call this function right before you would 935if necessary. You should only call this function right before you would
749have created an AnyEvent watcher anyway, that is, as late as possible at 936have created an AnyEvent watcher anyway, that is, as late as possible at
750runtime. 937runtime, and not e.g. during initialisation of your module.
938
939The effect of calling this function is as if a watcher had been created
940(specifically, actions that happen "when the first watcher is created"
941happen when calling detetc as well).
942
943If you need to do some initialisation before AnyEvent watchers are
944created, use C<post_detect>.
751 945
752=item $guard = AnyEvent::post_detect { BLOCK } 946=item $guard = AnyEvent::post_detect { BLOCK }
753 947
754Arranges for the code block to be executed as soon as the event model is 948Arranges for the code block to be executed as soon as the event model is
755autodetected (or immediately if this has already happened). 949autodetected (or immediately if that has already happened).
950
951The block will be executed I<after> the actual backend has been detected
952(C<$AnyEvent::MODEL> is set), but I<before> any watchers have been
953created, so it is possible to e.g. patch C<@AnyEvent::ISA> or do
954other initialisations - see the sources of L<AnyEvent::Strict> or
955L<AnyEvent::AIO> to see how this is used.
956
957The most common usage is to create some global watchers, without forcing
958event module detection too early, for example, L<AnyEvent::AIO> creates
959and installs the global L<IO::AIO> watcher in a C<post_detect> block to
960avoid autodetecting the event module at load time.
756 961
757If called in scalar or list context, then it creates and returns an object 962If called in scalar or list context, then it creates and returns an object
758that automatically removes the callback again when it is destroyed. See 963that automatically removes the callback again when it is destroyed (or
964C<undef> when the hook was immediately executed). See L<AnyEvent::AIO> for
759L<Coro::BDB> for a case where this is useful. 965a case where this is useful.
966
967Example: Create a watcher for the IO::AIO module and store it in
968C<$WATCHER>, but do so only do so after the event loop is initialised.
969
970 our WATCHER;
971
972 my $guard = AnyEvent::post_detect {
973 $WATCHER = AnyEvent->io (fh => IO::AIO::poll_fileno, poll => 'r', cb => \&IO::AIO::poll_cb);
974 };
975
976 # the ||= is important in case post_detect immediately runs the block,
977 # as to not clobber the newly-created watcher. assigning both watcher and
978 # post_detect guard to the same variable has the advantage of users being
979 # able to just C<undef $WATCHER> if the watcher causes them grief.
980
981 $WATCHER ||= $guard;
760 982
761=item @AnyEvent::post_detect 983=item @AnyEvent::post_detect
762 984
763If there are any code references in this array (you can C<push> to it 985If there are any code references in this array (you can C<push> to it
764before or after loading AnyEvent), then they will called directly after 986before or after loading AnyEvent), then they will be called directly
765the event loop has been chosen. 987after the event loop has been chosen.
766 988
767You should check C<$AnyEvent::MODEL> before adding to this array, though: 989You should check C<$AnyEvent::MODEL> before adding to this array, though:
768if it contains a true value then the event loop has already been detected, 990if it is defined then the event loop has already been detected, and the
769and the array will be ignored. 991array will be ignored.
770 992
771Best use C<AnyEvent::post_detect { BLOCK }> instead. 993Best use C<AnyEvent::post_detect { BLOCK }> when your application allows
994it, as it takes care of these details.
995
996This variable is mainly useful for modules that can do something useful
997when AnyEvent is used and thus want to know when it is initialised, but do
998not need to even load it by default. This array provides the means to hook
999into AnyEvent passively, without loading it.
1000
1001Example: To load Coro::AnyEvent whenever Coro and AnyEvent are used
1002together, you could put this into Coro (this is the actual code used by
1003Coro to accomplish this):
1004
1005 if (defined $AnyEvent::MODEL) {
1006 # AnyEvent already initialised, so load Coro::AnyEvent
1007 require Coro::AnyEvent;
1008 } else {
1009 # AnyEvent not yet initialised, so make sure to load Coro::AnyEvent
1010 # as soon as it is
1011 push @AnyEvent::post_detect, sub { require Coro::AnyEvent };
1012 }
1013
1014=item AnyEvent::postpone { BLOCK }
1015
1016Arranges for the block to be executed as soon as possible, but not before
1017the call itself returns. In practise, the block will be executed just
1018before the event loop polls for new events, or shortly afterwards.
1019
1020This function never returns anything (to make the C<return postpone { ...
1021}> idiom more useful.
1022
1023To understand the usefulness of this function, consider a function that
1024asynchronously does something for you and returns some transaction
1025object or guard to let you cancel the operation. For example,
1026C<AnyEvent::Socket::tcp_connect>:
1027
1028 # start a conenction attempt unless one is active
1029 $self->{connect_guard} ||= AnyEvent::Socket::tcp_connect "www.example.net", 80, sub {
1030 delete $self->{connect_guard};
1031 ...
1032 };
1033
1034Imagine that this function could instantly call the callback, for
1035example, because it detects an obvious error such as a negative port
1036number. Invoking the callback before the function returns causes problems
1037however: the callback will be called and will try to delete the guard
1038object. But since the function hasn't returned yet, there is nothing to
1039delete. When the function eventually returns it will assign the guard
1040object to C<< $self->{connect_guard} >>, where it will likely never be
1041deleted, so the program thinks it is still trying to connect.
1042
1043This is where C<AnyEvent::postpone> should be used. Instead of calling the
1044callback directly on error:
1045
1046 $cb->(undef), return # signal error to callback, BAD!
1047 if $some_error_condition;
1048
1049It should use C<postpone>:
1050
1051 AnyEvent::postpone { $cb->(undef) }, return # signal error to callback, later
1052 if $some_error_condition;
1053
1054=item AnyEvent::log $level, $msg[, @args]
1055
1056Log the given C<$msg> at the given C<$level>.
1057
1058Loads AnyEvent::Log on first use and calls C<AnyEvent::Log::log> -
1059consequently, look at the L<AnyEvent::Log> documentation for details.
1060
1061If you want to sprinkle loads of logging calls around your code, consider
1062creating a logger callback with the C<AnyEvent::Log::logger< function.
772 1063
773=back 1064=back
774 1065
775=head1 WHAT TO DO IN A MODULE 1066=head1 WHAT TO DO IN A MODULE
776 1067
787because it will stall the whole program, and the whole point of using 1078because it will stall the whole program, and the whole point of using
788events is to stay interactive. 1079events is to stay interactive.
789 1080
790It is fine, however, to call C<< ->recv >> when the user of your module 1081It is fine, however, to call C<< ->recv >> when the user of your module
791requests it (i.e. if you create a http request object ad have a method 1082requests it (i.e. if you create a http request object ad have a method
792called C<results> that returns the results, it should call C<< ->recv >> 1083called C<results> that returns the results, it may call C<< ->recv >>
793freely, as the user of your module knows what she is doing. always). 1084freely, as the user of your module knows what she is doing. Always).
794 1085
795=head1 WHAT TO DO IN THE MAIN PROGRAM 1086=head1 WHAT TO DO IN THE MAIN PROGRAM
796 1087
797There will always be a single main program - the only place that should 1088There will always be a single main program - the only place that should
798dictate which event model to use. 1089dictate which event model to use.
799 1090
800If it doesn't care, it can just "use AnyEvent" and use it itself, or not 1091If the program is not event-based, it need not do anything special, even
801do anything special (it does not need to be event-based) and let AnyEvent 1092when it depends on a module that uses an AnyEvent. If the program itself
802decide which implementation to chose if some module relies on it. 1093uses AnyEvent, but does not care which event loop is used, all it needs
1094to do is C<use AnyEvent>. In either case, AnyEvent will choose the best
1095available loop implementation.
803 1096
804If the main program relies on a specific event model - for example, in 1097If the main program relies on a specific event model - for example, in
805Gtk2 programs you have to rely on the Glib module - you should load the 1098Gtk2 programs you have to rely on the Glib module - you should load the
806event module before loading AnyEvent or any module that uses it: generally 1099event module before loading AnyEvent or any module that uses it: generally
807speaking, you should load it as early as possible. The reason is that 1100speaking, you should load it as early as possible. The reason is that
808modules might create watchers when they are loaded, and AnyEvent will 1101modules might create watchers when they are loaded, and AnyEvent will
809decide on the event model to use as soon as it creates watchers, and it 1102decide on the event model to use as soon as it creates watchers, and it
810might chose the wrong one unless you load the correct one yourself. 1103might choose the wrong one unless you load the correct one yourself.
811 1104
812You can chose to use a pure-perl implementation by loading the 1105You can chose to use a pure-perl implementation by loading the
813C<AnyEvent::Impl::Perl> module, which gives you similar behaviour 1106C<AnyEvent::Loop> module, which gives you similar behaviour
814everywhere, but letting AnyEvent chose the model is generally better. 1107everywhere, but letting AnyEvent chose the model is generally better.
815 1108
816=head2 MAINLOOP EMULATION 1109=head2 MAINLOOP EMULATION
817 1110
818Sometimes (often for short test scripts, or even standalone programs who 1111Sometimes (often for short test scripts, or even standalone programs who
831 1124
832 1125
833=head1 OTHER MODULES 1126=head1 OTHER MODULES
834 1127
835The following is a non-exhaustive list of additional modules that use 1128The following is a non-exhaustive list of additional modules that use
836AnyEvent and can therefore be mixed easily with other AnyEvent modules 1129AnyEvent as a client and can therefore be mixed easily with other AnyEvent
837in the same program. Some of the modules come with AnyEvent, some are 1130modules and other event loops in the same program. Some of the modules
838available via CPAN. 1131come as part of AnyEvent, the others are available via CPAN.
839 1132
840=over 4 1133=over 4
841 1134
842=item L<AnyEvent::Util> 1135=item L<AnyEvent::Util>
843 1136
844Contains various utility functions that replace often-used but blocking 1137Contains various utility functions that replace often-used blocking
845functions such as C<inet_aton> by event-/callback-based versions. 1138functions such as C<inet_aton> with event/callback-based versions.
846 1139
847=item L<AnyEvent::Socket> 1140=item L<AnyEvent::Socket>
848 1141
849Provides various utility functions for (internet protocol) sockets, 1142Provides various utility functions for (internet protocol) sockets,
850addresses and name resolution. Also functions to create non-blocking tcp 1143addresses and name resolution. Also functions to create non-blocking tcp
852 1145
853=item L<AnyEvent::Handle> 1146=item L<AnyEvent::Handle>
854 1147
855Provide read and write buffers, manages watchers for reads and writes, 1148Provide read and write buffers, manages watchers for reads and writes,
856supports raw and formatted I/O, I/O queued and fully transparent and 1149supports raw and formatted I/O, I/O queued and fully transparent and
857non-blocking SSL/TLS. 1150non-blocking SSL/TLS (via L<AnyEvent::TLS>).
858 1151
859=item L<AnyEvent::DNS> 1152=item L<AnyEvent::DNS>
860 1153
861Provides rich asynchronous DNS resolver capabilities. 1154Provides rich asynchronous DNS resolver capabilities.
862 1155
1156=item L<AnyEvent::HTTP>, L<AnyEvent::IRC>, L<AnyEvent::XMPP>, L<AnyEvent::GPSD>, L<AnyEvent::IGS>, L<AnyEvent::FCP>
1157
1158Implement event-based interfaces to the protocols of the same name (for
1159the curious, IGS is the International Go Server and FCP is the Freenet
1160Client Protocol).
1161
1162=item L<AnyEvent::Handle::UDP>
1163
1164Here be danger!
1165
1166As Pauli would put it, "Not only is it not right, it's not even wrong!" -
1167there are so many things wrong with AnyEvent::Handle::UDP, most notably
1168its use of a stream-based API with a protocol that isn't streamable, that
1169the only way to improve it is to delete it.
1170
1171It features data corruption (but typically only under load) and general
1172confusion. On top, the author is not only clueless about UDP but also
1173fact-resistant - some gems of his understanding: "connect doesn't work
1174with UDP", "UDP packets are not IP packets", "UDP only has datagrams, not
1175packets", "I don't need to implement proper error checking as UDP doesn't
1176support error checking" and so on - he doesn't even understand what's
1177wrong with his module when it is explained to him.
1178
863=item L<AnyEvent::HTTP> 1179=item L<AnyEvent::DBI>
864 1180
865A simple-to-use HTTP library that is capable of making a lot of concurrent 1181Executes L<DBI> requests asynchronously in a proxy process for you,
866HTTP requests. 1182notifying you in an event-based way when the operation is finished.
1183
1184=item L<AnyEvent::AIO>
1185
1186Truly asynchronous (as opposed to non-blocking) I/O, should be in the
1187toolbox of every event programmer. AnyEvent::AIO transparently fuses
1188L<IO::AIO> and AnyEvent together, giving AnyEvent access to event-based
1189file I/O, and much more.
867 1190
868=item L<AnyEvent::HTTPD> 1191=item L<AnyEvent::HTTPD>
869 1192
870Provides a simple web application server framework. 1193A simple embedded webserver.
871 1194
872=item L<AnyEvent::FastPing> 1195=item L<AnyEvent::FastPing>
873 1196
874The fastest ping in the west. 1197The fastest ping in the west.
875 1198
876=item L<AnyEvent::DBI>
877
878Executes L<DBI> requests asynchronously in a proxy process.
879
880=item L<AnyEvent::AIO>
881
882Truly asynchronous I/O, should be in the toolbox of every event
883programmer. AnyEvent::AIO transparently fuses L<IO::AIO> and AnyEvent
884together.
885
886=item L<AnyEvent::BDB>
887
888Truly asynchronous Berkeley DB access. AnyEvent::BDB transparently fuses
889L<BDB> and AnyEvent together.
890
891=item L<AnyEvent::GPSD>
892
893A non-blocking interface to gpsd, a daemon delivering GPS information.
894
895=item L<AnyEvent::IGS>
896
897A non-blocking interface to the Internet Go Server protocol (used by
898L<App::IGS>).
899
900=item L<AnyEvent::IRC>
901
902AnyEvent based IRC client module family (replacing the older Net::IRC3).
903
904=item L<Net::XMPP2>
905
906AnyEvent based XMPP (Jabber protocol) module family.
907
908=item L<Net::FCP>
909
910AnyEvent-based implementation of the Freenet Client Protocol, birthplace
911of AnyEvent.
912
913=item L<Event::ExecFlow>
914
915High level API for event-based execution flow control.
916
917=item L<Coro> 1199=item L<Coro>
918 1200
919Has special support for AnyEvent via L<Coro::AnyEvent>. 1201Has special support for AnyEvent via L<Coro::AnyEvent>.
920 1202
921=item L<IO::Lambda>
922
923The lambda approach to I/O - don't ask, look there. Can use AnyEvent.
924
925=back 1203=back
926 1204
927=cut 1205=cut
928 1206
929package AnyEvent; 1207package AnyEvent;
930 1208
931no warnings; 1209# basically a tuned-down version of common::sense
932use strict qw(vars subs); 1210sub common_sense {
1211 # from common:.sense 3.4
1212 ${^WARNING_BITS} ^= ${^WARNING_BITS} ^ "\x3c\x3f\x33\x00\x0f\xf0\x0f\xc0\xf0\xfc\x33\x00";
1213 # use strict vars subs - NO UTF-8, as Util.pm doesn't like this atm. (uts46data.pl)
1214 $^H |= 0x00000600;
1215}
933 1216
1217BEGIN { AnyEvent::common_sense }
1218
934use Carp; 1219use Carp ();
935 1220
936our $VERSION = 4.411; 1221our $VERSION = '6.01';
937our $MODEL; 1222our $MODEL;
938 1223
939our $AUTOLOAD;
940our @ISA; 1224our @ISA;
941 1225
942our @REGISTRY; 1226our @REGISTRY;
943 1227
944our $WIN32; 1228our $VERBOSE;
945 1229
946BEGIN { 1230BEGIN {
947 eval "sub WIN32(){ " . (($^O =~ /mswin32/i)*1) ." }"; 1231 require "AnyEvent/constants.pl";
1232
948 eval "sub TAINT(){ " . (${^TAINT}*1) . " }"; 1233 eval "sub TAINT (){" . (${^TAINT}*1) . "}";
949 1234
950 delete @ENV{grep /^PERL_ANYEVENT_/, keys %ENV} 1235 delete @ENV{grep /^PERL_ANYEVENT_/, keys %ENV}
951 if ${^TAINT}; 1236 if ${^TAINT};
952}
953 1237
954our $verbose = $ENV{PERL_ANYEVENT_VERBOSE}*1; 1238 $VERBOSE = $ENV{PERL_ANYEVENT_VERBOSE}*1;
1239}
1240
1241our $MAX_SIGNAL_LATENCY = 10;
955 1242
956our %PROTOCOL; # (ipv4|ipv6) => (1|2), higher numbers are preferred 1243our %PROTOCOL; # (ipv4|ipv6) => (1|2), higher numbers are preferred
957 1244
958{ 1245{
959 my $idx; 1246 my $idx;
960 $PROTOCOL{$_} = ++$idx 1247 $PROTOCOL{$_} = ++$idx
961 for reverse split /\s*,\s*/, 1248 for reverse split /\s*,\s*/,
962 $ENV{PERL_ANYEVENT_PROTOCOLS} || "ipv4,ipv6"; 1249 $ENV{PERL_ANYEVENT_PROTOCOLS} || "ipv4,ipv6";
963} 1250}
964 1251
1252our @post_detect;
1253
1254sub post_detect(&) {
1255 my ($cb) = @_;
1256
1257 push @post_detect, $cb;
1258
1259 defined wantarray
1260 ? bless \$cb, "AnyEvent::Util::postdetect"
1261 : ()
1262}
1263
1264sub AnyEvent::Util::postdetect::DESTROY {
1265 @post_detect = grep $_ != ${$_[0]}, @post_detect;
1266}
1267
1268our $POSTPONE_W;
1269our @POSTPONE;
1270
1271sub _postpone_exec {
1272 undef $POSTPONE_W;
1273
1274 &{ shift @POSTPONE }
1275 while @POSTPONE;
1276}
1277
1278sub postpone(&) {
1279 push @POSTPONE, shift;
1280
1281 $POSTPONE_W ||= AE::timer (0, 0, \&_postpone_exec);
1282
1283 ()
1284}
1285
1286sub log($$;@) {
1287 require AnyEvent::Log;
1288 # AnyEvent::Log overwrites this function
1289 goto &log;
1290}
1291
965my @models = ( 1292our @models = (
966 [EV:: => AnyEvent::Impl::EV::], 1293 [EV:: => AnyEvent::Impl::EV:: , 1],
967 [Event:: => AnyEvent::Impl::Event::], 1294 [AnyEvent::Loop:: => AnyEvent::Impl::Perl:: , 1],
968 [AnyEvent::Impl::Perl:: => AnyEvent::Impl::Perl::],
969 # everything below here will not be autoprobed 1295 # everything below here will not (normally) be autoprobed
970 # as the pureperl backend should work everywhere 1296 # as the pure perl backend should work everywhere
971 # and is usually faster 1297 # and is usually faster
1298 [Event:: => AnyEvent::Impl::Event::, 1],
1299 [Glib:: => AnyEvent::Impl::Glib:: , 1], # becomes extremely slow with many watchers
1300 [Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy
1301 [Irssi:: => AnyEvent::Impl::Irssi::], # Irssi has a bogus "Event" package
972 [Tk:: => AnyEvent::Impl::Tk::], # crashes with many handles 1302 [Tk:: => AnyEvent::Impl::Tk::], # crashes with many handles
973 [Glib:: => AnyEvent::Impl::Glib::], # becomes extremely slow with many watchers
974 [Event::Lib:: => AnyEvent::Impl::EventLib::], # too buggy
975 [Qt:: => AnyEvent::Impl::Qt::], # requires special main program 1303 [Qt:: => AnyEvent::Impl::Qt::], # requires special main program
976 [POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza 1304 [POE::Kernel:: => AnyEvent::Impl::POE::], # lasciate ogni speranza
977 [Wx:: => AnyEvent::Impl::POE::], 1305 [Wx:: => AnyEvent::Impl::POE::],
978 [Prima:: => AnyEvent::Impl::POE::], 1306 [Prima:: => AnyEvent::Impl::POE::],
1307 [IO::Async::Loop:: => AnyEvent::Impl::IOAsync::], # a bitch to autodetect
1308 [Cocoa::EventLoop:: => AnyEvent::Impl::Cocoa::],
1309 [FLTK:: => AnyEvent::Impl::FLTK2::],
979); 1310);
980 1311
981our %method = map +($_ => 1), 1312our @isa_hook;
1313
1314sub _isa_set {
1315 my @pkg = ("AnyEvent", (map $_->[0], grep defined, @isa_hook), $MODEL);
1316
1317 @{"$pkg[$_-1]::ISA"} = $pkg[$_]
1318 for 1 .. $#pkg;
1319
1320 grep $_ && $_->[1], @isa_hook
1321 and AE::_reset ();
1322}
1323
1324# used for hooking AnyEvent::Strict and AnyEvent::Debug::Wrap into the class hierarchy
1325sub _isa_hook($$;$) {
1326 my ($i, $pkg, $reset_ae) = @_;
1327
1328 $isa_hook[$i] = $pkg ? [$pkg, $reset_ae] : undef;
1329
1330 _isa_set;
1331}
1332
1333# all autoloaded methods reserve the complete glob, not just the method slot.
1334# due to bugs in perls method cache implementation.
982 qw(io timer time now now_update signal child idle condvar one_event DESTROY); 1335our @methods = qw(io timer time now now_update signal child idle condvar);
983 1336
984our @post_detect;
985
986sub post_detect(&) { 1337sub detect() {
987 my ($cb) = @_; 1338 return $MODEL if $MODEL; # some programs keep references to detect
988 1339
989 if ($MODEL) { 1340 local $!; # for good measure
990 $cb->(); 1341 local $SIG{__DIE__}; # we use eval
991 1342
992 1 1343 # free some memory
1344 *detect = sub () { $MODEL };
1345 # undef &func doesn't correctly update the method cache. grmbl.
1346 # so we delete the whole glob. grmbl.
1347 # otoh, perl doesn't let me undef an active usb, but it lets me free
1348 # a glob with an active sub. hrm. i hope it works, but perl is
1349 # usually buggy in this department. sigh.
1350 delete @{"AnyEvent::"}{@methods};
1351 undef @methods;
1352
1353 if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z0-9:]+)$/) {
1354 my $model = $1;
1355 $model = "AnyEvent::Impl::$model" unless $model =~ s/::$//;
1356 if (eval "require $model") {
1357 $MODEL = $model;
1358 AnyEvent::log 7 => "loaded model '$model' (forced by \$ENV{PERL_ANYEVENT_MODEL}), using it."
1359 if $VERBOSE >= 7;
993 } else { 1360 } else {
994 push @post_detect, $cb; 1361 AnyEvent::log warn => "unable to load model '$model' (from \$ENV{PERL_ANYEVENT_MODEL}):\n$@";
995 1362 }
996 defined wantarray
997 ? bless \$cb, "AnyEvent::Util::postdetect"
998 : ()
999 } 1363 }
1000}
1001 1364
1002sub AnyEvent::Util::postdetect::DESTROY { 1365 # check for already loaded models
1003 @post_detect = grep $_ != ${$_[0]}, @post_detect;
1004}
1005
1006sub detect() {
1007 unless ($MODEL) { 1366 unless ($MODEL) {
1008 no strict 'refs'; 1367 for (@REGISTRY, @models) {
1009 local $SIG{__DIE__}; 1368 my ($package, $model) = @$_;
1010 1369 if (${"$package\::VERSION"} > 0) {
1011 if ($ENV{PERL_ANYEVENT_MODEL} =~ /^([a-zA-Z]+)$/) {
1012 my $model = "AnyEvent::Impl::$1";
1013 if (eval "require $model") { 1370 if (eval "require $model") {
1014 $MODEL = $model; 1371 $MODEL = $model;
1015 warn "AnyEvent: loaded model '$model' (forced by \$PERL_ANYEVENT_MODEL), using it.\n" if $verbose > 1; 1372 AnyEvent::log 7 => "autodetected model '$model', using it."
1016 } else { 1373 if $VERBOSE >= 7;
1017 warn "AnyEvent: unable to load model '$model' (from \$PERL_ANYEVENT_MODEL):\n$@" if $verbose; 1374 last;
1375 }
1018 } 1376 }
1019 } 1377 }
1020 1378
1021 # check for already loaded models
1022 unless ($MODEL) { 1379 unless ($MODEL) {
1380 # try to autoload a model
1023 for (@REGISTRY, @models) { 1381 for (@REGISTRY, @models) {
1024 my ($package, $model) = @$_; 1382 my ($package, $model, $autoload) = @$_;
1383 if (
1384 $autoload
1385 and eval "require $package"
1025 if (${"$package\::VERSION"} > 0) { 1386 and ${"$package\::VERSION"} > 0
1026 if (eval "require $model") { 1387 and eval "require $model"
1388 ) {
1027 $MODEL = $model; 1389 $MODEL = $model;
1028 warn "AnyEvent: autodetected model '$model', using it.\n" if $verbose > 1; 1390 AnyEvent::log 7 => "autoloaded model '$model', using it."
1391 if $VERBOSE >= 7;
1029 last; 1392 last;
1030 }
1031 } 1393 }
1032 } 1394 }
1033 1395
1034 unless ($MODEL) {
1035 # try to load a model
1036
1037 for (@REGISTRY, @models) {
1038 my ($package, $model) = @$_;
1039 if (eval "require $package"
1040 and ${"$package\::VERSION"} > 0
1041 and eval "require $model") {
1042 $MODEL = $model;
1043 warn "AnyEvent: autoprobed model '$model', using it.\n" if $verbose > 1;
1044 last;
1045 }
1046 }
1047
1048 $MODEL 1396 $MODEL
1049 or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: EV, Event or Glib.\n"; 1397 or die "AnyEvent: backend autodetection failed - did you properly install AnyEvent?";
1050 }
1051 } 1398 }
1052
1053 push @{"$MODEL\::ISA"}, "AnyEvent::Base";
1054
1055 unshift @ISA, $MODEL;
1056
1057 require AnyEvent::Strict if $ENV{PERL_ANYEVENT_STRICT};
1058
1059 (shift @post_detect)->() while @post_detect;
1060 } 1399 }
1061 1400
1401 # free memory only needed for probing
1402 undef @models;
1403 undef @REGISTRY;
1404
1405 push @{"$MODEL\::ISA"}, "AnyEvent::Base";
1406
1407 # now nuke some methods that are overridden by the backend.
1408 # SUPER usage is not allowed in these.
1409 for (qw(time signal child idle)) {
1410 undef &{"AnyEvent::Base::$_"}
1411 if defined &{"$MODEL\::$_"};
1412 }
1413
1414 _isa_set;
1415
1416 # we're officially open!
1417
1418 if ($ENV{PERL_ANYEVENT_STRICT}) {
1419 require AnyEvent::Strict;
1420 }
1421
1422 if ($ENV{PERL_ANYEVENT_DEBUG_WRAP}) {
1423 require AnyEvent::Debug;
1424 AnyEvent::Debug::wrap ($ENV{PERL_ANYEVENT_DEBUG_WRAP});
1425 }
1426
1427 if (length $ENV{PERL_ANYEVENT_DEBUG_SHELL}) {
1428 require AnyEvent::Socket;
1429 require AnyEvent::Debug;
1430
1431 my $shell = $ENV{PERL_ANYEVENT_DEBUG_SHELL};
1432 $shell =~ s/\$\$/$$/g;
1433
1434 my ($host, $service) = AnyEvent::Socket::parse_hostport ($shell);
1435 $AnyEvent::Debug::SHELL = AnyEvent::Debug::shell ($host, $service);
1436 }
1437
1438 # now the anyevent environment is set up as the user told us to, so
1439 # call the actual user code - post detects
1440
1441 (shift @post_detect)->() while @post_detect;
1442 undef @post_detect;
1443
1444 *post_detect = sub(&) {
1445 shift->();
1446
1447 undef
1448 };
1449
1062 $MODEL 1450 $MODEL
1063} 1451}
1064 1452
1065sub AUTOLOAD { 1453for my $name (@methods) {
1066 (my $func = $AUTOLOAD) =~ s/.*://; 1454 *$name = sub {
1067 1455 detect;
1068 $method{$func} 1456 # we use goto because
1069 or croak "$func: not a valid method for AnyEvent objects"; 1457 # a) it makes the thunk more transparent
1070 1458 # b) it allows us to delete the thunk later
1071 detect unless $MODEL; 1459 goto &{ UNIVERSAL::can AnyEvent => "SUPER::$name" }
1072 1460 };
1073 my $class = shift;
1074 $class->$func (@_);
1075} 1461}
1076 1462
1077# utility function to dup a filehandle. this is used by many backends 1463# utility function to dup a filehandle. this is used by many backends
1078# to support binding more than one watcher per filehandle (they usually 1464# to support binding more than one watcher per filehandle (they usually
1079# allow only one watcher per fd, so we dup it to get a different one). 1465# allow only one watcher per fd, so we dup it to get a different one).
1080sub _dupfh($$$$) { 1466sub _dupfh($$;$$) {
1081 my ($poll, $fh, $r, $w) = @_; 1467 my ($poll, $fh, $r, $w) = @_;
1082 1468
1083 # cygwin requires the fh mode to be matching, unix doesn't 1469 # cygwin requires the fh mode to be matching, unix doesn't
1084 my ($rw, $mode) = $poll eq "r" ? ($r, "<") 1470 my ($rw, $mode) = $poll eq "r" ? ($r, "<&") : ($w, ">&");
1085 : $poll eq "w" ? ($w, ">")
1086 : Carp::croak "AnyEvent->io requires poll set to either 'r' or 'w'";
1087 1471
1088 open my $fh2, "$mode&" . fileno $fh 1472 open my $fh2, $mode, $fh
1089 or die "cannot dup() filehandle: $!,"; 1473 or die "AnyEvent->io: cannot dup() filehandle in mode '$poll': $!,";
1090 1474
1091 # we assume CLOEXEC is already set by perl in all important cases 1475 # we assume CLOEXEC is already set by perl in all important cases
1092 1476
1093 ($fh2, $rw) 1477 ($fh2, $rw)
1094} 1478}
1095 1479
1480=head1 SIMPLIFIED AE API
1481
1482Starting with version 5.0, AnyEvent officially supports a second, much
1483simpler, API that is designed to reduce the calling, typing and memory
1484overhead by using function call syntax and a fixed number of parameters.
1485
1486See the L<AE> manpage for details.
1487
1488=cut
1489
1490package AE;
1491
1492our $VERSION = $AnyEvent::VERSION;
1493
1494sub _reset() {
1495 eval q{
1496 # fall back to the main API by default - backends and AnyEvent::Base
1497 # implementations can overwrite these.
1498
1499 sub io($$$) {
1500 AnyEvent->io (fh => $_[0], poll => $_[1] ? "w" : "r", cb => $_[2])
1501 }
1502
1503 sub timer($$$) {
1504 AnyEvent->timer (after => $_[0], interval => $_[1], cb => $_[2])
1505 }
1506
1507 sub signal($$) {
1508 AnyEvent->signal (signal => $_[0], cb => $_[1])
1509 }
1510
1511 sub child($$) {
1512 AnyEvent->child (pid => $_[0], cb => $_[1])
1513 }
1514
1515 sub idle($) {
1516 AnyEvent->idle (cb => $_[0]);
1517 }
1518
1519 sub cv(;&) {
1520 AnyEvent->condvar (@_ ? (cb => $_[0]) : ())
1521 }
1522
1523 sub now() {
1524 AnyEvent->now
1525 }
1526
1527 sub now_update() {
1528 AnyEvent->now_update
1529 }
1530
1531 sub time() {
1532 AnyEvent->time
1533 }
1534
1535 *postpone = \&AnyEvent::postpone;
1536 *log = \&AnyEvent::log;
1537 };
1538 die if $@;
1539}
1540
1541BEGIN { _reset }
1542
1096package AnyEvent::Base; 1543package AnyEvent::Base;
1097 1544
1098# default implementations for many methods 1545# default implementations for many methods
1099 1546
1100BEGIN { 1547sub time {
1548 eval q{ # poor man's autoloading {}
1549 # probe for availability of Time::HiRes
1101 if (eval "use Time::HiRes (); Time::HiRes::time (); 1") { 1550 if (eval "use Time::HiRes (); Time::HiRes::time (); 1") {
1551 AnyEvent::log 8 => "AnyEvent: using Time::HiRes for sub-second timing accuracy."
1552 if $AnyEvent::VERBOSE >= 8;
1553 *time = sub { Time::HiRes::time () };
1102 *_time = \&Time::HiRes::time; 1554 *AE::time = \& Time::HiRes::time ;
1103 # if (eval "use POSIX (); (POSIX::times())... 1555 # if (eval "use POSIX (); (POSIX::times())...
1104 } else { 1556 } else {
1105 *_time = sub { time }; # epic fail 1557 AnyEvent::log critical => "using built-in time(), WARNING, no sub-second resolution!";
1558 *time = sub { CORE::time };
1559 *AE::time = sub (){ CORE::time };
1560 }
1561
1562 *now = \&time;
1563 };
1564 die if $@;
1565
1566 &time
1567}
1568
1569*now = \&time;
1570sub now_update { }
1571
1572sub _poll {
1573 Carp::croak "$AnyEvent::MODEL does not support blocking waits. Caught";
1574}
1575
1576# default implementation for ->condvar
1577# in fact, the default should not be overwritten
1578
1579sub condvar {
1580 eval q{ # poor man's autoloading {}
1581 *condvar = sub {
1582 bless { @_ == 3 ? (_ae_cb => $_[2]) : () }, "AnyEvent::CondVar"
1583 };
1584
1585 *AE::cv = sub (;&) {
1586 bless { @_ ? (_ae_cb => shift) : () }, "AnyEvent::CondVar"
1587 };
1588 };
1589 die if $@;
1590
1591 &condvar
1592}
1593
1594# default implementation for ->signal
1595
1596our $HAVE_ASYNC_INTERRUPT;
1597
1598sub _have_async_interrupt() {
1599 $HAVE_ASYNC_INTERRUPT = 1*(!$ENV{PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT}
1600 && eval "use Async::Interrupt 1.02 (); 1")
1601 unless defined $HAVE_ASYNC_INTERRUPT;
1602
1603 $HAVE_ASYNC_INTERRUPT
1604}
1605
1606our ($SIGPIPE_R, $SIGPIPE_W, %SIG_CB, %SIG_EV, $SIG_IO);
1607our (%SIG_ASY, %SIG_ASY_W);
1608our ($SIG_COUNT, $SIG_TW);
1609
1610# install a dummy wakeup watcher to reduce signal catching latency
1611# used by Impls
1612sub _sig_add() {
1613 unless ($SIG_COUNT++) {
1614 # try to align timer on a full-second boundary, if possible
1615 my $NOW = AE::now;
1616
1617 $SIG_TW = AE::timer
1618 $MAX_SIGNAL_LATENCY - ($NOW - int $NOW),
1619 $MAX_SIGNAL_LATENCY,
1620 sub { } # just for the PERL_ASYNC_CHECK
1621 ;
1106 } 1622 }
1107} 1623}
1108 1624
1109sub time { _time } 1625sub _sig_del {
1110sub now { _time } 1626 undef $SIG_TW
1111sub now_update { } 1627 unless --$SIG_COUNT;
1112
1113# default implementation for ->condvar
1114
1115sub condvar {
1116 bless { @_ == 3 ? (_ae_cb => $_[2]) : () }, "AnyEvent::CondVar"
1117} 1628}
1118 1629
1119# default implementation for ->signal 1630our $_sig_name_init; $_sig_name_init = sub {
1631 eval q{ # poor man's autoloading {}
1632 undef $_sig_name_init;
1120 1633
1121our ($SIGPIPE_R, $SIGPIPE_W, %SIG_CB, %SIG_EV, $SIG_IO); 1634 if (_have_async_interrupt) {
1635 *sig2num = \&Async::Interrupt::sig2num;
1636 *sig2name = \&Async::Interrupt::sig2name;
1637 } else {
1638 require Config;
1122 1639
1123sub _signal_exec { 1640 my %signame2num;
1124 sysread $SIGPIPE_R, my $dummy, 4; 1641 @signame2num{ split ' ', $Config::Config{sig_name} }
1642 = split ' ', $Config::Config{sig_num};
1125 1643
1126 while (%SIG_EV) { 1644 my @signum2name;
1127 for (keys %SIG_EV) { 1645 @signum2name[values %signame2num] = keys %signame2num;
1128 delete $SIG_EV{$_}; 1646
1129 $_->() for values %{ $SIG_CB{$_} || {} }; 1647 *sig2num = sub($) {
1648 $_[0] > 0 ? shift : $signame2num{+shift}
1649 };
1650 *sig2name = sub ($) {
1651 $_[0] > 0 ? $signum2name[+shift] : shift
1652 };
1130 } 1653 }
1131 } 1654 };
1132} 1655 die if $@;
1656};
1657
1658sub sig2num ($) { &$_sig_name_init; &sig2num }
1659sub sig2name($) { &$_sig_name_init; &sig2name }
1133 1660
1134sub signal { 1661sub signal {
1135 my (undef, %arg) = @_; 1662 eval q{ # poor man's autoloading {}
1663 # probe for availability of Async::Interrupt
1664 if (_have_async_interrupt) {
1665 AnyEvent::log 8 => "using Async::Interrupt for race-free signal handling."
1666 if $AnyEvent::VERBOSE >= 8;
1136 1667
1137 unless ($SIGPIPE_R) { 1668 $SIGPIPE_R = new Async::Interrupt::EventPipe;
1138 require Fcntl; 1669 $SIG_IO = AE::io $SIGPIPE_R->fileno, 0, \&_signal_exec;
1139 1670
1140 if (AnyEvent::WIN32) {
1141 require AnyEvent::Util;
1142
1143 ($SIGPIPE_R, $SIGPIPE_W) = AnyEvent::Util::portable_pipe ();
1144 AnyEvent::Util::fh_nonblocking ($SIGPIPE_R) if $SIGPIPE_R;
1145 AnyEvent::Util::fh_nonblocking ($SIGPIPE_W) if $SIGPIPE_W; # just in case
1146 } else { 1671 } else {
1672 AnyEvent::log 8 => "using emulated perl signal handling with latency timer."
1673 if $AnyEvent::VERBOSE >= 8;
1674
1675 if (AnyEvent::WIN32) {
1676 require AnyEvent::Util;
1677
1678 ($SIGPIPE_R, $SIGPIPE_W) = AnyEvent::Util::portable_pipe ();
1679 AnyEvent::Util::fh_nonblocking ($SIGPIPE_R, 1) if $SIGPIPE_R;
1680 AnyEvent::Util::fh_nonblocking ($SIGPIPE_W, 1) if $SIGPIPE_W; # just in case
1681 } else {
1147 pipe $SIGPIPE_R, $SIGPIPE_W; 1682 pipe $SIGPIPE_R, $SIGPIPE_W;
1148 fcntl $SIGPIPE_R, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_R; 1683 fcntl $SIGPIPE_R, AnyEvent::F_SETFL, AnyEvent::O_NONBLOCK if $SIGPIPE_R;
1149 fcntl $SIGPIPE_W, &Fcntl::F_SETFL, &Fcntl::O_NONBLOCK if $SIGPIPE_W; # just in case 1684 fcntl $SIGPIPE_W, AnyEvent::F_SETFL, AnyEvent::O_NONBLOCK if $SIGPIPE_W; # just in case
1150 1685
1151 # not strictly required, as $^F is normally 2, but let's make sure... 1686 # not strictly required, as $^F is normally 2, but let's make sure...
1152 fcntl $SIGPIPE_R, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC; 1687 fcntl $SIGPIPE_R, AnyEvent::F_SETFD, AnyEvent::FD_CLOEXEC;
1153 fcntl $SIGPIPE_W, &Fcntl::F_SETFD, &Fcntl::FD_CLOEXEC; 1688 fcntl $SIGPIPE_W, AnyEvent::F_SETFD, AnyEvent::FD_CLOEXEC;
1689 }
1690
1691 $SIGPIPE_R
1692 or Carp::croak "AnyEvent: unable to create a signal reporting pipe: $!\n";
1693
1694 $SIG_IO = AE::io $SIGPIPE_R, 0, \&_signal_exec;
1154 } 1695 }
1155 1696
1156 $SIGPIPE_R 1697 *signal = $HAVE_ASYNC_INTERRUPT
1157 or Carp::croak "AnyEvent: unable to create a signal reporting pipe: $!\n"; 1698 ? sub {
1699 my (undef, %arg) = @_;
1158 1700
1159 $SIG_IO = AnyEvent->io (fh => $SIGPIPE_R, poll => "r", cb => \&_signal_exec); 1701 # async::interrupt
1160 }
1161
1162 my $signal = uc $arg{signal} 1702 my $signal = sig2num $arg{signal};
1163 or Carp::croak "required option 'signal' is missing";
1164
1165 $SIG_CB{$signal}{$arg{cb}} = $arg{cb}; 1703 $SIG_CB{$signal}{$arg{cb}} = $arg{cb};
1704
1705 $SIG_ASY{$signal} ||= new Async::Interrupt
1706 cb => sub { undef $SIG_EV{$signal} },
1707 signal => $signal,
1708 pipe => [$SIGPIPE_R->filenos],
1709 pipe_autodrain => 0,
1710 ;
1711
1712 bless [$signal, $arg{cb}], "AnyEvent::Base::signal"
1713 }
1714 : sub {
1715 my (undef, %arg) = @_;
1716
1717 # pure perl
1718 my $signal = sig2name $arg{signal};
1719 $SIG_CB{$signal}{$arg{cb}} = $arg{cb};
1720
1166 $SIG{$signal} ||= sub { 1721 $SIG{$signal} ||= sub {
1167 local $!; 1722 local $!;
1168 syswrite $SIGPIPE_W, "\x00", 1 unless %SIG_EV; 1723 syswrite $SIGPIPE_W, "\x00", 1 unless %SIG_EV;
1169 undef $SIG_EV{$signal}; 1724 undef $SIG_EV{$signal};
1725 };
1726
1727 # can't do signal processing without introducing races in pure perl,
1728 # so limit the signal latency.
1729 _sig_add;
1730
1731 bless [$signal, $arg{cb}], "AnyEvent::Base::signal"
1732 }
1733 ;
1734
1735 *AnyEvent::Base::signal::DESTROY = sub {
1736 my ($signal, $cb) = @{$_[0]};
1737
1738 _sig_del;
1739
1740 delete $SIG_CB{$signal}{$cb};
1741
1742 $HAVE_ASYNC_INTERRUPT
1743 ? delete $SIG_ASY{$signal}
1744 : # delete doesn't work with older perls - they then
1745 # print weird messages, or just unconditionally exit
1746 # instead of getting the default action.
1747 undef $SIG{$signal}
1748 unless keys %{ $SIG_CB{$signal} };
1749 };
1750
1751 *_signal_exec = sub {
1752 $HAVE_ASYNC_INTERRUPT
1753 ? $SIGPIPE_R->drain
1754 : sysread $SIGPIPE_R, (my $dummy), 9;
1755
1756 while (%SIG_EV) {
1757 for (keys %SIG_EV) {
1758 delete $SIG_EV{$_};
1759 &$_ for values %{ $SIG_CB{$_} || {} };
1760 }
1761 }
1762 };
1170 }; 1763 };
1764 die if $@;
1171 1765
1172 bless [$signal, $arg{cb}], "AnyEvent::Base::signal" 1766 &signal
1173}
1174
1175sub AnyEvent::Base::signal::DESTROY {
1176 my ($signal, $cb) = @{$_[0]};
1177
1178 delete $SIG_CB{$signal}{$cb};
1179
1180 # delete doesn't work with older perls - they then
1181 # print weird messages, or just unconditionally exit
1182 # instead of getting the default action.
1183 undef $SIG{$signal} unless keys %{ $SIG_CB{$signal} };
1184} 1767}
1185 1768
1186# default implementation for ->child 1769# default implementation for ->child
1187 1770
1188our %PID_CB; 1771our %PID_CB;
1189our $CHLD_W; 1772our $CHLD_W;
1190our $CHLD_DELAY_W; 1773our $CHLD_DELAY_W;
1191our $WNOHANG;
1192 1774
1193sub _sigchld { 1775# used by many Impl's
1194 while (0 < (my $pid = waitpid -1, $WNOHANG)) { 1776sub _emit_childstatus($$) {
1777 my (undef, $rpid, $rstatus) = @_;
1778
1779 $_->($rpid, $rstatus)
1195 $_->($pid, $?) for (values %{ $PID_CB{$pid} || {} }), 1780 for values %{ $PID_CB{$rpid} || {} },
1196 (values %{ $PID_CB{0} || {} }); 1781 values %{ $PID_CB{0} || {} };
1197 }
1198} 1782}
1199 1783
1200sub child { 1784sub child {
1785 eval q{ # poor man's autoloading {}
1786 *_sigchld = sub {
1787 my $pid;
1788
1789 AnyEvent->_emit_childstatus ($pid, $?)
1790 while ($pid = waitpid -1, WNOHANG) > 0;
1791 };
1792
1793 *child = sub {
1201 my (undef, %arg) = @_; 1794 my (undef, %arg) = @_;
1202 1795
1203 defined (my $pid = $arg{pid} + 0) 1796 my $pid = $arg{pid};
1204 or Carp::croak "required option 'pid' is missing"; 1797 my $cb = $arg{cb};
1205 1798
1206 $PID_CB{$pid}{$arg{cb}} = $arg{cb}; 1799 $PID_CB{$pid}{$cb+0} = $cb;
1207 1800
1208 $WNOHANG ||= eval { local $SIG{__DIE__}; require POSIX; &POSIX::WNOHANG } || 1;
1209
1210 unless ($CHLD_W) { 1801 unless ($CHLD_W) {
1211 $CHLD_W = AnyEvent->signal (signal => 'CHLD', cb => \&_sigchld); 1802 $CHLD_W = AE::signal CHLD => \&_sigchld;
1212 # child could be a zombie already, so make at least one round 1803 # child could be a zombie already, so make at least one round
1213 &_sigchld; 1804 &_sigchld;
1214 } 1805 }
1215 1806
1216 bless [$pid, $arg{cb}], "AnyEvent::Base::child" 1807 bless [$pid, $cb+0], "AnyEvent::Base::child"
1217} 1808 };
1218 1809
1219sub AnyEvent::Base::child::DESTROY { 1810 *AnyEvent::Base::child::DESTROY = sub {
1220 my ($pid, $cb) = @{$_[0]}; 1811 my ($pid, $icb) = @{$_[0]};
1221 1812
1222 delete $PID_CB{$pid}{$cb}; 1813 delete $PID_CB{$pid}{$icb};
1223 delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} }; 1814 delete $PID_CB{$pid} unless keys %{ $PID_CB{$pid} };
1224 1815
1225 undef $CHLD_W unless keys %PID_CB; 1816 undef $CHLD_W unless keys %PID_CB;
1817 };
1818 };
1819 die if $@;
1820
1821 &child
1226} 1822}
1227 1823
1228# idle emulation is done by simply using a timer, regardless 1824# idle emulation is done by simply using a timer, regardless
1229# of whether the process is idle or not, and not letting 1825# of whether the process is idle or not, and not letting
1230# the callback use more than 50% of the time. 1826# the callback use more than 50% of the time.
1231sub idle { 1827sub idle {
1828 eval q{ # poor man's autoloading {}
1829 *idle = sub {
1232 my (undef, %arg) = @_; 1830 my (undef, %arg) = @_;
1233 1831
1234 my ($cb, $w, $rcb) = $arg{cb}; 1832 my ($cb, $w, $rcb) = $arg{cb};
1235 1833
1236 $rcb = sub { 1834 $rcb = sub {
1237 if ($cb) { 1835 if ($cb) {
1238 $w = _time; 1836 $w = AE::time;
1239 &$cb; 1837 &$cb;
1240 $w = _time - $w; 1838 $w = AE::time - $w;
1241 1839
1242 # never use more then 50% of the time for the idle watcher, 1840 # never use more then 50% of the time for the idle watcher,
1243 # within some limits 1841 # within some limits
1244 $w = 0.0001 if $w < 0.0001; 1842 $w = 0.0001 if $w < 0.0001;
1245 $w = 5 if $w > 5; 1843 $w = 5 if $w > 5;
1246 1844
1247 $w = AnyEvent->timer (after => $w, cb => $rcb); 1845 $w = AE::timer $w, 0, $rcb;
1248 } else { 1846 } else {
1249 # clean up... 1847 # clean up...
1250 undef $w; 1848 undef $w;
1251 undef $rcb; 1849 undef $rcb;
1850 }
1851 };
1852
1853 $w = AE::timer 0.05, 0, $rcb;
1854
1855 bless \\$cb, "AnyEvent::Base::idle"
1252 } 1856 };
1857
1858 *AnyEvent::Base::idle::DESTROY = sub {
1859 undef $${$_[0]};
1860 };
1253 }; 1861 };
1862 die if $@;
1254 1863
1255 $w = AnyEvent->timer (after => 0.05, cb => $rcb); 1864 &idle
1256
1257 bless \\$cb, "AnyEvent::Base::idle"
1258}
1259
1260sub AnyEvent::Base::idle::DESTROY {
1261 undef $${$_[0]};
1262} 1865}
1263 1866
1264package AnyEvent::CondVar; 1867package AnyEvent::CondVar;
1265 1868
1266our @ISA = AnyEvent::CondVar::Base::; 1869our @ISA = AnyEvent::CondVar::Base::;
1267 1870
1871# only to be used for subclassing
1872sub new {
1873 my $class = shift;
1874 bless AnyEvent->condvar (@_), $class
1875}
1876
1268package AnyEvent::CondVar::Base; 1877package AnyEvent::CondVar::Base;
1269 1878
1270use overload 1879#use overload
1271 '&{}' => sub { my $self = shift; sub { $self->send (@_) } }, 1880# '&{}' => sub { my $self = shift; sub { $self->send (@_) } },
1272 fallback => 1; 1881# fallback => 1;
1882
1883# save 300+ kilobytes by dirtily hardcoding overloading
1884${"AnyEvent::CondVar::Base::OVERLOAD"}{dummy}++; # Register with magic by touching.
1885*{'AnyEvent::CondVar::Base::()'} = sub { }; # "Make it findable via fetchmethod."
1886*{'AnyEvent::CondVar::Base::(&{}'} = sub { my $self = shift; sub { $self->send (@_) } }; # &{}
1887${'AnyEvent::CondVar::Base::()'} = 1; # fallback
1888
1889our $WAITING;
1273 1890
1274sub _send { 1891sub _send {
1275 # nop 1892 # nop
1893}
1894
1895sub _wait {
1896 AnyEvent->_poll until $_[0]{_ae_sent};
1276} 1897}
1277 1898
1278sub send { 1899sub send {
1279 my $cv = shift; 1900 my $cv = shift;
1280 $cv->{_ae_sent} = [@_]; 1901 $cv->{_ae_sent} = [@_];
1289 1910
1290sub ready { 1911sub ready {
1291 $_[0]{_ae_sent} 1912 $_[0]{_ae_sent}
1292} 1913}
1293 1914
1294sub _wait {
1295 AnyEvent->one_event while !$_[0]{_ae_sent};
1296}
1297
1298sub recv { 1915sub recv {
1916 unless ($_[0]{_ae_sent}) {
1917 $WAITING
1918 and Carp::croak "AnyEvent::CondVar: recursive blocking wait attempted";
1919
1920 local $WAITING = 1;
1299 $_[0]->_wait; 1921 $_[0]->_wait;
1922 }
1300 1923
1301 Carp::croak $_[0]{_ae_croak} if $_[0]{_ae_croak}; 1924 $_[0]{_ae_croak}
1302 wantarray ? @{ $_[0]{_ae_sent} } : $_[0]{_ae_sent}[0] 1925 and Carp::croak $_[0]{_ae_croak};
1926
1927 wantarray
1928 ? @{ $_[0]{_ae_sent} }
1929 : $_[0]{_ae_sent}[0]
1303} 1930}
1304 1931
1305sub cb { 1932sub cb {
1306 $_[0]{_ae_cb} = $_[1] if @_ > 1; 1933 my $cv = shift;
1934
1935 @_
1936 and $cv->{_ae_cb} = shift
1937 and $cv->{_ae_sent}
1938 and (delete $cv->{_ae_cb})->($cv);
1939
1307 $_[0]{_ae_cb} 1940 $cv->{_ae_cb}
1308} 1941}
1309 1942
1310sub begin { 1943sub begin {
1311 ++$_[0]{_ae_counter}; 1944 ++$_[0]{_ae_counter};
1312 $_[0]{_ae_end_cb} = $_[1] if @_ > 1; 1945 $_[0]{_ae_end_cb} = $_[1] if @_ > 1;
1317 &{ $_[0]{_ae_end_cb} || sub { $_[0]->send } }; 1950 &{ $_[0]{_ae_end_cb} || sub { $_[0]->send } };
1318} 1951}
1319 1952
1320# undocumented/compatibility with pre-3.4 1953# undocumented/compatibility with pre-3.4
1321*broadcast = \&send; 1954*broadcast = \&send;
1322*wait = \&_wait; 1955*wait = \&recv;
1323 1956
1324=head1 ERROR AND EXCEPTION HANDLING 1957=head1 ERROR AND EXCEPTION HANDLING
1325 1958
1326In general, AnyEvent does not do any error handling - it relies on the 1959In general, AnyEvent does not do any error handling - it relies on the
1327caller to do that if required. The L<AnyEvent::Strict> module (see also 1960caller to do that if required. The L<AnyEvent::Strict> module (see also
1354 1987
1355By default, AnyEvent will be completely silent except in fatal 1988By default, AnyEvent will be completely silent except in fatal
1356conditions. You can set this environment variable to make AnyEvent more 1989conditions. You can set this environment variable to make AnyEvent more
1357talkative. 1990talkative.
1358 1991
1359When set to C<1> or higher, causes AnyEvent to warn about unexpected 1992When set to C<5> or higher, causes AnyEvent to warn about unexpected
1360conditions, such as not being able to load the event model specified by 1993conditions, such as not being able to load the event model specified by
1361C<PERL_ANYEVENT_MODEL>. 1994C<PERL_ANYEVENT_MODEL>.
1362 1995
1363When set to C<2> or higher, cause AnyEvent to report to STDERR which event 1996When set to C<7> or higher, cause AnyEvent to report to STDERR which event
1364model it chooses. 1997model it chooses.
1998
1999When set to C<8> or higher, then AnyEvent will report extra information on
2000which optional modules it loads and how it implements certain features.
1365 2001
1366=item C<PERL_ANYEVENT_STRICT> 2002=item C<PERL_ANYEVENT_STRICT>
1367 2003
1368AnyEvent does not do much argument checking by default, as thorough 2004AnyEvent does not do much argument checking by default, as thorough
1369argument checking is very costly. Setting this variable to a true value 2005argument checking is very costly. Setting this variable to a true value
1370will cause AnyEvent to load C<AnyEvent::Strict> and then to thoroughly 2006will cause AnyEvent to load C<AnyEvent::Strict> and then to thoroughly
1371check the arguments passed to most method calls. If it finds any problems 2007check the arguments passed to most method calls. If it finds any problems,
1372it will croak. 2008it will croak.
1373 2009
1374In other words, enables "strict" mode. 2010In other words, enables "strict" mode.
1375 2011
1376Unlike C<use strict>, it is definitely recommended ot keep it off in 2012Unlike C<use strict> (or its modern cousin, C<< use L<common::sense>
1377production. Keeping C<PERL_ANYEVENT_STRICT=1> in your environment while 2013>>, it is definitely recommended to keep it off in production. Keeping
1378developing programs can be very useful, however. 2014C<PERL_ANYEVENT_STRICT=1> in your environment while developing programs
2015can be very useful, however.
2016
2017=item C<PERL_ANYEVENT_DEBUG_SHELL>
2018
2019If this env variable is set, then its contents will be interpreted by
2020C<AnyEvent::Socket::parse_hostport> (after replacing every occurance of
2021C<$$> by the process pid) and an C<AnyEvent::Debug::shell> is bound on
2022that port. The shell object is saved in C<$AnyEvent::Debug::SHELL>.
2023
2024This takes place when the first watcher is created.
2025
2026For example, to bind a debug shell on a unix domain socket in
2027F<< /tmp/debug<pid>.sock >>, you could use this:
2028
2029 PERL_ANYEVENT_DEBUG_SHELL=/tmp/debug\$\$.sock perlprog
2030
2031Note that creating sockets in F</tmp> is very unsafe on multiuser
2032systems.
2033
2034=item C<PERL_ANYEVENT_DEBUG_WRAP>
2035
2036Can be set to C<0>, C<1> or C<2> and enables wrapping of all watchers for
2037debugging purposes. See C<AnyEvent::Debug::wrap> for details.
1379 2038
1380=item C<PERL_ANYEVENT_MODEL> 2039=item C<PERL_ANYEVENT_MODEL>
1381 2040
1382This can be used to specify the event model to be used by AnyEvent, before 2041This can be used to specify the event model to be used by AnyEvent, before
1383auto detection and -probing kicks in. It must be a string consisting 2042auto detection and -probing kicks in.
1384entirely of ASCII letters. The string C<AnyEvent::Impl::> gets prepended 2043
2044It normally is a string consisting entirely of ASCII letters (e.g. C<EV>
2045or C<IOAsync>). The string C<AnyEvent::Impl::> gets prepended and the
1385and the resulting module name is loaded and if the load was successful, 2046resulting module name is loaded and - if the load was successful - used as
1386used as event model. If it fails to load AnyEvent will proceed with 2047event model backend. If it fails to load then AnyEvent will proceed with
1387auto detection and -probing. 2048auto detection and -probing.
1388 2049
1389This functionality might change in future versions. 2050If the string ends with C<::> instead (e.g. C<AnyEvent::Impl::EV::>) then
2051nothing gets prepended and the module name is used as-is (hint: C<::> at
2052the end of a string designates a module name and quotes it appropriately).
1390 2053
1391For example, to force the pure perl model (L<AnyEvent::Impl::Perl>) you 2054For example, to force the pure perl model (L<AnyEvent::Loop::Perl>) you
1392could start your program like this: 2055could start your program like this:
1393 2056
1394 PERL_ANYEVENT_MODEL=Perl perl ... 2057 PERL_ANYEVENT_MODEL=Perl perl ...
1395 2058
1396=item C<PERL_ANYEVENT_PROTOCOLS> 2059=item C<PERL_ANYEVENT_PROTOCOLS>
1426 2089
1427=item C<PERL_ANYEVENT_MAX_FORKS> 2090=item C<PERL_ANYEVENT_MAX_FORKS>
1428 2091
1429The maximum number of child processes that C<AnyEvent::Util::fork_call> 2092The maximum number of child processes that C<AnyEvent::Util::fork_call>
1430will create in parallel. 2093will create in parallel.
2094
2095=item C<PERL_ANYEVENT_MAX_OUTSTANDING_DNS>
2096
2097The default value for the C<max_outstanding> parameter for the default DNS
2098resolver - this is the maximum number of parallel DNS requests that are
2099sent to the DNS server.
2100
2101=item C<PERL_ANYEVENT_RESOLV_CONF>
2102
2103The file to use instead of F</etc/resolv.conf> (or OS-specific
2104configuration) in the default resolver. When set to the empty string, no
2105default config will be used.
2106
2107=item C<PERL_ANYEVENT_CA_FILE>, C<PERL_ANYEVENT_CA_PATH>.
2108
2109When neither C<ca_file> nor C<ca_path> was specified during
2110L<AnyEvent::TLS> context creation, and either of these environment
2111variables exist, they will be used to specify CA certificate locations
2112instead of a system-dependent default.
2113
2114=item C<PERL_ANYEVENT_AVOID_GUARD> and C<PERL_ANYEVENT_AVOID_ASYNC_INTERRUPT>
2115
2116When these are set to C<1>, then the respective modules are not
2117loaded. Mostly good for testing AnyEvent itself.
1431 2118
1432=back 2119=back
1433 2120
1434=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE 2121=head1 SUPPLYING YOUR OWN EVENT MODEL INTERFACE
1435 2122
1493 warn "read: $input\n"; # output what has been read 2180 warn "read: $input\n"; # output what has been read
1494 $cv->send if $input =~ /^q/i; # quit program if /^q/i 2181 $cv->send if $input =~ /^q/i; # quit program if /^q/i
1495 }, 2182 },
1496 ); 2183 );
1497 2184
1498 my $time_watcher; # can only be used once
1499
1500 sub new_timer {
1501 $timer = AnyEvent->timer (after => 1, cb => sub { 2185 my $time_watcher = AnyEvent->timer (after => 1, interval => 1, cb => sub {
1502 warn "timeout\n"; # print 'timeout' about every second 2186 warn "timeout\n"; # print 'timeout' at most every second
1503 &new_timer; # and restart the time
1504 }); 2187 });
1505 }
1506
1507 new_timer; # create first timer
1508 2188
1509 $cv->recv; # wait until user enters /^q/i 2189 $cv->recv; # wait until user enters /^q/i
1510 2190
1511=head1 REAL-WORLD EXAMPLE 2191=head1 REAL-WORLD EXAMPLE
1512 2192
1585 2265
1586The actual code goes further and collects all errors (C<die>s, exceptions) 2266The actual code goes further and collects all errors (C<die>s, exceptions)
1587that occurred during request processing. The C<result> method detects 2267that occurred during request processing. The C<result> method detects
1588whether an exception as thrown (it is stored inside the $txn object) 2268whether an exception as thrown (it is stored inside the $txn object)
1589and just throws the exception, which means connection errors and other 2269and just throws the exception, which means connection errors and other
1590problems get reported tot he code that tries to use the result, not in a 2270problems get reported to the code that tries to use the result, not in a
1591random callback. 2271random callback.
1592 2272
1593All of this enables the following usage styles: 2273All of this enables the following usage styles:
1594 2274
15951. Blocking: 22751. Blocking:
1643through AnyEvent. The benchmark creates a lot of timers (with a zero 2323through AnyEvent. The benchmark creates a lot of timers (with a zero
1644timeout) and I/O watchers (watching STDOUT, a pty, to become writable, 2324timeout) and I/O watchers (watching STDOUT, a pty, to become writable,
1645which it is), lets them fire exactly once and destroys them again. 2325which it is), lets them fire exactly once and destroys them again.
1646 2326
1647Source code for this benchmark is found as F<eg/bench> in the AnyEvent 2327Source code for this benchmark is found as F<eg/bench> in the AnyEvent
1648distribution. 2328distribution. It uses the L<AE> interface, which makes a real difference
2329for the EV and Perl backends only.
1649 2330
1650=head3 Explanation of the columns 2331=head3 Explanation of the columns
1651 2332
1652I<watcher> is the number of event watchers created/destroyed. Since 2333I<watcher> is the number of event watchers created/destroyed. Since
1653different event models feature vastly different performances, each event 2334different event models feature vastly different performances, each event
1674watcher. 2355watcher.
1675 2356
1676=head3 Results 2357=head3 Results
1677 2358
1678 name watchers bytes create invoke destroy comment 2359 name watchers bytes create invoke destroy comment
1679 EV/EV 400000 224 0.47 0.35 0.27 EV native interface 2360 EV/EV 100000 223 0.47 0.43 0.27 EV native interface
1680 EV/Any 100000 224 2.88 0.34 0.27 EV + AnyEvent watchers 2361 EV/Any 100000 223 0.48 0.42 0.26 EV + AnyEvent watchers
1681 CoroEV/Any 100000 224 2.85 0.35 0.28 coroutines + Coro::Signal 2362 Coro::EV/Any 100000 223 0.47 0.42 0.26 coroutines + Coro::Signal
1682 Perl/Any 100000 452 4.13 0.73 0.95 pure perl implementation 2363 Perl/Any 100000 431 2.70 0.74 0.92 pure perl implementation
1683 Event/Event 16000 517 32.20 31.80 0.81 Event native interface 2364 Event/Event 16000 516 31.16 31.84 0.82 Event native interface
1684 Event/Any 16000 590 35.85 31.55 1.06 Event + AnyEvent watchers 2365 Event/Any 16000 1203 42.61 34.79 1.80 Event + AnyEvent watchers
2366 IOAsync/Any 16000 1911 41.92 27.45 16.81 via IO::Async::Loop::IO_Poll
2367 IOAsync/Any 16000 1726 40.69 26.37 15.25 via IO::Async::Loop::Epoll
1685 Glib/Any 16000 1357 102.33 12.31 51.00 quadratic behaviour 2368 Glib/Any 16000 1118 89.00 12.57 51.17 quadratic behaviour
1686 Tk/Any 2000 1860 27.20 66.31 14.00 SEGV with >> 2000 watchers 2369 Tk/Any 2000 1346 20.96 10.75 8.00 SEGV with >> 2000 watchers
1687 POE/Event 2000 6328 109.99 751.67 14.02 via POE::Loop::Event 2370 POE/Any 2000 6951 108.97 795.32 14.24 via POE::Loop::Event
1688 POE/Select 2000 6027 94.54 809.13 579.80 via POE::Loop::Select 2371 POE/Any 2000 6648 94.79 774.40 575.51 via POE::Loop::Select
1689 2372
1690=head3 Discussion 2373=head3 Discussion
1691 2374
1692The benchmark does I<not> measure scalability of the event loop very 2375The benchmark does I<not> measure scalability of the event loop very
1693well. For example, a select-based event loop (such as the pure perl one) 2376well. For example, a select-based event loop (such as the pure perl one)
1705benchmark machine, handling an event takes roughly 1600 CPU cycles with 2388benchmark machine, handling an event takes roughly 1600 CPU cycles with
1706EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU 2389EV, 3100 CPU cycles with AnyEvent's pure perl loop and almost 3000000 CPU
1707cycles with POE. 2390cycles with POE.
1708 2391
1709C<EV> is the sole leader regarding speed and memory use, which are both 2392C<EV> is the sole leader regarding speed and memory use, which are both
1710maximal/minimal, respectively. Even when going through AnyEvent, it uses 2393maximal/minimal, respectively. When using the L<AE> API there is zero
2394overhead (when going through the AnyEvent API create is about 5-6 times
2395slower, with other times being equal, so still uses far less memory than
1711far less memory than any other event loop and is still faster than Event 2396any other event loop and is still faster than Event natively).
1712natively.
1713 2397
1714The pure perl implementation is hit in a few sweet spots (both the 2398The pure perl implementation is hit in a few sweet spots (both the
1715constant timeout and the use of a single fd hit optimisations in the perl 2399constant timeout and the use of a single fd hit optimisations in the perl
1716interpreter and the backend itself). Nevertheless this shows that it 2400interpreter and the backend itself). Nevertheless this shows that it
1717adds very little overhead in itself. Like any select-based backend its 2401adds very little overhead in itself. Like any select-based backend its
1718performance becomes really bad with lots of file descriptors (and few of 2402performance becomes really bad with lots of file descriptors (and few of
1719them active), of course, but this was not subject of this benchmark. 2403them active), of course, but this was not subject of this benchmark.
1720 2404
1721The C<Event> module has a relatively high setup and callback invocation 2405The C<Event> module has a relatively high setup and callback invocation
1722cost, but overall scores in on the third place. 2406cost, but overall scores in on the third place.
2407
2408C<IO::Async> performs admirably well, about on par with C<Event>, even
2409when using its pure perl backend.
1723 2410
1724C<Glib>'s memory usage is quite a bit higher, but it features a 2411C<Glib>'s memory usage is quite a bit higher, but it features a
1725faster callback invocation and overall ends up in the same class as 2412faster callback invocation and overall ends up in the same class as
1726C<Event>. However, Glib scales extremely badly, doubling the number of 2413C<Event>. However, Glib scales extremely badly, doubling the number of
1727watchers increases the processing time by more than a factor of four, 2414watchers increases the processing time by more than a factor of four,
1762(even when used without AnyEvent), but most event loops have acceptable 2449(even when used without AnyEvent), but most event loops have acceptable
1763performance with or without AnyEvent. 2450performance with or without AnyEvent.
1764 2451
1765=item * The overhead AnyEvent adds is usually much smaller than the overhead of 2452=item * The overhead AnyEvent adds is usually much smaller than the overhead of
1766the actual event loop, only with extremely fast event loops such as EV 2453the actual event loop, only with extremely fast event loops such as EV
1767adds AnyEvent significant overhead. 2454does AnyEvent add significant overhead.
1768 2455
1769=item * You should avoid POE like the plague if you want performance or 2456=item * You should avoid POE like the plague if you want performance or
1770reasonable memory usage. 2457reasonable memory usage.
1771 2458
1772=back 2459=back
1788In this benchmark, we use 10000 socket pairs (20000 sockets), of which 100 2475In this benchmark, we use 10000 socket pairs (20000 sockets), of which 100
1789(1%) are active. This mirrors the activity of large servers with many 2476(1%) are active. This mirrors the activity of large servers with many
1790connections, most of which are idle at any one point in time. 2477connections, most of which are idle at any one point in time.
1791 2478
1792Source code for this benchmark is found as F<eg/bench2> in the AnyEvent 2479Source code for this benchmark is found as F<eg/bench2> in the AnyEvent
1793distribution. 2480distribution. It uses the L<AE> interface, which makes a real difference
2481for the EV and Perl backends only.
1794 2482
1795=head3 Explanation of the columns 2483=head3 Explanation of the columns
1796 2484
1797I<sockets> is the number of sockets, and twice the number of "servers" (as 2485I<sockets> is the number of sockets, and twice the number of "servers" (as
1798each server has a read and write socket end). 2486each server has a read and write socket end).
1805it to another server. This includes deleting the old timeout and creating 2493it to another server. This includes deleting the old timeout and creating
1806a new one that moves the timeout into the future. 2494a new one that moves the timeout into the future.
1807 2495
1808=head3 Results 2496=head3 Results
1809 2497
1810 name sockets create request 2498 name sockets create request
1811 EV 20000 69.01 11.16 2499 EV 20000 62.66 7.99
1812 Perl 20000 73.32 35.87 2500 Perl 20000 68.32 32.64
1813 Event 20000 212.62 257.32 2501 IOAsync 20000 174.06 101.15 epoll
1814 Glib 20000 651.16 1896.30 2502 IOAsync 20000 174.67 610.84 poll
2503 Event 20000 202.69 242.91
2504 Glib 20000 557.01 1689.52
1815 POE 20000 349.67 12317.24 uses POE::Loop::Event 2505 POE 20000 341.54 12086.32 uses POE::Loop::Event
1816 2506
1817=head3 Discussion 2507=head3 Discussion
1818 2508
1819This benchmark I<does> measure scalability and overall performance of the 2509This benchmark I<does> measure scalability and overall performance of the
1820particular event loop. 2510particular event loop.
1822EV is again fastest. Since it is using epoll on my system, the setup time 2512EV is again fastest. Since it is using epoll on my system, the setup time
1823is relatively high, though. 2513is relatively high, though.
1824 2514
1825Perl surprisingly comes second. It is much faster than the C-based event 2515Perl surprisingly comes second. It is much faster than the C-based event
1826loops Event and Glib. 2516loops Event and Glib.
2517
2518IO::Async performs very well when using its epoll backend, and still quite
2519good compared to Glib when using its pure perl backend.
1827 2520
1828Event suffers from high setup time as well (look at its code and you will 2521Event suffers from high setup time as well (look at its code and you will
1829understand why). Callback invocation also has a high overhead compared to 2522understand why). Callback invocation also has a high overhead compared to
1830the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event 2523the C<< $_->() for .. >>-style loop that the Perl event loop uses. Event
1831uses select or poll in basically all documented configurations. 2524uses select or poll in basically all documented configurations.
1900 2593
1901Recently I was told about the benchmark in the IO::Lambda manpage, which 2594Recently I was told about the benchmark in the IO::Lambda manpage, which
1902could be misinterpreted to make AnyEvent look bad. In fact, the benchmark 2595could be misinterpreted to make AnyEvent look bad. In fact, the benchmark
1903simply compares IO::Lambda with POE, and IO::Lambda looks better (which 2596simply compares IO::Lambda with POE, and IO::Lambda looks better (which
1904shouldn't come as a surprise to anybody). As such, the benchmark is 2597shouldn't come as a surprise to anybody). As such, the benchmark is
1905fine, and shows that the AnyEvent backend from IO::Lambda isn't very 2598fine, and mostly shows that the AnyEvent backend from IO::Lambda isn't
1906optimal. But how would AnyEvent compare when used without the extra 2599very optimal. But how would AnyEvent compare when used without the extra
1907baggage? To explore this, I wrote the equivalent benchmark for AnyEvent. 2600baggage? To explore this, I wrote the equivalent benchmark for AnyEvent.
1908 2601
1909The benchmark itself creates an echo-server, and then, for 500 times, 2602The benchmark itself creates an echo-server, and then, for 500 times,
1910connects to the echo server, sends a line, waits for the reply, and then 2603connects to the echo server, sends a line, waits for the reply, and then
1911creates the next connection. This is a rather bad benchmark, as it doesn't 2604creates the next connection. This is a rather bad benchmark, as it doesn't
1912test the efficiency of the framework, but it is a benchmark nevertheless. 2605test the efficiency of the framework or much non-blocking I/O, but it is a
2606benchmark nevertheless.
1913 2607
1914 name runtime 2608 name runtime
1915 Lambda/select 0.330 sec 2609 Lambda/select 0.330 sec
1916 + optimized 0.122 sec 2610 + optimized 0.122 sec
1917 Lambda/AnyEvent 0.327 sec 2611 Lambda/AnyEvent 0.327 sec
1923 2617
1924 AnyEvent/select/nb 0.085 sec 2618 AnyEvent/select/nb 0.085 sec
1925 AnyEvent/EV/nb 0.068 sec 2619 AnyEvent/EV/nb 0.068 sec
1926 +state machine 0.134 sec 2620 +state machine 0.134 sec
1927 2621
1928The benchmark is also a bit unfair (my fault) - the IO::Lambda 2622The benchmark is also a bit unfair (my fault): the IO::Lambda/POE
1929benchmarks actually make blocking connects and use 100% blocking I/O, 2623benchmarks actually make blocking connects and use 100% blocking I/O,
1930defeating the purpose of an event-based solution. All of the newly 2624defeating the purpose of an event-based solution. All of the newly
1931written AnyEvent benchmarks use 100% non-blocking connects (using 2625written AnyEvent benchmarks use 100% non-blocking connects (using
1932AnyEvent::Socket::tcp_connect and the asynchronous pure perl DNS 2626AnyEvent::Socket::tcp_connect and the asynchronous pure perl DNS
1933resolver), so AnyEvent is at a disadvantage here as non-blocking connects 2627resolver), so AnyEvent is at a disadvantage here, as non-blocking connects
1934generally require a lot more bookkeeping and event handling than blocking 2628generally require a lot more bookkeeping and event handling than blocking
1935connects (which involve a single syscall only). 2629connects (which involve a single syscall only).
1936 2630
1937The last AnyEvent benchmark additionally uses L<AnyEvent::Handle>, which 2631The last AnyEvent benchmark additionally uses L<AnyEvent::Handle>, which
1938offers similar expressive power as POE and IO::Lambda (using conventional 2632offers similar expressive power as POE and IO::Lambda, using conventional
1939Perl syntax), which means both the echo server and the client are 100% 2633Perl syntax. This means that both the echo server and the client are 100%
1940non-blocking w.r.t. I/O, further placing it at a disadvantage. 2634non-blocking, further placing it at a disadvantage.
1941 2635
1942As you can see, AnyEvent + EV even beats the hand-optimised "raw sockets 2636As you can see, the AnyEvent + EV combination even beats the
1943benchmark", while AnyEvent + its pure perl backend easily beats 2637hand-optimised "raw sockets benchmark", while AnyEvent + its pure perl
1944IO::Lambda and POE. 2638backend easily beats IO::Lambda and POE.
1945 2639
1946And even the 100% non-blocking version written using the high-level (and 2640And even the 100% non-blocking version written using the high-level (and
1947slow :) L<AnyEvent::Handle> abstraction beats both POE and IO::Lambda, 2641slow :) L<AnyEvent::Handle> abstraction beats both POE and IO::Lambda
2642higher level ("unoptimised") abstractions by a large margin, even though
1948even thought it does all of DNS, tcp-connect and socket I/O in a 2643it does all of DNS, tcp-connect and socket I/O in a non-blocking way.
1949non-blocking way. 2644
2645The two AnyEvent benchmarks programs can be found as F<eg/ae0.pl> and
2646F<eg/ae2.pl> in the AnyEvent distribution, the remaining benchmarks are
2647part of the IO::Lambda distribution and were used without any changes.
1950 2648
1951 2649
1952=head1 SIGNALS 2650=head1 SIGNALS
1953 2651
1954AnyEvent currently installs handlers for these signals: 2652AnyEvent currently installs handlers for these signals:
1958=item SIGCHLD 2656=item SIGCHLD
1959 2657
1960A handler for C<SIGCHLD> is installed by AnyEvent's child watcher 2658A handler for C<SIGCHLD> is installed by AnyEvent's child watcher
1961emulation for event loops that do not support them natively. Also, some 2659emulation for event loops that do not support them natively. Also, some
1962event loops install a similar handler. 2660event loops install a similar handler.
2661
2662Additionally, when AnyEvent is loaded and SIGCHLD is set to IGNORE, then
2663AnyEvent will reset it to default, to avoid losing child exit statuses.
1963 2664
1964=item SIGPIPE 2665=item SIGPIPE
1965 2666
1966A no-op handler is installed for C<SIGPIPE> when C<$SIG{PIPE}> is C<undef> 2667A no-op handler is installed for C<SIGPIPE> when C<$SIG{PIPE}> is C<undef>
1967when AnyEvent gets loaded. 2668when AnyEvent gets loaded.
1979 2680
1980=back 2681=back
1981 2682
1982=cut 2683=cut
1983 2684
2685undef $SIG{CHLD}
2686 if $SIG{CHLD} eq 'IGNORE';
2687
1984$SIG{PIPE} = sub { } 2688$SIG{PIPE} = sub { }
1985 unless defined $SIG{PIPE}; 2689 unless defined $SIG{PIPE};
1986 2690
2691=head1 RECOMMENDED/OPTIONAL MODULES
2692
2693One of AnyEvent's main goals is to be 100% Pure-Perl(tm): only perl (and
2694its built-in modules) are required to use it.
2695
2696That does not mean that AnyEvent won't take advantage of some additional
2697modules if they are installed.
2698
2699This section explains which additional modules will be used, and how they
2700affect AnyEvent's operation.
2701
2702=over 4
2703
2704=item L<Async::Interrupt>
2705
2706This slightly arcane module is used to implement fast signal handling: To
2707my knowledge, there is no way to do completely race-free and quick
2708signal handling in pure perl. To ensure that signals still get
2709delivered, AnyEvent will start an interval timer to wake up perl (and
2710catch the signals) with some delay (default is 10 seconds, look for
2711C<$AnyEvent::MAX_SIGNAL_LATENCY>).
2712
2713If this module is available, then it will be used to implement signal
2714catching, which means that signals will not be delayed, and the event loop
2715will not be interrupted regularly, which is more efficient (and good for
2716battery life on laptops).
2717
2718This affects not just the pure-perl event loop, but also other event loops
2719that have no signal handling on their own (e.g. Glib, Tk, Qt).
2720
2721Some event loops (POE, Event, Event::Lib) offer signal watchers natively,
2722and either employ their own workarounds (POE) or use AnyEvent's workaround
2723(using C<$AnyEvent::MAX_SIGNAL_LATENCY>). Installing L<Async::Interrupt>
2724does nothing for those backends.
2725
2726=item L<EV>
2727
2728This module isn't really "optional", as it is simply one of the backend
2729event loops that AnyEvent can use. However, it is simply the best event
2730loop available in terms of features, speed and stability: It supports
2731the AnyEvent API optimally, implements all the watcher types in XS, does
2732automatic timer adjustments even when no monotonic clock is available,
2733can take avdantage of advanced kernel interfaces such as C<epoll> and
2734C<kqueue>, and is the fastest backend I<by far>. You can even embed
2735L<Glib>/L<Gtk2> in it (or vice versa, see L<EV::Glib> and L<Glib::EV>).
2736
2737If you only use backends that rely on another event loop (e.g. C<Tk>),
2738then this module will do nothing for you.
2739
2740=item L<Guard>
2741
2742The guard module, when used, will be used to implement
2743C<AnyEvent::Util::guard>. This speeds up guards considerably (and uses a
2744lot less memory), but otherwise doesn't affect guard operation much. It is
2745purely used for performance.
2746
2747=item L<JSON> and L<JSON::XS>
2748
2749One of these modules is required when you want to read or write JSON data
2750via L<AnyEvent::Handle>. L<JSON> is also written in pure-perl, but can take
2751advantage of the ultra-high-speed L<JSON::XS> module when it is installed.
2752
2753=item L<Net::SSLeay>
2754
2755Implementing TLS/SSL in Perl is certainly interesting, but not very
2756worthwhile: If this module is installed, then L<AnyEvent::Handle> (with
2757the help of L<AnyEvent::TLS>), gains the ability to do TLS/SSL.
2758
2759=item L<Time::HiRes>
2760
2761This module is part of perl since release 5.008. It will be used when the
2762chosen event library does not come with a timing source of its own. The
2763pure-perl event loop (L<AnyEvent::Loop>) will additionally load it to
2764try to use a monotonic clock for timing stability.
2765
2766=back
2767
1987 2768
1988=head1 FORK 2769=head1 FORK
1989 2770
1990Most event libraries are not fork-safe. The ones who are usually are 2771Most event libraries are not fork-safe. The ones who are usually are
1991because they rely on inefficient but fork-safe C<select> or C<poll> 2772because they rely on inefficient but fork-safe C<select> or C<poll> calls
1992calls. Only L<EV> is fully fork-aware. 2773- higher performance APIs such as BSD's kqueue or the dreaded Linux epoll
2774are usually badly thought-out hacks that are incompatible with fork in
2775one way or another. Only L<EV> is fully fork-aware and ensures that you
2776continue event-processing in both parent and child (or both, if you know
2777what you are doing).
2778
2779This means that, in general, you cannot fork and do event processing in
2780the child if the event library was initialised before the fork (which
2781usually happens when the first AnyEvent watcher is created, or the library
2782is loaded).
1993 2783
1994If you have to fork, you must either do so I<before> creating your first 2784If you have to fork, you must either do so I<before> creating your first
1995watcher OR you must not use AnyEvent at all in the child. 2785watcher OR you must not use AnyEvent at all in the child OR you must do
2786something completely out of the scope of AnyEvent.
2787
2788The problem of doing event processing in the parent I<and> the child
2789is much more complicated: even for backends that I<are> fork-aware or
2790fork-safe, their behaviour is not usually what you want: fork clones all
2791watchers, that means all timers, I/O watchers etc. are active in both
2792parent and child, which is almost never what you want. USing C<exec>
2793to start worker children from some kind of manage rprocess is usually
2794preferred, because it is much easier and cleaner, at the expense of having
2795to have another binary.
1996 2796
1997 2797
1998=head1 SECURITY CONSIDERATIONS 2798=head1 SECURITY CONSIDERATIONS
1999 2799
2000AnyEvent can be forced to load any event model via 2800AnyEvent can be forced to load any event model via
2014Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can 2814Similar considerations apply to $ENV{PERL_ANYEVENT_VERBOSE}, as that can
2015be used to probe what backend is used and gain other information (which is 2815be used to probe what backend is used and gain other information (which is
2016probably even less useful to an attacker than PERL_ANYEVENT_MODEL), and 2816probably even less useful to an attacker than PERL_ANYEVENT_MODEL), and
2017$ENV{PERL_ANYEVENT_STRICT}. 2817$ENV{PERL_ANYEVENT_STRICT}.
2018 2818
2819Note that AnyEvent will remove I<all> environment variables starting with
2820C<PERL_ANYEVENT_> from C<%ENV> when it is loaded while taint mode is
2821enabled.
2822
2019 2823
2020=head1 BUGS 2824=head1 BUGS
2021 2825
2022Perl 5.8 has numerous memleaks that sometimes hit this module and are hard 2826Perl 5.8 has numerous memleaks that sometimes hit this module and are hard
2023to work around. If you suffer from memleaks, first upgrade to Perl 5.10 2827to work around. If you suffer from memleaks, first upgrade to Perl 5.10
2026pronounced). 2830pronounced).
2027 2831
2028 2832
2029=head1 SEE ALSO 2833=head1 SEE ALSO
2030 2834
2031Utility functions: L<AnyEvent::Util>. 2835Tutorial/Introduction: L<AnyEvent::Intro>.
2032 2836
2033Event modules: L<EV>, L<EV::Glib>, L<Glib::EV>, L<Event>, L<Glib::Event>, 2837FAQ: L<AnyEvent::FAQ>.
2034L<Glib>, L<Tk>, L<Event::Lib>, L<Qt>, L<POE>. 2838
2839Utility functions: L<AnyEvent::Util> (misc. grab-bag), L<AnyEvent::Log>
2840(simply logging).
2841
2842Development/Debugging: L<AnyEvent::Strict> (stricter checking),
2843L<AnyEvent::Debug> (interactive shell, watcher tracing).
2844
2845Supported event modules: L<AnyEvent::Loop>, L<EV>, L<EV::Glib>,
2846L<Glib::EV>, L<Event>, L<Glib::Event>, L<Glib>, L<Tk>, L<Event::Lib>,
2847L<Qt>, L<POE>, L<FLTK>.
2035 2848
2036Implementations: L<AnyEvent::Impl::EV>, L<AnyEvent::Impl::Event>, 2849Implementations: L<AnyEvent::Impl::EV>, L<AnyEvent::Impl::Event>,
2037L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>, 2850L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>, L<AnyEvent::Impl::Perl>,
2038L<AnyEvent::Impl::EventLib>, L<AnyEvent::Impl::Qt>, 2851L<AnyEvent::Impl::EventLib>, L<AnyEvent::Impl::Qt>,
2852L<AnyEvent::Impl::POE>, L<AnyEvent::Impl::IOAsync>, L<Anyevent::Impl::Irssi>,
2039L<AnyEvent::Impl::POE>. 2853L<AnyEvent::Impl::FLTK>.
2040 2854
2041Non-blocking file handles, sockets, TCP clients and 2855Non-blocking handles, pipes, stream sockets, TCP clients and
2042servers: L<AnyEvent::Handle>, L<AnyEvent::Socket>. 2856servers: L<AnyEvent::Handle>, L<AnyEvent::Socket>, L<AnyEvent::TLS>.
2043 2857
2044Asynchronous DNS: L<AnyEvent::DNS>. 2858Asynchronous DNS: L<AnyEvent::DNS>.
2045 2859
2046Coroutine support: L<Coro>, L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>, 2860Thread support: L<Coro>, L<Coro::AnyEvent>, L<Coro::EV>, L<Coro::Event>.
2047 2861
2048Nontrivial usage examples: L<Net::FCP>, L<Net::XMPP2>, L<AnyEvent::DNS>. 2862Nontrivial usage examples: L<AnyEvent::GPSD>, L<AnyEvent::IRC>,
2863L<AnyEvent::HTTP>.
2049 2864
2050 2865
2051=head1 AUTHOR 2866=head1 AUTHOR
2052 2867
2053 Marc Lehmann <schmorp@schmorp.de> 2868 Marc Lehmann <schmorp@schmorp.de>

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines